Atrial Fibrillation: A Therapeutic Update

Emerg Med 38(4):35-44, 2006

The author explains how to evaluate a patient with the increasingly prevalent problem of atrial fibrillation in order to make three strategic decisions: whether to attempt to restore and then maintain sinus rhythm, how best to control ventricular rate acutely and long-term, and how to minimize the risk of thromboembolism in the pericardioversion and long-term settings.

By Barry Stults, MD

A 69-year-old man presents to his physicianÕs office with a three-day history of persistent, mild palpitations. His history is remarkable for hypertension and an anterior wall myocardial infarction (MI) ten years earlier; he also had a first episode of atrial fibrillation (AF) six months ago. An echocardiogram at that time revealed left atrial enlargement, left ventricular anterior wall hypokinesis, and an ejection fraction of 48%; a serum thyrotropin level was normal. After three weeks of anticoagulation with warfarin, he underwent successful electrical cardioversion to sinus rhythm and was discharged on warfarin, metoprolol, lisinopril, and atorvastatin.

He remained asymptomatic until he noted the onset of an irregular pulse rate three days before the visit to his physicianÕs office. His physical examination there is remarkable only for an irregularly irregular pulse rate of 100 to 110 bpm. An ECG showed AF with a ventricular rate of about 100 bpm and stable anterior Q waves.

How should this patientÕs AF be managed?

A 52-year-old woman with known hypertrophic cardiomyopathy presents to the emergency department with a two-hour history of sudden severe dyspnea and palpitations. Her only medication is sustained-release propranolol. The physical examination finds a middle-aged woman in respiratory distress with an irregularly irregular pulse rate of 130 to 140 bpm, a blood pressure of 130/78 mm Hg, bibasilar crackles in the posterior lung fields, and a harsh systolic ejection murmur at the left sternal border. An ECG shows new AF with a ventricular rate of about 130 bpm. Shortly after her arrival in the emergency department, the patient spontaneously converts to sinus rhythm with immediate relief of her palpitations and more gradual relief of her dyspnea over the next hour.

How should this patientÕs AF be managed?

NEARLY EPIDEMIC DISORDER

Atrial fibrillation is a common and nearly epidemic cardiovascular disorder. In 2002 there were an estimated 2.5 million persons with AF in the United States; this figure is expected to increase to 5.6 million cases by 2050. Multiple factors appear to contribute to the rapidly increasing prevalence of AF. The median age of the United States population continues to increase, and the prevalence of AF is strikingly related to age: 2.3% of the population over age 40, 5.9% over age 65, and nearly 10% over age 80 now have AF. Other independent risk factors for the development of AF include valvular heart disease, heart failure, coronary heart disease (CHD), obesity, hypertension, and diabetes. Improved survival rates for patients with valvular disease and CHD and heart failure and the increasing prevalence of obesity, hypertension, and diabetes are also likely contributors to the rising prevalence of AF.

The pathogenic property common to all the different risk factors for AF is that each contributes to diastolic dysfunction of the left ventricle, which in turn leads to left atrial dilatation, stretch, and fibrosis and subsequent vulnerability to the development of AF. About 10% to 20% of AF patients have none of the risk factors listed above; they are considered to have ÒloneÓ AF.

Atrial fibrillation is a considerable public health burden. Compared with sinus rhythm, AF independently increases total mortality twofold, heart failure threefold, and stroke rates fivefold. It is responsible for 15% of all strokes in the United StatesÑnearly 75,000 strokes annuallyÑand it may also cause clinically silent strokes that contribute to vascular dementia. While 10% to 30% of patients with AF may be entirely asymptomatic, others experience moderate or even disabling symptoms of dyspnea, palpitations, and fatigue with significant reductions in health-related quality of life. Overall, AF now contributes to 350,000 hospitalizations, 542,000 emergency department visits, and 7,000,000 office visits annually in the United States at a cost of $6.42 billion.

Substantial research has been conducted over the past 10 years to determine effective management strategies for AF. Definitive scientific evidence is still not available to guide all management decisions, so experienced clinical judgment remains an essential component of a sound management strategy.

