Haris M.Haqqani, MBBS (Hons), PhD, and David J. Callans, MD
A 35-year-old previously well man with no family history of cardiac disease presented with dyspnea, cough, recurrent palpitations, and syncope due to sustained monomorphic ventricular tachycardia (VT). He was in sinus rhythm with right bundle branch block (RBBB) and had moderate segmental left ventricular (LV) systolic dysfunction with an ejection fraction (EF) of 35%. Coronary angiography showed normal coronary arteries, and three monomorphic VTs were induced at EP study. Standard medical therapy was commenced and uptitrated for nonischemic cardiomyopathy (NICM), and a single chamber implantable cardioverter-defibrillator (ICD) was placed.
The patient was referred to after sustaining multiple shocks due to monomorphic VT despite treatment with amiodarone. He had unchanged LV systolic dysfunction and global right ventricular (RV) hypokinesis. Bilateral perihilar and mediastinal adenopathy was seen on thoracic CT, and cardiac PET scan showed uptake here as well as in the ventricles in a distribution consistent with sarcoidosis. Gluococortiocoids were commenced, and the patient underwent catheter ablation for ongoing episodes of VT.
Endocardial substrate mapping showed preserved LV voltage but a confluent area of anterior and septal RV scarring with widespread isolated and fractionated potentials (Figure 22-1A to C).
FIGURE 22-1 Bipolar right ventricular endocardial voltage maps from the patient in the case example are displayed in the right anterior oblique (RAO), left anterior oblique (LAO), and left lateral projections (A to C). A large confluent area of low bipolar voltage containing widespread isolated late potentials (black dots) is seen consistent with scarring. This is present in the anterior right ventricle and extends to the basal and mid septum. The unipolar right ventricular endocardial voltage map is displayed (D) and is adjusted for left ventricular septal wall thickness, rather than right ventricular free wall, with normal peak to peak amplitude defined as >8.3 mV. This suggests a much larger area of basal septal involvement.
Unipolar voltage analysis suggested a significantly larger area of intramural septal substrate (Figure 22-1D). Seven distinct VTs were inducible with programmed stimulation, only two of which were of RBBB configuration (Figure 22-2). Due to frequently changing morphologies as well as hemodynamic intolerance, only one of the VTs was mappable: VT 2, which had a cycle length of 431 ms. Entrainment mapping showed this VT to be a large macroreentrant circuit with an exit site on the infundibular septum of the RV and an entrance site on the inferoseptal RV (Figure 22-3). Central isthmus circuit components were seen on the septal RV in confluent area of endocardial scarring containing widespread isolated and fractionated potentials. Ablation here terminated VT, but other VTs were still inducible. Extensive substrate ablation was then performed from both sides of the interventricular septum, targeting these potentials and sites of good pace-maps with long stimulus to QRS delay. Following this, VT 3 was still inducible with a right bundle right inferior axis morphology and precordial transition pattern break in V2, strongly suggestive of a preaorticepicardial exit. Percutaneous pericardial access was obtained, and epicardial substrate mapping showed preserved voltages over the LV and RV free walls with the expected low-voltage area over the interventricular groove due to pericoronary fat. However, at the anteroseptal LV and at the LV summit, on either side of the left anterior descending artery, a cluster of isolated and fractioned potentials were seen, and pace-maps here, while not perfect, recreated the precordial and limb lead vectors of VT 3 (Figure 22-4). Extensive ablation here (after coronary angiography defined the LAD to be >5 mm from lesion sites) and in the preaortic region of the LV endocardium rendered this VT noninducible. Rapid VT, similar to VT 5, was still inducible with triple extra stimuli at the end of the procedure.
FIGURE 22-2 The seven ventricular tachycardia (VT) morphologies induced in this patient are displayed. VT 3 and VT 7 have left bundle branch block morphology. Only VT 2 was mappable.
FIGURE 22-3 Entrainment mapping of VT 2 from the patient in the case example, proving this VT to be due to a large loop reentrant circuit located in septal substrate. (A) This figure displays the endocardial right ventricular bipolar voltage map showing confluent septal scarring. (B) Exit site response is demonstrated, showing entrainment with concealed fusion (ECF), postpacing interval (PPI) equaling the tachycardia cycle length (TCL), and short stimulus to QRS (S-QRS) interval equaling the electrogram to QRS (egm-QRS). (C) Entrance site response is shown with concealed entrainment, PPI=TCL, and long S-QRS equaling the egm-QRS.
