Antithrombotic therapy in patients with heart failure

INTRODUCTION — Studies in patients with heart failure (HF), left ventricular systolic dysfunction, or both have noted high rates of thromboembolic events as compared with the general population. While more evidence is available on thromboembolic risk in patients with left ventricular systolic dysfunction (with or without HF), evidence suggests that HF with preserved ejection fraction is also be associated with excess thromboembolism and stroke risk. (See 'Epidemiology' below.)

Thrombus in the left heart chambers can lead to embolic stroke and other systemic embolic events, while pulmonary emboli and paradoxical emboli originate from either deep venous thrombus or thrombus in the right heart chambers. Low cardiac output, decreased physical activity, and peripheral edema all predispose to venous thrombi.

This topic will discuss the risk of thromboembolism and indications for anticoagulation in patients with HF.

Antithrombotic therapy in clinical settings that may be associated with HF is discussed separately:

●(See "Atrial fibrillation in adults: Use of oral anticoagulants".)

●(See "Amyloid cardiomyopathy: Treatment and prognosis".)

●(See "Isolated left ventricular noncompaction in adults: Management and prognosis", section on 'Management of complications'.)

●(See "Chronic Chagas cardiomyopathy: Management and prognosis".)

●(See "Hypertrophic cardiomyopathy in adults: Supraventricular tachycardias including atrial fibrillation".)

●(See "Left ventricular thrombus after acute myocardial infarction" and "Acute coronary syndrome: Oral anticoagulation in medically treated patients" and "Aspirin for the secondary prevention of atherosclerotic cardiovascular disease".)

EPIDEMIOLOGY

The epidemiology and pathophysiology of thromboembolism in HF in sinus rhythm were reviewed in a joint consensus document from the European Society of Cardiology (ESC) Heart Failure Association and the ESC Working Group on Thrombosis [1].

Incidence of thromboembolism or stroke in heart failure populations — High incidences of stroke and thromboembolic events have been observed in patients with HF. More evidence is available on the risk of stroke or thromboembolism in patients with HF with reduced ejection fraction (HFrEF; also known as HF due to systolic dysfunction or systolic HF) than in those with HF with preserved ejection fraction (HFpEF; also known as diastolic HF).

A limitation of the available data is that studies have generally not differentiated embolic from thrombotic stroke and thus it has not been determined whether the types of stroke differ in HFrEF versus HFpEF populations.

The incidence of thromboembolism in patients with HFrEF was evaluated by retrospective analyses of data from large trials, including V-HeFT, SOLVD, SAVE, and SCD-HeFT [2-5]. Thromboembolic events ranged from 1.5 to 2.7 percent per year and stroke rates ranged from 1.2 to 1.8 percent per year in these study populations. Similarly, a systematic review of studies of patients with chronic HFrEF (with study periods largely during the 1980s and 1990s) reported a stroke rate during the first year of HF of 1.8 percent; the risk of stroke was 4.7 with four to five years of follow-up [6]. Clinical factors that may affect these rates such as atrial fibrillation (AF) and antithrombotic therapy were not analyzed in the systematic review.

More contemporary data come from an analysis of the risk of stroke among patients with chronic HF (most with HFrEF) enrolled in two trials conducted in the 2000s (CORONA and GISSI-HF) [7]. The analysis included 9585 patients, 6054 of which did not have AF. Among patients without AF, the rate of stroke was 1.1 percent per year as compared with 1.7 percent among patients with AF.

Data in patients with HFpEF suggest that the risk of stroke or thromboembolism is similar to that in patients with HFrEF. Among participants in the ACTIVE trials on antiplatelet therapy (not anticoagulation), the risk of stroke, transient ischemic attack, or systemic embolism was similar in patients with HFpEF (4.3 percent per year) and patients with HFrEF (4.4 percent per year) [8]. Similarly, in the CHARM-Preserved study, the risk of stroke was unrelated to left ventricular EF (LVEF; eg, 1.2 percent per year for LVEF ≤22 percent versus 1.5 percent per year for LVEF >52 percent) [9]. In the I-PRESERVE and CHARM-Preserved studies, the percentage of deaths attributed to stroke (9 and 7 percent, respectively) were similar to or higher than rates seen in studies of patients with HFrEF (3 to 6 percent) [10].

