The direct surgical treatment of cardiac rhythm disturbances began in 1968, when the first successful division of the accessory pathway responsible for the Wolff–Parkinson–White syndrome was performed. Since then a number of surgical procedures have been developed to cure both supraventricular and ventricular tachyarrhythmias. These operative procedures have been demonstrated to be effective and in many instances they are preferred alternatives over a lifetime of drug therapy for many of these patients. This Section will discuss the pertinent anatomical, electrophysiological and surgical principles of the procedures that make up this relatively new area of cardiac surgery.

Candidates for surgical ablation of accessory pathways include patients with recurrent reciprocating tachycardia who are poorly controlled on medical therapy or have developed significant toxicity to an otherwise successful medical regimen; patients with Wolff–Parkinson–White syndrome who have symptomatic supraventricular arrhythmias and are undergoing cardiac surgery for other indications; and patients with atrial fibrillation or flutter who have an excessive ventricular rate, even if it can be attenuated with antiarrhythmic agents. The major indication for surgical intervention in these patients is refractoriness to medical treatment.

Patients with symptomatic arrhythmias due to other accessory atrioventricular connections (atrio–His, nodoventricular, and fasciculoventricular fibres) should undergo surgery if they are resistant or intolerant to medical therapy; however, these procedures should be performed by a surgeon with extensive experience of these more complicated types of accessory connections.

In the Wolff–Parkinson–White syndrome there is a congenital abnormal muscular connection between the atrium and ventricle, located somewhere in the atrioventricular groove of the heart. Anatomically, the heart can be sectioned at the level of the atrioventricular groove and divided into four discrete areas defined in this horizontal plane (Fig. 1) 1880. These four discrete areas are the left free wall, the right free wall, and the posterior septal and the anterior septal spaces.

The mitral and aortic valve annuli contribute significantly to the structural integrity of the fibrous skeleton and are further strengthened at their left junction to form the left fibrous trigone (Fig. 2) 1881. The atrioventricular groove between the left fibrous trigone and the right fibrous trigone (the anterior portion of the central fibrous body) represents the site of continuity between the anterior leaflet of the mitral valve and the aortic valve annulus. This is the only area in the atrioventricular groove where atrial muscle is not in juxtaposition to ventricular muscle, and for this reason accessory atrioventricular pathways are not found between the left and right fibrous trigones.

Both the preoperative catheter electrophysiological study and the intraoperative mapping procedure are directed towards localizing an accessory pathway to one of these four anatomical spaces. Because these pathways were initially thought to be macroscopically discrete and localized structures, an operative technique of limited dissection on either side of the localized area of earliest activation was developed. We have learned, however, that in order to surgically ablate a pathway in one of these regions with a success rate approaching 100 per cent, the entire anatomical space in which the pathway resides must be dissected in every patient.

Electrophysically, accessory pathways are the equivalent of an electrical cable that is capable of conducting electrical impulses between the atrium and ventricle. Histologically, these pathways resemble normal atrial myocardium. Because of the anatomical limitations imposed by the valve annulus on the inside of the atrioventricular groove and the epicardium on the outside of the atrioventricular groove in the vertical plane, all accessory pathways must connect atrium to ventricle somewhere between these two boundaries (Fig. 3) 1882. In the vertical plane, therefore, accessory pathways are confined to locations adjacent to or near the valve annulus, within the fat pad of the atrioventricular groove, or just beneath the epicardium overlying the atrioventricular groove. The ‘depth’ of the pathway within the atrioventricular groove in the vertical plane must be considered to be variable. In addition, when the horizontal and vertical planes are combined, it is important to understand that these accessory pathways can tangentially traverse this three-dimensional space.

Epicardial pacing and sensing electrodes are sutured on to the atrium and ventricle near the suspected site of the accessory pathway. An epicardial band electrode (Fig. 4) 1883 containing 16 bipolar button electrodes is placed on the ventricular side of the atrioventricular groove, and electrograms are recorded simultaneously from the 16 bipolar electrodes during normal sinus rhythm and during atrial pacing to assess anterograde ventricular activation during maximal pre-excitation. The analogue signals are filtered and digitized in the 160-channel front-end data acquisition system in the operating room, and then transmitted over a fibreoptic cable to the computer facility. A pre-excited QRS complex from the standard ECG is selected and the activation sequence of the 16 electrodes is determined by the computer. The local activation time (point of maximal deflection) for each of the 16 digitized electrograms is displayed by a vertical cursor on each electrogram. The electrogram recorded from the electrode located nearest the site of the ventricular insertion of the accessory pathway shows the earliest activation (Fig. 5) 1884.