FIGURE 22-4 The epicardial voltage map from the patient in the case example is displayed. A small patch of midanteroseptal scarring is displayed with networks of isolated late potentials (one is highlighted) in addition to low bipolar voltage (<0.5 mV), distinguishing it from epicardial pericoronary fat. VT 3 was pacemapped to this region as shown. The fluoroscopic proximity of the endocardial and epicardial catheters mapping VT 3 is shown. Only epicardial pacemapping was able to recreate the V1-V3 precordial transition pattern break as shown.
After this procedure, no ICD shocks occurred over short-term follow-up.
This patient with NICM due to cardiac sarcoidosis highlights some of the considerable challenges in dealing with VT in this context. The initial consideration of establishing a secondary cause for the NICM is well demonstrated as some etiologies other than idiopathic dilated cardiomyopathy (DCM) may have specific management or prognostic considerations. In the case of cardiac sarcoidosis, immunosuppression may be considered although even with treatment, prognosis is guarded with a considerable proportion of patients progressing to death or cardiac transplantation. Establishment of the diagnosis may require, in addition to a detailed clinical assessment, evaluation with biomarkers, serology, cardiac magnetic resonance imaging (MRI), PET scanning, and endomyocardial biopsy.
As seen in this patient, multiple VT morphologies are the rule in the setting of cardiac sarcoidosis, and indeed with any NICM, with the majority generally being unmappable by entrainment. Given the usually widespread, disparate regions of patchy, often subepicardial inflammation and fibrosis, both LBBB and RBBB VT configurations are seen. This patient displayed a less common pattern of confluent septal scarring, primarily manifest from the RV side of the septum with a patchier continuation to the epicardial aspect of the septum in the preaortic LV summit region. Such confluent scarring forms the electrophysiologic milieu necessary for the establishment of large-loop reentrant VT circuits, such as those that were proved by entrainment mapping in this patient. After ablating mappable tachycardias, extensive substrate ablation is necessary to target the usually widespread regions of endocardial, epicardial, and intramural scarring seen in these patients. Even after successful VT ablation, the prognosis for transplant-free survival is guarded in cardiac sarcoidosis and is worse than in other etiologies of NICM due to progressive heart failure and recurrent VT.1,2
EPIDEMIOLOGY AND CLINICAL SIGNIFICANCE
Although sustained monomorphic VT is an uncommon presentation in NICM, sudden death due to presumed ventricular arrhythmias is well described. The risk of this rises with worsening LV systolic function as the burden of ventricular fibrosis increases and the capacity to hemodynamically tolerate VT declines. The search for better risk stratifying tools remains the focus of much current investigation as this will lead to better targeted ICD therapy. Monitoring studies in patients suffering sudden death have demonstrated that monomorphic VT is the antecedent arrhythmia in many of these cases.3 Although sudden death can be effectively prevented with ICDs, such devices are not a cure for VT, and multiple shocks can confer considerable morbidity on patients and possibly even increase mortality.4-6Antiarrhythmic drug therapy has not proven to be an effective management option to eliminate shocks and comes with a significant risk of adverse events, particularly with the use of amiodarone.7
ETIOLOGY AND PATHOPHYSIOLOGY OF VT IN NICM
In the context of NICM and in direct comparison to the postinfarct situation, catheter ablation of scar-related VT can be a difficult procedure with poorer outcomes in general.2 Compared to other contexts, VT in NICM is more likely to be focal and to involve the His-Purkinje system.8 These latter VTs include typical and atypical forms of bundle-branch reentry and are very important to recognize as they are readily amenable to catheter ablation.9 Myocardial reentrant VT is seen when the interstitial and replacement fibrosis typically present in the early stages of NICM has progressed to the point of forming more confluent regions of scarring that allow for the fixed and functional barriers of VT circuits to be formed. Unlike a mature infarct scar, which extends inward from the endocardium, the scarring in NICM may lie deep to the myocardium surface, where it can be effectively shielded from currently available mapping and ablation technologies.
