Effectiveness and Safety of Direct Oral Anticoagulants vs. Warfarin and Recurrence After Discontinuation in Patients With Acute Venous Thromboembolism in the Real World

Abstract

Background: The efficacy of direct oral anticoagulants (DOACs) compared with warfarin for the treatment of venous thromboembolism (VTE), and the recurrence of VTE after discontinuation of anticoagulation therapy in research are limited.

Methods and Results: This retrospective study enrolled 893 patients with acute VTE between 2011 and 2019. The cohort was divided into the transient risk, unprovoked, continued cancer treatment, and cancer remission groups. The following were compared between DOACs and warfarin: composite outcome of all-cause death, VTE recurrence, bleeding and composite outcome of VTE-related death, recurrence and bleeding. In the continued cancer treatment group, more bleeding was seen in warfarin-treated patients than in patients treated with DOACs (53.2% vs. 31.2%, [P=0.048]). In addition, composite outcome of VTE-related death and recurrence after discontinuation of anticoagulation therapy (n=369) was evaluated. The continued cancer treatment group (multivariate analysis: HR: 3.62, 95% CI: 1.84–7.12, P<0.005) and bleeding-related discontinuation of therapy (HR: 2.60, 95% CI: 1.32–5.13, P=0.006) were independent predictors of the event after discontinuation of anticoagulation therapy. VTE recurrence after discontinuation of anticoagulation therapy in the cancer remission group was 1.6% and a statistically similar occurrence was found in the transient risk group (12.4%) (P=0.754).

Conclusions: DOACs may decrease bleeding incidence in patients continuing to receive cancer treatment. In patients with bleeding-related discontinuation of anticoagulation therapy, VTE recurrence may increase. Discontinuation of anticoagulant therapy might be a treatment option in patients who have completed their cancer treatment.

Venous thromboembolism (VTE) is a fatal disease,¹ and is on the rise in Japan and Asia.² Anticoagulation therapy is a vital part of treatment for early stage as well as chronic VTE, and warfarin has been the only oral anticoagulant used for this therapy since for a long time. However, warfarin requires regular blood examination for dosage adjustment, and interacts with other drugs and vitamin K. Due to the pharmacokinetic characteristics of warfarin, treatment with heparin is required prior to warfarin administration, and in approximately 40% of the cases, warfarin levels are reported to be below the target treatment range.³ In recent years, direct oral anticoagulants (DOACs) have been introduced as a treatment option. In contrast to warfarin, DOACs have relatively stable pharmacokinetics, so that regular monitoring and dose adjustment are unnecessary,⁴ although there are restrictions due to renal function disturbances. DOACs are also associated with a lower occurrence of cerebral hemorrhage than warfarin.⁵ Furthermore, because rivaroxaban and apixaban do not require prior heparin treatment, a single-drug approach has become possible, which has led to a recent dramatic increase in DOAC use.⁶ For VTE treatment effectiveness (to prevent recurrence), compared to warfarin, DOACs are non-inferior, and apixaban and edoxaban are superior; rivaroxaban is non-inferior to warfarin in terms of safety (bleeding).^7–9 Edoxaban was approved for insurance coverage for VTE treatment in Japan in September 2014, and rivaroxaban and apixaban were approved in 2015. However, 25% of patients in randomized controlled trials using DOACs were reported to meet one or more exclusion criteria.¹⁰ The target international normalized ratio prothrombin time (PT-INR) value in these trials was 2.0 to 3.0, which is higher than that set by the Japanese Circulation Society guidelines.

Editorial p ????

In this study, we aim to compare the safety and outcomes of DOACs and warfarin therapies for VTE treatment, as there have been few reports on real-world data pertaining to this subject. There are limited data on prognosis after discontinuation of anticoagulation treatment, so we also investigated VTE recurrence after completion of anticoagulation treatment.

Methods

Study Population

This study was a physician-initiated, retrospective, single-center cohort study. We reviewed patient data from the Japanese Red Cross Musashino Hospital medical records from January 2011 to December 2019. Patients with a thrombus found on contrast-enhanced CT or ultrasound images of the veins of the lower extremities were diagnosed with VTE. Of the 1,102 VTE cases, 893 consecutive cases were analyzed after applying exclusion criteria. Cases with creatinine clearance of <15 mL/min, dialysis requirement, loss to follow up, out-of-hospital cardiopulmonary arrest due to acute VTE, non-optimal dosage of DOACs, heparin-alone treatment, or VTE recurrence were excluded. The patients were divided into transient risk, unprovoked, and cancer groups, according to the American Heart Association (AHA) and American College of Chest Physicians guideline.^11,12

First, patients with associated active cancers were classified into the cancer group. Second, among the patients without active cancers, those without a transient risk factor for VTE were classified into the unprovoked group. Finally, patients with a transient risk factor for VTE were classified into the transient risk group. The cancer group was divided into the continued cancer treatment group (continuing cancer treatment 1 year after initiation of therapy for VTE) and the cancer remission group (completed cancer treatment) (Figure 1).

Figure 1.

Flowchart of participant inclusion. There were 369 patients (41.3%) with discontinuation of anticoagulant therapy. In the transient group, 43.0% of patients had completed anticoagulation, and in the unprovoked group, continued cancer treatment group, and cancer remission group, this was 29.5%, 33.4%, and 77.0%, respectively. Ccr, creatinine clearance; CPAOA, cardiopulmonary arrest on arrival; DOAC, direct oral anticoagulants; VTE, venous thromboembolism.

