Edoxaban vs. Warfarin in East Asian Patients With Atrial Fibrillation　– An ENGAGE AF-TIMI 48 Subanalysis –

Abstract

Background: In the multinational, double-blind, double-dummy ENGAGE AF-TIMI 48 phase 3 study, once-daily edoxaban was non-inferior to warfarin for prevention of stroke or systemic embolism event (SEE) in patients with non-valvular atrial fibrillation (AF). Here, we evaluated the efficacy and safety of edoxaban in patients from East Asia.

Methods and Results: Patients aged ≥21 years with documented AF and CHADS score ≥2 were randomized to receive once-daily edoxaban higher-dose (60 mg) or lower-dose (30 mg) regimen or warfarin dose-adjusted to an international normalized ratio of 2.0–3.0. Patients with a creatinine clearance of 30–50 ml/min, weighing ≤60 kg, or receiving strong p-glycoprotein inhibitors at randomization or during the study received a 50% dose reduction of edoxaban or matched placebo. This prespecified subanalysis included 1,943 patients from Japan, China, Taiwan, and South Korea. The annualized rate of stroke/SEE for higher-dose edoxaban was 1.35% vs. 2.62% for warfarin (hazard ratio [HR], 0.53; 95% confidence interval [CI]: 0.31–0.90, P=0.02) and 2.52% for lower-dose edoxaban (HR, 0.98; 95% CI: 0.63–1.54, P=0.93). Compared with warfarin (4.80%), major bleeding was significantly reduced for the higher-dose (2.86%; HR, 0.61; 95% CI: 0.41–0.89, P=0.011) and lower-dose regimens (1.59%; HR, 0.34; 95% CI: 0.21–0.54, P<0.001).

Conclusions: Once-daily edoxaban provided similar efficacy to warfarin while reducing major bleeding risk in the East Asian population.

Atrial fibrillation (AF) is the most common arrhythmia in adults and occurs in approximately 1% of the US population <60 years of age. Among those ≥80 years of age, this rate increases to >8%.¹ In Japan, the prevalence of AF is approximately 1.6% among those ≥60 years of age, increasing to 3.2% among those ≥80 years of age.² Overall, the estimated number of people in Japan with AF is 720,000, with a prevalence of 0.56%, which is approximately two-thirds of that in the USA.² In general, the prevalence of AF in East Asian countries is lower than that in the USA, but has been rising in recent years.^2–5 Community-based studies in East Asia have reported that the prevalence of stroke among patients with AF ranges from 13% to 15%.⁵ Asian patients are at a greater risk for intracranial hemorrhage. In a retrospective study of patients with ischemic stroke or intracranial hemorrhage, East Asian patients had a higher prevalence of intracranial hemorrhage compared with ischemic stroke (30% and 70%, respectively) in comparison with white patients (15% and 85%, respectively).⁶

Editorial p ????

Until recently, warfarin was the standard of care for stroke prevention in AF^7,8 and is recommended in patients with ≥2 risk factors for stroke to achieve a target international normalized ratio (INR) of 2.0–3.0 in Western countries, although the target INR is lower in elderly patients in some Asian countries, such as Japan, where the target INR is 1.5 to approximately 2.6 in AF patients ≥70 years of age.^7,9 Warfarin is associated with bleeding risk, and requires frequent monitoring and dose adjustments, and may not be suitable for all patients. Direct oral anticoagulants, which include the direct thrombin inhibitor dabigatran and the direct factor Xa inhibitors apixaban, edoxaban, and rivaroxaban, developed to address concerns associated with warfarin, are now also recommended as first-line anticoagulant therapy in Japanese patients with non-valvular AF with ≥2 risk factors for stroke.⁷

Edoxaban, a direct, oral factor Xa inhibitor with predictable pharmacokinetics,^10,11 is rapidly absorbed and achieves peak blood levels in 1–2 h. Approximately 50% of the edoxaban absorbed dose is excreted by the kidneys.¹² The Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation-Thrombolysis in Myocardial Infarction 48 (ENGAGE AF-TIMI 48; clinicaltrials.gov NCT no. NCT00781391) study was a multinational, double-blind, double-dummy trial comparing 2 once-daily dosing regimens of edoxaban with warfarin in >21,000 patients with non-valvular AF at moderate-to-high risk for stroke.¹³ In ENGAGE AF-TIMI 48, both edoxaban dose regimens were non-inferior to warfarin with respect to the prevention of stroke or systemic embolism and were associated with significantly lower rates of bleeding. Here we present the efficacy and safety results from a prespecified subanalysis of East Asian patients in ENGAGE AF-TIMI 48.

Methods

Study Design

This was a multinational, multi-center, randomized, double-blind, double-dummy trial comparing 2 dose regimens of edoxaban with warfarin.^13,14 This was an event-driven study, with patients enrolled from 19 November 2008 to 22 November 2010. The protocol was designed and led by an executive committee in coordination with an international steering committee; all protocols and amendments were approved by the ethics committee at each institution and conform to the ethics guidelines of the 1975 Declaration of Helsinki. All participating patients provided written informed consent.¹³

Patients

Study methods have been described previously.^13,14 Briefly, patients were eligible if they were >21 years of age, had documented AF within the prior 12 months, and a CHADS₂ risk score >2. Patients with a creatinine clearance (CrCl) <30 ml/min, increased risk of bleeding, need for dual antiplatelet therapy, or acute coronary syndrome or stroke in the prior 30 days were not eligible.

This subanalysis was prespecified to include patients from Japan, China, Taiwan, and South Korea. An additional posthoc analysis included patients from the Philippines and Thailand. All remaining regions were classified as non-East Asian. Patients were randomized to receive either warfarin dose-adjusted to achieve an INR 2.0–3.0 or edoxaban higher-dose regimen (60 mg) or lower-dose regimen (30 mg) once daily in a double-blind, double-dummy fashion. Patients with CrCl 30–50 ml/min, body weight <60 kg, or who were concomitantly treated with a strong p-glycoprotein inhibitor at randomization or during the study received a 50% dose reduction of edoxaban.

