Aspirin and Other Antithrombotic Agents

Aspirin does not prevent thromboembolic complications as effectively as warfarin or NOACs in patients with AF. In a meta-analysis of five randomized clinical trials, aspirin did not significantly reduce the risk of stroke compared to placebo in patients with AF.⁹ In a large cohort study of patients with non-valvular AF, aspirin had no therapeutic efficacy for preventing strokes.¹⁷ Therefore, if aspirin is used for prophylactic therapy, this should be only in patients at lowest risk of thromboembolic complications (CHA₂DS₂-VASc score of 0). However, the 2014 American College of Cardiology (ACC), American Heart Association (AHA), and /Heart Rhythm Society (HRS) guidelines for management of patients with AF recommend that for patients with nonvalvular AF and a CHA₂DS₂-VASc score of 0, it is reasonable to omit all antithrombotic therapy, including aspirin. When the CHA₂DS₂-VASc score is 1, the guidelines recommend that either no antithrombotic therapy or treatment with an oral anticoagulant or aspirin may be considered.² Because of the negligible therapeutic effect of aspirin, a risk of bleeding complications that is close to the risk of oral anticoagulants, and the ability of the CHA₂DS₂-VASc score to accurately identify low-risk patients, the most recent guidelines of the European Society of Cardiology (ESC) recommend no antithrombotic therapy when the CHA₂DS₂-VASc score is 0 and an individualized decision between no antithrombotic therapy versus an oral anticoagulant when the CHA₂DS₂-VASc score is 1.²⁰

If aspirin is used for stroke prevention in patients with AF, the appropriate daily dose is 81 to 325 mg/day. No data indicate superiority of a particular dose for prevention of thromboembolism.

In patients with a CHADS₂ score greater than 1 who are not able to tolerate anticoagulation with warfarin or a NOAC, combination therapy with aspirin and the platelet inhibitor clopidogrel is more efficacious than aspirin alone for prevention of thromboembolic complications.²¹ The potential benefit of combination therapy with aspirin plus clopidogrel may outweigh the increased risk of bleeding complications in high-risk patients who are not suitable candidates for warfarin or a NOAC.

Warfarin and Direct Thrombin Inhibitors

A meta-analysis of the major randomized clinical trials that compared warfarin with placebo for prevention of thromboembolism in patients with AF demonstrated that warfarin reduced the risk of all strokes (ischemic and hemorrhagic) by approximately 60%.⁹ The target INR should be 2.0 to 3.0. This range of INRs provides the best balance between stroke prevention and hemorrhagic complications. In clinical practice, maintenance of the INR in therapeutic range has been challenging, and a large proportion of patients often have an INR of less than 2.0. A large prospective study of community-based practices demonstrated that the mean time in therapeutic range in patients treated with warfarin was only 66% and that the time in therapeutic range was less than 60% in 34% of patients.²² Maintaining the INR at a level of 2.0 or higher is important because even a relatively small decrease in INR from 2.0 to 1.7 more than doubles the risk of stroke.

The annual risk of a major hemorrhagic complication during anticoagulation with warfarin is in the range of 1% to 2%, and a strong predictor of major bleeding events is an INR greater than 3.0. For example, the risk of intracranial bleeding is approximately twice as high at an INR of 4.0 than 3.0. This emphasizes the importance of maintaining the INR in the range of 2.0 to 3.0.

Some studies have indicated that advanced age can be a risk factor for intracranial hemorrhage in patients with AF treated with warfarin. The concern of hemorrhagic complications may lead some clinicians to favor the use of aspirin over warfarin or one of the NOACs in elderly patients. However, the available data indicate that the risk-to-benefit ratio of warfarin and the NOACs is more favorable than that of aspirin even in patients older than 75.²³

It is well established that genetic factors influence the dose of warfarin required to maintain the INR within therapeutic range. Several single-nucleotide polymorphisms that affect warfarin metabolism have been identified. Algorithms based on pharmacogenetic and clinical factors improve the accuracy of warfarin dose initiation compared with algorithms based only on clinical factors.²⁴ Additional studies are required to determine whether the clinical benefits of genotyping of warfarin candidates justify the cost of genetic testing (see Chapter 8).

Novel Oral Anticoagulants

Direct thrombin inhibitors and factor Xa inhibitors have several advantages over vitamin K antagonists such as warfarin, the most notable being a fixed dosing regimen that eliminates the need for monitoring of a laboratory test such as the INR. Dabigatran, an oral direct thrombin inhibitor, and rivaroxaban and apixaban, factor Xa inhibitors, are approved by the U.S. Food and Drug Administration (FDA) for prevention of stroke/embolism in patients with nonvalvular AF. Randomized clinical trials demonstrated that each of these three NOACs is noninferior or superior to warfarin in efficacy and safety in patients with nonvalvular AF who had risk factors for stroke.²⁵ One of the most serious risks of anticoagulation is intracranial hemorrhage. Recent studies have indicated that the risk of intracranial hemorrhage is about 50% lower with NOACs compared to warfarin.²⁶

The NOACs, in addition to eliminating the need for laboratory monitoring, have other advantages over warfarin: fewer drug interactions, no food interactions, and rapid onset of action that obviates the need for bridging therapy. However, they also have some disadvantages compared to warfarin: higher cost, more gastrointestinal side effects in the case of dabigatran, twice-daily dosing for dabigatran and apixaban, and the absence of a readily available laboratory test to verify compliance. Furthermore, these agents cannot be used safely in patients with severe renal disease. Another limitation is that there are no specific reversal agents for all the NOACs. The FDA recently approved idarucizumab, an antibody fragment that reverses the anticoagulant effects of dabigatran within minutes.²⁷ Prothrombin complex concentrate can reverse the anticoagulant effect of the NOACs,²⁸ but specific and rapid-acting antidotes for rivaroxaban and apixaban are not yet available. Nonetheless, for many patients with AF, the advantages of the newer anticoagulants outweigh the disadvantages.