Data Collection and Definition of Patient Characteristics

Transient risk factors for VTE included recent surgery (within 2 months prior to VTE), recent immobilization (non-surgical bed-ridden patients with bathroom privileges for >4 days within 2 months prior to VTE), recent leg trauma, fracture or burn (events requiring immobilization in the past 2 months), long-distance travel (travel lasting ≥6 h in the previous 3 weeks), pregnancy or puerperium, central venous catheter use, severe infection, and estrogen use.¹³ Patients with cancer-associated VTE included those receiving treatment for cancer such as chemotherapy or radiotherapy, those scheduled to undergo cancer surgery, those with metastases, and those with terminal cancer (expected life expectancy ≤6 months) at the time of diagnosis, according to a previous study.¹⁴ Cancer remission was defined as the absence of cancer in the body at 1 year after therapy initiation for cancer-associated VTE, and when at least 6 months have passed since the completion of surgery, chemotherapy, radiation therapy, or other treatments. Continued cancer treatment was defined in those receiving treatment for cancer such as radiotherapy or chemotherapy, those scheduled to undergo cancer surgery, those with metastases, and those with terminal cancer (expected life expectancy ≤6 months) at 1 year after the initiation of therapy for cancer-associated VTE. Creatinine clearance was calculated using the Cockcroft-Gault formula. Heart disease was defined as a history of heart failure, ischemic heart disease, non-ischemic cardiomyopathy, or arrhythmia. Brain disease was defined as a history of cerebral hemorrhage, cerebral infarction, brain tumor, degenerative disease, or epilepsy. Thrombophilia was defined as protein C deficiency, protein S deficiency, antithrombin deficiency, or antiphospholipid antibody positivity. The reasons for the discontinuation of anticoagulation therapy were hemorrhagic complications and others (physician’s discretion, drug side-effects other than bleeding, and patient’s self-interruption).

Anticoagulant Therapy and Additional Therapy

Anticoagulant therapy consisted of warfarin or DOACs. When heparin was used prior to oral medication, the heparin dose was adjusted so that the activated partial thromboplastin time (aPTT) was 1.5-fold that of the control value. Warfarin was adjusted to maintain PT-INR at 1.5 to 2.5. Rivaroxaban was administered twice daily at 15 mg for 21 days and then once daily at 15 mg. Apixaban was administered twice daily at 10 mg for 7 days and then twice daily at 5 mg. Edoxaban was administered once daily at 60 mg, or at 30 mg for patients weighing ≤60 kg, with a creatinine clearance ≤50 mL/min, who were taking quinidine sulfate hydrate, verapamil hydrochloride, erythromycin, or cyclosporine. At the doctor’s discretion, some patients began treatment with rivaroxaban 15 mg once daily or apixaban 5 mg twice daily. In addition, some patients were treated with warfarin and edoxaban without heparin. As per Japanese guidelines, the time in therapeutic range (TTR) was calculated using the Rosendaal method, according to a therapeutic INR range of 1.5–2.5. The choice of oral anticoagulants was as per the attending physician’s discretion and Japanese insurance practices.

If the attending physician decided that treatments other than anticoagulant therapy were necessary, an inferior vena cava filter (IVCF) and tissue plasminogen activator (tPA) were used. To prevent fatal pulmonary embolism (PE), IVCF was used in the case of a floating thrombus in the lower extremity vein of the proximal region, especially in the iliac vein; it was also used if the perioperative embolic risk was judged as high. An IVCF was used temporarily except when anticoagulation was not possible. Intravenous tPA was administered when the respiratory or hemodynamic parameters deteriorated.

Clinical Follow up and Endpoints

The primary endpoints during oral anticoagulant therapy included the composite incidence of all-cause death, recurrent VTE, and bleeding. VTE recurrence was defined as a new thrombus found on contrast-enhanced computed tomography (CT) or ultrasound images of the veins of the lower extremities. During the follow-up period, contrast-enhanced CT and ultrasound images of veins of the lower extremities were performed when thrombosis recurrence was suspected, such as shortness of breath, edema of the lower leg, and/or an increase in D-dimer levels. Contrast-enhanced CT was also performed to evaluate the status of cancer. Bleeding was evaluated based on the International Society on Thrombosis and Hemostasis criteria as clinically apparent acute hemorrhage, including at least one of the following: (1) fatal hemorrhage; (2) symptomatic hemorrhage at important sites or organs (intracranial, intrathecal, intraocular, retroperitoneal, intra-articular, pericardial, and intramuscular hemorrhage accompanied by compartment syndrome); or (3) decrease in hemoglobin by ≥2 g/dL and packed red blood cell transfusion of ≥2 units.¹⁵ Additionally, the composite endpoint of VTE-related death, VTE recurrence, and bleeding were examined. VTE-related death was defined as the cause of death if there was objective documentation of VTE, or if PE could not be categorically ruled out.⁷ We compared the outcomes of DOACs and warfarin for these endpoints.

We also examined the risk factors for the composite endpoint of VTE-related death and VTE recurrence after discontinuation of anticoagulation therapy, and evaluated them with or without bleeding as the reason for discontinuation. Definition of VTE recurrence and indications for imaging studies are the same as above.

The cases were followed up for a maximum of 3 years.

Statistical Analysis

Data normality was verified using the Kolmogorov-Smirnov test. Categorical data were expressed as absolute frequencies and percentages, and were compared using the chi-squared test. Continuous variables were expressed as mean±standard deviation for normally distributed variables, and as median values (25–75^th percentiles) for non-normally distributed variables; they were compared using the Student’s t-test. The Kaplan-Meier method was used to compare the composite risk of all-cause death, recurrent VTE, bleeding with DOACs or warfarin during anticoagulation therapy, and differences in the event-free curve were evaluated using a log-rank test. After discontinuation of anticoagulation therapy, the composite endpoint of VTE-related death and VTE recurrence, as well as whether the reason for discontinuation of anticoagulation therapy was bleeding, was calculated by using a log-rank test using the Kaplan-Meier method. Bonferroni correction was performed for post-hoc analysis. A Cox proportional hazard model analysis was performed to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for VTE recurrence after anticoagulation; the model was adjusted for sex, whether bleeding was the reason for discontinuation, continued cancer treatment group, PE or not. Statistical analyses were performed using the R statistical package version 3.1.0 (The R Foundation for Statistical Computing, Vienna, Austria; http://www.r-project.org/). Statistical significance was defined as a 2-sided P value <0.05. Using a Bonferroni test, statistical significance was defined as P value <0.0083.

