Indirect Methods

The first reported surgical cure of VT resulted from excision of a left ventricular aneurysm. When aneurysmectomy was used more widely as a treatment for VT, however, it became apparent that this procedure had low efficacy and high mortality rates (often because of intractable VT). CABG was also used in an attempt to control VT, with similarly poor results. These methods generally failed because although aneurysms are frequently present in patients with VT, the excised portion of the aneurysm does not contain critical portions of VT circuits, and VT episodes usually do not result from acute ischemia (which could be corrected by coronary bypass). CABG has definite application in relatively rare cases of polymorphic VT or VF that clearly occurs in the setting of acute ischemia. CABG and aneurysmectomy are both still performed in some cases of VT but only as adjunctive methods along with one of the more direct techniques discussed below. Neural modification, typically with left or bilateral stellate ganglionectomy, has benefit in selected patients with VTAs: patients with long QT syndrome (LQTS) who have recurrent syncope or cardiac arrest despite β-blocker and ICD therapies; patients with recurrent episodes of VF in the absence of SHD (idiopathic VF); and patients with catecholaminergic polymorphic VT. A variety of approaches have been used for the destruction or removal of the left or bilateral stellate ganglia (cervical, thoracotomy incisions). Thoracoscopic techniques are being more widely used for approaching the stellate ganglion, rather than open surgical exposure.¹

Direct Methods

After the failure of CABG and aneurysmectomy to reliably affect VT recurrence, several investigators began using intraoperative electrophysiological mapping studies to understand more about where VT was originating. The rationale was that a site in the ventricle from which VT arose, or exited from a circuit, should precede the onset of each surface QRS complex by greater than 40 ms, or even occur in mid-diastole. Epicardial mapping during VT usually did not reveal sites of activation that met this criterion; endocardial mapping, after ventriculotomy, routinely showed sites activated prior to the QRS during VT (Figure 97-1). These areas were typically near the border between the aneurysm or the dense infarct and more normal muscle. Several means of therapy targeting these regions were devised in the late 1970s, including excision (endocardial resection) an isolating incision (encircling ventriculotomy), and cryoablation. Additional modalities have been used, including intraoperative laser photoablation. Because these hearts have already suffered extensive damage from the infarction that caused the arrhythmia, a cardinal principle of surgical therapy is to remove or destroy only as much tissue as is necessary to eliminate the VT while damaging as little remaining normal myocardium as possible. Mapping tools and techniques have aided in the discrimination between normally and abnormally functioning tissue.

FIGURE 97-1 Surgical mapping during post-infarct ventricular tachycardia (VT). Surface electrocardiogram leads and endocardial electrograms recorded during VT. The dashed line shows the onset of the QRS complex in VT; the arrow indicates a recording in mid-diastole in a critical area of a VT circuit. Excision of the surrounding region eliminated VT. RV, Right ventricle; LV, left ventricle.

Preoperative Studies

Because most patients undergoing surgery for VT have prior infarction from coronary atherosclerosis, cardiac catheterization with coronary arteriography is an important part of the preoperative evaluation. Coronary stenoses that should undergo bypass grafting can thus be identified and any possible valvular lesions identified. Ventriculography (assessing overall left ventricular function) can be helpful in estimating operative risk. In the presence of coronary stenoses of uncertain flow-limiting significance, stress testing with some form of perfusion imaging can help determine whether bypass grafting of particular arteries would be beneficial. Preoperative electrophysiological study (EPS) is almost always indicated (except in cases of severe left main stenosis or three-vessel disease, in which the rapid heart rates encountered during induced VT could possibly be detrimental). This study can provide information as to how readily inducible VT is as a guide to how vigorous intraoperative stimulation will need to be. Endocardial catheter mapping can also be performed at this time to help focus attention on certain areas that need to be ablated during surgery. However, if mapping is performed in this setting, it is almost always accompanied by ablation attempts; if the ablation successfully eliminates inducibility of one or more morphologies of VT, they no longer need surgical attention. If ablation is not successful, the validity of the mapping information may be called into question. Thus preoperative mapping may have a more limited role than was once thought. Percutaneous access to the pericardial space has been used for epicardial mapping and ablation; this is more frequently necessary in cardiomyopathic cases than in post-MI VT. Recently, a hybrid procedure has been developed in which epicardial catheter mapping has been facilitated by surgical exposure of the pericardial space using a small subxiphoid incision. This is particularly helpful when prior cardiac surgery has resulted in extensive pericardial adhesions that would significantly restrict catheter mobility and access to all areas of the epicardial surface.²

