- Email: [email protected]
Safety and efficacy of rivaroxaban compared with warfarin in patients undergoing peripheral arterial procedures
From the Midwestern Vascular Surgical Society
Safety and efficacy of rivaroxaban compared with warfarin in patients undergoing peripheral arterial procedures Anjan Talukdar, MD,a S. Keisin Wang, MD,a Lauren Czosnowski, PharmD,a,b Nassim Mokraoui, MD,a Alok Gupta, MD,a Andres Fajardo, MD,a Michael Dalsing, MD,a and Raghu Motaganahalli, MD,a Indianapolis, Ind
ABSTRACT Objective: Rivaroxaban is a United States Food and Drug Administration-approved oral anticoagulant for venous thromboembolic disease; however, there is no information regarding the safety and its efficacy to support its use in patients after open or endovascular arterial interventions. We report the safety and efficacy of rivaroxaban vs warfarin in patients undergoing peripheral arterial interventions. Methods: This single-institution retrospective study analyzed all sequential patients from December 2012 to August 2014 (21 months) who were prescribed rivaroxaban or warfarin after a peripheral arterial procedure. Our study population was then compared using American College of Chest Physicians guidelines with patients then stratified as low, medium, or high risk for bleeding complications. Statistical analyses were performed using the Student t-test and c2 test to compare demographics, readmissions because of bleeding, and the need for secondary interventions. Logistic regression models were used for analysis of variables associated with bleeding complications and secondary interventions. The Fisher exact test was used for power analysis. Results: There were 44 patients in the rivaroxaban group and 50 patients in the warfarin group. Differences between demographics and risk factors for bleeding between groups or reintervention rate were not statistically significant (P ¼ .297). However, subgroup evaluation of the safety profile suggests that patients who were aged #65 years and on warfarin had an overall higher incidence of major bleeding (P ¼ .020). Patients who were aged >65 years, undergoing open operation, had a significant risk for reintervention (P ¼ .047) when they received rivaroxaban. Conclusions: Real-world experience using rivaroxaban and warfarin in patients after peripheral arterial procedures suggests a comparable safety and efficacy profile. Subgroup analysis of those requiring an open operation demonstrated a decreased bleeding risk when rivaroxaban was used (in those aged <65 years) but an increased risk for secondary interventions. Further studies with a larger cohort are required to validate our results. (J Vasc Surg 2017;-:1-6.)
Rivaroxaban, an oral direct anti-Xa agent, was approved by the United States Food and Drug Administration in 2011 for use in the United States. It is currently used for prevention of venous thromboembolism, deep venous thrombosis, stroke prevention in patients with nonvalvular atrial fibrillation, and secondary prevention of coronary events in patients with acute coronary syndrome.19 There are no data available to support its use in patients with peripheral arterial occlusive disease after surgical or From the Division of Vascular Surgerya and Department of Surgery,b Indiana University School of Medicine. Author conflict of interest: none. Presented at the Thirty-ninth Annual Meeting of the Midwestern Vascular Surgical Society, Chicago, Ill, September 10-12, 2015. Correspondence: Raghu Motaganahalli, MD, Division of Vascular Surgery, 1801 N Senate Blvd, MPC-2 #3500, Indianapolis, IN 46202 (e-mail: rmotagan@iupui. edu). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2017 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2017.02.052
endovascular intervention. Warfarin, an oral vitamin K antagonist, has been in use in general since 1954 as well as in patients deemed high risk for failure after peripheral arterial intervention. This benefit is offset by food and drug interactions as well as a bleeding risk, which is best controlled with frequent coagulation monitoring and dose adjustments as needed.8,10-13 Efficacy and safety of these drugs are compared in the ROCKET AF trial14 (The Rivaroxaban once daily oral Direct factor Xa inhibition compared with Vitamin K Antagonism for prevention of Stroke and Embolism Trial in Atrial Fibrillation) for recurrent stroke prevention in patients with nonvalvular atrial fibrillation. This randomized trial found that rivaroxaban was noninferior to warfarin in the primary analysis, which included patients in the per-protocol population, and also in the intention-totreat analysis. The primary safety analysis found no significant difference between rivaroxaban and warfarin with respect to rates of major or no major clinically relevant bleeding. The objective of our study was to compare the safety and efficacy of rivaroxaban vs warfarin in patients 1
2
Journal of Vascular Surgery
Talukdar et al
---
undergoing peripheral arterial procedures who are deemed of sufficiently high risk of failure to warrant anticoagulation.
ARTICLE HIGHLIGHTS d
d
METHODS The study was approved by the Indiana University Institutional Review Board as an exempt study, and hence, no patient consent was obtained. This single-center retrospective study included all sequential patients from December 2012 to August 2014 who were prescribed rivaroxaban after an open or endovascular intervention for lower extremity arterial occlusive disease or acute embolic occlusion. Comparisons were made with matched sequential patients who were prescribed warfarin during the same time for similar indications. Patients were identified from a database with Pharmacy, which kept a log of patients who were prescribed rivaroxaban or warfarin. Patients were prescribed warfarin or rivaroxaban according to physician preference, potential undesirable drug interactions, history of difficulty maintaining a therapeutic anticoagulation profile with warfarin, concern for hypercoagulable state, use of a nonautogenous conduit with poor runoff, and presence of a proximal embolic source. Patient demographics (Table I), readmission with cause, and need for secondary interventions were identified from inpatient record reviews. The American College of Chest Physicians (ACCP) guidelines were used to stratify patients into low, medium, or high risk for bleeding complications.15 The patient demographics that were recorded included age, any previous bleeding episodes, history of cancer, either primary or metastatic disease, including hematologic, history of chronic renal insufficiency or chronic liver disease, thrombocytopenia (<150,000/mm3), previous stroke, diabetes, anemia (<13.4 g/dL in men and <12 g/dL in women), use of antiplatelet (aspirin or clopidgrel), poor anticoagulant control, failure to thrive, recent surgery, frequent falls, and alcohol abuse. Low-risk patients had no risk factors, medium-risk patients had one risk factor, and high-risk patients had two or more risk factors. Major bleeding complications were defined as any bleeding requiring hospitalization or transfusion. Secondary interventions were defined as any intervention performed after the index operation to maintain graft/ stent patency. Interventions included open, endovascular, or hybrid procedures. Patients receiving rivaroxaban or warfarin were compared by age (#65 or >65 years), risk of bleeding, and type of primary and secondary interventions. Patients who required reintervention were managed case-by-case. Decision for an urgent intervention was dictated by an imminent threat to an extremity rather than the use of an anticoagulant. For patients requiring emergency intervention, we used either or a combination of reversal agents (vitamin K, fresh frozen plasma for warfarin) and blood transfusions. Agents to
2017
d
Type of Research: Retrospective case control study Take Home Message: In 94 patients undergoing peripheral arterial procedures rivaroxaban was equally safe and effective compared with warfarin but carried less risk of bleeding complications. Recommendation: The data suggest that in patients who require anticoagulation after lower extremity open or endovascular revascularization, rivaroxaban is preferred over warfarin.
