Chapter 12. Atrial Fibrillation

• Irregularly irregular ventricular rhythm.
• Absence of P waves on the electrocardiogram.

Atrial fibrillation, the most common sustained clinical arrhythmia, is diagnosed by finding an irregularly irregular ventricular rhythm without discrete P waves (Figure 12–1). The QRS complex is usually narrow, but it may be wide if aberrant conduction or bundle branch block is present. Atrial fibrillation associated with the Wolff-Parkinson-White syndrome may occur with very rapid ventricular rates and may be life-threatening. This arrhythmia is diagnosed by its very rapid irregular rate associated with wide preexcited QRS complexes and requires emergency treatment (see later section, Long-Term Approach).

Approximately 4% of the population over age 60 years has sustained an episode of atrial fibrillation, with a particularly steep increase in prevalence after the seventh decade of life. Risk factors for development of atrial fibrillation include heart failure, hypertensive cardiovascular disease, coronary artery disease, and valvular heart disease. Moreover, both sustained and paroxysmal atrial fibrillation have important implications for the development of a cerebrovascular accident (CVA) or other systemic emboli. It is estimated that 15–20% of CVAs in nonrheumatic patients are due to atrial fibrillation.

Symptoms & Signs

When called on to manage new-onset atrial fibrillation, it is important to establish the precipitating factors because the type of associated condition determines long-term prognosis. In some patients, episodes of atrial fibrillation may be initiated by caffeine, alcohol, or marijuana use. Atrial fibrillation may result from acute intercurrent ailments. For example, this arrhythmia may develop in patients with hyperthyroidism or lung disease, or after either cardiac or pulmonary surgery, especially in older patients. Atrial fibrillation is also seen in patients with acute pulmonary embolism, myocarditis, or acute myocardial infarction, particularly when the last condition is complicated by either occlusion of the right coronary artery or heart failure. When atrial fibrillation occurs in these settings, it almost always abates spontaneously if the patient recovers from the underlying problem. Hence, management usually involves administration of drugs to control the heart rate, and long-term antiarrhythmic therapy is generally not needed.

Alternatively, atrial fibrillation may occur in association with structural cardiac disease. Important associated conditions include rheumatic mitral stenosis, hypertension, hypertrophic cardiomyopathy, or chronic heart failure. In contrast to patients with acute intercurrent ailments, those with structural heart disease may expect (even with antiarrhythmic therapy) many recurrences, and chronic atrial fibrillation may supervene.

Lone fibrillation is the term used to describe patients with atrial fibrillation not associated with known cardiac conditions or noncardiac precipitants. The natural history of the atrial fibrillation for those with lone atrial fibrillation is similar to that in patients with structural cardiac disease, in that episodes of atrial fibrillation are likely to recur and, eventually, the arrhythmia may become sustained.

Initial Evaluation

The initial evaluation is mainly directed at diagnosing associated conditions (eg, valvular heart disease, hypertension, evidence of heart failure). The cardiac examination during atrial fibrillation will show an irregularly irregular rhythm, variability of the first heart sound, and augmentation of a systolic murmur after long pauses. The initial evaluation of new-onset atrial fibrillation also includes a detailed history focusing on possible precipitating factors as well as the presence of organic cardiac disease. As such, the initial evaluation includes, at a minimum, a careful physical examination, 12-lead electrocardiogram (ECG), chest radiograph, echocardiogram, and tests of thyroid function. Further testing will depend on various aspects of the history or physical examination. Ambulatory ECG monitoring may be helpful in diagnosing the patient with paroxysmal atrial fibrillation. For example, if atrial fibrillation is usually precipitated by exercise, then an exercise treadmill test is appropriate. In the patient with frequent episodes of paroxysmal atrial fibrillation, a 24- to 48-hour ambulatory ECG recording may discern whether atrial fibrillation was triggered by another arrhythmia, such as a premature atrial complex alone, or whether the fibrillation was preceded by an episode of supraventricular tachycardia. In addition, patients with vagally mediated fibrillation will typically have episodes either after heavy meals or during sleep. These clues may help identify those patients who may respond to specific approaches (see later section, Treatment).