We will return to the two patients introduced above, but first, a closer look at the evaluation of patients with AF.

INITIAL EVALUATION OF ATRIAL FIBRILLATION

The initial evaluation of AF requires a careful history, physical examination, ECG, transthoracic echocardiogram (TTE), and selected additional investigations. The evaluation should be designed to accomplish the following goals:
• Identify and treat any reversible precipitants of AF.
• Identify and optimally manage any underlying cardiovascular diseases.
• Identify pericardioversion and long-term stroke risk factors.
• Classify the predominant pattern of AF to guide subsequent management.
• Recognize AF patients with hemodynamic compromise, Wolff-Parkinson-White (WPW) syndrome, or the bradycardia-tachycardia syndrome since they will require special management strategies.

Potentially reversible precipitants of AF that should be identified and treated, if possible, include arterial hypoxemia, hypokalemia, hypomagnesemia, excessive alcohol consumption, pericarditis, and hyperadrenergic states such as the postoperative period, alcohol withdrawal, exacerbations of chronic obstructive pulmonary disease, pulmonary embolism, and pneumonia. While hyperthyroidism can also cause AF, screening thyroid function tests are of limited value except in patients suspected to have the disorder and in elderly patients in whom the typical clinical manifestations of hyperthyroidism may be masked. Obstructive sleep apnea may also facilitate the development of AF; patients suspected to have this disorder require further evaluation.

All patients with AF should have an ECG and a TTE to help identify and quantify any underlying cardiovascular disease and guide subsequent management. The ECG may reveal evidence of acute myocardial ischemia, pre-excitation, sinus node or conduction system disease, or QT interval prolongation that will alter the management of AF. The TTE determines left atrial size, a predictor for AF recurrence, and can also detect left ventricular hypertrophy and ventricular dysfunction that will influence subsequent management decisions.

Other cardiovascular testing should not be performed routinely. However, intracardiac electrophysiologic studies may be useful in young patients with idiopathic lone AF to detect underlying pre-excitation or the presence of regular supraventricular tachycardias that may induce AF. Stress testing to detect underlying CHD is indicated only if the initial evaluation suggests the presence of ischemic heart disease.

PERICARDIOVERSION AND LONG-TERM RISK FACTORS FOR STROKE

The initial evaluation should also determine the presence of any pericardioversion or long-term risk factors for AF-associated stroke. Cardioversion to sinus rhythm may cause an embolic stroke in up to 5% of patients who are not therapeutically anticoagulated. Risk factors for cardioversion-related stroke include duration of AF of more than 48 hours or AF of unknown duration, recurrent episodes of AF in the past few months, prior stroke or transient ischemic attack (TIA) related to AF, and mitral stenosis. The major independent long-term risk factors for stroke in patients with nonvalvular AF are listed in the table below. As discussed below, these risk factors have been incorporated into a variety of schemes to stratify long-term stroke risk and guide the use of antithrombotic therapy.

The classification scheme proposed by the 2001 American College of Cardiology/American Heart Association/European Society of Cardiology Board Task Force is a useful guide to management strategies for patients with AF. Paroxysmal AF is self-terminating and lasts less than seven days and usually less than 48 hours; it can be subdivided into first-episode paroxysmal AF and recurrent paroxysmal AF categories. Persistent AF is not self-terminating and lasts more than seven days unless cardioversion is performed; it can be subdivided into first-episode persistent AF and recurrent persistent AF categories. Either patients with permanent AF cannot be successfully cardioverted or a clinical decision is made not to attempt to cardiovert them.

The natural history and interrelationships of these different patterns of AF have not been fully clarified. There is evidence that over a five-year period about 25% of patients with paroxysmal AF will progress to persistent AF; the likelihood is increased in patients who are older or who have larger left atrial dimensions, CHD, or slower ventricular rates during their AF episodes.

KEY MANAGEMENT DECISIONS

After the initial evaluation, clinicians caring for patients with AF must make three key management decisions:
• whether to attempt to restore and then maintain sinus rhythm;
• how to optimally control ventricular rate acutely and long-term; and
• how to minimize the risk of thromboembolism in the pericardioversion and long-term settings.