Outcomes

The primary efficacy outcome was time to the first stroke (ischemic or hemorrhagic) or systemic embolic event (SEE). Secondary efficacy outcomes included the composite of stroke, SEE, or death from cardiovascular causes (including bleeding); major cardiovascular events; and the composite of stroke, SEE, or death from any cause. Additional efficacy endpoints were death from any cause, death from cardiovascular causes, and myocardial infarction (MI).

The primary safety outcome was major bleeding as defined by the International Society on Thrombosis and Haemostasis.¹⁵ Secondary safety outcomes included life-threatening bleeding, clinically relevant non major (CRNM) bleeding, minor bleeding, the composite of major and CRNM bleeding, and any overt bleeding. Prespecified net clinical endpoints included the composite of stroke, SEE, major bleeding, or death and the composite of disabling stroke, life-threatening bleeding, or death.

Adjudication of all deaths and suspected cerebrovascular, SEE, MI, bleeding, and hepatic events was performed by an independent clinical endpoint committee blinded to study drug. All adverse events (AE), reported by the subject or observed by the investigator, occurring after consent was given and for at least 30 days after the final dose of study drug, were recorded.

Surveillance and Follow-up

After randomization, study visits occurred at 8, 15, 29, and 60 days, and at 3 months. Patients were then evaluated at least every 3 months thereafter. At the end of the trial, patients were transitioned to open-label oral anticoagulation therapy per study protocol.¹³

Statistical Analysis

The primary efficacy analysis was performed using a Cox proportional-hazards model that included treatment groups and the 2 randomization stratification factors, CHADS₂ score and by factors requiring dose reduction of edoxaban. The primary efficacy analysis was conducted in the modified intention-to-treat (mITT) population. The mITT analysis included patients who underwent randomization and received ≥1 dose of the study drug. Events that occurred during the on-treatment period, which was defined as the period between administration of the first dose of the study drug and either 3 days after the receipt of the last dose or the end of double-blind therapy (whichever came first), were counted. Events that occurred during study-drug interruptions or discontinuations that lasted >3 days were not included in the mITT analysis. The analyses of secondary efficacy outcomes and net clinical endpoints were conducted in the intention-to-treat (ITT) population during the overall time period, defined as the period from randomization to the end of the treatment period. Thus, all events, whether on or off study drug, were counted in the ITT analyses. A two-sided 95% confidence interval (CI) for the hazard ratio (HR) was estimated using the proportional hazards model that included treatment groups and the 2 randomization stratification factors for the primary efficacy analysis. P-values were calculated using the log-rank test. The primary safety analysis was conducted for all randomized subjects who received ≥1 dose of randomized study drug (safety analysis set). Analyses were based on randomized treatment, even if the dosage of study drug was adjusted after randomization.

Results

Patients

Of the 21,105 total patients randomized in the ENGAGE AF-TIMI 48 trial, 1,943 patients were from East Asia (Figure 1). Of these patients, 1,010 were from Japan, 469 were from China, 234 were from Taiwan, and 230 were from South Korea. A total of 644 patients were randomized to warfarin, 646 to the edoxaban higher-dose regimen, and 653 to the edoxaban lower-dose regimen, with 592, 609, and 612 patients completing the end of study visit from the warfarin, edoxaban higher-dose, and edoxaban lower-dose regimens, respectively.

Figure 1.

Patient disposition. mITT, modified intention to treat.

Demographic and baseline clinical characteristics were well balanced between the warfarin and edoxaban groups (Table 1). Compared with patients outside of East Asia, the East Asian population was twice as likely to require dose reduction at randomization, had a lower number of females, and weighed nearly 20 kg less on average.

Table 1. Demographic and Clinical Patient Characteristics

	East Asian (n=1,943)	Non-East Asian (n=19,162)	P-value†
Age (years)	70.1±8.7	70.7±9.5	0.0085
Female	545 (28.0)	7,495 (39.1)	<0.0001
Weight (kg)	67.0±12.6	85.6±20.0	<0.0001
Dose reduction at randomization	912 (46.9)	4,444 (23.2)	<0.0001
CrCl ≤50 ml/min	583 (30.0)	3,491 (18.2)	<0.0001
Weight ≤60 kg	594 (30.6)	1,489 (7.8)	<0.0001
Use of verapamil or quinidine	128 (6.6)	633 (3.3)	<0.0001
CHADS2 score	2.9±1.0	2.8±1.0	0.17
≤3	1,487 (76.5)	14,850 (77.5)	0.33
4–6	456 (23.5)	4,312 (22.5)
Congestive heart failure	920 (47.3)	11,204 (58.5)	<0.0001
Hypertension	1,596 (82.1)	18,158 (94.8)	<0.0001
Age ≥75	729 (37.5)	7,745 (40.4)	0.013
Diabetes mellitus	680 (35.0)	6,944 (36.2)	0.28
Prior stroke or TIA	824 (42.4)	5,149 (26.9)	<0.0001
Previous use of VKA for ≥60 days	1,153 (59.3)	11,288 (58.9)	0.71
Paroxysmal atrial fibrillation	373 (19.2)	4,993 (26.1)	<0.0001
Medication at time of randomization
Aspirin	543 (27.9)	5,637 (29.4)	0.17
Thienopyridine	60 (3.1)	427 (2.2)	0.016
Amiodarone	85 (4.4)	2,407 (12.6)	<0.0001
Digoxin or digitalis preparation	576 (29.6)	5,751 (30.0)	0.74

Data given as mean±SD or n (%). ^†Quantitative data, t-test; qualitative data, χ² test. CHADS₂, congestive heart failure, hypertension, age, diabetes, prior stroke or TIA; CrCl, creatinine clearance; SD, standard deviation; TIA, transient ischemic attack; VKA, vitamin K antagonist.

For patients receiving warfarin therapy, the median time in therapeutic range was not significantly different (P≥0.05) between East Asian (INR 2.0–3.0, 67.1%; INR 1.8–3.2, 82.6%) and non-East Asian patients (INR 2.0–3.0, 68.6%; INR 1.8–3.2, 83.1%).