The ECG diagnosis of atrial fibrillation is usually straightforward and is characterized by an irregularly irregular rhythm with a narrow QRS complex unless there is associated bundle branch block. The chief diagnostic problem relates to patients showing short bursts of irregular atrial tachycardia (which are common on random ambulatory ECG recordings in the elderly). In addition, patients with atrial flutter may have varying atrioventricular (AV) conduction, which may serve to mimic atrial fibrillation. Careful evaluation of all leads looking for the characteristic flutter pattern is important. Less commonly, patients presenting with multifocal atrial tachycardia may mimic the irregularity found in those with atrial fibrillation. Careful analysis in the latter group shows an irregular atrial tachycardia with at least three P-wave morphologies. A very uncommon arrhythmia that may mimic atrial fibrillation is seen in patients with dual AV nodal physiology with dual conduction over the fast and slow AV nodal pathways. For this diagnosis, identify that a single sinus P wave is associated with two QRS complexes. These patients are readily cured with slow pathway ablation.

The objectives of therapy include (1) achieving rate control, (2) restoring sinus rhythm (where feasible), and (3) decreasing the risk of CVA. The principles of treatment discussed in this chapter largely follow those promulgated in the recent American College of Cardiology (ACC)/American Heart Association (AHA)/European Society of Cardiology (ESC) guidelines.

Rate Control

If the patient has atrial fibrillation and a rapid rate associated with severe heart failure or cardiogenic shock, emergency direct-current cardioversion is indicated. For patients with atrial fibrillation associated with rapid rate but with stable hemodynamics, attempts to achieve acute rate control are indicated. Drugs to slow the ventricular rate in patients with atrial fibrillation (Table 12–1) include digitalis preparations, calcium channel blockers (verapamil or diltiazem), and β-blockers. If rapid rate control is desired, then calcium channel blockers and β-blockers are far more effective than digitalis. In addition, a common misconception is that digitalis therapy is associated with acute conversion to sinus rhythm, but carefully controlled studies have shown that conversion to sinus rhythm is no more likely with digoxin than with placebo. As emphasized later, digitalis and intravenous calcium channel blocker therapy are contraindicated in patients with Wolff-Parkinson-White syndrome and atrial fibrillation. Intravenous diltiazem has been shown to be safe and effective for patients with atrial fibrillation and a modest degree of heart failure. In patients refractory to these agents, intravenous amiodarone may be effective but requires many hours before rate control is achieved. In patients with a known history of congestive heart failure, use of intravenous β-blockers or calcium channel blockers may aggravate cardiac failure. In this subset, digitalis or intravenous amiodarone would be the preferred agent for rate control.

Long-Term Antiarrhythmic Therapy and Elective Cardioversion

For patients who have had a single, initial episode of atrial fibrillation with no significant hemodynamic problems, no specific therapy is required because repeat episodes may not occur for many years. In contrast, patients who manifest frequent recurrences may be candidates for long-term antiarrhythmic therapy with class Ia (quinidine, procainamide, and disopyramide), class Ic (propafenone and flecainide), or class III (sotalol, amiodarone, dronedarone, and dofetilide) agents, all of which are more effective than placebo in maintaining sinus rhythm (Figure 12–2).

Antiarrhythmic Drug Therapy for Atrial Fibrillation

For patients with lone atrial fibrillation, use of any of the antiarrhythmic drugs listed is appropriate. In general, the class Ic agents (flecainide or propafenone) are the first choice in terms of efficacy and lowest incidence of side effects. It would be wise, for example, to withhold amiodarone as a first-line drug in view of the potential for adverse effects. Only two drugs have been proved safe for patients with severe congestive heart failure: dofetilide and amiodarone.

For patients with atrial fibrillation associated with coronary artery disease, consider use of sotalol as initial drug therapy. This agent has class III antiarrhythmic effects and is a potent β-blocker. Class Ic drugs should not be used in patients with significant structural cardiac disease or in those with ischemic heart disease. Sotalol may be safe and effective for patients with hypertension and atrial fibrillation without significant left ventricular hypotrophy. Similarly, class Ic drugs may be used in well-controlled hypertension, but without left ventricular hypertrophy.