The band electrode is then moved to the atrial side of the atrioventricular groove, and retrograde atrial activation is assessed during orthodromic reciprocating tachycardia induced with programmed electrical stimulation or with ventricular pacing. Only a few cycles of tachycardia are allowed to occur since haemodynamic compromise is common and the patients are not on cardiopulmonary bypass during mapping. Because only retrograde atrial data are of interest, the portion of the ECG containing the retrograde p wave is selected for analysis (Fig. 6) 1885. This display of atrial data is especially important because it demonstrates unsuspected concealed accessory pathways, undetected until this point in the mapping procedure (Fig. 7) 1886. In addition, rapid identification of multiple pathways and pathways manifesting intermittent conduction is greatly facilitated. In most circumstances, complete atrial and ventricular mapping and data analysis can be completed in approximately 10 min, and only a single beat is required to perform analysis of ventricular or atrial activation times.

The anatomical location of accessory atrioventricular connections may be classified into the four locations, as described earlier. In decreasing order of frequency, accessory pathways are located in the left free-wall, posterior septal, right free-wall, and anterior septal positions. Approximately 20 per cent of patients in our series have multiple (two to four) pathways.

Two surgical approaches are now commonly employed to divide accessory atrioventricular connections. The endocardial technique is designed to divide the ventricular end of the accessory pathway, and the epicardial technique is directed toward division of the atrial end of the pathway (Fig. 8) 1887. Thus the surgical objective is to divide the accessory connection, either at the atrial or the ventricular end of the pathway. Excellent results may be obtained with both techniques.

Since 1981 we have used an endocardial approach and an anatomically based operation for division of all accessory pathways (Figs. 9,10) 1888,1889,1890,1891. The principles of this operative approach are: (1) accurate intraoperative localization of the pathway(s) to one of the four anatomical areas in the horizontal plane; (2) appreciation that the location of the pathway in the vertical plane may be variable; (3) appreciation that the endocardial dissection technique divides the ventricular insertion of the pathway and does nothing to the atrial insertion of the pathway; (4) complete dissection of the appropriate anatomical space(s) in every patient regardless of the location of the pathway within that space as determined by intraoperative mapping; (5) appreciation that certain pathways may exist as ‘broad bands’ and that when the ventricular insertion site is located at the junction of two anatomical areas complete dissection of both anatomical spaces should be performed; (6) that isolation of the atrial rim of tissue above the annulus of the valve to prevent a juxta-annular pathway from retrogradely activating the atrium.

In 330 patients operated upon for the Wolff–Parkinson–White syndrome and/or other accessory pathways since 1981, surgical correction of the former syndrome using the techniques described was successful in nearly all patients at the the initial operation, with an operative mortality for elective, uncomplicated cases of 0.5 per cent. There have been no early or late recurrences following surgery using the endocardial technique described here.

In Guiraudon's series, similar results were obtained using the epicardial dissection technique for right free-wall, left free wall, and certain posterior septal pathways.

Re-entry within the atrioventricular node is the most common cause of paroxysmal supraventricular arrhythmia in adults. The anatomical and tachycardia electrophysiological reason for this arrhythmia is the presence of ‘dual atrioventrical nodal conduction pathways’, one fast and one slow, through the node or the perinodal tissues. When this arrhythmogenic characteristic has been demonstrated by appropriate electrophysiological study, and when it is refractory or unresponsive to medical management, surgery should be considered. In addition, when surgery is being performed to interrupt an accessory pathway in a patient with atrioventricular nodal re-entry or a dual atrioventricular node, surgical correction as described below should be undertaken.