Table I. Patient demographics and risk for bleeding Variablea
Rivaroxaban (n ¼ 44)
Warfarin (n ¼ 44)
P value
Age, years
60.5 6 15
63.8 6 14
.278
Male
27
32
Female
17
18
Gender
.833
Race Caucasian
.541 41
44
African American
3
5
Other
0
1
0
0
Risk category Low Medium High Risk factors, No.
.297 6
3
38
47
2.4 6 0.8
2.7 6 0.9
Intervention type Endovascular Open Hybrid
.199 .312
7
11
36
35
1
4
Antiplatelet use
31
45
.016
Diabetes mellitus
14
14
.686
Anemia
32
33
.481
Stroke
3
2
.544
Cancer
4
1
.126
Thrombocytopenia
5
4
.58
Continuous data are shown as the mean 6 standard deviation and categoric data as number of patients.
a
reverse effects of rivaroxaban were not available commercially during the time of our data collection. Statistical analyses were performed with IBM SPSS 23 software (IBM Corp, Armonk, NY) using the Student t-test and c2 test to compare demographics, readmissions due to bleeding, and need for secondary interventions. Logistic regression models were used to evaluate variables associated with major bleeding complications and secondary interventions. A P value of <.05 was defined as statistically significant. A post hoc power
Journal of Vascular Surgery Volume
-,
Number
Talukdar et al
3
-
Table II. Risk of bleeding and reintervention for patients using rivaroxaban and warfarin Rivaroxaban All ages (N ¼ 44)
Age #65 (n ¼ 27)
All
20/44 (45.5)
Open
16/36 (44.4)
Variablea
P value
Warfarin Age >65 (n ¼ 17)
All ages (N ¼ 50)
Age #65 (n ¼ 25)
Age >65 (n ¼ 25)
All ages
Age #65
Age >65
13/27 (48.1)
7/17 (41.2)
16/50 (32)
9/25 (36)
7/25 (28)
.207
.413
.508
9/22 (40.9)
7/14 (50)
8/35 (22.9)
5/19 (26.3)
3/16 (18.8)
.047
.258
.077
Reintervention
Major bleeding
a
All
3/44 (6.8)
0/27 (0)
3/17 (17.6)
8/50 (16)
5/25 (20)
3/25 (12)
.209
.02
.672
Open
3/36 (8.3)
0/22 (0)
3/14 (21.4)
6/35 (17.1)
5/19 (26.3)
1/16 (6.3)
.225
.016
.249
Data are presented as number/number of patients available (%).
analysis was performed to detect a clinically meaningful difference of 0.10 with the Fisher exact conditional test for two proportions. Continuos data are reported as the mean 6 standard deviation.
RESULTS There were 44 patients in the rivaroxaban cohort and 50 patients in the warfarin cohort. Patients were aged 60.5 6 15 years in the rivaroxaban group and 63.8 6 14 years (P ¼ .278) in the warfarin group. There were 27 men and 17 women in the rivaroxaban group and 32 men and 18 women in the warfarin group (P ¼ .833), and 90.4% of the patients in both groups were Caucasian. Neither group had any patients who were low risk for bleeding. Most patients (86.3% in rivaroxaban and 94% in warfarin) in the study were at high risk for bleeding, and both groups were statistically similar by bleeding risk (P ¼ .297). The number of risk factors were 2.4 6 0.8 in the rivaroxaban cohort and 2.7 6 0.9 in the warfarin cohort (P ¼ .199). Power analysis using the Fisher exact conditional test for two proportions yielded a value of 0.249. Concomitant antiplatelets were used in 70% (n ¼ 31) of patients prescribed rivaroxaban and in 94% (n ¼ 47) of patients prescribed warfarin (P ¼ .016). We did not collect specific information regarding the type and dosage of the antiplatelet use. None of the patients in either cohort had deep venous thrombosis or pulmonary embolism complications. The type of index interventions performed between two groups was similar (P ¼ .199). In the rivaroxaban group, 82% (n ¼ 36) of patients underwent an open vascular reconstruction, with endovascular procedures in 16% (n ¼ 7) and hybrid procedures in 2% (n ¼ 1). This procedural distribution was statistically similar to the group receiving warfarin, with 70% (n ¼ 35) of patients undergoing open, 22% (n ¼ 11) endovascular, and 8% (n ¼ 4) having a hybrid procedure. Endovascular procedures included iliac or infrainguinal arterial stents or balloon angioplasties. Open procedures included aortobifemoral bypass, infrainguinal bypass, and thrombectomy/embolectomy.