Results

Patient Characteristics

Table 1 shows the baseline characteristics of the patients. The DOAC usage rate was 86.8% after September 2014 when edoxaban became available, and 90.6% after December 2015 when rivaroxaban and apixaban also became available. For this reason, the use of warfarin was found mostly in cases prior to September 2014. The median age was 71 years for the DOACs group and 70 years for the warfarin group; 37.8% and 37.9% of patients in the DOACs and warfarin groups, respectively, were men. TTR for PT-INR was 61.1% in the continued cancer treatment group, whereas it was >70% in the other groups. In addition, the usage rate of IVCF was lower in the DOACs group than in the warfarin group. The DOAC-treated patients of the transient risk group had a higher age and hemoglobin level than the warfarin-treated patients. The DOAC-treated patients of the unprovoked group had a lower D-dimer than the warfarin-treated patients. No difference was observed in the continued cancer treatment groups. The DOAC-treated patients of the cancer remission group had lower PE and higher deep vein thrombosis levels than the warfarin-treated patients.

Table 1. Baseline Characteristics of Study Participants

	Total (n=893)		P value	Transient risk (n=316)		P value	Unprovoked (n=234)		P value	Continued cancer treatment group (n=230)		P value	Cancer remission group (n=113)		P value
	DOAC (n=603)	Warfarin (n=290)		DOAC (n=193)	Warfarin (n=123)		DOAC (n=145)	Warfarin (n=89)		DOAC (n=171)	Warfarin (n=59)		DOAC (n=94)	Warfarin (n=19)
Age, years [IQR]	71 [58, 79]	70 [55, 79]	0.212	73 [58, 81]	66 [45, 80]	0.015	72 [58, 80]	73 [59, 80]	0.971	69 [62, 77]	70 [58, 75]	0.525	66 [54, 75]	68 [64, 76]	0.433
Male (%)	228 (37.8)	110 (37.9)	1	121 (62.7)	85 (69.1)	0.276	83 (57.2)	45 (50.6)	0.345	97 (56.7)	36 (61.0)	0.647	74 (78.7)	14 (73.7)	0.762
BW, kg [IQR]	56.0 [48.0, 66.0]	56.6 [48.1, 64.5]	0.817	56 [49.0, 66.0]	54.9 [48.9, 66.0]	0.745	60.0 [50.0, 68.0]	59.40 [53.0, 64.1]	0.557	54.0 [46.0, 65.0]	55.0 [45.0, 63.5]	0.605	54.0 [47.2, 65.0]	58.0 [47.9, 61.0]	0.945
Comorbidities
Hypertension (%)	364 (60.4)	183 (63.1)	0.463	68 (35.2)	34 (27.6)	0.176	60 (41.4)	41 (46.1)	0.499	70 (40.9)	23 (39.0)	0.878	41 (43.6)	9 (47.4)	0.804
DM (%)	79 (13.1)	27 (9.3)	0.122	23 (11.9)	7 (5.7)	0.077	18 (12.4)	9 (10.1)	0.677	27 (15.8)	9 (15.3)	1	11 (11.7)	2 (10.5)	1
Dyslipidemia (%)	112 (18.6)	49 (16.9)	0.578	31 (16.1)	22 (17.9)	0.758	34 (23.4)	16 (18.0)	0.412	26 (15.2)	5 (8.5)	0.269	21 (22.3)	6 (31.6)	0.389
Heart disease (%)	60 (10.0)	37 (12.8)	0.208	30 (15.5)	16 (13.0)	0.624	12 (8.3)	12 (13.5)	0.267	13 (7.6)	6 (10.2)	0.585	5 (5.3)	3 (15.8)	0.130
Brain disease (%)	36 (6.0)	12 (4.1)	0.341	18 (9.3)	8 (6.5)	0.410	7 (4.8)	1 (1.1)	0.160	7 (4.1)	2 (3.4)	1	4 (4.3)	1 (5.3)	1
Thrombophilia (%)	29 (4.8)	10 (3.4)	0.388	5 (2.6)	0 (0.0)	0.161	18 (12.4)	9 (10.1)	0.677	5 (2.9)	1 (1.7)	1	1 (1.1)	0 (0)	1
Thrombophilia patients tested (%)	415 (68.8)	224 (77.2)		135 (70.0)	93 (75.6)		145 (100)	89 (100)		80 (46.8)	30 (50.1)		55 (58.5)	12 (63.2)
Smoking (%)	114 (18.9)	42 (14.6)	0.131	114 (18.9)	42 (14.6)	0.131	17 (11.7)	14 (15.9)	0.427	52 (30.4)	11 (18.6)	0.092	25 (26.6)	5 (26.3)	1
Antiplatelet therapy (%)	22 (3.7)	11 (3.8)	1	22 (3.7)	11 (3.8)	1	4 (2.8)	3 (3.4)	1	6 (3.5)	4 (6.8)	0.284	2 (2.1)	0 (0)	1
PE (%)	273 (45.3)	150 (51.7)	0.074	273 (45.3)	150 (51.7)	0.491	86 (59.3)	52 (58.4)	0.892	74 (43.3)	26 (44.1)	1	25 (26.6)	11 (57.9)	0.013
DVT without PE (%)	330 (54.7)	140 (48.3)		105 (54.4)	62 (50.4)		59 (40.7)	37 (41.6)		97 (56.7)	33 (55.9)		69 (73.4)	8 (42.1)
Symptomatic (%)	495 (82.1)	255 (87.9)	0.025	174 (90.2)	108 (7.8)	0.578	143 (98.6)	88 (98.9)	1	115 (67.3)	44 (74.6)	0.33	63 (67.0)	15 (78.9)	0.418
Laboratory data
WBC count, /μL [IQR]	6,800 [5,200, 8,750]	6,900 [5,425, 9,700]	0.079	7,000 [5,200, 9,100]	7,400 [5,700, 10,800]	0.079	6,600 [5,300, 8,700]	6,800 [5,300, 9,800]	0.468	7,000 [5,350, 8,700]	6,400 [5,250, 8,050]	0.308	6,150 [4,800, 8,000]	6,700 [5,650, 9,050]	0.248
Hemoglobin, g/dL [IQR]	12.5 [11.0, 13.7]	12.0 [10.4, 13.2]	0.500	12.6 [10.9, 13.9]	11.7 [10.2, 13.0]	0.002	13.5 [12.3, 14.7]	13.1 [12.0, 14.6]	0.791	11.8 [10.7, 13.0]	11.4 [10.3, 12.8]	0.200	12.1 [10.8, 13.0]	11.8 [10.7, 12.3]	0.106
Platelet, ×104/μL [IQR]	21.7 [17.2, 27.8]	21.1 [17.0, 27.3]	0.536	22.5 [17.7, 27.9]	21.4 [17.0, 29.2]	0.461	20.2 [16.8, 24.8]	20.6 [17.0, 24.0]	0.859	22.3 [17.8, 28.3]	23.1 [18.6, 32.7]	0.382	24.2 [17.0, 23.0]	21.5 [20.0, 30.2]	0.812
D dimer, μg/mL [IQR]	6.0 [2.1, 8.6]	6.1 [2.1, 11.7]	0.078	6.0 [2.1, 8.7]	3.8 [1.9, 10.6]	0.883	5.0 [1.4, 8.0]	3.8 [2.3, 10.6]	0.023	7.7 [3.7, 9.0]	7.5 [2.3, 17.3]	0.203	5.0 [1.6, 8.0]	3.8 [2.1, 10.6]	0.298
Creatinine clearance, mL/min [IQR]	73.6 [57.9, 90.0]	71.0 [53.1, 95.8]	0.419	77.0 [57.0, 95.3]	75.4 [49.1, 105.0]	0.745	68.5 [57.0, 86.0]	66.2 [54.0, 83.0]	0.304	74.0 [58.5, 87.0]	72.0 [63.0, 92.5]	0.470	75.0 [60.3, 88.5]	74.0 [38.3, 87.0]	0.343
Anticoagulation
Edoxaban (%)	267 (44.3)			87 (27.5)			41 (28.3)			90 (52.6)			49 (52.1)
Apixaban (%)	219 (36.3)			33 (17.1)			53 (36.6)			19 (11.1)			12 (12.8)
Rivaroxaban (%)	117 (19.4)			73 (37.8)			51 (35.2)			62 (36.3)			33 (35.1)
TTR for PT-INR 1.5–2.5 (%)		76.5 [56.2, 92.2]			72.2 [54.5, 89.2]			75.6 [57.5, 93.1]			61.1 [40.2, 83.5]			72.5 [50.5, 91.2]
IVCF (%)	11 (1.8)	13 (4.5)	0.027	2 (1.0)	6 (4.9)	0.060	5 (3.4)	3 (3.4)	1	3 (1.8)	3 (5.1)	0.177	1 (1.1)	1 (5.3)	0.309
tPA (%)	13 (2.2)	13 (4.5)	0.058	6 (3.1)	10 (8.1)	0.064	7 (4.8)	2 (2.2)	0.489	0 (0)	1 (1.7)	0.257	0	0	1
Duration of anticoagulation, days [IQR]	334 [122, 668]	405 [180, 1,248]		387 [126, 744]	388 [180, 994]		456 [200, 899]	1,099 [368, 1,963]		180 [90, 400]	154 [60, 467]		213 [126, 448]	1,198 [376, 2,140]