The anatomy of the triangle of Koch is the important consideration in the surgical treatment of atrioventricular nodal re-entrant tachycardia. This triangle is bounded by the tendon of Todaro superiorly, the annulus of the tricuspid valve inferiorly, and the os of the coronary sinus posteriorly. The membranous septum marks the anterior tip of the triangle, and within this area lie the compact atrioventricular node, the perinodal tissues, and the bundle of His (Fig. 11) 1892. Cryosurgical modification of the perinodal tissue to interrupt the electrophysiological substrate for atrioventricular nodal re-entrant tachycardia is the goal of our surgical approach.

Surgical exposure for the discrete cryosurgical technique is the same as for a posterior septal pathway dissection. During application of the cryolesions the patient is atrially paced at a constant cycle length and atrioventricular conduction is monitored on a beat-to-beat basis. After setting the pacing and recording parameters, a nitrous oxide cryoprobe with a 3 mm tip is used to place cryolesions along the tendon of Todaro at a temperature of −60°C for 2 min, or until transient heart block occurs. These four cryolesions extend from the os of the coronary sinus to the apex of the Triangle (Fig. 12(a)) 1893.

Cryolesions are then placed along the annulus of the tricuspid valve beginning just beneath the os of the coronary sinus (Fig. 12(b)) 1893. Prolongation of the atrioventricular interval usually occurs first during applications of cryothermia at sites 7 and 8. It is important to apply cryothermia to each of these sites for the full 2 min if possible, since permanent tissue injury cannot be assured otherwise. Impending complete heart block is heralded by a prolongation of the atrioventricular interval by 200 to 300 m. When block occurs cryothermia is terminated instantly and the cryolesion is irrigated with copious amounts of warm saline. As the lesion thaws the atrioventricular interval shortens back to baseline, often on beat-to-beat basis. The freeze in that area is completed by moving the cryoprobe a few millimetres to either side until the cryolesion can be applied for the full 2 min at that site without causing complete heart block. The first nine cryolesions are designed to outline the border of the Triangle and to encircle the atrioventricular nodal tissue; subsequently, as many cryolesions as possible are placed within the Triangle of Koch without causing block (Fig. 13) 1894. Used in this way, the cryoprobe acts as a ‘reversible knife’. In essence, the objective of this operation is to cryoablate as much of the perinodal tissue as possible without causing permanent atrioventricular conduction block.

Two other surgical techniques have been developed for the treatment of this arrhythmia. Ross and Johnson first used their surgical dissection technique for atrioventricular node re-entry in 1983, with good results (Fig. 14) 1895. This approach has recently been modified slightly by Guiraudon. These techniques, while effective, have been complicated by a small but finite incidence of permanent heart block, and a rather high late recurrent rate.

The discrete cryosurgical procedure has been performed on 35 patients at our institution to date. In all cases postoperative electrophysiological study has demonstrated the persistence of only a single atrioventricular conduction pathway, and the re-entrant tachycardia could not be induced. There have been no instances of permanent heart block, and no late recurrences.

As mentioned, Johnson and Guiraudon have described surgical dissection techniques for this arrhythmia, with good initial results. However, a finite incidence of permanent heart block and a relatively high late recurrence rate has been associated with these procedures.

Automatic (ectopic) atrial tachycardias are often incessant. They usually originate from the body of the right or left atrium, but occasionally arise from the interatrial septum. The ventricular rate depends upon conduction through the atrioventricular note during the atrial tachycardia. Reversible abnormalities that may precipitate atrial tachycardias, such as hyperthyroidism, electrolyte imbalance, and digitalis toxicity, should be excluded before considering surgical treatment.

Preoperative electrophysiological evaluation is necessary to discern the mechanism of the arrhythmia, establish the region of origin of the tachycardia, and exclude concomitant electrophysiological derangements that may contribute to the rapid ventricular rate. Accurate preoperative localization is particularly important in patients with automatic atrial tachycardias if surgical ablation of the focus is contemplated: general anaesthesia frequently suppresses the ectopic focus; intraoperative mapping without sophisticated computerized multipoint systems can be prohibitively difficult and time-consuming; and ectopic tachyarrhythmias are not inducible by standard programmed stimulation techniques.