Principles of Cardiac Mapping

If mapping is to be performed, several pieces of equipment are necessary beyond the usual requirements for open-heart surgery. These include electrodes, a mapping system (consisting of amplifiers and a recording and analysis system), and a stimulator with which to initiate VT. Several commercial mapping systems are available, capable of recording from 64 to 256 electrodes simultaneously, online analysis of activation times, and generation of isochronal maps. Most record onto magnetic or optical media for data archiving. At the time of surgery, the electrocardiogram (ECG) leads from the patient are connected to the mapping system. The heart is exposed through a median sternotomy, and a pericardial cradle is formed. Cannulae are placed for cardiopulmonary bypass, and reference electrodes are inserted and tested. The heart is inspected for regions of infarct or aneurysm, through which the ventricle may be entered without damaging the viable myocardium; sometimes, this area is only evident after the patient is on cardiopulmonary bypass and the left ventricle (LV) has been vented. Cardiopulmonary bypass is then established while maintaining a perfusate temperature of 37° C to 38° C. This is necessary because cardiac electrical activity deteriorates at cooler temperatures and arrhythmias may not be inducible or mappable. Radiant heat loss of the heart surface during mapping may necessitate even slightly higher perfusate temperatures. Epicardial mapping during sinus rhythm or VT can be performed with the heart closed, but this is usually omitted in the interests of time unless specific circumstances suggest its usefulness. In most cases, once the heart-lung machine is running well, the LV is then opened through the previously identified infarct or aneurysm. The endocardial surface is inspected for the presence of any adherent thrombus, which is then removed. An electrode or multipolar electrode array is placed on the endocardial surface in an area of obvious scar tissue. At this time, sinus rhythm mapping can be performed to designate areas with abnormal electrograms indicating damaged myocardium; this step usually adds little and is omitted except in special situations. In most cases, VT is then initiated using previously placed electrodes, and endocardial mapping is performed during VT. Some tissue is activated at every instant during a re-entrant arrhythmia, even during the diastolic interval between discrete QRS complexes; sites from which diastolic activation is recorded during VT are of particular interest, since these regions are often in “protected” corridors that are critical for continued re-entry. Computerized mapping systems are able to quickly display these areas of diastolic activation (Figure 97-2). Because most patients with post-infarct VT have more than one ECG morphology of VT that may arise from the same or different regions, attempts are made to initiate and map other morphologies of VT. Because of time limitations, rigorous entrainment studies (overdrive pacing during VT) to validate the activation mapping data are often not performed. Once all inducible morphologies of VT have been mapped, the electrode array is removed and the ablation portion of the procedure begins. Several different procedures have been used for this purpose (Figure 97-3), including:

FIGURE 97-2 Isochronal maps during ventricular tachycardia (VT). The endocardial surface is shown in a polar projection with the apex at the center, septum at on the left, lateral wall on the right. Lines of isochronal activation during VT are shown at 10-ms intervals; thick dark lines denote arcs of block; shaded curved arrows denote the direction of wavefront propagation. Four electrocardiogram leads are shown above with a timeline. A nearly complete figure-of-8 re-entrant pattern is shown; mid-diastolic recordings are made from the narrow zone between the arcs of block. No electrical activity could be recorded from the shaded areas.

FIGURE 97-3 Surgical procedures for ventricular tachycardia (VT). In each panel, the opened left ventricle is depicted as viewing the septal endocardial surface showing papillary muscles and a thin, smooth aneurysm as opposed to the trabeculated normal endocardium. A VT circuit is shown, some of which is near the endocardial surface (solid line) and some in deeper layers (dotted line). Lower left, Encircling endocardial ventriculotomy incises near the perimeter of the endocardial scar and transects a portion of the circuit. Upper right, Subendocardial resection removes a superficial portion of the circuit, and cryoablation destroys a portion by freezing. Of note, to prevent VT, none of these procedures needs to destroy or remove the entire circuit, only a critical portion of it.

