Clinical Characteristics and Outcomes of Patients With Venous Thromboembolism Receiving Edoxaban in the Real World

Abstract

Background: The effectiveness and safety of edoxaban for venous thromboembolism (VTE) in unselected real-world patients have not been fully evaluated.

Methods and Results: In the Japanese nationwide administrative database, we identified 6,262 VTE patients in whom edoxaban was initiated; these patients were divided into 3 groups based on their index doses: 15 mg/day (n=235), 30 mg/day (n=4,532), and 60 mg/day (n=1,495). We evaluated patient characteristics, recurrent VTEs, and a composite endpoint of intracranial hemorrhage (ICH) and gastrointestinal (GI) bleeding. Patient characteristics among the 15-, 30-, and 60-mg edoxaban groups varied widely regarding several aspects, including age (mean 81.0, 76.2, and 65.0 years, respectively) and body weight (mean 49.5, 51.8, and 70.3 kg, respectively). At 180 days, the cumulative incidence of recurrent VTEs in the 15-, 30-, and 60-mg edoxaban groups was 4.4%, 2.6%, and 1.8%, respectively, whereas that of ICH or GI bleeding was 7.3%, 5.4%, and 3.3%, respectively. Subgroup analyses showed that the cumulative incidence of ICH or GI bleeding in patients in the 15-mg edoxaban group was 3.6% for patients aged ≥80 years, 8.4% for those with a body weight <60 kg, and 31.3% for those with renal dysfunction.

Conclusions: Only a minority of patients with VTEs received a super low dose (15 mg) of edoxaban, and these patients may be at higher risk of bleeding as well as VTE recurrence.

Venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE), is a major health problem worldwide.¹ VTE has a long-term risk of recurrence, which can be prevented by anticoagulation therapy.² Current guidelines recommend prolonged anticoagulation therapy for patients with VTE, except for patients at a low risk of recurrence.^3–5 Recently, direct oral anticoagulants (DOACs) have become widely available for VTE^6–10 and they could be safer than a vitamin K antagonist (VKA) in terms of bleeding risk, thus favoring a longer duration of anticoagulation therapy.¹¹ Although prolonged anticoagulation therapy could be effective for preventing VTE recurrence, clinicians are sometimes reluctant to continue anticoagulation therapy for patients at a high risk of bleeding.

In the Hokusai-VTE trial, both a high (60 mg) and low (30 mg) once-daily dose of edoxaban were non-inferior to VKAs for the prevention of recurrent VTE and were favorable in terms of fewer bleeding events compared with VKAs.¹⁰ Based on results of randomized clinical trials (RCT) and ease of use, edoxaban has prevailed in daily clinical practice in Japan.¹² However, RCTs are conducted under idealized and rigorously controlled conditions, with highly selected patients, which may not be broadly representative of real-world practice. Because the clinical outcomes of edoxaban in the real world may not necessarily coincide with the clinical outcomes in RCTs, the effectiveness and safety of edoxaban in unselected real-world patients could be important. Thus, the aim of the present study was to investigate clinical characteristics and outcomes of VTE patients receiving edoxaban in a large-scale all-comer cohort, using a large nationwide administrative database in Japan.

Methods

Study Design and Data Source

We conducted a new-user cohort study using a nationwide hospital administrative database provided by Medical Data Vision Co., Ltd. (MDV), covering the period April 1, 2008 to July 31, 2020. As of 2021, the database included data on over 36 million patients treated in 449 acute care hospitals adopting the Diagnosis Procedure Combination/Per-Diem Payment System (DPC/PDPS), which is a Japanese case-mix classification system for inpatient reimbursement.¹³

The structure of the database has been reported in detail elsewhere.^14–16 Briefly, the database contains longitudinal information on patient demographics, diagnoses coded using the International Classification of Diseases, 10th Revision (ICD-10), drug prescriptions coded using the Anatomical Therapeutic Chemical (ATC) classification system, procedures coded using the National Health Insurance reimbursement system, and laboratory results. The database can link inpatient and outpatient data, allowing for tracking of patients as long as they are treated in the same hospital.

The present study was conducted in accordance with the principles of the Declaration of Helsinki. The research protocol was approved by the Ethics Committee of Kyoto University Graduate School and Faculty of Medicine (Approval no. R3817). In accordance with the current guidelines for epidemiological studies issued by the Ministry of Health, Labor, and Welfare in Japan, the need for written informed consent from each patient was waived because the data were anonymous.

Study Population

We identified patients hospitalized with VTE using ICD-10 codes (Supplementary Appendix 1). Patients were required to undergo at least 1 diagnostic imaging during hospitalization. Imaging modalities included computed tomography for DVT and PE, ultrasound examination of leg veins for DVT, or ventilation-perfusion lung scanning for PE.^14,15,17 Patients aged 18–99 years in whom edoxaban was initiated between September 26, 2014 (approval date of edoxaban for VTE in Japan) and February 2, 2020 (180 days before the end of the study period) were included in the present study. The cohort entry date was the date of the first prescription of edoxaban, and patients were required to have started the index dose of edoxaban within 30 days after the onset of VTE. In addition, all patients were required to have ≥90 days of continuous enrollment in the database before entry into the study cohort.