Heart failure as stroke risk factor — The impact of HF on the risk of stroke is illustrated by the following community-based observational studies:

●Early data on 5184 Framingham Heart Study participants followed for 24 years found that clinical evidence of HF was associated with an increased risk of stroke. HF was associated with adjusted risk ratios for stroke of 4.1 in men and 2.8 in women [11]. However, stroke risk was not adjusted for presence of AF. LVEF data were not available.

●A study from Olmsted County of 630 persons with new HF identified ischemic stroke in 102 (16 percent) during 4.3-year follow-up. During the first 30 days after HF diagnosis, the risk of stroke was 17.4-fold higher than in the general population. The stroke risk remained elevated during five years of follow-up (risk ratio 2.9) [12]. AF was not a significant predictor of stroke in this HF population and LVEF was not associated with stroke in a multivariable analysis [12]. However LVEF data were missing in nearly half of the subjects and the mean LVEF among HF subjects was relatively high (43 percent).

●In the Rotterdam study, there was a more than a threefold increase in the risk of ischemic stroke in the first month after diagnosis of HF after adjustment for age, sex, and cardiovascular risk factors (adjusted hazard ratio 3.59, 95% CI 2.15-15.62), but the risk attenuated over time, becoming nonsignificant six months after diagnosis [13].

●In the Diet, Cancer, and Health study, incident HF was a major risk factor for stroke, death, and the composite of stroke and death, particularly in the 30 days following initial diagnosis of HF [14]. Of note, the use of vitamin K antagonist therapy was associated with lower risks of death and of the composite end point.

PATHOGENESIS OF THROMBOEMBOLISM IN HEART FAILURE — The pathogenesis of thromboembolism with HF with reduced ejection fraction (HFrEF) can be considered by reference to Virchow’s triad for thrombogenesis, first proposed in the mid-19^th century [15].

●"Abnormal flow" caused by low cardiac output, dilated cardiac chambers, ventricular aneurysms, and poor contractility may contribute to thrombus formation in the left and right heart chambers. In addition, atrial fibrillation is frequently present in patients with HF and is frequently associated with slow flow in the atria.

Low cardiac output, decreased physical activity, and peripheral edema all predispose to venous thrombi.

●There is laboratory evidence of a hypercoagulable state ("abnormal blood constituents") in patients with HFrEF in sinus rhythm as indicated by increases in platelet activation and in plasma viscosity, fibrinogen, von Willebrand factor, and fibrin D-dimer [15-18]. The degree of hypercoagulability is related to the severity of the HF, but not to the left ventricular ejection fraction.

Hemostatic abnormalities can be improved by anticoagulation with warfarin [19]. Angiotensin converting enzyme inhibitors, which are routinely used in patients with HF, may reduce some components of the prothrombotic state; a similar effect does not appear to occur with beta blockers [17].

●Endothelial dysfunction ("vessel wall abnormalities") may also contribute to thromboembolism in patients with HF.

These abnormalities contribute to a prothrombotic state, which increases the risk of thromboembolism in patients with HFrEF.

Less information is available on the pathogenesis of thromboembolism in patients with HF with preserved ejection fraction (HFpEF). There is evidence of endothelial dysfunction in patients with HFpEF [20], though a link to risk of thromboembolism has not been established.

RISK FACTORS FOR THROMBOEMBOLISM — A role for risk stratification in identifying patients with HF who may benefit from antithrombotic therapy has not been established. Stratification of risk of thromboembolism in patients with HF is evolving. While atrial fibrillation (AF) is a risk factor, the predictive value of clinical characteristics such as left ventricular (LV) systolic dysfunction and presence of LV thrombus on echocardiogram is uncertain. Other risk factors, such as diabetes mellitus treated with insulin, predict risk of stroke in patients with HF but the relative contributions of thrombotic versus embolic stroke is uncertain.

Atrial fibrillation — AF is a risk factor for stroke or thromboembolism among patients with HF [21,22]. Among patients with AF, HF is an important risk factor for embolization as reflected in risk scores used to guide therapy (eg, the CHA₂DS₂-VASc score). (See "Atrial fibrillation in adults: Use of oral anticoagulants".)

Left ventricular systolic dysfunction — An association between LV systolic dysfunction and stroke or thromboembolism has been found in some studies [3,5] but not in other studies in patients with [8,23] or without AF [7,9,24].