Older studies showed underutilization of warfarin in patients with AF and risk factors for stroke. The inconvenience and potential risks of warfarin likely contributed to its underutilization. However, underutilization of an oral anticoagulant in patients with AF continues to be the case even with the advent of NOACs and even in patients with a CHA₂DS₂-VASc score 3 or higher.²⁹

The major professional societies have incorporated recommendations regarding the use of the factor Xa and direct thrombin inhibitors into their most recent updated guidelines for the management of AF. The ACC/AHA/HRS practice guidelines recommend dabigatran, rivaroxaban, and apixaban as useful alternatives to warfarin for prevention of stroke or systemic embolism in patients with nonvalvular paroxysmal or persistent AF and risk factors for stroke.² This recommendation is limited to patients without a prosthetic valve, with creatinine clearance higher than 15 mL/min, and without impaired clotting function from advanced liver disease. The ESC guidelines recommend dabigatran, rivaroxaban, or apixaban for patients with AF in whom maintenance of a therapeutic INR during treatment with warfarin is difficult and state that one of these NOACs should be considered instead of dose-adjusted warfarin for most patients with nonvalvular AF, based on their net clinical benefits.¹³ These guidelines recommend that the NOACs not be used in patients with a creatinine clearance of less than 30 mL/min.

The results of a prospective randomized clinical trial and post hoc analysis of three other randomized clinical trials indicate that the NOACs are as effective as warfarin for prevention of thromboembolic complications associated with cardioversion.³⁰ This is the case regardless of whether a transesophageal echocardiogram is performed before cardioversion to look for left atrial thrombus.

The onset of action of dabigatran, rivaroxaban, and apixaban is approximately 1.5 to 2 hours after a dose. The half-life of dabigatran and apixaban ranges between 10 and 16 hours, and that of rivaroxaban is 5 to 9 hours. These anticoagulants lose most of their effect by 24 hours after discontinuation. The rapid onset of action and washout eliminates the need for bridging therapy with heparin when treatment with one of the NOACs is interrupted for a surgical or invasive procedure. Recent data indicate that the risk of major periprocedural complications does not differ significantly between patients who undergo radiofrequency catheter ablation of AF during uninterrupted therapy with warfarin and patients anticoagulated with a NOAC when the last dose of the NOAC is 1 to 2 days before the procedure and dosing is restarted 6 hours after the procedure, without bridging therapy.³¹

Low-Molecular-Weight Heparin

Low-molecular-weight heparin (LMWH) has a longer half-life than unfractionated heparin and a predictable antithrombotic effect that is attained with a fixed dosage administered subcutaneously twice daily. Because LMWH can be self-injected outside the hospital, it is a practical alternative to unfractionated heparin for initiation of anticoagulation with warfarin in patients with AF. Bridging therapy with LMWH be continued until the INR is 2.0 or higher.

Because of its high cost, LMWH rarely is used in clinical practice as a substitute for long-term conventional anticoagulation. LMWH typically is used as a temporary bridge to therapeutic anticoagulation when therapy with warfarin is initiated or in high-risk patients for a few days before and after a medical or dental procedure when anticoagulation with warfarin has been suspended.

Excision or Closure of the Left Atrial Appendage

Approximately 90% of left atrial thrombi form in the left atrial appendage (LAA), and therefore successful excision or closure of the LAA should greatly reduce the risk of thromboembolic complications in patients with AF. Surgical techniques consist of either excision or closure by suturing or stapling. The efficacy of these techniques is variable and probably dependent on both the technique and the operator.³² Transesophageal echocardiography (TEE) should be performed after surgical closure of the LAA to confirm successful closure before discontinuation of anticoagulation.

In recent years, several percutaneous LAA occlusion and suturing devices have been developed as alternatives to surgical closure techniques. These devices have their greatest utility in high-risk AF patients who cannot tolerate or who refuse to take an oral anticoagulant.

The only percutaneous occlusion device approved by the FDA specifically for stroke prevention as an alternative to warfarin is the WATCHMAN (Boston Scientific, Marlborough, Massachusetts). This nitinol plug covered with fenestrated fabric became widely available for clinical use after FDA approval in 2015. After implantation of the WATCHMAN using femoral vein access and transseptal catheterization, anticoagulation with warfarin is necessary for at least 45 days, at which time anticoagulation can be discontinued if there is no TEE evidence of peridevice flow. The most recent clinical trial and registry data indicated an implantation success rate of 95% and a procedural complication rate of approximately 2% to 3%, the most common complication being pericardial effusion requiring drainage.³² Approximately 95% of patients are able to discontinue warfarin at 45 days after implant. Clinical trial data have demonstrated noninferiority of the WATCHMAN compared to warfarin for stroke prevention and a significantly lower risk of hemorrhagic stroke during long-term follow-up.³²

Another device used in the United States for LAA occlusion is the LARIAT (Sentreheart, Redwood City, California). This device has FDA approval as a method for soft tissue approximation (not stroke prevention) and has been used “off label” in clinical practice in the United States and Europe for LAA occlusion. A guidewire with a magnetic tip is inserted into the left atrium after transseptal catheterization and is positioned at the tip of the LAA and functions as a rail for an epicardial snare. Entry into the pericardial space is attained using a percutaneous approach. A snare with a pretied suture is inserted into the pericardial space and guided toward the LAA. The pretied suture then is tightened to occlude the LAA. In a large multicenter registry, complete LAA closure was achieved in 94% of 712 patients.³³ There was one procedure-related death, and cardiac perforation occurred in 3.4% of patients, with open heart surgery required to repair the perforation in 1.4% of patients.³³ Clinical trial data demonstrating efficacy of the LARIAT for stroke prevention are lacking.