The data are expressed as numbers (%), means and standard deviations, or medians and interquartile ranges. BW, body weight; DM, diabetes mellitus; DOAC, direct oral anticoagulant; DVT, deep vein thrombosis; IQR, interquartile range; IVCF, inferior vena cava filter; PE, pulmonary embolism; PT-INR, prothrombin time-international normalized ratio; tPA, tissue plasminogen activator; TTR, time in therapeutic range; WBC, white blood cell.

Clinical Outcomes of Anticoagulant Therapy

The primary endpoint incidence in the transient risk group was 23.8% at 3 years. The incidence at 3 years was 9.1% in the unprovoked group, 89.3% in the continued cancer treatment group, and 4.7% in the cancer remission group. The continued cancer treatment group had more events than the other 3 groups (P<0.005). The transient risk group had more events than the unprovoked and cancer remission group (P<0.005). There was no difference between the unprovoked and the cancer remission group (P=0.214) (Figure 2A).

Figure 2.

(A) Kaplan-Meier curve for primary endpoint (composite incidence of all-cause death, recurrent VTE, and bleeding) (a), composite incidence of VTE-related death and VTE recurrence (b), and bleeding (c) during anticoagulation therapy by group. VTE, venous thromboembolism; DOAC, direct oral anticoagulants; (B) Kaplan-Meier curve for primary endpoint (composite incidence of all-cause death, recurrent VTE, and bleeding), composite incidence of VTE-related death and VTE recurrence, and bleeding during anticoagulation therapy. DOACs vs. Warfarin by the transient group, the unprovoked group, continued cancer treatment group, and cancer remission group. VTE, venous thromboembolism; DOAC, direct oral anticoagulants.

The incidence of the composite endpoint of VTE-related death and VTE recurrence in the transient risk, unprovoked, continued cancer treatment, and cancer remission groups was 2.5%, 1.1%, 4.8%, and 0%, respectively. There was no difference between the 4 group. Bleeding incidence in the transient risk, unprovoked, continued cancer treatment, and cancer remission groups was 10.4%, 3.2%, 40.4%, and 0.9%, respectively. The continued cancer treatment group had more bleeding events than the other 3 groups (P<0.005). There was no difference between the other 3 groups. The follow-up rate was 72.3% at 1 year.

