Atrial fibrillation (AF) is the most common sustained arrhythmia worldwide with an increasing incidence with age. Adults at the age of 40 years have a 25 percent risk of developing AF during their lives.¹ The actual incidence and prevalence of AF is likely underestimated due to undetected asymptomatic occurrences or patients ignoring paroxysmal episodes. Experts predict that the number of AF-related hospitalizations will increase and almost double to 3.5 million by the year 2025 in the United States.

Although AF itself is not considered a life-threatening arrhythmia, it is associated with significant morbidity and mortality secondary to hemodynamic compromise and tachycardia-induced cardiomyopathy in some patients. Besides palpitations resulting in discomfort and anxiety, the loss of synchronous atrioventricular contractility may cause various degrees of ventricular dysfunction, exercise intolerance or congestive heart failure in patients with AF. Furthermore, stasis of blood flow in the fibrillating left atrium increases the risk of thrombus formation. Thromboembolic events and stroke remain the most feared complication in these patients. AF accounts for about 25 percent of strokes in patients older than 80 years and increases a person’s risk of stroke by 5-fold.² Overall, AF has been identified as an independent risk factor for mortality. Using the data from the Framingham study, the established risk factor–adjusted odds-ratio for death in men and women with AF was 1.5 and 1.9, respectively.³

The pharmacological treatment of AF has been disappointing with failure rates as high as 60 percent. Besides the significant side effects associated with antiarrhythmic drugs, their long-term efficacy in maintaining normal sinus rhythm is poor. In the AFFIRM-trial, rhythm control strategies were compared to rate control and anticoagulation. There was no survival benefit of rhythm over rate control. This appeared to be due to the use of antiarrhythmic drugs to maintain sinus rhythm as the AFFIRM-trial confirmed that the presence of normal sinus rhythm significantly decreased the risk of death with a hazard ratio of 0.53 (p <0.001).⁴

The persistent limitations of pharmacological therapy of AF have led to the development and proliferation of interventional approaches in the treatment of AF over the past two decades. The most promising and successful options have included catheter ablation and surgery. Strategies for interventional treatment have been based on the type of AF. A recent consensus statement of the Heart Rhythm Society, the European Heart Rhythm Association, the American College of Cardiology, the American Heart Association, and the Society of Thoracic Surgery defined three different types of AF including paroxysmal, persistent, and longstanding persistent AF (Table 43-1).⁵ The foundation of catheter ablation has been the isolation of the pulmonary veins (PVI). Results have been variable with single-procedure success rates between 16 and 84 percent.⁵ Although PVI has had good success rates in most patients with paroxysmal AF, certain patient subgroups have done poorly, such as patients with longstanding persistent AF and large atria. A recent report from Haïssaguerre’s group, who pioneered PVI, reported a single procedure success rate as low as 29 percent after 5 years.⁶ The surgical treatment of AF has had higher success rates in preventing AF recurrence. The biatrial Cox-Maze procedure (CMP) III has been the gold standard for more than one decade, with a freedom from symptomatic AF of 97 percent, and has been equally effective in all types of AF.

AF is a manifestation of a complex and progressive disease of the atria with electrical, contractile and structural remodeling that, in most incidences, gradually worsens over time.⁷

There are four main components that play an important role in the initiation and sustention of AF: (1) a trigger, which can be a single premature or a run of focal ectopic depolarization; (2) the effective atrial refractory period; (3) the conduction velocity; and (4) the geometry or anatomy of the atrium. Although the precise mechanisms that cause AF are incompletely understood, AF appears to require both an initiating event and a permissive atrial substrate. There is supportive data for multiple wavelets, mother waves, fixed or moving rotors, and macro-reentrant circuits playing a role in AF, and in a given patient, multiple mechanisms may coexist at any given time.