Pharmacologic Rate Control

An excessively rapid ventricular rate during AF often results in uncomfortable symptoms and decreased effort tolerance and can cause a tachycardia-induced cardiomyopathy if it is sustained for several weeks to months. Optimal heart rates during AF vary with age and should be similar to the heart rates that a patient would have at a particular degree of exertion during sinus rhythm. Heart rate control must be assessed both at rest and during exertion. At rest, the ideal ventricular rate during AF is in the range of 60 to 80 beats/min. During mild to moderate exertion (e.g., rapid walking), the target rate should be 90 to 115 beats/min. During strenuous exercise, the ideal rate is in the range of 120 to 160 beats/min. Optimal assessment of the degree of heart rate control is provided by an ambulatory 24-hour Holter recording or an exercise test.

Oral agents available for long-term heart rate control in patients with AF are digitalis, beta blockers, calcium channel antagonists, and amiodarone. The first-line agents for rate control are beta blockers and the calcium channel antagonists verapamil and diltiazem. A combination is often used to improve efficacy or to limit side effects by allowing the use of smaller dosages of the individual drugs. In patients with sinus node dysfunction and tachycardia-bradycardia syndrome, the use of a beta blocker with intrinsic sympathomimetic activity (pindolol, acebutolol) may provide rate control without aggravating sinus bradycardia.

Digitalis may adequately control the rate at rest but often does not provide adequate rate control during exertion. However, its use for rate control of AF is controversial because digitalis has been shown to increase the risk of all-cause mortality, particularly among patients with AF.³⁶

Amiodarone is much less frequently used for rate control than the other negative dromotropic agents because of the risk of organ toxicity associated with long-term therapy. Amiodarone can be an appropriate choice for rate control if the other agents are not tolerated or are ineffective. For example, amiodarone would be an appropriate choice for a patient with persistent AF, heart failure, and reactive airway disease who cannot tolerate either a calcium channel antagonist or a beta blocker and who has a rapid ventricular rate despite treatment with digitalis.

Strict heart rate control can be difficult to achieve pharmacologically. Based on the results of a single randomized study that demonstrated no significant differences in major outcomes between a lenient rate-control strategy (resting rate <110 beats/min) and a strict rate-control strategy (resting heart rate <80 beats/min, rate during moderate exercise <110 beats/min), a lenient rate-control strategy is reasonable if the patient remains asymptomatic and left ventricular systolic function is not compromised.² However, strict rate control often still is an appropriate goal for relief of symptoms, improvement in functional capacity, and avoidance of tachycardia-induced cardiomyopathy during long-term follow-up.

Pharmacologic Rhythm Control

The results of published studies on the efficacy of AADs for AF suggest that all the available drugs except amiodarone have similar efficacy and are associated with a 50% to 60% reduction in the odds of recurrent AF during 1 year of treatment. The one drug that stands out as having higher efficacy than the others is amiodarone. In studies that directly compared amiodarone with sotalol or class I drugs, amiodarone was 60% to 70% more effective in suppressing AF. However, because of the risk of organ toxicity, amiodarone is not appropriate first-line drug therapy for many patients with AF. Because the efficacy of rhythm-control agents other than amiodarone is in the same general range, the selection of an AAD to prevent AF often is dictated by the issues of safety and side effects.

Ventricular proarrhythmia from class IA agents (quinidine, procainamide, disopyramide) and class III agents (sotalol, dofetilide, dronedarone, amiodarone) is manifested as QT prolongation and polymorphic ventricular tachycardia (torsades de pointes). Risk factors for this type of proarrhythmia include female gender, left ventricular dysfunction, and hypokalemia. The risk of torsades de pointes appears to be much lower with dronedarone and amiodarone than with the other class III drugs. The ventricular proarrhythmia from class IC agents (flecainide and propafenone) manifests as monomorphic ventricular tachycardia, sometimes associated with widening of the QRS complex during sinus rhythm but not QT prolongation. Drugs most likely to result in ventricular proarrhythmia are quinidine, flecainide, sotalol, and dofetilide. In controlled studies, these agents increased the risk of ventricular tachycardia by a factor of 2 to 6.

Adverse drug events or side effects resulting in discontinuation of drug therapy are fairly common with rhythm-control drugs, with discontinuation rates reported to be as high as 40%.³⁷

The best options for drug therapy to suppress AF depend on the patient's comorbidities. In patients with lone AF or minimal heart disease (e.g., mild left ventricular hypertrophy), flecainide, propafenone, sotalol, and dronedarone are reasonable first-line drugs, and amiodarone and dofetilide can be considered if the first-line agents are ineffective or not tolerated. In patients with substantial left ventricular hypertrophy (left ventricular wall thickness >15 mm), the hypertrophy heightens the risk of ventricular proarrhythmia, and the safest choices for drug therapy are amiodarone and dronedarone.² In patients with coronary artery disease (CAD), several of the class I drugs have been found to increase the risk of death, and the safest first-line options are dofetilide and sotalol, with amiodarone reserved for use as a second-line agent. In patients with heart failure, several AADs have been associated with increased mortality, and the only two drugs known to have a neutral effect on survival are amiodarone and dofetilide (see Chapter 36).

At the time of FDA approval, it was known that dronedarone increases mortality in NYHA Class IV heart failure or patients with a recent episode of decompensated heart failure. After approval, the categories of patients in which dronedarone is contraindicated expanded based on the results of a randomized clinical trial that was discontinued prematurely because of major adverse drug effects.³⁸ Dronedarone should be avoided in patients age 65 or older with permanent AF and CAD, prior stroke, or symptomatic heart failure and in those age 75 or older with hypertension and diabetes.