FIGURE 97-4 Effect of endocardial resection on electrograms. Electrodes 1 through 4 are closely spaced recordings from a multipolar array. All three panels show a single complex in sinus rhythm; dashed line is the end of QRS. A, Recordings before resection in an area from which diastolic activation occurred in VT show isolated potentials, some of which occur after the QRS end. B, Recordings made after resection show loss of delayed potentials and increase in amplitude of signals in the QRS. C, Recordings from same site after resection and replacement of specimen show decrease in signal amplitude similar to A, but no late potentials. Thus resection removes tissue from which the late potentials derive as well as attenuates amplitude of recordings from subjacent tissue.

After the ablation portion, attempts are usually made to re-initiate VT with the same type of stimulation used during mapping. If VT can still be initiated, mapping is repeated; this usually reveals sites just outside the area ablated or deep to it (intramural). Ablation is performed again (often cryoablation of deeper sites), and cycles of stimulation, mapping and ablation are repeated until VT can no longer be induced. This technique results in postoperative noninducibility of VT approaching 90%. In inferior infarction, cryoablation of the isthmus of the ventricular myocardium between the ventriculotomy and the mitral annulus has increased surgical success rates.⁸ The ventriculotomy is closed and other procedures performed, such as CABG or valve surgery. If no other procedures are needed, the patient is weaned from cardiopulmonary bypass. Recent innovations in how the ventriculotomy is closed (aneurysmorrhaphy) have led to improvements in postoperative left ventricular function. Patients typically undergo EPS 5 to 7 days after surgery to assess its antiarrhythmic efficacy.

In some cases, no obvious site for making a ventriculotomy to enter the ventricle (no infarct or aneurysm) exists. This is especially true in inferior infarctions and cases in which early reperfusion of the infarct (pharmacologic or mechanical) yielded only a partial success. In these cases, an electrode-studded inflatable latex balloon can be inserted through the mitral orifice after a purse-string left atriotomy. Once the balloon is in the ventricular cavity, it is inflated such that its electrodes contact the endocardial surface. Mapping can then be carried out as indicated above but without ventriculotomy, and ablation can be performed either with electrical energy (DC shock or radiofrequency energy) delivered through appropriately located electrodes, or using a cryoprobe advanced through the mitral orifice as the array had been.⁹

Because of the presence of diseased myocardium, many areas have abnormal electrograms during sinus rhythm or VT but do not actually participate in the re-entrant wavefront. This low specificity, as well as the fact that up to 15% of VTs have been mapped to regions with relatively normal electrograms during sinus rhythm, limits the use of simple ablation of areas with abnormal electrograms. Furthermore, some sites have diastolic activation during VT but represent side branches that are not in the re-entrant path. Some of these display 2 : 1 conduction of potentials during VT; obviously, these sites are not in the re-entrant circuit because VT continues despite their absence on alternate beats, but they are typically recorded from regions near the actual circuit.

Operative Risk

Because of pre-existing extensive myocardial damage, the operative mortality rates of VT surgery are higher than in other forms of elective cardiac surgery. Mortality rates range from 3% to 20% in different patient subsets. Because of this, several investigators have evaluated the risk factors for VT surgery and have identified left ventricular wall motion score, age greater than 65 years, and emergency surgery as factors that increase risk of operative mortality. A major cause of perioperative mortality is heart failure; in the early surgical experience, the aneurysm was removed (or plicated) and its cut edges sutured together. This could result in an inordinately small left ventricular cavity and compromise cardiac output. In the 1990s, a more physiological closure method, introduced by Dor and colleagues, more closely approximated the normal ventricular shape.¹¹ Lower operative mortality rates and better functional capacity have been reported with this method (Figure 97-5).

FIGURE 97-5 Ventricular remodeling after aneurysm surgery. Left, A normal left ventricular contour (right anterior oblique view). Infarction results in aneurysm formation (center). Earlier work with aneurysm resection (or plication) made a simple closure that compromised cavity size and thus stroke volume. More recent developments include repair of the aneurysm, involving partial resection and use of prosthetic patch material to restore more normal ventricular geometry. LV, Left ventricle.