Automaticity and Triggered Activity

Enhanced automaticity occurs when myocytes possessing pacemaker activity increase their rate of spontaneous discharge. It can be due to a lowered threshold of the action potential upstroke, a less negative maximal diastolic potential, or an increase in the slope of spontaneous diastolic depolarization. Triggered activity arises from membrane oscillations following normal action potentials. If such oscillation reaches threshold of depolarizing currents, they can provoke new action potentials. Under certain circumstances, such triggered responses can in turn elicit new action potentials, resulting in self-sustaining runs of triggered activity. After Haïssaguerre and colleagues reported on the role of pulmonary veins (PVs) in initiating AF, most interventional treatment strategies focused on the isolation of triggers from these anatomic sources.⁸ While paroxysmal AF often initiates in the PVs, biatrial mapping studies have shown that the premature beats that trigger AF were located in the PVs in only 53 percent of the time. However, data about the presence of pacemaker cells in the PVs or other atrial areas outside the sinus node are controversial. The conduction velocity within the PVs is slower (0.3 m/s) compared with the surrounding atrial tissue (0.9 m/s). The complex fiber orientation spiraling perpendicularly around the vein with other fibers running parallel to the vein, and the lack of Connexin 40 in the veins, a major determinant of conduction velocity in the atria, probably contribute to this slower conduction. This creates heterogeneity of electrical conduction around the PVs, which is theorized to promote reentry and sustained AF. Although the PVs have the substrates needed to initiate and maintain AF in some patients, in almost 50 percent of cases there are sources present outside the PVs. In the group of patients undergoing concomitant cardiac surgery for structural heart disease, it is important to remember that the underlying mechanisms cannot be necessarily translated from experiences drawn from the electrophysiological laboratory with lone AF.

Mechanisms of Reentry

There are multiple theories for the development of reentrant circuits in the atrium.⁹ Circus movement reentry is characterized by an activation that can travel around a preformed anatomical structure, like the orifice of the PVs, and reactivate previously excited tissue. A short refractory period and a low conduction velocity make circus movement reentry more likely. The minimal path length for circus movement reentry can be calculated as the product of conduction velocity and refractory period (wavelength). If the path of the circuit is longer than the wavelength, there is a delay between recovery of the tissue and the moment of reexcitation, which is called the “excitable gap” (Fig. 43-1A). Initiation of circus movement reentry requires unidirectional conduction block often occurring in regions with long refractory periods.

In the leading circle concept, the reentrant circuit does not require an anatomical obstacle but rather moves around a core that is not fully activated. The dynamics of reentry are determined by the smallest possible loop in which the impulse can continue to circulate. Under this condition, the wavefront must propagate through relatively refractory tissue, making it a relatively unstable reentrant since there is no fully excitable gap (Fig. 43-1B). In the spiral wave theory, the rotor rotates around an unexcited yet excitable core. Along the wavefront, there is a decline in conduction velocity until block occurs (dotted line in Fig. 43-1C). One way of visualizing such a wave is to think of a broken wave that curls at its broken end and begins to rotate. Multiple unstable rotors resemble conduction patterns similar to that of the multiple wavelet hypothesis. According to Moe’s computer simulation, wavefronts continuously undergo wavefront–wavetail interactions resulting in wave breaks and generation of new wavefronts (Fig. 43-1D).

As the underlying distribution of refractory periods becomes more heterogeneous, unidirectional block can occur. This is a necessary condition for the initiation of reentry. When unidirectional block occurs, reentrant arrhythmia will initiate only if a critical mass of tissue is present. The critical mass is determined by the wavelength of the reentrant circuit. As atrial size increases, conduction velocity is slowed or the atrial refractory period is decreased, and the probability of initiating and sustaining AF increases. Although numerous non-pharmacologic approaches to treat AF, like isolating triggers in PVs, have involved altering one of these substrates, it must be remembered that the other substrates can still be active.