Rhythm Control with Agents Other Than Antiarrhythmic Drugs

Experimental studies indicate that angiotensin-converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) have favorable effects on electrical and structural remodeling (see Chapter 34). This explains why some studies have shown that ACE inhibitors and ARBs prevent AF. However, other studies have demonstrated that these agents do not prevent AF. A meta-analysis demonstrated that the strongest beneficial effects have been observed with ARBs and in patients with heart failure.³⁹ However, the quality of the data in the meta-analysis studies was low, and at present, there is insufficient evidence to support the use of ACE inhibitors and ARBs for the sole purpose of preventing AF.

Some evidence indicates that statins prevent AF, perhaps because of their anti-inflammatory effects. A systematic review of 16 observational studies demonstrated a 12% reduction in the relative risk of new-onset AF and a 15% reduction in recurrent AF in patients treated with statins.⁴⁰ However, a meta-analysis of 11 randomized clinical trials concluded that statins do not prevent AF in patients who have not undergone open heart surgery.⁴⁰ Therefore, the available data do not support the use of statins for the prevention of AF.

Omega-3 polyunsaturated fatty acids (PUFAs) have anti-inflammatory and antioxidant effects and also can have direct ion channel effects. Epidemiologic studies suggested that fish oil could prevent AF, which lead to several randomized clinical trials that assessed the efficacy of PUFAs for AF prevention. These trials reported conflicting results, but the predominance of the data indicate that the daily intake of PUFAs or fish oil does not prevent recurrent AF after cardioversion or episodes of paroxysmal AF.⁴¹

Catheter Ablation of Atrial Fibrillation

Catheter ablation reliably and permanently eliminates several types of arrhythmias, such as atrioventricular nodal reentrant tachycardia (AVNRT) and accessory pathway–mediated tachycardias (see Chapters 36 and 37). Success rates greater than 95% are attainable when the arrhythmia substrate is well defined, localized, and temporally stable. In contrast, the arrhythmia substrate of AF is not well understood, usually is widespread, is variable between patients, and may be progressive. Furthermore, several factors that promote AF cannot be addressed simply by catheter ablation, including comorbidities (e.g., hypertension, obesity, obstructive sleep apnea), structural remodeling of the atria, inflammatory factors, and genetic factors (see Chapter 7). Therefore, whereas late recurrences of AVNRT or accessory pathway conduction are very rare, AF can recur more than 2 or 3 years after an initially successful ablation procedure.

Radiofrequency Catheter Ablation

A commonly used energy for catheter ablation of AF is radiofrequency (RF) energy. RF energy is delivered point by point, typically in association with a three-dimensional electroanatomic mapping system as a navigation guide and to create a visual record of the sites that already have been ablated (Fig. 38.7). To improve anatomic accuracy, the electroanatomic map of the left atrium can be merged with a computed tomography scan or magnetic resonance image of the left atrium and pulmonary veins (PVs) or with an ultrasound image generated by intracardiac echocardiography. An important determinant of lesion depth and durability is contact force, and the newest generation of RF ablation catheters provides the operator with immediate feedback on contact force.

Because of their important role in triggering and maintaining episodes of AF, almost all ablation strategies include electrical isolation of the PVs (Fig. 38.8). This is often sufficient for patients with paroxysmal AF and sometimes sufficient for patients with persistent AF. Pulmonary vein (PV) isolation can be accomplished by either ostial ablation or wide-area ablation 1 to 2 cm away from the ostia, in the antral regions of the PVs. Most of the available data indicate that wide-area ablation is more effective than ostial ablation, probably because it also targets drivers that are in the antrum, outside the PV itself.⁴³ Triggers of AF can also arise from other thoracic veins, such as the superior vena cava, coronary sinus, and the vein of Marshall. After the PVs have been isolated, infusion of isoproterenol is helpful to determine whether any non–PV triggers are present.

A variety of ablation strategies have been used for persistent AF after the PVs have been isolated: linear ablation across the left atrial roof, mitral isthmus, or cavotricuspid isthmus; ablation of complex fractionated atrial electrograms (CFAEs) in the left atrium, coronary sinus, or right atrium; various combinations of linear and CFAE ablation; and ablation of ganglionated plexi.⁴³ The endpoint of catheter ablation of persistent AF is either completion of a prespecified lesion set (in which case sinus rhythm is restored by cardioversion) or stepwise ablation until the AF converts to sinus rhythm. Recent randomized studies have demonstrated that antral PV isolation is as effective as PV isolation plus linear ablation and/or ablation of CFAEs.^44,45

A novel approach to ablation of AF is based on the hypothesis that AF is sustained by localized sources, either rotors or focal impulses.⁵ Signal processing using proprietary software allows identification of focal impulses and rotors that are targeted for RF ablation during ongoing episodes of AF. In one of the first clinical studies using this approach, AF termination or slowing was successfully achieved by ablation of the localized sources in 86% of cases. At a median of 9 months of follow-up, 82% of patients were free of AF, compared to 45% of patients in a control group who underwent conventional ablation.⁵ These early results suggested that focal impulse and rotor modulation can improve outcomes of catheter ablation of AF. However, more recent studies in patients with paroxysmal, persistent, and longstanding persistent AF have not confirmed the initial results of ablation of rotors and/or focal impulses. The single-procedure freedom from AF/atrial tachycardia (AT) off AADs at 6 to 18 months of follow-up was only 17% to 21% in these studies.^46,47 Therefore, although focal impulse and rotor modification is potentially an important step forward in understanding the mechanism of AF, its clinical value in patients undergoing catheter ablation of AF is unsettled.