Patients were excluded from the study if they had been prescribed any oral anticoagulant (warfarin, dabigatran, apixaban, and rivaroxaban) before study entry; had received parenteral anticoagulation (unfractionated heparin or fondaparinux) in the 90 days before the onset VTE; had another indication for edoxaban (atrial fibrillation, atrial flutter, or prophylaxis of VTE after total knee arthroplasty, total hip arthroplasty, or hip fracture surgery); had endocarditis, heart valve disease, or were pregnant; or had missing data on body weight in the 90 days before study entry (Supplementary Appendix 1).¹⁴ In addition, patients were excluded if they had been prescribed edoxaban by a department of orthopedic surgery at cohort entry (because orthopedic surgeons often use edoxaban for VTE prophylaxis) and if they had an intracranial hemorrhage (ICH) or gastrointestinal (GI) bleeding at study entry. Patients were allowed to have received parenteral anticoagulation between VTE onset and study entry.

Patients included in the present study were divided into 3 mutually exclusive groups based on the index dose of edoxaban: 15, 30, or 60 mg/day. Although 15 mg edoxaban for the prevention of recurrent VTE has not been approved in Japan, some patients with VTE received 15 mg edoxaban off-label.

Patient Characteristics and Definitions

We evaluated baseline patient characteristics, including demographics, comorbidities, medications, and laboratory test results, where available.¹⁴ Active cancer was defined as at least one record of cancer that was treated with either pharmacotherapy, radiotherapy, or surgery. Renal dysfunction was defined as at least one record of renal disease or dialysis. Detailed definitions of patient characteristics are provided in Supplementary Appendix 2.

Clinical Outcomes and Definitions

The primary effectiveness outcome was recurrent VTE, defined as a DVT and/or PE occurring subsequent to the index VTE (Supplementary Appendix 3). Based on previous studies,^14,15,17 recurrent VTE were determined by the combination of the following 2 components: inpatient diagnosis of VTE; and diagnostic imaging during hospitalization. We excluded recurrent VTE during the hospitalization for the index VTE to ensure that there was only one VTE episode per hospitalization. Following a previous study,¹⁸ we did not include outpatient VTE diagnoses because they may not have indicated a recurrence, but rather a follow-up visit.¹⁹

The primary safety outcome was the composite of ICH or GI bleeding, as in previous studies (Supplementary Appendix 3).^18,20 ICH and GI bleeding were determined by a combination of a diagnosis and either a procedure or hospitalization.^14,21 Secondary safety outcomes were separate incidences of ICH and GI bleeding.

Clinical Follow-up

Follow-up for the study outcomes began on the day after study entry and continued until Day 180,²² discontinuation of edoxaban, switching to another dose from the index dose of edoxaban, a treatment switch to any other oral anticoagulant, transferal to another hospital or clinic, or death, whichever came first. Discontinuation of edoxaban was defined as no subsequent prescription for edoxaban within a 14-day grace period after the end of the supply provided in the last prescription. To account for medication stockpiling, when the period covered by 2 prescriptions overlapped, the number of overlapping days was added to the end date of the latter prescription.¹⁴ Separate follow-up times were calculated for each outcome.

Statistical Analysis

Categorical variables are presented as numbers and percentages. Continuous variables are presented as the mean±SD or as a median with interquartile range according to data distribution. To examine the significance of differences in patient characteristics among the 3 groups, we calculated standardized mean differences (SMDs) for 3 pairwise comparisons and averaged them for each patient characteristic. Patient characteristics with an SMD ≥0.2 were considered significantly different.^23,24

We estimated the cumulative incidence function with 95% confidence intervals (CIs) for each outcome accounting for the competing risk of death. No statistical tests were performed because of the large differences in the patient characteristics among the 3 groups. Subgroup analyses were conducted by age (<80, ≥80 years), body weight (<60, ≥60 kg), active cancer (with/without), and renal dysfunction (with/without). Furthermore, sensitivity analyses were performed by changing the grace period from 14 to 7 or 28 days to assess for any potential misclassification of the exposure. All analyses were performed by one author (T.F.) using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA) and R version 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria).

Results

We identified 27,397 patients with VTE in whom edoxaban was initiated between September 26, 2014 and February 2, 2020, and 15,650 patients met the inclusion criteria (Figure 1). After excluding 9,388 patients who met the exclusion criteria, the study population consisted of 6,262 patients with VTE in whom edoxaban was initiated. Among these patients, 235 (3.8%), 4,532 (72%), and 1,495 (24%) received 15, 30, and 60 mg edoxaban, respectively, as the index dose.

Figure 1.

Study flow chart. VTE, venous thromboembolism.

Patient Characteristics

The patient characteristics varied widely regarding several aspects among the 3 groups (Table 1). Specifically, patients in the 15-mg edoxaban group were older than those in the 30- and 60-mg groups (mean age 81.0, 76.2, and 65.0 years, respectively), and the proportion of patients aged ≥80 years was highest in the 15-mg edoxaban group (62.6%, 45.7%, and 12.4%, respectively). Patients in the 15-mg edoxaban group had a lower body weight than those in the 30- and 60-mg groups (mean 49.5, 51.8, and 70.3 kg, respectively), and the proportion of patients with body weight <60 kg was highest in the 15-mg edoxaban group (82.1%, 81.5%, and 11.0%, respectively). Patients in the 15-mg edoxaban group more often presented with DVT as a VTE presentation than those in the 30- and 60-mg groups (78.7%, 65.4%, and 50.8%, respectively). Active cancer trended to be least frequent in patients in the 15-mg edoxaban group than in the 30- and 60-mg groups (21.7%, 25.4%, and 29.7%, respectively), whereas renal dysfunction trended to be most frequent in the 15-mg edoxaban group (5.1%, 2.2%, and 0.5%, respectively).

