Atrial Fibrillation and Flutter

AF is the most common serious arrhythmia and is a major risk factor for stroke and death. It accounts for nearly one-half of all cardiac causes of stroke and about one-quarter of strokes in the elderly. Strokes associated with AF tend to be more severe than strokes due to other etiologies, and the 30-day mortality is approximately 25 percent.

The prevalence of AF is age dependent, ranging from 0.1 percent among adults younger than 55 years to 9 percent in those 80 years or older. The population prevalence of AF is increasing due to an aging population. AF typically occurs in patients with underlying cardiac disease (e.g,. hypertension, valvular heart disease, congestive heart failure, coronary disease, cardiomyopathy, mitral valve prolapse, mitral annular calcification, and cardiac tumors), but may also occur as “lone AF” in young patients who have no cardiac disease. AF may be paroxysmal (defined as a self-terminating episode lasting less than 7 days), recurrent (two or more episodes), persistent (more than 7 days), or permanent (cardioversion failed or not attempted). Reversible or temporary causes include alcohol, surgery, hyperthyroidism, acute myocardial infarction, pulmonary embolism, and pericarditis, among others.

The average annual risk of stroke in individuals with AF is 5 percent and is heavily dependent on age and the presence of additional risk factors ( Table 5-3 ). The most important predictor of stroke risk in patients with AF is a history of thromboembolism (i.e., previous TIA, stroke, or systemic arterial embolism). Other independent risk factors for stroke in patients with AF are increasing age, hypertension, congestive heart failure, and diabetes mellitus. Other factors that have been associated with increased stroke risk in some studies include female sex, systolic hypertension, and left ventricular dysfunction. Commonly used risk stratification tools are the CHADS ₂ scale and the CHA ₂ DS ₂ -VASC scale, which predict the risk of stroke in patients with AF based on the presence of additional risk factors. The CHADS ₂ scale ranges from 0 (low stroke risk, 1.9% per year) to 6 (high stroke risk, 18.2% per year) points. Online calculators are available that show the estimated annual stroke risk for patients without anticoagulant therapy and the risk reduction afforded by aspirin.

Echocardiographic features that have been used for risk stratification in patients with AF include left ventricular systolic dysfunction, atrial thrombus, dense spontaneous echo contrast or reduced blood flow velocity within the left atrium or left atrial appendage on TEE, and aortic arch atheroma. Left atrial size was previously considered not to predict the risk of thromboembolism, although a large observational study found that left atrial diameter measured on TTE was a predictor of all-cause mortality and of ischemic stroke (the latter in women only). TEE is the method of choice for evaluating the left atrial appendage, the site at which most thrombi form, as well as the left atrium. In a prospective study of patients with AF considered on clinical grounds to be at high risk, stroke occurred at a rate of 18 percent per year in those with dense spontaneous echo contrast who were treated with low-dose warfarin (international normalized ratio [INR] 1.2 to 1.5) plus aspirin compared to 4.5 percent for those on monotherapy with dose-adjusted warfarin with a goal INR of 2.0 to 3.0. The prevalence of thrombus in the left atrial appendage was similar initially in the two treatment groups (10 to 12%) when TEE was performed more than 2 weeks after study entry, but atrial thrombus was present in 6 percent of those on warfarin compared to 18 percent of those on combination therapy. Absence of thrombus predicted a low rate of ischemic events (2.3% per year), and the presence of thrombus predicted a high rate (18%).

The risk of stroke in AF is significantly reduced by anticoagulation. A meta-analysis showed that warfarin reduced stroke risk by 62 percent overall compared with placebo. Absolute risk reductions were higher for secondary prevention (8.4% per year) than primary prevention (2.7%). These percentages translate into numbers needed to treat (NNT) of 12 and 37, respectively. Although more intracranial hemorrhages (ICHs) occurred in the warfarin group (0.3% per year) compared to the placebo group (0.1%), this was not statistically significant. Major extracranial hemorrhage occurred in 0.6 percent per year of patients on placebo, with a relative risk for those on warfarin of 2.4 (absolute risk increase, 0.3% per year). The total number of patients in six trials that assessed dose-adjusted warfarin to placebo was 2,900, with an average follow-up of 1.7 years. The risk reduction with warfarin was based on intention-to-treat analyses; the on-treatment analysis showed a relative risk reduction in stroke of more than 80 percent.

