Anti-Xa Activity and Event Risk in Patients With Direct Factor Xa Inhibitors Initiated Early After Stroke

Abstract

Background: Measuring anti-Xa activity (AXA) has been reported as useful for predicting future risk of hemorrhagic and ischemic events in stroke patients taking direct factor Xa inhibitors. We evaluated AXA levels of rivaroxaban or apixaban in acute stroke patients with non-valvular atrial fibrillation.

Methods and Results: This was a single-center, prospective, observational study. Consecutive patients with acute ischemic stroke or transient ischemic attack who were admitted within 7 days of onset and started taking rivaroxaban or apixaban for NVAF between January 2012 and April 2017 were enrolled. AXA was measured at 2 time points: just before (AXA_trough) and 4 h after (AXA_peak) taking rivaroxaban or apixaban on the 2nd day or later of administration. Of 156 patients taking rivaroxaban, hemorrhagic events occurred in 13. Patients with hemorrhagic events had higher AXA_peak than those without [median (interquartile range): 1.93 (1.11–3.75) vs. 1.35 (0.80–2.00) IU/mL; P<0.01]. Multivariable-adjusted Cox models showed that AXA_peak was independently related to the incidence of hemorrhagic events. Of 169 patients taking apixaban, hemorrhagic events occurred in 11. Patients with hemorrhagic events had higher AXA_trough [2.78 (1.90–3.53) vs. 1.42 (0.93–2.08) IU/mL, P<0.01] and AXA_peak [4.05 (3.44–4.72) vs. 2.43 (1.79–3.35) IU/mL, P<0.01] than those without. Both AXA_trough and AXA_peak were independently related to the incidence of hemorrhagic events.

Conclusions: In these patients who started rivaroxaban or apixaban early after stroke, AXA levels in the early period were related to future hemorrhagic events.

Direct factor Xa inhibitors (e.g., rivaroxaban and apixaban) are effective for preventing stroke and systemic embolism caused by non-valvular atrial fibrillation (NVAF).^1–4 Although Xa inhibitors do not require routine monitoring of their anticoagulation effects in daily practice,⁵ laboratory measurement of their anticoagulant activity may be desirable in special clinical settings, including bleeding, indication for thrombolysis, overdosing, emergency surgery, etc.^6–13 It has been reported that patients who had hemorrhagic and ischemic events tended to have relatively high and low plasma concentrations, respectively, on admission after onset.¹⁴ Anti-Xa activity (AXA) was previously reported to be useful for monitoring the plasma Xa inhibitor concentration.⁹ Measuring the AXA of both rivaroxaban and apixaban could be useful for predicting future hemorrhagic or ischemic events.^6,15–21 The aim of the present study was to clarify whether the AXA levels of rivaroxaban or apixaban in acute stroke patients were associated with subsequent hemorrhagic and ischemic events.

Methods

All authors have full access to the data, analytic methods, and study materials. The data that support the findings of this study are available from the correspondence author on reasonable request.

Participants

Consecutive patients with an ischemic stroke or transient ischemic attack (TIA) who had NVAF and had secondary prevention with rivaroxaban or apixaban initiated during acute hospitalization from January 2012 to April 2017 were enrolled from a prospective database of patients admitted to a stroke center within 7 days of onset (National Cerebral and Cardiovascular Center Stroke Registry, ClinicalTrials.gov: NCT02251665).^22–24 Approval for this study was obtained from the Ethics Committee of the National Cerebral and Cardiovascular Center, and written, informed consent was given by the patients or their families.

Procedures

The indication and choice of oral anticoagulant, including Xa inhibitors, depended on the discretion of the physicians in charge without randomization. The official dosage of rivaroxaban in Japan was based on a domestic preapproval trial: 15 mg daily for patients with creatinine clearance (CCr) ≥50 mL/min, and 10 mg daily for those with CCr 15–49 mL/min. The apixaban dosage was also chosen according to the officially recommended regimen, which is 5 mg twice daily for patients with 0 or 1 of the following factors and 2.5 mg twice daily for patients with 2 or 3 of the following factors: (1) age ≥80 years; (2) weight ≤60 kg; and (3) serum creatinine ≥1.5 mg/dL.

Baseline clinical characteristics, risk indices (CHADS₂, CHA₂DS₂-VASc, HAS-BLED),^25–27 and outcomes were recorded. Renal function was expressed as CCr using the Cockcroft and Gault equation.²⁸ Blood sampling for AXA was performed at least 2 days after initiating rivaroxaban or apixaban to ensure the steady-state concentration had been achieved. Two venous blood samples were collected each time in citrate-containing tubes using a 21-gauge needle just before (trough) and 4 h (peak) after drug administration. Because the maximum concentration of rivaroxaban in the fasting state has been reported to occur 2–4 h after tablet intake and to be delayed by 2 h if taken with food²⁹ and for apixaban it is 1–4 h and a delay of 2 h with food, the 4-h sampling point was used to capture the maximum concentrations of rivaroxaban and apixaban.^30,31 For 1 of the 2 tubes, following double centrifugation at 2,500 g for 15 min, platelet-poor plasma was collected, quick-frozen, and stored at −80℃ until the AXA analysis was performed [anti-factor Xa chromogenic assay, HemosIL Heparin Calibrators (Instrumentation Laboratory, Milano, Italy)]. AXA chromogenic assays have previously been shown to have acceptable accuracy and precision.³²