Anatomically, ectopic or automatic atrial tachycardias can originate from foci occurring anywhere within the left or right atrial tissue or atrial septum. In addition, there may be be multiple foci. Lowe has collected data on the 125 patients reported in the literature to date; of the 89 in whom the the location was specified, 61 (68 per cent) originated in right atrial tissue, five (6 per cent) in the atrial septum, and 23 (26 per cent) in left atrial tissue (Fig. 15) 1896.

If the tachycardia focus can be localized, a variety of techniques has been advocated for surgical treatment. Cryoblation of the ectopic focus, with or without cardiopulmonary bypass, has been used. Alternatively, wide excision of foci located on the right atrial free wall with a pericardial patch repair can be used (Fig. 16) 1897. Foci located on the atrial appendage may be treated by simple excision and oversewing of the line of resection. Others have used a combination of cryoblation and resection.

Foci on the left atrium have tended to be near the left superior pulmonary vein; localized isolation procedures have been described for these tachyarrhythmias, but have had limited success. In addition, a significant number of these ectopic foci are localized to the left atrium on the preoperative study, but cannot be localized intraoperatively, and a localized isolation procedure cannot be performed. In both instances, these ectopic foci should be excluded from the remainder of the heart using the left atrial isolation procedure described by Williams et al. (Fig. 17) 1898. Following the left atrial isolation procedure patients remain in normal sinus rhythm despite the presence of an incessant tachycardia confined to the left atrium. No adverse sequelae have been noted over a 7-year follow-up period with this procedure. This therapy is preferable to the other therapeutic alternative—elective His bundle ablation and pacemaker insertion.

Right atrial tachycardias are usually confined to the body of the right atrium, and may be multifocal. If the arrhythmia circuit or focus cannot be localized at surgery, then a right atrial isolation procedure that isolates the body of the right atrium while leaving the atrial pacemaker complex in continuity with the atrial septum and ventricles should be performed (Fig. 18) 1899.

If the focus of origin for ectopic atrial tachycardias can be adequately localized in the operation room, operative procedures for isolation and/or ablation of the arrhythmia should be uniformly successful. The left atrial isolation procedure has been performed on six patients, and the right atrial isolation procedure has been performed on three patients; there have been no adverse sequelae from these operations to date.

The surgical treatment of atrial flutter and fibrilation is still considered experimental, but a brief discussion is included here because of the potential impact it will have on cardiac arrhythmia surgery in the future.

Experimental studies have documented that atrial flutter is probably due to macro re-entrant circuits in the atrial tissue. Using sophisticated multi-point computerized mapping, intraoperative localization of the re-entrant circuit responsible for the flutter is possible, with subsequent surgical intervention using dissection, cryoblation, or a combination of the two to interrupt the re-entry (Fig. 19) 1900.

Chronic or intermittent atrial fibrillation results in three untoward sequelae for patients who suffer from this arrhythmia: loss of atrial transport function with loss of atrial contraction; unpleasant subjective symptomatology due to the irregular heart beat; and increased risk of thromboembolism. Experimental studies have confirmed that atrial fibrillation is due to intra-atrial re-entry in which multiple re-entrant wavelets are present. As atrial size increases, the number of wavefronts during atrial fibrillation and the duration of atrial fibrillation increase. Intraoperative mapping studies in patients have indicated that the only way to ablate atrial fibrillation surgically is to create a specific pathway of depolarization from the sinoatrial node to the atrioventricular node; this pathway would have to activate all functional atrial myocardium in order for the atrium to maintain its normal transport function and to eliminate the problem of thromboembolism. A procedure that meets these requirements has been developed, and is termed the maze procedure (Fig. 20) 1901. Following surgery, a normally generated impulse propagates from the sinoatrial node to activate the entire atrial myocardium, apart from the excised atrial appendages and the pulmonary veins. At the same time, it is impossible for a large macro re-entrant circuit to exist since the atrial impulses are precluded from turning back on themselves. Regardless of the number of macro re-entrant circuits responsible for the development and perpetuation of atrial fibrillation, they cannot occur following the maze procedure.

With adequate intraoperative mapping, the macro re-entrant circuits responsible for atrial flutter can be interrupted with uniform success. The maze procedure has been performed on 12 patients, all of whom are in normal sinus rhythm and require no postoperative antiarrhythmic medications.