Adjusted-dose warfarin was also compared to aspirin in five trials including 2,837 individuals. Excluding one study because the range of the INR was wide (2.0 to 4.5), the relative risk reduction for warfarin compared to aspirin was 46 percent. Patients with AF at high risk of stroke generally still benefit from anticoagulation even after sinus rhythm has been restored.

The use of aspirin compared to placebo has also been addressed in several trials. The prescribed dose of aspirin has ranged from 25 to 1,200 mg daily, with more than 3,000 patients studied; average follow-up was 1.5 years. In patients receiving placebo, the stroke incidence was 5.2 percent per year for primary prevention and 12.9 percent for secondary prevention; aspirin reduced stroke risk by 22 percent, resulting in numbers needed to treat of 67 and 40, respectively. All-cause mortality was not reduced. Aspirin’s benefit in these patients may be to prevent nondisabling stroke that is not of cardioembolic origin. Therefore, published guidelines strongly recommend warfarin rather than aspirin for stroke prevention in individuals with AF who are at high risk.

In practice, despite the clear benefit of warfarin for stroke prevention in patients with AF, this medication is frequently underutilized. Warfarin is a difficult medication for patients because of the inconvenience of INR monitoring, drug and food interactions, and bleeding risks. Physicians also frequently overestimate the bleeding risks but underestimate the benefits of warfarin compared with aspirin.

Individual patient preferences, knowledge, and attitudes affect compliance with long-term anticoagulation therapy. Among AF patients taking warfarin in one study, about one-half did not know that AF was a risk factor for stroke and could not state why they were taking warfarin; ethnic differences in knowledge about their diagnosis and treatment were also identified. Methods to encourage compliance with appropriate antithrombotic prophylaxis include use of a patient decision aid. Home INR finger-stick devices for self-monitoring may increase the time that patients spend in the therapeutic INR range.

Bleeding is the major concern with anticoagulant therapy. The average risk of major bleeding in clinical trials is 1.3 percent per year with warfarin compared to 1 percent with aspirin or placebo. The Stroke Prevention in Atrial Fibrillation study had a higher rate of major bleeding at 2.3 percent on warfarin and 1.1 percent per year on aspirin. Rates of ICH were 0.9 percent and 0.3 percent per year, respectively. Age older than 75 years increased the risk of major hemorrhage to 4.2 percent per year, and only in this age group was intensity of anticoagulation predictive of risk.

With the exception of some patients with lone AF, all patients with AF (regardless of whether paroxysmal, persistent, or permanent) require some form of antithrombotic therapy, unless contraindicated. It remains necessary to individualize management strategies for specific patients, taking into account compliance, risk of bleeding complications, and other medical conditions. Risk stratification is essential to determine optimal treatment with warfarin or aspirin. Many schemes have been devised for identifying patients with AF unassociated with valvular heart disease who are at high, moderate, or low risk of stroke. According to the American Heart Association guidelines, high-risk factors are previous stroke, TIA, or systemic embolism; mitral stenosis; and prosthetic heart valves. Moderate risk factors include age older than 75 years; hypertension; heart failure; left ventricular ejection fraction less than 35 percent; and diabetes. Warfarin is recommended for patients with any high-risk factor or more than one moderate-risk factor. This means that all patients with a previous ischemic stroke or TIA are considered at high risk and require anticoagulation for secondary stroke prevention unless contraindicated. Warfarin or aspirin (81 to 325 mg) is recommended for those with only one moderate-risk factor. Aspirin alone (81 to 325 mg) is considered sufficient for patients without any of these risk factors.

