- Email: [email protected]
Risk for Venous Thromboembolism Recurrence Among Rivaroxaban-treated Patients Who Continued Versus Discontinued Therapy: Analyses Among Patients with VTE
Clinical Therapeutics/Volume ], Number ], 2017
Risk for Venous Thromboembolism Recurrence Among Rivaroxaban-Treated Patients Who Continued Versus Discontinued Therapy: Analyses Among Patients with VTE Alok A. Khorana, MD1; Jeffrey S. Berger, MD2; Philip S. Wells, MD3; Roger Seheult, MD4; Veronica Ashton, MPH5; Franc¸ois Laliberte´, MA6; Concetta Crivera, PharmD, MPH5; Dominique Lejeune, MSc6; Jeff Schein, DrPH, MPH5; Peter Wildgoose, PhD5; Patrick Lefebvre, MA6; and Scott Kaatz, DO, MSc7 1
Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio; 2New York University School of Medicine, New York, New York; 3The Ottawa Hospital, Ottawa, Ontario, Canada; 4Loma Linda University School of Medicine, Loma Linda, California; 5Janssen Scientific Affairs, LLC, Raritan, New Jersey; 6Groupe d’analyse, Lte´e, Montre´al, Que´bec, Canada; and 7Henry Ford Hospital, Detroit, Michigan ABSTRACT Purpose: The EINSTEIN-Extension trial showed that an extended rivaroxaban treatment significantly reduced the risk for venous thromboembolic (VTE) recurrence. The present study assessed the risk for VTE recurrence and major bleeding associated with extended rivaroxaban treatment in a clinical practice setting among patients with VTE. Methods: A retrospective study was conducted using claims data from February 2011 to April 2015. It included adult patients who initiated rivaroxaban therapy within 7 days after their first VTE and who continuously used rivaroxaban for at least 3 months (index date: end of initial 3-month treatment). Categorized into discontinued and continued cohorts, patients were followed up from the index date until the end of continuous treatment (continued cohort) or end of data or reinitiation of oral anticoagulant therapy (discontinued cohort). Using inverse probability of treatment weights controlling for confounders, adjusted Kaplan-Meier rates of recurrent VTE and major bleeding events were compared. Findings: The analysis showed that, compared with the discontinued cohort (n ¼ 1,536), the continued cohort (n ¼ 5,933) had a significantly lower VTE recurrence rate after an additional 3 months (0.70% vs 1.70%), 6 months (1.41% vs 2.34%), 9 months (1.82% vs 3.01%), and 12 months (1.97% vs 3.01%) of treatment (all, p o 0.05). The difference in the cumulative event rates for major bleeding was not statistically significant. Similar results were obtained
] 2017
in an analysis among patients with VTE receiving rivaroxaban for Z6 months. Implications: Our results suggest that, in clinical practice settings, patients with VTE who continued rivaroxaban therapy after the initial 3- or 6-month treatment period had a significantly lower risk for VTE recurrence without a statistically significant increased risk for major bleeding. (Clin Ther. 2017;]:]]]–]]]) & 2017 Elsevier HS Journals, Inc. All rights reserved. Key words: anticoagulant, extended treatment, recurrent, rivaroxaban, venous thromboembolism.
INTRODUCTION Venous thromboembolism (VTE) is a medical condition that encompasses deep vein thrombosis or pulmonary embolism (PE). First-time VTE affects 100 per 100,000 persons (0.1%) in the United States each year, with the incidence rising substantially to 450 to 600 per 100,000 ( 0.5%) among individuals over the age of 80 years.1 Patients with a history of VTE are at an increased risk for VTE recurrence compared with patients Parts of this article were presented at the 58th Annual Meeting of the American Society of Hematology, December 3–6, 2016, San Diego, Calif. Accepted for publication May 25, 2017. http://dx.doi.org/10.1016/j.clinthera.2017.05.357 0149-2918/$ - see front matter & 2017 Elsevier HS Journals, Inc. All rights reserved.
1
Clinical Therapeutics without prior VTE.1–4 The risk varies with time and is the highest in the first 6 to 12 months after the initial episode.1,5 Patients with cancer as well as those not provoked by transient risk factors have a further increased risk for recurrence.1,5 The importance of anticoagulation therapy in patients with VTE is well recognized, but the optimal treatment duration is uncertain due to an increased risk for bleeding associated with long-term anticoagulation. The American College of Chest Physicians guideline on antithrombotic therapy for VTE disease recommends 3 months of therapy in most patients with VTE, while extended therapy beyond 3 months is reserved for those with a high risk for recurrence— patients with first or second unprovoked VTE and active cancer.6 A recent meta-analysis by Bova et al7 concluded that longer anticoagulation would be beneficial against intermediate risk for recurrence, while Cohen et al8 suggested that direct oral anticoagulants would be an effective alternative to the current standard of care for extended treatment (i.e., warfarin or aspirin). EINSTEIN-Ext (Oral Rivaroxaban for Symptomatic Venous Thromboembolism)9 assessed the benefit of prolonged (additional 6 or 12 months) rivaroxaban treatment in patients with VTE who completed the first 6 to 12 months of anticoagulation with a vitamin K antagonist or rivaroxaban. The study showed that rivaroxaban reduced the risk for recurrent VTE by 82%, with a small increased risk for major bleeding (0.7% in the rivaroxaban cohort vs none in the placebo cohort), suggesting that rivaroxaban may provide a welltolerated and effective approach to extended treatment of VTE. While these findings on extended anticoagulant treatment in the EINSTEIN-Ext study are promising, data assessing the duration of treatment with anticoagulants and associated outcomes in clinical practice are limited. The objective of the present study was to assess the association between extended rivaroxaban treatment and the risk for VTE recurrence and major bleeding in a clinical practice setting among patients with VTE.
PATIENTS AND METHODS Data Source Data dated February 2011 to April 2015 in the Truven Health Analytics MarketScan Databases were used in this study. This Commercial Claims and
2
Encounters database includes claims from employees and their dependents covered by employer-sponsored private health insurance plans from 100 employers and a number of health plans. The Medicare Supplemental and Coordination of Benefits database focuses on patients aged Z65 years with Medicare coverage plus employer-paid commercial plans. The data included in both MarketScan databases are de-identified and are in compliance with the Health Insurance Portability and Accountability Act of 1996 to preserve participant anonymity and confidentiality.
Study Design A retrospective longitudinal study design was used for comparing the risk for recurrent VTE and major bleeding events among patients who continued and patients who discontinued treatment with rivaroxaban. All adult (aged Z18 years) patients initiated on rivaroxaban treatment within 7 days after a first VTE (index VTE) were included in the study. Patients were required to have had data available from at least 12 months before the index date and not to have received anticoagulant therapy prior to the index VTE. VTEs were identified from hospitalization discharge or outpatient/emergency department visit claims with a primary or secondary VTE diagnosis (International Classification of Disease, Ninth Revision–Clinical Modification [ICD-9-CM] codes: deep vein thrombosis, 451.1, 451.2, 453.4, 453.8, 453.9; or PE, 415.1). The study population included patients who received, for at least 3 months, continuous rivaroxaban treatment, defined as rivaroxaban use, without a gap of 430 days, after the first VTE event. The end of the initial 3-month rivaroxaban treatment period following the first VTE event defined the index date. Patients whose continuous treatment ended within the 30-day window after the index date were assigned to the discontinued cohort, while patients whose continuous treatment ended after this 30-day window were assigned to the continued cohort. Figure 1 depicts the study design. A similar design was used for identifying the population treated for 6 months, except that the index date was defined as the end of the 6-month period and this subgroup included only patients treated for at least 6 months. Patients were excluded if they had one or more claims for any anticoagulant agent (ie, warfarin, lowmolecular-weight heparin, rivaroxaban, dabigatran, apixaban, or fondaparinux), a diagnosis of atrial fibrillation Volume ] Number ]
A.A. Khorana et al.
First day of eligibility
First VTE diagnosis
Initiation of rivaroxaban
Index date after 3 or 6 months of rivaroxaban treatment
End of observation
3- or 6-month treatment
12-month baseline period
Observation (follow-up) period
Study period Continuous insurance coverage
Figure 1. Study design scheme.
(ICD-9-CM 427.31), or a major bleeding event10 during the 12-month baseline period; or had a recurrent VTE event between the index VTE and the index date. Baseline characteristics were measured during the 12-month period prior to the index date, including demographic and clinical characteristics, risk factors for VTE (including risk factors from the American College of Chest Physicians Evidence-Based Clinical Practice guideline11), risk factors for bleeding,12 and the Registro Informatizado de Enfermedad TromboEmbólica (RIETE) bleeding risk assessment model score (having a diagnosis of chronic kidney disease was used as the proxy for the abnormal creatinine level).13
Outcomes Measures The primary end point of the study was recurrent VTE, defined as hospitalization with a primary diagnosis of VTE. Major bleeding events were identified based on primary ICD-9-CM diagnostic codes during hospitalization using a validated algorithm developed by Cunningham et al.10
Statistical Analysis Descriptive analyses were conducted for comparing baseline demographic and clinical characteristics between cohorts. Mean (SD) and median values were reported for continuous variables; frequency and percentage were reported for categorical variables. Statistical differences between cohorts were assessed using standardized differences. A standardized difference of o10% was considered an indication of wellbalanced distribution of the given variable.14 The inverse probability of treatment weight (IPTW), based on a propensity score method, was used for assessing the relationship between the continued (vs
] 2017
discontinued) treatment and outcomes in each population. Weighting patients by the inverse probability of the treatment received creates a synthetic sample independent of baseline coviariates.15 Each patient was assigned a weight such that in the weighted pseudopopulation, the distribution of measured confounders was similar between the continued and discontinued cohorts. In order to calculate IPTWs, the probability (ie, the propensity score) of receiving continued treatment (vs discontinued treatment) was first estimated using a multivariate logistic regression model conditional on baseline covariates, including age, sex, region of residence, insurance type, year of index date, type of first VTE, time from first VTE to first rivaroxaban dispensing, baseline risk factors for bleeding and VTE,11,12 Quan-Charlson comorbidity index, RIETE bleeding score, and health care resource utilization and costs. IPTWs were then calculated as the inverse of patients’ estimated probabilities of having their observed treatment; that is, IPTWs were calculated as 1/(Propensity score) in the continued cohort and as 1/(1 – [Propensity score]) in the discontinued cohort. Finally, the IPTWs were normalized within each cohort by dividing each IPTW by the mean of the IPTWs in each cohort. Kaplan-Meier curves for time to first recurrent VTE and time to first major bleeding event from the index date were compared between cohorts in the IPTWweighted populations. Kaplan-Meier rates for VTE recurrences and major bleeding events at 3, 6, 9, and 12 months after the index date were compared between cohorts using log-rank tests. All of the analyses in the study were performed with SAS software version 9.4 (SAS Institute, Cary, North Carolina).