In addition, extracardiac factors are very important in the choice of antiarrhythmic drugs. For example, dose adjustments are mandatory for patients with renal insufficiency. This is especially true for sotalol and dofetilide. Dofetilide, for example, requires hospital admission, calculation of the creatinine clearance, and drug titration according to the QT corrected for heart rate as well as renal function. An algorithm for antiarrhythmic drug usage is summarized in Figure 12–2, and drug doses are listed in Table 12–2.

Even with drug therapy, recurrence rates for atrial fibrillation approach 50% per year (as opposed to recurrences with placebo therapy of 75% per year). In addition, these agents may be associated with significant side effects. For class Ia drugs, these include induction of torsades de pointes, especially for those with congestive heart failure. For example, a meta-analysis compared quinidine with placebo for patients with atrial fibrillation and found that death from all causes was higher in the groups treated with quinidine. In addition, in the Stroke Prevention in Atrial Fibrillation (SPAF) trials, substantial numbers of patients were treated with antiarrhythmic agents; in patients with heart failure, those treated with class I drugs had significantly increased mortality rates compared with those not treated with antiarrhythmic drugs. Great care must be exercised in the use of these agents, balancing the benefits against the potential for adverse effects. General rules include avoidance of all class Ia drugs or sotalol for patients with congestive heart failure and avoidance of class Ic agents for patients with structural heart disease. In addition, sotalol is contraindicated for patients with severe depression of the left ventricular ejection fraction and/or severe left ventricular hypertrophy. Patients with significant sinus node or AV conduction disease may require pacemaker therapy before use of antiarrhythmic drugs because these drugs may further depress sinus node or AV conduction. The only drugs that appear to be both effective and safe for patients with heart failure and atrial fibrillation are amiodarone and dofetilide. Amiodarone is associated with a host of both cardiac (eg, severe sinus bradycardia or arrest or AV block) and noncardiac (eg, thyroid abnormalities, pulmonary fibrosis) adverse effects, but low-dose amiodarone (ie, 200 mg/day) appears to be effective and very well tolerated. Dofetilide has a narrow therapeutic window and can cause life-threatening arrhythmias; it can be used in patients with atrial fibrillation and congestive heart failure but requires a 2- to 3-day in-hospital stay for monitoring of the agent. Dronedarone is the newest agent approved for treatment of atrial fibrillation. This drug is a derivative of amiodarone and has most of the physiologic effects of amiodarone. This drug has limited efficacy and should not be used in patients with severe or worsening heart failure or in those with persistent atrial fibrillation and cardiovascular risk factors (see later discussion of CHADS2 score in Anticoagulant Therapy section).

Chemical Cardioversion

Recent studies have emphasized the use of drugs for acute conversion of atrial fibrillation. It has been shown that intravenous ibutilide and intravenous dofetilide (not available in the United States) are effective for conversion of approximately 50% of patients with recent-onset atrial fibrillation. It should be emphasized that ibutilide should be used only in a monitored environment. The usual dose is 1 mg over 10 minutes, followed by a 10-minute interlude, followed by an additional 1 mg over 10 minutes if necessary. A defibrillator should be readily available, because the incidence of sustained torsades de pointes is 1–2%. Ibutilide should be avoided for patients with severe heart failure bradycardia or a prolonged QT interval.

Other Drugs for Chemical Cardioversion

Other drug combinations have also been found effective. For example, it has been found that use of large oral doses of either flecainide (300 mg) or propafenone (600 mg) may terminate up to 80% of episodes of atrial fibrillation within 2 hours (pill-in-the-pocket). This approach should be used only in patients who are pretreated with β-blocking drugs and in the absence of significant cardiac disease or heart failure. It is best to assess first-time use of the “pill-in-the-pocket” approach in a hospital setting.

