Andexanet Alfa for the Reversal of Factor Xa Inhibitor Activity: Prespecified Subgroup Analysis of the ANNEXA-4 Study in Japan

Abstract

Aims: Andexanet alfa, a specific antidote to factor Xa (FXa) inhibitors, has been approved for clinical use in several countries, including Japan, based on the results from the phase 3 trial ANNEXA-4. We aimed to assess the efficacy and safety of andexanet alfa treatment in FXa inhibitor–related acute major bleeding in patients enrolled for ANNEXA-4 in Japan.

Methods: This prespecified analysis included patients enrolled at Japanese sites in the prospective, open-label, single-arm ANNEXA-4 trial. Eligible patients had major bleeding within 18 hours of oral FXa inhibitor administration. The coprimary efficacy endpoints were percent change in anti-FXa activity and proportion of patients achieving excellent or good hemostatic efficacy 12 hours post-treatment.

Results: A total of 19 patients were enrolled, all of whom had intracranial hemorrhage; 16 patients were evaluable for efficacy. Median percent reduction in anti-FXa activity from baseline to nadir was 95.4% in patients taking apixaban, 96.1% in patients taking rivaroxaban, and 82.2% in patients taking edoxaban. Overall, 14/16 patients (88%) achieved excellent or good hemostasis (apixaban, 5/5; rivaroxaban, 6/7; edoxaban, 3/4). Within 30 days, treatment-related adverse events (AEs) and serious AEs occurred in 2 and 5 patients, respectively. One patient died during follow-up, and 2 patients experienced thrombotic events.

Conclusion: Treatment with andexanet alfa rapidly reduced anti-FXa activity with favorable hemostatic efficacy in Japanese patients with acute major bleeding. Serious AEs of thrombotic events during rapid reversal of anti-FXa activity arose as particular safety concerns in this population as with previous studies.

Trial Registration: NCT02329327

Introduction

Factor Xa (FXa) inhibitors are widely used and effective oral anticoagulants for the prevention and treatment of embolic events. Apixaban, rivaroxaban, and edoxaban are available in Japan for the prevention of stroke in patients with nonvalvular atrial fibrillation and for the prevention and treatment of venous thromboembolism¹⁾. The current rate of FXa inhibitor use is estimated at approximately 60% of Japanese patients with nonvalvular atrial fibrillation, a population predicted to reach 1.03 million by 2050 ^{2, 3)}. However, oral anticoagulants may lead to or worsen acute major bleeding, which in turn can be associated with significant complications, including death^4-9). In the Japanese post-marketing surveillance XAPASS study in patients with nonvalvular atrial fibrillation treated with rivaroxaban, the incidence of major bleeding was 1.8/100 person-years¹⁰⁾. The incidence of major bleeding among Japanese clinical trial participants with nonvalvular atrial fibrillation receiving edoxaban and apixaban has been reported at 1.02% and 2.36% per year, respectively^{11, 12)}. A method of rapidly reversing the anticoagulant effect in emergency situations is, therefore, desirable. Andexanet alfa is a modified, recombinant form of FXa that is catalytically inactive but retains the ability to bind FXa inhibitors, thereby neutralizing their anticoagulant effects¹³⁾.

In the international, multicenter, open-label, single-arm ANNEXA-4 (ClinicalTrials.gov Identifier: NCT02329327) study, andexanet alfa was studied in patients treated with FXa inhibitors who presented with acute major bleeding (primarily intracranial and gastrointestinal)^{14, 15)}. In this study, andexanet alfa was associated with median percent reductions in anti-FXa activity of 94% in patients taking rivaroxaban, 93% in patients taking apixaban, 75% in patients taking enoxaparin, and 71% in patients taking edoxaban¹⁵⁾; these results were similar with those of healthy volunteers without major bleeding¹⁶⁾. In the ANNEXA-4 study, most patients achieved excellent or good hemostasis (79% in patients with intracranial hemorrhage [ICH] and 82% in patients with gastrointestinal bleeding). Based on the results from this pivotal study, andexanet alfa was granted accelerated approval in the United States and conditional approval in Europe as a reversal agent for apixaban and rivaroxaban in patients with uncontrolled or life-threatening bleeding. Andexanet alfa has also been approved in Japan for patients treated with apixaban, rivaroxaban, or edoxaban.

It is relevant to evaluate the risk-benefit profile of andexanet alfa, specifically in Japanese patients given certain characteristics of this population, including a higher prevalence of ICH compared to Western countries^{17, 18)}, differences in medical practice (eg, lower officially approved dosage of rivaroxaban¹⁹⁾, relatively frequent use of edoxaban, and relatively infrequent use of enoxaparin), and possible differences in pharmacodynamics of andexanet alfa in Japanese patients relative to other patient populations. Safety and the reversal effect of andexanet alfa on the FXa inhibitor–mediated anticoagulant activity in healthy adults have been confirmed in a phase 2 study including Japanese participants²⁰⁾. Herein, we determined the efficacy and safety of andexanet alfa in a subgroup of ANNEXA-4 patients enrolled at study sites in Japan.

Aims

The aim of this prespecified subanalysis of the ANNEXA-4 study was to assess the efficacy and safety of andexanet alfa treatment in FXa inhibitor–related acute major bleeding in patients enrolled at study sites in Japan.

Methods

Study Design, Participants, and Procedures

ANNEXA-4 was a prospective, open-label, single-arm study of andexanet alfa in patients who had acute major bleeding following administration of a FXa inhibitor. The present prespecified subgroup analysis of ANNEXA-4 included patients enrolled at 10 sites in Japan. Full details of the study design have been previously published^{14, 21)}. Briefly, patients aged ≥ 18 years were required to have an episode of acute major bleeding (as defined by the International Society on Thrombosis and Haemostasis²²⁾) and to have received apixaban, rivaroxaban, or edoxaban or enoxaparin at a dose of ≥ 1 mg/kg/day, within the preceding 18 hours. A protocol amendment was made to exclude Japanese patients taking enoxaparin, as the enoxaparin dose of ≥ 1 mg/kg/day specified in the protocol was not approved in Japan. Ethical approval was obtained from the institutional review boards, and all patients provided informed consent prior to the study.