RESULTS Patients’ Baseline Characteristics A total of 7,469 patients who received rivaroxaban for at least 3 months after the initial VTE event were included in the 3-month analysis; 5,933 and 1,536 patients formed the continued and discontinued cohorts, respectively. For the 6-month analysis, 2,676 patients were included in the continued cohort and 1,127 in the discontinued cohort. Between-cohort comparisons of patients’ baseline characteristics and baseline risk factors for bleeding and VTE are presented in Tables I and II, respectively. After weighting, the continued and discontinued cohorts were wellbalanced across baseline characteristics and risk factors.
3
At Least 3 mo of Therapy Characteristic
Volume ] Number ]
Duration, mean (SD) [median], d Observation period† Total treatment (including 3 mo of rivaroxaban therapy)‡ Propensity score variables Demographics Age, mean (SD) [median], y Female, no. (%) Insurance type, no. (%) Consumer-directed Exclusive provider organization Comprehensive Health maintenance organization High-deductible Preferred provider organization Point-of-service Point-of-service capitated Not specified
At Least 6 mo of Therapy
Discontinued Cohort (n ¼ 1,536)
Continued Cohort (n ¼ 5,933)
Standardized Discontinued Cohort Difference, % (n ¼ 1,127)
211.1 (191.6) [162] 105.7 (8.5) [107]
149.3 (124.4) [108] 242.0 (126.2) [199]
– –
206.5 (171.5) [158] 196.0 (7.9) [197]
57.1 (15.4) [58]
56.8 (14.7) [57]
2.1
57.7 (14.3) [58]
Continued Cohort (n ¼ 2,676)
Standardized Difference, %
158.5 (130.6 ) [114] 341.3 (132.0 ) [297]
– –
57.8 (14.2) [58]
0.3
701 (45.6)
2,735 (46.1)
0.9
487 (43.2)
1,152 (43.1)
0.4
892 (58.1) 212 (13.8)
3,432 (57.9) 805 (13.6)
0.4 0.6
650 (57.6) 163 (14.5)
1,534 (57.3) 385 (14.4)
0.6 0.3
125 (8.1) 118 (7.7)
467 (7.9) 459 (7.7)
1.0 0.3
95 (8.4) 87 (7.7)
208 (7.8) 215 (8.0)
2.3 1.0
92 (6.0) 70 (4.6)
371 (6.2) 279 (4.7)
1.2 0.7
65 (5.8) 45 (4.0)
163 (6.1) 115 (4.3)
1.3 1.6
5 (0.3) 3 (0.2)
22 (0.4) 24 (0.4)
0.7 4.1
4 (0.4) 8 (0.7)
10 (0.4) 8 (0.3)
0.1 5.4
21 (1.3)
73 (1.2)
1.1
10 (0.9)
38 (1.4)
5.0 (continued)
Clinical Therapeutics
4 Table I. Demographic and clinical characteristics assessed during the 12-month baseline period, weighted population.*
] 2017
Table I. (continued). At Least 3 mo of Therapy Characteristic
6 412 866 252 705
(0.4) (26.9) (56.4) (16.4) (45.9)
Continued Cohort (n ¼ 5,933)
(0.4) (26.4) (57.0) (16.3) (46.3)
0.1 1.1 1.2 0.2 0.7
3 205 714 205 586
807 (52.5)
3,171 (53.4)
1.8
442 (28.7)
1,639 (27.6)
287 (18.7)
1,123 (18.9)
0.69 (1.18) [0] 949 365 101 54 67
(61.8) (23.8) (6.6) (3.5) (4.4)
22 1,564 3,381 967 2,746
Standardized Discontinued Cohort Difference, % (n ¼ 1,127)
0.68 (1.17) [0] 3,674 1,435 364 201 258
(61.9) (24.2) (6.1) (3.4) (4.4)
(0.2) (18.2) (63.4) (18.2) (52.0)
Continued Cohort (n ¼ 2,676)
(0.2) (18.7) (62.4) (18.7) (50.7)
0.1 1.3 2.0 1.1 2.5
539 (47.8)
1,307 (48.8)
2.0
2.5
311 (27.6)
748 (27.9)
0.9
0.6
277 (24.6)
622 (23.2)
3.2
0.68 (1.20) [0]
2.6
0.8
0.65 (1.14) [0]
0.3 1.0 1.7 0.7 0.1
710 264 64 44 46
(63.0) (23.4) (5.7) (3.9) (4.1)
6 500 1,670 500 1,358
Standardized Difference, %
1,667 629 159 104 117
(62.3) (23.5) (5.9) (3.9) (4.4)
1.4 0.3 1.0 0.0 1.7 (continued)
5
A.A. Khorana et al.
Year of index date, no. (%) 2012 2013 2014 2015 First VTE diagnosis identified in-hospital, no. (%) Type of first VTE diagnosis, no. (%) Deep vein thrombosis Pulmonary embolism Both diagnoses on the same day Time from first VTE to first rivaroxaban dispensing Mean (SD) [median], d Same day, no. (%) 1 d, no. (%) 2 d, no. (%) 3 d, no. (%) 4–7 d, no. (%)
Discontinued Cohort (n ¼ 1,536)
At Least 6 mo of Therapy
At Least 3 mo of Therapy Characteristic
Discontinued Cohort (n ¼ 1,536)
Continued Cohort (n ¼ 5,933)
Volume ] Number ]
Comorbidity index scores Quan-Charlson Mean (SD) [median] 1.66 (2.41) [1] 1.57 (2.30) [1] Comorbidities, no. (%) 0 650 (42.3) 2,581 (43.5) 1 337 (21.9) 1,337 (22.5) 2 209 (13.6) 759 (12.8) Z3 340 (22.1) 1,256 (21.2) RIETE Mean (SD) [median] 1.43 (1.35) [1] 1.38 (1.35) [1] Comorbidities, no. (%) 0 467 (30.4) 1,847 (31.1) 1–4 999 (65.1) 3,814 (64.3) 44 69 (4.5) 272 (4.6) Baseline health care resource utilization, mean (SD) [median] Outpatient visits 20.62 (16.96) [16] 20.11 (16.66) [15] ED visits 1.11 (1.59) [1] 1.06 (1.48) [1] Hospitalizations 0.71 (0.74) [1] 0.71 (0.74) [1] Baseline health care costs, mean (SD), year-2015 $US Total health care 37,207 (74,562) 36,028 (57,866) costs Hospitalizations 16,448 (41,767) 16,166 (41,435)
At Least 6 mo of Therapy Standardized Discontinued Cohort Difference, % (n ¼ 1,127)
3.7
1.64 (2.22) [1]
2.4 1.5 2.4 2.4
449 257 168 252
3.8
1.47 (1.35) [1]
1.5 1.6 0.4
309 (27.5) 758 (67.3) 59 (5.2)
3.0 3.4 0.0
(39.8) (22.8) (14.9) (22.4)
21.15 (16.20) [17] 1.01 (1.46) [1] 0.75 (0.79) [1]
Continued Cohort (n ¼ 2,676)
1.66 (2.33) [1] 1,088 619 372 597
Standardized Difference, %
0.8
(40.7) (23.1) (13.9) (22.3)
1.7 0.6 2.9 0.2
1.47 (1.38) [1]
0.2
747 (27.9) 1,787 (66.8) 142 (5.3)
1.0 1.1 0.3
21.13 (17.39) [16] 1.03 (1.59) [1] 0.73 (0.71) [1]
0.1 1.4 2.8
1.8
48,597 (93,979)
38,068 (61,852)
13.2
0.7
16,432 (32,119)
15,969 (42,297)
1.2 (continued)
Clinical Therapeutics
6 Table I. (continued).
] 2017
ED ¼ emergency department; RIETE ¼ Registro Informatizado de Enfermedad TromboEmbólica; VTE ¼ venous thromboembolism. * The propensity score of receiving continued treatment (vs discontinued treatment) was estimated using a multivariate logistic regression model conditional on baseline covariates including age, sex, region, insurance type, year of index date, episode type of first VTE, type of first VTE, time from first VTE to first rivaroxaban dispensing, baseline risk factors for bleeding and VTE, Quan-Charlson comorbidity index, RIETE score, and health care resource utilization and costs. † From the index date to the earliest date between initiation of a new anticoagulant therapy, end of data availability (April 2015), or end of insurance coverage as well as the end of the rivaroxaban therapy for the continued cohort. ‡ Duration of treatment was defined as the duration of continuous use of rivaroxaban from the first dispensing until a 30-day interruption or end of follow-up.
15.5 5.3 0.9 12,961 (30,870) 6,067 (10,277) 2,244 (4,880) 22,435 (80,858) 6,711 (13,869) 2,203 (4,468) 2.2 0.6 2.0 12,026 (27,875) 4,767 (9,342) 2,195 (4,370) 12,736 (36,978) 4,824 (9,042) 2,287 (4,841) Outpatient visits Pharmacy ED visits
Continued Cohort (n ¼ 2,676) Standardized Discontinued Cohort Difference, % (n ¼ 1,127) Continued Cohort (n ¼ 5,933) Discontinued Cohort (n ¼ 1,536) Characteristic
Table I. (continued).
At Least 3 mo of Therapy
At Least 6 mo of Therapy
Standardized Difference, %
A.A. Khorana et al.