Left ventricular thrombus — The prognostic importance of LV thrombi identified by echocardiography in patients with HF or LV systolic dysfunction is uncertain. Identification of LV thrombus by echocardiography (particularly mobile or protruding) was associated with risk of embolic complications in some studies [25,26], but not in others [22,27,28]. One confounding factor may be limited sensitivity and specificity of echocardiographic detection of LV thrombus [29], particularly in older studies not using contrast agents. (See "Echocardiography in detection of cardiac and aortic sources of systemic embolism".)

Patients with LV thrombus in the setting of an acute myocardial infarction have a significantly increased risk of embolic events. LV thrombus in acute myocardial infarction, stress-induced cardiomyopathy, and Chagas cardiomyopathy is discussed separately. (See "Left ventricular thrombus after acute myocardial infarction" and "Chronic Chagas cardiomyopathy: Management and prognosis", section on 'Thromboembolism treatment' and "Management and prognosis of stress (takotsubo) cardiomyopathy", section on 'Thromboembolism'.)

The LV ejection fraction is an important risk factor for thrombus formation, while severe mitral regurgitation may have a protective effect by decreasing stasis [30].

Scant data are available on the predictive value of spontaneous echocardiographic contrast (known as smoke) as a risk factor for stroke in patients with HF and/or LV systolic dysfunction [31].

Clinical risk models — Models incorporating clinical characteristics may help predict the risk of thromboembolism or stroke in patients with HF. However, further study is needed to determine whether such models can help identify patients most likely to benefit from antiplatelet or anticoagulant therapy.

The CHA₂DS₂-VASc score commonly used to predict risk of stroke in patients with AF (table 1 and table 2) may also have predictive value in patients with HF (with or without AF). In a study using registry data on 42,987 patients with HF (21,986 with concomitant AF), the CHA₂DS₂-VASc score was associated with ischemic stroke, thromboembolism, and death at one year [32]. Thromboembolism included ischemic stroke, transient ischemic attack, systemic embolism, pulmonary embolism, or acute myocardial infarction.

For scores 1 through 6, respectively:

●Among patients with concomitant AF

•The risk of ischemic stroke was 1.5, 1.0, 1.1, 1.7, 2.2, 3.6

•The risk of thromboembolism was 3.3, 2.9, 3.0, 3.4, 4.7, 6.9

•The risk of death was 2.0, 4.2, 9.5, 13.8, 15.0, 18.9

●Among patients without AF

•The risk of ischemic stroke was 0.4, 0.4, 0.6, 1.0, 1.3, 2.6

•The risk of thromboembolism was 1.6, 2.0, 2.6, 3.5, 4.5, 7.5

•The risk of death was 2.1, 2.4, 5.7, 8.7, 9.8, 12.9

A study using data from two large HF trials proposed two clinical models for stroke risk in patients without AF [7]:

●One model included five independent risk factors for stroke: age, New York Heart Association class, diabetes mellitus treated with insulin, previous stroke, and body mass index up to 30 (the last being inversely related to risk of stroke).

•This model was validated using the CHARM trial data (which did not include natriuretic peptide data).

•For patients classified in the highest tertile of risk by this model, the stroke rate was 1.98 percent per year in the derivation cohort and 1.79 percent per year in the validation cohort. (For comparison, the rate of stroke generally was 1.7 per year in patients with AF and 1.1 percent per year in patients without AF.)

●One model included three independent risk factors for stroke: N-terminal pro brain natriuretic peptide, diabetes mellitus treated with insulin, and history of stroke. For patients classified in the highest tertile of risk, the stroke rate was 2.29 percent per year.

Thus, the above models identified patients with HF without AF with rates of stroke similar to those for patients with HF and AF. However, neither study assessed the distribution of thrombotic and embolic events within overall stroke or thromboembolic rates. The utility of such models in managing patients with HF remains to be determined.

ROLE OF ANTITHROMBOTIC THERAPY

Our approach — Our recommendations for antithrombotic therapy are based on the available evidence on thromboembolic risk and risks and benefits of antiplatelet and anticoagulant therapy in patients with HF or left ventricular (LV) systolic dysfunction.

●The following recommendations apply to patients with LV systolic dysfunction:

•For patients in sinus rhythm with LV systolic dysfunction (with or without HF) without coronary artery disease, prior thromboembolic event, acute LV thrombus, atrial fibrillation, or other indication for antithrombotic therapy, we recommend against administering aspirin or vitamin K antagonist.