Overall, demographics, risk factors for bleeding, and intervention types were similar in both groups (Table I). Patients receiving warfarin and rivaroxaban had a similar open procedure profile. Infrainguinal procedures in the warfarin group included 12 bypasses using vein (n ¼ 7), prosthetic (n ¼ 4), or cryograft (n ¼ 1) and suprainguinal procedures included 10 bypasses using vein (n ¼ 1), prosthetic (n ¼ 6), or cryograft (n ¼ 3). Similarly, infrainguinal procedures in patients receiving rivaroxaban included 14 bypasses with vein (n ¼ 8), prosthetic (n ¼ 2), or cryograft (n ¼ 4) and suprainguinal procedures included nine bypasses with vein (n ¼ 2), prosthetic (n ¼ 7), or cryograft (n ¼ 0). Overall reintervention rates, which included hematoma evacuation and vascular reintervention, between the two cohorts were similar (Table II). The patients receiving warfarin underwent 16 reinterventions: five patients had hematoma evacuation, which included one patient with intracranial bleeding, and 11 reinterventions were for bypass occlusions. In the patients receiving rivaroxaban, there were 20 reinterventions, which included hematoma evacuations in three patients and bypass occlusions in 17. Further subgroup analysis of all patients requiring only open vascular surgical procedures showed a significantly (P ¼ .047; Fig) increased rate of reintervention in the rivaroxaban group (44.4%) compared with the warfarin group (22.9%). Although not statistically significant (P ¼ .077), a similar trend for reintervention was observed in patients aged >65 years in the two groups (rivaroxaban: 50%; warfarin: 18.8%). There was no significant difference in overall major bleeding complications (rivaroxaban: 6.8%; warfarin: 16%; P ¼ .209) or in the transfusion requirements for packed red blood cells between the two groups (rivaroxaban: 7 units; warfarin: 8 units). However in patients aged #65 years, the rivaroxaban group had no bleeding complications compared with 20% in the warfarin group (P ¼ .020). Further subgroup analysis of patients aged <65 years and who underwent open surgical procedures demonstrated that warfarin was associated with an increased bleeding risk (rivaroxaban: 0%; warfarin: 26.3%; P ¼ .016; Table II; Fig). One patient in
4
Talukdar et al
Journal of Vascular Surgery ---
Fig. Comparison of bleeding risk and reintervention.
the rivaroxaban group had minor epistaxis that was selflimiting. There were no minor bleeding episodes in the warfarin group. There were three amputations in each group. Two patients died in the rivaroxaban group and one in the warfarin cohort. These patients did not have any significant bleeding complications. One patient in the warfarin group had a major bleeding episode secondary to a supratherapeutic international normalized ratio (INR) of 4. Post hoc power analysis revealed a power of 0.204, indicating the need for a larger study with more statistical power.
DISCUSSION Newer anticoagulants are increasingly used for venous thromboembolism treatment and for the prevention of recurrent stroke with the current guidelines recommendations that are based on patient comorbidities, including the risk of bleeding.1-9,15 The safety and efficacy of these newer anticoagulants is not well known when they are used for peripheral arterial indications. Our study is a retrospective review of patients undergoing peripheral arterial procedures treated postoperatively with oral anticoagulation for maintenance of graft patency or patency of an endovascular intervention. All of the patients included in our series were considered to be moderate to high risk for bleeding according to the ACCP criteria. They received rivaroxaban or warfarin as the oral anticoagulant. Previous safety studies comparing warfarin and rivaroxaban showed similar safety and efficacy.8,10,16 Subgroup analysis of the ROCKET AF trial did not show any significant difference for bleeding risk in patients with peripheral arterial disease.10 The ROCKET AF study was not designed to analyze the adverse outcomes of patients who had undergone an arterial revascularization procedure. Our study cohort, in contrast, included patients who
2017
had undergone an open, endovascular, or hybrid peripheral arterial intervention. We used published ACCP guidelines to stratify risk for bleeding in our study population. The 2016 CHEST guidelines15 reported eight fewer major bleeding per 1000 patients while taking rivaroxaban compared with warfarin. As seen in our study cohorts overall, patients receiving rivaroxaban or warfarin were similar with respect to bleeding risk. Similar to other studies, we found that warfarin had an overall higher incidence of major bleeding with a trend towards statistical significance. These bleeding episodes were distributed not just during the perioperative period but also during the span of the study period of 21 months. This included gastrointestinal, nasal, and surgical site bleeding requiring hospitalization or transfusion of packed cells or reversal agents. The only patient with a major bleeding episode had a supratherapeutic INR of 4. On further subanalysis, we found that patients aged #65 years were more prone to bleeding when warfarin was used. In contrast, patients >65 years did not demonstrate any significant difference in bleeding. This finding was even more significant when patients undergoing an open surgical procedure were evaluated. Clinicians should remember to adjust the dosage of rivaroxaban in the elderly due to lower creatinine clearance because the half-life of rivaroxaban increases to 11 to 13 hours.1,17-20 Subgroup analysis of the ROCKET AF trial showed similar safety profile for these oral anticoagulants in elderly patients.21 Given the limitations of our study we cannot generalize the observations we have made to those made by large prospective trials that were conducted predominantly for venous indications or secondary prevention of those complications resulting from nonvalvular atrial fibrillation. When the efficacy of the oral anticoagulants was evaluated according to the type of intervention performed, reinterventions were significantly higher in patients who had an open index operation in the rivaroxaban cohort. We hypothesize that this observation is likely associated with medication compliance. In addition, we did not study the use of other medications that may affect the bioavailability. In the subset of patients who are aged >65 years, we observed a similar risk for bleeding irrespective of the agent. These observations should suggest that warfarin may be a better anticoagulant in elderly patients who require an open arterial reconstruction despite the inconvenience associated with this regimen. We do not have an explanation for this finding but postulate that it could be related to medication compliance or cost associated with the medication, although we did not examine these factors in our analyses. Use of warfarin requires frequent monitoring of the coagulation profile, which may reinforce the need for patient compliance.15
Journal of Vascular Surgery Volume
-,
Number
Talukdar et al
5
-
The ROCKET AF trial found the efficacy of both of these medications was similar, but the participants were volunteers in a controlled environment with frequent followup. Warfarin may continue to be an appropriate first choice of anticoagulation for older patients after an open surgical procedure. In patients aged #65 years, rivaroxaban may be used as the first-line agent if these patients are in the high-risk category for bleeding; however, there is an increased incidence of reintervention. Limitations to our study findings are that it is limited to a single institution, is retrospective in nature, and carries the inherent bias associated with physician’s choice of anticoagulation after arterial intervention. Despite the lack of a control group and underpowered nature, our study patients were less heterogenous and had similar risk factors for bleeding, with associated comorbidity based on ACCP risk stratification, as well as procedure profile. We were unable to perform a multivariate analysis owing to the small sample size with no matched control group. We did not perform cost analysis comparing the use of rivaroxaban with warfarin. We did not have detailed information regarding the prothrombin time or INR or on medication compliance. Nonetheless, our results provide real-world information on a newer anticoagulant that might be considered as a substitute for warfarin use in patients at high risk of lower extremity arterial intervention failure.