Andexanet alfa was administered as a bolus over a period of 15 to 30 minutes, followed by a 2-hour infusion. Patients received either a low dose (400 mg bolus＋480 mg infusion) or a high dose (800 mg bolus＋960 mg infusion) depending on the type, amount, and timing of the last dose of FXa inhibitor (Supplemental Table 1).

Supplemental Table 1.Andexanet Alfa Dosing Regimens

FXa inhibitor	FXa inhibitor last dose	Timing of FXa inhibitor last dose before andexanet alfa initiation
		＜8 hours or unknown	≥ 8 hours
Rivaroxaban	≥ 10 mg/unknown	High dose
Apixaban	≤ 5 mg	Low dose
	＞5 mg/unknown	High dose	Low dose
Edoxaban	≥ 30 mg/unknown	High dose
Unknown	Unknown	High dose

FXa, factor Xa.

High dose: 800 mg bolus＋960 mg infusion.

Low dose: 400 mg bolus＋480 mg infusion.

The coprimary efficacy endpoints were percent change in anti-FXa activity and proportion of patients who achieved excellent or good hemostatic efficacy 12 hours after treatment, as assessed by an independent adjudication committee. Excellent hemostasis was defined as a ≤ 20% increase in hematoma volume (for intracerebral hemorrhage) or maximum thickness (for subarachnoid bleeding or subdural hematoma) on a repeat computed tomography (CT) or magnetic resonance imaging (MRI) scan at both 1 and 12 hours after the end of treatment versus baseline, and good hemostasis as a ＞ 20% and ≤ 35% increase in hematoma volume/thickness on a CT or MRI scan at 12 hours compared with baseline. The secondary efficacy endpoints included indicators for thrombin generation and monitoring for antibodies to andexanet alfa, factor X, and FXa. Safety outcomes included death, occurrence of thrombotic events within 30 days, and treatment-emergent adverse events (TEAEs).

Statistical Analysis

The safety population consisted of all patients receiving any dose of andexanet alfa. The efficacy population included all patients with baseline anti-FXa activity ≥ 75 ng/mL for apixaban and rivaroxaban or ≥ 40 ng/mL for edoxaban and adjudicated major bleeding at presentation. Baseline demographic characteristics, ICH characteristics, and outcomes were summarized using descriptive statistics of frequencies for categorical data and mean (standard deviation) and median (range) for continuous data. Due to the small sample size, no statistical tests of significance were performed, and the analysis of all study outcomes is descriptive. Adverse events (AEs) were coded using the Medical Dictionary for Regulatory Activities (MedDRA), version 18.0.

Results

Study Patients

Of the 479 patients included in ANNEXA-4, 19 patients (8 women; median age of 78 years) were enrolled in Japan (Table 1). Six patients were taking apixaban, 8 were taking rivaroxaban, and 5 were taking edoxaban. All 19 patients presented with ICH. All but 1 fatal patient, who was taking edoxaban, completed 30-day follow-up.

Table 1.Baseline Characteristics Stratified by FXa Inhibitor

Parameter	Apixaban (n = 6)	Rivaroxaban (n = 8)	Edoxaban (n = 5)	All patients (n = 19)
Age, years, median (range)	77.5 (58.0, 89.0)	74.0 (63.0, 86.0)	80.0 (71.0, 91.0)	78 (58, 91)
Sex, n (%)
Male	4 (66.7)	6 (75.0)	1 (20.0)	11 (58)
Female	2 (33.3)	2 (25.0)	4 (80.0)	8 (42)
BMI, kg/m2, median (range)	23.1 (17.9, 30.7)	22.9 (20.6, 29.3)	20.9 (18.0, 25.0)	23.0 (17.9, 30.7)
Medical history, n (%)
Atrial fibrillation	6 (100)	8 (100)	4 (80)	18 (95)
Hypertension	6 (100)	8 (100)	5 (100)	19 (100)
Stroke	4 (67)	3 (38)	1 (20)	8 (42)
Deep vein thrombosis	0	0	2 (40)	2 (11)
Pulmonary embolism	0	0	0	0
Congestive heart failure	0	1 (13)	0	1 (5)
Myocardial infarction	0	1 (13)	0	1 (5)
Transient ischemic attack	0	0	0	0
Severe peripheral vascular disease	0	0	0	0
Hepatic laboratory data, mean (SD)
Albumin, g/L	38.3 (3.88)	40.4 (2.20)	40.2 (4.97)	39.7 (3.5)
Alkaline phosphatase, U/L	243.3 (62.12)	221.5 (94.00)	272.8 (81.15)	241.9 (80.1)
Alanine aminotransferase, U/L	13.7 (5.61)	32.9 (21.32)	13.4 (4.39)	21.7 (16.9)
Aspartate aminotransferase, U/L	22.3 (4.84)	29.3 (11.40)	24.0 (15.28)	25.7 (10.9)
Bilirubin, μmol/L	12.8 (3.56)	16.0 (5.62)	13.7 (4.37)	14.4 (4.7)
eGFR, mL/min, median (range)	50.1 (35.8, 128.1)	80.3 (38.3, 90.9)	39.0 (19.7, 83.4)	66.5 (19.7, 128.1)
Prothrombin time, seconds	15.3 (2.50)	22.4 (4.79)	14.4 (1.82)	17.8 (5.0)
Intracranial compartment site of bleeding, n (%)
Intracerebral§	6 (100)	5 (63)	3 (60)	14 (74)
Subarachnoid	0	2 (25)	0	2 (11)
Subdural	0	1 (13)	2 (40)	3 (16)
Baseline hematoma volume/thickness
Intracerebral hemorrhage, median (range), cc	8.3 (1.8, 33.5)	2.9 (0.5, 22.4)	4.4 (2.6, 26.2)	4.2 (0.5, 33.5)
Subdural hemorrhage, median (range), cc	0	24.4 (0.6, 48.1)	59.8 (25.3, 94.3)	36.7 (0.6, 94.3)
Andexanet alfa dose, n (%)
Low dose (400 mg bolus＋480 mg IV)	6 (100)	5 (63)	3 (60)	14 (74)
High dose (800 mg bolus＋960 mg IV)	0	3 (38)	2 (40)	5 (26)
Time to start of andexanet alfa treatment, hours, median (range)
From last dose of FXa inhibitors	9.9 (4.7, 17.5)	9.0 (3.8, 10.9)†	10.4 (3.9, 16.5)†	9.85 (3.78, 17.47)
From baseline CT scan	1.6 (1.3, 3.0)†	1.7 (0.6, 1.9)†	1.5 (0.5, 3.0)†	1.5 (0.5, 3.00)

BMI, body mass index; FXa, factor Xa; eGFR, estimated glomerular filtration rate; SD, standard deviation; IV, intravenous; CT, computed tomography.