•For patients in sinus rhythm with HF with reduced ejection fraction without coronary artery disease, atrial fibrillation, or other indication for direct oral anticoagulant therapy, we suggest against using a direct oral anticoagulant. Indications for use of direct oral anticoagulant in patients with coronary artery disease are discussed separately. (See "Prevention of cardiovascular disease events in those with established disease (secondary prevention) or at very high risk" and "Acute coronary syndrome: Oral anticoagulation in medically treated patients".)

•For patients in sinus rhythm HF with reduced ejection fraction and a prior thromboembolic event who lack another cause of thromboembolism, we suggest oral anticoagulant therapy (eg, warfarin), although supportive data are limited. (See "Overview of secondary prevention of ischemic stroke", section on 'Other risk factors'.)

•Antithrombotic therapy for patients with coronary artery disease and LV systolic dysfunction should be governed by the applicable recommendations for coronary artery disease. The role of antithrombotic therapy in patients with coronary artery disease is discussed separately. (See "Acute non-ST-elevation acute coronary syndromes: Early antiplatelet therapy" and "Acute ST-elevation myocardial infarction: Antiplatelet therapy" and "Prevention of cardiovascular disease events in those with established disease (secondary prevention) or at very high risk" and "Aspirin for the secondary prevention of atherosclerotic cardiovascular disease".)

●For patients with LV thrombus:

•The risk of thromboembolism in patients with chronic LV thrombus and HF and/or LV systolic dysfunction is uncertain (see 'Left ventricular thrombus' above) and evidence is lacking to support anticoagulation in such patients.

•For patients in sinus rhythm with acute LV thrombus and LV systolic dysfunction not caused by acute myocardial infarction, we recommend oral anticoagulant therapy (eg, warfarin). Similar recommendations for patients with acute LV thrombus with acute myocardial infarction (see "Left ventricular thrombus after acute myocardial infarction") or stress cardiomyopathy are discussed separately. (See "Clinical manifestations and diagnosis of stress (takotsubo) cardiomyopathy".)

●Specific recommendations for antithrombotic therapy apply to certain types of cardiomyopathy (with or without LV thrombus) as discussed separately. (See "Isolated left ventricular noncompaction in adults: Clinical manifestations and diagnosis" and "Amyloid cardiomyopathy: Treatment and prognosis" and "Chronic Chagas cardiomyopathy: Management and prognosis" and "Hypertrophic cardiomyopathy in adults: Supraventricular tachycardias including atrial fibrillation".)

●Given the lack of demonstrated benefit of antithrombotic therapy in patients with HF with reduced ejection fraction, we do not use antiplatelet or anticoagulant therapy in patients with HF with preserved LV ejection fraction in the absence of a specific indication for such therapy.

●Antithrombotic therapy in patients with HF and AF is governed by applicable guidelines for nonvalvular AF (or for valvular AF, when appropriate). HF is a thromboembolic risk factor among patients with AF. (See "Atrial fibrillation in adults: Use of oral anticoagulants".)

The risk of bleeding should be evaluated in all patients undergoing consideration for warfarin therapy. Candidates for anticoagulation must have minimal risk factors for bleeding (such as abnormal liver function) and a stable hemodynamic profile. Assessment of risk of bleeding is discussed in detail separately.

The optimal degree of anticoagulation has not been established for patients with sinus rhythm, HF, and LV systolic dysfunction with a prior thromboembolic event. We aim for an International Normalized Ratio (INR) between 2.0 and 3.0 in patients with prior thromboembolic event. Once the patient is stable, the INR is measured monthly. We reassess the safety of continued anticoagulation if the heart disease progresses.

We do not consider direct oral anticoagulants, such as dabigatran or rivaroxaban, alternatives to warfarin in this setting. A randomized controlled trial comparing oral rivaroxaban with placebo in patients with HF and significant coronary artery disease failed to show benefit [33]. (See 'Direct oral anticoagulant' below.)

Evidence on antithrombotic therapy in heart failure

Antithrombotic therapy compared with placebo — The available evidence has not demonstrated an overall clinical benefit from antithrombotic therapy (vitamin K antagonist, direct oral anticoagulant, or aspirin) in patients with HF in sinus rhythm [34].

Effect of vitamin K antagonist — A systematic review concluded that based upon two multicenter randomized trials (HELAS [24] and WASH [35]), there is no convincing evidence that vitamin K antagonist reduces mortality or vascular events in patients with HF and sinus rhythm [34].

