Anticoagulation and Antiplatelet Drugs

Patients with HF are recognized to be at increased risk for arterial or venous thromboembolic events. In addition to atrial fibrillation and poor ventricular function, which promote stasis and increase the risk of thrombus formation, patients with HF have other manifestations of hypercoagulability. Evidence of heightened platelet activation, increased plasma and blood viscosity, and increased plasma levels of fibrinopeptide A, beta-thromboglobulin, D-dimer, and von Willebrand factor have been found in many patients.112-114 Despite a predisposition, estimates regarding the incidence of thromboemboli in patients with HF vary substantially between 1.4% and 4.2% per 100 patient years.115-117 Although variability in the reported incidence likely results from differences in the populations studied and the methodology used to identify these events, the consensus is that pulmonary and systemic emboli are not common in HF patients in sinus rhythm. Traditionally, discussion of anticoagulation in patients with HF has centered on warfarin. Antiplatelet agents are often used in patients with HF from ischemic heart disease.

Previous guidelines have recommended warfarin anticoagulation in patients with HF complicated by atrial fibrillation or prior thromboembolic events.118 Warfarin anticoagulation was specifically not recommended in patients with HF in the absence of these indications.

Recommendations regarding warfarin use, in the absence of atrial fibrillation or clinically overt systemic or pulmonary thromboemboli, must be made on the basis of cohort data and expert opinion. The likely incidence of thromboembolic events and the possibility of averting them with warfarin are important considerations for any guideline recommendation. In addition, the potential beneficial effects of warfarin on coronary thrombotic events, independent of embolic phenomena, must be taken into account. The substantial clinical trial data reflecting the beneficial effects of antiplatelet therapy in patients with ischemic heart disease suggest that the role of this therapy in patients with reduced LVEF should be addressed.

Previous guideline recommendations have been positive concerning warfarin therapy in patients with HF complicated by atrial fibrillation, a common clinical presentation. The benefit of warfarin anticoagulation in this setting is well established through several randomized trials.119 Warfarin anticoagulation should be implemented in these patients unless clear contraindications exist.

LV thrombus is a frequent finding in patients with dilated dysfunctional ventricles, especially in patients who have suffered a large anterior MI, although the incidence appears to be declining with modern therapies.120-122 LV thrombus is associated with thromboembolism, especially cerebral embolism.123-125 Two-thirds of these embolic events occur in the first week after MI.124,125 When LV mural thrombus is present, anticoagulation does appear to reduce the incidence of subsequent embolic events.123 There are no randomized trials of anticoagulation for LV thrombus, but the data presented have led to a recommendation for short-term (3 months) anticoagulation in patients with a large anterior MI and wall motion abnormality or in patients with LV thrombus.126

Background

Cohort analyses examining the relationship between warfarin use and noncoronary thromboembolism in patients with HF have not yielded consistently positive findings.115,117,127-130 It is possible that the lack of consistent benefit was related to the low incidence of identifiable embolic events in these populations. Other retrospective evaluations of the use of anticoagulation in patients with HF have also yielded conflicting results.131-133 Thromboembolic events were not different in HF patients who were taking warfarin as compared to those who were not in several retrospective analyses.117,134,135 Warfarin was associated with a reduction in cardiovascular events and death in a retrospective analysis of the Studies of Left Ventricular Dysfunction (SOLVD) studies,136 whereas no difference in antiplatelet or anticoagulant therapy were observed in another analysis.137

A recent review suggested that anticoagulation with warfarin in patients with HF reduced death and cardiovascular events but that the data were insufficient to recommend routine use.138 Two prospective randomized trials of anticoagulation have been published since that review but both were underpowered. The Warfarin/Aspirin Study in Heart Failure (WASH) randomized 279 patients with HF to warfarin (INR target 2.5), 300 mg aspirin, or no treatment.139 There were no differences in the combined primary outcomes of death, MI, or stroke. However, significantly more patients randomized to aspirin were hospitalized for ADHF or serious adverse gastrointestinal events.

In the Warfarin and Antiplatelet Therapy in Chronic Heart Failure (WATCH) trial, patients with symptomatic HF and reduced LVEF were randomized to aspirin 162 mg/day, clopidogrel 75 mg/day, or open-label warfarin to achieve an INR of 2.5 to 3.140 The primary endpoint of the study was the composite of all-cause mortality, non-fatal MI, and non-fatal stroke. The majority of patients had an ischemic etiology of HF, although the study population was not limited to patients with coronary artery disease. There were no statistically significant differences in the primary endpoint for warfarin versus aspirin, for clopidogrel versus aspirin, or for warfarin versus clopidogrel.140 However, as in WASH, fewer patients randomized to warfarin were hospitalized for HF. A recent retrospective analysis of 290 patients with HF and LVEF <35% and idiopathic dilated cardiomyopathy reported an odds ratio of 3.4 (P=.027) for stroke in those with LV thrombus but no difference in mortality.141 In the absence of strong data, the decision to anticoagulate must be an individual one. There are insufficient data for or against the use of warfarin in patients with dilated cardiomyopathy and LVEF <=35%.

The analysis of the SOLVD population mentioned above focused on the relation between warfarin use and the risk of all-cause mortality rather than risk for embolic events.136 After adjustment for baseline differences, patients treated with warfarin at baseline had a 24% lower risk of mortality during follow-up. Warfarin use also was associated with an 18% reduction in the combined endpoint of death or hospitalization for HF. In the SOLVD population, the benefit associated with warfarin use was not significantly influenced by (1) presence or absence of symptoms, (2) randomization to enalapril or placebo, (3) gender, (4) presence or absence of atrial fibrillation, (5) age, (6) LVEF, (7) NYHA class, or (8) etiology.