Anticoagulant Therapy

The risk of CVA in patients with nonrheumatic atrial fibrillation is 4–7% per year. Patients at particularly high risk include those over age 75 years or with hypertension, a history of heart failure, increased left atrial size, diabetes, or prior CVA. The risks of CVA are similar in patients with paroxysmal versus chronic atrial fibrillation. Risk of stroke is assessed by noting the CHADS2 score: C (congestive heart failure), H (hypertension), A (age > 75 years), D (diabetes), S (previous stroke). According to current AHA and ACC guidelines, aspirin therapy is acceptable for those with CHADS2 score of 0 or 1, whereas anticoagulants are advised for those with CHADS2 score ≥ 2 who have no contraindications for anticoagulant therapy. More recent risk criteria also include female sex, age over 65, and evidence of coronary or peripheral vascular disease (CHA2DS2-Vasc score). When compared with CHADS2, CHA2DS2-Vasc performed better in predicting those at high risk or those at very low risk for stroke. Warfarin anticoagulation is recommended for those with a CHA2DS2-Vasc score ≥ 2 (Table 12–3). Numerous studies have documented the remarkable efficacy of warfarin in decreasing the risk of emboli by 45–85% in patients with nonrheumatic atrial fibrillation with a low risk of significant hemorrhage, provided the international normalized ratio (INR) is in the range of 2.0–2.5. However, some patients have a considerable risk of bleeding, which can be estimated by scores such as the HAS-BLED (see Table 12–3). In general, if the HAS-BLED score exceeds the CHADS2 score, the benefits of anticoagulation may be outweighed by the risk of bleeding. Risk associated with bleeding exceeds the possible benefit of anticoagulant therapy in younger patients with lone atrial fibrillation because the risk of emboli is very low in this group.

Several newer anticoagulant drugs (dabigatran, rivaroxaban, and apixaban) have recently been approved for anticoagulant therapy of patients with nonvalvular atrial fibrillation. These drugs have proved to be either equivalent or actually superior in efficacy and safety compared with warfarin. The advantages of the newer agents include obviating the need for INR blood tests and fewer drug–drug or drug–food interactions compared with warfarin. All of the newer agents require careful monitoring of renal function per serial measurement of creatinine clearance (CrCl) because dose adjustment is required in the face of reduced renal function (Table 12–4). In addition, specific inhibitors (in the event of bleeding) are not currently available.

The role of aspirin therapy for patients with atrial fibrillation remains controversial. In one study, 75 mg of aspirin failed to decrease the stroke risk compared with placebo (5.5%/year). In contrast, the SPAF I trials showed that a higher dose of aspirin, 325 mg, appeared to be of benefit in patients under 75 years of age. In a follow-up study (SPAF II), the incidence of stroke was higher with aspirin (4.8%) compared with warfarin (3.6%). The SPAF III trials demonstrated that aspirin (325 mg/day) and fixed low-dose warfarin (1, 2, or 3 mg) were ineffective for stroke prevention. Therefore, the weight of current data favors use of warfarin with an INR of 2.0–3.0 as the best strategy to prevent systemic embolization. Recent studies showed that a newer anticoagulant (apixaban) was more effective than aspirin in those with nonvalvular atrial fibrillation who could not take warfarin.

Direct-Current Cardioversion

Direct-current cardioversion is a very effective technique for restoration of sinus rhythm. Because of the benefits of sinus rhythm in terms of improved cardiac output and decreased risk of embolic phenomena, in general, at least one attempt should be made to restore sinus rhythm. Several precautions are in order. If the patient has a history of recurrent episodes of atrial fibrillation, then he or she should be pretreated with an antiarrhythmic agent because reversion to atrial fibrillation after shock therapy is very high. Use of antiarrhythmic drugs before direct-current shock, however, is inappropriate for the patient with an initial episode of well-tolerated atrial fibrillation. Unless urgent cardioversion is required because of hemodynamic decompensation, severe ischemia, or congestive heart failure, it is imperative to follow one of several options for reducing the risk of systemic embolization:

1. For patients with atrial fibrillation of less than 48 hours in duration, it would appear to be safe to proceed with application of direct-current shock.

