**3. Electrophysiology of AF**

In recent years there has been considerable progress in understanding the pathophysiology of AF. It was initially thought that AF was caused by random, multiple wavelets generated throughout the atria that propagated new wavelets to cause re-entry mechanism. In the 1980s, several procedures were developed aimed at curing atrial fibrillation. However, most of these were abandoned due to their inability to address all three detrimental sequelae of atrial fibrillation. It was discovered that electrical impulses were incapable of crossing areas of the heart that had been incised and sutured.16,17

This fostered the development of the Maze procedure, designed and first reported by Cox and colleagues in which multiple incisions were made and sutured in a manner that blocked the aberrant impulses within the atrium (figure 2). These incisions directed the SA node impulse to initiate and propagate throughout the atria to the AV node along a specified route created by the Maze incisional lesions18. However, recently studies have demonstrated that the initiation of paroxysmal but not necessarily persistent or permanent AF are generated by electrical waves from focal sources, particularly near the pulmonary veins. Haissaguerre et al.19 mapped the triggers of paroxysmal AF to originate around the orifices of the pulmonary veins in 94% of patients with atrial diameters of less than 5cm. These findings have directed the development of more precisely targeted procedures that can be performed less invasively with less destruction of normal tissue.

Fig. 2. The pathway taken in the atria by the aberrant impulses in atrialfibrillation (arrows). The thick lines are the ablation lesions used in the Maze procedure. Key- SAN: sino-atrial node, AVN: atrio-ventricular node, RAA: right atrial appendage, LAA: left atrial appendage, SVC: superior vena cava, IVC: inferior vena cava, PVs: pulmonary veins.

Surgery for Atrial Fibrillation 261

restored.37 In addition patients in whom anticoagulation is contraindicated may also be suitable candidates that may benefit from surgical intervention following failure of catheter based ablation. Patients that continue to experience thromboembolic events despite

Several procedures were developed in the 1980s aimed at finding a cure to atrial fibrillation. Hoewever, most of these procedures were subsequently abandoned due to their inability to address all three of the detrimental sequelae of AF. Early attempts at surgical treatment of AF attempted to isolate and confine AF to a specific region of the atria and thereby stopping it from propagating its effects upon the ventricles. The left atrial isolation procedure developed by Williams and colleagues was successful in confining AF to the left atrium and thus restoring sinus rhythm to the rest of the heart.39 It also removed two of the 3 detrimental consequences attributed to AF namely, irregular heart rate and compromised haemodynamics. The latter was achieved because restoring sinus rhythm on the right side permitted a normal right-sided cardiac output that was delivered to the left side of the heart. The left ventricle responded to the normal cardiac output on the right side by delivering a normal cardiac output. Since the left atrium continued to fibrillate this procedure did not reduce the risk of thromboembolism. The 'Corridor' procedure was introduced in 1985 that isolated a strip of atrium that contained both the SA node and AV node from the rest of the atria to create a continuous pathway (corridor) directing the impulses from the SA node to the AV node to maintain sinus rhythm. Since parts of the right and left atrial were free to fibrillate it did not eliminate the risk of thromboembolism and nor did it restore

Cox and colleagues described a series of experiments that attempted to cure AF in dogs. A single incision across both atria successfully prevented AF and atrial flutter. Further investigations by Cox and colleagues led to the Cox-Maze procedure in 1987.16,17,40 The procedure itself was based upon a cut and sew technique whereby multiple incisions were made in the atria. This created lines of scar that interrupted the conduction routes of the most common re-entrant circuits, thus preventing AF or atrial flutter by directing the sinus node impulses along a specified route. It was based around the concept of a maze and as a result was called the Cox-Maze procedure. In contrast to the previous surgical techniques, this was the first that addressed all three sequelae of AF and restored sinus rhythm, AV synchrony and thus significantly reducing the risk of thromboembolism and stroke.41 The original procedure, known as Cox-Maze I was complicated with a high incidence of heart block requiring pacemaker implantation. It also resulted in the late incidence of two problems. Firstly it led to the frequent inability of patients to generate an appropriate sinus tachycardia and secondly left atrial dysfunction. This was modified to the Cox Maze II procedure which despite decreased incidence of conduction system injury was technically difficult. It was therefore modified to the Cox-Maze III procedure that was associated with a higher incidence of sinus rhythm and improved long-term sinus node function and atrial transport function.40 In this procedure several dead-end "alleyways "create a maze-like pathway and permit the depolarization of all the atrial tissue. The Cox Maze III procedure can be performed both through median sternotomy as well as a partial lower sternotomy. The patient is fully heparinized and the surgeon cannulates the patient for cardiopulmonary bypass after dividing the sternum. Bicaval cannulation is achieved. The right atrial appendage is excised and a series of incisions are made to the right atrium including a cryolesion. The aorta is occluded preparing for the left atrial portions of the operation. Cold blood potassium cardioplegia is administered via retrograde perfusion of the coronary

adequate anticoagulation may also benefit from surgery.38

atrioventricular synchrony.

A key concept in understanding the development of persistent atrial fibrillation is that the atria undergo electrical remodelling. Atrial electrical remodelling results in shortening of the atrial refractory period, myocyte calcium overload, decreased conduction velocity, dispersion of conduction and increased sensitivity to catecholamines4,20,21,22. This phenomenon may be reversed after maintenance of sinus rhythm23. The more a patient experiences AF, the more susceptible they are to continue fibrillating as a result of remodelling. Therefore AF can become a self-sustaining arrhythmia once atrial remodelling has occurred. Micro-reentrant triggers are then no longer necessary and limited to pulmonary venous impulse-triggering sites.24 Therefore, treatment for persistent or longstanding AF directed at the pulmonary veins only is likely to be unsuccessful.