Atrial Remodeling

Studies of Alessie in the early 1990s advanced our understanding of AF by the observation that AF itself induces changes in the atrium that result in an increased susceptibility to further AF.¹⁰ He demonstrated that, during atrial pacing, action potential duration and atrial refractory period both decreased. This electrical remodeling led to sustained AF. However, structural changes of the atrium contribute to this remodeling process as well. Age, hypertension, congestive heart failure, or heart valve diseases are all strong predictors for the development of AF. These characteristics all increase atrial wall stress and some degree of dilatation in the atria by increased volume or pressure load. All these conditions can be the cause of the development of fibrosis and myocyte hypertrophy. These pathological changes can result in slow conduction or conduction block in some areas of the atrium, further increasing susceptibility to AF.

AF affects about 2 percent of the general population and almost 3 million people in the United States. It is strongly correlated to age and the prevalence increases from 0.1 percent in persons younger than 55 years to about 5 percent in persons 60 years or older and approaches 10 to 15 percent for patients over the age of 80. Overall, significantly more men than women are affected by AF.

There are a number of cardiovascular disorders associated with AF including hypertension, heart failure, valvular heart disease, and ischemic heart disease. All of these comorbidities can cause increased stress on the atrial wall resulting in interstitial fibrosis and atrial dilatation. However, in 10 to 15 percent of patients, there is no underlying cardiac pathology present (lone AF).

Stasis of atrial blood and resultant thrombus formation associated with AF increases a person’s risk of stroke by 5-fold and the rate of stroke averages about 5 percent/year in affected patients. It increases substantially in the presence of other cardiovascular diseases. While the prevalence of stroke is low in patients under the age of 60 at less than 0.5 percent, it approaches 30 percent for patients between 80 and 90 years. The intensity of anticoagulation involves a balance between prevention of ischemic stroke and avoidance of hemorrhagic complications. An international normalized ratio (INR) of 2 to 3 is recommended in patients with AF.

AF is a costly public health issue. A study analyzing three federally funded databases in the United States in 2005 suggested a total annual cost of US$6.65 billion for the treatment of AF, including US$2.93 billion (44 percent) for hospitalizations with a principal discharge diagnosis of AF, US$1.95 billion (29 percent) for the incremental inpatient cost of AF as a comorbid diagnosis, US$1.53 billion (23 percent) for outpatient treatment of AF, and US$235 million (4 percent) for prescription drugs.¹¹ In all analyses, AF was a significant contributor to hospital cost.

The indications for surgical ablation have been defined in a recent consensus statement and include⁵:

1. 
   

   All symptomatic patients with documented AF undergoing other cardiac surgical procedures

   
   
2. 
   

   Selected asymptomatic patients with AF undergoing cardiac surgery in which the ablation can be performed with minimal risk in experienced centers

   
   
3. 
   

   Stand-alone surgery should be considered for symptomatic patients with AF who either prefer a surgical approach, have failed one or more attempts of catheter ablation, or are not candidates for catheter ablation.

There is still controversy regarding the referral of patients for surgery with medically refractory, symptomatic lone AF in lieu of less invasive catheter ablation. In our opinion, there are relative indications for surgery that were not included or expressed in the consensus statement:

1. 
   

   AF patients who develop a contraindication to long-term anticoagulation and have a high risk for stroke (CHADS score ≥2) are excellent candidates for surgery. The CMP not only is able to eliminate AF in most of these patients, but also amputates the left atrial appendage, which is known to be one of the main sources of atrial thrombus formation. The stroke rate following the procedure off anticoagulation has been remarkably low, even in patients with high CHADS scores.

   
   
2. 
   

   Patients with longstanding AF who have suffered a cerebrovascular accident despite adequate anticoagulation are at high risk for repeat neurological events. In our experience of over 200 patients with a stand-alone CMP, there was only one late stroke, with over 80 percent of patients off anticoagulation at last follow-up. This is remarkable in that about 20 percent of patients had experienced at least one cerebrovascular accident before surgery.