Efficacy

During the treatment period, stroke or SEE occurred at a rate of 2.62% per year in the East Asian mITT population for warfarin as compared with 1.34% per year for patients in the edoxaban higher-dose regimen (HR vs. warfarin, 0.53; 95% CI: 0.31–0.90; P=0.02; interaction for higher-dose regimen, P=0.64) and 2.52% per year for the edoxaban lower-dose regimen (HR vs. warfarin, 0.98; 95% CI: 0.63–1.54; P=0.93; interaction for lower-dose regimen, P=0.11; Figure 2; Table 2). Consistent results were observed in the non-East Asian population, with rates of stroke or SEE of 1.16% per year for the higher-dose regimen (HR, 0.84; 95% CI: 0.68–1.04; P=0.10) and 1.51% per year for the edoxaban lower-dose regimen (HR, 1.09; 95% CI: 0.90–1.33; P=0.38), compared with a rate of 1.38% for warfarin in the non-East Asian population. Results did not differ significantly between the East Asian and non-East Asian populations. There were no gender-based differences in efficacy between the East Asian and non-East Asian populations. Similar results were observed when patients from the Philippines and Thailand were included in the East Asian cohort in a post-hoc analysis (Tables S1,S2). Compared with warfarin, higher-dose edoxaban reduced the rate of stroke or SEE by 45%, while rates were similar in the lower-dose edoxaban group.

Figure 2.

Rate of stroke or systemic embolic event in East Asian patients treated with warfarin or edoxaban. On-treatment period, first dose of study drug and either 3 days after receipt of last dose or end of double-blind therapy. CI, confidence interval; HR, hazard ratio.

Table 2. Efficacy Endpoints

	East Asian					Non-East Asian
	Warfarin	Edoxaban				Warfarin	Edoxaban
	n=644 (ITT) n=641 (mITT)	Higher-dose n=646 (ITT) n=642 (mITT)		Lower-dose n=653 (ITT) n=652 (mITT)		n=6,392 (ITT) n=6,371 (mITT)	Higher-dose n=6,389 (ITT) n=6,370 (mITT)		Lower-dose n=6,381 (ITT) n=6,350 (mITT)
	n (% patients/ year)	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	HR† (95% CI), P-value
Primary endpoint
mITT, on treatment period‡	38 (2.62)	20 (1.34)	0.53 (0.31–0.90), 0.02	38 (2.52)	0.98 (0.63–1.54), 0.93	194 (1.38)	162 (1.16)	0.84 (0.68–1.04), 0.10	215 (1.51)	1.09 (0.90–1.33), 0.38
ITT, overall study period§	47 (2.71)	34 (1.89)	0.70 (0.45–1.09), 0.11	52 (2.94)	1.09 (0.73–1.62), 0.67	290 (1.71)	262 (1.53)	0.90 (0.76–1.06), 0.20	331 (1.95)	1.14 (0.97–1.33), 0.11
Stroke (mITT)	36 (2.48)	18 (1.20)	0.50 (0.28–0.88), 0.02	38 (2.52)	1.04 (0.66–1.64), 0.87	183 (1.30)	156 (1.12)	0.86 (0.69–1.06), 0.16	206 (1.45)	1.11 (0.91–1.35), 0.31
Hemorrhagic	18 (1.23)	7 (0.46)	0.37 (0.16–0.89), 0.03	4 (0.26)	0.22 (0.07–0.64), <0.01	58 (0.41)	33 (0.24)	0.57 (0.38–0.88), 0.01	14 (0.10)	0.24 (0.13–0.43), <0.001
Ischemic	19 (1.31)	12 (0.80)	0.64 (0.31–1.32), 0.23	34 (2.26)	1.77 (1.01–3.10), 0.05	125 (0.89)	123 (0.88)	0.99 (0.77–1.27), 0.94	192 (1.35)	1.51 (1.21–1.90), <0.001
Fatal	4 (0.27)	2 (0.13)	0.48 (0.09–2.63), 0.40	2 (0.13)	0.48 (0.09–2.67), 0.40	39 (0.28)	43 (0.31)	1.11 (0.72–1.71), 0.63	38 (0.26)	0.96 (0.61–1.49), 0.85
SEE (mITT)	2 (0.14)	2 (0.13)	1.09 (0.16–7.51), 0.93	0 (0.00)	–	11 (0.08)	6 (0.04)	0.55 (0.20–1.49), 0.24	11 (0.08)	0.98 (0.43–2.26), 0.96
Key secondary endpoints (ITT, overall study period)
Stroke, SEE, or death from CV causes	70 (4.02)	44 (2.44)	0.61 (0.42–0.89), 0.01	66 (3.72)	0.93 (0.66–1.30), 0.67	761 (4.47)	684 (4.00)	0.89 (0.80–0.99), 0.03	730 (4.28)	0.95 (0.86–1.06), 0.37
MACE	76 (4.38)	48 (2.67)	0.61 (0.43–0.88), 0.01	71 (4.02)	0.92 (0.67–1.27), 0.62	850 (5.04)	779 (4.59)	0.91 (0.82–1.00), 0.05	842 (4.99)	0.99 (0.90–1.09), 0.80
Stroke, SEE, or death	88 (5.05)	58 (3.22)	0.64 (0.46–0.90), 0.01	78 (4.40)	0.87 (0.64–1.18), 0.37	958 (5.63)	891 (5.20)	0.92 (0.84–1.01), 0.08	907 (5.32)	0.94 (0.86–1.03), 0.20
Other endpoints (ITT, overall study period)
Death, any cause	50 (2.77)	32 (1.73)	0.63 (0.40–0.98), 0.04	34 (1.84)	0.66 (0.42–1.02), 0.06	789 (4.51)	741 (4.23)	0.93 (0.85–1.03), 0.19	703 (4.00)	0.88 (0.80–0.98), 0.02
Death, CV causes	32 (1.77)	15 (0.81)	0.46 (0.25–0.84), 0.01	20 (1.08)	0.61 (0.35–1.06), 0.08	579 (3.31)	515 (2.94)	0.89 (0.79–1.00), 0.04	507 (2.89)	0.87 (0.77–0.98), 0.02
MI	8 (0.45)	7 (0.38)	0.83 (0.30–2.30), 0.72	6 (0.33)	0.74 (0.26–2.12), 0.57	133 (0.78)	126 (0.73)	0.94 (0.74–1.20), 0.64	163 (0.95)	1.22 (0.97–1.53), 0.09

^†Edoxaban dose vs. warfarin. ^‡For higher-dose edoxaban, interaction between East Asian and non-East Asian, P=0.64; for lower-dose edoxaban, interaction between East Asian and non-East Asian, P=0.11. ^§HR, 95% CI derived from a Cox model that included only patients who received the randomly assigned treatment; P-value calculated using log-rank test. CI, confidence interval; CV, cardiovascular; HR, hazard ratio; ITT, intention-to-treat; MACE, major adverse cardiac event; MI, myocardial infarction; mITT, modified ITT; SEE, systemic embolic event.