When the efficacy of catheter ablation of AF is evaluated, recurrences of AF in the first 3 months after ablation usually are ignored. A 3-month blanking period excludes early recurrences that are caused by a transient inflammatory response or incomplete lesion maturation. Even in patients with symptomatic AF, postablation recurrences can be asymptomatic. Therefore, accurate assessment of efficacy requires monitoring for at least 7 days and preferably for 1 month with a device capable of detecting asymptomatic episodes of AF. Ideally, the monitoring should be performed at 6 and 12 months. Continuous monitoring for recurrent AF is possible in patients who have a pacemaker or ICD and an atrial lead. A miniaturized insertable cardiac monitor with wireless telemetry for remote monitoring (Reveal LINQ, Medtronic, Minneapolis, Minnesota) also allows for intense monitoring for postablation AF.⁴⁸

A wide range of success rates has been reported for RF catheter ablation of AF. When the latest-generation ablation catheters are used by experienced operators, 12-month freedom from AF/AT off AADs after a single procedure generally is 70% to 75% for paroxysmal AF^49,50 and 50% to 60% for persistent AF.^44,45

Recurrent AF after PV isolation in patients with paroxysmal AF most often is attributable to electrical reconnection of one or more PVs but sometimes is caused by triggers or drivers of AF originating in structures such as the superior vena cava or coronary sinus or in the atria themselves. After reisolation of the PVs and ablation of triggers outside the PVs, 12-month efficacy of 85% to 90% can be expected.

Redo ablation procedures in patients with persistent AF usually consist of re-isolation of the PVs if necessary, and additional ablation directed at substrate modification by targeting CFAEs or creating lines of conduction block at sites such as the left atrial roof or mitral isthmus.⁴³ A 12-month success rate of 75% to 85% typically is achieved after multiple ablation procedures for persistent AF.

Most recurrences of AF after RF catheter ablation occur within the first year of follow-up. However, recurrences continue to occur at a rate of approximately 10% per year at 1 to 3 years, then approximately 4% to 5%/year at 3 to 12 years after ablation.⁵¹ The predictors of late recurrences of AF include a history of persistent AF before the initial ablation procedure, age, left atrial size, diabetes, valvular heart disease, cardiomyopathy, and sleep apnea.⁴³ In addition, obesity and lack of cardiovascular conditioning have been reported to predispose to recurrences of postablation AF. Treatment of sleep apnea, weight loss, and improvement in cardiovascular fitness have been found independently to reduce the risk of recurrent AF after an ablation procedure.^52,53

Atrial tachyarrhythmias that occur after catheter ablation of AF can take the form of atrial tachycardia/flutter that can be either focal or reentrant. When the ablation strategy consists only of PV isolation, postablation focal ATs often are attributable to partial recovery of PV conduction. The incidence of reentrant AT/flutter after ablation is related to the extent of ablation at atrial sites other than around the PVs. When extensive ablation is performed in the left atrium or in both atria, AT/flutter can occur during follow-up in up to 50% or more of patients. These arrhythmias do not respond well to AAD therapy and often require another ablation procedure for elimination.

The risk of a major complication from RF catheter ablation of AF is reported to be 5% to 6%.^54,55 The most common major complications are cardiac tamponade, PV stenosis, and cerebral thromboembolism, each with a prevalence of approximately 1%. The risk of a femoral vascular injury is reported to be 1% to 2%. The risk of a major complication is more than twofold higher when the annual operator volume is less than 25 cases compared to more than 25 cases.⁵⁴

The risk of esophageal perforation is reported in the range of 0.01% to 0.02%.⁵⁶ Despite its rarity, this complication is of great concern because it often is lethal. Patients typically present 3 to 14 days after ablation with one of more of the following: dysphagia, odynophagia, fever, leukocytosis, bacteremia, and septic, thrombotic, or air emboli. Computed tomography of the chest with intravenous contrast is the diagnostic test of choice. The presence of contrast in the esophagus or air in the mediastinum or cardiac chambers is indicative of esophageal perforation or fistula formation. Instrumentation of the esophagus should be avoided.

Monitoring of the position of the esophagus and intraluminal esophageal temperature monitoring have been used to prevent esophageal injury during ablation along the posterior wall. Although these maneuvers may reduce the risk, they clearly do not prevent all cases of esophageal injury, since 90% of patients with an esophageal perforation had undergone monitoring of the esophageal position or temperature.⁵⁶ There is evidence that limiting the power of RF energy applications to 20 to 25 watts for less than 30 seconds when ablating along the posterior left atrial wall and the use of periprocedural proton pump inhibitors reduce the risk of esophageal injury.⁵⁷

Based on the results of a recent global survey, 72% of patients with an esophageal perforation had evidence of an atrial-esophageal fistula, and mortality among these patients was 79%. In contrast, among the 28% of patients with an esophageal perforation who did not have an atrial-esophageal fistula, mortality was 13%.⁵⁶ This highlights the importance of early diagnosis and treatment of esophageal perforations. Early surgical intervention is appropriate regardless of whether an atrial-esophageal fistula is present.

A recently recognized complication of RF ablation is silent cerebral ischemic lesions. Cerebral magnetic resonance imaging has demonstrated silent cerebral ischemic lesions in 2% to 14% of patients in whom an irrigated-tip RF ablation catheter is used.⁵⁸ The long-term clinical significance of these lesions is unclear.

Recent randomized clinical trials have demonstrated that RF catheter ablation of AF is superior to AAD therapy for prevention of AF. In previously untreated patients with paroxysmal AF, there was an absolute difference of 18% in the 12-month risk of recurrent AF in favor of RF catheter ablation over AAD therapy.⁵⁹ Among patients with persistent AF, there was an absolute difference of 27% in the 12-month risk of recurrent AF in favor of RF catheter ablation over AAD therapy.⁶⁰

Cryoballoon Ablation

In 2010 a cryoballoon catheter designed to isolate PVs became widely available for use in the United States. In contrast to point-by-point RF ablation around the PVs, the cryoballoon was designed to fit into the antrum of a PV and to create a circumferential ablation lesion using cryoenergy. Cryoenergy is delivered through the entire distal half of the second-generation cryoballoon catheter currently in clinical use. Complete occlusion of the PV by the inflated balloon is necessary for reliable PV isolation (Fig. 38.9). Two 4-minute applications of cryoenergy initially were recommended for each of the PVs. With the second-generation cryoballoon catheter, a single 3- or 4-minute application of cryoenergy often is sufficient to create durable PV isolation in a majority of PVs (Fig. 38.10). Among patients with paroxysmal AF, with a mean of 1.1 to 1.7 applications of cryoenergy per PV, 1-year freedom from recurrent AF was 80% to 82% using a single-freeze protocol.^61,62