Multiple pharmacologic and nonpharmacologic options are available for each of these three management decisions (see algorithm).

Recent studies suggest that the decision to restore and maintain sinus rhythm should be carefully individualized among AF patients. For many years the accepted strategy to manage AF was to routinely restore sinus rhythm with cardioversion and then maintain it with antiarrhythmic drugs. This strategy of rhythm control was presumed to reduce mortality, heart failure, and stroke rates as well as improve symptoms and health-related quality of life. However, the benefits of this strategy had never been proved, and it was frequently ineffective and occasionally toxic.

Recent randomized clinical trials (AFFIRM, RACE, PIAF, STAF, and CAFƒ) compared a strategy of rhythm control using serial cardioversion, antiarrhythmic drugs, and optional anticoagulation with a strategy of ventricular rate control to eliminate symptoms of AF (with no attempt to restore or maintain sinus rhythm) and routine anticoagulation. The patients included in these trials were primarily elderly patients with recurrent persistent AF or at high risk for recurrent AF who were minimally symptomatic and without significant heart failure. In these patients, the rhythm- and rate-control strategies were equivalent with respect to mortality, stroke rates, and symptom control. However, there were fewer hospitalizations and adverse drug effects and lower costs with the rate-control strategy. Of note, in the AFFIRM trial, patients age 65 and over with no history of heart failure had a significantly lower mortality with a rate-control strategy.

These clinical trials also demonstrated that a rhythm-control strategy does not prevent AF-related stroke. Patients at high risk for stroke require antithrombotic therapy (with warfarin, unless contraindicated) regardless of whether a rhythm- or rate-control strategy is pursued.

An initial strategy of rate control with anticoagulation, if indicated, is acceptable for many elderly patients at high risk for recurrent AF. Other subsets of AF patients may benefit from a rhythm-control strategy, particularly patients with paroxysmal or persistent AF who remain symptomatic despite attempts at rate control. Some clinicians also favor a limited rhythm-control strategy for most patients with a first episode of persistent AF. These patients would be cardioverted to sinus rhythm but would not receive maintenance therapy with antiarrhythmic drugs. The rationale for this approach is that about 25% of patients with a first episode of AF will not have a recurrence for several years or longer; this clinical course is more likely in younger patients without valvular heart disease or hypertension who have a relatively normal left atrial size and an initial episode of AF of short duration (less than a few days). A third subset of AF patients who may possibly benefit from a rhythm-control strategy are those with significant left ventricular systolic dysfunction. Atrial fibrillation doubles the mortality rate in patients with systolic heart failure; a large clinical trial is in progress to determine the benefit of restoring and maintaining sinus rhythm in these patients.

The algorithms below suggest an approach for selecting rhythm- versus rate-control strategies for different subsets of AF patients.

If a rhythm-control strategy is selected, clinicians must first determine if anticoagulation is indicated prior to elective cardioversion of patients with persistent AF. About 50% to 60% of patients whose duration of AF is less than 48 hours will spontaneously cardiovert to sinus rhythm; drugs that block the atrioventricular (AV) node to control ventricular rate do not facilitate cardioversion. Patients who do not cardiovert spontaneously may undergo synchronized electrical cardioversion or pharmacologic cardioversion with antiarrhythmic drugs; patients with AF of 48 to 72 hours or longer duration will generally require electrical cardioversion along with an anticoagulation strategy to prevent stroke. Specific guidelines for electrical and pharmacological cardioversion have recently been published (Canadian Cardiovascular Society, 2004).

LOW EFFICACY WITH ANTIARRHYTHMIC DRUGS

Maintenance of sinus rhythm with antiarrhythmic drugs is complicated by low efficacy and a considerable rate of significant side effects. Flecainide and propafenone have one-year efficacy rates of about 50% in preventing AF; these drugs cannot be used by patients with underlying heart disease because of their potential for proarrhythmic effects. These drugs can also precipitate atrial flutter with 1:1 AV nodal conduction and a very rapid ventricular rate; digoxin, a beta blocker, verapamil, or diltiazem must be prescribed concurrently. Sotalol has similar efficacy in preventing AF and may also be used in patients with CHD if systolic function is normal.