In the East Asian population, the overall rate of stroke was decreased with the edoxaban higher-dose regimen compared with warfarin (1.20% vs. 2.48%; P=0.02), whereas the stroke rate with the edoxaban lower-dose regimen (2.52%) was similar to warfarin. The decrease in overall stroke rate was primarily due to significant decreases in hemorrhagic stroke, with a rate of 0.46% for the edoxaban higher-dose regimen (P=0.03) and 0.26% for the lower-dose regimen (P=0.01) compared with a rate of 1.23% for warfarin. Patients receiving the edoxaban higher-dose regimen also had a numerically lower but not significantly different rate of ischemic stroke relative to warfarin (0.80% vs. 1.30%). Rate of ischemic stroke was increased in patients receiving the edoxaban lower-dose regimen (2.26% vs. 1.31%; Table 2). As in the East Asian population, hemorrhagic stroke rate in the non-East Asian population was significantly reduced by the edoxaban higher-dose (0.24%, P=0.01) and lower-dose regimens (0.10%, P<0.001) relative to warfarin (0.41%; Table 2). Ischemic stroke was increased in patients receiving the edoxaban lower-dose regimen in the non-East Asian population (1.35% vs. 0.89%).

Among East Asian patients who received a 50% dose reduction, the annualized rate of stroke or SEE in the warfarin group was higher than in the edoxaban higher-dose regimen and similar to that in the lower-dose regimen (Table 3). Annualized rates of stroke or SEE in the non-East Asian patients who received a 50% dose reduction were similar across the treatment regimens (interaction for higher-dose regimen, P=0.93; interaction for lower-dose regimen, P=0.17).

Table 3. Efficacy in Patients Who Received a Dose Reduction

	East Asian					Non-East Asian
	Warfarin (dose reduced)	Edoxaban (dose reduced)				Warfarin (dose reduced)	Edoxaban (dose reduced)
	n=301 (ITT) n=300 (mITT)	Higher-dose n=296 (ITT) n=294 (mITT)		Lower-dose n=315 (ITT) n=315 (mITT)		n=1,486 (ITT) n=1,480 (mITT)	Higher-dose n=1,488 (ITT) n=1,482 (mITT)		Lower-dose n=1,470 (ITT) n=1,459 (mITT)
	n (% patients/ year)	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	HR† (95% CI), P-value
Primary endpoint (stroke or SEE)
mITT, on treatment period‡	19 (2.92)	9 (1.41)	0.50 (0.23–1.11), 0.09	22 (3.16)	1.10 (0.60–2.04), 0.75	58 (2.05)	53 (1.87)	0.91 (0.63–1.32), 0.62	63 (2.17)	1.06 (0.74–1.51), 0.76
ITT, overall study period§	25 (3.11)	17 (2.09)	0.68 (0.37–1.27), 0.23	28 (3.31)	1.07 (0.62–1.83), 0.81	95 (2.59)	89 (2.37)	0.92 (0.69–1.22), 0.55	115 (3.10)	1.19 (0.91–1.57), 0.21
Stroke (mITT)	17 (2.60)	7 (1.09)	0.43 (0.18–1.05), 0.06	22 (3.16)	1.24 (0.66–2.34), 0.51	54 (1.91)	51 (1.80)	0.94 (0.64–1.38), 0.76	61 (2.10)	1.10 (0.76–1.59), 0.61
SEE (mITT)	2 (0.31)	2 (0.31)	NC	0 (0.00)	NC	95 (2.59)	89 (2.37)	0.92 (0.69–1.22), 0.55	115 (3.10)	1.19 (0.91–1.57), 0.21

^†Edoxaban dose vs. warfarin. ^‡For higher-dose edoxaban, interaction between East Asian and non-East Asian population, P=0.93; for lower-dose edoxaban, interaction between East Asian and non-East Asian, P=0.17. ^§HR and 95% CI were derived from a Cox model that included only patients who received the randomly assigned treatment; P-value calculated using log-rank test. NC, not calculated. Other abbreviations as in Table 2.

Safety

Among East Asian patients, both regimens of edoxaban led to significant reductions in major bleeding compared with warfarin. The annualized rate of major bleeding in the warfarin group was 4.80% compared with 2.86% in the edoxaban higher-dose regimen (HR vs. warfarin, 0.61; 95% CI: 0.41–0.89; P=0.011; P=0.12 for interaction) and 1.59% in the edoxaban lower-dose regimen (HR vs. warfarin 0.34; 95% CI: 0.21–0.54; P<0.001; P=0.12 for interaction; Figure 3). Similar to East Asian patients, the rate of major bleeding for edoxaban was significantly reduced in non-East Asian patients, with a rate of 2.74% for the edoxaban higher-dose regimen (HR, 0.83; 95% CI: 0.73–0.96; P=0.009) and 1.62% for the edoxaban lower dose regimen (HR, 0.49; 95% CI: 0.42–0.58; P<0.001), vs. 3.29% for warfarin (Table 4). Similar results were observed when patients from the Philippines and Thailand were included in the East Asian cohort in a post-hoc analysis (Tables S1,S2).

Figure 3.

Rate of major bleeding in East Asian patients treated with warfarin or edoxaban. CI, confidence interval; HR, hazard ratio.