^§Thalamic hemorrhage (n = 6), putaminal hemorrhage (n = 4), pontine hemorrhage (n = 1), parietal cortical hemorrhage (n = 1), cerebellar hemorrhage (n = 1), and not specified (n = 1).

^†One patient is missing.

Efficacy

Efficacy assessments were performed in 16 patients in whom the baseline anti-FXa level exceeded the efficacy threshold, 5 of these were taking apixaban, 7 were taking rivaroxaban, and 4 were taking edoxaban. Anti-FXa activity promptly decreased following administration of andexanet alfa, regardless of the FXa inhibitor taken. The change in anti-FXa activity from the screening period through 14 hours after starting andexanet alfa (12 hours after the end of infusion) is shown in Fig.1. The median percent reduction in anti-FXa activity from baseline to nadir was 95.4% in patients taking apixaban, 96.1% in patients taking rivaroxaban, and 82.2% in patients taking edoxaban (Fig.1 and Supplemental Table 2).

Fig.1. Anti-FXa Activity Over Time in Patients Who Received A) Apixaban, B) Rivaroxaban, and C) Edoxaban (Efficacy Population)

FXa, factor Xa; CT, computed tomography; MRI, magnetic resonance imaging.

Good/excellent hemostasis was defined as a ≤ 35% increase in hematoma volume/thickness on a CT or MRI scan at 12 hours compared with baseline. Poor/no hemostasis was defined as a ＞ 35% increase in volume/thickness on a CT or MRI scan at 12 hours compared with baseline.

Supplemental Table 2.Anti-FXa Activity

	Apixaban (n = 5)	Rivaroxaban (n = 7)	Edoxaban (n = 4)
Baseline, ng/mL, median (range)	215.1 (86.4, 335.0)	180.8 (79.9, 392.3)	87.9 (51.5, 131.5)
On-treatment nadir, ng/mL, median (range)	12.7 (4.0, 18.3)	6.2 (4.0, 17.7)	14.8 (11.2, 18.3)
Change from baseline, ng/mL, median (range)	−201.7 (−322.3, −82.4)	−176.8 (−374.6, −75.5)	−73.1 (−113.2, −40.3)
Percent change, median (range)	−95.4 (−96.2, −93.7)	−96.1 (−98.3, −94.3)	−82.2 (−87.9, −77.7)

FXa, factor Xa.

Administration of andexanet alfa resulted in excellent or good hemostasis in 14 of 16 patients in the overall efficacy population (88%; Table 2), all 5 patients taking apixaban, 6 of 7 taking rivaroxaban, and 3 of 4 taking edoxaban. Excellent or good hemostasis was achieved for all 3 patients receiving high-dose andexanet alfa and 11 of 13 receiving low-dose andexanet alfa.

Table 2.Hemostatic Efficacy at 12 Hours After Andexanet Alfa Treatment (Efficacy Population)

	Patients, n	Excellent	Good	Poor/none
Overall	16	13	1	2§
FXa inhibitor
Apixaban	5	5	0	0
Rivaroxaban	7	5	1	1
Edoxaban	4	3	0	1
Andexanet alfa dose
Low dose	13	10	1	2
High dose	3	3	0	0

FXa, factor Xa.

^§In 1 male patient 78 years of age with subarachnoid hemorrhage, infusion was started later than acceptable; the time interval between end of bolus until start of infusion was 21.0 minutes. Atrial fibrillation was the indication for FXa inhibitor treatment, and time from last dose to andexanet alfa initiation was 9 hours and 15 minutes. In 1 other female patient 91 years of age with intracerebral hemorrhage, indication for FXa inhibitor treatment was venous thromboembolism and time from last dose to andexanet alfa initiation was 16.52 hours.

Thrombin generation increased during treatment with andexanet alfa, demonstrated by reduced endogenous thrombin potential (ETP) lag time and time to peak and increased ETP peak height and velocity index observed after the andexanet alfa bolus that was maintained during infusion (Supplemental Table 3). ETP values decreased to within the normal ranges following cessation of andexanet alfa.