Because of the electrophysiological complexity of ventricular arrhythmias and difficulties with preoperative and intraoperative mapping, the development of surgical procedures for the relief of ventricular tachycardia has proceeded more slowly than for supraventricular arrhythmias. The efficacy of surgical treatments for these arrhythmias has recently been demonstrated, however, by analysis of surgical results with 10-year follow-up.

Patients with both ischaemic and non-ischaemic ventricular tachycardias who are surgical candidates undergo complete electrophysiological, angiographic, and ventriculographic evaluation. Catheter electrophysiological study is performed to confirm that the arrhythmia is ventricular and not supraventricular in origin, to demonstrate that the arrhythmia is re-entrant by induction and termination with programmed electrical stimulation techniques, and to identify the earliest site of origin of all morphologically distinct tachycardias using ‘catheter mapping’ techniques.

Angiographic and ventriculographic evaluation is particularly helpful in the evaluation of three of the non-ischaemic forms of ventricular tachycardia. In patients with diffuse cardiomyopathy, due to patchy myocardial fibrosis, angiographic and haemodynamic data usually indicate some type of abnormal myocardial contractility associated with recurrent tachycardia. Ventriculography demonstrates diffuse digitation of both ventricles. This same finding on ventriculography can also be present in patients with idiopathic ventricular tachycardia, due to repeated bouts of tachycardia; however, in this latter entity, pathological evidence of primary cardiac disease is absent. In arrhythmogenic right ventricular dysplasia, due to transmural infiltration of adipose tissue resulting in weakness and aneurysmal bulging of three areas of the right ventricle, ventriculography demonstrates diffuse digitation, depressed contractility, and delayed emptying of the right ventricle. Frank aneurysms of the infundibulum, apex, and/or posterior basilar region are seen, and hypertrophic muscular bands in the infundibulum and anterior right ventricular wall result in a feathering appearance of the outflow tract. Since the origin of the tachycardia is the right ventricle, the 12-lead ECG shows a pattern consistent with left bundle branch block during the tachycardia. Right ventriculography should therefore be performed in any patient with ventricular tachycardia and this QRS complex configuration.

The final decision regarding surgical therapy for ventricular tachycardias of both ischaemic and non-ischaemic origin is based upon a variety of preoperative factors. The primary indication for surgery is refractoriness to medical therapy.

These tachycardias usually arise in the right ventricular free wall or septum and are in general extremely resistant to medical therapy. Localized surgical isolation techniques are usually employed for tachycardias arising in the right ventricular free wall, while multipoint map-guided cryoblative techniques are used for arrhythmias localized to the septum.

The surgical procedure developed for arrhythmogenic right ventricular dysplasia is designed to isolate the arrhythmogenic myocardium from the remainder of the heart. For well-localized ventricular tachycardias, focal isolation or ablation procedures are recommended. When the area of arrhythmogenic myocardium is larger, a transmural encircling ventriculotomy is created with the surgically isolated pedicle is based on a vascular supply originating from the right coronary artery (Fig. 21) 1902. Two cryolesions are placed at the proximal and distal aspect of the incision at the level of the tricuspid annulus to ensure complete separation of all ventricular muscle fibres on either side of the incision. Successful isolation has been demonstrated in a small number of patients. In certain instances, intraoperative mapping has suggested that the entire right ventricular free wall may be arrhythmogenic, giving rise to multiple morphological types of tachycardia. Surgical isolation of the entire right ventricular free wall was previously performed in such cases. Postoperatively, however, the right ventricle has on occasion undergone a progressive dilatation, and cardiac transplantation in a suitable patient with this arrhythmia is currently the most feasible surgical approach.

Patients with ventricular tacycardia occurring in association with the long QT interval syndrome frequently have ‘torsades de pointes’ as the manifestation of their tachycardia. Medical therapy consists of &bgr;-adrenergic blockade; this has recently been coupled with permanent atrial or ventricular pacing. Surgical therapy has consisted of left cervicothoracic sympathectomy with removal of the left stellate ganglion and the first three to four left thoracic sympathetic ganglia. In our experience, this procedure is associated with early success but late failure, and as a result implantation of an automatic implantable cardioverter-defibrillator at the time of sympathectomy is now recommended in patients with a history of life-threatening arrhythmias.