Table 4. Safety and Net Clinical Endpoints

Endpoint	East Asian					Non-East Asian
	Warfarin (n=641)	Edoxaban				Warfarin (n=6,371)	Edoxaban
		Higher-dose (n=642)		Lower-dose (n=652)			Higher-dose (n=6,370)		Lower-dose (n=6,350)
	n (% patients/ year)	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	HR† (95% CI), P-value
Major bleeding‡	68 (4.80)	42 (2.86)	0.61 (0.41–0.89), 0.011	24 (1.59)	0.34 (0.21–0.54), <0.001	456 (3.29)	376 (2.74)	0.83 (0.73–0.96), 0.009	230 (1.62)	0.49 (0.42–0.58), <0.001
Fatal	4 (0.27)	1 (0.07)	0.24 (0.03–2.20), 0.209	2 (0.13)	0.49 (0.09–2.67), 0.408	55 (0.39)	31 (0.22)	0.57 (0.37–0.88), 0.012	19 (0.13)	0.34 (0.20–0.57), <0.001
Intracranial bleeding	28 (1.92)	9 (0.60)	0.31 (0.15–0.66), 0.002	7 (0.46)	0.24 (0.11–0.56), <0.001	104 (0.74)	52 (0.37)	0.51 (0.36–0.70), <0.001	34 (0.24)	0.32 (0.22–0.47), <0.001
Fatal intracranial bleeding	4 (0.27)	1 (0.07)	0.24 (0.03–2.19), 0.208	1 (0.07)	0.25 (0.03–2.31), 0.221	38 (0.27)	23 (0.16)	0.61 (0.36–1.03), 0.062	11 (0.08)	0.28 (0.15–0.56), <0.001
Gastrointestinal bleeding	16 (1.11)	15 (1.01)	0.91 (0.45–1.85), 0.802	11 (0.72)	0.67 (0.31–1.45), 0.312	174 (1.24)	217 (1.57)	1.27 (1.04–1.54), 0.021	118 (0.83)	0.67 (0.53–0.84), <0.001
Life-threatening bleeding	28 (1.93)	10 (0.66)	0.34 (0.17–0.71), 0.004	6 (0.39)	0.21 (0.09–0.50), <0.001	94 (0.67)	52 (0.37)	0.56 (0.40–0.78), <0.001	34 (0.24)	0.36 (0.24–0.53), <0.001
CRNM bleeding	203 (17.07)	170 (13.33)	0.79 (0.64–0.97), 0.024	138 (10.10)	0.60 (0.49–0.75), <0.001	1,193 (9.50)	1,044 (8.21)	0.87 (0.80–0.94), <0.001	831 (6.24)	0.66 (0.61–0.72), <0.001
Major or CRNM bleeding	248 (21.38)	202 (16.15)	0.77 (0.64–0.92), 0.005	155 (11.50)	0.55 (0.45–0.67), <0.001	1,513 (12.23)	1,326 (10.60)	0.87 (0.81–0.94), <0.001	1,006 (7.61)	0.63 (0.58–0.68), <0.001
Any overt bleeding	281 (25.49)	231 (19.35)	0.77 (0.65–0.92), 0.004	187 (14.36)	0.58 (0.48–0.69), <0.001	1,833 (15.55)	1,634 (13.64)	0.88 (0.82–0.94), <0.001	1,312 (10.31)	0.67 (0.63–0.72), <0.001
Net clinical outcome
Primary§	145 (8.80)	97 (5.60)	0.64 (0.50–0.83), <0.001	103 (5.96)	0.68 (0.53–0.87), 0.003	1,317 (8.04)	1,226 (7.43)	0.92 (0.85–1.00), 0.039	1,145 (6.87)	0.85 (0.79–0.92), <0.001
Secondary¶	86 (4.94)	46 (2.52)	0.51 (0.36–0.73), <0.001	53 (2.91)	0.59 (0.42–0.83), 0.002	901 (5.26)	837 (4.86)	0.92 (0.84–1.01), 0.083	784 (4.54)	0.86 (0.78–0.95), 0.002

Data are from the safety cohort during the treatment period (from first dose of study drug), with interval censoring of events during study-drug interruptions lasting >3 days, except for net clinical outcomes, which are presented for the overall treatment period (which began at the time of randomization). ^†Edoxaban dose vs. warfarin. ^‡For higher-dose edoxaban, interaction between East Asian and non-East Asian population, P=0.12; for lower-dose edoxaban, interaction between East Asian and non-East Asian, P=0.12. ^§Composite of stroke, systemic embolic event, major bleeding, or death from any cause. ^¶Composite of disabling stroke, life-threatening bleeding, or death from any cause. CRNM, clinically relevant non-major bleeding. Other abbreviations as in Table 2.

East Asian patients receiving edoxaban also had a significantly reduced rate of intracranial bleeding: 0.60% for the edoxaban higher-dose regimen (HR vs. warfarin, 0.31; 95% CI: 0.15–0.66; P=0.002) and 0.46% for the lower-dose regimen (HR vs. warfarin, 0.24; 95% CI: 0.11–0.56; P<0.001), vs. 1.92% for warfarin. There were no significant differences between warfarin and either dose regimen of edoxaban in rates of fatal, fatal intracranial, or gastrointestinal bleeding in East Asian patients (Table 4). Both regimens of edoxaban significantly reduced life-threatening bleeding, CRNM bleeding, major or CRNM bleeding, and any overt bleeding in comparison with warfarin. Unlike the East Asian population, both regimens of edoxaban significantly decreased the rates of fatal bleeding, and the edoxaban higher-dose regimen significantly increased gastrointestinal bleeding compared with warfarin in the non-East Asian patients (Table 4).

In East Asian and non-East Asian patients who received a 50% dose reduction, both edoxaban dose regimens significantly reduced the annualized rates of major bleeding compared with warfarin (P=0.99 for higher-dose regimen interaction; P=0.09 for lower-dose regimen interaction; Table 5).