●The Warfarin/Aspirin Study in Heart failure (WASH) trial randomized 279 patients with HF and to warfarin, aspirin, or no therapy over mean follow-up of 27 months [35]. There were no significant differences between the treatment groups in frequency of the primary composite outcome of death, nonfatal myocardial infarction, or nonfatal stroke. Hospitalization for cardiovascular reasons (especially worsening HF) was more frequent in patients randomized to aspirin.

●The Heart failure Long-term Antithrombotic Study (HELAS) evaluated antithrombotic treatment in 197 patients with HF and LVEF <35 percent [24]. Patients with ischemic heart disease were randomized to receive either aspirin 325 mg or warfarin. Patients with dilated cardiomyopathy were randomized to receive either warfarin or placebo. There were 2.2 embolic events per 100 patient-years, with no significant differences between treatment groups.

Earlier observational studies (including retrospective analyses of clinical trial data) provided conflicting data on the effectiveness of warfarin therapy in this population [2-4,36,37].

The uncertain benefit associated with warfarin therapy must be evaluated in relation to the risk of bleeding. In AF trials in patients with normal LV function, for example, the incidence of major bleeding ranges between 0.4 and 2.5 percent per year (see "Atrial fibrillation in adults: Use of oral anticoagulants"). The risk of hemorrhage with warfarin varies with the clinical indication and with the presence of comorbid conditions [37,38]. In general, anticoagulation is more difficult to regulate and bleeding complications are more frequent in patients with HF [39-41]. (See "Management of warfarin-associated bleeding or supratherapeutic INR", section on 'Mitigating bleeding risk'.)

Direct oral anticoagulant — There is no established role for direct oral anticoagulant therapy in patients with HF in sinus rhythm in the absence of another indication for anticoagulant therapy. In the COMMANDER HF trial, 5022 patients with chronic HF, left ventricular ejection fraction ≤40 percent, coronary artery disease, elevated plasma natriuretic peptide level, and no atrial fibrillation were randomly assigned to receive rivaroxaban 2.5 mg twice daily or placebo [33]. During the median follow-up of 21 months, there was no significant difference in all-cause mortality between the rivaroxaban and placebo groups (21.8 versus 22.1 percent). The composite outcome of death from any cause, myocardial infarction, or stroke occurred at similar rates in the rivaroxaban and placebo groups (25.0 versus 26.2 percent). There was a significant reduction in the risk of stroke (2.0 versus 3.0 percent; HR 0.66, 95% CI 0.47-0.96) but this reduction was not sufficient to cause a significant difference in the composite outcome. The risk of major bleed was higher in the rivaroxaban group (3.3 versus 2.0 percent).

Evidence on the use of direct oral anticoagulant in patients with coronary artery disease is discussed separately. (See "Prevention of cardiovascular disease events in those with established disease (secondary prevention) or at very high risk" and "Acute coronary syndrome: Oral anticoagulation in medically treated patients".)

Effect of aspirin — While selected patients with HF may benefit from antithrombotic therapy (eg, patients with coronary artery disease), a benefit from routine aspirin therapy in patients with HF has not been established, and some studies have suggested possible harm.

●Results of randomized trials in patients with HF with reduced ejection fraction (HFrEF) have been mixed. The WASH trial reported a greater rate of hospitalization for HF with aspirin compared with placebo or warfarin therapy and the WATCH trial showed a greater rate of hospitalization for HF with aspirin compared with warfarin therapy [35,42]. However, the larger WARCEF trial found no significant difference in hospitalization rates between aspirin and warfarin groups [43].

●In a study using nationwide Danish registries of patients with HF with no history of atrial fibrillation, 3840 patients using low-dose aspirin were propensity matched to an equal number of patients not taking aspirin [44]. An increased risk of HF readmission was observed in the aspirin group (HR 1.25, 95% CI 1.17-1.33). The aspirin and nonaspirin groups had similar rates of mortality and of stroke, as well as the composite outcome (mortality, myocardial infarction, and stroke). Aspirin use was associated with an increased risk of myocardial infarction, but possible confounding by indication and misclassification could contribute to this observation.

A potential interaction between aspirin and angiotensin converting enzyme (ACE) inhibitors has been investigated. While aspirin may attenuate some of the acute hemodynamic effects of ACE inhibitors, most of the available evidence does not support an inhibitory effect of aspirin on the long-term outcome benefits of ACE inhibitors in HF.

Limited data suggest that aspirin use may interfere with the beneficial effects of beta blocker therapy on LVEF in patients with HFrEF [45]. The analysis was adjusted for other factors that may affect remodeling, but confounding cannot be excluded.