2. If atrial fibrillation persists for more than 48 hours, then the risk of embolization increases and anticoagulants are required prior to ablation. One recommended option for patients with atrial fibrillation of more than 48 hours in duration is to perform transesophageal echocardiography (TEE), which is excellent for detecting clots in the left atrium or the left atrial appendage. Evidence from several studies indicates that the finding of either a clot or spontaneous echocardiographic contrast in the left atrium is associated with higher risks of systemic embolization. In the absence of such findings on TEE, systemic emboli are rare. Therefore, patients with recent-onset atrial fibrillation with no evidence of atrial clots or spontaneous contrast by TEE may undergo direct-current cardioversion after initiation of heparin therapy. A recent report from the ACUTE trial showed that patients with atrial fibrillation treated based on TEE-guided therapy versus a group treated with a 3-week course of anticoagulant therapy had similar rates of thromboembolism (< 1%). It must be appreciated that atrial function is depressed (atrial stunning) after cardioversion and that anticoagulant therapy is recommended for at least 1 month after cardioversion. This is true whether the duration of atrial fibrillation was either less than or greater than 48 hours. For patients with clot or dense spontaneous echocardiographic contrast with TEE, full anticoagulant therapy with an INR of 2.0–3.0 is recommended for at least 2–3 weeks before cardioversion.

3. An alternative approach is that patients with atrial fibrillation of greater than 48 hours be fully anticoagulated for at least 3 consecutive weeks before attempting direct-current cardioversion and for about 4 weeks afterward to decrease the risk of an embolism after successful reversion to sinus rhythm. This approach tends to be less efficient than the TEE-guided approach for recent-onset atrial fibrillation, but is an acceptable alternative treatment for atrial fibrillation.

Direct-current external shock is usually performed in a monitored area under supervision of an anesthesiologist. Pads are placed in an anterior-posterior orientation in order to maximize current delivered to the atrium. It is wise to check the arterial oxygen saturation and serum potassium level before cardioversion. Direct-current shocks beginning with at least 150 J biphasic are used in an attempt to achieve sinus rhythm. Multiple shocks of lesser energy are to be avoided. If the patient fails to revert after maximal external shocks (360 J biphasic), then successful cardioversion can almost always be achieved by the use of supplemental doses of ibutilide. Ibutilide has been shown to lower the atrial defibrillation threshold. An attempt at internal cardioversion using small-energy shocks delivered between the coronary sinus and the right atrium is seldom necessary because the previously described treatments are almost always effective.

Long-Term Approach

Clinicians should be especially careful to identify patients whose atrial fibrillation might be cured. Examples include patients with hyperthyroidism as well as those in whom other cardiac arrhythmias appear to trigger atrial fibrillation. For example, patients with atrial flutter or paroxysmal supraventricular tachycardia may experience atrial premature impulses during tachycardia that trigger atrial fibrillation. In selected patients, it is possible to apply catheter ablation to cure the underlying supraventricular arrhythmia and, hence, prevent the trigger for atrial fibrillation. Therefore, in the evaluation of patients with atrial fibrillation, initial testing should include obtaining a thyroid-stimulating hormone assay, an echocardiogram, and a 48-hour ambulatory ECG recording for those with paroxysmal atrial fibrillation. In analyses of these recordings, the clinician seeks evidence for triggering arrhythmias. In addition, the clinician looks for vagal triggers of atrial fibrillation, such as sinus bradycardia associated with sleep or heavy meals, that initially may be treated with vagolytic antiarrhythmic agents such as disopyramide. Alternatively, if atrial fibrillation appears only with enhanced sympathetic tone, such as with exercise, a trial of β-blocker therapy is appropriate.

One important special circumstance is that of atrial fibrillation in the patient with Wolff-Parkinson-White syndrome. These patients may have a very rapid irregular rate and wide complex tachycardia owing to conduction over the accessory pathway (Figure 12–3). After recognition of this entity, appropriate immediate therapy includes use of intravenous ibutilide or procainamide or direct-current cardioversion. It is important to remember that intravenous digoxin and calcium channel blockers are contraindicated. In addition, use of lidocaine, β-blockers, or adenosine is not effective and is contraindicated because they delay appropriate therapy. After the rhythm is stabilized, these patients should undergo catheter ablation of the accessory pathway.