Avoidance of entry of the cryoballoon into the luminal portion of a PV is important to avoid PV stenosis. The most commonly used cryoballoon catheter has a 28-mm diameter when the balloon is fully inflated. The relatively large size of the balloon typically allows occlusion of a PV from the antrum. With use of the 28-mm balloon, up to 40% of the posterior wall between the left and right PVs is rendered electrically silent.⁶³

A multielectrode catheter inserted through a central lumen of the cryoballoon catheter often allows recording of PV potentials during an application of cryoenergy. Disappearance or dissociation of PV potentials within the first minute of a cryoenergy application is a strong independent predictor of durable PV isolation⁶⁴ (Fig. 38.10). Other independent predictors are achieving a balloon temperature of −40°C in less than 60 seconds during an application of cryoenergy and an interval thaw time to 0°C of longer than 10 seconds on completion of a cryoenergy application.^64,65

There is a lower clinical success rate with use of the cryoballoon catheter in patients with persistent AF than in those with paroxysmal AF. Among patients with persistent AF who undergo PV isolation using the 28-mm cryoballoon catheter, freedom from AF at 12 months of follow-up was 60%.⁶⁶ This is similar to the clinical results achieved in similar patients when wide-area circumferential PV isolation is performed by point-by-point RF catheter ablation.⁶⁶

Possible complications associated with the cryoballoon catheter have the same risk as with RF catheter ablation of AF, including cardiac tamponade (approximately 1%) and femoral vascular injury (1% to 2%).

The most common complication of cryoballoon ablation is phrenic nerve injury. During early experience with the cryoballoon catheter, the incidence of right phrenic nerve injury was approximately 10%, with the injury resolving within 12 months in almost all patients.⁶⁷ Injury to the left phrenic nerve is possible but rare. It now is standard clinical practice to pace the phrenic nerve with a catheter positioned in the superior vena cava or right subclavian vein during cryoballoon ablation at the right PVs. Various strategies are available to monitor diaphragmatic contraction or phrenic nerve function during phrenic nerve pacing, including direct palpation of diaphragmatic contraction and monitoring the diaphragmatic compound motor action potential.⁶⁸ The immediate discontinuation of an application of cryoenergy on the first evidence of phrenic nerve injury greatly reduces the risk of long-lasting or permanent injury. In recent experience with the 28-mm cryoballoon catheter, the risk of right phrenic nerve injury is as low as 1.5% to 2%.^62,69

A small number of case reports have made it clear that death from an atrial-esophageal fistula is a potential complication of cryoballoon ablation. Measures to minimize the risk of esophageal injury are appropriate, including the periprocedural use of a proton pump inhibitor and monitoring of intraluminal esophageal temperature during cryoablation. It is common practice to discontinue cryoablation if the esophageal temperature drops to 30°C. This provides a large safety margin for avoiding esophageal injury.⁷⁰

Ablation of the Atrioventricular Node

Radiofrequency catheter ablation of the AV node results in complete AV nodal block and substitutes a regular, paced rhythm for an irregular and rapid native rhythm. It is a useful strategy in patients who are symptomatic from AF because of a rapid ventricular rate that cannot be adequately controlled pharmacologically as a result of inefficacy or intolerance to rate-control drugs, and who either are not good candidates for AF ablation or already have undergone unsuccessful attempts at ablation. AV node ablation also can be helpful in patients with heart failure and AF to maximize the benefits of cardiac resynchronization therapy (CRT) if there already is not 100% ventricular pacing.⁷³

In patients with AF and an uncontrolled ventricular rate, AV node ablation improves the left ventricular ejection fraction (EF) if there is a tachycardia-induced cardiomyopathy. AV node ablation also has been shown to improve symptoms, quality of life, and functional capacity and to reduce the use of health care resources.⁷⁴

The disadvantages of AV node ablation are that it creates a lifelong need for ventricular pacing and does not restore AV synchrony. Although symptoms and functional capacity typically improve after AV node ablation in patients with AF and an uncontrolled ventricular rate, some patients may not feel as well as during sinus rhythm.

Atrioventricular node ablation is a technically simple procedure with an acute and long-term success rate of 98% or higher and a very low risk of complications. In patients with persistent AF, a ventricular pacemaker is implanted. A dual-chamber pacemaker is appropriate if the AF is paroxysmal. Most patients have a good clinical outcome with right ventricular pacing, but in patients with left ventricular dysfunction, biventricular pacing for CRT is appropriate.⁷⁵ In patients with an ischemic or nonischemic cardiomyopathy and EF of 30% to 35% or lower, an ICD may be appropriate for primary prevention of sudden death. However, a pacemaker without the ICD often is adequate for patients with a borderline EF (30% to 35%) and a rapid ventricular rate because the EF is likely to improve after the ventricular rate has been controlled by AV node ablation.⁷⁴ His bundle pacing can be considered in some patients to avoid problems with right ventricular pacing.

Surgical Approaches to Atrial Fibrillation

The most effective surgical procedure for AF is the “cut-and-sew” maze procedure developed by Cox in 1987. This operation involves multiple atrial incisions to isolate the PVs and to create lines of block in the left atrium and right atrium. In addition, the left and right atrial appendages are excised. In a study from a highly experienced surgical center in patients who underwent periodic 24-hour Holter monitoring during follow-up, there was 83% freedom from AF off drugs at a median follow-up of 5.9 years after the most recent version of the cut-and-sew maze procedure (Cox maze III).⁷⁶ However, because continuous monitoring for only 24 hours at a time often is insufficient to detect recurrent episodes of AF, the 83% long-term freedom from AF likely was an overestimate.