Supplemental Table 3.Endogenous Thrombin Potential Parameters

ETP parameter	Baseline	EOB	EOI	EOI＋ 4 hours	EOI＋ 8 hours	EOI＋ 12 hours	EOI＋ 18 hours	EOI＋ 24 hours	Day 3	Day 30
ETP lag time, minutes
Apixaban
Mean (SD)	6.7 (3.43)	2.5 (0.75)	2.5 (0.74)	3.2 (1.20)	3.4 (1.12)	3.4 (1.13)	3.6 (1.34)	3.4 (1.08)	4.8 (1.35)	7.7 (4.91)
Median (IQR)	5.7 (4.3, 9.6)	2.3 (2.2, 2.8)	2.4 (2.0, 2.7)	3.4 (2.0, 4.0)	3.6 (2.3, 4.3)	3.1 (3.0, 4.0)	3.0 (2.7, 4.8)	3.0 (2.7, 4.0)	4.4 (3.8, 5.9)	6.1 (4.3, 7.9)
Rivaroxaban
Mean (SD)	7.3 (1.57)	2.5 (0.48)	2.4 (0.43)	3.7 (0.92)	3.8 (0.95)	4.0 (1.06)	3.8 (1.19)	3.7 (1.40)	4.3 (2.83)	8.0 (8.71)
Median (IQR)	7.0 (6.3, 9.1)	2.3 (2.1, 3.1)	2.3 (2.0, 3.0)	4.0 (2.6, 4.5)	4.0 (3.2, 4.6)	4.2 (3.0, 5.0)	4.0 (2.4, 4.8)	3.2 (2.5, 4.9)	2.7 (2.6, 4.8)	3.3 (3.0,12.8)
Edoxaban
Mean (SD)	7.1 (3.59)	3.0 (1.37)	3.0 (1.20)	3.7 (1.59)	4.1 (1.51)	3.1 (0.33)	3.7 (1.56)	3.7 (1.21)	4.8 (1.86)	7.9 (3.37)
Median (IQR)	5.9 (4.6, 9.7)	2.5 (2.2, 3.8)	2.7 (2.2, 3.8)	3.1 (2.7, 4.7)	3.4 (3.0, 5.8)	3.1 (2.9, 3.3)	3.2 (2.7, 4.7)	3.3 (3.0, 4.4)	4.2 (3.6, 5.9)	6.4 (5.6,11.8)
ETP peak height, nmol
Apixaban
Mean (SD)	60.7 (56.20)	282.1 (40.50)	266.4 (24.96)	151.6 (89.70)	156.3 (90.72)	137.8 (90.69)	175.0 (90.35)	157.5 (77.51)	122.2 (91.64)	95.5 (104.81)
Median(IQR)	31.1 (28.7, 84.8)	315.4)	255.7 (248.4, 281.1)	92.7 (92.0, 216.1)	103.6 (91.0, 224.5)	98.9 (70.6, 234.2)	148.9 (103.0, 264.1)	138.5 (94.0, 224.5)	96.0 (55.5, 145.0)	34.2 (16.3, 205.8)
Rivaroxaban
Mean (SD)	27.9 (14.53)	224.4 (60.46)	240.6 (48.57)	100.3 (31.90)	95.1 (40.22)	107.5 (47.96)	122.0 (54.94)	121.9 (57.46)	147.6 (72.12)	141.2 (99.68)
Median(IQR)	23.6 (15.3, 45.2)	224.1 (169.9, 281.8)	225.3 (197.3, 293.2)	109.0 (80.1, 120.0)	97.1 (63.3, 142.8)	100.3 (69.5, 165.9)	117.3 (87.9, 185.0)	93.7 (69.4, 190.3)	183.9 (65.9, 192.8)	179.9 (14.6, 189.7)
Edoxaban
Mean (SD)	52.6 (32.49)	232.0 (44.28)	232.2 (40.98)	178.7 (52.02)	170.2 (55.12)	217.8 (15.46)	208.0 (45.63)	213.9 (36.74)	151.7 (68.79)	71.3 (56.01)
Median(IQR)	47.5 (28.2, 76.9)	234.9 (197.2, 266.8)	233.2 (200.4, 263.9)	195.9 (145.4, 212.0)	154.0 (125.1, 231.7)	217.8 (206.9, 228.7)	215.8 (170.1, 246.0)	210.3 (183.1, 244.6)	164.6 (109.4, 194.0)	48.4 (30.3, 135.1)
ETP time to peak, minutes
Apixaban
Mean (SD)	11.1 (4.48)	4.9 (0.73)	5.1 (0.85)	5.9 (1.34)	6.0 (1.22)	6.3 (1.17)	6.4 (1.61)	6.2 (1.06)	9.3 (4.33)	13.4 (7.17)
Median(IQR)	8.9 (8.2, 14.7)	4.6 (4.3, 5.3)	4.7 (4.4, 5.3)	6.0 (5.2, 6.8)	6.1 (5.0, 6.9)	5.9 (5.8, 7.0)	5.7 (5.3, 7.7)	6.0 (5.6, 6.9)	7.5 (6.8, 9.5)	10.1 (7.6, 19.6)
Rivaroxaban
Mean (SD)	19.8 (4.16)	5.5 (1.10)	5.1 (0.68)	11.2 (4.16)	11.8 (4.60)	11.1 (4.23)	9.8 (3.86)	9.3 (3.81)	9.6 (6.62)	12.6 (9.90)
Median (IQR)	19.4 (16.7, 20.6)	5.3 (4.7, 6.0)	5.1 (4.4, 5.9)	10.4 (9.0,14.0)	12.3 (7.0,15.4)	11.0 (6.0,14.3)	10.7 (5.7,11.8)	9.9 (5.0,12.1)	8.3 (5.2, 9.0)	8.0 (6.0, 24.4)
Edoxaban
Mean (SD)	17.3 (5.38)	6.2 (1.98)	6.1 (1.82)	7.9 (3.49)	8.6 (3.02)	6.3 (0.88)	7.1 (2.57)	7.1 (1.86)	8.7 (2.03)	12.1 (3.29)
Median(IQR)	17.2 (12.9, 21.6)	5.4 (5.0, 7.4)	5.5 (4.8, 7.3)	6.7 (5.7, 10.2)	8.3 (5.8, 11.8)	6.3 (5.7, 6.9)	6.2 (5.5, 8.8)	6.5 (6.0, 8.1)	8.2 (7.1, 10.3)	11.8 (9.0, 15.6)
ETP velocity index, nmol/ min
Apixaban
Mean (SD)	16.1 (14.65)	120.4 (26.97)	105.2 (12.36)	60.9 (46.07)	62.8 (41.37)	47.4 (30.75)	66.1 (38.61)	55.2 (27.88)	40.2 (38.09)	27.6 (34.51)
Median (IQR)	9.7 (7.8, 23.1)	110.8 (103.3,118.3)	112.6 (95.9,113.6)	36.5 (33.2,67.2)	40.6 (35.7,81.1)	37.1 (23.6,78.7)	55.8 (35.8,99.0)	46.2 (32.7,77.9)	31.4 (14.6,48.3)	4.1 (2.9,58.1)
Rivaroxaban
Mean (SD)	2.5 (1.54)	81.2 (32.14)	90.5 (24.86)	17.6 (11.97)	18.3 (16.49)	23.5 (22.32)	31.1 (28.50)	33.7 (30.73)	44.5 (32.62)	43.3 (32.67)
Median(IQR)	2.1 (1.5, 3.6)	92.0 (46.3, 101.3)	96.2 (65.8, 112.2)	17.2 (8.6, 20.3)	11.4 (6.2, 37.9)	14.2 (7.5, 53.3)	21.9 (12.9, 69.1)	17.6 (8.0, 74.6)	46.0 (11.2, 77.6)	54.0 (3.0, 71.1)
Edoxaban
Mean (SD)	6.2 (5.53)	75.7 (22.85)	78.6 (24.14)	51.6 (28.33)	45.2 (33.41)	70.3 (16.98)	65.3 (25.92)	65.7 (17.73)	41.6 (23.67)	15.8 (8.72)
Median(IQR)	4.6 (2.4,10.0)	81.3 (60.8,90.7)	982.2 (63.7,3.6)	54.7 (31.5,71.7)	31.5 (20.9,83.3)	70.3 (58.3,82.3)	69.3 (46.5,84.2)	69.5 (53.2,78.2)	39.8 (25.9,57.4)	14.1 (8.0,25.2)