Outcome measures were hemorrhagic and ischemic events. Hemorrhagic events were defined as major hemorrhagic events according to the classification of the International Society on Thrombosis and Haemostasis and non-major clinically relevant hemorrhagic events. Major hemorrhagic events included: fatal outcome; involvement of a critical anatomic site (intracranial, spinal, ocular, pericardial, articular, retroperitoneal, or intramuscular with compartment syndrome); decrease in hemoglobin concentration ≥2 g/dL; and transfusion of ≥2 units of whole blood or packed red blood cells.³³ Non-major hemorrhagic events were also included as outcome measures, defined as overt bleeding not meeting criteria for major bleeding but requiring medial intervention, unscheduled contact (visit or telephone) with a physician, temporary interruption of the study drug (i.e., delayed dosing), pain, or impairment of daily activities.^1,34 Ischemic events were defined as recurrence of ischemic stroke or TIA, systemic embolism, acute coronary syndrome, aortic dissection, aortic aneurysm rupture, peripheral artery disease requiring hospitalization, venous thromboembolism, and revascularization, such as carotid endarterectomy, carotid artery stenting, and percutaneous coronary intervention.³⁵ Events were ascertained at outpatient clinics (or by phone survey for patients with too severe aftereffects to visit the clinic) for at least 1 year in principle.

Statistical Analysis

The clinical characteristics of the study participants were summarized as medians [interquartile range (IQR)], mean±standard deviation, or numbers (%) by Xa inhibitor (rivaroxaban and apixaban). Intergroup differences between patients with and without hemorrhagic/ischemic events in each Xa inhibitor group were compared using the Mann-Whitney U test, unpaired Student’s t-test, or the chi-squared test. Univariate factors with P<0.05 were entered into multiple-adjusted Cox models to identify independent predictors for each event according to the timing of AXA measurement (i.e., trough or peak), and hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated. AXA_trough and AXA_peak were categorized into quartiles for both rivaroxaban and apixaban and the rates of patients with hemorrhagic events were compared by Cochran-Armitage test. Cumulative hemorrhagic rates were estimated by the Kaplan-Meier method, and differences between groups were evaluated using the log-rank method. Differences were considered significant at the 5% level (2-sided P<0.05). All data were analyzed with JMPversion 12.0.1 (SAS Institute, Cary, NC, USA).

Results

Analysis of Overall Cohort

A total of 325 patients (age, 78±9 years; 126 women) participated in this study. During the follow-up period from starting Xa inhibitors [median 360, IQR 103–667 days], 24 patients had a hemorrhagic events. They had higher National Institutes of Health Stroke Scale (NIHSS) on admission, took anti-Xa inhibitors in the form of fine granules and with concurrent antiplatelet agents (Table S1) compared with the other 301 patients. AXA_peak was significantly higher in patients with hemorrhagic events than in those without [3.46 (IQR 1.54–4.07) vs. 1.92 (IQR 1.17–2.74) IU/mL, P<0.01, Figure S1]. On multivariable-adjusted Cox models, AXA_peak was significantly related to hemorrhagic events (HR 1.94 per 1.0 IU/mL; 95% CI, 1.31–2.87; Table S2). When patients were divided into quartiles based on AXA, 13 of 24 (54.2%) patients with hemorrhagic events were in the upper 4th quartile of the AXA_peak (Table S3). Although 21 patients experienced ischemic events, there was no relationship between AXA_peak and ischemic events (Figure S1).

Analysis of Patients Taking Rivaroxaban

Of a total of 325 patients enrolled in this study, 156 (age, 77±10 years; 55 women) started rivaroxaban at a median of 5 (IQR 3–10) days from stroke onset (15 mg daily in 83, 10 mg daily in 73). The period from starting rivaroxaban to collecting blood sample was a median of 5 (IQR 4–6) days. The baseline characteristics of the patients are shown in Table 1. In the 15-mg daily group, the AXA_trough was 0.13 (IQR 0.05–0.33) IU/mL and AXA_peak was 1.53 (IQR 0.99–2.50) IU/mL, while in the 10-mg daily group the AXA_trough was 0.05 (IQR 0.01–0.28) IU/mL and AXA_peak was 1.24 (IQR 0.68–1.71) IU/mL. The median follow-up period was 366 (IQR 87–812) days from starting rivaroxaban.