Comparison of aspirin to warfarin — Randomized trials have shown that warfarin therapy does not improve overall outcomes among patients with HF compared with aspirin therapy [24,35,42,43]. Meta-analyses of these trials found no significant difference between warfarin and aspirin therapy in all-cause mortality [46-48]. While warfarin therapy reduced the rates of stroke (RR 0.56, 95% CI 0.38-0.82) and ischemic stroke (RR 0.45, 95% CI 0.24-0.86) compared with aspirin, it increased the risk of major bleeding (RR 1.95, 95% CI 1.37-2.76) [46].

The best data comparing aspirin and warfarin in patients with HF in sinus rhythm come from the WATCH and WARCEF trials:

●In the WATCH trial, 1587 adults in sinus rhythm with symptomatic HF and LVEF ≤35 percent were randomly assigned to aspirin (162 mg daily), clopidogrel (75 mg daily), or warfarin titrated to a target INR of 2.5 to 3.0 [42]. A majority (74 percent) of the subjects had coronary artery disease.

•There were no significant differences among the three treatment arms for the primary end point of time to first occurrence of death, nonfatal myocardial infarction, or nonfatal stroke.

•Warfarin was associated with significantly fewer nonfatal strokes than clopidogrel or aspirin (0.2 versus 2.1 and 1.7 percent). However, when central nervous system bleeds and fatal strokes were included, the effect of warfarin was no longer significant (1.7 versus 2.5 and 2.9 percent, p>0.17 to 0.35).

•Hospitalization for worsening HF occurred more frequently in patients treated with aspirin than in those treated with warfarin (22.2 versus 16.5 percent).

●In the WARCEF trial, 2305 adults in sinus rhythm with an LVEF ≤35 percent were randomly assigned to aspirin (325 mg daily) or warfarin (titrated to a target INR of 2.75 with acceptable range of 2.0 to 3.5) and followed for up to six years [43]. Less than half (43 percent) of the subjects had coronary artery disease.

•There was no significant difference between warfarin and aspirin for the primary end point of ischemic stroke, intracerebral hemorrhage, or death from any cause (7.47 versus 7.93 per 100 patient-years). The hazard ratio (HR) changed over time with a marginally significant advantage of warfarin over aspirin by the fourth year of follow-up (p = 0.046).

•Warfarin significantly reduced the rate of ischemic stroke compared with aspirin (0.72 versus 1.36 events per 100 patient-years). However, major hemorrhage was more frequent in the warfarin group (1.78 versus 0.87 events per 100 patient-years). Rates of intracerebral and intracranial hemorrhage were similar in the two treatment groups (0.27 versus 0.22 events per 100 patient-years).

•There was a nonsignificant trend (p = 0.053) toward higher rate of hospitalization for HF with warfarin than with aspirin. This result contrasts with the WASH and WATCH trial results of greater hospitalization for HF in patients treated with aspirin as compared with those treated with warfarin [35,42].

●An exploratory subgroup analysis of the WARCEF trial found that age was a significant treatment effect modifier [49]. In patients <60 years old, the primary outcome (the first to occur of ischemic stroke, intracerebral hemorrhage, or death) was lower with warfarin compared with aspirin (4.81 versus 6.76 events per 100 patient-years, HR 0.63, 95% CI 0.48-0.84). By contrast, in patients ≥60 years old, the event rate for the primary outcome was similar with warfarin and aspirin (9.91 versus 9.01 events per 100 patient-years, HR 1.09, 95% CI 0.88-1.35). With major hemorrhage added as an event, in patients <60 years old, the event rate remained significantly lower with warfarin compared with aspirin, but in patients ≥60 years old, the event rate was lower with aspirin than with warfarin. Further study is needed to determine whether warfarin is beneficial in younger patients with HF.

Recommendations of others — Our approach is similar to that described in major society guidelines. The 2012 American College of Chest Physicians guidelines suggest not using antiplatelet therapy or warfarin for patients with LV systolic dysfunction without established coronary artery disease and no LV thrombus [50]. For patients with systolic LV dysfunction without established coronary artery disease with identified acute LV thrombus (eg, stress-induced cardiomyopathy), warfarin therapy (target INR 2.0 to 3.0) for at least three months is suggested. For patients with systolic LV dysfunction and established coronary artery disease, recommendations are as per the coronary artery disease recommendations. Recommendations for patients with established coronary artery disease are discussed separately. (See "Left ventricular thrombus after acute myocardial infarction" and "Aspirin for the secondary prevention of atherosclerotic cardiovascular disease".)