ETP, endogenous thrombin potential; EOB, end of bolus; EOI, end of infusion; SD, standard deviation; IQR, interquartile range.

Of the 19 patients who received andexanet alfa, 12 received ≥ 1 anticoagulation dose within 30 days after treatment. Median time to resumption of anticoagulant therapy was 3.5 (interquartile range: 2.0, 8.5) days.

Neurologic status remained relatively similar throughout the study (Supplemental Table 4).

Supplemental Table 4.Neurologic Status for Patients with ICH (Efficacy Population)

Parameter	Baseline	1 hour	12 hours	Day 30
GCS score
n	16	16	16	15
Median (range)	14.5 (11, 15)	14.5 (3, 15)	14.0 (3, 15)	15.0 (6, 15)
NIHSS score
n	16	16	16	15
Median (range)	5.5 (0, 25)	7.0 (0, 38)	5.5 (0, 38)	3.0 (0, 36)
mRS score at Day 30
n	16§
Median (range)	3.0 (0, 6)

ICH, intracranial hemorrhage; GCS, Glasgow Coma Scale; NIHSS, National Institutes of Health Stroke Scale; mRS, modified Rankin Scale.

Note: Baseline GCS scores and mRS scores were obtained in the 3-hour time window prior to the start of the andexanet alfa bolus; baseline NIHSS scores were obtained in the 15-minute time window prior to the start of the andexanet alfa bolus.

^§mRS scores: 0 (n = 3), 1 (n = 2), 2 (n = 2), 3 (n = 2), 4 (n = 2), 5 (n = 4), and 6 (n = 1).

Safety

Over a follow-up period of 30 days, treatment with andexanet alfa was generally safe and well tolerated. TEAEs occurred in 13 out of 19 (68%) patients (Supplemental Table 5). The most commonly reported TEAE was liver dysfunction shown on a blood test (n=4, 21%). The abnormalities of the liver panel blood tests were considered mild for all 4 patients. Two (11%) patients had a thrombotic event; 1 patient developed a mild transient ischemic attack on Day 1, and another patient developed a severe ischemic stroke on Day 2. TEAEs classified as possibly related to treatment occurred in 2 (11%) patients: liver dysfunction and hyponatremia in 1 patient and ischemic stroke and liver dysfunction in the other patient. All other TEAEs were classified as unlikely or not related to treatment. Serious AEs were reported in 5 (26%) patients, all of whom were taking low-dose andexanet alfa: subarachnoid hematoma, status epilepticus, ischemic stroke, and subdural hematoma in 1 patient each and pneumonia and respiratory failure in 1 patient. The serious AEs of pneumonia and respiratory failure led to the 1 death (5.3%) reported in this study; those serious AEs occurred in a 91-year-old female taking edoxaban 30 mg once daily for the prevention of venous thromboembolism. This patient developed pneumonia as a complication of intracerebral hemorrhage, and death occurred on Day 28 during the follow-up period.

Supplemental Table 5.Treatment-emergent Adverse Events

TEAE＊ (MedDRA preferred term)	Apixaban (n = 6)	Rivaroxaban (n = 8)	Edoxaban (n = 5)
Total TEAEs, n (%)	4 (67)	5 (63)	4 (80)
Liver dysfunction＊	0	2 (25)	2 (40)
Constipation	1 (17)	2 (25)	0
Hyponatremia＊	0	0	2 (40)
Urinary tract infection	2 (33)	0	0
Aphasia	0	1 (13)	0
Atrial fibrillation	0	1 (13)	0
Brain edema	0	1 (13)
Intracerebral hemorrhage	1 (17)	0	0
Cerebral infarction＊, †	0	0	1 (20)
Delirium	0	1 (13)	0
Dysuria	0	1 (13)	0
Hyperkalemia	0	0	1 (20)
Hypoalbuminemia	0	0	1 (20)
Laceration	1 (17)	0	0
Meningitis	0	1 (13)	0
Pneumonia†	0	0	1 (20)
Pyrexia	0	1 (13)	0
Respiratory failure†	0	0	1 (20)
Skin mass	0	0	1 (20)
Status epilepticus†	1 (17)	0	0
Subarachnoid hemorrhage†	0	1 (13)	0
Subdural hematoma†	0	0	1 (20)
Transient ischemic attack	0	0	1 (20)
Vomiting	0	0	1 (20)

MedDRA, Medical Dictionary for Regulatory Activities; TEAE, treatment-emergent adverse event.

^＊TEAEs classified as possibly related to treatment were reported in 2 (11%) patients: liver disorder in 1 patient and liver dysfunction and hypernatremia in 1 patient.

^†Serious TEAEs were reported in 5 (26%) patients: subarachnoid hematoma, status epilepticus, cerebral infarction, and subdural hematoma in 1 patient each and pneumonia and respiratory failure in 1 patient.

Antibodies against andexanet alfa were identified in 1 patient (Supplemental Table 6).