Table 5. Safety Endpoints in Patients Receiving a Dose Reduction

Endpoint	East Asian					Non-East Asian
	Warfarin (dose reduced) (n=300)	Edoxaban (dose reduced)				Warfarin (dose reduced) (n=1,480)	Edoxaban (dose reduced)
		Higher-dose (n=294)		Lower-dose (n=315)			Higher-dose (n=1,482)		Lower-dose (n=1,459)
	n (% patients/ year)	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	n (% patients/ year)	HR† (95% CI), P-value	n (% patients/ year)	HR† (95% CI), P-value
Major bleeding‡	33 (5.18)	12 (1.91)	0.38 (0.20–0.74), 0.004	11 (1.58)	0.31 (0.16–0.61), <0.001	133 (4.78)	92 (3.30)	0.69 (0.53–0.90), 0.007	43 (1.48)	0.31 (0.22–0.44), <0.001
Life-threatening bleeding	11 (1.68)	3 (0.47)	NC	1 (0.14)	NC	22 (0.78)	12 (0.42)	0.55 (0.27–1.10), 0.091	6 (0.21)	0.27 (0.11–0.66), 0.004
CRNM bleeding	100 (19.67)	76 (13.89)	0.74 (0.55–1.00), 0.046	62 (9.75)	0.52 (0.38–0.72), <0.001	322 (12.94)	216 (8.33)	0.65 (0.55–0.78), <0.001	186 (6.87)	0.54 (0.45–0.65), <0.001
Major or CRNM bleeding	122 (23.84)	87 (16.23)	0.69 (0.53–0.91), 0.009	70 (11.24)	0.49 (0.36–0.65), <0.001	415 (16.94)	291 (11.44)	0.68 (0.59–0.80), <0.001	219 (8.14)	0.49 (0.42–0.58), <0.001
Any overt bleeding	139 (28.73)	99 (19.29)	0.69 (0.53– 0.89), 0.004	88 (14.66)	0.53 (0.40–0.69), <0.001	486 (20.79)	360 (14.75)	0.72 (0.63–0.83), <0.001	293 (11.33)	0.56 (0.49–0.65), <0.001

Data are from the safety cohort during the treatment period (from the first dose of study drug), with interval censoring of events during study-drug interruptions lasting >3 days. ^†Edoxaban dose vs. warfarin. ^‡For higher-dose edoxaban, interaction between East Asian and non-East Asian population P=0.99; for lower-dose edoxaban, interaction between East Asian and non-East Asian P=0.09. Abbreviations as in Tables 2,4.

The annualized rate of the primary net clinical outcome (death from any cause, stroke, SEE, or major bleeding) was significantly lower for both regimens of edoxaban as compared with warfarin in the East Asian population. The annualized rate was 8.80% for warfarin and 5.60% for the edoxaban higher-dose regimen (HR vs. warfarin, 0.64; 95% CI: 0.50–0.83; P<0.001), and 5.96% for the edoxaban lower-dose regimen (HR vs. warfarin, 0.68; 95% CI: 0.53–0.87; P=0.003). Compared with warfarin, both edoxaban regimens were associated with a significantly lower rate of the secondary net clinical outcome. For the non-East Asian population, rates for both the primary and secondary net clinical outcomes were significantly lower for both regimens of edoxaban compared with warfarin, except for the secondary net clinical outcome for the higher-dose edoxaban regimen (Table 4).

AE

Analyses of adverse effects by treatment group showed that there were no differences between the East Asian and non-East Asian populations in AE, serious AE (SAE) including fatal AE, or drug-related SAE either in the overall or on-treatment period (Table S3). There were no differences in AE leading to drug interruptions during the on-treatment period.

Discussion

In East Asian patients, both the edoxaban higher-dose and lower-dose regimens were at least as effective as warfarin in preventing stroke and SEE, with significantly lower rates observed with higher-dose edoxaban as compared with warfarin for the on-treatment analysis. Hemorrhagic stroke rates were reduced for both edoxaban dosage regimens compared with warfarin while the lower-dose edoxaban regimen was not as effective in preventing ischemic stroke. Rates of major bleeding and intracranial bleeding were significantly reduced for both regimens of edoxaban. Overall, results in the East Asian population were consistent with those in the non-East Asian population, despite some significant differences in population demographics. Of note, a significantly greater proportion of East Asian patients than non-East Asian patients qualified for a 50% dose reduction. There was a greater reduction in bleeding with no apparent differences in the rate of stroke and SEE between the dose-reduced and non-reduced patients in both the East Asian and non-East Asian populations.

Rate of hemorrhagic stroke was higher in each treatment group of East Asian patients compared with that observed in the non-East Asian patients. Higher-dose edoxaban, however, may provide a greater reduction for East Asian patients than for non-East Asian patients in the risk for hemorrhagic stroke compared with warfarin; compared with warfarin, higher-dose edoxaban reduced the risk for hemorrhagic stroke by 63% in East Asian patients and by 43% in non-East Asian patients. The efficacy of low-dose edoxaban vs. warfarin for the prevention of hemorrhagic stroke was similar between the 2 patient populations. Similarly, while rates for intracranial bleeding were higher across the treatment groups for East Asian patients compared with non-East Asian patients, higher-dose and lower-dose edoxaban reduced the risk for intracranial bleeding in East Asian patients by 69% and 76%, respectively, vs. warfarin and by 49% and 68%, respectively, in non-East Asian patients.