Table 1. Baseline Characteristics of Acute Stroke Patients

	Rivaroxaban				Apixaban
	Overall (n=156)	Without hemorrhagic events (n=143)	With hemorrhagic events (n=13)	P value	Overall (n=169)	Without hemorrhagic events (n=158)	With hemorrhagic events (n=11)	P value
Women	55 (35)	50 (35)	5 (38)	0.80	71 (42)	66 (42)	5 (45)	0.81
Age, years	77±10	77±10	82±9	0.07	79±9	79±9	80±7	0.68
Weight, kg	57.3±11.1	57.4±11.2	56.2±10.5	0.72	55.0±12.3	55.0±12.3	56.0±12.4	0.79
Prestroke CHADS2 score	2 [1–3]	2 [1–3]	2 [2–4]	0.36	2 [2–3]	2 [2–3]	2 [2–3]	0.63
Prestroke CHA2DS2-VASc score	3 [1–4]	3 [1–4]	3 [2–4]	0.29	3 [2–4]	3 [2–4]	4 [2–5]	0.26
Prestroke HAS-BLED score	2 [1–3]	2 [1–2]	3 [1–3]	0.14	2 [2–3]	2 [2–3]	3 [2–3]	0.98
Prestroke medication with DOACs	15 (10)	13 (9)	2 (15)	0.49	22 (13)	22 (14)	0	0.07
Prestroke medication with warfarin	37 (24)	34 (24)	3 (23)	0.95	36 (21)	33 (21)	3 (27)	0.63
NIHSS score on admission	10 [3–19]	9 [3–18]	18 [10–21]	0.07	6 [3–14]	6 [3–14]	10 [1–21]	0.17
Creatinine clearance, mL/min	58.1±19.8	58.2±19.9	57.0±19.0	0.84	53.0±24.1	53.4±24.6	47.1±15.4	0.40
Paroxysmal AF	85 (54)	78 (55)	7 (54)	0.96	115 (68)	108 (68)	7 (64)	0.75
Regular (higher) dosage of Xa inhibitor	83 (53)	76 (53)	7 (54)	0.96	90 (53)	83 (53)	7 (64)	0.47
Crushed tablet	31 (20)	30 (21)	1 (8)	0.21	0	0	0
Fine granules as dosage form	27 (17)	22 (15)	5 (38)	0.06	0	0	0
Concomitant use with antiplatelet agents	19 (12)	17 (12)	2 (15)	0.72	33 (20)	26 (16)	7 (64)	<0.01
Days from onset to starting Xa inhibitor	5 [3–10]	5 [3–10]	6 [3–12]	0.79	5 [2–8]	5 [2–8]	5 [3–7]	0.42

n (%), mean±standard deviation, or median [interquartile range]. AF, atrial fibrillation; DOAC, direct oral anticoagulant; NIHSS, National Institutes of Health Stroke Scale.

Hemorrhagic events occurred in 13 patients (8/100 person-years, Table 2). The median period from starting rivaroxaban to a hemorrhagic event was 11 (IQR 7–172) days; 8 of 13 events (61.5%) occurred within the initial 14 days. AXA_peak was significantly higher in patients with hemorrhagic events than in those without [1.93 (IQR 1.11–3.75) vs. 1.35 (IQR 0.80–2.00) IU/mL, P<0.01, Figure 1]. In the multivariable-adjusted Cox models, AXA_peak was significantly related to hemorrhagic events (HR 2.07 per 1.0 IU/mL; 95% CI, 1.18–3.65; Table 3). When patients with rivaroxaban were divided into quartiles based on the AXA_peak, 4 of the 8 patients having a hemorrhagic event within the initial 14 days were in the upper 4th quartile (AXA_peak >2.055 IU/mL; Table 4). The cumulative rate of hemorrhagic events was not significantly higher in patients with AXA_peak >2.055 IU/mL than in those with AXA_peak ≤2.055 IU/mL (P=0.20; Figure S2). However, AXA_peak >2.055 IU/mL tended to be related to hemorrhagic events within the initial 14 days on multivariable-adjusted Cox models adjusted by age, sex, and factors with P<0.05 in Table 1 (HR 3.54; 95% CI, 0.81–15.50, P=0.099). AXA_trough showed no significant difference between patients with and without hemorrhagic events on either univariate [0.15 (IQR 0.02–0.53) vs. 0.09 (IQR 0.03–0.26) IU/mL, Figure 1] or multivariate analyses.

Table 2. Hemorrhagic/Ischemic Events in Acute Stroke Patients

	Rivaroxaban (n=156)	Apixaban (n=169)
Hemorrhagic events	13 (8)	11 (8)
Major hemorrhagic event	1 (1)	2 (1)
Intracranial hemorrhage	1 (1)	2 (1)
Non-major hemorrhagic event	12 (7)	9 (6)
Gastrointestinal hemorrhage	6 (4)	6 (4)
Subcutaneous hemorrhage	2 (1)	1 (1)
Musculoskeletal hemorrhage	2 (1)	0
Urinary tract hemorrhage	2 (1)	2 (1)
Time from starting drug to event, days	11 [7–172]	260 [89–711]
Time from stroke onset to event, days	19 [11–184]	266 [88–716]
Ischemic events	14 (8)	7 (5)
Ischemic stroke	13 (8)	7 (5)
Systemic embolism	1 (1)	0
Time from starting drug to event, days	263 [74–699]	249 [162–661]
Time from stroke onset to event, days	269 [79–702]	261 [166–664]

n (per 100 person-years) or median [interquartile range].

Figure 1.

Relationships between anti-Xa activity (AXA) of rivaroxaban and hemorrhagic/ischemic events. Relationship between AXA_trough and (A) hemorrhagic events and (B) ischemic events. Relationship between AXA_peak and (C) hemorrhagic events and (D) ischemic events. The box plots indicate the 25th, 50th (median), and 75th percentiles. Tops and bottoms of the bars indicate the 10th and 90th percentiles. White dots indicate the AXA of patients who experienced events within 14 days of taking rivaroxaban. Median (interquartile range).