Clinical Outcomes of DOACs vs. Warfarin Use

In the transient risk group, the primary endpoint incidence was 24.3% at 3 years in the DOACs group, and 23.8% in the warfarin group (P=0.772). The incidence of the composite endpoint of VTE-related death and VTE recurrence was 1.9% in the DOACs group, and 3.5% in the warfarin group (P=0.342). Bleeding incidence was 8.7% in the DOACs group, and 13.2% in the warfarin group (P=0.367). In the unprovoked group, the primary endpoint was 6.3% in the DOACs group, 13.0% in the warfarin group (P=0.143). The incidence of the composite endpoint of VTE-related death and VTE recurrence was 2.3% in the DOACs group, and 0% in the warfarin group (P=0.301). Bleeding incidence was 2.7% in the DOACs group, and 3.9% in the warfarin group (P=0.608). In the continued cancer treatment group, the event rate of the primary endpoint was 90.6% in the DOACs group, and 85.8% in the warfarin group (P=0.825). The incidence of the composite endpoint of VTE-related death and VTE recurrence was 6.7% in the DOACs group, and 0% in the warfarin group (P=0.178). Bleeding incidence was 31.2% in the DOACs group, and 53.2% in the warfarin group (P=0.048). In the cancer remission group, the event rate of the primary endpoint was 8.7% in the DOACs group, and 0% in the warfarin group (P=0.238). The incidence of the composite endpoint of VTE-related death and VTE recurrence was 0% in the DOACs group and 0% in the warfarin group (P=1.000). Bleeding incidence was 1.1% in the DOACs group, and 0% in the warfarin group (P=0.646) (Figure 2B). In addition, the analysis was divided into fatal bleeding, intracranial, and others in all patients and in the continued cancer treatment group with the most bleeding complications (Table 2). Intracranial bleeding incidence was 0.3% in the DOACs group and 2.1% in the warfarin group among all patients (P=0.017) and 1.2% in the DOACs group and 6.8% in the warfarin group among continued cancer treatment group (P=0.039).

Table 2. Bleeding Rates for DOACs and Warfarin by Bleeding Type

	All patients			Continued cancer treatment group
	DOAC (n=603)	Warfarin (n=290)	P value	DOAC (n=171)	Warfarin (n=59)	P value
Fatal bleeding (%)	4 (0.7)	5 (1.7)	0.159	4 (2.3)	5 (8.5)	0.051
Intracranial (%)	2 (0.3)	6 (2.1)	0.017	2 (1.2)	4 (6.8)	0.039
Other (%)	44 (7.3)	26 (9.0)	0.425	30 (17.5)	15 (25.4)	0.189

DOAC, direct oral anticoagulant.

The continued cancer treatment group had a higher rate of edoxaban use. Therefore, further analysis was performed separately for edoxaban and other DOACs. The patient background characteristics are shown in the Supplementary Table. There was no difference in patient background among the 4 groups. There were no significant differences for primary endpoint (P=1), VTE-related death and recurrence (P=1), and also bleeding event (P=0.265) among patients treated with edoxaban or other DOACs.

Clinical Outcomes After Discontinuation of Anticoagulant Therapy

There were 369 patients (41.3%) with discontinuation of anticoagulant therapy and we examined the incidence of a VTE-related event in the subset of the 369 patients. Composite of VTE-related death and VTE recurrence after discontinuation of anticoagulation therapy in the transient risk group was 12.4% at 3 years. The rates in the unprovoked, continued cancer treatment, and cancer remission groups were 28.9%, 39.0%, and 1.6% at 3 years, respectively (Figure 3). There was no significant difference of incidence of events between the transient group and the cancer remission group (P=0.754), but the cancer remission group had less composite of VTE-related death and VTE recurrence than the unprovoked group and continued cancer treatment group (P<0.005). The follow-up rate at 1 year was 85.5%. Many of the reasons for termination of anticoagulation therapy in the unprovoked group was due to a physician’s judgment that the risk of bleeding was judged to outweigh the risk of VTE recurrence.

Figure 3.

Kaplan-Meier curve for the composite incidence of VTE-related death and VTE recurrence after discontinuation of anticoagulation therapy by the transient group, the unprovoked group, continued cancer treatment group, and cancer remission group.

Table 3 shows the background information in patients with and without recurrence. Both groups received anticoagulant therapy for 180 days (median). More recurrence was observed among patients who had bleeding-related discontinuation of therapy (35.1% vs. 17.2%, P=0.014). There was no difference by the type of anticoagulant drugs used (P=0.850). Table 4 indicates the independent factors that correlated with the composite endpoint of VTE-related death and recurrence after adjustments in the Cox regression analysis. In multivariate analysis, the event rate was significantly higher in the continued cancer treatment group (HR: 3.62, 95% CI: 1.84–7.12, P<0.005) and in the group who had bleeding-related discontinuation of therapy (HR: 2.60, 95% CI: 1.32–5.13, P=0.006). Furthermore, we performed Kaplan-Meier analysis between patients who stopped anticoagulant therapy due to bleeding or other reasons in order to examine the incidence of VTE-related events in the transient risk group, unprovoked group, continued cancer treatment group, and cancer remission group. In the transient risk group, the event rate of the composite endpoint of VTE-related death and VTE recurrence was 9.0% at 3 years in the bleeding group, and 12.5% in the other groups (P=0.565) (Figure 4A). In the unprovoked group, it was 20.0% at 3 years in the bleeding group, and 34.8% in the other groups (P=0.393) (Figure 4B). In the continued cancer treatment group, it was 100% at 3 years in the bleeding group, and 11.8% in other groups (P<0.005) (Figure 4C). In the cancer remission group, it was 0% at 3 years in the bleeding group, and 2.0% in other groups (P=0.614) (Figure 4D).