For most patients receiving warfarin for AF (excluding mechanical heart valves), the target INR is 2.5 (range 2.0 to 3.0). The INR should be monitored closely; usually weekly initially and then monthly once stable. Stroke risk increases exponentially as the intensity of anticoagulation declines below 2.0.

In addition to protecting against stroke, antithrombotics can attenuate stroke severity. Patients taking warfarin at the time of stroke have, on average, less-disabling strokes compared to individuals taking aspirin or no antithrombotic therapy, and stroke severity is negatively correlated with INR at stroke onset. Table 5-4 gives a summary of the indications for warfarin in secondary stroke prevention for patients with selected cardiac conditions.

Dual antiplatelet therapy (aspirin plus clopidogrel) was investigated in a randomized trial and found to be inferior to warfarin for stroke prevention in AF, but associated with a similar rate of adverse bleeding events compared to warfarin; the combination was superior to aspirin alone and therefore could be considered in some high-risk patients in whom warfarin is contraindicated.

If warfarin therapy needs to be interrupted for surgical procedures, temporary discontinuation for up to 1 week is usually considered reasonable for patients without mechanical heart valves. However, since this practice is associated with increased stroke risk, it must be individualized. Bridging heparin therapy may be substituted in high-risk patients during periods of warfarin interruption.

A major development in AF treatment has been the recent arrival of a new generation of oral anticoagulant drugs that have emerged as alternatives to warfarin: the direct thrombin inhibitor dabigatran and the factor Xa inhibitors rivaroxaban and apixaban. There have been four pivotal phase III randomized controlled trials testing these new agents in patients with non-valvular AF. All three agents appear to be at least as effective as warfarin for stroke prevention and are associated with a lower incidence of intracranial hemorrhage. In contrast to warfarin, these drugs have a rapid onset of action, short half-life, fewer drug interactions, lack of food interactions, and do not require INR monitoring. Regular patient follow-up is still necessary to monitor adherence and renal function. No specific antidotes to these drugs are available yet to reverse bleeding. Contraindications to the use of these agents include severe renal failure and mechanical heart valves. If the outcomes of the new anticoagulants prove to be as good in real-world practice as in clinical trials, these agents represent a major therapeutic advance. Other new anticoagulants are under investigation.

In addition to medical therapy for stroke prevention in AF, interventional techniques are being investigated. These include percutaneously implanted left atrial appendage occlusive devices and surgical resection of the left atrial appendage, given that 91 percent of thrombi are localized at that site. Carotid artery endovascular devices to filter emboli are also under investigation.

It is clear that cardioversion of AF to sinus rhythm (either pharmacologic or electric) does not reduce the risk of stroke and therefore does not obviate the need for continued anticoagulation therapy for stroke prevention.

AF occurring in the postoperative setting following cardiac surgery is fairly common and usually self-limited. Anticoagulation is reasonable if AF persists for more than 48 hours, but it may not need to be continued long-term if sinus rhythm is restored. Similarly, other conditions associated with transient AF (e.g., alcohol, thyrotoxicosis) usually do not need long-term antithrombotic prophylaxis.

In patients with atrial flutter, the risk of thromboembolism is less than that of AF but higher than for patients in sinus rhythm. These patients often eventually develop AF. For practical purposes, the antithrombotic treatment recommendations are similar to those for AF.

Brief subclinical AF or atrial tachyarrhythmias are emerging risk factors for stroke, as demonstrated by pacemaker studies. For example, one study monitored 2,580 patients without known AF in whom a pacemaker or defibrillator had been implanted. Subclinical episodes of high atrial rate (>190 beats per minute for>6 minutes) were found in 10.1 percent within 3 months of monitoring, and this finding was a significant independent predictor of clinical AF and ischemic stroke or systemic embolism during follow-up.

In patients with ischemic stroke presenting in sinus rhythm, Holter monitoring for 24 to 72 hours detects paroxysmal AF in about 5 percent of patients. However, AF can be difficult to detect because it is frequently intermittent and asymptomatic. There is increasing evidence that prolonged electrocardiographic (ECG) monitoring, through external or implanted recording devices, can improve the detection of occult paroxysmal AF in patients with strokes of undetermined etiology ; several studies show a rate of detection of 5 to 20 percent.