Table 1.

Patient Characteristics

	Edoxaban (mg)			SMD
	15 (n=235)	30 (n=4,532)	60 (n=1,495)
Demographics
Age (years)	81.0±10.4	76.2±11.8	65.0±13.2	0.89
Age category (tears)
18–64	17 (7.2)	635 (14.0)	633 (42.3)	0.86
65–79	71 (30.2)	1,828 (40.3)	677 (45.3)
80–99	147 (62.6)	2,069 (45.7)	185 (12.4)
Female sex	174 (74.0)	3,132 (69.1)	502 (33.6)	0.59
Body weight (kg)	49.5±11.6	51.8±10.2	70.3±12.0	1.21
<60 kg	193 (82.1)	3,694 (81.5)	164 (11.0)	1.35
≥60 kg	42 (17.9)	838 (18.5)	1,331 (89.0)
BMI (kg/m2)	21.54±3.87	21.85±3.71	26.15±4.40	0.75
BMI category (kg/m2)
<18.5	49 (20.9)	780 (17.2)	15 (1.0)	0.70
18.5–25	140 (59.6)	2,858 (63.1)	637 (42.6)
25–30	35 (14.9)	670 (14.8)	598 (40.0)
≥30	6 (2.6)	106 (2.3)	216 (14.4)
Missing	5 (2.1)	118 (2.6)	29 (1.9)
Calendar time of the index date
September 2014–December 2015	24 (10.2)	589 (13.0)	209 (14.0)	0.13
January 2016–December 2016	28 (11.9)	628 (13.9)	230 (15.4)
January 2017–December 2017	44 (18.7)	830 (18.3)	294 (19.7)
January 2018–December 2018	61 (26.0)	1,130 (24.9)	348 (23.3)
January 2019–February 2020	78 (33.2)	1,355 (29.9)	414 (27.7)
Type of index VTE
DVT only	185 (78.7)	2,962 (65.4)	759 (50.8)	0.40
PE with or without a DVT	50 (21.3)	1,570 (34.6)	736 (49.2)
Acute VTE treatment
Initial parenteral anticoagulation	101 (43.0)	2,389 (52.7)	1,001 (67.0)	0.33
Thrombolysis	3 (1.3)	159 (3.5)	90 (6.0)	0.17
Inferior vena cava filter placement	8 (3.4)	294 (6.5)	126 (8.4)	0.14
Comorbidities
Active cancer	51 (21.7)	1,152 (25.4)	444 (29.7)	0.12
Fracture or trauma	51 (21.7)	611 (13.5)	149 (10.0)	0.22
Moderate- to high-risk surgery for PEs	19 (8.1)	408 (9.0)	180 (12.0)	0.09
Previous VTE	18 (7.7)	306 (6.8)	100 (6.7)	0.03
COPD	17 (7.2)	351 (7.7)	107 (7.2)	0.02
Heart failure	53 (22.6)	829 (18.3)	212 (14.2)	0.15
Stroke or transient ischemic attack	41 (17.4)	782 (17.3)	190 (12.7)	0.09
Varicose veins of lower extremities	4 (1.7)	65 (1.4)	29 (1.9)	0.03
Peptic ulcer disease	51 (21.7)	811 (17.9)	249 (16.7)	0.09
Previous bleeding	39 (16.6)	640 (14.1)	212 (14.2)	0.05
Previous gastrointestinal bleeding	10 (4.3)	220 (4.9)	69 (4.6)	0.02
Hypertension	134 (57.0)	1,951 (43.0)	610 (40.8)	0.22
Hyperlipidemia	52 (22.1)	923 (20.4)	347 (23.2)	0.05
Diabetes	51 (21.7)	998 (22.0)	388 (26.0)	0.07
Renal dysfunction	12 (5.1)	99 (2.2)	8 (0.5)	0.19
Liver dysfunction	36 (15.3)	559 (12.3)	202 (13.5)	0.06
Concomitant medication
Antiplatelet agents	56 (23.8)	796 (17.6)	202 (13.5)	0.18
NSAIDs	88 (37.4)	1,525 (33.6)	574 (38.4)	0.07
H2 receptor antagonists	45 (19.1)	782 (17.3)	234 (15.7)	0.06
Proton pump inhibitors	135 (57.4)	2,426 (53.5)	774 (51.8)	0.08
Statins	51 (21.7)	709 (15.6)	239 (16.0)	0.10
Laboratory testing
eGFR (mL/min/1.73 m2)	63.6±33.4	70.2±25.0	75.2±18.5	0.29
eGFR results available	27 (11.5)	373 (8.2)	121 (8.1)	0.08

Unless indicated otherwise, data are given as the mean±SD or n (%). BMI, body mass index; COPD, chronic obstructive pulmonary disease; DVT, deep vein thrombosis; eGFR, estimated glomerular filtration rate; NSAIDs, non-steroidal anti-inflammatory drugs; PE, pulmonary embolism; SMD, standardized mean difference; VTE, venous thromboembolism.