Table 3. Baseline Characteristics With or Without VTE Recurrence After Discontinuation of Anticoagulation Therapy

	Recurrence (n=37)	No recurrence (n=332)	P value
Age, years [IQR]	72 [58, 76]	68 [51, 77]	0.278
Male (%)	17 (45.9)	110 (33.1)	0.144
Body weight, kg [IQR]	61.0 [51.0, 67.0]	56.0 [49.0, 65.0]	0.236
Comorbidities
Hypertension (%)	11 (29.7)	109 (32.8)	0.854
Diabetes mellitus (%)	1 (2.7)	36 (10.8)	0.152
Dyslipidemia (%)	7 (18.9)	59 (17.8)	0.823
Heart disease (%)	2 (5.4)	18 (5.4)	1.000
Brain disease (%)	2 (5.4)	11 (3.3)	0.628
Thrombophilia (%)	1 (2.7)	5 (1.5)	0.472
Smoking (%)	10 (27.0)	61 (18.4)	0.270
Antiplatelet therapy (%)	2 (5.4)	9 (2.7)	0.304
PE (%)	18 (48.6)	113 (34.0)	0.102
Anticoagulation
DOAC	26 (70.3)	237 (71.4)	0.850
Warfarin	11 (29.7)	95 (28.6)
Duration of anticoagulation, days [IQR]	180 [91, 334]	180 [90, 349]	0.933
Reason for discontinuation
Bleeding	13 (35.1)	57 (17.2)	0.014
Others	24 (64.9)	275 (82.8)
Group
Transient risk group	10 (27.0)	126 (38.0)	<0.005
Unprovoked group	12 (32.4)	57 (17.2)
Continued cancer treatment group	14 (37.8)	63 (19.0)
Cancer remission group	1 (2.7)	86 (25.9)

DOAC, direct oral anticoagulant; IQR, interquartile range; PE, pulmonary embolism.

Table 4. Factors Correlating With VTE-Related Death and VTE Recurrence

Variables	Univariate		Multivariate
	HR (95% CI)	P value	HR (95% CI)	P value
Male	1.78 (1.05–2.98)	0.029	0.58 (0.30–1.11)	0.098
Continued cancer treatment group	3.28 (1.68–6.38)	<0.005	3.62 (1.84–7.12)	<0.005
Bleeding was the reason for discontinuation	2.23 (1.16–4.30)	0.017	2.60 (1.32–5.13)	0.006
PE	1.52 (0.80–2.90)	0.204	1.73 (0.90–3.35)	0.100

CI, confidence interval; HR, hazard ratio; PE, pulmonary embolism; VTE, venous thromboembolism.

Figure 4.

Kaplan-Meier curve for the composite incidence of VTE-related death and VTE recurrence after discontinuation of anticoagulation therapy. Bleeding-related discontinuation of anticoagulant therapy by the transient group, the unprovoked group, continued cancer treatment group, and cancer remission group. VTE, venous thromboembolism.

Discussion

Clinical Outcomes of Anticoagulation Therapy

The continued cancer treatment group had a high cumulative incidence of the primary endpoint and bleeding. Cancer patients have been reported to have a high VTE incidence rate,¹⁶ and VTE complications affect vital prognosis.¹⁷ Although anticoagulation therapy is the basic treatment, VTE recurrence and hemorrhagic complications are common in patients with cancer.¹⁸ Bladder, stomach, and pancreatic cancers have been reported to be associated with a higher incidence of bleeding,¹⁹ and the factors correlated with bleeding were reported to include stage 4, low performance status, diabetes mellitus, liver dysfunction, and low hemoglobin level.²⁰ In this study, the number of VTE-related deaths and VTE recurrence during oral anticoagulation therapy was not high, whereas there were many cases of non-VTE-related deaths (primarily cancer-related) and bleeding. The treatment results differed greatly between the group that continued to receive cancer treatment 1 year after the start of VTE treatment and the group that had completed cancer treatment.

It has been reported that there was no difference between the transient risk and unprovoked groups;¹⁴ however, in this study, the transient risk group had more events than the unprovoked group. The transient risk group includes diseases that may cause a relatively long-term decline in activities of daily living, such as bone fractures and aspiration pneumonia; results are expected to vary depending on the frequency of these diseases.

Additionally, IVCF use tended to be higher in the warfarin group. However, the warfarin group had a higher number of patients from an earlier treatment era, and it has been reported that the IVCF usage rate decreased after 2010.²¹ In this study, IVCF use is thought to have a trend similar to that. Thus, IVCF use was higher in the warfarin group because warfarin was more used in older periods.

Clinical Outcomes of DOACs vs. Warfarin

The transient risk, unprovoked, and cancer remission groups showed no difference in the primary endpoint, composite endpoint of VTE-related death and recurrence, and bleeding. The warfarin and DOACs groups were comparable in terms of all-cause mortality; therefore, it was reported that other factors such as patient preference, cost, risk of recurrent VTE, and risk of bleeding, need to be used for comparative asssessment.²² Although apixaban and edoxaban were reported to result in less bleeding,^8,9 their target PT-INR is 2–3, which is different from the target in Japanese individuals. In the Apixaban for the treatment of Japanese subjects with acute venous thromboembolism (AMPLIFY-J Study) in Japanese subjects with a target PT-INR of 1.5–2.5, the DOACs group had fewer cases of major bleeding and clinically important non-major bleeding,²³ but the study had a small sample size. Some reports suggested that there was no difference in bleeding in clinical practice,^24,25 whereas others reported that the DOACs group had reduced bleeding.^26,27 The target PT-INR in the present study was 1.5–2.5, which may be one of the reasons for the lack of difference in bleeding incidence between the groups. DOACs have been reported to improve quality of life²⁸ and shorten the length of hospital stay;²⁹ this therapy also does not require frequent dose adjustments like warfarin. In contrast, it was reported that warfarin is preferred in patients with a creatinine clearance of <30, a history of poor medication adherence, in breastfeeding mothers, or in the antiphospholipid antibody syndrome.³⁰

In this study, however, the DOACs group of the continued cancer treatment group had fewer cases of bleeding. One possible reason that the DOACs group had fewer cases of bleeding is that PT-INR control by warfarin were unpredictable in the continued cancer treatment group. This may be due to weakness associated with cancer progression, decreased food intake, or due to the interaction of intermittent chemotherapy and warfarin. In addition, as previously reported,⁵ intracranial bleeding was lower with DOAC use in all patients and in the continued cancer treatment group.