Supplemental Table 6.Antibodies to Andexanet Alfa, FX, FXa, and Host Cell Proteins

Antibody	Results	n (%)	Number tested
Anti–andexanet alfa
Baseline	Positive	0	0
	Titer 1:10	0	0
Day 30/45	Positive	1 (5.6)	18
	Titer 1:10	1 (5.6)	18
Anti-FX
Baseline	Positive	0	0
Day 30	Positive	0	0
Anti-FXa
Baseline	Positive	0	0
Day 30	Positive	0	0
Anti–host cell protein
Baseline	Positive	0	0
Day 30	Positive	0	0

FX, factor X; FXa, factor Xa.

Discussion

This prespecified subgroup analysis of ANNEXA-4 included 19 Japanese patients with acute major bleeding. Andexanet alfa rapidly reduced anti-FXa activity with favorable hemostatic efficacy in the majority of those patients. Andexanet alfa was generally safe and well-tolerated, with 1 death, 2 thrombotic events, and 5 serious AEs reported.

Few agents have been developed for anticoagulant therapy reversal in patients experiencing acute major bleeding^{14, 23, 24)}. Andexanet alfa is the only agent approved in Japan and by the US Food and Drug Administration under Accelerated Approval, with conditional approval by the European Medicines Agency, for specific anticoagulation reversal in patients with life-threatening or uncontrolled bleeding associated with direct FXa inhibitors^{25, 26)}. There are, however, important differences in the epidemiologic characteristics and clinical practice of stroke management between Japan and western countries to consider, including a higher prevalence of ICH and more frequent administration of edoxaban relative to enoxaparin in Japan. These differences provided the rationale for investigating hemostatic outcomes of Japanese patients enrolled in ANNEXA-4.

In the present Japanese cohort of ANNEXA-4, percent reductions in anti-FXa activity and hemostasis results are consistent with those of the overall study population^{14, 15)}; anti-FXa activity decreased by 77.7% to 98.3%, and 88% of patients achieved excellent or good hemostasis. ETP values returned to within normal ranges following andexanet alfa treatment.

A total of 2 (11%) patients experienced thrombotic events, both of which occurred within the initial 2 days. The rate was similar to the global results within 30 days (50 patients [10.4%], 19 of whom had events within the first 6 days)¹⁵⁾. These early thrombotic events would not be due to andexanet alfa itself but instead would be due to underlying atrial fibrillation, withdrawal of FXa inhibitors, and activation of coagulation concomitant with bleeding. The rate of thromboembolic events in the global study of andexanet alfa was numerically higher than that observed in the REVERSE-AD (NCT02104947) study, in which patients on dabigatran were treated with idarucizumab for life-threatening or uncontrolled bleeding (4.8% within 30 days and 6.8% within 90 days)²³⁾ and than that reported in the global study of 4-factor prothrombin complex concentrate in patients on warfarin (7.8%)²⁴⁾. Similar trends have been observed in Asian populations, with reported rates of thrombotic events ranging from approximately 3% to 6% in patients treated with idarucizumab for the reversal of the effects of dabigatran^{27, 28)}.

For the safety profile, liver dysfunction was identified in 4 (21%) patients, although none of the cases were severe. The reason that liver dysfunction was only reported in the Japanese population may be attributed to differences in protocol-defined laboratory measurements and routine clinical practice in Japan. Laboratory measurements must be tested several times during the course of a hospital stay in Japan, whereas the ANNEXA-4 protocol did not specify post-exposure clinical chemistry, which includes the liver function test. The threshold to report an abnormality in a laboratory test as an AE could also have varied by investigator. Therefore, caution should be applied to interpretation regarding a comparison with other participants in ANNEXA-4. Andexanet alfa has no known hepatic route of metabolism, a biologic profile that is inconsistent with a potential relationship to liver dysfunction. As with other protein preparations, andexanet alfa is expected to be metabolized into peptides or amino acids and is not expected to be catabolized by liver enzymes. No liver dysfunction AEs were reported from the phase 2 study conducted in healthy Japanese participants²⁹⁾. Given that elevated liver enzymes can commonly occur following an ICH event³⁰⁾, it is plausible that the rate of liver dysfunction observed in this study could be due to the index event of ICH rather than andexanet alfa treatment.

Limitations of this study include those inherent to the open-label, single-arm study design, including the lack of a control group. Additionally, the sample size of 19 patients was too small to assess the rates of safety events as compared to other larger size studies. Because patients with a hematoma volume ＞60 mL were excluded from the analysis, further investigation is warranted to verify outcomes in patients with greater hematoma volume. Patients who were scheduled to undergo surgery ＜12 hours after the end of andexanet alfa infusion were also excluded. Along with these considerations, future studies should also aim to confirm these findings in Japanese patients experiencing gastrointestinal bleeding and bleeding at other sites.

Conclusion

This subgroup analysis of ANNEXA-4 demonstrated that treatment with andexanet alfa rapidly reduced anti-FXa activity, with favorable hemostatic efficacy in Japanese patients with ICH. The potential for thromboembolic events should be considered after rapid reversal of anti-FXa activity by andexanet alfa, as similarly stressed in previous studies, although there is no clear causal relationship between those events and andexanet alfa use.

Acknowledgements

This study was funded by Alexion, AstraZeneca Rare Disease. Under the direction of the authors, Melanie Chen, PharmD (Lumanity Communications Inc.) provided medical writing assistance. Alexion, AstraZeneca Rare Disease (the sponsor) funded the medical writing assistance and provided a formal review of the publication. Authors retained control and final authority of publication content and decisions, including the choice of journal.

Conflicts of Interest

Kazunori Toyoda: Personal fees from Daiichi Sankyo, Otsuka, Novartis, Bayer Yakuhin, and Bristol Myers Squibb outside the submitted work.

Shuji Arakawa: Personal fees from Daiichi Sankyo and Bayer.

Masayuki Ezura: No conflicts of interest.

Rei Kobayashi: No conflicts of interest.

Yoshihide Tanaka: No conflicts of interest.

Shu Hasegawa: No conflicts of interest.

Shigeo Yamashiro: No conflicts of interest.

Yoji Komatsu: No conflicts of interest.