Follow-up Information

The respective median follow-up times for the effectiveness and safety outcomes were 47 and 44 days for the 15-mg edoxaban group; 49 and 48 days for the 30-mg edoxaban group; and 60 and 60 days for the 60-mg edoxaban group. The reasons for censoring differed somewhat across the 3 groups (Supplementary Table 1). Treatment discontinuation, death, and transfer to another hospital or clinic were the most common reasons for censoring in the 15-mg edoxaban group.

Effectiveness Outcomes

At 180 days, the cumulative incidence of VTE recurrence in the 15-, 30-, and 60-mg edoxaban groups was 4.4% (95% CI 1.3–10.7%), 2.6% (95% CI 1.9–3.4%), and 1.8% (95% CI 1.0–2.8%), respectively (Table 2). The time course of VTE recurrence is shown in Figure 2A. Subgroup analyses stratified by age, body weight, active cancer, and renal dysfunction showed that the cumulative incidence of recurrent VTE among patients in the 15-mg edoxaban group was 7.2% (95% CI 1.7–18.3%) in those aged ≥80 years, 5.6% (95% CI 1.6–13.4%) in those with a body weight <60 kg, 7.0% (95% CI 0.9–22.2%) in those with active cancer, and 20.0% (95% CI 0.4–62.1%) in those with renal dysfunction (Table 3).

Table 2.

Clinical Outcomes According to Patients Receiving 15, 30, or 60 mg Edoxaban

Outcome	Edoxaban (mg)	No. patients	No. events in 7 days	7-day CIF (95% CI)	No. events in 30 days	30-day CIF (95% CI)	No. events in 90 days	90-day CIF (95% CI)	No. events in 180 days	180-day CIF (95% CI)
Recurrent VTE	15	235	1	0.4 (0.0–2.2)	1	0.4 (0.0–2.2)	2	1.5 (0.2–5.2)	4	4.4 (1.3–10.7)
	30	4,532	1	0.0 (0.0–0.1)	16	0.4 (0.2–0.6)	38	1.3 (0.9–1.8)	55	2.6 (1.9–3.4)
	60	1,495	0	0.0 (0.0–0.0)	8	0.7 (0.3–1.3)	15	1.6 (0.9–2.6)	16	1.8 (1.0–2.8)
ICH or GI bleeding	15	235	3	1.3 (0.4–3.5)	9	4.2 (2.0–7.5)	11	6.1 (3.1–10.6)	12	7.3 (3.7–12.4)
	30	4,532	58	1.3 (1.0–1.7)	132	3.1 (2.6–3.7)	157	4.3 (3.6–5.0)	171	5.4 (4.5–6.3)
	60	1,495	13	0.9 (0.5–1.5)	29	2.2 (1.5–3.1)	35	2.9 (2.0–4.0)	37	3.3 (2.3–4.5)
ICH	15	235	0	0.0 (0.0–0.0)	4	2.0 (0.6–4.6)	4	2.0 (0.6–4.6)	4	2.0 (0.6–4.6)
	30	4,532	21	0.5 (0.3–0.7)	47	1.1 (0.8–1.5)	51	1.3 (1.0–1.7)	54	1.5 (1.1–2.0)
	60	1,495	3	0.2 (0.1–0.6)	11	0.8 (0.4–1.5)	11	0.8 (0.4–1.5)	12	1.1 (0.5–1.9)
GI bleeding	15	235	3	1.3 (0.4–3.5)	5	2.2 (0.8–4.8)	7	4.2 (1.7–8.4)	8	5.3 (2.3–10.3)
	30	4,532	38	0.9 (0.6–1.2)	86	2.1 (1.7–2.5)	107	3.0 (2.4–3.6)	119	4.0 (3.2–4.8)
	60	1,495	10	0.7 (0.4–1.3)	19	1.4 (0.9–2.2)	25	2.1 (1.4–3.1)	26	2.3 (1.5–3.4)

CI, confidence interval; CIF, cumulative incidence function; GI, gastrointestinal; ICH, intracranial hemorrhage; VTE, venous thromboembolism.

Figure 2.

Cumulative incidence curves for (A) recurrent venous thromboembolism (VTE) and (B) intracranial hemorrhage (ICH) or gastrointestinal (GI) bleeding over a period of 180 days among patients receiving 15, 30, or 60 mg edoxaban.

Table 3.