Table 3. Multivariable Analysis Related to Hemorrhagic Events in the Rivaroxaban Group

	HR (95% CI)	P value
Trough
Anti-Xa activity (/1.0 IU/mL)	3.96 (0.69–22.82)	0.15
Age (/1 year)	1.07 (0.99–1.15)	0.06
Women	1.43 (0.39–5.17)	0.58
Peak
Anti-Xa activity (/1.0 IU/mL)	2.07 (1.18–3.65)	0.01
Age (/1 year)	1.08 (1.00–1.17)	0.04
Women	1.70 (0.46–6.36)	0.42

Adjusted by age, sex, P<0.05 factors in Table 1, and anti-Xa activity (trough/peak). CI, confidence interval; HR, hazard ratio.

Table 4. Quartile Comparison of Hemorrhagic Events and No Hemorrhagic Events in Rivaroxaban Group According to AXA_trough or AXA_peak Quartiles

	With hemorrhagic events (total)	P value	With hemorrhagic events (within 14 days)	P value
Trough
Q1 (n=47)	4 (8.5)	0.20	2 (4.3)	0.17
Q2 (n=31)	0		0
Q3 (n=39)	3 (7.7)		2 (5.1)
Q4 (n=39)	6 (15.4)		4 (10.3)
Peak
Q1 (n=39)	3 (7.7)	0.24	1 (2.6)	0.05
Q2 (n=39)	1 (2.6)		0
Q3 (n=39)	4 (10.3)		3 (7.7)
Q4 (n=39)	5 (12.8)		4 (10.3)

Data are n (%). The total patient population was grouped by trough or peak, and the anti-Xa activity values were divided into quartiles (Trough: Q1 ≤0.03 IU/mL, 0.03<Q2≤0.095 IU/mL, 0.095<Q3≤0.2975 IU/mL, Q4 >0.2975 IU/mL. Peak: Q1 ≤0.805 IU/mL, 0.805<Q2≤1.425 IU/mL, 1.425<Q3≤2.055 IU/mL, Q4 >2.055 IU/mL).

Ischemic events occurred in 14 patients (8/100 person-years, Table 2). Patients with ischemic stroke had higher prestroke CHADS₂ and HAS-BLED scores (Table S4). No significant relationship was seen between AXA_trough or AXA_peak and ischemic events (Figure 1).

Analysis of Patients Taking Apixaban

A total of 169 patients (age, 79±9 years; 71 women) started apixaban at a median of 5 (IQR 2–8) days from stroke onset (5 mg twice daily in 90, 2.5 mg twice daily in 79, Table 1). The period from starting apixaban to collecting blood sample was a median of 5 (IQR 3–6) days. In 5-mg twice daily group the AXA_trough was 1.80 (IQR 1.13–2.33) IU/mL and AXA_peak was 2.93 (IQR 2.18–3.70) IU/mL, while in the 2.5-mg twice daily group the AXA_trough was 1.16 (IQR 0.78–1.99) IU/mL and AXA_peak was 1.92 (IQR 1.45–2.75) IU/mL. The median follow-up was 354 (IQR 115–443) days from starting apixaban.

Hemorrhagic events occurred in 11 patients (8/100 person-years). The median period from starting apixaban to a hemorrhagic event was 260 (IQR 89–711) days; 2 of 11 events (18%) occurred within the initial 14 days. Patients with hemorrhagic events more commonly took antiplatelet agents than those without (P<0.01 Table 1). Both AXA_trough [2.78 (IQR 1.90–3.53) IU/mL and 1.42 (IQR 0.93–2.08) IU/mL, P<0.01] and AXA_peak [4.05 (IQR 3.44–4.72) IU/mL and 2.43 (IQR 1.79–3.35) IU/mL, P<0.01, Figure 2] were significantly higher in patients with than in those without hemorrhagic events. On multivariable-adjusted Cox models, both AXA_trough and AXA_peak were significantly higher in patients with than in those without hemorrhagic events (AXA_trough: HR 2.22 per 1.0 IU/mL; 95% CI, 1.16–4.24, AXA_peak: HR 2.47 per 1.0 IU/mL; 95% CI, 1.28–4.75; Table 5). When patients with apixaban were divided into quartiles based on AXA, 7 of 11 (63.6%) patients with hemorrhagic events were in the upper 4th quartile for the trough (AXA_trough >2.095 IU/mL) and 8 patients (72.7%) were in the upper 4th quartile for the peak (AXA_peak >3.45 IU/mL; Table 6). Moreover, AXA_peak >3.45 IU/mL was significantly related to hemorrhagic events (HR 6.85; 95% CI, 1.51–30.95, P<0.01) and AXA_trough >2.095 IU/mL tended to be related to the events on multivariable-adjusted Cox models adjusted by age, sex, and P<0.05 factors in Table 1 (HR 3.94; 95% CI, 0.97–15.96, P=0.05). The cumulative rate of hemorrhagic events was significantly higher in patients with AXA_trough >2.095 IU/mL (P=0.047) or AXA_peak >3.45 IU/mL (P<0.01; Figure S3).

Figure 2.