CONCLUSIONS This study provides early real-world experience with use of two anticoagulants. The safety and efficacy profile of rivaroxaban is comparable to warfarin when used in patients after peripheral arterial procedures. Subgroup analysis of patients aged <65 years requiring an open operation demonstrate a decreased bleeding risk when rivaroxaban was used but carried an increased risk of secondary interventions. Further studies with a larger cohort are required to validate our results. We acknowledge contributions by Teresa Bell, PhD, with Center for Outcomes Research in Surgery (CORe-S), Indiana University School of Medicine.
AUTHOR CONTRIBUTIONS Conception and design: AT, KW, LC, NM, AG, AF, MD, RM Analysis and interpretation: AT, LC, NM, AG, AF, MD, RM Data collection: AT, KW, LC, NM, RM Writing the article: AT, KW, LC, NM, AG, MD, RM Critical revision of the article: AT, KW, AG, AF, MD, RM Final approval of the article: AT, KW, LC, NM, AG, AF, MD, RM Statistical analysis: AT, KW Obtained funding: Not applicable Overall responsibility: RM
REFERENCES 1. Turpie AG, Kreutz R, Llau J, Norrving B, Haas S. Management consensus guidance for the use of rivaroxabanean oral, direct factor Xa inhibitor. Thromb Haemost 2012;108:876-86. 2. Eriksson BI, Borris LC, Friedman RJ, Haas S, Huisman MV, Kakkar AK, et al. Rivaroxaban versus enoxaparin for thromboprophylaxis after hip arthroplasty. N Engl J Med 2008;358: 2765-75. 3. Kakkar AK, Brenner B, Dahl OE, Ericksson BI, Mouret P, Muntz J, et al. Extended duration rivaroxaban versus short-term enoxaparin for the prevention of venous thromboembolism after total hip arthroplasty: a double-blind, randomized controlled trial. Lancet 2008;372:31-9. 4. Lassen MR, Ageno W, Borris LC, Lieberman JR, Rosencher N, Bandel TJ, et al. Rivaroxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty. N Engl J Med 2008;358:2776-86. 5. Turpie AG, Lassen MR, Davidson BL, Bauer KA, Gent M, Kwong LM, et al. Rivaroxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty (RECORD4): a randomized trial. Lancet 2009;373:1673-80. 6. Nunnelee JD. Review of an article: oral rivaroxaban for symptomatic venous thromboembolism. The EINSTEIN investigators et al. N Engl J Med 2010;363:2499-510. 7. Buller HR, Prins MH, Lensin AW, Decousus H, Jacobson BF, Minar E, et al; The EINSTEIN-PE Investigators. Oral rivaroxaban for the treatment of symptomatic pulmonary embolism. N Engl J Med 2012;366:1287-97. 8. Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, et al; ROCKET AF Investigators. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med 2011;365:883-91. 9. Mega JL, Braunwald E, Wiviott SD, Bassand JP, Bhatt DL, Bode C, et al. Rivaroxaban in patients with a recent acute coronary syndrome. N Engl J Med 2012;366:9-19. 10. Albers GW, Yim JM, Belew KM, Bittar N, Hattemer CR, Phillips BG, et al. Status of antithrombotic therapy for patients with atrial fibrillation in university hospitals. Arch Intern Med 1996;156:2311-6. 11. Go AS, Hylek EM, Borowsky LH, Phillips KA, Selby JV, Singer DE. Warfarin use among ambulatory patients with nonvalvular atrial fibrillation: the Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) study. Ann Intern Med 1999;131:927-34. 12. Piccini JP, Hernandez AF, Zhao X, Patel MR, Lewis WR, Peterson ED, et al. Quality of care for atrial fibrillation among patients hospitalized for heart failure. J Am Coll Cardiol 2009;54:1280-9. 13. Shamoun FE, Martin EN, Money SR. The novel anticoagulants: the surgeons’ prospective. Surgery 2013;153:303-7. 14. Jones WS, Hellkamp AS, Halperin J, Piccini JP, Breithardt G, Singer DE, et al. Efficacy and safety of rivaroxaban compared with warfarin in patients with peripheral artery disease and non-valvular atrial fibrillation: insights from ROCKET AF. Eur Heart J 2014;35:242-9. 15. Kearon C, Akl EA, Ornelas J, Blaivas A, Jimenez D, Bounameaux H, et al. Antithrombotic therapy for VTE disease: CHEST Guideline and Expert Panel Report. Chest 2016;149:315-52. 16. Snipelisky D, Ray JC, Ung R, Duart M, Kauffman C, Kusumoto F. A comparison of bleeding complications between warfarin, dabigatran, and rivaroxaban in patients undergoing cryoballoon ablation. J Interv Card Electrophysiol 2014;41:231-6. 17. López-Jiménez L, Montero M, González-Fajardo JA, Arcelus JI, Suarez C, Lobo JL, et al. Venous thromboembolism in very elderly patients: findings from a prospective registry (RIETE). Haematologica 2006;91:1046-51.
6
Talukdar et al
Journal of Vascular Surgery ---
18. Spencer FA, Gore JM, Lessard D, Emery C, Pacifico L, Reed G, et al. Venous thromboembolism in the elderly. A community-based perspective. Thromb Haemost 2008;100: 780-8. 19. Glassock RJ, Winearls C. Ageing and the glomerular filtration rate: truths and consequences. Trans Am Clin Climatol Assoc 2009;120:419-28. 20. Turpie AG, Lassen MR, Eriksson BI, Gent M, Berkowitz SD, Misselwitz F, et al. Rivaroxaban for the prevention of venous thromboembolism after hip or knee arthroplasty. Pooled analysis of four studies. Thromb Haemost 2011;105:444-53.
2017
21. Halperin JL, Hankey GJ, Wojdyla DM, Piccini JP, Lokhnygina Y, Patel MR, et al; ROCKET AF Steering Committee and Investigators. Efficacy and safety of rivaroxaban compared with warfarin among elderly patients with nonvalvular atrial fibrillation in the rivaroxaban Once Daily, Oral, and Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF). Circulation 2014;130:138-46.