Yuka Terasawa: No conflicts of interest.

Tomohiko Masuno: No conflicts of interest.

Hiroshi Kobayashi: Employee of Alexion, AstraZeneca Rare Disease.

Suzuko Oikawa: No conflicts of interest.

Masahiro Yasaka: Lecture fees from Nippon, Boehringer Ingelheim, Bayer, and Daiichi Sankyo.

References

• 1)  Takagi M, Tanaka K, Miwa K, Sasaki M, Koga M, Hirano T, Kamiyama K, Yagita Y, Nagakane Y, Hoshino H, Terasaki T, Yakushiji Y, Kudo K, Ihara M, Yoshimura S, Yamaguchi Y, Shiozawa M, Toyoda K and BAT2 Investigators: The bleeding with antithrombotic therapy study 2: Rationale, design, and baseline characteristics of the participants. Eur Stroke J, 2020; 5: 423-431
• 2)  Ono K, Iwasaki YK, Akao M, Ikeda T, Ishii K, Inden Y, Kusano K, Kobayashi Y, Koretsune Y, Sasano T, Sumitomo N, Takahashi N, Niwano S, Hagiwara N, Hisatome I, Furukawa T, Honjo H, Maruyama T, Murakawa Y, Yasaka M, Watanabe E, Aiba T, Amino M, Itoh H, Ogawa H, Okumura Y, Aoki-Kamiya C, Kishihara J, Kodani E, Komatsu T, Sakamoto Y, Satomi K, Shiga T, Shinohara T, Suzuki A, Suzuki S, Sekiguchi Y, Nagase S, Hayami N, Harada M, Fujino T, Makiyama T, Maruyama M, Miake J, Muraji S, Murata H, Morita N, Yokoshiki H, Yoshioka K, Yodogawa K, Inoue H, Okumura K, Kimura T, Tsutsui H, Shimizu W and Japanese Circulation Society and Japanese Heart Rhythm Society Joint Working Group: JCS/JHRS 2020 guideline on pharmacotherapy of cardiac arrhythmias. Circ J, 2022; 86: 1790-1924
• 3)  Yamashita T, Suzuki S, Inoue H, Akao M, Atarashi H, Ikeda T, Okumura K, Koretsune Y, Shimizu W, Tsutsui H, Toyoda K, Hirayama A, Yasaka M, Yamaguchi T, Teramukai S, Kimura T, Kaburagi J and Takita A: Two-year outcomes of more than 30 000 elderly patients with atrial fibrillation: results from the All Nippon AF In the Elderly (ANAFIE) Registry. Eur Heart J Qual Care Clin Outcomes, 2022; 8: 202-213
• 4)  Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, Waldo AL, Ezekowitz MD, Weitz JI, Spinar J, Ruzyllo W, Ruda M, Koretsune Y, Betcher J, Shi M, Grip LT, Patel SP, Patel I, Hanyok JJ, Mercuri M and Antman EM: Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med, 2013; 369: 2093-2104
• 5)  Granger CB, Alexander JH, McMurray JJV, Lopes RD, Hylek EM, Hanna M, Al-Khalidi HR, Ansell J, Atar D, Avezum A, Bahit MC, Diaz R, Easton JD, Ezekowitz JA, Flaker G, Garcia D, Geraldes M, Gersh BJ, Golitsyn S, Goto S, Hermosillo AG, Hohnloser SH, Horowitz J, Mohan P, Jansky P, Lewis BS, Lopez-Sendon JL, Pais P, Parkhomenko A, Verheugt FWA, Zhu J, Wallentin L and ARISTOTLE Committees and Investigators: Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med, 2011; 365: 981-992
• 6)  Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, Breithardt G, Halperin JL, Hankey GJ, Piccini JP, Becker RC, Nessel CC, Paolini JF, Berkowitz SD, Fox KA, Califf RM and ROCKET AF Investigators: Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med, 2011; 365: 883-891
• 7)  Agnelli G, Buller HR, Cohen A, Curto M, Gallus AS, Johnson M, Masiukiewicz U, Pak R, Thompson J, Raskob GE, Weitz JI and Caravaggio Investigators: Oral apixaban for the treatment of acute venous thromboembolism. N Engl J Med, 2013; 369: 799-808
• 8)  Einstein Investigators, Bauersachs R, Berkowitz SD, Brenner B, Buller HR, Decousus H, Gallus AS, Lensing AW, Misselwitz F, Prins MH, Raskob GE, Segers A, Verhamme P, Wells P, Agnelli G, Bounameaux H, Cohen A, Davidson BL, Piovella F and Schellong S: Oral rivaroxaban for symptomatic venous thromboembolism. N Engl J Med, 2010; 363: 2499-2510
• 9)  Hokusai VTEI, Buller HR, Decousus H, Grosso MA, Mercuri M, Middeldorp S, Prins MH, Raskob GE, Schellong SM, Schwocho L, Segers A, Shi M, Verhamme P and Wells P: Edoxaban versus warfarin for the treatment of symptomatic venous thromboembolism. N Engl J Med, 2013; 369: 1406-1415
• 10)  Ikeda T, Ogawa S, Kitazono T, Nakagawara J, Minematsu K, Miyamoto S, Murakawa Y, Takeichi M, Ohashi Y, Okayama Y, Sunaya T and Yamanaka S: Real-world outcomes of the Xarelto Post-Authorization Safety & Effectiveness Study in Japanese Patients with Atrial Fibrillation (XAPASS). J Cardiol, 2019; 74: 60-66
• 11)  Inoue H, Umeyama M, Yamada T, Hashimoto H, Komoto A and Yasaka M: Safety and effectiveness of apixaban in Japanese patients with nonvalvular atrial fibrillation in clinical practice: a regulatory postmarketing surveillance, the STANDARD study. J Arrhythm, 2019; 35: 506-514
• 12)  Yamashita T, Koretsune Y, Nagao T and Shiosakai K: Safety and effectiveness of edoxaban in Japanese patients with nonvalvular atrial fibrillation: final report of a two-year postmarketing surveillance study (ETNA-AF-Japan). J Arrhythm, 2021; 37: 370-383