Patient Selection

Because of the above features and the availability of lower-risk alternative therapies, surgery is rarely performed for VT. The benefits and risks of surgery, as well as the availability of the necessary mapping equipment and electrophysiological and surgical expertise, must be carefully weighed before proceeding.

Other Substrates of Ventricular Tachycardia

Re-entrant VT can occur in other settings such as idiopathic cardiomyopathy, arrhythmogenic right ventricular dysplasia (ARVD), and Chagas disease or after right ventriculotomy for repair of congenital heart disease (ventriculoseptal defect, tetralogy of Fallot). There is relatively little surgical experience with cardiomyopathic VT. The arrhythmogenic substrate appears to be located on the endocardial aspect of the LV in most cases, as in post-infarct VT. Unlike the latter, however, there is usually no readily available region of scar or aneurysm through which the ventricle can be opened to allow endocardial access. It is generally undesirable to make a large incision through normal-appearing myocardium, the closure of which could itself conceivably become a source of re-entrant arrhythmias postoperatively. Some cases of VT caused by ARVD were managed surgically in the 1970s and 1980s; the substrate in these cases was relatively thin strands of surviving normal myocardium separated by large zones of adipose and fibrous tissues that had replaced portions of the right ventricular free wall. Three major regions were identified, in which this transformation was most commonly observed: the apex, the outflow tract, and the inferolateral basal right ventricle (RV) near the tricuspid annulus. The first surgical efforts in this disorder were directed at incising visually identifiable zones of fatty replacement of the right ventricular wall. However, postoperative VT recurrences in some cases were determined to be caused by re-entry in other regions of the RV that had not been sources of arrhythmia earlier. This led to the right ventricular disarticulation procedure, in which the entire right ventricular free wall was incised from its septal attachments, and the myocardium that could not be completely incised (infundibulum, near the valve annuli) was frozen with a cryoprobe. The right ventricular free wall was then reattached with a long suture line. In this way, VT could still occur in the isolated RV, but not spread to the LV (which remained in sinus rhythm). Although conceptually elegant, the results of this procedure were disappointing; arrhythmia “cure” rates were high, but poor hemodynamic performance (with the RV as merely a passive conduit) caused significant morbidity and even death. As a result, this procedure is rarely performed now; patients with ARVD most commonly receive ICD therapy.

In cases of VT after right ventriculotomy for repair of congenital heart disease, mapping studies have usually shown a re-entrant wave front circulating around the incision that acts as a nonconducting barrier. This can be treated by continuing the incision to another nonconducting barrier such as the pulmonic valve annulus. This is usually accomplished with extensive cryoablation. In some cases, VT appears to be due to re-entry around the ventriculoseptal defect patch.

Rarely, VT occurs in the absence of structural heart disease. Several syndromes of VT in this setting have been relatively well characterized, including exercise-provoked right ventricular outflow tract (RVOT) VT of focal origin, and a re-entrant arrhythmia involving a portion of the left ventricular Purkinje system, right bundle branch block with left-axis deviation (RBBB-LAD) VT. An uncommon variant of RVOT VT arises in the LV, often near the aortic or mitral valve annulus. Although some of the earliest reports of surgical therapy of VT were for these arrhythmias, surgery is rarely needed currently because of the very high success rates of catheter ablation (>90%). Several patients underwent surgery for RVOT VT before the development of catheter ablation techniques. General anesthesia can have a profound quieting effect on these types of VT; catecholamine infusion may restore firing of the focus but can also simply increase the sinus rate over the intrinsic rate of the VT. We have had to pack the sinus node in ice to prevent this latter effect and allow the VT to become manifest so that its origin could be mapped. Recently, minimally invasive surgical therapy for epicardial foci of VT that have not responded to catheter ablation has been reported. Only rare cases of RBBB-LAD VT have undergone surgical therapy; in one fascinating case, the tachycardia terminated following incision of a left ventricular false tendon. Histologic examination revealed a large number of Purkinje cells, strengthening the association of the specialized conduction system with this unusual arrhythmia.

Current Status

Because of the high efficacy and low morbidity and mortality of ICD implantation and catheter ablation, surgical therapy for VT has largely been relegated to the second or even third tier of therapy, after these other methods have failed to effect adequate arrhythmia control. However, in some situations, surgical therapy warrants earlier consideration: