Does Warfarin Reduce Stroke?

abstract & commentary

Source: Pullicino PM, et al. Stroke in patients with heart failure and reduced left ventricular ejection fraction. Neurology 2000;54:288-294.

The role of warfarin in the prevention of stroke from atrial fibrillation (AF) is well established. Anticoagulation for stroke prevention in cardiac failure, an even more common condition than AF, is also commonly used. No randomized trial has been done to support this practice, however. Pullicino and colleagues review the literature outlining the risk of stroke in cardiac failure, the possible benefits of prophylactic warfarin therapy, and the proposed randomized studies to further explore these questions.

Stasis of blood in a poorly functioning left ventricle may lead to thrombus formation and embolic stroke. The relative risk of stroke is as much as 4.1 times higher among patients with cardiac failure than those without. This risk increases proportionately with decreased ejection fraction. In the Survival and Ventricular Enlargement (SAVE) Trial, there was an 18% increment in stroke risk for every 5% decline in ejection fraction (EF). In the SOLVD Trial (among women), there was a 58% increase in the risk of thromboembolic events for every 10% decrease in EF. No significant increase in stroke rate among men was observed. Comorbid factors such as advanced age, hypertension, and diabetes magnify stroke risk in cardiac failure just as these factors also increase stroke risk in AF.

Cardiac failure is increasingly common, in part due to prolonged survival rates. While patients with severe cardiac failure previously had mortality rates as high as 17% per year, the addition of drug therapy with angiotensin-converting enzyme inhibitors and beta-blockers has reduced this to approximately 5% per year. First-ever strokes attributable to cardiac failure may be as many as 72,000 per year.

Studies of warfarin in patients with a history of myocardial infarction (MI) show relative decreases in stroke risks in the range of 40-55% compared with placebo, about twice the efficacy of aspirin. Data from the SOLVD study suggest that warfarin may also reduce mortality by about 20%. This benefit is likely to be present whether the cardiomyopathy is ischemic or not and is probably independent of the coexistence of AF.

As Pullicino et al outline, warfarin in patients with cardiac failure may have benefit in three main areas: overall mortality reduction, primary stroke prevention, and secondary stroke prevention. But is warfarin justified over aspirin given its possible serious hemorrhagic complications? For first-ever strokes, with an occurrence rate of 1.5% per year and a relative risk benefit of 30%, the absolute risk reduction would be 0.45% per year. This is not significantly higher than the rate of intracerebral hemorrhage (ICH) with warfarin therapy (about 0.3%). For recurrent stroke, the data are more convincing. Because this event is more frequent (about 9% per year), the absolute risk reduction would be 3%, outweighing the risk of ICH by 10-fold.


Given the relatively small potential benefit of warfarin for first-ever stroke and the infrequent occurrence of this event among the population at risk, randomized studies would require as many as 10,000 patients to achieve statistical significance. Such limitations have led investigators to suggest using composite endpoints (including cardiac outcomes and overall mortality) or to restrict study to the patients at highest risk (such as those with prior stroke). The table outlines the main features of the proposed Antiplatelet Therapy in Chronic Heart Failure (WATCH) and the Wafarin Aspirin in Reduced Cardiac Ejection Fraction (WARCEF) studies. To decrease sample size, WATCH will include MI in a composite endpoint and WARCEF will include only patients with prior stroke. Pooled data from these two studies combined may have the statistical power to answer specific questions about stroke risk. Neurology Alert eagerly awaits these results. —azs

Table-Main Features of the WARCEF and WATCH Studies
Blinding Aspirin and clopidogrel blinded; warfarin unblinded Blinded
Study Three (warfarin,aspirin,clopidogrel) (warfarin, aspirin)
Target INR 2.5-3.0 2.5-3.0
NYHA class for entry II, III, or IV I, II, or III
Entry objection fraction £ 30% £ 30%
Echo entry criteria LV end diastolic dimension £ 6 cm (men) or £ 5.6 cm (women) and fractional shortening < 22% Wall motion index £ 2
Primary endpoint Death, stroke, and myocardial infraction All-cause mortality and stroke
Study duration 5-y with 3-y enrollment 5-y with 3-y enrollment
Sample size 4500 patients 2860 patients
INR= International Normalized Ratio; LV = left ventricular; NYHA = New York Heart Association; WARCEF = Warfarin Versus Aspirin in Reduced Cardiac Ejection Fraction; WATCH = Warfarin and Antiplatelet Therapy in Chronic Heart Failure
Source: Pullicino PM, et al. Neurology 2000;54:293.

The relative risk of stroke is higher among patients with cardiac failure than those without it. The risk increases proportionately with:

a. gender.

b. age.

c. decreased ejection fraction.