ENDOCARDIAL ELECTROPHYSIOLOGIC SUBSTRATE
Although patients with NICM often have diffuse interstitial and replacement fibrosis as well as variable degrees of myocyte hypertrophy and myofibril disarray,10 those presenting with monomorphic VT generally have confluent areas of scar which form the substrate for reentrant and triggered arrhythmias.8 Remarkably, in a theme echoed across the various etiologies of nonischemic left and right ventricular cardiomyopathies, the fibrosis is generally centered around the basal, periannular regions.11,12 However, it is important to highlight the marked heterogeneity observed, with some patients having no detectable endocardial abnormality at all. The precise characteristics of this electrophysiologic substrate can be defined with the use of electroanatomic mapping as demonstrated in the previous case example. Normal bipolar voltage parameters are extrapolated from mapping completed in ischemic cardiomyopathy, and normal voltage is measured as peak-to-peak signal amplitude of >1.5 mV, although pathologic correlation studies have not been performed in NICM. On the resulting electroanatomic substrate map, fractionated and isolated late potentials (ILPs) are tagged, and areas of slow conduction and putative VT exits are located by pace-mapping. In LV NICM, the basolateral periannular regions are much more commonly involved than the basal septal areas.11 The pathologic process may progress apically as the disease evolves, thus generating apical VT morphologies, and this may predict a worse prognosis as the total scar burden rises.13
EPICARDIAL ELECTROPHYSIOLOGIC SUBSTRATE
Although there is considerable heterogeneity, many patients with NICM-related VT have a greater burden of scarring on the epicardium than on the endocardium.14 This characteristically mirrors the regional pattern seen on the endocardium with a predilection for the basolateral areas overlying the endocardial scarring, as well as the preaortic region of the LV summit. The presence of epicardial fat, particularly near the coronary arteries in the atrioventricular and interventricular sulci, confounds the assessment of epicardial fibrosis due to the attenuating effect it has on signal amplitude. Consequently, a bipolar voltage cutoff for epicardial substrate of <1.0 mV has been shown to be a better discriminator, in addition to requiring the low voltage area to contain fractionated and isolated late potentials (which are not caused by fat but are reflective of slow conduction within scar).14 The summated effects of the endocardial and epicardial substrate in the basolateral LV can be seen on the surface ECG in the R:S ratios of V1and V6.15
ELECTROCARDIOGRAPHIC CLUES TO EPICARDIAL VT EXIT
The predilection for the electrophysiologic substrate in NICM to involve the epicardium means that a considerable proportion of the VTs seen in these patients will have epicardial origins. In the case of VT due to intramyocardial reentry, this means that some or all of the circuit components may be epicardial.16 If the exit of the reentrant wavefront from the constrained diastolic isthmus is located on the epicardium, several characteristic surface ECG features may be seen during VT. Broadly, these can be divided into two categories: (1) interval criteria and (2) morphology criteria. The former are QRS onset criteria and include a pseudo-delta wave duration of ≥34 ms, a V2 intrinsicoid deflection of ≥85 ms, and a shortest RS complex duration of ≥121 ms.17 These are based on the assumption that epicardial VT exits are relatively late in engaging the His-Purkinje system compared to endocardial exit sites. The second group, the morphology criteria, is based on the fact that the ventricular myocardium is thick enough for its transmural depolarization to register on the surface ECG. In sinus rhythm, this can be seen as small Q waves in V6reflective of left-to-right transmural septal activation. The presence of QS waves in focus leads during VT (lead I for basal, and the inferior leads for inferior sites of origin suggest an epicardial site of VT circuit exit. With epicardial VT exits on the basolateral LV, the inferior leads do not register a small initial Q wave because all activation is directed toward them, with no initial endocardial-to-epicardial basolateral LV vector as is seen with an endocardial exit (Figure 22-5). Similar morphology criteria can be applied to other exit regions.18 Valles et al described an algorithm combining both interval and morphology criteria with an accuracy of 90%.19
FIGURE 22-5 A 68-year-old man with idiopathic NICM who had undergone three prior endocardial and epicardial VT ablations and one surgical VT ablation presented with the clinical VT shown. Both interval and morphology criteria strongly suggest an epicardial exit from the basolateral left ventricle with long intrinsicoid deflections, pseudo-delta waves, and RS complexes in the precordium, as well as the absence of inferior Q waves and the absence of R waves in lead I and aVL. Extensive adhesions from prior sternotomy and VT ablation were encountered, limiting epicardial mapping, but the clinical VT was pace-mapped to the region shown and successfully ablated there.