Submitted Oct 11, 2016; accepted Feb 27, 2017.
Safety and efficacy of rivaroxaban compared with warfarin in patients undergoing peripheral arterial procedures Anjan Talukdar, MD,a S. Keisin Wang, MD,a Lauren Czosnowski, PharmD,a,b Nassim Mokraoui, MD,a Alok Gupta, MD,a Andres Fajardo, MD,a Michael Dalsing, MD,a and Raghu Motaganahalli, MD,a Indianapolis, Ind
ABSTRACT Objective: Rivaroxaban is a United States Food and Drug Administration-approved oral anticoagulant for venous thromboembolic disease; however, there is no information regarding the safety and its efficacy to support its use in patients after open or endovascular arterial interventions. We report the safety and efficacy of rivaroxaban vs warfarin in patients undergoing peripheral arterial interventions. Methods: This single-institution retrospective study analyzed all sequential patients from December 2012 to August 2014 (21 months) who were prescribed rivaroxaban or warfarin after a peripheral arterial procedure. Our study population was then compared using American College of Chest Physicians guidelines with patients then stratified as low, medium, or high risk for bleeding complications. Statistical analyses were performed using the Student t-test and c2 test to compare demographics, readmissions because of bleeding, and the need for secondary interventions. Logistic regression models were used for analysis of variables associated with bleeding complications and secondary interventions. The Fisher exact test was used for power analysis. Results: There were 44 patients in the rivaroxaban group and 50 patients in the warfarin group. Differences between demographics and risk factors for bleeding between groups or reintervention rate were not statistically significant (P ¼ .297). However, subgroup evaluation of the safety profile suggests that patients who were aged #65 years and on warfarin had an overall higher incidence of major bleeding (P ¼ .020). Patients who were aged >65 years, undergoing open operation, had a significant risk for reintervention (P ¼ .047) when they received rivaroxaban. Conclusions: Real-world experience using rivaroxaban and warfarin in patients after peripheral arterial procedures suggests a comparable safety and efficacy profile. Subgroup analysis of those requiring an open operation demonstrated a decreased bleeding risk when rivaroxaban was used (in those aged <65 years) but an increased risk for secondary interventions. Further studies with a larger cohort are required to validate our results. (J Vasc Surg 2017;-:1-6.)
Rivaroxaban, an oral direct anti-Xa agent, was approved by the United States Food and Drug Administration in 2011 for use in the United States. It is currently used for prevention of venous thromboembolism, deep venous thrombosis, stroke prevention in patients with nonvalvular atrial fibrillation, and secondary prevention of coronary events in patients with acute coronary syndrome.19 There are no data available to support its use in patients with peripheral arterial occlusive disease after surgical or From the Division of Vascular Surgerya and Department of Surgery,b Indiana University School of Medicine. Author conflict of interest: none. Presented at the Thirty-ninth Annual Meeting of the Midwestern Vascular Surgical Society, Chicago, Ill, September 10-12, 2015. Correspondence: Raghu Motaganahalli, MD, Division of Vascular Surgery, 1801 N Senate Blvd, MPC-2 #3500, Indianapolis, IN 46202 (e-mail: rmotagan@iupui. edu). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2017 by the Society for Vascular Surgery. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jvs.2017.02.052
endovascular intervention. Warfarin, an oral vitamin K antagonist, has been in use in general since 1954 as well as in patients deemed high risk for failure after peripheral arterial intervention. This benefit is offset by food and drug interactions as well as a bleeding risk, which is best controlled with frequent coagulation monitoring and dose adjustments as needed.8,10-13 Efficacy and safety of these drugs are compared in the ROCKET AF trial14 (The Rivaroxaban once daily oral Direct factor Xa inhibition compared with Vitamin K Antagonism for prevention of Stroke and Embolism Trial in Atrial Fibrillation) for recurrent stroke prevention in patients with nonvalvular atrial fibrillation. This randomized trial found that rivaroxaban was noninferior to warfarin in the primary analysis, which included patients in the per-protocol population, and also in the intention-totreat analysis. The primary safety analysis found no significant difference between rivaroxaban and warfarin with respect to rates of major or no major clinically relevant bleeding. The objective of our study was to compare the safety and efficacy of rivaroxaban vs warfarin in patients 1
2
Journal of Vascular Surgery
Talukdar et al
---
undergoing peripheral arterial procedures who are deemed of sufficiently high risk of failure to warrant anticoagulation.
ARTICLE HIGHLIGHTS d
d
METHODS The study was approved by the Indiana University Institutional Review Board as an exempt study, and hence, no patient consent was obtained. This single-center retrospective study included all sequential patients from December 2012 to August 2014 who were prescribed rivaroxaban after an open or endovascular intervention for lower extremity arterial occlusive disease or acute embolic occlusion. Comparisons were made with matched sequential patients who were prescribed warfarin during the same time for similar indications. Patients were identified from a database with Pharmacy, which kept a log of patients who were prescribed rivaroxaban or warfarin. Patients were prescribed warfarin or rivaroxaban according to physician preference, potential undesirable drug interactions, history of difficulty maintaining a therapeutic anticoagulation profile with warfarin, concern for hypercoagulable state, use of a nonautogenous conduit with poor runoff, and presence of a proximal embolic source. Patient demographics (Table I), readmission with cause, and need for secondary interventions were identified from inpatient record reviews. The American College of Chest Physicians (ACCP) guidelines were used to stratify patients into low, medium, or high risk for bleeding complications.15 The patient demographics that were recorded included age, any previous bleeding episodes, history of cancer, either primary or metastatic disease, including hematologic, history of chronic renal insufficiency or chronic liver disease, thrombocytopenia (<150,000/mm3), previous stroke, diabetes, anemia (<13.4 g/dL in men and <12 g/dL in women), use of antiplatelet (aspirin or clopidgrel), poor anticoagulant control, failure to thrive, recent surgery, frequent falls, and alcohol abuse. Low-risk patients had no risk factors, medium-risk patients had one risk factor, and high-risk patients had two or more risk factors. Major bleeding complications were defined as any bleeding requiring hospitalization or transfusion. Secondary interventions were defined as any intervention performed after the index operation to maintain graft/ stent patency. Interventions included open, endovascular, or hybrid procedures. Patients receiving rivaroxaban or warfarin were compared by age (#65 or >65 years), risk of bleeding, and type of primary and secondary interventions. Patients who required reintervention were managed case-by-case. Decision for an urgent intervention was dictated by an imminent threat to an extremity rather than the use of an anticoagulant. For patients requiring emergency intervention, we used either or a combination of reversal agents (vitamin K, fresh frozen plasma for warfarin) and blood transfusions. Agents to
2017
d
Type of Research: Retrospective case control study Take Home Message: In 94 patients undergoing peripheral arterial procedures rivaroxaban was equally safe and effective compared with warfarin but carried less risk of bleeding complications. Recommendation: The data suggest that in patients who require anticoagulation after lower extremity open or endovascular revascularization, rivaroxaban is preferred over warfarin.
