Chronic atrial fibrillation (AF) is the second most common cardiac arrhythmia after premature ventricular beats and the most common sustained cardiac arrhythmia. Incidence rises markedly with age, and AF is seen in five percent of patients over the age of 69.^, There are an estimated 2.2 million people in the United States with AF, with a median age of about 75 years. Patients with AF are at increased risk of stroke and of death. In the Framingham cohort study, the risk of stroke in patients with nonvalvular (nonrheumatic) AF was 5.6 times greater than in comparably aged patients in sinus rhythm; in patients with AF and rheumatic valvular disease it was 17 times higher. AF is associated with reduced life expectancy with an approximate doubling of all-cause mortality.^2, Analysis of the SOLVD trials found that the presence of atrial fibrillation in patients with asymptomatic and symptomatic left ventricular systolic dysfunction is associated with an increased risk for all-cause mortality [relative risk 1.34]. Treatment of patients with AF has two principle objectives. One is to relieve symptoms through restoration of sinus rhythm or control of ventricular rate; the other is to use prophylactic therapy to reduce the risk of stroke.^,

Two treatment strategies have been suggested for the management of atrial fibrillation, rhythm control (i.e. maintenance of sinus rhythm with electrical/pharmacologic cardioversion and catheter based therapies) and rate control without attempts to maintain sinus rhythm. The effect of each strategy on overall mortality is the focus of a large randomized ongoing clinical trial – AFFIRM, which recently reported a 16% increase risk of death (95% CI 0–32%) in the patients randomized to rhythm control. A smaller study, the Pharmacological Intervention in Atrial Fibrillation (PIAF) trial was also recently completed and published. It was a randomized trial in 252 patients with atrial fibrillation of between 7 days and 360 days duration, which compared rate (group A, 125 patients) with rhythm control (group B, 127 patients). The primary study endpoint was improvement in symptoms related to atrial fibrillation. Over the entire observation period of 1 year, a similar proportion of patients reported improvement in symptoms in both groups (76 responders at 12 months in group A vs 70 responders in group B, p=0.317). Walking distance in a 6 min walk test was better in group B compared with group A, but assessment of quality of life showed no differences between groups. The incidence of hospital admission was higher in group B (87 [69%] out of 127 vs 30 [24%] out of 125 in group A, p=0.001). Adverse drug effects more frequently led to a change in therapy in group B (31 [25%] patients compared with 17 [14%] in group A, p=0.036). Thus, based on the preliminary, unpublished report of the AFFIRM trial and the PIAF study, the therapeutic strategy of rate control and anticoagulation is the treatment strategy of choice for a majority of subjects.

The following is standardized nomenclature for delineating different forms of atrial fibrillation. Acute atrial fibrillation [onset within 24-48 hours] is distinguished by a high chance of successful cardioversion, either spontaneously or with electrical/pharmacologic therapy, and the lack of a requirement for anticoagulation prior to cardioversion. Acute atrial fibrillation also includes patients who develop atrial fibrillation after cardiac surgery. Chronic atrial fibrillation is defined by the presence of the arrhythmia for greater than 48 hours and includes patients with atrial fibrillation for an unknown duration. Chronic atrial fibrillation may be paroxysmal, persistent/chronic or permanent/resistant.

The initial evaluation of a patient with atrial fibrillation should include (1) historical information with attention to the presence of concurrent symptoms of congestive heart failure, angina and palpitations, previous thromboembolic events, and a history of hypertension, diabetes, congestive heart failure or valvular heart disease (2) physical examination (3) electrocardiogram (4) initial laboratory examination including an electrolyte panel, coagulation profile and TSH and (5) transthoracic echocardiogram to assess valve function and left atrial size.

AF may present as an asymptomatic arrhythmia or, at the other extreme, in a patient who is severely symptomatic with hemodynamic compromise. The previous focus of published literature on antithrombotic therapy for AF has been on demonstrating its value in decreasing stroke and death. The effects of individual symptomatic treatments, including rate control medicines, atrioventricular node ablation, arrhythmia termination and prevention therapies (cardioversion, antiarrhythmic medications or operation) on morbidity and mortality are unknown.^, In the absence of clinical trial data on optimal therapy for arrhythmia related symptoms, assessment of the setting and severity of such symptoms is critical for determining the appropriateness of therapy for an individual patient.

Symptoms usually consist of palpitations and lightheadedness. The loss of synchronous atrial contraction and decreased diastolic filling time during tachycardia can lead to a hemodynamically significant decrease in cardiac output in patients with left ventricular dysfunction, valvular or coronary heart disease. Angina pectoris may be precipitated as a result of increased myocardial oxygen demand and shortened diastolic coronary blood flow. In SPAF-III, among 1,411 patients, atrial fibrillation was initially asymptomatic in 44 percent, while 32 percent had palpitations, 10 percent had congestive heart failure, 2 percent presented with a stroke, and 2 percent with syncope. In 11 percent, the presenting symptom was unknown. While these study subjects were not necessarily representative of the general population with AF, the data support the conclusion that symptoms from AF are usually not disabling.

Physical findings in patients with AF include an irregularly irregular ventricular rhythm, a variation in intensity of the first heart sound and absence of a waves in the jugular venous pulse. At faster ventricular rates, the auscultated apical rate may exceed the palpable radial rate, owing to failure of many of the ventricular contractions to generate a palpable peripheral pulse. Since AF is associated with underlying heart disease and heart failure, the physical examination should focus on the presence of murmurs and gallops. Occult or manifest thyrotoxicosis should be considered in patients with recent onset AF and physical examination should include careful assessment of the thyroid.

Electrolyte abnormalities may precipitate or exacerbate an underlying tendency towards AF and since correction of these abnormalities is simple and effective, routine evaluation is indicated. A screening TSH may identify hyperthyroidism as the etiology of AF. This can be particularly important in the elderly, in whom clinical symptoms and signs of hyperthyroidism may be absent. In one large survey of patients who presented with AF, occult thyroid disease (especially hyperthyroidism) has been identified as a principle underlying cause in 3.5 percent. Thus, routine thyroid function tests, while having a low yield, may identify a potentially treatable cause of the dysrhythmia and should be performed. Finally, before anticoagulation is begun, knowledge of the patient’s coagulation status is important in judging response to therapy and in preventing complications.

The twelve lead EKG confirms the diagnosis of AF and provides information about underlying heart disease and coexisting conduction disease; it should be performed in all patients with AF.

A transthoracic echocardiogram provides anatomic and functional information about underlying valvular heart disease, the presence and degree of left ventricular hypertrophy and an assessment of left ventricular function. Left atrial size may predict the outcome of cardioversion and subsequent maintenance of sinus rhythm.^, Echocardiographic data are also useful in making an assessment of embolic risk. Moderate to severe left ventricular systolic dysfunction, as demonstrated by two-dimensional echocardiography, is a strong predictor of stroke risk [RR 2.5], independent of clinical risk factors. Additionally, normal cardiac function implies that the risk of systemic embolism is low and long-term anticoagulation may not be required (see Anticoagulation, below).

A treatment approach guided by the results of transesophogeal echocardiography was a focus of a recent large clinical trial. The conventional treatment strategy for patients with atrial fibrillation who are to undergo electrical cardioversion is to prescribe warfarin for anticoagulation for three weeks before cardioversion. It has been proposed that if transesophageal echocardiography reveals no atrial thrombus, cardioversion may be performed safely after only a short period of anticoagulant therapy: In this multicenter, randomized, prospective clinical trial, 1222 patients with atrial fibrillation of more than two days' duration were assigned them to either treatment guided by the findings on transesophageal echocardiography or conventional treatment. The composite primary end point was cerebrovascular accident, transient ischemic attack, and peripheral embolism within eight weeks. Secondary end points were functional status, successful restoration and maintenance of sinus rhythm, hemorrhage, and death. There was no significant difference between the two treatment groups in the rate of embolic events. At eight weeks, there were no significant differences between the two groups in the rates of death or maintenance of sinus rhythm or in functional status. Thus, for a majority of outpatients with atrial fibrillation who are scheduled to undergo cardioversion a TEE guided approach does not appear to be superior to standard therapy.

Control of resting and exercise heart rate during AF is an important therapeutic goal, which may markedly reduce symptomatology. Restoration and maintenance of sinus rhythm may not be indicated or possible in many patients, in whom rate control and anticoagulation are essential. Patients with AF should have their heart rates reduced to physiologically normal values. In the middle aged population, this might be a resting heart rate < 90 beats/min and an exercise heart rate < 140 beats/min. Chronic rate control can be achieved with several different classes of agents, none of which is clearly preferable, and patient symptoms and co-morbid conditions should be taken into account when selecting an agent for rate control.