VTE Recurrence In the weighted population, 89 recurrent VTEs were identified after 12 months of follow-up. The continued cohort had significantly lower rates of recurrent VTEs than did the discontinued cohort at 3 months (0.70% vs 1.70%; P = 0.003), 6 months (1.41% vs 2.34%; P = 0.016), and 12 months (1.97% vs 3.01%; P = 0.017) of follow-up (Figure 2A). Comparisons also showed that, among the 3803 patients who received rivaroxaban treatment for at least 6 months, continued treatment was associated with a reduced risk for recurrent VTEs (Figure 2B).
Major Bleeding Events In the weighted population, 57 major bleeding events were identified at 12 months of follow-up. The rates of major bleeding (gastrointestinal or other bleeding) in the continued and the discontinued cohorts, respectively, were 0.58% and 0.82% at 3 months (P ¼ 0.367), 0.91% and 0.88% at 6 months (P ¼ 0.754), and 1.44% and 1.44% at 12 months (P ¼ 0.813) (Figure 3A). Results in patients treated for at least 6 months were consistent with those treated at least 3 months (Figure 3B).
DISCUSSION The present study, using claims data from clinical practice, assessed the impact of extended treatment with rivaroxaban on the risks for recurrent VTE and major bleeding. Continuing therapy for 43 months was associated with a 35% relative risk reduction of recurrent VTE at 12 months (rate of recurrent VTE, 1.97% vs 3.01%), without a significant difference in the risk for major bleeding. In patients who received rivaroxaban treatment for at least 6 months, continuing therapy was associated with a 54% relative risk reduction in VTE recurrence (rate of recurrent VTE, 1.72% vs 3.70%). The results from the present study are in line with results from the EINSTEIN-Ext randomized clinical trial, in which extended treatment with rivaroxaban was associated with a lower risk for VTE recurrence without any significant impact on the risk for bleeding. The mean age in EINSTEIN-Ext was 58 years, 38% of patients had a PE, 73% had an unprovoked event, and 4% to 5% had active cancer. In the present study, the mean age was 57 years, 48% had an initial diagnosis of PE, 67% had an unprovoked event, and
7
At least 3 months of therapy Characteristics, n (%)
Volume ] Number ]
VTE and bleeding Hypertension Diabetes Cancer Cerebrovascular accident (stroke) VTE Multiple trauma Hyperlipidemia Other serious infections Obesity Abdominal surgery Arrhythmia Major surgery Pneumonia Contraceptive pill (use of oral) Hip, pelvis, or leg fracture Varicose veins Congestive heart failure COPD Thrombophilia Myocardial infarction Pregnancy Total knee replacement Left ventricular dysfunction Rheumatoid arthritis Inflammatory bowel disease Treatment with aromatase inhibitors Surgical resection of abdominal or pelvic cancer Total hip replacement Treatment with SERMs
Discontinued Cohort (N ¼ 1,536)
Continued Cohort (N ¼ 5,933)
At least 6 months of therapy Standardized Difference
Discontinued Cohort (N ¼ 1,127)
Continued Cohort (N ¼ 2,676)
Standardized Difference
808 290 267 60
(52.6%) (18.9%) (17.4%) (3.9%)
3,044 1,101 956 206
(51.3%) (18.6%) (16.1%) (3.5%)
2.6% 0.8% 3.4% 2.3%
603 226 205 48
(53.5%) (20.0%) (18.2%) (4.3%)
1,443 551 442 98
(53.9%) (20.6%) (16.5%) (3.7%)
0.8% 1.4% 4.5% 3.1%
540 664 306 277 211 200 190 177 98
(35.2%) (43.2%) (19.9%) (18.0%) (13.8%) (13.0%) (12.4%) (11.5%) (6.4%)
2,082 2,523 1,193 1,074 780 750 692 680 406
(35.1%) (42.5%) (20.1%) (18.1%) (13.2%) (12.6%) (11.7%) (11.5%) (6.8%)
0.1% 1.4% 0.5% 0.2% 1.8% 1.2% 2.2% 0.2% 1.8%
363 494 212 233 151 157 125 138 65
(32.2%) (43.8%) (18.8%) (20.6%) (13.4%) (14.0%) (11.1%) (12.3%) (5.8%)
869 1,203 518 511 363 364 306 312 159
(32.5%) (45.0%) (19.4%) (19.1%) (13.6%) (13.6%) (11.4%) (11.6%) (5.9%)
0.4% 2.3% 1.3% 3.9% 0.5% 1.1% 1.1% 2.0% 0.6%
89 75 111 72 63 61 24 19 28 32 23 19
(5.8%) (4.9%) (7.3%) (4.7%) (4.1%) (4.0%) (1.5%) (1.2%) (1.8%) (2.1%) (1.5%) (1.3%)
(5.6%) (5.1%) (6.8%) (4.6%) (4.1%) (3.7%) (1.7%) (1.3%) (2.0%) (2.1%) (1.6%) (1.2%)
0.9% 0.9% 1.8% 0.3% 0.0% 1.5% 1.0% 0.0% 1.5% 0.2% 0.7% 1.0%
47 63 80 47 74 53 35 8 23 25 16 11
(4.1%) (5.6%) (7.1%) (4.2%) (6.5%) (4.7%) (3.1%) (0.7%) (2.1%) (2.2%) (1.4%) (0.9%)
(4.2%) (5.8%) (6.9%) (4.4%) (5.3%) (3.9%) (1.6%) (0.7%) (2.0%) (2.2%) (1.5%) (1.2%)
0.3% 0.9% 0.6% 0.8% 5.4% 3.6% 10.2% 0.0% 0.1% 0.3% 0.5% 2.5%
18 (1.2%)
80 (1.4%)
1.3%
14 (1.3%)
33 (1.2%)
0.2%
13 (0.9%) 7 (0.5%) 8 (0.5%)
38 (0.6%) 35 (0.6%) 27 (0.4%)
2.4% 1.6% 1.4%
8 (0.7%) 8 (0.7%) 5 (0.4%)
17 (0.6%) 17 (0.6%) 13 (0.5%)
0.9% 1.2% 0.8%
331 302 403 274 245 220 99 74 120 126 95 68
112 156 185 117 141 105 42 19 55 60 39 32
(continued)
Clinical Therapeutics
8
Table II. Risk Factors assessed during the 12-month baseline period, weighted population*.
] 2017
Table II. (continued). At least 3 months of therapy Characteristics, n (%) Treatment with erythropoiesis stimulating agents Spinal cord injury Immobility Bleeding NSAID use Anemia Renal disease Excessive fall risk (Parkinson’s disease, etc.) Chronic kidney disease Hepatic disease Coagulation defect Central venous catheter Thrombocytopenia (low platelet count) Ethanol abuse Peptic ulcer Bleeding diathesis
Discontinued Cohort (N ¼ 1,536)
6 (0.4%) 2 (0.1%) 456 283 279 247
(29.7%) (18.4%) (18.2%) (16.1%)
110 111 84 37 50
(7.2%) (7.2%) (5.5%) (2.4%) (3.3%)
24 (1.5%) 16 (1.0%) 0 (0.0%)
Continued Cohort (N ¼ 5,933)
18 (0.3%) 8 (0.1%) 1,764 1,070 1,045 949
At least 6 months of therapy Standardized Difference
Discontinued Cohort (N ¼ 1,127)
1.3% 0.9%
2 (0.2%) 0 (0.0%)
Continued Cohort (N ¼ 2,676)
5 (0.2%) 2 (0.1%)
Standardized Difference
0.6% 3.9%
(29.7%) (18.0%) (17.6%) (16.0%)
0.1% 1.0% 1.4% 0.3%
292 204 185 174
(25.9%) (18.1%) (16.5%) (15.4%)
699 492 462 421
(26.1%) (18.4%) (17.3%) (15.7%)
0.5% 0.6% 2.2% 0.8%
(6.6%) (7.0%) (5.3%) (2.4%) (3.1%)
2.2% 0.9% 0.8% 0.2% 0.7%
75 82 86 22 44
(6.6%) (7.3%) (7.6%) (2.0%) (3.9%)
183 206 175 56 96
(6.8%) (7.7%) (6.5%) (2.1%) (3.6%)
0.7% 1.4% 4.3% 0.7% 1.8%
95 (1.6%) 47 (0.8%) 3 (0.0%)
0.4% 2.6% 3.1%
15 (1.3%) 9 (0.8%) 2 (0.1%)
38 (1.4%) 20 (0.8%) 1 (0.0%)
1.1% 0.3% 3.5%
393 415 314 144 186
Abbreviations: VTE: venous thromboembolism; NSAID: nonsteroidal anti-inflammatory drugs; COPD: chronic obstructive pulmonary disease; SERM: selective estrogen receptor modulator Notes: ⁎ The propensity score of receiving continued treatment (vs. discontinued treatment) was estimated using a multivariate logistic regression model conditional on baseline covariates including age, gender, region, insurance type, year of index date, episode type of first VTE, type of first VTE, time from first VTE to first rivaroxaban dispensing, baseline risk factors for bleeding and VTE, Quan-Charlson comorbidity index, RIETE score, and healthcare resource utilization and costs.
A.A. Khorana et al.
9
Clinical Therapeutics
% of Patients with a Recurrent VTE
A
8%
After 3 months of therapy
Discontinued Continued
Log -rank P-value = 0.017
6%
Log -rank P-value = 0.012 3-month period of initial rivaroxaban treatment
4%
Log -rank P -value = 0.016 Log -rank P -value = 0.003 1.70%
3.01% 2.34%
1.82%
3.01% 1.97%
1.41%
0.70% 2%
0% Discontinued Continued
% of Patients with a Recurrent VTE
B
0 Days Number at risk 1,536 Number of events Number at risk 5,933 Number of events -
8%
90 964 20 3,645 37
180 699 26 1,487 53
270 471 30 806 58
After 6 months of therapy
Discontinued Continued
Log -rank P -value = 0.024
6%
4%
365 305 30 417 59
Log -rank P-value = 0.023 6-month period of initial rivaroxaban treatment
3.70%
Log -rank P-value = 0.023 Log-rank P-value = 0.133 1.41%
3.02% 2.69%
1.72%
1.35%
1.22%
0.82%
2%
0% Discontinued Continued
0 Days Number at risk 1,127 Number of events Number at risk 2,676 Number of events -
90 748 14 1,567 17
180 511 22 844 22
270 347 23 445 23
365 221 25 243 25
Figure 2. Inverse probability of treatment weight–adjusted Kaplan-Meier rates of recurrent venous thromboembolism (VTE) after 3 (A) and 6 (B) months of treatment among rivaroxaban users who continued versus discontinued therapy.