Similarly, the 2012 American College of Cardiology/American Heart Association (ACC/AHA) guidelines noted that anticoagulation is not recommended in patients with systolic HF without AF, a prior thromboembolic event, or a cardioembolic source [51].

The 2010 Heart Failure Society of America (HFSA) guidelines for HF with LV systolic dysfunction recommend anticoagulation (goal INR 2.0 to 3.0) for all patients with HF a history of systemic or pulmonary emboli (including stroke or transient ischemic attack) or AF [52].

The 2012 joint consensus document from the European Society of Cardiology ESC Heart Failure Association and Working Group on Thrombosis noted that in the absence of a specific indication (such as documented coronary artery disease), aspirin should not be initiated in patients with HF in sinus rhythm [1]. The document also noted that there is currently no compelling reason for routine warfarin therapy in HF patients in sinus rhythm.

The ACC/AHA, HFSA, and ESC HF guidelines also include recommendations for anticoagulation for patients with HF and AF [51-53]. Management of such patients is discussed separately. (See "Atrial fibrillation in adults: Use of oral anticoagulants".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Heart failure in adults".)

SUMMARY AND RECOMMENDATIONS

●Patients with systolic heart failure (HF) are at increased risk for thromboembolic events. (See 'Incidence of thromboembolism or stroke in heart failure populations' above.)

●Risk factors for thromboembolism in patients with left ventricular (LV) systolic dysfunction have not been definitively established. Severity of systolic dysfunction, presence of protuberant thrombus in the LV cavity, a hypercoagulable state, and presence of atrial fibrillation (AF) may be factors but study results have been mixed. (See 'Risk factors for thromboembolism' above.)

●Data are inconclusive on the utility of antithrombotic therapy (antiplatelet therapy or anticoagulation) to reduce thromboembolic events or mortality in patients with systolic HF who are in sinus rhythm. (See 'Antithrombotic therapy compared with placebo' above.)

●The results of randomized trials show that warfarin therapy does not improve overall outcomes among patients with HF compared with aspirin therapy. (See 'Comparison of aspirin to warfarin' above.)

●For patients in sinus rhythm with LV systolic dysfunction (with or without HF) without coronary artery disease, prior thromboembolic event, acute LV thrombus, atrial fibrillation or other indication for antithrombotic therapy, we recommend against administering aspirin or vitamin K antagonist (Grade 1B). (See 'Our approach' above and 'Effect of vitamin K antagonist' above.)

●For patients in sinus rhythm with HF with reduced ejection fraction without coronary artery disease and without atrial fibrillation or other indication for a direct oral anticoagulant therapy, we suggest against use of a direct oral anticoagulant. (Grade 2B). Indications for use of direct oral anticoagulant in patients with coronary artery disease are discussed separately. (See "Prevention of cardiovascular disease events in those with established disease (secondary prevention) or at very high risk" and "Acute coronary syndrome: Oral anticoagulation in medically treated patients".)

●For patients with HF with reduced ejection fraction (HFrEF) in sinus rhythm and prior thromboembolic event who lack another cause of thromboembolism, we suggest a vitamin K antagonist (eg, warfarin) versus no anticoagulant (Grade 2C). Due to lack of data, we do not consider newer anticoagulants such as dabigatran or rivaroxaban alternatives to warfarin in this setting. (See 'Our approach' above.)

●Given the lack of demonstrated benefit of antithrombotic therapy in patients with HFrEF, we do not use antiplatelet or anticoagulant therapy in patients with HF with preserved LV ejection fraction in the absence of a specific indication for such therapy.

●Treatment of patients with HF and AF should be governed by applicable guidelines for nonvalvular AF (or for valvular AF, when appropriate). (See "Atrial fibrillation in adults: Use of oral anticoagulants".)

●Antithrombotic therapy for patients with coronary artery disease and LV systolic dysfunction should be governed by the applicable recommendations for coronary artery disease. The role of antithrombotic therapy in patients with coronary artery disease is discussed separately. (See "Prevention of cardiovascular disease events in those with established disease (secondary prevention) or at very high risk" and "Aspirin for the secondary prevention of atherosclerotic cardiovascular disease" and "Coronary artery disease patients requiring combined anticoagulant and antiplatelet therapy".)