Clinical Outcomes in Subgroup Analyses According to Edoxaban Dose

Outcome / Subgroup		Edoxaban (mg)	No. patients	No. events in 180 days	180-day CIF (95% CI)
Recurrent VTE
Age (years)	<80	15	88	1	1.1 (0.1–5.6)
		30	2,463	37	2.9 (2.0–4.0)
		60	1,310	15	1.7 (1.0–2.8)
	≥80	15	147	3	7.2 (1.7–18.3)
		30	2,069	18	1.9 (1.1–3.1)
		60	185	1	2.4 (0.2–11.2)
Body weight (kg)	<60	15	193	4	5.6 (1.6–13.4)
		30	3,694	43	2.4 (1.7–3.3)
		60	164	1	1.2 (0.1–6.0)
	≥60	15	42	0	0.0 (0.0–0.0)
		30	838	12	3.2 (1.6–5.7)
		60	1,331	15	1.8 (1.0–2.9)
Active cancer	With	15	51	2	7.0 (0.9–22.2)
		30	1,152	24	3.5 (2.2–5.2)
		60	444	9	3.0 (1.5–5.5)
	Without	15	184	2	3.1 (0.6–9.6)
		30	3,380	31	2.2 (1.4–3.2)
		60	1,051	7	1.1 (0.5–2.3)
Renal dysfunction	With	15	12	1	20.0 (0.4–62.1)
		30	99	2	2.4 (0.5–7.6)
		60	8	0	0.0 (0.0–0.0)
	Without	15	223	3	3.5 (0.8–9.8)
		30	4,433	53	2.6 (1.9–3.4)
		60	1,487	16	1.8 (1.0–2.8)
ICH or GI bleeding
Age (years)	<80	15	88	7	12.1 (5.0–22.7)
		30	2,463	98	5.3 (4.3–6.6)
		60	1,310	32	3.2 (2.2–4.6)
	≥80	15	147	5	3.6 (1.3–7.8)
		30	2,069	73	5.4 (4.0–7.1)
		60	185	5	3.4 (1.2–7.4)
Body weight (kg)	<60	15	193	11	8.4 (4.2–14.6)
		30	3,694	130	5.2 (4.2–6.2)
		60	164	6	4.4 (1.8–8.9)
	≥60	15	42	1	2.4 (0.2–10.9)
		30	838	41	6.3 (4.4–8.6)
		60	1,331	31	3.1 (2.1–4.4)
Active cancer	With	15	51	2	4.0 (0.7–12.2)
		30	1,152	42	5.1 (3.6–6.9)
		60	444	11	2.9 (1.5–5.1)
	Without	15	184	10	8.4 (4.0–15.1)
		30	3,380	129	5.4 (4.4–6.5)
		60	1,051	26	3.5 (2.3–5.2)
Renal dysfunction	With	15	12	2	31.3 (1.5–72.4)
		30	99	2	3.3 (0.6–10.6)
		60	8	0	0.0 (0.0–0.0)
	Without	15	223	10	6.1 (2.9–11.0)
		30	4,433	169	5.4 (4.5–6.4)
		60	1,487	37	3.3 (2.3–4.6)

Abbreviations as in Table 2.

Safety Outcomes

At 180 days, the cumulative incidence of ICH or GI bleeding in the 15-, 30-, and 60-mg edoxaban groups was 7.3% (95% CI 3.7–12.4%), 5.4% (95% CI 4.5–6.3%), and 3.3% (95% CI 2.3–4.5%), respectively (Table 2). The time course of ICH or GI bleeding is shown in Figure 2B. These results were generally consistent for the individual outcomes of ICH (2.0%, 1.5%, and 1.1%) and GI bleeding (5.3%, 4.0%, and 2.3%) at 180 days in the 15-, 30-, and 60-mg edoxaban groups (Table 2). Subgroup analyses stratified by age, body weight, active cancer, and renal dysfunction showed that the cumulative incidence of ICH or GI bleeding in patients with an index dose of 15 mg edoxaban was 3.6% (95% CI 1.3–7.8%) in those aged ≥80 years, 8.4% (95% CI 4.2–14.6%) in those with a body weight <60 kg, 4.0% (95% CI 0.7–12.2%) in those with active cancer, and 31.3% (95% CI 1.5–72.4%) in those with renal dysfunction (Table 3).

Sensitivity Analyses

The sensitivity analyses that changed the grace period from 14 to 7 or 28 days showed consistent results with the main analysis (Supplementary Table 2).

Discussion

The main findings of the present study are as follows: (1) patients with VTE frequently have characteristics of high bleeding risk, including an old age, low body weight, and renal dysfunction, and most patients with VTE received 30 mg edoxaban as current daily clinical practice; (2) the risks of recurrent VTE and ICH or GI bleeding differed in patients according to different doses of edoxaban; and (3) patients with a low body weight and renal dysfunction appeared to be at a higher risk of bleeding.

The mainstay of the prevention of recurrent VTE is anticoagulation therapy, which could also cause major concern regarding bleeding events, especially in patients at high bleeding risk. Thus, it may be important to obtain a good balance between the risk of recurrence and bleeding through long-term anticoagulation strategies including the duration and dose of anticoagulation therapy. In the Hokusai-VTE trial, edoxaban was given once daily at a dose of 60 mg, which was reduced to 30 mg in patients with a creatinine clearance of 30–50 mL/min, body weight ≤60 kg, or receiving certain P-glycoprotein inhibitors, and the reduced dose of 30 mg was reported to have a similar efficacy and safety as the 60-mg dose in patients meeting the criteria for a dose reduction.²⁵ However, clinicians can sometimes be reluctant to prescribe even the reduced dose of 30 mg edoxaban, especially for patients at high bleeding risk, including those with very old age, severe renal dysfunction, very low body weight, or an overlapping of those risk factors. In fact, the present study showed that patients with VTE frequently had characteristics of high bleeding risk in daily clinical practice. Although a super low dose of edoxaban of 15 mg for the prevention of recurrent VTE has not been approved anywhere in the world, including in Japan, the ELDERCARE-AF trial reported that a 15-mg super low dose of edoxaban could be a potential alternative for the prevention of stroke in very elderly patients with atrial fibrillation who are at high bleeding risk.²⁶ The large nationwide administrative database used in the present study revealed that only a minority of patients with VTE received 15 mg edoxaban as off-label use, and old age, low body weight, and renal dysfunction seemed to be the clinical characteristics for receiving 15 mg edoxaban.