Relationships between anti-Xa activity (AXA) of apixaban and hemorrhagic/ischemic events. Relationship between AXA_trough and (A) hemorrhagic events and (B) ischemic events. Relationship between AXA_peak and (C) hemorrhagic events and (D) ischemic events. The box plots indicate the 25th, 50th (median), and 75th percentiles. Tops and bottoms of the bars indicate the 10th and 90th percentiles. White dots indicate the AXA of patients who experienced events within 14 days of taking apixaban. Median (interquartile range).

Table 5. Multivariable Analysis Related to Hemorrhagic Events in the Apixaban Group

	HR (95% CI)	P value
Trough
Anti-Xa activity (/1.0 IU/mL)	2.22 (1.16–4.24)	0.01
Age (/1 year)	1.01 (0.93–1.10)	0.74
Women	1.36 (0.32–5.73)	0.67
Concomitant use with antiplatelet agents	4.95 (1.16–21.16)	0.03
Peak
Anti-Xa activity (/1.0 IU/mL)	2.47 (1.28–4.75)	<0.01
Age (/1 year)	1.04 (0.95–1.14)	0.37
Women	1.08 (0.26–4.60)	0.91
Concomitant use with antiplatelet agents	4.72 (1.13–19.63)	0.03

Adjusted by age, sex, P<0.05 factors in Table 1, and anti-Xa activity (trough/peak). CI, confidence interval; HR, hazard ratio.

Table 6. Quartile Comparison of Hemorrhagic Events and No Hemorrhagic Events in Apixaban Group According to AXA_trough or AXA_peak Quartiles

	With hemorrhagic events (total)	P value	With hemorrhagic events (within 14 days)	P value
Trough
Q1 (n=43)	1 (2.3)	<0.01	0	0.20
Q2 (n=42)	1 (2.4)		0
Q3 (n=42)	2 (4.8)		1 (2.4)
Q4 (n=42)	7 (16.7)		1 (2.4)
Peak
Q1 (n=43)	1 (2.3)	<0.01	0	0.20
Q2 (n=42)	0		0
Q3 (n=42)	2 (4.8)		1 (2.4)
Q4 (n=42)	8 (19.0)		1 (2.4)

Data are n (%). The total patient population was grouped by trough or peak, and the anti-Xa activity values were divided into quartiles (Trough: Q1 ≤0.94 IU/mL, 0.94<Q2≤1.47 IU/mL, 1.47<Q3≤2.095 IU/mL, Q4 >2.095 IU/mL. Peak: Q1 ≤1.81 IU/mL, 1.81<Q2≤2.48 IU/mL, 2.48<Q3≤3.45 IU/mL, Q4 >3.45 IU/mL).

Ischemic events occurred in 7 patients (5/100 person-years). There was no significant difference in the background characteristics of patients with and without ischemic events (Table S5). No significant relationship was seen between AXA_trough or AXA_peak and ischemic events after starting apixaban (Figure 2).

Discussion

In this study, rivaroxaban or apixaban was started during the acute stage of ischemic stroke or TIA, at a median of 5 days after onset. The major finding was that the AXA_peak of rivaroxaban and both the AXA_trough and AXA_peak of apixaban measured early after starting medication were associated with future hemorrhagic events. There were no relationships between any level of AXA and ischemic events in the rivaroxaban or the apixaban group. Unique points of the present study were that all patients took Xa inhibitors for secondary stroke prevention, and that the initiation of anticoagulation was quite soon after stroke onset. Thus, early stroke conditions might affect AXA levels and event occurrence.

As compared with previous studies,^13,19 the AXAs of rivaroxaban were lower in the present patients (Table S6). Various factors may contribute to AXA level, including age, weight, renal function, measurement method, ethnicity, drug-drug interactions, and food effects.³⁶ For example, decreased food intake early after stroke onset in the present severe and elderly patients might have prolonged the time to maximum concentration as compared with young, healthy volunteer men in a fasting state.^29,37 Patients with a nasogastric tube who were administered a crushed rivaroxaban tablet were included, which could also have affected AXA levels.^29,38,39 In this study, half of patients with a hemorrhagic event within the initial 14 days had an AXA_peak >2.055 IU/mL. The value was similar to the 2.19 IU/mL reported as a cutoff value of AXA_peak to predict hemorrhagic events in a previous report.¹⁹ The cutoff of AXA_peak could be a useful indicator of short-term hemorrhagic events. The AXA_trough for rivaroxaban was not predictive of hemorrhagic events, presumably because the level of AXA was very low 24 h after dosing.

The AXAs of apixaban in the present patients were higher than in previous ones (Table S6).^13,20 Some patients with hemorrhagic events had remarkably high AXA_trough and AXA_{peak values}. In particular, 8 of 11 patients with hemorrhagic events had an AXA_peak >3.45 IU/mL, which converted to a concentration 390ng/mL measured by STA-Liquid Anti-Xa reagents (Diagnostica Stago, Asnieres, France). These values could be useful for predicting hemorrhagic events because a recent study involving 85 Korean patients who started taking apixaban immediately after admission with acute ischemic stroke reported 2 patients developing intracranial hemorrhage with concentrations of 450 ng/mL and 491 ng/mL, respectively.²¹ In addition, the AXA_trough in the present patients also seemed to be a meaningful predictor of hemorrhagic events, because the half-life time of apixaban (12 h) is similar to the timing of the trough and somewhat reflects anticoagulant intensity. However, the AXA_peak could be more useful for predicting hemorrhagic events than the AXA_trough when comparing the results of multivariate analysis using quartile values.