Cardioversion in Atrial Fibrillation or Flutter

Cardioversion (electric or pharmacologic) undertaken to convert AF back to sinus rhythm is associated with an increased risk of thromboembolism. It is therefore recommended that warfarin (INR 2.0 to 3.0) be given for at least 3 weeks prior to elective cardioversion of patients who have been in AF for 2 days or more or when the duration of AF is unknown; warfarin should be continued until normal sinus rhythm has been maintained for 4 weeks.

For patients requiring immediate cardioversion, intravenous heparin is recommended concurrently followed by warfarin for at least 4 weeks. Alternatively, TEE prior to cardioversion can be performed; if no left atrial appendage thrombus is detected, cardioversion can occur as soon as the patient is anticoagulated and continue for at least 4 weeks. If a left atrial thrombus is detected on TEE, warfarin is recommended for at least 3 weeks prior to cardioversion and may need to be continued for a longer duration afterward.

The recommendations for cardioversion in atrial flutter are the same as for AF. Atrial flutter has been studied less extensively than AF. The total incidence of acute and chronic events was found to be 7 percent over an average period of 26 months. Prior TEE is not an adequate predictor of those at risk; a total of 3 of 41 patients who had no left atrial clot developed ischemic neurologic syndromes within 48 hours of elective cardioversion.

Chronic Sinoatrial Disorder (Sick Sinus Syndrome)

Similar to atrioventricular block, chronic sinoatrial disease (sick sinus syndrome) usually presents with syncope and dizziness; however, those with sinoatrial disorder have a much higher rate of systemic emboli than patients with atrioventricular block, and this rate is not mitigated by pacemaker insertion. There is no significant difference in death from any cause between treatment with single-lead atrial pacing (AAIR) and dual-chamber pacing (DDDR). Single-lead atrial pacing is associated with a higher incidence of paroxysmal AF and a twofold increased risk of pacemaker reoperation.

A Cochrane review concluded that physiologic (primarily dual-chamber) pacing had a statistically significant benefit in preventing the development of AF compared with ventricular pacing. Patients with the “brady-tachy” form of the disorder are at higher risk of developing AF and stroke, and warfarin treatment should be considered for these patients.

Cardiomyopathies

A new definition and classification scheme for cardiomyopathy was proposed in 2006 and updated in 2008. Cardiomyopathies are defined as a heterogeneous group of diseases of the myocardium associated with mechanical or electric dysfunction (or both) that usually, but not invariably, exhibit inappropriate ventricular hypertrophy or dilatation and are due to a variety of causes that frequently are genetic. Specifically excluded are those diseases of the myocardium secondary to congenital or valvular heart disease, systemic hypertension, or atherosclerotic coronary disease.

The cardiomyopathies are divided into two major groups based on predominant organ involvement. The primary cardiomyopathies are those solely or predominantly confined to heart muscle; genetic, mixed, and acquired forms are recognized. Both hypertrophic and dilated cardiomyopathies are considered primary diseases. Also included are the ion-channel disorders, in which there is a primary electric disturbance without structural cardiac pathology; these disorders are further considered in the section devoted to syncope.

Secondary cardiomyopathies involve skeletal or smooth muscle in addition to cardiac muscle. Neuromuscular or neurologic causes include Friedreich ataxia, Duchenne or Becker muscular dystrophy, Emery–Dreifuss muscular dystrophy, neurofibromatosis, and tuberous sclerosis. The secondary cardiomyopathy classification does not include infective processes, such as Chagas disease or infection with human immunodeficiency virus.