A limited number of studies have addressed the safety and efficacy of non-vitamin K antagonist oral anticoagulants in East Asian patients. The efficacy of dabigatran 110 mg and 150 mg twice daily and rivaroxaban 15 mg or 20 mg once daily have been shown to be at least similar to warfarin in Asian patients of varying composition and non-East Asian patients with non-valvular AF.^16–18 In the Randomized Evaluation of Long-Term Anticoagulation Therapy (RE-LY) trial, a trend toward greater bleeding in a broad population of Asian patients (China, Hong Kong, Japan, South Korea, Taiwan, India, Malaysia, Philippines, Singapore, and Thailand) vs. non-Asian patients was noted, with statistically greater rates of total bleeding and hemorrhagic stroke following warfarin treatment compared with dabigatran.¹⁷ In 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), East Asian patients (China, Korea, Taiwan, Hong Kong) also had higher rates of major or CRNM bleeding, regardless of treatment, compared with patients from other geographical regions.¹⁸ In the J-ROCKET AF trial,¹⁶ which specifically evaluated in Japanese patients with non-valvular AF the safety and efficacy of rivaroxaban 15 mg once daily compared with dose-adjusted warfarin, rivaroxaban was non-inferior to warfarin for the prevention of stroke and SEE and for rates of major or CRNM bleeding, and these results were consistent with those observed in the global ROCKET AF trial.¹⁹ Consistent with these findings, numerically higher numbers of life-threatening bleeding, CRNM bleeding, major or CRNM bleeding, and overt bleeding in the East Asian population were also noted in the current analysis, although it was not designed for direct comparisons between the East Asian and non-East Asian populations. These data suggest a greater propensity for bleeding in Asian patients. In fact, Asian patients receiving vitamin K antagonist therapy had a higher rate of intracranial hemorrhage compared with other ethnic groups.²⁰ The risk of bleeding is exacerbated in elderly patients, and maintenance of an INR 1.6–2.6 in Japanese patients has been proposed to reduce the risk of hemorrhagic events, although it may be less effective in reducing stroke occurrence.²¹ Data from the J-RHYTHM Registry, a prospective, observational, nationwide study, demonstrated that maintaining warfarin at an INR 1.6–2.6 provided similar efficacy and greater reductions in major bleeding rates compared with other INR ranges for Japanese patients ≥70 years of age and this benefit is evident even in those ≥85 years of age.^22,23 Intracerebral or subarachnoid hemorrhages comprise 15%–20% of all strokes in Caucasian patients, and the percentage may be 1.5–2-fold higher in Asian patients.¹⁸ Thus, the significant reduction in both the rates of intracranial hemorrhage and hemorrhagic stroke with either regimen of edoxaban may be of particular importance in this patient population.

Interestingly, despite a trend toward increased rates of bleeding in the East Asian population, the rates of gastrointestinal bleeding with edoxaban were similar to warfarin, but greater with the edoxaban higher-dose regimen (1.57%) compared with warfarin (1.24%) among non-East Asian patients. Similarly, in a subanalysis of RE-LY, Asian patients had numerically lower rates of major gastrointestinal bleeding compared with non-Asian patients receiving dabigatran 150 mg, and a significant interaction between treatment and region.¹⁶ In a study of rivaroxaban in Japanese patients, major gastrointestinal bleeding rates were also lower in patients receiving rivaroxaban compared with warfarin.²⁰ The reasons for this are not understood.

A limitation of this subanalysis was that the number of patients in the East Asian group was relatively small compared with the non-East Asian group and did not provide sufficient power for statistical comparisons. This study also was not designed to directly compare the East Asian population to the non-East Asian population. Another limitation of this subanalysis was our definition of “East Asian,” which was prespecified in the ENGAGE AF-TIMI 48 statistical analysis plan to include patients from China, South Korea, Taiwan, and Japan. Other countries that may be considered part of this geographical region, such as Thailand and the Philippines, were not included, which could suggest a bias in the definition of “East Asian”. This was addressed, however, in post-hoc analysis. When patients from the Philippines and Thailand who participated in the ENGAGE AF trial were included in the analysis of the primary efficacy and safety endpoints, there was no difference in results compared with the original East Asian cohort, indicating a lack of bias in the definition of the study cohort.

Conclusions

Both regimens of edoxaban were at least as effective as warfarin in preventing stroke and SEE in East Asian patients with non-valvular AF while significantly reducing the rate of major bleeding and intracranial hemorrhage. The results in the East Asian population were consistent with those outside East Asia.

Acknowledgments

This study was sponsored by Daiichi Sankyo Co, Ltd (Tokyo, Japan). Medical writing and editorial support was provided by Terri Schochet, PhD, and Meryl Gersh, PhD, of AlphaBioCom, LLC (King of Prussia, PA, USA), and funded by Daiichi Sankyo.

Conflicts of Interest

T.Y. has received honoraria from Daiichi Sankyo, Boehringer Ingelheim, Bayer Healthcare, Pfizer, Bristol-Myers Squibb, Eisai, Tanabe-Mitsubishi, and Ono Pharmaceutical; and has received research funding from Daiichi Sankyo, Boehringer Ingelheim, and Tanabe-Mitsubishi. Y.K. has received lecture fees from Daiichi Sankyo, Boehringer Ingelheim, Bayer, Bristol-Myers Squibb, and Pfizer, and research funding from Daiichi Sankyo and Boehringer Ingelheim. R.P.G. has served as a consultant and has received honoraria from Bristol-Myers Squibb, Janssen, Portola, Pfizer, Daiichi Sankyo, Merck & Co, and Sanofi-Aventis; and has received research funding through his institution from Daiichi Sankyo, Merck & Co, Johnson & Johnson, Sanofi-Aventis, and AstraZeneca. T.K., K.M., and K.A. are employees of and have stock options in Daiichi Sankyo Co, Ltd, Tokyo, Japan. M.M. is an employee of and has stock options in Daiichi Sankyo Pharma Development, Edison, NJ, USA. C.T.R. has served as a consultant and has received honoraria from Daiichi Sankyo, Boehringer Ingelheim, Bayer, and Portola. N.C. has received consultation fees from Daiichi Sankyo as a steering committee member. Y.J.S. has received consultant fees from Daiichi Sankyo. Y.Y. and S.-A.C. declare no conflicts of interest.

Author Contributions

R.P.G., M.M., and C.T.R. were involved in the concept and design of the study and in the analysis and interpretation of the data. T.Y., Y.K., Y.Y., S.-A.C., N.C., Y.J.S., T.K., K.M., and K.A. were involved in the analysis and interpretation of the data. Each author critically reviewed the manuscript and provided final approval to submit this manuscript for publication.