12% had a diagnosis of cancer in the year prior to the index date. While both studies found a significant impact of extended therapy on the reduction in the risk for recurrence, the magnitude of the reduction in the 12-month rate of VTE recurrence was higher in EINSTEIN-Ext—7.1% versus 1.3% for rivaroxaban versus placebo. The smaller effect observed in our study may have been related to potential underestimation of the number of recurrent VTEs: Our study captured only recurrent VTEs diagnosed during hospitalization and not in the outpatient setting. However, because patients with PE are more likely to be diagnosed and treated in hospitals, our approach was
10
less likely to have underestimated PE recurrence. The findings from the 6-month sensitivity analysis are consistent with those from the EINSTEIN-Ext study and others, which have shown that extended anticoagulation treatment had efficacy superior (i.e., recurrent VTE) to that of placebo in patients with VTE who had received 6 to 12 months of prior anticoagulant treatment.8,9,16,17 No statistically significant differences in the risk for major bleeding were observed between the continued and discontinued cohorts in either the 3- or 6-month analysis. The bleeding rates at 12 months were similar between the continued and discontinued cohorts.
Volume ] Number ]
A.A. Khorana et al.
A
8%
After 3 months of therapy
% of Patients with Major Bleeding
Discontinued Continued Log-rank P-value = 0.813
6%
4%
3-month period of initial rivaroxaban treatment
Log-rank P-value = 0.870 Log-rank P-value = 0.754 Log-rank P-value = 0.367
2%
0.82%
0.88%
0.58%
0.91%
90 964 10 3,657 30
180 699 11 1,497 37
1.18%
1.44%
1.33%
1.44%
0%
Discontinued Continued
% of Patients with Major Bleeding
B
8%
0 Days Number at risk 1,536 Number of events Number at risk 5,933 Number of events -
270 466 13 807 42
365 299 14 416 43
After 6 months of therapy
Discontinued Continued
6%
4%
Log-rank P-value = 0.794 6-month period of initial rivaroxaban treatment
Log-rank P-value = 0.544 Log-rank P-value = 0.854 1.32%
Log-rank P-value = 0.903 0.80% 0.37%
2%
1.32% 0.82%
1.53%
0.72%
0.38% 0%
Discontinued Continued
0 Days Number at risk 1,127 Number of events Number at risk 2,676 Number of events -
90 750 4 1,573 8
180 511 6 845 12
270 348 9 444 13
365 222 9 241 15
Figure 3. Inverse probability of treatment weight–adjusted Kaplan-Meier rates of major bleeding events after 3 (A) and 6 (B) months of treatment among rivaroxaban users who continued versus discontinued therapy. Although it was expected that patients in the continued cohort would have a higher risk for bleeding because the continued cohort remained on bloodthinning medications longer than did the other cohort, our study was underpowered for detecting the association between a major bleeding event with extended treatment because of the short duration of follow-up (up to 12 months). Results from clinical trials have shown that anticoagulation treatment for at least 3 months reduces the risk for VTE recurrence.18 The American College of Chest Physicians suggests extended anticoagulant therapy beyond 3 months among patients with a first
] 2017
unprovoked VTE and who have a low or moderate bleeding risk.19 A systematic review conducted by Kearon and Akl20 found that indefinite anticoagulation (or extended, referring to continued treatment without a scheduled stopping date) is often chosen if there is a low risk for bleeding, whereas anticoagulation is usually stopped at 3 months if there is a high risk for bleeding. Studies included in that systematic review showed that indefinite anticoagulation with a vitamin K antagonist was associated with a reduced risk for recurrent VTE.20 Newer oral anticoagulants have shown efficacy in the prevention of recurrent VTE. The results from 10 Phase III studies demonstrated that rivaroxaban and
11
Clinical Therapeutics dabigatran (1) provided effective and convenient shortterm treatment for VTE compared with the standard of care and (2) showed superiority in the long-term prevention of recurrent VTE compared with placebo.21 The current study was subject to certain limitations. First, billing inaccuracies and missing data may have occurred. It is possible that VTE events were under-coded (ie, false-negatives). Second, recurrent VTE in this study was determined based on only the primary diagnosis during hospitalization. The use of this strict definition on the outcome warrants that the "true" VTE events were captured, while it likely led to underestimation of the number of recurrent VTE events because VTE may have been treated in outpatient settings. Finally, as in all observational studies, adjustments in the multivariate analyses could account for only observable factors.
CONCLUSIONS Our study results suggest that patients with VTE who continued rivaroxaban therapy after the first 3- or 6-month treatment period had a significantly lower risk for VTE recurrence without an increased risk for major bleeding. These results suggest that gains in efficacy observed in extended-therapy trials might translate to clinical practice and provide support for extended therapy to clinicians and patients.
ACKNOWLEDGMENTS Dr. Khorana acknowledges research support from the Sondra and Stephen Hardis Chair in Oncology Research.
CONFLICTS OF INTEREST This research was funded by Janssen Scientific Affairs, LLC (Titusville, New Jersey). The authors are employees of (V.A., C.C., J.S. and P.W.) or have received research funding (A.K.), consultant’s fees (A.K., R.S., J.B., P.W., S.K.), or honoraria (A.K., S.K.) from Janssen. F.L., D.L., and P.L. are employees of Groupe d’analyse, Ltée, a consulting company that has received research grants from Janssen. A.K. has also received research funding from Sanofi and Amgen; and consultant’s fees and honoraria from Leo, Sanofi, Amgen, Bayer, Halozyme, Pfizer, and Roche. J.B. has received research funding from AstraZeneca and has been a member of the scientific
12
advisory board for Merck. P.W. has received research funding from Bristol-Myers Squibb/Pfizer; has been a member of the advisory board of Bayer Health Care; and has served on a writing committee for Itreas. C.C., J.S., and P.W. own stock options in Johnson & Johnson. S.K. has received consultant’s fees from Boehringer Ingelheim, Daiichi Sankyo, Bristol-Myers Squibb, Portola, and Pfizer; and honoraria from Boehringer Ingelheim, Bristol-Myers Squibb, Pfizer, Daiichi Sankyo, and CSL Behring. The authors have indicated that they have no other conflicts of interest with regard to the content of this article.
REFERENCES 1. White RH. The epidemiology of venous thromboembolism. Circulation. 2003;107(Suppl 1):I4–I8. 2. Hansson PO, Sörbo J, Eriksson H. Recurrent venous thromboembolism after deep vein thrombosis: incidence and risk factors. Arch Intern Med. 2000;160:769–774. 3. Kearon C. Natural history of venous thromboembolism. Circulation. 2003;107(Suppl 1):I22–I30. 4. Zhu T, Martinez I, Emmerich J. Venous thromboembolism: risk factors for recurrence. Arterioscler Thromb Vasc Biol. 2009;29:298–310. 5. Heit JA, Spencer FA, White RH. The epidemiology of venous thromboembolism. J Thromb Thrombolysis. 2016;41: 3–14. 6. Kearon C, Akl EA, Ornelas J, et al. Antithrombotic therapy for VTE disease. Chest. 2016;149:315–352. 7. Bova C, Bianco A, Mascaro V, Nobile CG. Extended anticoagulation and mortality in venous thromboembolism. a meta-analysis of six randomized trials. Thromb Res. 2016;139:22–28. 8. Cohen AT, Hamilton M, Bird A, et al. Comparison of the non-VKA oral anticoagulants apixaban, dabigatran, and rivaroxaban in the extended treatment and prevention of venous thromboembolism: systematic review and network meta-analysis. PLoS ONE. 2016;11:e0160064. 9. The EINSTEIN Investigators. Oral rivaroxaban for symptomatic venous thromboembolism. N Engl J Med. 2010; 363:2499–2510. 10. Cunningham A, Stein CM, Chung CP, et al. An automated database case definition for serious bleeding related to oral anticoagulant use. Pharmacoepidemiol Drug Saf. 2011;20: 560–566. 11. Kahn SR, Lim W, Dunn AS, et al. Prevention of VTE in nonsurgical patients. Chest. 2012;141:e195S–e226S. 12. Laliberté F, Cloutier M, Nelson WW, et al. Real-world comparative effectiveness and safety of rivaroxaban and
Volume ] Number ]
A.A. Khorana et al.
13.
14.
15.
16.
17.
18.
19.
20.
21.
warfarin in nonvalvular atrial fibrillation patients. Curr Med Res Opin. 2014;30:1317–1325. Ruíz-Giménez N, Suárez C, González R, 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. Austin P. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38:1228–1234. Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med. 2015;34:3661–3679. Romualdi E, Donadini M, Ageno W. Oral rivaroxaban after symptomatic venous thromboembolism: the continued treatment study (EINSTEINextension study). N Engl J Med. 2010;363:2499–2510. Cohen AT, Dobromirski M. The use of rivaroxaban for short- and longterm treatment of venous thromboembolism. Thromb Haemost. 2012; 107:1035–1043. Smilowitz N, Mega J, Berger J. Duration of anticoagulation for venous thromboembolic events. Circulation. 2014;130:2343–2348. Kearon C, Akl EA, Comerota AJ, et al. Antithrombotic therapy for VTE disease: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141(2 Suppl):e419S–e4194S. Kearon C, Akl EA. Duration of anticoagulant therapy for deep vein thrombosis and pulmonary embolism. Blood. 2014;123:1794–1801. Cohen AT. Long-term benefits of preventing venous thromboembolic events. Curr Med Res Opin. 2012;28:877–889.