In North America, the most common cardiomyopathy is hypertrophic cardiomyopathy, which is an autosomal dominant disease affecting 1 in 500 persons. The disorder is a major cause of sudden cardiac death in athletes but is compatible with survival until old age. Mortality rates have been estimated at around 1 percent for persons age 16 to 65, 4 percent over the next decade, and 5 percent for those older than 75 years. Stroke risk in hypertrophic cardiomyopathy was studied in a group of 900 patients. Stroke occurred in 44 patients over a period of 7 years, with an annual incidence of 0.8 percent. In patients with hypertrophic cardiomyopathy, left ventricular outflow tract obstruction at rest predicts the development of heart failure and death.

There are considerable geographic variations in the causes of cardiomyopathy. In Latin America, American trypanosomiasis (Chagas disease) is common. Cardioembolic stroke has been increasingly well documented as a complication, and most are in the anterior circulation. In Chagas cardiomyopathy, the apical region of the left ventricle is the typical site for formation of thrombosis or aneurysm. Echocardiography reveals an apical aneurysm in around one-third of patients and a mural thrombus in about 10 percent. Left ventricular diastolic dysfunction is present in nearly one-half of patients. The ECG is abnormal in two-thirds, including right bundle branch block, left anterior fascicular block, and AF. Oral anticoagulation has been recommended for all individuals with Chagas disease and risk factors for cardioembolism.

In Africa, the major cardiomyopathy is the dilated type, but peripartum cardiomyopathy is ubiquitous with an incidence between 1 in 100 and 1 in 1,000. Regional variations include endomyocardial fibrosis restricted to the tropical regions of East, Central, and West Africa. The incidence of human immunodeficiency virus–associated cardiac disease, including cardiomyopathy, is increasing, in contrast to developing countries where the availability of antiretroviral therapy has reduced the incidence of myocarditis.

In young adults, arrhythmogenic right ventricular dysplasia with cardiomyopathy is another rare hereditary disorder causing sudden death. In a study of the natural history in 130 patients, age at onset of symptoms averaged 32 years.

Patients with dilated cardiomyopathy have a higher incidence of embolic events including systemic embolism and stroke secondary to ventricular mural thrombi, and therefore warfarin or antiplatelet therapy should be considered for secondary stroke prevention.

Myocardial Infarction and Left Ventricular Dysfunction

Patients with coronary artery disease have an increased stroke risk, particularly within the first month after myocardial infarction (MI). Mechanisms include embolism from left ventricular mural thrombus and the development of AF (which occurs in up to 20% of patients following MI).

A community-based study of 2,160 patients hospitalized between 1979 and 1998 found stroke risk during the 30 days after a first MI to be increased 44-fold, and it remained two to three times higher than expected during the subsequent 3 years. Of note, the 20-year duration of the study enabled the conclusion to be drawn that acute MI treatment with thrombolysis did not reduce stroke risk. Overall, stroke risk following MI is approximately 1 percent during the first month and about 2 percent at 1 year. For a non-ST elevation acute coronary syndrome, the early stroke risk is only 0.7 percent at 3 months. In large randomized trials of aspirin versus the combination of aspirin and clopidogrel in patients with MI or acute coronary syndrome, the stroke rate ranged between 0.9 and 1.7 percent.

In a meta-analysis, predictors of stroke following MI included advanced age, diabetes, hypertension, previous stroke or MI, anterior MI, AF, heart failure, and nonwhite race. Left ventricular thrombus develops in about one-third of individuals in the first 2 weeks following an anterior MI, posing an increased risk of embolization which is reduced by anticoagulation.

The current recommendation, in the absence of thrombolytic therapy, is that after acute MI, heparin should be initiated and followed by warfarin for 3 months in patients considered to be at increased risk of embolism. High-risk patients are those with severe left ventricular dysfunction, congestive heart failure, a history of pulmonary or systemic embolism, echocardiographic evidence of mural thrombosis, or the presence of AF. Because of the increased frequency of mural thrombus in anterior as opposed to inferior myocardial infarctions, patients with an anterior Q-wave infarction should also receive heparin followed by warfarin.

In patients with TIA or ischemic stroke related to an acute MI in which left ventricular mural thrombus is identified, oral anticoagulation is recommended for at least 3 months and up to 1 year (goal INR, 2.0 to 3.0), perhaps in addition to aspirin for coronary artery disease (up to 162 mg/day).