The natural history of atrial fibrillation associated with structural cardiac disease or in lone atrial fibrillation patients is spontaneous recurrence of the arrhythmia. Unfortunately, no drug is universally effective, and the decision of how many drugs to try before a judgment is made to terminate antiarrhythmic drugs and focus on rate control depends on how symptomatic the patient is during atrial fibrillation. If the episodes are poorly tolerated, then multiple drug trials or even various ablative procedures may be required (see later section, Nonpharmacologic Treatment of Atrial Fibrillation). On the other hand, if rate control can be readily achieved with drugs, such as digoxin, β-blockers, or calcium antagonists that block AV nodal conduction and the patient has a good symptomatic outcome, then an acceptable alternative is to use drugs that control rate combined with long-term anticoagulant treatment. A large, randomized trial (AFFIRM) compared the strategy of rate control and anticoagulation versus maintaining sinus rhythm with antiarrhythmic drugs. The AFFIRM trial randomized over 4000 patients with atrial fibrillation to either rate or rhythm control. The patient cohort consisted in large measure of older (mean age 69 years) patients who were not very symptomatic. They found no difference in mortality or quality of life between groups. The rhythm control group had a higher incidence of hospitalizations and episodes of torsades de pointes. In addition, stroke risk was related to presence of no or inadequate anticoagulant treatment. This study showed that for most patients with atrial fibrillation, rate control was equally effective as rhythm control and that long-term anticoagulation therapy is required for both groups.

Nonpharmacologic Treatment of Atrial Fibrillation

Because pharmacologic therapy for atrial fibrillation is not ideal, a number of nonpharmacologic treatment modalities have been introduced. Current ACC/AHA/ESC guidelines allow for catheter ablation for treatment of atrial fibrillation after one failed drug trial. Recent studies used wide-area ablative lesions around the pulmonary veins in order to create entrance and exit block from those triggering foci. Other foci may evacuate from the coronary sinus, vein of Marshall, or the superior vena cava and may require ablation as well. In addition, some groups advocate ablation of complex fractionated potentials (CAFÉ) or use of a roof line between the upper right and left superior pulmonary vein or an ablation line from the mitral annulus to the left inferior pulmonary vein (mitral isthmus line). The initial success rate for ablation of paroxysmal atrial fibrillation is 70–90%, but the 5-year success rate is approximately 50%.

Patients with persistent tachycardia may suffer from a tachycardia-induced cardiomyopathy with left ventricular failure superimposed on their native cardiac disease. Hence, in the management of chronic atrial fibrillation, rate control is an important objective that must be achieved either via AV nodal blocking drugs or, failing these, with catheter ablative procedures. Catheter ablation of the AV junction involves insertion of an electrode catheter in the region of the His bundle with application of radiofrequency energy in order to destroy AV conduction. The chief benefit of this technique is achievement of perfect rate control without need for drugs. The drawbacks include the need for permanent pacing and a continued need for anticoagulant therapy.

For most of these patients, especially for those with left ventricular ejection fraction greater than 40%, single chamber pacing appears to be sufficient. In some patients, left ventricular function may deteriorate, and biventricular pacing may be helpful (PAVE trial).

It has been shown that atrial-based pacing systems will decrease the incidence of atrial fibrillation in patients with the tachycardia-bradycardia syndrome. In addition, pacing may allow for safe use of antiarrhythmic drugs. In patients with vagally mediated atrial fibrillation, atrial pacing may be effective in decreasing episodes of atrial fibrillation. Experimental studies have shown that either dual-site atrial pacing (ie, from coronary sinus and right atrium) or pacing from the atrial septum in conjunction with antiarrhythmic therapy may suppress atrial fibrillation, but this technique has not been shown to be clinically useful.

Atrial fibrillation, particularly associated with rapid rates, carries significant negative prognostic implications for those with associated structural cardiac disease or heart failure. For example, loss of the atrial booster pump function for those with less compliant ventricles (ie, hypertrophic myopathy) may result in hemodynamic collapse. In addition, persistent, rapid rates may induce a tachycardia-mediated cardiomyopathy with similar negative prognostic implications. This is why, depending on the clinical scenario, maintenance of sinus rhythm or rate control is such an important therapeutic concern.

Although anticoagulants and perhaps aspirin can reduce the risk of stroke or other systemic emboli, the risk is never zero, whether the patient has paroxysmal, persistent, or permanent atrial fibrillation. Also, in patients with cardiovascular disease, the presence of atrial fibrillation always increases the risk of major adverse cardiac events. It is not clear whether various treatment approaches substantially alter this adverse prognosis. However, maintenance of sinus rhythm compared to rate control does not seem to improve long-term survival.