The benefit associated with warfarin use in the cohort analysis of the SOLVD population was related to a reduction in cardiac mortality. Specifically, there was a significant reduction among warfarin users in deaths that were identified as sudden, in deaths associated with HF, and in fatal MI There was no significant difference in deaths considered cardiovascular but non-cardiac, including pulmonary embolism and fatal stroke. Some caution is needed related to this finding as the number of cardiovascular deaths that were non-cardiac was far smaller than the number of cardiac deaths.

Reduction in ischemic events is one potential explanation for the apparent benefit from warfarin in the SOLVD study. Warfarin users showed a reduced rate of hospitalization for unstable angina or nonfatal MI. Prior investigations in patients following acute MI showed that warfarin anticoagulation, when begun within 4 weeks, reduced the incidence of fatal and non-fatal coronary events, as well as pulmonary emboli and strokes.139

As with other post-hoc cohort analyses, it is possible that the findings from the SOLVD study may result from unidentified differences between the treatment groups, for which statistical correction could not adequately adjust. For this reason, evidence from any cohort study must be considered less powerful than that derived from randomized, controlled trials.

Combined Use of Aspirin and an ACE Inhibitor. Strong evidence supports the clinical benefit of both aspirin and ACE inhibitors in ischemic heart disease and atherosclerosis.142-145 However, post-hoc analyses of large randomized trials involving ACE inhibitors in HF and post-MI have raised the possibility of an adverse drug interaction between aspirin and ACE inhibitors.146

It is critical to understand the possible nature of the adverse interaction raised by these retrospective analyses. Because both aspirin and ACE inhibitors are beneficial in ischemic heart disease, patients taking both agents might be expected to do better than patients on either agent alone. However, if the 2 drugs have similar mechanisms of action, then additive benefit would not be expected. Another possibility is that one drug might antagonize the effects of the other, resulting in reduced benefit from the combination.

Post-MI. Early work concerning the nature of the interaction in ischemic heart disease, using data from the Cooperative New Scandinavian Enalapril Survival Study II (CONSENSUS II) and the Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries (GUSTO-1) studies in post-MI patients, suggested not only lack of additive benefit, but also the possibility of a negative effect on mortality from the combination of aspirin and ACE inhibition. A large-scale meta-analysis of patients after acute MI failed to confirm an adverse interaction, with evidence of significant benefit from ACE inhibition in patients taking and not taking aspirin.147 However, the point estimate for the reduction in mortality in patients taking the combination of aspirin and ACE inhibition, whereas not statistically less than for aspirin alone, was lower, providing no support for additive benefit from the 2 drugs.

Heart Failure. A retrospective cohort analysis of the SOLVD study found that patients on antiplatelet therapy (assumed to be aspirin in the great majority of cases) derived no additional survival benefit from the addition of enalapril.146 Other studies have shown no clear evidence of harm from the combination of aspirin and ACE inhibitors in patients with HF.148

Relationship to Dose. There is also some evidence that the potential interaction between aspirin and ACE inhibitor may be dose-related. A meta-analysis of all hypertension and HF patients who have received both aspirin and ACE inhibitors suggests that aspirin at doses <=100 mg did not interact with ACE inhibitors.149 Any interaction, if observed, occurred at higher doses of aspirin. A more recent meta-analysis could not confirm or exclude a modest effect of aspirin on the benefits of ACE inhibitors.147

A potential mechanism for the hypothesized adverse interaction between aspirin and ACE inhibitors in patients with HF involves prostaglandin synthesis. ACE inhibition is felt to augment bradykinin, which in turn stimulates the synthesis of various prostaglandins that may contribute vasodilatory and other salutary effects. In the presence of aspirin, the bradykinin-induced increase in prostaglandins should be attenuated or blocked, potentially reducing the benefits of ACE inhibition. Invasive hemodynamic monitoring has demonstrated that the acute hemodynamic effect of enalapril is blunted by concomitant administration of aspirin.150 Another possibility is that aspirin and ACE inhibitors act in a similar fashion in HF so that no added benefit is gained from the combination. ACE inhibitors appear to reduce ischemic events in HF patients possibly through antithrombotic effects, which could mimic those of antiplatelet agents. Recent study results suggesting that aspirin may have independent beneficial action on ventricular remodeling support the hypothesis of similar mechanisms of action for ACE inhibitors and aspirin.151

Development of the adenosine diphosphate antagonists, ticlopidine and clopidogrel, provide alternative therapy for platelet inhibition that does not appear to influence prostaglandin synthesis.152 In direct comparison with aspirin, large-scale clinical trial results have established the efficacy of clopidogrel in the prevention of vascular events in patients with arteriosclerotic disease.153 In WATCH, there were no statistically significant differences in death, MI, or stroke for warfarin versus aspirin, for clopidogrel versus aspirin, or for warfarin versus clopidogrel.140 A hemodynamic study found a similar reduction in systemic vascular resistance in HF patients treated with the combination of ACE inhibitor and ticlopidine versus ACE inhibitor alone, suggesting no adverse hemodynamic interaction between ACE inhibition and this type of antiplatelet compound.154 Definitive resolution of the therapeutic implications of the aspirin-ACE inhibitor interaction and determination of alternative therapy, if any, in HF awaits the results of additional studies.