Stroke risk is inversely proportional to left ventricular ejection fraction. In a study of 2,231 patients with left ventricular dysfunction following acute MI, those with an ejection fraction less than 29 percent had a stroke risk that was nearly double that of patients with ejection fraction exceeding 35 percent; the annual stroke rate was 1.5 percent overall.

Congestive heart failure carries a two- to threefold increase in the relative risk of stroke. Among patients enrolled into heart failure trials, the overall annual stroke risk ranged between 1.3 and 3.5 percent; most patients were taking aspirin or warfarin in these trials. In the absence of clinically overt heart failure or MI, the presence of asymptomatic left ventricular systolic dysfunction is also an independent risk factor for stroke.

The WARCEF Trial compared warfarin (target INR, 2.0 to 3.5) versus aspirin (325 mg per day) in 2,305 patients with reduced left ventricular ejection fraction (35% or less) who were in sinus rhythm and followed for up to 6 years. There was no significant overall difference between groups in the composite primary outcome of ischemic stroke, intracerebral hemorrhage, or death from any cause. A reduced risk of ischemic stroke with warfarin was offset by an increased risk of major hemorrhage. Similarly, in a meta-analysis of four randomized trials comprising 4,187 patients with heart failure in sinus rhythm, warfarin was found to reduce ischemic stroke risk compared to aspirin by 0.74 percent per year but increased major bleeding by 0.99 percent per year. There is therefore no convincing evidence that warfarin is superior to aspirin in stroke prevention for patients with reduced left ventricular ejection fraction.

Rheumatic Heart Disease

There is a well-established association between stroke and rheumatic heart disease (especially mitral valve stenosis), particularly in the setting of atrial fibrillation and atrial thrombus. Current Class I recommendations strongly favor the use of long-term warfarin (target INR, 2.0 to 3.0) in patients with rheumatic mitral valve disease who have a history of systemic embolism or who develop AF, either chronic or paroxysmal. It is also recommended that warfarin be given to patients in normal sinus rhythm if the left atrial diameter is in excess of 5.5 cm, regardless of a history of embolism.

Atrial Myxoma

Atrial myxomas have long been recognized as an uncommon cause of cerebral embolism. A French hospital reviewed experience with 112 cases collected over a period of 40 years. Women outnumbered men nearly 2 to 1, and ages ranged from 5 to 84 years. The presenting symptoms were cardiac (67%), constitutional (34%), and embolic (29%). Younger and male patients were more likely to experience embolic events. A literature review identified ischemic stroke as the most common neurologic manifestation. Syncope, psychiatric symptoms, headache, and seizures also occur. A rare delayed complication is that of multiple cerebral aneurysm formation.

Marantic (Nonbacterial Thrombotic) Endocarditis

There are several causes of marantic (nonbacterial thrombotic) endocarditis, a condition characterized by platelet aggregates or vegetations on previously undamaged heart valves (most often aortic and mitral) without evidence of bacteremia. It is a rare condition often associated with hypercoagulable states or advanced malignancies such as adenocarcinomas.

The widespread availability of echocardiography has facilitated recognition of this disorder in patients with cancer. A prospective study of 200 unselected ambulatory patients with solid tumors found vegetations in 19 percent compared to 2 percent in controls. Vegetations were seen in 50 percent of patients with pancreatic cancer, 28 percent of those with lung cancer, and 19 percent of patients with lymphoma. Only two patients had cerebral events. Brain MRI typically shows numerous lesions of various sizes in multiple arterial territories.

At one cancer center, 96 stroke patients were assessed, and TTE was performed in two-thirds. An embolic mechanism was thought to be causative in over half; the heart was implicated in 14 patients, but nonbacterial thrombotic endocarditis in only 3. Stroke of embolic origin carried a dismal prognosis with life expectancy of just over 2 months; treatment had no apparent influence on outcome.

Other Echocardiographic Abnormalities Linked to Stroke