When considering an optimal dosage of edoxaban from a pharmacological point of view, identifying patients at high bleeding risk during anticoagulation therapy could be particularly important. Previous studies reported several risk factors for bleeding in VTE patients, including old age, sex, cancer, uncontrolled hypertension, previous bleeding, renal dysfunction, anemia, thrombocytopenia, a labile international normalized ratio for warfarin users, and the use of platelet inhibitors.^27–29 In addition, several factors related to the pharmacological metabolism of DOACs could be especially important for optimal doses of DOACs, including age, body weight, renal function, and concomitant drugs.^30–33 In line with previous reports, the present study revealed that the risk of ICH or GI bleeding in patients receiving 15 mg edoxaban was numerically greater in patients with low body weight and renal dysfunction. Low body weight and renal dysfunction may be especially notable factors in terms of the bleeding risk related to the use of edoxaban.

No RCT to date has evaluated the efficacy and safety of a 15-mg super low dose of edoxaban for the prevention of recurrent VTE. The present observational study revealed that patients receiving 15 mg edoxaban had a numerically higher incidence of both recurrent VTE and ICH or GI bleeding compared with the other groups, which suggests that patients at potential high bleeding risk may also be at a high risk of recurrent VTE. Considering the quite different baseline characteristics of patients receiving 15 mg edoxaban compared with the other groups, including old age, low body weight, and renal dysfunction, these clinical features could have a considerable influence on bleeding risk and the risk of VTE recurrence. Although the present analysis was primarily exploratory due to the observational study design, making it subject to various biases, the risk of recurrent VTE should be noted when prescribing the super low dose of 15 mg edoxaban. Thus, it may be clinically relevant to identify which patients would most likely benefit from the super low dose of 15 mg edoxaban. Considering the clinical relevance of the issues, future RCTs evaluating the efficacy and safety of super low-dose edoxaban would be warranted.

Study Limitations

The present study has several limitations. First, this study was an observational study, making it subject to various biases inherent to an observational study design. In particular, the therapeutic decision making was left to the discretion of the attending physicians, which could have had a considerable influence on the clinical outcomes. Second, the MDV database collected data only from acute care hospitals, and data were unavailable for patients who had also been treated in other clinics and hospitals. Thus, the present results may be only partially applicable to patients treated in clinics and long-term care hospitals. In addition, this may have led to an outcome misclassification and underestimation of cumulative incidences. However, the majority of VTE patients are seen in acute care hospitals, and the treatment for VTE is likely to continue in the same hospital. Third, the MDV database lacked body weight data for <5% of the total cohort, and we excluded those patients. However, when key confounders are missing in <5% of cases, a complete-case analysis may be acceptable because the results would be nearly identical, regardless of how missing confounders are treated.³⁴ Fourth, we excluded recurrent VTE during the hospitalization for the index VTE to ensure the validity and accuracy of the recurrent VTE, which could underestimate recurrent VTE in the acute phase. Fifth, we defined baseline characteristics based on ICD-10 codes. Renal dysfunction was defined as at least one record of renal disease or dialysis. However, considering the relatively low prevalence of renal dysfunction, there could be some concerns regarding the accuracy of the identification of renal dysfunction, meaning the results should be interpreted with caution. Active cancer was defined as at least one record of cancer that was treated with either pharmacotherapy, radiotherapy, or surgery. However, if a patient refused any treatment for cancer, we would not have included them as patients with active cancer. Sixth, we could not evaluate patients with edoxaban under- or overdosing based on the packet insert in Japan, because there were only a few patients in the MDV database with data on serum creatine. Finally, the demographics, practice patterns, and clinical outcomes for patients with VTE in Japan may differ from those outside Japan.

Conclusions

In the Japanese nationwide administrative database, patients with VTE frequently had characteristics of high bleeding risk, with risks of clinical outcomes differing according to the different doses of edoxaban. Only a minority of patients with VTE received the super low dose of 15 mg edoxaban, and these patients may be at higher risk of bleeding and VTE recurrence, and patients with a low body weight and renal dysfunction appeared to be at a markedly higher risk of bleeding.

Acknowledgments

The authors are grateful to Medical Data Vision Co., Ltd. for the generous provision of the medical data and Mr. John Martin (Exclusive Consultant at Japan Lifeline Co., Ltd.) for his grammatical assistance.

Sources of Funding

This study was supported, in part, by research funding for the investigator-initiated studies by Daiichi Sankyo Co., Ltd. The research funder had no role in the design and conduct of the study; the collection, management, analysis, and interpretation of the data; and in the preparation, review, or approval of the manuscript.

Disclosures

K.O. is a member of Circulation Journal’s Editorial Team.

Conflict of Interest

Y.Y. has received lecture fees from Bayer Healthcare, Bristol-Myers Squibb, Pfizer, and Daiichi Sankyo, and grant support from Bayer Healthcare and Daiichi Sankyo. T.F. has been employed by the Department of Digital Health and Epidemiology, an Industry-Academia Collaboration Course supported by Eisai Co., Ltd., Kyowa Kirin Co., Ltd., and Real World Data Co., Ltd.; and has received consulting fees from Real World Data Co., Ltd. and speaker fees from Asahi Kasei Pharma Corporation and EPS Corporation. K.K. has received research grants from Eisai Co., Ltd., Kyowa Kirin Co., Ltd., Sumitomo Pharma Co., Ltd., and Real World Data Co., Ltd.; consulting fees from LEBER Inc., JMDC Inc., Shin Nippon Biomedical Laboratories Ltd., and Advanced Medical Care Inc.; executive compensation from Cancer Intelligence Care Systems, Inc.; and speaker fees from Pharma Business Academy and Toppan Inc. All other authors have no relationships relevant to the contents of this paper to disclose.