The cut-and-sew Cox maze procedure has not been widely performed because it requires cardiopulmonary bypass, is technically difficult, and is associated with a mortality rate of 1% to 2%.

A large variety of surgical ablation tools have been developed to simplify the Cox maze III procedure. These tools allow the surgeon to substitute an ablation line for a surgical incision. Some surgeons use a minimally invasive approach in which the ablation tools are inserted through small incisions between the ribs, and thoracoscopic video-assisted epicardial ablation is performed. Several different types of energy have been used for surgical ablation: RF energy, cryoenergy, microwave, laser, and high-intensity focused ultrasound. The tool that most consistently produces transmural ablation lesions is a clamp device developed to isolate the PVs using bipolar RF energy.

Various surgical ablation strategies have been used, including PV isolation, left atrial ablation, and the Cox maze lesion set (Cox maze IV) in which a combination of RF and cryothermal ablation lines replace most of the surgical incisions. In a large series of patients with AF who underwent the Cox maze IV procedure, freedom from AF off AADs at 5 years was 66%, with no difference in efficacy between patients with paroxysmal versus persistent AF or between patients with a stand-alone procedure versus a concomitant procedure.⁷⁷

Surgical therapy for AF is appropriate as a concomitant procedure in patients with symptomatic AF undergoing open heart surgery for CAD or valvular disease. A stand-alone surgical procedure for AF is an option for patients who have not had a successful outcome from catheter ablation, who are not good candidates for catheter ablation, or who prefer a surgical approach over catheter ablation.

Hybrid Approach to Ablation of Atrial Fibrillation

Some AF ablation centers use a hybrid approach that combines surgical and catheter ablation, either on the same day or staged, with the two components separated by days to weeks. The rationale for the hybrid approach is that it capitalizes on the strengths of the two approaches and minimizes their limitations.⁷⁸ The surgical component is performed first, allowing exclusion of the left atrial appendage, direct visualization of the antrum of the PVs, access to epicardial structures such as the ligament of Marshall and ganglionated plexi, and avoidance of damage to the esophagus and phrenic nerve. A subsequent catheter ablation procedure consists of additional ablation as needed, for PV isolation, conduction block across epicardial ablation lines, ablation of complex fractionated electrograms or rotors, and ablation of sites that are not accessible from the epicardium, such as the tricuspid or mitral isthmus.

The 12-month freedom from AF off AADs after the hybrid approach has varied widely, from as low as 37% to as high as approximately 80%.⁷⁸ This wide range of success rates likely is attributable to differences in patient selection, operator skill, specific technologies used for ablation, and lesion sets.

In the absence of data from prospective randomized studies, the incremental value of the hybrid approach over surgical or catheter ablation alone has remained unclear.

Wolff-Parkinson-White Syndrome

Patients with the WPW syndrome and an accessory pathway with a short refractory period can experience a very rapid ventricular rate during AF (see Chapters 36 and 37). Ventricular rates greater than 250 to 300 beats/min can result in loss of consciousness or precipitate ventricular fibrillation and a cardiac arrest. Patients with WPW syndrome who present in AF with a rapid ventricular rate should undergo transthoracic cardioversion if there is hemodynamic instability. If the patient is hemodynamically stable, intravenous procainamide or ibutilide can be used for pharmacologic cardioversion. Procainamide may be preferable to ibutilide because it blocks accessory pathway conduction and slows the ventricular rate before AF has converted to sinus rhythm. Digitalis and calcium channel antagonists are contraindicated in patients with WPW syndrome and AF. These agents selectively block conduction in the AV node and can result in acceleration of conduction through the accessory pathway.

The preferred therapy for patients with WPW syndrome and AF with a rapid ventricular rate is catheter ablation of the accessory pathway. The efficacy of catheter ablation is 95% or higher for most types of accessory pathways, and the risk of a major complication is very low. AF typically no longer recurs after successful accessory pathway ablation, probably because AF in the WPW syndrome often is induced by tachycardia and a result of atrioventricular reciprocating tachycardia (AVRT).

Congestive Heart Failure

Atrial fibrillation is a common arrhythmia in patients with heart failure, with a prevalence ranging from 10% in patients with New York Heart Association (NYHA) functional Class I up to 50% in Class IV patients (see Chapters 24 and 25). AF may be the cause of heart failure in patients who present with a nonischemic cardiomyopathy and AF with a rapid ventricular rate. Occasionally, AF causes left ventricular dysfunction and heart failure even when the ventricular rate is not rapid. In patients with structural heart disease and preexisting left ventricular dysfunction, AF can worsen the heart failure. The deleterious hemodynamic effects of AF are mediated by a rapid and/or irregular ventricular rate and loss of AV synchrony.

The most appropriate rate-control drugs in patients with systolic heart failure are beta blockers and digitalis. If necessary, amiodarone also can be used for rate control. In patients with diastolic heart failure, nondihydropyridine calcium antagonists can be used safely for rate control. Amiodarone and dofetilide are the only two rhythm-control drugs that are not associated with an increased risk of death in patients with heart failure.