It is important to stress that all the ECG criteria predicting epicardial VT origins are only able to give information about VT exit sites. The absence of such criteria does not exclude the possible presence of other critical VT circuit components on the epicardium (or intramurally), even with endocardial exit sites.
SEPTAL AND INTRAMURAL SUBSTRATE
While some degree of septal scarring is probably present in the majority of NICM patients with VT, a small but important minority of patients display only isolated septal substrate with sparing of the basolateral LV (Figure 22-6).20These patients may have low bipolar voltage with abnormal electrograms on either the LV or RV aspects of the interventricular septum (or both), as well as on the epicardial surfaces at the top and bottom of the septum, namely the preaortic region on the LV summit or at the crux. Rarely, there may be no bipolar voltage abnormality on any cardiac surface in patients with intramurally confined substrate, but low unipolar voltage may suggest its presence deep to the endocardial surface.21 Patches of intramural septal delayed gadolinium enhancement are a well-described finding on cardiac MRI in NICM, and the electrophysiologic substrate described in this subgroup of NICM VT patients is likely to correspond to the same pathology as visualized on MRI.22
FIGURE 22-6 Endocardial voltage maps from a 47-year-old man with NICM and isolated septal scarring who presented with VT storm and 52 appropriate ICD shocks. Left ventricular endocardial voltage was essentially preserved, but septal infundibular RV scarring was seen. Epicardial substrate map (not shown) was normal. Pre-ICD MRI had suggested septal delayed enhancement with LV free wall sparing. His clinical and four other induced VT morphologies were all mapped to the septum and successfully ablated. He was noninducible at the end of the case and has had no further VT over a 15-month follow-up period.
The most common indication for catheter ablation of VT in the setting of NICM remains recurrent appropriate ICD therapies despite medical therapy, often including high-dose amiodarone. There is some data to suggest that earlier ablation may be associated with better outcomes and lower VT recurrences.23 The risks of both endocardial and epicardial ablation are discussed with the patient, including death, stroke, myocardial infarction, groin complications, tamponade, valvular damage, atrioventricular block with septal ablation, abdominal or pericardial bleeding with pericardial access, and phrenic nerve or coronary artery injury with epicardial ablation.24 Adequate procedural preparation is essential including an analysis of all surface ECGs and ICD electrograms from clinical VTs. Intracardiac thrombus is excluded with echocardiography prior to the procedure. Patients can be studied under general anaesthesia or conscious sedation, but the latter has the distinct advantage of allowing more induced VTs to be mapped as hemodynamic tolerance is improved. Additional depth of sedation or conversion to general anesthetic may be required for percutaneous pericardial access. Depending on the degree of LV systolic dysfunction and dilatation, consideration should be given to mechanical hemodynamic support with an intraaortic balloon pump or, less frequently, with a percutaneous left ventricular assist device. Also, intracardiac echocardiography may be useful to define anatomy (valvular, papillary muscle, etc.), image ventricular scarring,25 monitor for catheter contact and lesion formation, exclude complications such as tamponade, and potentially assist with pericardial access. After vascular access is obtained, systemic anticoagulation is achieved with heparin and endocardial substrate mapping commences, often of both LV and RV. Any sustained VTs induced during this process are mapped by entrainment if hemodynamically and morphologically stable, or else they are terminated by overdrive pacing or DC cardioversion. During the substrate mapping process, endocardial low voltage zones are defined, both bipolar and unipolar, and isolated and fractioned potentials are tagged. Pacing at these sites during sinus rhythm is used to define potential slow zones, anatomically constrained conducting channels, and regions of good pace-maps. Programmed ventricular stimulation is then performed, and induced VTs are compared for cycle length and morphology (both on surface ECG and ICD electrograms) to clinical tracings. Bundle branch and interfascicular reentry VT mechanisms are excluded upfront by His bundle electrogram and Purkinje potential analysis, as well as RV apical entrainment. Mappable VTs have their circuit components (entrance, central isthmus, exit) defined by entrainment and ablated at critical sites. Unmappable VTs have their putative critical machinery defined in sinus rhythm by analysis of the voltage map and by pace-mapping. Extensive endocardial substrate ablation is then performed. In some cases, VT circuit components may be best ablated from the left aortic sinus of Valsalva.26 Programmed stimulation is then repeated, and ongoing inducibility of sustained monomorphic VT (particularly clinical morphologies) is an indication to proceed to epicardial mapping.