Author Contributions

Y.Y. and T.F. contributed equally to this work. Y.Y., T.F., and C.T. formulated the study concept and design. T.F. performed the statistical analysis and Y.Y. wrote the manuscript. All authors contributed to the discussion, reviewed, edited, and approved the final manuscript, and agreed to submission.

IRB Information

This study was conducted in accordance with the principles of the Declaration of Helsinki. The research protocol was approved by the Ethics Committee of Kyoto University Graduate School and Faculty of Medicine (Approval no. R3817). In accordance with the current guidelines for epidemiological studies issued by the Ministry of Health, Labor, and Welfare in Japan, The need for written informed consent from each patient was waived because the data are anonymous.

Data Availability

Deidentified participant data will not be shared.

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-23-0818

References

• 1.   Di Nisio M, van Es N, Buller HR. Deep vein thrombosis and pulmonary embolism. Lancet 2016; 388: 3060–3073, doi:10.1016/S0140-6736(16)30514-1.
• 2.   Kyrle PA, Rosendaal FR, Eichinger S. Risk assessment for recurrent venous thrombosis. Lancet 2010; 376: 2032–2039, doi:10.1016/S0140-6736(10)60962-2.
• 3.   Stevens SM, Woller SC, Kreuziger LB, Bounameaux H, Doerschug K, Geersing GJ, et al. Antithrombotic therapy for VTE disease: Second update of the CHEST guideline and expert panel report. Chest 2021; 160: e545–e608, doi:10.1016/j.chest.2021.07.055.
• 4.   Konstantinides SV, Meyer G, Becattini C, Bueno H, Geersing GJ, Harjola VP, et al. 2019 ESC guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS). Eur Heart J 2020; 41: 543–603, doi:10.1093/eurheartj/ehz405.
• 5.   Witt DM, Nieuwlaat R, Clark NP, Ansell J, Holbrook A, Skov J, et al. American Society of Hematology 2018 guidelines for management of venous thromboembolism: Optimal management of anticoagulation therapy. Blood Adv 2018; 2: 3257–3291, doi:10.1182/bloodadvances.2018024893.
• 6.   Schulman S, Kearon C, Kakkar AK, Mismetti P, Schellong S, Eriksson H, et al. Dabigatran versus warfarin in the treatment of acute venous thromboembolism. N Engl J Med 2009; 361: 2342–2352, doi:10.1056/NEJMoa0906598.
• 7.   EINSTEIN Investigators; Bauersachs R, Berkowitz SD, Brenner B, Buller HR, Decousus H, Gallus AS, et al. Oral rivaroxaban for symptomatic venous thromboembolism. N Engl J Med 2010; 363: 2499–2510, doi:10.1056/NEJMoa1007903.
• 8.   EINSTEIN-PE Investigators; Büller HR, Prins MH, Lensin AW, Decousus H, Jacobson BF, Minar E, et al. Oral rivaroxaban for the treatment of symptomatic pulmonary embolism. N Engl J Med 2012; 366: 1287–1297, doi:10.1056/NEJMoa1113572.
• 9.   Agnelli G, Buller HR, Cohen A, Curto M, Gallus AS, Johnson M, et al. Oral apixaban for the treatment of acute venous thromboembolism. N Engl J Med 2013; 369: 799–808, doi:10.1056/NEJMoa1302507.
• 10.   Hokusai-VTE Investigators; Büller HR, Décousus H, Grosso MA, Mercuri M, Middeldorp S, Prins MH, et al. Edoxaban versus warfarin for the treatment of symptomatic venous thromboembolism. N Engl J Med 2013; 369: 1406–1415, doi:10.1056/NEJMoa1306638.
• 11.   Gomez-Outes A, Terleira-Fernandez AI, Lecumberri R, Suarez-Gea ML, Vargas-Castrillon E. Direct oral anticoagulants in the treatment of acute venous thromboembolism: A systematic review and meta-analysis. Thromb Res 2014; 134: 774–782, doi:10.1016/j.thromres.2014.06.020.
• 12.   Yamashita Y, Morimoto T, Yoshikawa Y, Yaku H, Sumita Y, Nakai M, et al. Temporal trends in the practice pattern for venous thromboembolism in Japan: Insight from JROAD-DPC. J Am Heart Assoc 2020; 9: e014582, doi:10.1161/JAHA.119.014582.
• 13.   Hayashida K, Murakami G, Matsuda S, Fushimi K. History and profile of Diagnosis Procedure Combination (DPC): Development of a real data collection system for acute inpatient care in Japan. J Epidemiol 2021; 31: 1–11, doi:10.2188/jea.JE20200288.
• 14.   Fukasawa T, Seki T, Nakashima M, Kawakami K. Comparative effectiveness and safety of edoxaban, rivaroxaban, and apixaban in patients with venous thromboembolism: A cohort study. J Thromb Haemost 2022; 20: 2083–2097, doi:10.1111/jth.15799.
• 15.   Shinkawa K, Yoshida S, Seki T, Yanagita M, Kawakami K. Risk factors of venous thromboembolism in patients with nephrotic syndrome: A retrospective cohort study. Nephrol Dial Transplant 2020, doi:10.1093/ndt/gfaa134.