Our current algorithm for the optimal surgical treatment of refractory ischaemic ventricular tachycardia is shown in Fig. 22 1903. With this algorithm as a guide, the following points can be emphasized.

1.This evaluation regarding surgical intervention for medically refractory ischaemic ventricular tachycardia should be made prior to the institution of amiodarone therapy. The depressant effect of amiodarone on left ventricular function is aggravated by ischaemic cardioplegic arrest in the majority of patients, and therapy with the drug prior to surgical intervention can significantly complicate the operative procedure.

2.The evaluation for surgical intervention should be based primarily on the determination that the patient has a sufficient degree of normal left ventricular function to survive operative intervention. Because most patients with ischaemic ventricular tachycardia have a left ventricular aneurysm, accurate determination of the ejection fraction is difficult, and the absolute number is not an accurate predictor of operative mortality. Patients with discrete apical or posterior aneurysms and myocardium in the remainder of the heart that contracts in a normal or near-normal fashion are the best candidates from a functional point of view.

3.If the patient has global ventricular dysfunction and is not a direct surgical candidate, an automatic implantable cardioverter-defibrillator should be implanted, according to the criteria described below. If the patient is not a candidate for such therapy, amiodarone should be given.

4.If the patient's tachycardia is uncontrolled with an automatic implantable cardioverter-defibrillator or amiodarone, cardiac transplantation should be considered. If the patient is not a transplant candidate, then ventricular tachycardia surgery is the only therapeutic option available.

5.If the patient's left ventricular function is acceptable for surgery, the surgical approach outlined below is performed.

All 160 channels of the computerized system are used to map the heart in patients with ventricular tachycardia. This system can record 160 simultaneous signals, analyse the data, and display it in various forms with 2 min after acquisition of data for editing. An epicardial map employing a 96-electrode sock array (Fig. 23) 1904 is recorded, and this information is used to guide the subsequent placement of plunge needle electrodes to delineate further the specific site of arrhythmogenesis.

Subsequently, multiple plunge needle electrodes containing four bipolar pairs of electrodes are inserted into the ventricle in the region of earliest epicardial activation (Fig. 24) 1905. If the tachycardia appears to arise from the intraventricular septum, a right atriotomy is performed and up to 15 plunge needle electrodes are inserted into the septum from the right side. A total of up to 160 endocardial, intramural, and epicardial data points can be recorded simultaneously from the septum and free wall. A ventriculotomy, which can prevent further inducibility of the ventricular tachycardia and necessitate performance of a procedure that is not map-guided, does not have to be performed in order to obtain an endocardial map.

After completion of intraoperative mapping, with the patient on cardiopulmonary bypass, the ventricle is opened through the infarct or aneurysm. This is performed with the heart in the normothermic beating state, and preferably during ventricular tachycardia. All of the associated endocardial fibrosis is resected except that which extends on to the base of the papillary muscles (Fig. 25) 1906. About 10 per cent of patients will still have inducible tachycardia following resection of the endocardial fibrosis, indicating that the actual site of origin of the tachycardia in these patients is deeper in the myocardium than the visible gross border of the fibrosis. Endocardial cryolesions are applied with a 2.5-cm nitrous oxide cryoprobe to the site(s) of origin of the tachycardia(s) as determined by the intraoperative mapping, thus ablating the myocardium beneath the visible fibrosis that is responsible for the tachycardia. Arrhythmogenic tissue at the base of the papillary muscles is cryoblated not resected. These techniques are applicable to both anterior and posteroinferior infarct/aneurysms.

Following completion of the resection and cryoablation procedure, programmed electrical stimulation is applied in an attempt to reinduce the arrhythmia. If ventricular tachycardia is still inducible, mapping is again performed and the remaining arrhythmogenic myocardium is again cryoblated. If the arrhythmia is no longer inducible there is a 98 per cent change that is has been permanently ablated. Coronary artery bypass grafting or other procedures that may be required are performed after completion of the antiarrhythmic portion of the operation. Cardioplegic solution must not be administered until the antiarrhythmic portion of the operative procedure has been successfully completed: cardioplegia may itself temporarily alter the delicate re-entry circuits causing the tachycardia. If the antitachycardia procedure is performed under cardioplegic arrest, it is impossible to determine intraoperatively whether or not the surgical procedure has ablated the arrhythmias.