The present incidence of hemorrhagic events was equivalent to known trials and post-marketing surveys.^2,3,40–43 Time from starting Xa inhibitor to hemorrhagic event tended to be shorter in the rivaroxaban group. In both groups, AXA levels seemed to identify patients developing hemorrhagic events within the initial 14 days. Early initiation of Xa inhibitors was recently reported to be generally safe,^34,44–46 but safety has not been ascertained by randomized, comparative studies with aspirin or other antithrombotics. However, measuring AXA seems to be useful for identifying patients with early hemorrhagic risk.

Compared with other studies, the incidence of ischemic events in the present patients was high,^2,3,40–43 partly because of their higher CHADS₂ scores. The relationship between ischemic events and AXA measured in the early stage was evaluated without regard to medication adherence after hospital discharge. Poor adherence after discharge can decrease AXA and may cause ischemic events, not hemorrhagic events. This might be a reason for the lack of relationships between AXA and ischemic events in both the rivaroxaban and apixaban groups.

Study Limitations

First, this study was an observational investigation and the choice of rivaroxaban and apixaban was not randomized. There may be potential bias for selecting either of them. The AXA_trough and AXA_peak of rivaroxaban seemed lower than for apixaban. However, this does not mean the anticoagulation effect of rivaroxaban is inferior to apixaban, because baseline characteristics and pharmacokinetics were different between the 2 groups. Second, the blood sampling time was set at 4 h after medication, the plausible peak time of the agents after food intake. However, there might be wide variation in the peak time among patients according to renal function, age, feeding conditions, and other factors. Furthermore, differences in the number of days from starting Xa inhibitors to the measurement AXA may also influence the results because the time to reach steady state differed between rivaroxaban and apixaban.^29,30 Third, we measured AXA only once during the acute stroke phase, and the level of AXA might be affected by stroke condition.^47,48

Conclusions

A high AXA_peak was related to future hemorrhagic events in patients with stroke/TIA who started secondary prevention using rivaroxaban or apixaban relatively early after event onset. A high AXA_trough of apixaban was also associated with hemorrhagic events. Determination of the timing and dose of Xa inhibitors according to the level of AXA may be useful for decreasing hemorrhagic risk.

Authorship

S.W. contributed to the concept and rationale for the study. S.W., T. Matsuki, T.O., M. Kumamoto, N.T., and A.O. contributed to data collection. S.W. and S.S. contributed to data analysis and interpretation of the results and were responsible for the first draft of the manuscript. K.T. supervised the study and provided financial support. All authors participated in the drafting and approval of the final manuscript and take responsibility for the content and interpretation of this article.

Funding

Cardiovascular Diseases of the National Cerebral and Cardiovascular Center, and Grants from the Japan Agency for Medical Research and Development (AMED: 17ek0210091 h0001, 17ek0210055 h0001).

Acknowledgments

None.

Declaration of Conflict Interests

K.T. received honoraria (modest, <1,000,000 JPY/y) from Bayer Yakuhin and Bristol-Myers Squibb. M. Ihara received honoraria (modest) from Pfizer and Bristol-Myers Squibb.

M. Koga received honoraria (modest) from Pfizer, Bristol-Myers Squibb and Bayer Yakuhin. S.S. reports employment by Bayer HealthCare Pharmaceuticals’ Japanese subsidiary, Bayer Yakuhin, from January 2018, but all of his contribution to this study was made during his employment at the National Cerebral and Cardiovascular Center. This study was not funded by Bayer HealthCare Pharmaceuticals or Bayer Yakuhin.

Supplementary Files

Supplementary File 1

Figure S1. Relationships between anti-Xa activity (AXA) of overall patient group and hemorrhagic/ischemic events.

Figure S2. Kaplan-Meier curves for hemorrhagic events in the rivaroxaban group.

Figure S3. Kaplan-Meier curves for hemorrhagic events in the apixaban group.

Table S1. Baseline characteristics of overall cohort of acute stroke patients with or without hemorrhagic events

Table S2. Multivariable analysis related to hemorrhagic events in overall cohort of acute stroke patients

Table S3. Quartile comparison of hemorrhagic events and no hemorrhagic events in overall cohort of acute stroke patients according to AXA_peak quartiles

Table S4. Baseline characteristics of patients taking rivaroxaban with and without ischemic events

Table S5. Baseline characteristics of patients taking apixaban with and without ischemic events

Table S6. Comparison of baseline characteristics and AXA of present study patients with past studies