In patients with heart failure, a left ventricular ejection fraction below 35%, or significant left ventricular hypertrophy, only amiodarone and dofetilide are sufficiently free of proarrhythmic effects to be safe for use. Amiodarone has a one-year efficacy rate of about 70% in preventing AF, but significant pulmonary, thyroid, and hepatic toxicity limit its widespread use. In patients with no underlying heart disease, antiarrhythmic drugs may be initiated in the outpatient setting. Because of its low risk of proarrhythmia, amiodarone may also be initiated in the outpatient setting even in patients with left ventricular dysfunction. All other antiarrhythmic drugs for patients with structural heart disease must be initiated in the inpatient setting with several days of ECG monitoring.

Nonpharmacologic approaches to rhythm control are now under investigation, although their ultimate role in the management of AF remains to be defined. Device-based therapy such as burst or dual-site atrial pacing and implantable atrial defibrillation have low efficacy and poor tolerance, respectively. The Maze procedure to surgically interrupt re-entrant circuits in the left atrium is effective but requires open-heart surgery. Percutaneous radiofrequency ablation of arrhythmogenic foci in and around the junction of the pulmonary veins and the left atrium has shown considerable promise in the prevention of recurrent AF. In a recent one-year randomized clinical trial, 63% of patients treated with antiarrhythmic drugs had a symptomatic recurrence of AF compared to just 13% of patients treated with radiofrequency ablation. However, even in experienced centers the rate of major complications is 1% to 6%. For now, the use of radiofrequency ablation as a rhythm-control intervention may best be limited to patients with WPW syndrome or with AF and minimal to moderate underlying heart disease who remain symptomatic despite the use of antiarrhythmic drugs and pharmacologic attempts at rate control.

VENTRICULAR RATE CONTROL

For AF patients in acute settings who do not have decompensated heart failure or the WPW syndrome, beta blockers, diltiazem, and verapamil are the most effective drugs to control rapid ventricular rates and improve symptoms. Beta blockers are preferred in patients with ischemic heart disease or those who have high sympathetic tone from precipitating factors such as the postoperative setting, hyperthyroidism, alcohol withdrawal, pulmonary embolism, pericarditis, or systemic infection. Diltiazem or verapamil may be particularly useful in patients with contraindications to beta blockers such as reactive airways disease. Patients with minimal symptoms may initially be treated with oral, short-acting preparations for ease of dose titration and then converted to the appropriate dose of an extended-release preparation once their ventricular rate is controlled. More symptomatic patients may require intravenous (IV) administration of one of these medications that can later be converted to an extended-release preparation.

Digoxin is less useful in acute settings because its peak effect on heart rate is delayed for several hours. However, in patients with decompensated heart failure, it is the initial drug of choice; in some cases, it may be temporarily supplemented by low doses of diltiazem, if needed. Beta blockers should be avoided until heart failure has improved.

Although conclusive scientific evidence is lacking, the recommended ventricular rate in patients with permanent AF is 60 to 80 bpm at rest and less than 110 bpm with daily activities such as a six-minute walk or stair climbing. Despite optimal control of the resting ventricular rate, some AF patients will have excessive tachycardia and symptoms of palpitations and dyspnea with their usual daytime activities. Beta blockers, diltiazem, and verapamil can control the ventricular rate during exercise more effectively than digoxin. Beta blockers are preferred as initial therapy in patients with ischemic heart disease or compensated heart failure. In patients without heart failure who are physically active or who have reactive airways disease, diltiazem or verapamil may be better tolerated.

If the ventricular rate is not adequately controlled with the initial drug or if there are side effects, digoxin may be useful as adjunctive therapy to improve rate control or allow the use of a lower dose of the initial drug. In elderly patients with limited physical activity, digoxin alone may provide adequate rate control. A few patients with permanent AF do not respond to rate-control therapy with drugs; they may require ablation of the AV node with placement of a pacemaker to adequately control the ventricular rate.