Table I. Patient demographics and risk for bleeding Variablea
Rivaroxaban (n ¼ 44)
Warfarin (n ¼ 44)
P value
Age, years
60.5 6 15
63.8 6 14
.278
Male
27
32
Female
17
18
Gender
.833
Race Caucasian
.541 41
44
African American
3
5
Other
0
1
0
0
Risk category Low Medium High Risk factors, No.
.297 6
3
38
47
2.4 6 0.8
2.7 6 0.9
Intervention type Endovascular Open Hybrid
.199 .312
7
11
36
35
1
4
Antiplatelet use
31
45
.016
Diabetes mellitus
14
14
.686
Anemia
32
33
.481
Stroke
3
2
.544
Cancer
4
1
.126
Thrombocytopenia
5
4
.58
Continuous data are shown as the mean 6 standard deviation and categoric data as number of patients.
a
reverse effects of rivaroxaban were not available commercially during the time of our data collection. Statistical analyses were performed with IBM SPSS 23 software (IBM Corp, Armonk, NY) using the Student t-test and c2 test to compare demographics, readmissions due to bleeding, and need for secondary interventions. Logistic regression models were used to evaluate variables associated with major bleeding complications and secondary interventions. A P value of <.05 was defined as statistically significant. A post hoc power
Journal of Vascular Surgery Volume
-,
Number
Talukdar et al
3
-
Table II. Risk of bleeding and reintervention for patients using rivaroxaban and warfarin Rivaroxaban All ages (N ¼ 44)
Age #65 (n ¼ 27)
All
20/44 (45.5)
Open
16/36 (44.4)
Variablea
P value
Warfarin Age >65 (n ¼ 17)
All ages (N ¼ 50)
Age #65 (n ¼ 25)
Age >65 (n ¼ 25)
All ages
Age #65
Age >65
13/27 (48.1)
7/17 (41.2)
16/50 (32)
9/25 (36)
7/25 (28)
.207
.413
.508
9/22 (40.9)
7/14 (50)
8/35 (22.9)
5/19 (26.3)
3/16 (18.8)
.047
.258
.077
Reintervention
Major bleeding
a
All
3/44 (6.8)
0/27 (0)
3/17 (17.6)
8/50 (16)
5/25 (20)
3/25 (12)
.209
.02
.672
Open
3/36 (8.3)
0/22 (0)
3/14 (21.4)
6/35 (17.1)
5/19 (26.3)
1/16 (6.3)
.225
.016
.249
Data are presented as number/number of patients available (%).
analysis was performed to detect a clinically meaningful difference of 0.10 with the Fisher exact conditional test for two proportions. Continuos data are reported as the mean 6 standard deviation.
RESULTS There were 44 patients in the rivaroxaban cohort and 50 patients in the warfarin cohort. Patients were aged 60.5 6 15 years in the rivaroxaban group and 63.8 6 14 years (P ¼ .278) in the warfarin group. There were 27 men and 17 women in the rivaroxaban group and 32 men and 18 women in the warfarin group (P ¼ .833), and 90.4% of the patients in both groups were Caucasian. Neither group had any patients who were low risk for bleeding. Most patients (86.3% in rivaroxaban and 94% in warfarin) in the study were at high risk for bleeding, and both groups were statistically similar by bleeding risk (P ¼ .297). The number of risk factors were 2.4 6 0.8 in the rivaroxaban cohort and 2.7 6 0.9 in the warfarin cohort (P ¼ .199). Power analysis using the Fisher exact conditional test for two proportions yielded a value of 0.249. Concomitant antiplatelets were used in 70% (n ¼ 31) of patients prescribed rivaroxaban and in 94% (n ¼ 47) of patients prescribed warfarin (P ¼ .016). We did not collect specific information regarding the type and dosage of the antiplatelet use. None of the patients in either cohort had deep venous thrombosis or pulmonary embolism complications. The type of index interventions performed between two groups was similar (P ¼ .199). In the rivaroxaban group, 82% (n ¼ 36) of patients underwent an open vascular reconstruction, with endovascular procedures in 16% (n ¼ 7) and hybrid procedures in 2% (n ¼ 1). This procedural distribution was statistically similar to the group receiving warfarin, with 70% (n ¼ 35) of patients undergoing open, 22% (n ¼ 11) endovascular, and 8% (n ¼ 4) having a hybrid procedure. Endovascular procedures included iliac or infrainguinal arterial stents or balloon angioplasties. Open procedures included aortobifemoral bypass, infrainguinal bypass, and thrombectomy/embolectomy.