   
   
3. 
   

   AF patients with a clot in the left atrial appendage are not candidates for catheter ablation and should be considered for surgical ablation.

In the era of minimally invasive approaches, including thoracoscopic and hybrid procedures, a team approach including both electrophysiologists and surgeons should be established to ensure appropriate patient and treatment selection.

It has not been possible to locate the precise mechanism of AF in the majority of patients prior to surgery due to the limitations of endocardial non-contact mapping and the invasiveness of epicardial mapping. Traditional epicardial activation sequence mapping is difficult and time-consuming and has not been suitable to provide useful real-time information during the procedure. Improvements in technology and further experimentation may provide the means for faster data analysis; a map-directed approach may be feasible in the future. Electrocardiographic imaging (ECGI) can be used to compute epicardial potentials noninvasively and reconstruct a 3-dimensional anatomical map of atrial electrical activity using multiple surface electrodes and anatomic data obtained through CT.

Besides documentation of the patient’s heart rhythm prior to surgery, a preoperative transthoracic and an intraoperative transesophageal echocardiogram should be performed to determine the left atrial diameter and to evaluate for the presence of a left atrial thrombus. Left atrial size is a significant predictor of failure in our series and is important to define prior to surgery.¹² It is our policy that in patients with a left atrial size ≥7 cm, a left atrial reduction procedure should be considered. A cardiac catheterization will provide information about the presence of coronary artery disease. Moreover, the anatomical location of the circumflex coronary artery is important to define in order to safely perform the left atrial isthmus lesion and the ablation of the coronary sinus. In patients who have failed catheter ablation, a chest CT is indicated to assess for PV stenosis. It is also important to determine the precise atrial tachyarrhythmia in patients who have failed catheter ablation; that may require an electrophysiological study. Atrial tachycardia is usually not amenable to surgical ablation and atypical atrial flutter usually necessitates a full Cox-Maze lesion set.

In 1987, Dr. James Cox introduced the Maze procedure for the surgical treatment of AF at Washington University in Saint Louis. Following extensive experimental investigation, his surgical approach was designed to block the multiple macroreentrant circuits which were believed to be the cause of AF.¹³ Unlike other historical attempts to address AF, such as the left atrial isolation procedure, His bundle ablation, the corridor procedure, or the atrial transection procedure, the CMP successfully restored both AV synchrony and sinus rhythm, thus potentially reducing the risk of thromboembolism and stroke. The operation consisted of creating surgical incisions in both the right and left atria. These incisions were placed in a pattern that allowed the propagation of the sinoatrial node impulse throughout both atria (Fig. 43-2A). This resulted in a preserved activation of most of the atrial myocardium, resulting in preservation of atrial transport function in most patients.

The original version, the CMP I, was modified due to problems with late chronotropic incompetence and a high incidence of pacemaker implantations. The CMP II was technically difficult to perform and was soon replaced by the final iteration known as the CMP III (Fig. 43-2B).¹⁴ The CMP III, often referred to as the “cut-and-sew” Maze, became the gold standard for the surgical treatment of AF with excellent long-term success rates and a low postoperative risk of stroke.¹⁵ In our series at Washington University, 97 percent of 198 consecutive patients who underwent this procedure with a mean follow-up of 5.4 years were free from symptomatic AF. There was no difference in success rates between patients undergoing a stand-alone CMP and those undergoing concomitant procedures and it was also equally efficacious in paroxysmal and persistent or longstanding persistent AF.¹⁵

Despite its proven efficacy, the CMP III was technically challenging and invasive, and few cardiac surgeons still perform this operation today. At most centers, the surgical incisions have been replaced with linear lines of ablation using a variety of different energy sources. These ablation-assisted procedures have greatly expanded the field of AF surgery in the last decade. With present ablation technology, surgery can be performed with low mortality and limited access incisions while preserving the high success rates of the traditional procedure.