MINIMIZING THROMBOEMBOLIC RISK

When AF has persisted for 48 hours or longer or is of unknown duration and patients have not received prior anticoagulation therapy, pharmacologic or electrical cardioversion is associated with a 5% to 7% risk of arterial thromboembolism over the following month. Either of two strategies can effectively reduce this risk to less than 1%. The most commonly used strategy is therapeutic anticoagulation with warfarin to achieve an INR target goal of between 2 and 3 for at least three weeks before and at least four weeks after elective cardioversion. It is essential to document that the INR is therapeutic on a weekly basis before and after as well as at the time of cardioversion.

Alternatively, if prompt cardioversion is desired, therapeutic anticoagulation with IV unfractionated heparin can be initiated and a multiplane transesophageal echocardiogram (TEE) performed. If no thrombus is present in the left atrium or the left atrial appendage, cardioversion can be safely performed provided that IV heparin is continued until the patient is successfully transitioned to warfarin with an INR target goal between 2 and 3 and anticoagulation is continued for at least four weeks after cardioversion. If a thrombus is found, cardioversion must be delayed and long-term anticoagulation with warfarin should be initiated. Cardioversion should not be performed until a repeat TEE documents resolution of the thrombus.

When AF is known to have been present for less than 48 hours, cardioversion can be performed in most patients without a prior three-week period of anticoagulation with warfarin. Some clinicians recommend pericardioversion use of therapeutic doses of IV unfractionated heparin or subcutaneous low-molecular-weight heparin for these patients if there are no contraindications. If the AF patient is at very high risk for strokeÑfor example, if there is a history of prior stroke or mitral stenosis or multiple stroke risk factors are present, or if there have been recurrent episodes of AF in the past few monthsÑthe use of IV heparin and TEE should be considered even if the duration of AF is less than 48 hours.

Depending on the presence and number of stroke risk factors, the long-term risk of stroke in AF varies from less than 1% per year to more than 10% per year (see table on page 37). Given equivalent risk factors, patients with paroxysmal AF have the same risk of stroke as patients with permanent AF. The use of a pharmacologic or nonpharmacologic rhythm-control strategy, even if it is apparently successful and patients are entirely asymptomatic, does not adequately prevent stroke in patients who have stroke risk factors. Prolonged ECG monitoring studies have demonstrated that asymptomatic (or ÒsilentÓ) episodes of AFÑsome with a duration of up to 48 hoursÑare surprisingly common in these patients.

Meta-analyses of clinical trials have demonstrated that warfarin therapy with an INR target goal of 2 or higher reduces the incidence of stroke in AF patients by 68% compared to control and by about 50% compared to aspirin (75-325 mg/day). In contrast, aspirin reduces stroke incidence in AF patients by only 21% compared to control. Warfarin is associated with a higher annual rate of major bleeding than aspirin (2.2% versus 1.3% per year), but the rate of intracranial hemorrhage is similar (0.5% versus 0.3% per year). Prior intracranial hemorrhage, uncontrolled hypertension (blood pressure above 160/90 mm Hg), thrombocytopenia (platelet count below 50,000/mm3), and poor adherence are absolute contraindications to warfarin therapy. A predisposition to falls or a prior history of peptic ulcer disease (provided that Helicobacter pylori infection has been effectively treated and any nonsteroidal anti-inflammatory drug use has been discontinued) are only relative contraindications. Atrial fibrillation patients with these conditions and a high risk for stroke may still be considered for warfarin therapy.

ACCP STRATIFICATION SCHEME FOR LONG-TERM STROKE RISK

Over the past 10 years, eight different schemes have been published to stratify stroke risk and guide long-term antithrombotic therapy in AF. Some schemes have been based on AF patient cohorts from clinical trials, administrative databases, or community-based samples while others are based in part on expert consensus. None of the schemes has been prospectively validated on an independent sample of AF patients, so their relative accuracy is unknown.