] 2017
Address correspondence to: Dominique Lejeune, MSc, Groupe d’analyse, Ltée, 1000, rue de La Gauchetière Ouest, Bureau 1200, Montréal H3B 4W5, Québec, Canada. E-mail: [email protected]
13
Risk for Venous Thromboembolism Recurrence Among Rivaroxaban-Treated Patients Who Continued Versus Discontinued Therapy: Analyses Among Patients with VTE Alok A. Khorana, MD1; Jeffrey S. Berger, MD2; Philip S. Wells, MD3; Roger Seheult, MD4; Veronica Ashton, MPH5; Franc¸ois Laliberte´, MA6; Concetta Crivera, PharmD, MPH5; Dominique Lejeune, MSc6; Jeff Schein, DrPH, MPH5; Peter Wildgoose, PhD5; Patrick Lefebvre, MA6; and Scott Kaatz, DO, MSc7 1
Taussig Cancer Institute, Cleveland Clinic, Cleveland, Ohio; 2New York University School of Medicine, New York, New York; 3The Ottawa Hospital, Ottawa, Ontario, Canada; 4Loma Linda University School of Medicine, Loma Linda, California; 5Janssen Scientific Affairs, LLC, Raritan, New Jersey; 6Groupe d’analyse, Lte´e, Montre´al, Que´bec, Canada; and 7Henry Ford Hospital, Detroit, Michigan ABSTRACT Purpose: The EINSTEIN-Extension trial showed that an extended rivaroxaban treatment significantly reduced the risk for venous thromboembolic (VTE) recurrence. The present study assessed the risk for VTE recurrence and major bleeding associated with extended rivaroxaban treatment in a clinical practice setting among patients with VTE. Methods: A retrospective study was conducted using claims data from February 2011 to April 2015. It included adult patients who initiated rivaroxaban therapy within 7 days after their first VTE and who continuously used rivaroxaban for at least 3 months (index date: end of initial 3-month treatment). Categorized into discontinued and continued cohorts, patients were followed up from the index date until the end of continuous treatment (continued cohort) or end of data or reinitiation of oral anticoagulant therapy (discontinued cohort). Using inverse probability of treatment weights controlling for confounders, adjusted Kaplan-Meier rates of recurrent VTE and major bleeding events were compared. Findings: The analysis showed that, compared with the discontinued cohort (n ¼ 1,536), the continued cohort (n ¼ 5,933) had a significantly lower VTE recurrence rate after an additional 3 months (0.70% vs 1.70%), 6 months (1.41% vs 2.34%), 9 months (1.82% vs 3.01%), and 12 months (1.97% vs 3.01%) of treatment (all, p o 0.05). The difference in the cumulative event rates for major bleeding was not statistically significant. Similar results were obtained
] 2017
in an analysis among patients with VTE receiving rivaroxaban for Z6 months. Implications: Our results suggest that, in clinical practice settings, patients with VTE who continued rivaroxaban therapy after the initial 3- or 6-month treatment period had a significantly lower risk for VTE recurrence without a statistically significant increased risk for major bleeding. (Clin Ther. 2017;]:]]]–]]]) & 2017 Elsevier HS Journals, Inc. All rights reserved. Key words: anticoagulant, extended treatment, recurrent, rivaroxaban, venous thromboembolism.
INTRODUCTION Venous thromboembolism (VTE) is a medical condition that encompasses deep vein thrombosis or pulmonary embolism (PE). First-time VTE affects 100 per 100,000 persons (0.1%) in the United States each year, with the incidence rising substantially to 450 to 600 per 100,000 ( 0.5%) among individuals over the age of 80 years.1 Patients with a history of VTE are at an increased risk for VTE recurrence compared with patients Parts of this article were presented at the 58th Annual Meeting of the American Society of Hematology, December 3–6, 2016, San Diego, Calif. Accepted for publication May 25, 2017. http://dx.doi.org/10.1016/j.clinthera.2017.05.357 0149-2918/$ - see front matter & 2017 Elsevier HS Journals, Inc. All rights reserved.
1
Clinical Therapeutics without prior VTE.1–4 The risk varies with time and is the highest in the first 6 to 12 months after the initial episode.1,5 Patients with cancer as well as those not provoked by transient risk factors have a further increased risk for recurrence.1,5 The importance of anticoagulation therapy in patients with VTE is well recognized, but the optimal treatment duration is uncertain due to an increased risk for bleeding associated with long-term anticoagulation. The American College of Chest Physicians guideline on antithrombotic therapy for VTE disease recommends 3 months of therapy in most patients with VTE, while extended therapy beyond 3 months is reserved for those with a high risk for recurrence— patients with first or second unprovoked VTE and active cancer.6 A recent meta-analysis by Bova et al7 concluded that longer anticoagulation would be beneficial against intermediate risk for recurrence, while Cohen et al8 suggested that direct oral anticoagulants would be an effective alternative to the current standard of care for extended treatment (i.e., warfarin or aspirin). EINSTEIN-Ext (Oral Rivaroxaban for Symptomatic Venous Thromboembolism)9 assessed the benefit of prolonged (additional 6 or 12 months) rivaroxaban treatment in patients with VTE who completed the first 6 to 12 months of anticoagulation with a vitamin K antagonist or rivaroxaban. The study showed that rivaroxaban reduced the risk for recurrent VTE by 82%, with a small increased risk for major bleeding (0.7% in the rivaroxaban cohort vs none in the placebo cohort), suggesting that rivaroxaban may provide a welltolerated and effective approach to extended treatment of VTE. While these findings on extended anticoagulant treatment in the EINSTEIN-Ext study are promising, data assessing the duration of treatment with anticoagulants and associated outcomes in clinical practice are limited. The objective of the present study was to assess the association between extended rivaroxaban treatment and the risk for VTE recurrence and major bleeding in a clinical practice setting among patients with VTE.
PATIENTS AND METHODS Data Source Data dated February 2011 to April 2015 in the Truven Health Analytics MarketScan Databases were used in this study. This Commercial Claims and
2
Encounters database includes claims from employees and their dependents covered by employer-sponsored private health insurance plans from 100 employers and a number of health plans. The Medicare Supplemental and Coordination of Benefits database focuses on patients aged Z65 years with Medicare coverage plus employer-paid commercial plans. The data included in both MarketScan databases are de-identified and are in compliance with the Health Insurance Portability and Accountability Act of 1996 to preserve participant anonymity and confidentiality.
Study Design A retrospective longitudinal study design was used for comparing the risk for recurrent VTE and major bleeding events among patients who continued and patients who discontinued treatment with rivaroxaban. All adult (aged Z18 years) patients initiated on rivaroxaban treatment within 7 days after a first VTE (index VTE) were included in the study. Patients were required to have had data available from at least 12 months before the index date and not to have received anticoagulant therapy prior to the index VTE. VTEs were identified from hospitalization discharge or outpatient/emergency department visit claims with a primary or secondary VTE diagnosis (International Classification of Disease, Ninth Revision–Clinical Modification [ICD-9-CM] codes: deep vein thrombosis, 451.1, 451.2, 453.4, 453.8, 453.9; or PE, 415.1). The study population included patients who received, for at least 3 months, continuous rivaroxaban treatment, defined as rivaroxaban use, without a gap of 430 days, after the first VTE event. The end of the initial 3-month rivaroxaban treatment period following the first VTE event defined the index date. Patients whose continuous treatment ended within the 30-day window after the index date were assigned to the discontinued cohort, while patients whose continuous treatment ended after this 30-day window were assigned to the continued cohort. Figure 1 depicts the study design. A similar design was used for identifying the population treated for 6 months, except that the index date was defined as the end of the 6-month period and this subgroup included only patients treated for at least 6 months. Patients were excluded if they had one or more claims for any anticoagulant agent (ie, warfarin, lowmolecular-weight heparin, rivaroxaban, dabigatran, apixaban, or fondaparinux), a diagnosis of atrial fibrillation Volume ] Number ]
A.A. Khorana et al.
First day of eligibility
First VTE diagnosis
Initiation of rivaroxaban
Index date after 3 or 6 months of rivaroxaban treatment
End of observation
3- or 6-month treatment
12-month baseline period
Observation (follow-up) period
Study period Continuous insurance coverage
Figure 1. Study design scheme.
(ICD-9-CM 427.31), or a major bleeding event10 during the 12-month baseline period; or had a recurrent VTE event between the index VTE and the index date. Baseline characteristics were measured during the 12-month period prior to the index date, including demographic and clinical characteristics, risk factors for VTE (including risk factors from the American College of Chest Physicians Evidence-Based Clinical Practice guideline11), risk factors for bleeding,12 and the Registro Informatizado de Enfermedad TromboEmbólica (RIETE) bleeding risk assessment model score (having a diagnosis of chronic kidney disease was used as the proxy for the abnormal creatinine level).13
Outcomes Measures The primary end point of the study was recurrent VTE, defined as hospitalization with a primary diagnosis of VTE. Major bleeding events were identified based on primary ICD-9-CM diagnostic codes during hospitalization using a validated algorithm developed by Cunningham et al.10
Statistical Analysis Descriptive analyses were conducted for comparing baseline demographic and clinical characteristics between cohorts. Mean (SD) and median values were reported for continuous variables; frequency and percentage were reported for categorical variables. Statistical differences between cohorts were assessed using standardized differences. A standardized difference of o10% was considered an indication of wellbalanced distribution of the given variable.14 The inverse probability of treatment weight (IPTW), based on a propensity score method, was used for assessing the relationship between the continued (vs
] 2017
discontinued) treatment and outcomes in each population. Weighting patients by the inverse probability of the treatment received creates a synthetic sample independent of baseline coviariates.15 Each patient was assigned a weight such that in the weighted pseudopopulation, the distribution of measured confounders was similar between the continued and discontinued cohorts. In order to calculate IPTWs, the probability (ie, the propensity score) of receiving continued treatment (vs discontinued treatment) was first estimated using a multivariate logistic regression model conditional on baseline covariates, including age, sex, region of residence, insurance type, year of index date, type of first VTE, time from first VTE to first rivaroxaban dispensing, baseline risk factors for bleeding and VTE,11,12 Quan-Charlson comorbidity index, RIETE bleeding score, and health care resource utilization and costs. IPTWs were then calculated as the inverse of patients’ estimated probabilities of having their observed treatment; that is, IPTWs were calculated as 1/(Propensity score) in the continued cohort and as 1/(1 – [Propensity score]) in the discontinued cohort. Finally, the IPTWs were normalized within each cohort by dividing each IPTW by the mean of the IPTWs in each cohort. Kaplan-Meier curves for time to first recurrent VTE and time to first major bleeding event from the index date were compared between cohorts in the IPTWweighted populations. Kaplan-Meier rates for VTE recurrences and major bleeding events at 3, 6, 9, and 12 months after the index date were compared between cohorts using log-rank tests. All of the analyses in the study were performed with SAS software version 9.4 (SAS Institute, Cary, North Carolina).