Supplementary Files

Supplementary File 1

Table S1. Primary endpoint (East Asian vs. East Asian including the Philippines and Thailand)

Table S2. Primary endpoint (East Asian including the Philippines and Thailand vs. Non-East Asian)

Table S3. Non-bleeding AE

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-15-1082

References

• 1.    Snow  V,  Weiss  KB,  LeFevre  M,  McNamara  R,  Bass  E,  Green  LA, et al. Management of newly detected atrial fibrillation: A clinical practice guideline from the American Academy of Family Physicians and the American College of Physicians. Ann Intern Med 2003; 139: 1009–1017.
• 2.    Inoue  H,  Fujiki  A,  Origasa  H,  Ogawa  S,  Okumura  K,  Kubota  I, et al. Prevalence of atrial fibrillation in the general population of Japan: An analysis based on periodic health examination. Int J Cardiol 2009; 137: 102–107.
• 3.    Jeong  JH. Prevalence of and risk factors for atrial fibrillation in Korean adults older than 40 years. J Korean Med Sci 2005; 20: 26–30.
• 4.    Zhou  Z,  Hu  D. An epidemiological study on the prevalence of atrial fibrillation in the Chinese population of mainland China. J Epidemiol 2008; 18: 209–216.
• 5.    Lip  GY,  Brechin  CM,  Lane  DA. The global burden of atrial fibrillation and stroke: A systematic review of the epidemiology of atrial fibrillation in regions outside North America and Europe. Chest 2012; 142: 1489–1498.
• 6.    Khan  NA,  Quan  H,  Hill  MD,  Pilote  L,  McAlister  FA,  Palepu  A, et al. Risk factors, quality of care and prognosis in South Asian, East Asian and White patients with stroke. BMC Neurol 2013; 13: 74.
• 7.   JCS Joint Working Group. Guidelines for pharmacotherapy of atrial fibrillation (JCS 2013): Digest version. Circ J 2014; 78: 1997–2021.
• 8.   Warfarin [package insert]. Tokyo, Japan: Eisai Co, Ltd (in Japanese), 2013.
• 9.    January  CT,  Wann  LS,  Alpert  JS,  Calkins  H,  Cigarroa  JE,  Cleveland  JC Jr, et al. 2014 AHA/ACC/HRS Guideline for the management of patients with atrial fibrillation: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society. J Am Coll Cardiol 2014; 64: 2246–2280.
• 10.    Furugohri  T,  Isobe  K,  Honda  Y,  Kamisato-Matsumoto  C,  Sugiyama  N,  Nagahara  T, et al. DU-176b, a potent and orally active factor Xa inhibitor: In vitro and in vivo pharmacological profiles. J Thromb Haemost 2008; 6: 1542–1549.
• 11.    Ogata  K,  Mendell-Harary  J,  Tachibana  M,  Masumoto  H,  Oguma  T,  Kojima  M, et al. Clinical safety, tolerability, pharmacokinetics, and pharmacodynamics of the novel factor Xa inhibitor edoxaban in healthy volunteers. J Clin Pharmacol 2010; 50: 743–753.
• 12.    Matsushima  N,  Lee  F,  Sato  T,  Weiss  D,  Mendell  J. Bioavailability and safety of the factor Xa inhibitor edoxaban and the effects of quinidine in healthy subjects. Clin Pharm Drug Dev 2013; 2: 358–366.
• 13.    Giugliano  RP,  Ruff  CT,  Braunwald  E,  Murphy  SA,  Wiviott  SD,  Halperin  JL, et al. Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2013; 369: 2093–2104.
• 14.    Ruff  CT,  Giugliano  RP,  Antman  EM,  Crugnale  SE,  Bocanegra  T,  Mercuri  M, et al. Evaluation of the novel factor Xa inhibitor edoxaban compared with warfarin in patients with atrial fibrillation: Design and rationale for the Effective aNticoaGulation with factor xA next GEneration in Atrial Fibrillation-Thrombolysis In Myocardial Infarction study 48 (ENGAGE AF-TIMI 48). Am Heart J 2010; 160: 635–641.
• 15.    Schulman  S,  Kearon  C; Subcommittee on Control of Anticoagulation of the Scientific Standardization Committee of the International Society on Thrombosis and Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost 2005; 3: 692–694.
• 16.    Hori  M,  Matsumoto  M,  Tanahashi  N,  Momomura  S,  Uchiyama  S,  Goto  S, et al. Rivaroxaban vs. warfarin in Japanese patients with atrial fibrillation: The J-ROCKET AF study. Circ J 2012; 76: 2104–2111.
• 17.    Hori  M,  Connolly  SJ,  Zhu  J,  Liu  LS,  Lau  CP,  Pais  P, et al. Dabigatran versus warfarin: Effects on ischemic and hemorrhagic strokes and bleeding in Asians and non-Asians with atrial fibrillation. Stroke 2013; 44: 1891–1896.
• 18.    Wong  KS,  Hu  DY,  Oomman  A,  Tan  RS,  Patel  MR,  Singer  DE, et al. Rivaroxaban for stroke prevention in East Asian patients from the ROCKET AF trial. Stroke 2014; 45: 1739–1747.
• 19.    Patel  MR,  Mahaffey  KW,  Garg  J,  Pan  G,  Singer  DE,  Hacke  W, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med 2011; 365: 883–891.
• 20.    Shen  AY,  Yao  JF,  Brar  SS,  Jorgensen  MB,  Chen  W. Racial/ethnic differences in the risk of intracranial hemorrhage among patients with atrial fibrillation. J Am Coll Cardiol 2007; 50: 309–315.
• 21.    Yasaka  M,  Minematsu  K,  Yamaguchi  T. Optimal intensity of international normalized ratio in warfarin therapy for secondary prevention of stroke in patients with non-valvular atrial fibrillation. Intern Med 2001; 40: 1183–1188.
• 22.    Inoue  H,  Okumura  K,  Atarashi  H,  Yamashita  T,  Origasa  H,  Kumagai  N, et al. Target international normalized ratio values for preventing thromboembolic and hemorrhagic events in Japanese patients with non-valvular atrial fibrillation: Results of the J-RHYTHM Registry. Circ J 2013; 77: 2264–2270.
• 23.    Kodani  E,  Atarashi  H,  Inoue  H,  Okumura  K,  Yamashita  T,  Origasa  H, et al. Use of warfarin in elderly patients with non-valvular atrial fibrillation: Subanalysis of the J-RHYTHM Registry. Circ J 2015; 79: 2345–2352.