The long-term success rate for patients who survive ventricular tachycardia surgery is 87 per cent at 9 years in our series of patients operated upon for ischaemic ventricular tachycardia. Long-term survival in this series is 72 per cent at 9 years following successful operative intervention.

The current role of the automatic implantable cardioverter-defibrillator in the surgical treatment of ventricular tachycardias is as follows:

1.If an institution does not have the capability of performing computerized intraoperative mapping or cryosurgery, if a procedure that is not dependent upon intraoperative mapping is performed, or if the surgical procedure is performed under cardioplegic arrest, it is appropriate to implant automatic implantable cardioverter-defibrillator patches at the time of the ventricular tachycardia surgery. Under these circumstances, approximately 25 per cent of patients will have inducible tachycardia following the surgical procedure.

2.If the tachycardia is inducible at the time of the postoperative electrophysiological study, the automatic implantable cardioverter-defibrillator should be implanted.

3.The only absolute contraindications to implantation include the inability of the patient to tolerate the operative procedure and an episodic rate of the patient's tachycardia such that the battery supply of the device becomes exhausted or makes the patient's life prohibitively uncomfortable due to an excessive number of device discharges.

4.The operative and long-term follow-up results for ventricular tachycardia surgery argue strongly against routine implantation of the device and performance of coronary bypass surgery as the primary treatment for ventricular tachycardia in patients who are otherwise candidates for a map-directed, wide-resection/cryoablative procedure as described above.

5.It should be recognized that the device provides a viable therapeutic option for those patients in whom performance of ventricular tachycardia surgery has been fraught with an excessive mortality. Judicious selection of these patients for implantation should reduce the operative mortality rate in this subset of ventricular tachycardia patients, thereby also reducing the overall operative mortality rate for ventricular tachycardia procedures.

The list of supraventricular arrhythmias amenable to surgical correction has grown over the past decade, due to advances in our understanding of the anatomical and electrophysiological principles pertaining to these arrhythmias. Treatment for the surgical cure of patients with the Wolff–Parkinson–White syndrome and other accessory atrioventricular connections is well-established, and should be considered in any patient with a symptomatic form of this congenital lesion. Similarly, surgical intervention in patients with atrioventricular nodal re-entry should be considered as a therapeutic alternative in patients with a sympomatic form of the arrhythmia. Over the next decade, the newer surgical techniques described here for the treatment of atrial tachycardias, atrial flutter, and atrial fibrillation should become a regular part of the surgical armamentarium available for cure of these common arrhythmia problems.

These promising results of surgical treatment of ventricular tachyarrhythmias provide an extremely strong impetus to optimize the preoperative evaluation, operative selection, and intraoperative management of this critically ill but potentially curable group of patients. In the future, the automatic implanted cardioverter-defibrillator will play a critical role in reducing the overall operative mortality for patients undergoing surgery for ventricular tachycardia.

Allessie MA, et al. Experimental evaluation of Moe's multiple wavelet hypothesis of atrial fibrillation. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology and Arrhythmias. Orlando: Grune and Stratton, 1985: 265–75.

Boineau JP, et al. Mechanisms of atrial flutter as determined from spontaneous and experimental models. In: Josephson ME, Wellens HJJ, eds. Tachycardias: Mechanisms, Diagnosis and Treatment. Philadelphia: Lea and Febiger, 1984: 91–111.

Boineau JP, Schuessler RB, Cain ME, Corr PB, Cox JL. Activation mapping during normal atrial rhythms and atrial flutter. In: Zipes DP, Jalife J, eds. Cardiac Electrophysiology. Philadelphia: W.B. Saunders and Co., 1990: 537–48.

Cain ME, Lindsay BD. The preoperative electrophysiologic study. In: Cox JL, ed. Cardiac Surgery: State of the Art Reviews. Philadelphia: Hanley and Belfus, Inc., 1990: 1–39.

Canavan TE, Schuessler RB, Boineau JP, Corr PB, Cain ME, Cox JL. Computerized global electrophysiological mapping of the atrium in patients with the Wolff-Parkinson-White Syndrome. Ann Thoracic Surg 1989; 46: 223–31.