• 13)  Lu G, DeGuzman FR, Hollenbach SJ, Karbarz MJ, Abe K, Lee G, Luan P, Hutchaleelaha A, Inagaki M, Conley PB, Phillips DR and Sinha U: A specific antidote for reversal of anticoagulation by direct and indirect inhibitors of coagulation factor Xa. Nat Med, 2013; 19: 446-451
• 14)  Connolly SJ, Crowther M, Eikelboom JW, Gibson CM, Curnutte JT, Lawrence JH, Yue P, Bronson MD, Lu G, Conley PB, Verhamme P, Schmidt J, Middeldorp S, Cohen AT, Beyer-Westendorf J, Albaladejo P, Lopez-Sendon J, Demchuk AM, Pallin DJ, Concha M, Goodman S, Leeds J, Souza S, Siegal DM, Zotova E, Meeks B, Ahmad S, Nakamya J and Milling TJ Jr: Full study report of andexanet alfa for bleeding associated with factor Xa inhibitors. N Engl J Med, 2019; 380: 1326-1335
• 15)  Milling TJ Jr, Middeldorp S, Xu L, Koch B, Demchuk A, Eikelboom JW, Verhamme P, Cohen AT, Beyer-Westendorf J, Gibson CM, Lopez-Sendon J, Crowther M, Shoamanesh A, Coppens M, Schmidt J, Albaladejo P, Connolly SJ and ANNEXA-4 Investigators: Final study report of andexanet alfa for major bleeding with factor Xa inhibitors. Circulation, 2023; 147: 1026-1038
• 16)  Siegal DM, Curnutte JT, Connolly SJ, Lu G, Conley PB, Wiens BL, Mathur VS, Castillo J, Bronson MD, Leeds JM, Mar FA, Gold A and Crowther MA: Andexanet alfa for the reversal of factor Xa inhibitor activity. N Engl J Med, 2015; 373: 2413-2424
• 17)  Toyoda K, Inoue M and Koga M: Small but steady steps in stroke medicine in Japan. J Am Heart Assoc, 2019; 8: e013306
• 18)  Venketasubramanian N, Yoon BW, Pandian J and Navarro JC: Stroke epidemiology in South, East, and South-East Asia: a review. J Stroke, 2017; 19: 286-294
• 19)  Hori M, Matsumoto M, Tanahashi N, Momomura S, Uchiyama S, Goto S, Izumi T, Koretsune Y, Kajikawa M, Kato M, Ueda H, Iwamoto K and Tajiri M: Rivaroxaban vs. warfarin in Japanese patients with atrial fibrillation the J-ROCKET AF study. Circ J, 2012; 76: 2104-2111
• 20)  Yajima T, Higashimori M, Takata C and Sasabe T: Andexanet alfa (ONDEXXYA® for intravenous injection 200 mg), a reversal agent for direct factor Xa inhibitors: pharmacological characteristics and clinical evidence. Nihon Yakurigaku Zasshi, 2023; 158: 89-100
• 21)  Connolly SJ, Milling TJ Jr, Eikelboom JW, Gibson CM, Curnutte JT, Gold A, Bronson MD, Lu G, Conley PB, Verhamme P, Schmidt J, Middeldorp S, Cohen AT, Beyer-Westendorf J, Albaladejo P, Lopez-Sendon J, Goodman S, Leeds J, Wiens BL, Siegal DM, Zotova E, Meeks B, Nakamya J, Lim WT, Crowther M and ANNEXA-4 Investigators: Andexanet alfa for acute major bleeding associated with factor Xa inhibitors. N Engl J Med, 2016; 375: 1131-1141
• 22)  Schulman S and Kearon C: Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost, 2005; 3: 692-694
• 23)  Pollack CV Jr, Reilly PA, van Ryn J, Eikelboom JW, Glund S, Bernstein RA, Dubiel R, Huisman MV, Hylek EM, Kam CW, Kamphuisen PW, Kreuzer J, Levy JH, Royle G, Sellke FW, Stangier J, Steiner T, Verhamme P, Wang B, Young L and Weitz JI: Idarucizumab for dabigatran reversal - Full cohort analysis. N Engl J Med, 2017; 377: 431-441
• 24)  Sarode R, Milling TJ Jr, Refaai MA, Mangione A, Schneider A, Durn BL and Goldstein JN: Efficacy and safety of a 4-factor prothrombin complex concentrate in patients on vitamin K antagonists presenting with major bleeding: a randomized, plasma-controlled, phase IIIb study. Circulation, 2013; 128: 1234-1243
• 25)  ANDEXXA® (coagulation factor Xa (recombinant), inactivated-zhzo). Lyophilized powder for solution for intravenous injection [package insert]. South San Francisco, CA: Portola Pharmaceuticals, Inc.; 2018
• 26)  ONDEXXYA® 200 mg powder for solution for infusion [summary of product characteristics]. Dublin, Ireland: Alexion Pharma Operations Ltd.; 2019
• 27)  Yasaka M, Yokota H, Suzuki M, Asakura H, Yamane T, Ogi Y, Ochiai K and Nakayama D: Idarucizumab for emergency reversal of anticoagulant effects of dabigatran: interim results of a Japanese post-marketing surveillance study. Cardiology and Therapy, 2020; 9: 167-188
• 28)  Dai JW, Wang CH, Chu CL and Liao SC: Effectiveness and safety of dabigatran reversal with idarucizumab in the Taiwanese population: A comparison based on eligibility for inclusion in clinical trials. Medicina (Kaunas), 2023; 59: 881
• 29)  ClinicalTrials.gov: NCT03310021. A healthy volunteer pharmacokinetics (PK)/pharmacodynamics (PD), safety and tolerability study of andexanet in healthy Japanese and Caucasian subjects. 2023; https://clinicaltrials.gov/study/NCT03310021?term=NCT03310021&rank=1. Accessed March 9, 2023
• 30)  Meythaler JM, Hazlewood J, DeVivo MJ and Rosner M: Elevated liver enzymes after nontraumatic intracranial hemorrhages. Arch Phys Med Rehabil, 1998; 79: 766-771