RESULTS Patients’ Baseline Characteristics A total of 7,469 patients who received rivaroxaban for at least 3 months after the initial VTE event were included in the 3-month analysis; 5,933 and 1,536 patients formed the continued and discontinued cohorts, respectively. For the 6-month analysis, 2,676 patients were included in the continued cohort and 1,127 in the discontinued cohort. Between-cohort comparisons of patients’ baseline characteristics and baseline risk factors for bleeding and VTE are presented in Tables I and II, respectively. After weighting, the continued and discontinued cohorts were wellbalanced across baseline characteristics and risk factors.
3
At Least 3 mo of Therapy Characteristic
Volume ] Number ]
Duration, mean (SD) [median], d Observation period† Total treatment (including 3 mo of rivaroxaban therapy)‡ Propensity score variables Demographics Age, mean (SD) [median], y Female, no. (%) Insurance type, no. (%) Consumer-directed Exclusive provider organization Comprehensive Health maintenance organization High-deductible Preferred provider organization Point-of-service Point-of-service capitated Not specified
At Least 6 mo of Therapy
Discontinued Cohort (n ¼ 1,536)
Continued Cohort (n ¼ 5,933)
Standardized Discontinued Cohort Difference, % (n ¼ 1,127)
211.1 (191.6) [162] 105.7 (8.5) [107]
149.3 (124.4) [108] 242.0 (126.2) [199]
– –
206.5 (171.5) [158] 196.0 (7.9) [197]
57.1 (15.4) [58]
56.8 (14.7) [57]
2.1
57.7 (14.3) [58]
Continued Cohort (n ¼ 2,676)
Standardized Difference, %
158.5 (130.6 ) [114] 341.3 (132.0 ) [297]
– –
57.8 (14.2) [58]
0.3
701 (45.6)
2,735 (46.1)
0.9
487 (43.2)
1,152 (43.1)
0.4
892 (58.1) 212 (13.8)
3,432 (57.9) 805 (13.6)
0.4 0.6
650 (57.6) 163 (14.5)
1,534 (57.3) 385 (14.4)
0.6 0.3
125 (8.1) 118 (7.7)
467 (7.9) 459 (7.7)
1.0 0.3
95 (8.4) 87 (7.7)
208 (7.8) 215 (8.0)
2.3 1.0
92 (6.0) 70 (4.6)
371 (6.2) 279 (4.7)
1.2 0.7
65 (5.8) 45 (4.0)
163 (6.1) 115 (4.3)
1.3 1.6
5 (0.3) 3 (0.2)
22 (0.4) 24 (0.4)
0.7 4.1
4 (0.4) 8 (0.7)
10 (0.4) 8 (0.3)
0.1 5.4
21 (1.3)
73 (1.2)
1.1
10 (0.9)
38 (1.4)
5.0 (continued)
Clinical Therapeutics
4 Table I. Demographic and clinical characteristics assessed during the 12-month baseline period, weighted population.*
] 2017
Table I. (continued). At Least 3 mo of Therapy Characteristic
6 412 866 252 705
(0.4) (26.9) (56.4) (16.4) (45.9)
Continued Cohort (n ¼ 5,933)
(0.4) (26.4) (57.0) (16.3) (46.3)
0.1 1.1 1.2 0.2 0.7
3 205 714 205 586
807 (52.5)
3,171 (53.4)
1.8
442 (28.7)
1,639 (27.6)
287 (18.7)
1,123 (18.9)
0.69 (1.18) [0] 949 365 101 54 67
(61.8) (23.8) (6.6) (3.5) (4.4)
22 1,564 3,381 967 2,746
Standardized Discontinued Cohort Difference, % (n ¼ 1,127)
0.68 (1.17) [0] 3,674 1,435 364 201 258
(61.9) (24.2) (6.1) (3.4) (4.4)
(0.2) (18.2) (63.4) (18.2) (52.0)
Continued Cohort (n ¼ 2,676)
(0.2) (18.7) (62.4) (18.7) (50.7)
0.1 1.3 2.0 1.1 2.5
539 (47.8)
1,307 (48.8)
2.0
2.5
311 (27.6)
748 (27.9)
0.9
0.6
277 (24.6)
622 (23.2)
3.2
0.68 (1.20) [0]
2.6
0.8
0.65 (1.14) [0]
0.3 1.0 1.7 0.7 0.1
710 264 64 44 46
(63.0) (23.4) (5.7) (3.9) (4.1)
6 500 1,670 500 1,358
Standardized Difference, %
1,667 629 159 104 117
(62.3) (23.5) (5.9) (3.9) (4.4)
1.4 0.3 1.0 0.0 1.7 (continued)
5
A.A. Khorana et al.
Year of index date, no. (%) 2012 2013 2014 2015 First VTE diagnosis identified in-hospital, no. (%) Type of first VTE diagnosis, no. (%) Deep vein thrombosis Pulmonary embolism Both diagnoses on the same day Time from first VTE to first rivaroxaban dispensing Mean (SD) [median], d Same day, no. (%) 1 d, no. (%) 2 d, no. (%) 3 d, no. (%) 4–7 d, no. (%)
Discontinued Cohort (n ¼ 1,536)
At Least 6 mo of Therapy
At Least 3 mo of Therapy Characteristic
Discontinued Cohort (n ¼ 1,536)
Continued Cohort (n ¼ 5,933)
Volume ] Number ]
Comorbidity index scores Quan-Charlson Mean (SD) [median] 1.66 (2.41) [1] 1.57 (2.30) [1] Comorbidities, no. (%) 0 650 (42.3) 2,581 (43.5) 1 337 (21.9) 1,337 (22.5) 2 209 (13.6) 759 (12.8) Z3 340 (22.1) 1,256 (21.2) RIETE Mean (SD) [median] 1.43 (1.35) [1] 1.38 (1.35) [1] Comorbidities, no. (%) 0 467 (30.4) 1,847 (31.1) 1–4 999 (65.1) 3,814 (64.3) 44 69 (4.5) 272 (4.6) Baseline health care resource utilization, mean (SD) [median] Outpatient visits 20.62 (16.96) [16] 20.11 (16.66) [15] ED visits 1.11 (1.59) [1] 1.06 (1.48) [1] Hospitalizations 0.71 (0.74) [1] 0.71 (0.74) [1] Baseline health care costs, mean (SD), year-2015 $US Total health care 37,207 (74,562) 36,028 (57,866) costs Hospitalizations 16,448 (41,767) 16,166 (41,435)
At Least 6 mo of Therapy Standardized Discontinued Cohort Difference, % (n ¼ 1,127)
3.7
1.64 (2.22) [1]
2.4 1.5 2.4 2.4
449 257 168 252
3.8
1.47 (1.35) [1]
1.5 1.6 0.4
309 (27.5) 758 (67.3) 59 (5.2)
3.0 3.4 0.0
(39.8) (22.8) (14.9) (22.4)
21.15 (16.20) [17] 1.01 (1.46) [1] 0.75 (0.79) [1]
Continued Cohort (n ¼ 2,676)
1.66 (2.33) [1] 1,088 619 372 597
Standardized Difference, %
0.8
(40.7) (23.1) (13.9) (22.3)
1.7 0.6 2.9 0.2
1.47 (1.38) [1]
0.2
747 (27.9) 1,787 (66.8) 142 (5.3)
1.0 1.1 0.3
21.13 (17.39) [16] 1.03 (1.59) [1] 0.73 (0.71) [1]
0.1 1.4 2.8
1.8
48,597 (93,979)
38,068 (61,852)
13.2
0.7
16,432 (32,119)
15,969 (42,297)
1.2 (continued)
Clinical Therapeutics
6 Table I. (continued).
] 2017
ED ¼ emergency department; RIETE ¼ Registro Informatizado de Enfermedad TromboEmbólica; VTE ¼ venous thromboembolism. * The propensity score of receiving continued treatment (vs discontinued treatment) was estimated using a multivariate logistic regression model conditional on baseline covariates including age, sex, region, insurance type, year of index date, episode type of first VTE, type of first VTE, time from first VTE to first rivaroxaban dispensing, baseline risk factors for bleeding and VTE, Quan-Charlson comorbidity index, RIETE score, and health care resource utilization and costs. † From the index date to the earliest date between initiation of a new anticoagulant therapy, end of data availability (April 2015), or end of insurance coverage as well as the end of the rivaroxaban therapy for the continued cohort. ‡ Duration of treatment was defined as the duration of continuous use of rivaroxaban from the first dispensing until a 30-day interruption or end of follow-up.
15.5 5.3 0.9 12,961 (30,870) 6,067 (10,277) 2,244 (4,880) 22,435 (80,858) 6,711 (13,869) 2,203 (4,468) 2.2 0.6 2.0 12,026 (27,875) 4,767 (9,342) 2,195 (4,370) 12,736 (36,978) 4,824 (9,042) 2,287 (4,841) Outpatient visits Pharmacy ED visits
Continued Cohort (n ¼ 2,676) Standardized Discontinued Cohort Difference, % (n ¼ 1,127) Continued Cohort (n ¼ 5,933) Discontinued Cohort (n ¼ 1,536) Characteristic
Table I. (continued).
At Least 3 mo of Therapy
At Least 6 mo of Therapy
Standardized Difference, %
A.A. Khorana et al.