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-18-0506

References

• 1.   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.
• 2.   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.
• 3.   Granger CB, Alexander JH, McMurray JJ, Lopes RD, Hylek EM, Hanna M, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med 2011; 365: 981–992.
• 4.   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.
• 5.   Mueck W, Eriksson BI, Bauer KA, Borris L, Dahl OE, Fisher WD, et al. Population pharmacokinetics and pharmacodynamics of rivaroxaban – an oral, direct factor Xa inhibitor – in patients undergoing major orthopaedic surgery. Clin Pharmacokinet 2008; 47: 203–216.
• 6.   Samama MM, Contant G, Spiro TE, Perzborn E, Guinet C, Gourmelin Y, et al. Evaluation of the anti-factor Xa chromogenic assay for the measurement of rivaroxaban plasma concentrations using calibrators and controls. Thromb Haemost 2012; 107: 379–387.
• 7.   Douxfils J, Mullier F, Loosen C, Chatelain C, Chatelain B, Dogne JM. Assessment of the impact of rivaroxaban on coagulation assays: Laboratory recommendations for the monitoring of rivaroxaban and review of the literature. Thromb Res 2012; 130: 956–966.
• 8.   Hankey GJ. Unanswered questions and research priorities to optimise stroke prevention in atrial fibrillation with the new oral anticoagulants. Thromb Haemost 2014; 111: 808–816.
• 9.   Cuker A, Siegal DM, Crowther MA, Garcia DA. Laboratory measurement of the anticoagulant activity of the non-vitamin K oral anticoagulants. J Am Coll Cardiol 2014; 64: 1128–1139.
• 10.   Cuker A, Siegal D. Monitoring and reversal of direct oral anticoagulants. Hematology Am Soc Hematol Educ Program 2015; 2015: 117–124.
• 11.   Samama MM, Amiral J, Guinet C, Le Flem L, Seghatchian J. Monitoring plasma levels of factor Xa inhibitors: How, why and when? Expert Rev Hematol 2013; 6: 155–164.
• 12.   Barton J, Wong A, Graudins A. Anti-Xa activity in apixaban overdose: Case report. Clin Toxicol 2016; 54: 871–873.
• 13.   Ikeda K, Tachibana H. Clinical implication of monitoring rivaroxaban and apixaban by using anti-factor Xa assay in patients with non-valvular atrial fibrillation. J Arrhythm 2016; 32: 42–50.
• 14.   Seiffge DJ, Kagi G, Michel P, Fischer U, Bejot Y, Wegener S, et al. Rivaroxaban plasma levels in acute ischemic stroke and intracerebral hemorrhage. Ann Neurol 2018; 83: 451–459.
• 15.   Kanemoto M, Kuhara H, Ueda T, Shinohara T, Oda T, Nakao F, et al. Association of apixaban therapy and prothrombin time in patients with atrial fibrillation. Circ J 2014; 78: 2651–2656.
• 16.   Barrett YC, Wang Z, Frost C, Shenker A. Clinical laboratory measurement of direct factor Xa inhibitors: Anti-Xa assay is preferable to prothrombin time assay. Thromb Haemost 2010; 104: 1263–1271.
• 17.   Gouin-Thibault I, Flaujac C, Delavenne X, Quenet S, Horellou MH, Laporte S, et al. Assessment of apixaban plasma levels by laboratory tests: Suitability of three anti-Xa assays: A multicentre French GEHT study. Thromb Haemost 2014; 111: 240–248.
• 18.   Ruff CT, Giugliano RP, Braunwald E, Morrow DA, Murphy SA, Kuder JF, et al. Association between edoxaban dose, concentration, anti-Factor Xa activity, and outcomes: An analysis of data from the randomised, double-blind ENGAGE AF-TIMI 48 trial. Lancet 2015; 385: 2288–2295.
• 19.   Sakaguchi T, Osanai H, Murase Y, Ishii H, Nakashima Y, Asano H, et al. Monitoring of anti-Xa activity and factors related to bleeding events: A study in Japanese patients with nonvalvular atrial fibrillation receiving rivaroxaban. J Cardiol 2017; 70: 244–249.
• 20.   Osanai H, Ajioka M, Masutomi T, Kuwayama T, Ishihama S, Sakamato Y, et al. Measurement of anti-factor Xa activity in patients on apixaban for non-valvular atrial fibrillation. Circ J 2015; 79: 2584–2590.
• 21.   Shin H, Cho MC, Kim RB, Kim CH, Choi NC, Kim SK, et al. Laboratory measurement of apixaban using anti-factor Xa assays in acute ischemic stroke patients with non-valvular atrial fibrillation. J Thromb Thrombolysis 2018; 45: 250–256.
• 22.   Sato S, Uehara T, Ohara T, Suzuki R, Toyoda K, Minematsu K, et al. Factors associated with unfavorable outcome in minor ischemic stroke. Neurology 2014; 83: 174–181.
• 23.   Sakamoto Y, Koga M, Kimura K, Nagatsuka K, Okuda S, Kario K, et al. Intravenous thrombolysis for patients with reverse magnetic resonance angiography and diffusion-weighted imaging mismatch: SAMURAI and NCVC rt-PA Registries. Eur J Neurol 2014; 21: 419–426.
• 24.   Toyoda K, Okada Y, Kobayashi S. Early recurrence of ischemic stroke in Japanese patients: The Japan standard stroke registry study. Cerebrovasc Dis 2007; 24: 289–295.
• 25.   Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: Results from the National Registry of Atrial Fibrillation. JAMA 2001; 285: 2864–2870.