The Seventh American College of Chest PhysiciansÕ Conference on Antithrombotic and Thrombolytic Therapy (2004) suggests the following approach to risk stratification and antithrombotic therapy:
¥ In the absence of contraindications, high-risk patients should be treated indefinitely with warfarin to an INR target goal of between 2 and 3. (High-risk patients are those with prior stroke, TIA, or systemic embolism, age over 75 years, moderate or severe left ventricular dysfunction or heart failure, history of hypertension, or diabetes mellitus.)
¥ Patients aged 65 to 75 years with none of the above risk factors are at intermediate risk of stroke and may be treated either with warfarin to an INR of between 2 and 3 or with aspirin, 325 mg daily, depending on bleeding risk and patient preference.
¥ Low-risk patients under age 65 with no risk factors should be treated with aspirin 325 mg/day.

Unfortunately, only about 50% of eligible AF patients at high risk for stroke are currently being treated with warfarin in the United States.

SPECIAL MANAGEMENT STRATEGIES

Patients with hemodynamic compromise, WPW syndrome, or the bradycardia-tachycardia syndrome must be promptly recognized because they require a special management strategy compared to other patients with AF. Patients who present with hypotension, acute coronary syndrome, or severe pulmonary edema should be considered for prompt electrical cardioversion, generally followed by therapeutic anticoagulation with heparin. Atrial fibrillation in the patient with WPW syndrome may conduct anterograde through an accessory pathway rather than through the AV node. This conduction pathway results in a wide QRS complex (1.2 ms or wider) with a delta wave and the potential for a very rapid ventricular rate of 180 to 200 bpm or higher that can degenerate into ventricular fibrillation. If hemodynamic compromise is present in a patient with AF and WPW syndrome, emergent electrical cardioversion should be performed.

In hemodynamically stable patients with AF and pre-excitation, it is essential to avoid AV nodal blocking drugs such as beta blockers, verapamil, diltiazem, and digoxin. These drugs do not slow conduction through the accessory pathway but may actually augment it, resulting in a further increase in ventricular rate and subsequent ventricular fibrillation. Pharmacologic cardioversion may be accomplished in these patients with IV procainamide or ibutilide.

Atrial fibrillation patients with the bradycardia-tachycardia or sick sinus syndrome also require special management. These patients may have a history of sinus bradycardia or other conduction system abnormalities. During episodes of AF they may have a ventricular rate below 100 bpm in the absence of rate-controlling medications; this spontaneous rate control reflects concurrent disease of the AV node that may be associated with similar disease of the sinoatrial node. Patients with the bradycardia-tachycardia syndrome may experience prolonged sinus pauses after spontaneous or pharmacologic cardioversion of AF, resulting in syncope or presyncopal symptoms. Pacemaker implantation should therefore be available at the time of pharmacologic cardioversion. These patients may also experience serious bradyarrhythmias with the use of rate- or rhythm-controlling drugs for AF.

RETURNING TO OUR PATIENTS

The first patient presented at the beginning of this article has recurrent persistent AF in the setting of hypertension, CHD, and slightly reduced left ventricular systolic function. Given his stroke risk factors of hypertension, CHD, and age over 65 years, he requires long-term anticoagulation with warfarin regardless of whether a rhythm- or rate-control strategy is chosen. His first episode of persistent AF was managed with a limited rhythm-control strategyÑcardioversion but no antiarrhythmic drugsÑbecause he had a 25% chance of remaining in sinus rhythm for several years. Now that he has recurrent persistent AF and few or no symptoms, it is reasonable to manage him with a rate-control strategy and reduce his resting ventricular rate to 60 to 80 bpm with an increased dose of metoprolol.

The second patient has had a first episode of paroxysmal AF in the setting of known hypertrophic cardiomyopathy. Because of her severe left ventricular diastolic dysfunction, the episode was complicated by acute pulmonary edema. She is at high risk for recurrent paroxysmal AF and pulmonary edema, so a rhythm-control strategy is essential. Amiodarone and dofetilide are the only antiarrhythmic drugs likely to be sufficiently free of proarrhythmic effects for use in this patient, and amiodarone can be initiated in the outpatient setting. Atrial fibrillation in patients with hypertrophic cardiomyopathy is associated with a high annual risk for stroke. Long-term warfarin therapy to an INR of between 2 and 3 is indicated.

back to top

Printable version of this article

E-mail this article