Systemic anticoagulation must first be fully reversed with protamine, and a normalized activated clotting time must be documented. Percutaneous pericardial access is obtained in the manner first described by Sosa et al,27 and after an intracardiac wire position has been excluded by fluoroscopy in the left anterior oblique projection, a deflectable sheath is introduced into the pericardial space and aspirated to assess for pericardial bleeding. Epicardial substrate mapping is then performed, and VT induction, mapping, and ablation are repeated as on the endocardium. Phrenic nerve capture with high output pacing must be assessed prior to ablation at sites on the LV free wall. Likewise, coronary artery imaging is necessary to ensure lesions are delivered a safe distance from major epicardial coronary vessels.
If minimal epicardial substrate is found or if no VT circuit components can be mapped to the epicardium, a detailed assessment for the presence of intramural scarring is important, particularly on the septum.20 In these cases, conventional irrigated radiofrequency ablation may be ineffective in creating lesions of sufficient depth to abolish conduction through deep intramural VT isthmuses. Bipolar radiofrequency energy delivery or intracoronary ethanol ablation may both be considered in such instances.
Despite significant limitations, the currently accepted end point for the ablation procedure remains VT noninducibility.
Compared to the postinfarct context where the substrate is subendocardial in the majority of cases, scar-related VT ablation in NICM is a more difficult undertaking. Monomorphic VT is rarer in this context, thus large series and randomized trials of ablation strategies do not exist. The largest published experiences suggest that the majority of patients can achieve a reasonable VT-free survival and reduction in recurrent ICD shocks, but the outcomes depend significantly on the cause of NICM with sarcoidosis patients faring the worst.2 If catheter ablation fails, open surgical cryoablation may offer an additional alternative in select cases.28 Additionally, a substantial number of NICM patients may succumb to progressive congestive cardiac failure, even after successful VT ablation.
1. Koplan BA, Soejima K, Baughman K, Epstein LM, Stevenson WG. Refractory ventricular tachycardia secondary to cardiac sarcoid: electrophysiologic characteristics, mapping, and ablation. Heart Rhythm. 2006;3(8):924-929.
2. Tokuda M, Tedrow UB, Kojodjojo P, et al. Catheter ablation of ventricular tachycardia in nonischemic heart disease. Circ Arrhythm Electrophysiol. 2012;5(5):992-1000.
3. Bayes de Luna A, Coumel P, Leclercq JF. Ambulatory sudden cardiac death: mechanisms of production of fatal arrhythmia on the basis of data from 157 cases. Am Heart J. 1989;117(1):151-159.
4. Bardy GH, Lee KL, Mark DB, et al. Amiodarone or an implantable cardioverter-defibrillator for congestive heart failure. N Engl J Med. 2005;352(3):225-237.
5. Bradfield JS, Buch E, Shivkumar K. Interventions to decrease the morbidity and mortality associated with implantable cardioverter-defibrillator shocks. Curr Opin Crit Care. 2012(5);18:432-437.
6. Sweeney MO. The contradiction of appropriate shocks in primary prevention ICDs: increasing and decreasing the risk of death. Circulation. 2010(25);122:2638-2641.
7. Connolly SJ, Dorian P, Roberts RS, et al. Comparison of beta-blockers, amiodarone plus beta-blockers, or sotalol for prevention of shocks from implantable cardioverter defibrillators: the OPTIC Study: a randomized trial. JAMA. 2006;295(2):165-171.
8. Soejima K, Stevenson WG, Sapp JL, Selwyn AP, Couper G, Epstein LM. Endocardial and epicardial radiofrequency ablation of ventricular tachycardia associated with dilated cardiomyopathy: the importance of low-voltage scars. J Am Coll Cardiol. 2004;43(10):1834-1842.
9. Lopera G, Stevenson WG, Soejima K, et al. Identification and ablation of three types of ventricular tachycardia involving the His-Purkinje system in patients with heart disease. J Cardiovasc Electrophysiol. 2004;15(1):52-58.
10. Roberts WC, Siegel RJ, McManus BM. Idiopathic dilated cardiomyopathy: analysis of 152 necropsy patients. Am J Cardiol. 1987;60(16):1340-1355.
11. Hsia HH, Callans DJ, Marchlinski FE. Characterization of endocardial electrophysiological substrate in patients with nonischemic cardiomyopathy and monomorphic ventricular tachycardia. Circulation. 2003;108(6):704-710.