Overall, demographics, risk factors for bleeding, and intervention types were similar in both groups (Table I). Patients receiving warfarin and rivaroxaban had a similar open procedure profile. Infrainguinal procedures in the warfarin group included 12 bypasses using vein (n ¼ 7), prosthetic (n ¼ 4), or cryograft (n ¼ 1) and suprainguinal procedures included 10 bypasses using vein (n ¼ 1), prosthetic (n ¼ 6), or cryograft (n ¼ 3). Similarly, infrainguinal procedures in patients receiving rivaroxaban included 14 bypasses with vein (n ¼ 8), prosthetic (n ¼ 2), or cryograft (n ¼ 4) and suprainguinal procedures included nine bypasses with vein (n ¼ 2), prosthetic (n ¼ 7), or cryograft (n ¼ 0). Overall reintervention rates, which included hematoma evacuation and vascular reintervention, between the two cohorts were similar (Table II). The patients receiving warfarin underwent 16 reinterventions: five patients had hematoma evacuation, which included one patient with intracranial bleeding, and 11 reinterventions were for bypass occlusions. In the patients receiving rivaroxaban, there were 20 reinterventions, which included hematoma evacuations in three patients and bypass occlusions in 17. Further subgroup analysis of all patients requiring only open vascular surgical procedures showed a significantly (P ¼ .047; Fig) increased rate of reintervention in the rivaroxaban group (44.4%) compared with the warfarin group (22.9%). Although not statistically significant (P ¼ .077), a similar trend for reintervention was observed in patients aged >65 years in the two groups (rivaroxaban: 50%; warfarin: 18.8%). There was no significant difference in overall major bleeding complications (rivaroxaban: 6.8%; warfarin: 16%; P ¼ .209) or in the transfusion requirements for packed red blood cells between the two groups (rivaroxaban: 7 units; warfarin: 8 units). However in patients aged #65 years, the rivaroxaban group had no bleeding complications compared with 20% in the warfarin group (P ¼ .020). Further subgroup analysis of patients aged <65 years and who underwent open surgical procedures demonstrated that warfarin was associated with an increased bleeding risk (rivaroxaban: 0%; warfarin: 26.3%; P ¼ .016; Table II; Fig). One patient in
4
Talukdar et al
Journal of Vascular Surgery ---
Fig. Comparison of bleeding risk and reintervention.
the rivaroxaban group had minor epistaxis that was selflimiting. There were no minor bleeding episodes in the warfarin group. There were three amputations in each group. Two patients died in the rivaroxaban group and one in the warfarin cohort. These patients did not have any significant bleeding complications. One patient in the warfarin group had a major bleeding episode secondary to a supratherapeutic international normalized ratio (INR) of 4. Post hoc power analysis revealed a power of 0.204, indicating the need for a larger study with more statistical power.
DISCUSSION Newer anticoagulants are increasingly used for venous thromboembolism treatment and for the prevention of recurrent stroke with the current guidelines recommendations that are based on patient comorbidities, including the risk of bleeding.1-9,15 The safety and efficacy of these newer anticoagulants is not well known when they are used for peripheral arterial indications. Our study is a retrospective review of patients undergoing peripheral arterial procedures treated postoperatively with oral anticoagulation for maintenance of graft patency or patency of an endovascular intervention. All of the patients included in our series were considered to be moderate to high risk for bleeding according to the ACCP criteria. They received rivaroxaban or warfarin as the oral anticoagulant. Previous safety studies comparing warfarin and rivaroxaban showed similar safety and efficacy.8,10,16 Subgroup analysis of the ROCKET AF trial did not show any significant difference for bleeding risk in patients with peripheral arterial disease.10 The ROCKET AF study was not designed to analyze the adverse outcomes of patients who had undergone an arterial revascularization procedure. Our study cohort, in contrast, included patients who
2017
had undergone an open, endovascular, or hybrid peripheral arterial intervention. We used published ACCP guidelines to stratify risk for bleeding in our study population. The 2016 CHEST guidelines15 reported eight fewer major bleeding per 1000 patients while taking rivaroxaban compared with warfarin. As seen in our study cohorts overall, patients receiving rivaroxaban or warfarin were similar with respect to bleeding risk. Similar to other studies, we found that warfarin had an overall higher incidence of major bleeding with a trend towards statistical significance. These bleeding episodes were distributed not just during the perioperative period but also during the span of the study period of 21 months. This included gastrointestinal, nasal, and surgical site bleeding requiring hospitalization or transfusion of packed cells or reversal agents. The only patient with a major bleeding episode had a supratherapeutic INR of 4. On further subanalysis, we found that patients aged #65 years were more prone to bleeding when warfarin was used. In contrast, patients >65 years did not demonstrate any significant difference in bleeding. This finding was even more significant when patients undergoing an open surgical procedure were evaluated. Clinicians should remember to adjust the dosage of rivaroxaban in the elderly due to lower creatinine clearance because the half-life of rivaroxaban increases to 11 to 13 hours.1,17-20 Subgroup analysis of the ROCKET AF trial showed similar safety profile for these oral anticoagulants in elderly patients.21 Given the limitations of our study we cannot generalize the observations we have made to those made by large prospective trials that were conducted predominantly for venous indications or secondary prevention of those complications resulting from nonvalvular atrial fibrillation. When the efficacy of the oral anticoagulants was evaluated according to the type of intervention performed, reinterventions were significantly higher in patients who had an open index operation in the rivaroxaban cohort. We hypothesize that this observation is likely associated with medication compliance. In addition, we did not study the use of other medications that may affect the bioavailability. In the subset of patients who are aged >65 years, we observed a similar risk for bleeding irrespective of the agent. These observations should suggest that warfarin may be a better anticoagulant in elderly patients who require an open arterial reconstruction despite the inconvenience associated with this regimen. We do not have an explanation for this finding but postulate that it could be related to medication compliance or cost associated with the medication, although we did not examine these factors in our analyses. Use of warfarin requires frequent monitoring of the coagulation profile, which may reinforce the need for patient compliance.15
Journal of Vascular Surgery Volume
-,
Number
Talukdar et al
5
-
The ROCKET AF trial found the efficacy of both of these medications was similar, but the participants were volunteers in a controlled environment with frequent followup. Warfarin may continue to be an appropriate first choice of anticoagulation for older patients after an open surgical procedure. In patients aged #65 years, rivaroxaban may be used as the first-line agent if these patients are in the high-risk category for bleeding; however, there is an increased incidence of reintervention. Limitations to our study findings are that it is limited to a single institution, is retrospective in nature, and carries the inherent bias associated with physician’s choice of anticoagulation after arterial intervention. Despite the lack of a control group and underpowered nature, our study patients were less heterogenous and had similar risk factors for bleeding, with associated comorbidity based on ACCP risk stratification, as well as procedure profile. We were unable to perform a multivariate analysis owing to the small sample size with no matched control group. We did not perform cost analysis comparing the use of rivaroxaban with warfarin. We did not have detailed information regarding the prothrombin time or INR or on medication compliance. Nonetheless, our results provide real-world information on a newer anticoagulant that might be considered as a substitute for warfarin use in patients at high risk of lower extremity arterial intervention failure.