• 16.   Yoshida S, Takeuchi M, Tanaka-Mizuno S, Mizuno K, Nakashima M, Fukasawa T, et al. Clinical epidemiology and pharmacoepidemiology studies with real-world databases. Proc Jpn Acad Ser B Phys Biol Sci 2022; 98: 517–528, doi:10.2183/pjab.98.026.
• 17.   Alotaibi GS, Wu C, Senthilselvan A, McMurtry MS. The validity of ICD codes coupled with imaging procedure codes for identifying acute venous thromboembolism using administrative data. Vasc Med 2015; 20: 364–368, doi:10.1177/1358863X15573839.
• 18.   Dawwas GK, Leonard CE, Lewis JD, Cuker A. Risk for recurrent venous thromboembolism and bleeding with apixaban compared with rivaroxaban: An analysis of real-world data. Ann Intern Med 2022; 175: 20–28, doi:10.7326/M21-0717.
• 19.   White RH, Garcia M, Sadeghi B, Tancredi DJ, Zrelak P, Cuny J, et al. Evaluation of the predictive value of ICD-9-CM coded administrative data for venous thromboembolism in the United States. Thromb Res 2010; 126: 61–67, doi:10.1016/j.thromres.2010.03.009.
• 20.   Fralick M, Colacci M, Schneeweiss S, Huybrechts KF, Lin KJ, Gagne JJ. Effectiveness and safety of apixaban compared with rivaroxaban for patients with atrial fibrillation in routine practice: A cohort study. Ann Intern Med 2020; 172: 463–473, doi:10.7326/M19-2522.
• 21.   Fujihara K, Yamada-Harada M, Matsubayashi Y, Kitazawa M, Yamamoto M, Yaguchi Y, et al. Accuracy of Japanese claims data in identifying diabetes-related complications. Pharmacoepidemiol Drug Saf 2021; 30: 594–601, doi:10.1002/pds.5213.
• 22.   Sindet-Pedersen C, Staerk L, Pallisgaard JL, Gerds TA, Berger JS, Torp-Pedersen C, et al. Safety and effectiveness of rivaroxaban and apixaban in patients with venous thromboembolism: A nationwide study. Eur Heart J Cardiovasc Pharmacother 2018; 4: 220–227, doi:10.1093/ehjcvp/pvy021.
• 23.   Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med 2009; 28: 3083–3107, doi:10.1002/sim.3697.
• 24.   Lau WC, Chan EW, Cheung CL, Sing CW, Man KK, Lip GY, et al. Association between dabigatran vs warfarin and risk of osteoporotic fractures among patients with nonvalvular atrial fibrillation. JAMA 2017; 317: 1151–1158, doi:10.1001/jama.2017.1363.
• 25.   Verhamme P, Wells PS, Segers A, Ageno W, Brekelmans MP, Cohen AT, et al. Dose reduction of edoxaban preserves efficacy and safety for the treatment of venous thromboembolism: An analysis of the randomised, double-blind HOKUSAI VTE trial. Thromb Haemost 2016; 116: 747–753, doi:10.1160/TH16-03-0244.
• 26.   Okumura K, Akao M, Yoshida T, Kawata M, Okazaki O, Akashi S, et al. Low-dose edoxaban in very elderly patients with atrial fibrillation. N Engl J Med 2020; 383: 1735–1745, doi:10.1056/NEJMoa2012883.
• 27.   den Exter PL, Woller SC, Robert-Ebadi H, Masias C, Morange PE, Castelli D, et al. Management of bleeding risk in patients who receive anticoagulant therapy for venous thromboembolism: Communication from the ISTH SSC Subcommittee on Predictive and Diagnostic Variables in Thrombotic Disease. J Thromb Haemost 2022; 20: 1910–1919, doi:10.1111/jth.15776.
• 28.   Ruiz-Gimenez N, Suarez C, Gonzalez R, Nieto JA, Todoli JA, Samperiz AL, et al. Predictive variables for major bleeding events in patients presenting with documented acute venous thromboembolism: Findings from the RIETE Registry. Thromb Haemost 2008; 100: 26–31, doi:10.1160/TH08-03-0193.
• 29.   Klok FA, Hosel V, Clemens A, Yollo WD, Tilke C, Schulman S, et al. Prediction of bleeding events in patients with venous thromboembolism on stable anticoagulation treatment. Eur Respir J 2016; 48: 1369–1376, doi:10.1183/13993003.00280-2016.
• 30.   Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med 2009; 361: 1139–1151, doi:10.1056/NEJMoa0905561.
• 31.   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, doi:10.1056/NEJMoa1009638.
• 32.   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, doi:10.1056/NEJMoa1107039.
• 33.   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, doi:10.1056/NEJMoa1310907.
• 34.   Lee KJ, Tilling KM, Cornish RP, Little RJA, Bell ML, Goetghebeur E, et al. Framework for the treatment and reporting of missing data in observational studies: The Treatment And Reporting of Missing data in Observational Studies framework. J Clin Epidemiol 2021; 134: 79–88, doi:10.1016/j.jclinepi.2021.01.008.