VTE Recurrence In the weighted population, 89 recurrent VTEs were identified after 12 months of follow-up. The continued cohort had significantly lower rates of recurrent VTEs than did the discontinued cohort at 3 months (0.70% vs 1.70%; P = 0.003), 6 months (1.41% vs 2.34%; P = 0.016), and 12 months (1.97% vs 3.01%; P = 0.017) of follow-up (Figure 2A). Comparisons also showed that, among the 3803 patients who received rivaroxaban treatment for at least 6 months, continued treatment was associated with a reduced risk for recurrent VTEs (Figure 2B).
Major Bleeding Events In the weighted population, 57 major bleeding events were identified at 12 months of follow-up. The rates of major bleeding (gastrointestinal or other bleeding) in the continued and the discontinued cohorts, respectively, were 0.58% and 0.82% at 3 months (P ¼ 0.367), 0.91% and 0.88% at 6 months (P ¼ 0.754), and 1.44% and 1.44% at 12 months (P ¼ 0.813) (Figure 3A). Results in patients treated for at least 6 months were consistent with those treated at least 3 months (Figure 3B).
DISCUSSION The present study, using claims data from clinical practice, assessed the impact of extended treatment with rivaroxaban on the risks for recurrent VTE and major bleeding. Continuing therapy for 43 months was associated with a 35% relative risk reduction of recurrent VTE at 12 months (rate of recurrent VTE, 1.97% vs 3.01%), without a significant difference in the risk for major bleeding. In patients who received rivaroxaban treatment for at least 6 months, continuing therapy was associated with a 54% relative risk reduction in VTE recurrence (rate of recurrent VTE, 1.72% vs 3.70%). The results from the present study are in line with results from the EINSTEIN-Ext randomized clinical trial, in which extended treatment with rivaroxaban was associated with a lower risk for VTE recurrence without any significant impact on the risk for bleeding. The mean age in EINSTEIN-Ext was 58 years, 38% of patients had a PE, 73% had an unprovoked event, and 4% to 5% had active cancer. In the present study, the mean age was 57 years, 48% had an initial diagnosis of PE, 67% had an unprovoked event, and
7
At least 3 months of therapy Characteristics, n (%)
Volume ] Number ]
VTE and bleeding Hypertension Diabetes Cancer Cerebrovascular accident (stroke) VTE Multiple trauma Hyperlipidemia Other serious infections Obesity Abdominal surgery Arrhythmia Major surgery Pneumonia Contraceptive pill (use of oral) Hip, pelvis, or leg fracture Varicose veins Congestive heart failure COPD Thrombophilia Myocardial infarction Pregnancy Total knee replacement Left ventricular dysfunction Rheumatoid arthritis Inflammatory bowel disease Treatment with aromatase inhibitors Surgical resection of abdominal or pelvic cancer Total hip replacement Treatment with SERMs
Discontinued Cohort (N ¼ 1,536)
Continued Cohort (N ¼ 5,933)
At least 6 months of therapy Standardized Difference
Discontinued Cohort (N ¼ 1,127)
Continued Cohort (N ¼ 2,676)
Standardized Difference
808 290 267 60
(52.6%) (18.9%) (17.4%) (3.9%)
3,044 1,101 956 206
(51.3%) (18.6%) (16.1%) (3.5%)
2.6% 0.8% 3.4% 2.3%
603 226 205 48
(53.5%) (20.0%) (18.2%) (4.3%)
1,443 551 442 98
(53.9%) (20.6%) (16.5%) (3.7%)
0.8% 1.4% 4.5% 3.1%
540 664 306 277 211 200 190 177 98
(35.2%) (43.2%) (19.9%) (18.0%) (13.8%) (13.0%) (12.4%) (11.5%) (6.4%)
2,082 2,523 1,193 1,074 780 750 692 680 406
(35.1%) (42.5%) (20.1%) (18.1%) (13.2%) (12.6%) (11.7%) (11.5%) (6.8%)
0.1% 1.4% 0.5% 0.2% 1.8% 1.2% 2.2% 0.2% 1.8%
363 494 212 233 151 157 125 138 65
(32.2%) (43.8%) (18.8%) (20.6%) (13.4%) (14.0%) (11.1%) (12.3%) (5.8%)
869 1,203 518 511 363 364 306 312 159
(32.5%) (45.0%) (19.4%) (19.1%) (13.6%) (13.6%) (11.4%) (11.6%) (5.9%)
0.4% 2.3% 1.3% 3.9% 0.5% 1.1% 1.1% 2.0% 0.6%
89 75 111 72 63 61 24 19 28 32 23 19
(5.8%) (4.9%) (7.3%) (4.7%) (4.1%) (4.0%) (1.5%) (1.2%) (1.8%) (2.1%) (1.5%) (1.3%)
(5.6%) (5.1%) (6.8%) (4.6%) (4.1%) (3.7%) (1.7%) (1.3%) (2.0%) (2.1%) (1.6%) (1.2%)
0.9% 0.9% 1.8% 0.3% 0.0% 1.5% 1.0% 0.0% 1.5% 0.2% 0.7% 1.0%
47 63 80 47 74 53 35 8 23 25 16 11
(4.1%) (5.6%) (7.1%) (4.2%) (6.5%) (4.7%) (3.1%) (0.7%) (2.1%) (2.2%) (1.4%) (0.9%)
(4.2%) (5.8%) (6.9%) (4.4%) (5.3%) (3.9%) (1.6%) (0.7%) (2.0%) (2.2%) (1.5%) (1.2%)
0.3% 0.9% 0.6% 0.8% 5.4% 3.6% 10.2% 0.0% 0.1% 0.3% 0.5% 2.5%
18 (1.2%)
80 (1.4%)
1.3%
14 (1.3%)
33 (1.2%)
0.2%
13 (0.9%) 7 (0.5%) 8 (0.5%)
38 (0.6%) 35 (0.6%) 27 (0.4%)
2.4% 1.6% 1.4%
8 (0.7%) 8 (0.7%) 5 (0.4%)
17 (0.6%) 17 (0.6%) 13 (0.5%)
0.9% 1.2% 0.8%
331 302 403 274 245 220 99 74 120 126 95 68
112 156 185 117 141 105 42 19 55 60 39 32
(continued)
Clinical Therapeutics
8
Table II. Risk Factors assessed during the 12-month baseline period, weighted population*.
] 2017
Table II. (continued). At least 3 months of therapy Characteristics, n (%) Treatment with erythropoiesis stimulating agents Spinal cord injury Immobility Bleeding NSAID use Anemia Renal disease Excessive fall risk (Parkinson’s disease, etc.) Chronic kidney disease Hepatic disease Coagulation defect Central venous catheter Thrombocytopenia (low platelet count) Ethanol abuse Peptic ulcer Bleeding diathesis
Discontinued Cohort (N ¼ 1,536)
6 (0.4%) 2 (0.1%) 456 283 279 247
(29.7%) (18.4%) (18.2%) (16.1%)
110 111 84 37 50
(7.2%) (7.2%) (5.5%) (2.4%) (3.3%)
24 (1.5%) 16 (1.0%) 0 (0.0%)
Continued Cohort (N ¼ 5,933)
18 (0.3%) 8 (0.1%) 1,764 1,070 1,045 949
At least 6 months of therapy Standardized Difference
Discontinued Cohort (N ¼ 1,127)
1.3% 0.9%
2 (0.2%) 0 (0.0%)
Continued Cohort (N ¼ 2,676)
5 (0.2%) 2 (0.1%)
Standardized Difference
0.6% 3.9%
(29.7%) (18.0%) (17.6%) (16.0%)
0.1% 1.0% 1.4% 0.3%
292 204 185 174
(25.9%) (18.1%) (16.5%) (15.4%)
699 492 462 421
(26.1%) (18.4%) (17.3%) (15.7%)
0.5% 0.6% 2.2% 0.8%
(6.6%) (7.0%) (5.3%) (2.4%) (3.1%)
2.2% 0.9% 0.8% 0.2% 0.7%
75 82 86 22 44
(6.6%) (7.3%) (7.6%) (2.0%) (3.9%)
183 206 175 56 96
(6.8%) (7.7%) (6.5%) (2.1%) (3.6%)
0.7% 1.4% 4.3% 0.7% 1.8%
95 (1.6%) 47 (0.8%) 3 (0.0%)
0.4% 2.6% 3.1%
15 (1.3%) 9 (0.8%) 2 (0.1%)
38 (1.4%) 20 (0.8%) 1 (0.0%)
1.1% 0.3% 3.5%
393 415 314 144 186
Abbreviations: VTE: venous thromboembolism; NSAID: nonsteroidal anti-inflammatory drugs; COPD: chronic obstructive pulmonary disease; SERM: selective estrogen receptor modulator Notes: ⁎ The propensity score of receiving continued treatment (vs. discontinued treatment) was estimated using a multivariate logistic regression model conditional on baseline covariates including age, gender, region, insurance type, year of index date, episode type of first VTE, type of first VTE, time from first VTE to first rivaroxaban dispensing, baseline risk factors for bleeding and VTE, Quan-Charlson comorbidity index, RIETE score, and healthcare resource utilization and costs.
A.A. Khorana et al.
9
Clinical Therapeutics
% of Patients with a Recurrent VTE
A
8%
After 3 months of therapy
Discontinued Continued
Log -rank P-value = 0.017
6%
Log -rank P-value = 0.012 3-month period of initial rivaroxaban treatment
4%
Log -rank P -value = 0.016 Log -rank P -value = 0.003 1.70%
3.01% 2.34%
1.82%
3.01% 1.97%
1.41%
0.70% 2%
0% Discontinued Continued
% of Patients with a Recurrent VTE
B
0 Days Number at risk 1,536 Number of events Number at risk 5,933 Number of events -
8%
90 964 20 3,645 37
180 699 26 1,487 53
270 471 30 806 58
After 6 months of therapy
Discontinued Continued
Log -rank P -value = 0.024
6%
4%
365 305 30 417 59
Log -rank P-value = 0.023 6-month period of initial rivaroxaban treatment
3.70%
Log -rank P-value = 0.023 Log-rank P-value = 0.133 1.41%
3.02% 2.69%
1.72%
1.35%
1.22%
0.82%
2%
0% Discontinued Continued
0 Days Number at risk 1,127 Number of events Number at risk 2,676 Number of events -
90 748 14 1,567 17
180 511 22 844 22
270 347 23 445 23
365 221 25 243 25
Figure 2. Inverse probability of treatment weight–adjusted Kaplan-Meier rates of recurrent venous thromboembolism (VTE) after 3 (A) and 6 (B) months of treatment among rivaroxaban users who continued versus discontinued therapy.