Clinical Outcomes After Discontinuation of Anticoagulant Therapy

The continued cancer treatment group had a high number of recurrence cases, whereas the cancer remission group had few cases. The most important factor was whether the cause of VTE was eliminated. The incidence of VTE is lower in patients whose cancer is in remission, and the course of cancer treatment has a significant impact on the incidence.³¹ The rate of VTE recurrence after discontinuation of anticoagulation therapy in the cancer remission group was similar to that in the transient risk group, for which the Japanese Circulation Society (JCS) guidelines recommend discontinuing anticoagulation therapy at 3 months. The guidelines of the JCS, European Society of Cardiology,³² and American Society of Clinical Oncology³³ recommend discontinuing anticoagulant therapy after cancer is in remission, but there are few evidences on this.

Bleeding as a reason of discontinuation was a risk factor for recurrence, and it was found mostly in the continued cancer treatment group. Anticoagulation therapy was performed for 180 days (median), and anticoagulation therapy was administered for a defined time period as per guidelines in other groups. Therefore, there was no difference in the recurrence rate after completion of anticoagulation therapy due to bleeding. The continued cancer treatment group with persistent thrombogenic tendency will have higher recurrence of VTE after discontinuation of anticoagulation therapy. The usefulness of DOACs for cancer-associated VTE has been reported in recent years.^34–36 The interaction between anticancer drugs and warfarin was unfavorable for cancer patients, and furthermore, bleeding was followed by increased mortality.²⁰ Therefore, considering the results of the present study, which showed that the DOACs group had few cases of hemorrhagic complications, DOACs should be used for VTE in patients undergoing cancer treatment.

Study Limitations

First, this was a retrospective study conducted at a single institution with a relatively small sample size. Therefore, selection bias was inevitable. A multicenter randomized study with a larger sample size would be required to increase the generalizability of results. Second, the study did not have a high follow-up rate. Third, the evaluation of VTE as per its severity was not performed. In addition, the number of cases was too low for examination by type of cancer. The frailty of patients also could not be evaluated, and congenital coagulation factors were not sufficiently evaluated.

Conclusions

DOACs may decrease bleeding incidence in patients continuing to receive cancer treatment. In the group whose cancer was in remission, VTE recurrence and bleeding due to anticoagulation were less common than in the group whose cancer was still being treated. Discontinuation of anticoagulation therapy following bleeding may lead to recurrence of VTE. DOACs should be used for VTE in patients undergoing cancer treatment. Those who complete cancer treatment may show low VTE incidence even on discontinuation of anticoagulant therapy. Discontinuation of anticoagulant therapy might be a treatment option in patients who have completed cancer treatment.

Funding Sources

There was no financial support associated with this study.

Disclosures

The authors declare no conflicts of interest.

IRB Information

This study conformed to the ethical principles of the Declaration of Helsinki. The requirement for informed consent was waived because all data were cataloged anonymously. The institutional review board of the Japanese Red Cross Musashino Hospital approved the study protocol (protocol number: 2058; approval date: October 1, 2020). The disclosure document is available on the hospital’s website. Patients were notified about their participation in the study, and were informed that they were free to opt out of the study at any time.

Data Availability

Deidentified participant data will not be shared.