As in other patients with AF, the decision to pursue a rate-control or a rhythm-control strategy in patients with heart failure should be individualized. If the heart failure is a result of AF, a rhythm-control strategy should be employed. In patients with heart failure who develop AF, there are no significant differences in all-cause mortality, cardiovascular mortality, or worsening heart failure between patients treated with a rate-control strategy or a pharmacologic rhythm-control strategy.² However, although symptoms improve with both a rate- and rhythm-control strategy, sinus rhythm is associated with greater improvement in NYHA functional class and quality of life.⁸²

A limitation of a pharmacologic rhythm-control strategy is that many patients continue to experience episodes of AF. Catheter ablation is more effective than AADs for preventing recurrent AF and therefore could have incremental benefit over drug therapy in patients with heart failure and AF. In patients with heart failure and persistent AF who were randomly assigned to either catheter ablation or treatment with amiodarone, freedom from AF at 24 months was 70% in the catheter ablation group compared to 34% in the amiodarone group, and the relative risks of unplanned hospitalizations and mortality were reduced by 45% to 55% in the catheter ablation group.⁸³ The EF increases by approximately 13% when AF is successfully eliminated in patients with left ventricular dysfunction.⁸⁴

Four small-scale randomized studies (sample sizes of 41 to 81 patients) compared catheter ablation to either pharmacologic rate control or AV junction ablation in patients with left ventricular dysfunction, and NYHA Class II to IV heart failure.⁸⁵ A majority of the patients had persistent AF. Freedom from AF ranged from 50% to 81% at 6 months and was 88% at 12 months after catheter ablation. Compared to rate control, catheter ablation was associated with significantly greater improvements in EF, functional capacity, brain natriuretic peptide levels, and quality of life.

These results suggest that maintenance of sinus rhythm is preferable to rate control in patients with heart failure and that catheter ablation of the AF should be considered if sinus rhythm is not maintained by amiodarone or dofetilide. Large-scale multicenter trials currently are in progress and will be helpful in clarifying the role of catheter ablation in patients with heart failure and AF.

A rate-control strategy is appropriate for patients who do not respond adequately to amiodarone or dofetilide and either are not suitable candidates for catheter ablation of the AF or have had an unsuccessful outcome from ablation. AV node ablation is appropriate for patients when ventricular rate during AF is not adequately controlled by drug therapy. Because left ventricular dysfunction and heart failure can be aggravated by right ventricular pacing, biventricular pacing should be performed after AV node ablation. The decision to implant a biventricular pacemaker versus a biventricular ICD is based on clinical judgment. If it seems likely that the EF will remain less than 30% to 35% after optimal heart rate control, a biventricular ICD is appropriate for primary prevention of sudden cardiac death.

In patients with heart failure who undergo CRT, AF often results in intrinsic and/or fused QRS complexes even when the rate is considered to be adequately controlled. This can limit the extent of biventricular capture. If this is the case in a patient with heart failure and AF, AV node ablation can maximize the benefit of CRT by resulting in 100% biventricular capture.⁷³

Hypertrophic Cardiomyopathy

Atrial fibrillation occurs in approximately 25% of patients with hypertrophic cardiomyopathy (HCM) and can cause severe hemodynamic impairment because of an inadequate diastolic filling time and loss of atrial-ventricular synchrony. Because of a high risk of thromboembolic complications, anticoagulation is indicated in AF patients with HCM, independent of the CHA₂DS₂-VASc score.

Severe left ventricular hypertrophy increases the risk of drug-induced torsade de pointes, and the only rhythm-control agents recommended for patients with HCM and a wall thickness greater than 1.5 cm are dronedarone and amiodarone.² Catheter ablation of AF also is an option. However, the responses to drug therapy and catheter ablation often are suboptimal because of the extensive atrial remodeling that may be present in patients with HCM.

Eight observational cohort studies reported on the results of RF catheter ablation of AF in patients with HCM.⁸⁶ The median follow-up in these studies was 18 to 19 months, and the rate of redo ablation procedures was 43%. Freedom from AF in these studies was a median of 66% to 86% for patients with paroxysmal AF and 42% to 65% for patients with persistent AF. These results indicate that catheter ablation is a reasonable option in patients with HCM and AF who do not respond well to AAD therapy. However, because of the prominent structural atrial abnormalities that can be associated with HCM, AF is likely to recur during long-term follow-up in many of these patients.

Percutaneous Coronary Intervention for Coronary Artery Disease

In patients with stable ischemic heart disease, treatment with aspirin plus clopidogrel (double-antiplatelet [-antithrombotic] therapy, DAPT) is recommended for a minimum of 1 month after placement of a bare metal stent and for a minimum of 6 months after placement of a drug-eluting stent. In patients with an acute coronary syndrome who undergo percutaneous coronary intervention with either a bare metal stent or a drug-eluting stent, DAPT is recommended for at least 12 months.

In patients who require long-term anticoagulation for AF, the addition of DAPT increases the risk of serious bleeding. The results of a randomized clinical trial indicated that therapy with a vitamin K inhibitor plus clopidogrel after stent placement reduces the risk of major bleeding events compared to therapy with a vitamin K inhibitor plus DAPT, without increasing the risk of adverse outcomes such as myocardial infarction, death, or stent thrombosis.⁸⁷

Pregnancy

The prevalence of AF during pregnancy is very low, approximately 60/100,000 pregnancies. When it occurs, AF often is in the setting of underlying congenital or valvular heart disease, thyrotoxicosis, or electrolyte abnormalities. Pregnancy is associated with a hypercoagulable state, but no data indicate that pregnancy increases the risk of thromboembolic complications related to AF. In women with paroxysmal AF before pregnancy, the frequency of episodes may or may not increase during pregnancy.

The decision to anticoagulate a pregnant woman with AF should be made using the same criteria as in nonpregnant women. If anticoagulation is deemed necessary, warfarin (not an NOAC) is recommended from the second trimester until 1 month before the due date, and subcutaneous low-molecular-weight heparin is recommended during the first trimester and during the final month of pregnancy.

Transthoracic cardioversion is considered safe at all stages of pregnancy. The recommended pharmacologic agents for acute management of AF consist of intravenous metoprolol for rate control and flecainide or sotalol for conversion to sinus rhythm. If ongoing therapy is deemed necessary, the recommended rate-control drug is digoxin. If ineffective, a beta blocker can be used, but only after the first trimester. If there is no structural heart disease, flecainide and sotalol are recommended for long-term rhythm control.⁸⁸ In the patient with structural heart disease, amiodarone is recommended for rhythm control.