12% had a diagnosis of cancer in the year prior to the index date. While both studies found a significant impact of extended therapy on the reduction in the risk for recurrence, the magnitude of the reduction in the 12-month rate of VTE recurrence was higher in EINSTEIN-Ext—7.1% versus 1.3% for rivaroxaban versus placebo. The smaller effect observed in our study may have been related to potential underestimation of the number of recurrent VTEs: Our study captured only recurrent VTEs diagnosed during hospitalization and not in the outpatient setting. However, because patients with PE are more likely to be diagnosed and treated in hospitals, our approach was
10
less likely to have underestimated PE recurrence. The findings from the 6-month sensitivity analysis are consistent with those from the EINSTEIN-Ext study and others, which have shown that extended anticoagulation treatment had efficacy superior (i.e., recurrent VTE) to that of placebo in patients with VTE who had received 6 to 12 months of prior anticoagulant treatment.8,9,16,17 No statistically significant differences in the risk for major bleeding were observed between the continued and discontinued cohorts in either the 3- or 6-month analysis. The bleeding rates at 12 months were similar between the continued and discontinued cohorts.
Volume ] Number ]
A.A. Khorana et al.
A
8%
After 3 months of therapy
% of Patients with Major Bleeding
Discontinued Continued Log-rank P-value = 0.813
6%
4%
3-month period of initial rivaroxaban treatment
Log-rank P-value = 0.870 Log-rank P-value = 0.754 Log-rank P-value = 0.367
2%
0.82%
0.88%
0.58%
0.91%
90 964 10 3,657 30
180 699 11 1,497 37
1.18%
1.44%
1.33%
1.44%
0%
Discontinued Continued
% of Patients with Major Bleeding
B
8%
0 Days Number at risk 1,536 Number of events Number at risk 5,933 Number of events -
270 466 13 807 42
365 299 14 416 43
After 6 months of therapy
Discontinued Continued
6%
4%
Log-rank P-value = 0.794 6-month period of initial rivaroxaban treatment
Log-rank P-value = 0.544 Log-rank P-value = 0.854 1.32%
Log-rank P-value = 0.903 0.80% 0.37%
2%
1.32% 0.82%
1.53%
0.72%
0.38% 0%
Discontinued Continued
0 Days Number at risk 1,127 Number of events Number at risk 2,676 Number of events -
90 750 4 1,573 8
180 511 6 845 12
270 348 9 444 13
365 222 9 241 15
Figure 3. Inverse probability of treatment weight–adjusted Kaplan-Meier rates of major bleeding events after 3 (A) and 6 (B) months of treatment among rivaroxaban users who continued versus discontinued therapy. Although it was expected that patients in the continued cohort would have a higher risk for bleeding because the continued cohort remained on bloodthinning medications longer than did the other cohort, our study was underpowered for detecting the association between a major bleeding event with extended treatment because of the short duration of follow-up (up to 12 months). Results from clinical trials have shown that anticoagulation treatment for at least 3 months reduces the risk for VTE recurrence.18 The American College of Chest Physicians suggests extended anticoagulant therapy beyond 3 months among patients with a first
] 2017
unprovoked VTE and who have a low or moderate bleeding risk.19 A systematic review conducted by Kearon and Akl20 found that indefinite anticoagulation (or extended, referring to continued treatment without a scheduled stopping date) is often chosen if there is a low risk for bleeding, whereas anticoagulation is usually stopped at 3 months if there is a high risk for bleeding. Studies included in that systematic review showed that indefinite anticoagulation with a vitamin K antagonist was associated with a reduced risk for recurrent VTE.20 Newer oral anticoagulants have shown efficacy in the prevention of recurrent VTE. The results from 10 Phase III studies demonstrated that rivaroxaban and
11
Clinical Therapeutics dabigatran (1) provided effective and convenient shortterm treatment for VTE compared with the standard of care and (2) showed superiority in the long-term prevention of recurrent VTE compared with placebo.21 The current study was subject to certain limitations. First, billing inaccuracies and missing data may have occurred. It is possible that VTE events were under-coded (ie, false-negatives). Second, recurrent VTE in this study was determined based on only the primary diagnosis during hospitalization. The use of this strict definition on the outcome warrants that the "true" VTE events were captured, while it likely led to underestimation of the number of recurrent VTE events because VTE may have been treated in outpatient settings. Finally, as in all observational studies, adjustments in the multivariate analyses could account for only observable factors.
CONCLUSIONS Our study results suggest that patients with VTE who continued rivaroxaban therapy after the first 3- or 6-month treatment period had a significantly lower risk for VTE recurrence without an increased risk for major bleeding. These results suggest that gains in efficacy observed in extended-therapy trials might translate to clinical practice and provide support for extended therapy to clinicians and patients.
ACKNOWLEDGMENTS Dr. Khorana acknowledges research support from the Sondra and Stephen Hardis Chair in Oncology Research.
CONFLICTS OF INTEREST This research was funded by Janssen Scientific Affairs, LLC (Titusville, New Jersey). The authors are employees of (V.A., C.C., J.S. and P.W.) or have received research funding (A.K.), consultant’s fees (A.K., R.S., J.B., P.W., S.K.), or honoraria (A.K., S.K.) from Janssen. F.L., D.L., and P.L. are employees of Groupe d’analyse, Ltée, a consulting company that has received research grants from Janssen. A.K. has also received research funding from Sanofi and Amgen; and consultant’s fees and honoraria from Leo, Sanofi, Amgen, Bayer, Halozyme, Pfizer, and Roche. J.B. has received research funding from AstraZeneca and has been a member of the scientific
12
advisory board for Merck. P.W. has received research funding from Bristol-Myers Squibb/Pfizer; has been a member of the advisory board of Bayer Health Care; and has served on a writing committee for Itreas. C.C., J.S., and P.W. own stock options in Johnson & Johnson. S.K. has received consultant’s fees from Boehringer Ingelheim, Daiichi Sankyo, Bristol-Myers Squibb, Portola, and Pfizer; and honoraria from Boehringer Ingelheim, Bristol-Myers Squibb, Pfizer, Daiichi Sankyo, and CSL Behring. The authors have indicated that they have no other conflicts of interest with regard to the content of this article.
REFERENCES 1. White RH. The epidemiology of venous thromboembolism. Circulation. 2003;107(Suppl 1):I4–I8. 2. Hansson PO, Sörbo J, Eriksson H. Recurrent venous thromboembolism after deep vein thrombosis: incidence and risk factors. Arch Intern Med. 2000;160:769–774. 3. Kearon C. Natural history of venous thromboembolism. Circulation. 2003;107(Suppl 1):I22–I30. 4. Zhu T, Martinez I, Emmerich J. Venous thromboembolism: risk factors for recurrence. Arterioscler Thromb Vasc Biol. 2009;29:298–310. 5. Heit JA, Spencer FA, White RH. The epidemiology of venous thromboembolism. J Thromb Thrombolysis. 2016;41: 3–14. 6. Kearon C, Akl EA, Ornelas J, et al. Antithrombotic therapy for VTE disease. Chest. 2016;149:315–352. 7. Bova C, Bianco A, Mascaro V, Nobile CG. Extended anticoagulation and mortality in venous thromboembolism. a meta-analysis of six randomized trials. Thromb Res. 2016;139:22–28. 8. Cohen AT, Hamilton M, Bird A, et al. Comparison of the non-VKA oral anticoagulants apixaban, dabigatran, and rivaroxaban in the extended treatment and prevention of venous thromboembolism: systematic review and network meta-analysis. PLoS ONE. 2016;11:e0160064. 9. The EINSTEIN Investigators. Oral rivaroxaban for symptomatic venous thromboembolism. N Engl J Med. 2010; 363:2499–2510. 10. Cunningham A, Stein CM, Chung CP, et al. An automated database case definition for serious bleeding related to oral anticoagulant use. Pharmacoepidemiol Drug Saf. 2011;20: 560–566. 11. Kahn SR, Lim W, Dunn AS, et al. Prevention of VTE in nonsurgical patients. Chest. 2012;141:e195S–e226S. 12. Laliberté F, Cloutier M, Nelson WW, et al. Real-world comparative effectiveness and safety of rivaroxaban and
Volume ] Number ]
A.A. Khorana et al.
13.
14.
15.
16.
17.
18.
19.
20.
21.
warfarin in nonvalvular atrial fibrillation patients. Curr Med Res Opin. 2014;30:1317–1325. Ruíz-Giménez N, Suárez C, González R, 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. Austin P. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38:1228–1234. Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med. 2015;34:3661–3679. Romualdi E, Donadini M, Ageno W. Oral rivaroxaban after symptomatic venous thromboembolism: the continued treatment study (EINSTEINextension study). N Engl J Med. 2010;363:2499–2510. Cohen AT, Dobromirski M. The use of rivaroxaban for short- and longterm treatment of venous thromboembolism. Thromb Haemost. 2012; 107:1035–1043. Smilowitz N, Mega J, Berger J. Duration of anticoagulation for venous thromboembolic events. Circulation. 2014;130:2343–2348. Kearon C, Akl EA, Comerota AJ, et al. Antithrombotic therapy for VTE disease: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest. 2012;141(2 Suppl):e419S–e4194S. Kearon C, Akl EA. Duration of anticoagulant therapy for deep vein thrombosis and pulmonary embolism. Blood. 2014;123:1794–1801. Cohen AT. Long-term benefits of preventing venous thromboembolic events. Curr Med Res Opin. 2012;28:877–889.
] 2017
Address correspondence to: Dominique Lejeune, MSc, Groupe d’analyse, Ltée, 1000, rue de La Gauchetière Ouest, Bureau 1200, Montréal H3B 4W5, Québec, Canada. E-mail: [email protected]
13