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-21-0588

References

• 1.   Kasper W, Konstantinides S, Geibel A, Olschewski M, Heinrich F, Grosser FD, et al. Management strategies and determinants of outcome in acute major pulmonary embolism: Results of a multicenter registry. J Am Coll Cardiol 1997; 30: 1165–1171.
• 2.   Gonsalves WI, Pruthi RK, Patnaik MM. The new oral anticoagulants in clinical practice. Mayo Clin Proc 2013; 88: 495–511.
• 3.   Nakamura M, Miyata T, Ozeki Y, Takayama M, Komori K, Yamada N, et al. Current venous thromboembolism management and outcomes in Japan: Nationwide the Japan venous thromboembolism treatment registry observational study. Circ J 2014; 78: 708–717.
• 4.   Cohen A, Chiu KM, Park K, Jeyaindran S, Tambunan KL, Ward C, et al. Managing venous thromboembolism in Asia: Winds of change in the era of new oral anticoagulants. Thromb Res 2012; 130: 291–301.
• 5.   Hori M, Matsumoto M, Tanahashi N, Momomura S, Uchiyama S, Goto S, et al; on behalf of the J-ROCKET AF study investigators. Rivaroxaban vs. warfarin in Japanese patients with atrial fibrillation: The J-ROCKET AF Study. Circ J 2012; 76: 2104–2111.
• 6.   Lutsey PL, Walker RF, MacLehose RF, Alonso A, Adam TJ, Zakai NA. Direct oral anticoagulants and warfarin for venous thromboembolism treatment: Trends from 2012 to 2017. Res Pract Thromb Haemost 2019; 3: 668–673.
• 7.   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.
• 8.   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.
• 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.
• 10.   Monreal M, Suárez C, Fajardo JA, Barba R, Uresandi F, Valle R, et al. Management of patients with acute venous thromboembolism: Findings from the RIETE registry. Pathophysiol Haemost Thromb 2003; 33: 330–334.
• 11.   Kearon C, Akl EA, Comerota AJ, Prandoni P, Bounameaux H, Goldhaber SZ, et al. Antithrombotic therapy for VTE disease: Antithrombotic therapy and prevention of thrombosis, 9th edn: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012; 141: e419S–e494S.
• 12.   Jaff MR, McMurtry MS, Archer SL, Cushman M, Goldenberg N, Goldhaber SZ, et al. Management of massive and submassive pulmonary embolism, iliofemoral deep vein thrombosis, and chronic thromboembolic pulmonary hypertension: A scientific statement from the American Heart Association. Circulation 2011; 123: 1788–1830.
• 13.   Laporte S, Mismetti P, Decousus H, Uresandi F, Otero R, Lobo JL, et al. Clinical predictors for fatal pulmonary embolism in 15,520 patients with venous thromboembolism: Findings from the Registro Informatizado de la Enfermedad TromboEmbolica venosa (RIETE) registry. Circulation 2008; 117: 1711–1716.
• 14.   Yamashita Y, Morimoto T, Amano H, Takase T, Hiramori S, Kim K, et al. Anticoagulation therapy for venous thromboembolism in the real world: From the COMMAND VTE Registry. Circ J 2018; 82: 1262–1270.
• 15.   Schulman S, Kearon C, Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost 2005; 3: 692–694.
• 16.   Prandoni P, Lensing AW, Harry RB, Alberto C, Martin HP, Anna MC, et al. Deep-vein thrombosis and the incidence of subsequent symptomatic cancer. N Engl J Med 1992; 327: 1128–1133.
• 17.   Sörensen HT, Melemkjær L, Olsen JH, Baron JA. Prognosis of cancers associated with venous thromboembolism. N Engl J Med 2000; 343: 1846–1850.
• 18.   Prandoni P, Lensing AW, Piccioli A, Bernardi E, Simioni P, Girolami B, et al. Recurrent venous thromboembolism and bleeding complications during anticoagulant treatment in patients with cancer and venous thrombosis. Blood 2002; 100: 3484–3488.
• 19.   Khorana AA, Dalal M, Lin J, Connolly GC. Incidence and predictors of venous thromboembolism (VTE) among ambulatory high-risk cancer patients undergoing chemotherapy in the United States. Cancer 2013; 119: 648–655.
• 20.   Ogino Y, Ishigami T, Minamimoto Y, Kimura Y, Akiyama E, Okada K, et al. Direct oral anticoagulant therapy for cancer-associated venous thromboembolism in routine clinical practice. Circ J 2020; 84: 1330–1338.
• 21.   Saeed MJ, Turner TE, Brown DL. Trends in inferior vena cava filter placement by indication in the United States from 2005 to 2014. JAMA Intern Med 2017; 177: 1861–1862.
• 22.   Roetker NS, Lutsey PL, Zakai NA, Alonso A, Adam TJ, MacLehose RF. All-cause mortality risk with direct oral anticoagulants and warfarin in the primary treatment of venous thromboembolism. Thromb Haemost 2018; 118: 1637–1645.
• 23.   Nakamura M, Nishikawa M, Komuro I, Kitajima I, Uetsuka Y, Yamagami T, et al. Apixaban for the treatment of Japanese subjects with acute venous thromboembolism (AMPLIFY-J Study). Circ J 2015; 79: 1230–1236.
• 24.   Costa OS, Thompson S, Ashton V, Palladino M, Bunz TJ, Coleman CI. Rivaroxaban versus warfarin for treatment and prevention of recurrence of venous thromboembolism in African American patients: A retrospective cohort analysis. Thromb J 2020; 18: 6.
• 25.   Jun M, Lix LM, Durand M, Dahl M, Paterson JM, Dormuth CR, et al. Comparative safety of direct oral anticoagulants and warfarin in venous thromboembolism: Multicentre, population based, observational study. BMJ 2017; 359: j4323.
• 26.   Hlavacek P, Guo JD, Rosenblatt L, Keshishian A, Russ C, Mardekian J, et al. Safety, effectiveness, and health care cost comparisons among elderly patients with venous thromboembolism prescribed warfarin or apixaban in the United States Medicare population. Curr Med Res Opin 2019; 35: 2043–2051.
• 27.   Kohn CG, Bunz TJ, Beyer-Westendorf J, Coleman CI. Comparative risk of major bleeding with rivaroxaban and warfarin: Population-based cohort study of unprovoked venous thromboembolism. Eur J Haematol 2019; 102: 143–149.
• 28.   Ng DL, Gan GG, Chai CS, Chee KH, Tan KL, Tan SB, et al. Comparing quality of life and treatment satisfaction between patients on warfarin and direct oral anticoagulants: A cross-sectional study. Patient Prefer Adherence 2019; 13: 1363–1373.
• 29.   Basto AN, Fewel NP, Vo K, Stock EM, Ta M. Initiation of direct oral anticoagulants versus warfarin for venous thromboembolism: Impact on time to hospital discharge. J Thromb Thrombolysis 2018; 45: 51–55.
• 30.   Witt DM, Clark NP, Kaatz S, Schnurr T, Ansell JE. Guidance for the practical management of warfarin therapy in the treatment of venous thromboembolism. J Thromb Thrombolysis 2016; 41: 187–205.
• 31.   Lyman GH. Venous thromboembolism in the patient with cancer. Focus on burden of disease and benefits of thromboprophylaxis. Cancer 2011; 117: 1334–1349.
• 32.   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.
• 33.   Key NS, Khorana AA, Kuderer NM, Bohlke K, Lee AYY, Arcelus JY, et al. Venous thromboembolism prophylaxis and treatment in patients with cancer: ASCO Clinical Practice Guideline Update. J Clin Oncol 2020; 38: 496–520.
• 34.   Agnelli G, Becattini C, Meyer G, Muñoz A, Huisman MV, Connors JM, et al. Apixaban for the treatment of venous thromboembolism associated with cancer. N Engl J Med 2020; 382: 1599–1607.
• 35.   Raskob GE, van Es N, Verhamme P, Carrier M, Di Nisio M, Garcia D, et al. Edoxaban for the treatment of cancer-associated venous thromboembolism. N Engl J Med 2018; 378: 615–624.
• 36.   Young AM, Marshall A, Thirlwall J, Chapman O, Lokare A, Hill C, et al. Comparison of an oral factor Xa inhibitor with low molecular weight heparin in patients with cancer with venous thromboembolism: Results of a randomized trial (SELECT-D). J Clin Oncol 2018; 36: 2017–2023.