Digoxin has been used for over a century to acutely reduce the heart rate in AF. Numerous studies have demonstrated, however, that the predominant effect of digoxin is on the resting heart rate and that its mechanism of action is by activating the parasympathetic nervous system, causing sinus slowing and AV node inhibition. Despite controlled resting heart rates, patients on digoxin can have very rapid responses during exertion or with paroxysms of fibrillation when sympathetic tone is high. In fact, there is very little difference in exercise heart rates with or without digoxin therapy, and increasing digoxin dosage (and blood levels) does not seem to add any benefit with respect to nonresting rate control. Sedentary patients may be well controlled on digoxin alone, but in many patients the drug is ineffective in controlling heart rate, particularly during exercise.

Calcium channel blockers have been shown to reduce resting and exercise heart rates in AF. In a placebo-controlled crossover trial of verapamil, Lang demonstrated reduction in exercise and resting heart rates and improved exercise capacity. Several uncontrolled studies have shown improved heart rate response at rest and with exercise using diltiazem. High dose calcium channel blockers, with or without digoxin, may cause prolonged pauses, which may become symptomatic. Drug combinations at lower doses may affect adequate rate control without symptomatic bradycardia. These agents should be used with caution in the presence of heart failure. Diltiazem should be preferred to verapamil in patients with mild to moderate left ventricular dysfunction. Because of concern about the cardiovascular toxicity of short-acting calcium channel blockers, long-acting preparations should be used. These are the preferred agents for rate control in patients with symptomatic obstructive lung disease.

Several controlled and uncontrolled trials have shown that beta blockers effectively reduce exercise heart rates in patients with AF.^, In a randomized, double-blind crossover trial, Dibianco demonstrated that adding naldolol to digoxin effectively controlled exercise heart rate in patients with AF. No study has conclusively demonstrated improved exercise capacity with beta blockers. As with calcium channel blockers, symptomatic bradycardia is a concern. A meta-analysis suggests that for prophylactic use of beta-adrenergic blockers reduces the incidence of postoperative atrial fibrillation and should be administered before and after cardiac surgery to all patients without contraindication.

The use of amiodarone solely for the chronic management of rapid ventricular response in AF has not been evaluated in a controlled trial. However, in several trials evaluating patients with refractory or resistant AF, a consistent and significant slowing of ventricular response has been uniformly reported. While amiodarone is clearly efficacious in maintaining sinus rhythm in patients with AF, especially if they are resistant to other therapies, its toxicity precludes recommending it simply for rate control when other less toxic agents are available. Nonetheless, in patients with left ventricular dysfunction who cannot tolerate standard therapy for rate control, amiodarone may be beneficial. Prophylactic amiodarone should be considered in patients at high risk for postoperative atrial fibrillation (for example, patients with previous atrial fibrillation or mitral valve surgery).

Ablation of the AV node or His bundle using radiofrequency energy delivered by catheter and insertion of a permanent rate responsive pacemaker is an alternative to medication. Since the procedure has not been studied in a controlled trial, and since it creates permanent heart block, it should be considered only for patients who have failed all other forms of medical therapy.

Prospective trials have confirmed that the risk of stroke in patients with nonvalvular AF is large - about five percent per year. More importantly, these trials have convincingly demonstrated that low dose warfarin therapy can substantially reduce stroke with a minimal risk of significant hemorrhage in properly chosen patients.^,,,, In these randomized trials, warfarin consistently decreased the risk of stroke in patients with AF - a 68 percent reduction in risk - with virtually no increased incidence of major bleeding. Several recent trials have demonstrated that adjusted dose warfarin [i.e. to maintain an INR 2.0 – 3.0] is superior to fixed mini-dose warfarin [1.25 mg/day] and the combination of fixed mini-dose warfarin plus aspirin.^,, In spite of this documented efficacy, warfarin is underused in the prevention of thromboembolism among patients with chronic atrial fibrillation,^,, especially the elderly.^,

Pooled analysis of the studies demonstrates four clinical features that independently predict higher stroke risk: previous transient ischemic attack, thromboembolism or stroke (RR 2.5), increasing age (RR = 1.4 for each decade), history of hypertension (RR=1.6) and diabetes (RR=1.7). Those at highest risk were patients with a prior history of stroke or TIA, who had an annual stroke rate of 12 percent - which fell to 5.1 percent per year with warfarin. Fifteen percent of the patients in these trials had none of these risk factors and their risk of stroke receiving no anticoagulant therapy was 1.0 percent per year. Although not significant in the multivariate analysis, the presence of either congestive heart failure or clinical coronary artery disease did increase the risk of stroke. Patients with either of these cardiac disorders had stroke rates approximately three times higher than patients without any of the risk factors.