Cox JL. Intraoperative computerized mapping techniques. In: Brugada P, Wellens HJJ, eds. Cardiac Arrhythmias: Where To Go From Here? Mt. Kisco, NY: Futura Publishing Co., 1987: 631–7.

Cox JL. Patient selection criteria and results of surgery for refractory ischemic ventricular tachycardia. Circulation 1989; 79 (Suppl I): I163-7.

Cox JL, Ferguson TB Jr. Cardiac arrhythmia surgery. In: Wells SA, ed. Current Problems in Surgery. 26th edn. Chicago: Year Book Medical Publishers, Inc., 1989: 193–278.

Cox JL, Gallagher JJ, Cain ME. Experience with 118 consecutive patients undergoing surgery for the Wolff-Parkinson-White syndrome. J Thoracic Cardiovasc Surg 1985; 90: 490–501.

Cox JL, et al. Surgery for atrial fibrillation. Seminic Thoracic Cardiovasc Surg 1989; 1: 67–73.

Cox JL, Ferguson TB Jr, Lindsay BD, Cain ME. Peri-nodal cryosurgery for AV nodal reentrant tachycardia in 23 patients. J Thoracic Cardiovasc Surg 1990; 99: 440–50.

Ferguson TB Jr. The role of the automatic implantable cardioverter-defibrillator (AICD) in the treatment of medically-refractory ventricular arrhythmias. Trends Cardiovasc Med 1991; 1: 131–9.

Ferguson TB Jr. Anatomic and electrophysiologic principles in the surgical treatment of cardiac arrhythmias. In: Cox JL, ed. Cardiac Arrhythmia Surgery. Philadelphia: Hanley and Belfus, 1990: 41–73.

Ferguson TB Jr, Cox JL. Surgical therapy for patients with supraventricular tachycardia. In: Scheinman MM, ed. Cardiology Clinics. 8–3. Philadelphia: W.B. Saunders & Co., 1990: 535–56.

Ferguson TB Jr, Cox JL. Antiarrhythmic surgery: ventricular arrhythmias. In: Horowitz LN, ed. Current Management of Arrhythmias. Philadelphia: B.C. Decker, 1990: 382–91.

Ferguson TB Jr, Cox JL. Complications related to the surgical treatment of supraventricular and ventricular cardiac arrhythmias. In: Waldhausen JA, Orringer MB, eds. Complications in Cardiothoracic Surgery. 1st edn. St. Louis: Mosby Year Book, 1990: 303–15.

Guiraudon GM, et al. Surgery for the Wolff-Parkinson-White syndrome: the epicardial approach. Semin Thoracic Cardiovasc Surg 1989, 1: 21–33.

Guiraudon GM, Klein GJ, Sharma AD, Yee R, Kaushik RJ, Fujimura O. Skeletonization of the atrioventricular node for AV node reentrant tachycardia: experience in 32 patients. Ann Thoracic Surg 1990; 49: 565–73.

Harada A, D'Agostino HJ, Schuessler RB, Boineau JP, Cox JL. Right atrial isolation: a new surgical treatment for supraventricular tachycardia. J Thoracic Cardiovasc Surg 1988; 95: 643–50.

Johnson DC, Nunn GR, Meldrum-Hanna W. Surgery for atrioventricular node reentry tachycardia: the surgical dissection technique. Semin Thoracic Cardiovasc Surg 1989; 1: 53–7.

Kramer JB, et al. Arrhythmia and conduction disturbances: simultaneous computer mapping to facilitate intraoperative localization of accessory pathways in patients with Wolff-Parkinson-White syndrome. Am J Cardiol 1985; 56: 571–6.

Lawrie GM, Lin H, Wyndham CRC, DeBakey ME. Surgical treatment of supraventricular arrhythmias-results in 67 patients. Ann Surg 1987; 205: 700–11.

Lowe JE, Hendry PJ, Packer DL, Tang AS. Surgical management of chronic ectopic atrial tachycardia. Semin Thoracic Cardiovasc Surg 1989; 1: 58–66.

Williams JM, et al. Left atrial isolation. New technique for the treatment of supraventricular arrhythmias. J Thoracic Cardiovasc Surg 1980; 80: 373–80.