12. Marchlinski FE, Zado E, Dixit S, et al. Electroanatomic substrate and outcome of catheter ablative therapy for ventricular tachycardia in setting of right ventricular cardiomyopathy. Circulation. 2004;110(16):2293-2298.
13. Frankel DS, Tschabrunn CM, Cooper JM, et al. Apical ventricular tachycardia morphology in left ventricular nonischemic cardiomyopathy predicts poor transplant-free survival. Heart Rhythm. 2013;10(5):621-626.
14. Cano O, Hutchinson MD, Lin D, et al. Electroanatomic substrate and ablation outcome for suspected epicardial ventricular tachycardia in left ventricular nonischemic cardiomyopathy. J Am Coll Cardiol. 2009;54(9):799-808.
15. Tzou WS, Zado ES, Lin D, et al. Sinus rhythm ECG criteria associated with basal-lateral ventricular tachycardia substrate in patients with nonischemic cardiomyopathy. J Cardiovasc Electrophysiol. 2011;22(12):1351-1358.
16. Swarup V, Morton JB, Arruda M, Wilber DJ. Ablation of epicardial macroreentrant ventricular tachycardia associated with idiopathic nonischemic dilated cardiomyopathy by a percutaneous transthoracic approach. J Cardiovasc Electrophysiol. 2002;13(11):1164-1168.
17. Berruezo A, Mont L, Nava S, Chueca E, Bartholomay E, Brugada J. Electrocardiographic recognition of the epicardial origin of ventricular tachycardias. Circulation. 2004;109(15):1842-1847.
18. Bazan V, Gerstenfeld EP, Garcia FC, et al. Site-specific twelve-lead ECG features to identify an epicardial origin for left ventricular tachycardia in the absence of myocardial infarction. Heart Rhythm. 2007;4(11):1403-1410.
19. Valles E, Bazan V, Marchlinski FE. ECG criteria to identify epicardial ventricular tachycardia in nonischemic cardiomyopathy. Circ Arrhythm Electrophysiol. 2010;3(1):63-71.
20. Haqqani HM, Tschabrunn CM, Tzou WS, et al. Isolated septal substrate for ventricular tachycardia in nonischemic dilated cardiomyopathy: incidence, characterization, and implications. Heart Rhythm. 2011;8(8):1169-1176.
21. Hutchinson MD, Gerstenfeld EP, Desjardins B, et al. Endocardial unipolar voltage mapping to detect epicardial ventricular tachycardia substrate in patients with nonischemic left ventricular cardiomyopathy. Circ Arrhythm Electrophysiol. 2011;4(1):49-55.
22. McCrohon JA, Moon JC, Prasad SK, et al. Differentiation of heart failure related to dilated cardiomyopathy and coronary artery disease using gadolinium-enhanced cardiovascular magnetic resonance. Circulation. 2003;108(1):54-59.
23. Frankel DS, Mountantonakis SE, Robinson MR, Zado ES, Callans DJ, Marchlinski FE. Ventricular tachycardia ablation remains treatment of last resort in structural heart disease: argument for earlier intervention. J Cardiovasc Electrophysiol. 2011;22(10):1123-1128.
24. Sacher F, Roberts-Thomson K, Maury P, et al. Epicardial ventricular tachycardia ablation a multicenter safety study. J Am Coll Cardiol. 2010;55(21):2366-2372.
25. Bala R, Ren JF, Hutchinson MD, et al. Assessing epicardial substrate using intracardiac echocardiography during VT ablation. Circ Arrhythm Electrophysiol. 2011;4(5):667-673.
26. Yokokawa M, Good E, Crawford T, et al. Ventricular tachycardia originating from the aortic sinus cusp in patients with idiopathic dilated cardiomyopathy. Heart Rhythm. 2011;8(3):357-360.
27. Sosa E, Scanavacca M, d’Avila A, Pilleggi F. A new technique to perform epicardial mapping in the electrophysiology laboratory. J Cardiovasc Electrophysiol. 1996;7(6):531-536.
28. Anter E, Hutchinson MD, Deo R, et al. Surgical ablation of refractory ventricular tachycardia in patients with nonischemic cardiomyopathy. Circ Arrhythm Electrophysiol. 2011;4(4):494-500.