While echocardiographic features that predict stroke were not included in the pooled analysis, the SPAF trial found that left atrial size and left ventricular dysfunction measured on two-dimensional echocardiography were associated with increased stroke risk. In addition, mitral annular calcification was associated with increased stroke risk in the BAATAF study. Transesophageal echocardiography may be more sensitive for identifying predictors of stroke, but its clinical importance in risk stratification of patients with atrial fibrillation awaits confirmation.

Table 1 indicates absolute contraindications to initiation of warfarin therapy in patients with atrial fibrillation. In the absence of such contraindications, consideration of anticoagulation requires assessment of concomitant risks and benefits. As demonstrated in table 2 below, risk-stratification using the data outlined above allows recommendations for warfarin, aspirin or no antithrombotic prophylaxis:

Patients with "lone AF" - those under 60 with no history of previous thromboembolic event, diabetes, hypertension, valvular disease or prosthesis, heart failure, thyrotoxicosis, ischemic heart disease, or echocardiographic risk factors (impaired LV function, valvular disease or intracardiac thrombus) - have a very low risk of stroke and routine anticoagulation may not be warranted. However, many clinicians use aspirin as prophylaxis in these patients.¹¹

Advancing age and more intense anticoagulation increases the risk of major hemorrhage in patients given warfarin for stroke prevention. The present recommendation in all patients with chronic AF is to give warfarin therapy in doses that prolong the INR to 2.0-3.0. Patients with recent thromboembolic events may benefit from a slightly higher INR and some investigators recommend a target INR of 3.0.

Two of the primary prevention trials^,and a secondary prevention study investigated whether antiplatelet therapy with aspirin reduced the risk of stroke compared to placebo. Taken overall, the studies suggest that aspirin, at 325 mg, prevents stroke less effectively than warfarin but that it is also less likely to cause major hemorrhage. A more recent primary prevention study directly compared aspirin 325 mg daily with warfarin. Compared with aspirin, warfarin reduced the incidence of ischemic stroke by about 30 percent. However, aspirin-treated patients under 75 with no risk factors other than AF had a low rate of thromboembolism (0.5 percent annually) and treatment with warfarin was no more effective. The recent SPAF III trial has shown that patients with AF categorized as "low risk" based on the absence of four prespecified thromboembolic risk factors: (1) recent congestive heart failure or left ventricular fractional shortening of 25% or less, (2) previous thromboembolism, (3) systolic blood pressure greater than 160 mm Hg, or (4) female gender at age older than 75 years have relatively low rates of ischemic stroke during treatment with aspirin. Thus, in patients at low risk for stroke, warfarin may confer no benefit in terms of absolute risk and aspirin therapy may be beneficial.¹¹

The goals of therapy in patients with atrial fibrillation are directed at controlling the patient’s symptoms and reducing the risk of thromboembolic complications. Conversion of AF to sinus rhythm will directly accomplish the first goal over the short term and theoretically will accomplish the second goal if sinus rhythm can be maintained. Conversion of AF to sinus rhythm is associated with increased cardiac output and improved exercise capacity. An attempt at cardioversion should be considered in all patients with chronic AF and electrical cardioversion is the preferred method with which to reestablish sinus rhythm. For patients in chronic atrial fibrillation, anticoagulation with warfarin for 3 to 4 weeks prior to cardioversion is required to reduce thromboembolic risk. Among patients presenting with atrial fibrillation that was clinically estimated to have lasted less than 48 hours [i.e. acute atrial fibrillation], the likelihood of cardioversion-related clinical thromboembolism is low [0.8%] and early cardioversion in these patients is safe.

Cardioversion can be achieved with antiarrhythmic drug therapy, electrical cardioversion or both. There are no data to prove that maintenance of sinus rhythm with antiarrhythmic agents reduces thromboembolic risk. In addition, there is increasing concern that some antiarrhythmic drugs may confer a substantial risk of proarrhythmia in patients with AF. Initiation of antiarrhythmic therapy is associated with a significant risk of adverse events. Therefore, observation with electrocardiographic monitoring is advisable for the first 24 – 48 hours, especially in the elderly or those with previous myocardial infarction. Thus, the benefits of chronic drug therapy need to be weighed against each patient’s individual risks and the decision to institute therapy should be strongly guided by the patient’s symptoms.