• 26.   Olesen JB, Lip GY, Hansen ML, Hansen PR, Tolstrup JS, Lindhardsen J, et al. Validation of risk stratification schemes for predicting stroke and thromboembolism in patients with atrial fibrillation: Nationwide cohort study. BMJ 2011; 342: d124.
• 27.   Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJ, Lip GY. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: The Euro Heart Survey. Chest 2010; 138: 1093–1100.
• 28.   Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976; 16: 31–41.
• 29.   Xarelto [package insert]. Bayer Schering Pharma AG, 2013.
• 30.   Eliquis [package insert]. Pfizer AG, 2013.
• 31.   Yamahira N, Frost C, Fukase H, Yu Z, Wang J, Pursley J, et al. Safety, tolerability, pharmacokinetics, and pharmacodynamics of multiple doses of apixaban in healthy Japanese male subjects. Int J Clin Pharmacol Ther 2014; 52: 564–573.
• 32.   Samama MM, Martinoli JL, LeFlem L, Guinet C, Plu-Bureau G, Depasse F, et al. Assessment of laboratory assays to measure rivaroxaban: An oral, direct factor Xa inhibitor. Thromb Haemost 2010; 103: 815–825.
• 33.   Schulman S, Kearon C. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost 2005; 3: 692–694.
• 34.   Goodman SG, Wojdyla DM, Piccini JP, White HD, Paolini JF, Nessel CC, et al. Factors associated with major bleeding events: Insights from the ROCKET AF trial (rivaroxaban once-daily oral direct factor Xa inhibition compared with vitamin K antagonism for prevention of stroke and embolism trial in atrial fibrillation). J Am Coll Cardiol 2014; 63: 891–900.
• 35.   Arihiro S, Todo K, Koga M, Furui E, Kinoshita N, Kimura K, et al. Three month risk-benefit profile of anticoagulation after stroke with atrial fibrillation: The SAMURAI-Nonvalvular Atrial Fibrillation (NVAF) study. Int J Stroke 2016; 11: 565–574.
• 36.   Gong IY, Kim RB. Importance of pharmacokinetic profile and variability as determinants of dose and response to dabigatran, rivaroxaban, and apixaban. Can J Cardiol 2013; 29: S24–S33.
• 37.   Kaneko M, Tanigawa T, Hashizume K, Kajikawa M, Tajiri M, Mueck W. Confirmation of model-based dose selection for a Japanese phase III Study or rivaroxaban in non-valvular atrial fibrillation patients. Drug Metab Pharmacokinet 2013; 28: 321–331.
• 38.   Mueck W, Stampfuss J, Kubitza D, Becka M. Clinical pharmacokinetic and pharmacodynamic profile of rivaroxaban. Clin Pharmacokinet 2014; 53: 1–16.
• 39.   Okata T, Toyoda K, Okamoto A, Miyata T, Nagatsuka K, Minematsu K. Anticoagulation intensity of rivaroxaban for stroke patients at a special low dosage in Japan. PLoS One 2014; 9: e113641.
• 40.   Mahaffey KW, Wojdyla D, Hankey GJ, White HD, Nessel CC, Piccini JP, et al. Clinical outcomes with rivaroxaban in patients transitioned from vitamin K antagonist therapy: A subgroup analysis of a randomized trial. Ann Intern Med 2013; 158: 861–868.
• 41.   Camm AJ, Amarenco P, Haas S, Hess S, Kirchhof P, Kuhls S, et al. XANTUS: A real-world, prospective, observational study of patients treated with rivaroxaban for stroke prevention in atrial fibrillation. Eur Heart J 2016; 37: 1145–1153.
• 42.   Senoo K, Lau YC, Dzeshka M, Lane D, Okumura K, Lip GY. Efficacy and safety of non-vitamin K antagonist oral anticoagulants vs. warfarin in Japanese patients with atrial fibrillation: Meta-analysis. Circ J 2015; 79: 339–345.
• 43.   Hori M, Matsumoto M, Tanahashi N, Momomura S, Uchiyama S, Goto S, et al. Rivaroxaban versus warfarin in Japanese patients with nonvalvular atrial fibrillation in relation to the CHADS2 score: A subgroup analysis of the J-ROCKET AF trial. J Stroke Cerebrovasc Dis 2014; 23: 379–383.
• 44.   Deguchi I, Tanahashi N, Takao M. Timing of treatment initiation with oral anticoagulant for acute ischemic stroke in patients with nonvalvular atrial fibrillation. Circ J 2017; 81: 180–184.
• 45.   Gioia LC, Kate M, Sivakumar L, Hussain D, Kalashyan H, Buck B, et al. Early rivaroxaban use after cardioembolic stroke may not result in hemorrhagic transformation: A prospective magnetic resonance imaging study. Stroke 2016; 47: 1917–1919.
• 46.   Seiffge DJ, Traenka C, Polymeris A, Hert L, Peters N, Lyrer P, et al. Early start of DOAC after ischemic stroke: Risk of intracranial haemorrhage and recurrent events. Neurology 2016; 87: 1856–1862.
• 47.   Testa S, Tripodi A, Legnani C, Pengo V, Abbate R, Dellanoce C, et al. Plasma levels of direct oral anticoagulants in real life patients with atrial fibrillation: Results observed in four anticoagulation clinics. Thromb Res 2016; 137: 178–183.
• 48.   Kreutz R, Persson PB, Kubitza D, Thelen K, Heitmeier S, Schwers S, et al. Dissociation between the pharmacokinetics and pharmacodynamics of once-daily rivaroxaban and twice-daily apixaban: A randomized crossover study. J Thromb Haemost 2017; 15: 2017–2028.