CONCLUSIONS This study provides early real-world experience with use of two anticoagulants. The safety and efficacy profile of rivaroxaban is comparable to warfarin when used in patients after peripheral arterial procedures. Subgroup analysis of patients aged <65 years requiring an open operation demonstrate a decreased bleeding risk when rivaroxaban was used but carried an increased risk of secondary interventions. Further studies with a larger cohort are required to validate our results. We acknowledge contributions by Teresa Bell, PhD, with Center for Outcomes Research in Surgery (CORe-S), Indiana University School of Medicine.
AUTHOR CONTRIBUTIONS Conception and design: AT, KW, LC, NM, AG, AF, MD, RM Analysis and interpretation: AT, LC, NM, AG, AF, MD, RM Data collection: AT, KW, LC, NM, RM Writing the article: AT, KW, LC, NM, AG, MD, RM Critical revision of the article: AT, KW, AG, AF, MD, RM Final approval of the article: AT, KW, LC, NM, AG, AF, MD, RM Statistical analysis: AT, KW Obtained funding: Not applicable Overall responsibility: RM
REFERENCES 1. Turpie AG, Kreutz R, Llau J, Norrving B, Haas S. Management consensus guidance for the use of rivaroxabanean oral, direct factor Xa inhibitor. Thromb Haemost 2012;108:876-86. 2. Eriksson BI, Borris LC, Friedman RJ, Haas S, Huisman MV, Kakkar AK, et al. Rivaroxaban versus enoxaparin for thromboprophylaxis after hip arthroplasty. N Engl J Med 2008;358: 2765-75. 3. Kakkar AK, Brenner B, Dahl OE, Ericksson BI, Mouret P, Muntz J, et al. Extended duration rivaroxaban versus short-term enoxaparin for the prevention of venous thromboembolism after total hip arthroplasty: a double-blind, randomized controlled trial. Lancet 2008;372:31-9. 4. Lassen MR, Ageno W, Borris LC, Lieberman JR, Rosencher N, Bandel TJ, et al. Rivaroxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty. N Engl J Med 2008;358:2776-86. 5. Turpie AG, Lassen MR, Davidson BL, Bauer KA, Gent M, Kwong LM, et al. Rivaroxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty (RECORD4): a randomized trial. Lancet 2009;373:1673-80. 6. Nunnelee JD. Review of an article: oral rivaroxaban for symptomatic venous thromboembolism. The EINSTEIN investigators et al. N Engl J Med 2010;363:2499-510. 7. Buller HR, Prins MH, Lensin AW, Decousus H, Jacobson BF, Minar E, et al; The EINSTEIN-PE Investigators. Oral rivaroxaban for the treatment of symptomatic pulmonary embolism. N Engl J Med 2012;366:1287-97. 8. Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, et al; ROCKET AF Investigators. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med 2011;365:883-91. 9. Mega JL, Braunwald E, Wiviott SD, Bassand JP, Bhatt DL, Bode C, et al. Rivaroxaban in patients with a recent acute coronary syndrome. N Engl J Med 2012;366:9-19. 10. Albers GW, Yim JM, Belew KM, Bittar N, Hattemer CR, Phillips BG, et al. Status of antithrombotic therapy for patients with atrial fibrillation in university hospitals. Arch Intern Med 1996;156:2311-6. 11. Go AS, Hylek EM, Borowsky LH, Phillips KA, Selby JV, Singer DE. Warfarin use among ambulatory patients with nonvalvular atrial fibrillation: the Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) study. Ann Intern Med 1999;131:927-34. 12. Piccini JP, Hernandez AF, Zhao X, Patel MR, Lewis WR, Peterson ED, et al. Quality of care for atrial fibrillation among patients hospitalized for heart failure. J Am Coll Cardiol 2009;54:1280-9. 13. Shamoun FE, Martin EN, Money SR. The novel anticoagulants: the surgeons’ prospective. Surgery 2013;153:303-7. 14. Jones WS, Hellkamp AS, Halperin J, Piccini JP, Breithardt G, Singer DE, et al. Efficacy and safety of rivaroxaban compared with warfarin in patients with peripheral artery disease and non-valvular atrial fibrillation: insights from ROCKET AF. Eur Heart J 2014;35:242-9. 15. Kearon C, Akl EA, Ornelas J, Blaivas A, Jimenez D, Bounameaux H, et al. Antithrombotic therapy for VTE disease: CHEST Guideline and Expert Panel Report. Chest 2016;149:315-52. 16. Snipelisky D, Ray JC, Ung R, Duart M, Kauffman C, Kusumoto F. A comparison of bleeding complications between warfarin, dabigatran, and rivaroxaban in patients undergoing cryoballoon ablation. J Interv Card Electrophysiol 2014;41:231-6. 17. López-Jiménez L, Montero M, González-Fajardo JA, Arcelus JI, Suarez C, Lobo JL, et al. Venous thromboembolism in very elderly patients: findings from a prospective registry (RIETE). Haematologica 2006;91:1046-51.
6
Talukdar et al
Journal of Vascular Surgery ---
18. Spencer FA, Gore JM, Lessard D, Emery C, Pacifico L, Reed G, et al. Venous thromboembolism in the elderly. A community-based perspective. Thromb Haemost 2008;100: 780-8. 19. Glassock RJ, Winearls C. Ageing and the glomerular filtration rate: truths and consequences. Trans Am Clin Climatol Assoc 2009;120:419-28. 20. Turpie AG, Lassen MR, Eriksson BI, Gent M, Berkowitz SD, Misselwitz F, et al. Rivaroxaban for the prevention of venous thromboembolism after hip or knee arthroplasty. Pooled analysis of four studies. Thromb Haemost 2011;105:444-53.
2017
21. Halperin JL, Hankey GJ, Wojdyla DM, Piccini JP, Lokhnygina Y, Patel MR, et al; ROCKET AF Steering Committee and Investigators. Efficacy and safety of rivaroxaban compared with warfarin among elderly patients with nonvalvular atrial fibrillation in the rivaroxaban Once Daily, Oral, and Direct Factor Xa Inhibition Compared with Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF). Circulation 2014;130:138-46.
Submitted Oct 11, 2016; accepted Feb 27, 2017.