Radiofrequency Catheter Ablation of Ventricular Tachyarrhythmia Under Navigation Using EnSite Array

Circulation Journal Official Journal of the Japanese Circulation Society http://www.j-circ.or.jp Radiofrequency Catheter Ablation of Ventricular Tachyarrhythmia Under Navigation Using EnSite Array Koji Miyamoto, MD; Takeshi Tsuchiya, MD; Sumito Narita, MD; Yasutsugu Nagamoto, MD; Takanori Yamaguchi, MD; Shin-ichi Ando, MD*; Kiyoshi Hayashida, MD**; Yoshito Tanioka, MD ; Naohiko Takahashi, MD Background: EnSite array (EA) provides virtual activation of ventricular tachycardia (VT) and premature ventricular contraction (PVC) on a beat-to-beat basis. Methods and Results: Fifty-five consecutive patients (age 52±16 years) with 79 VTs/PVCs undergoing EAguided radiofrequency catheter ablation (RFA) were studied, of whom 7 patients had organic heart diseases. A virtual activation map showed that 66 VTs/PVCs originated from the right ventricle (RV), including the RV outflow tract in 57, lateral wall of RV in 4, His bundle region in 3 and tricuspid annulus in 2. Ten VTs/PVCs originated from the left ventricle (LV), including the LV endocardium in 7 and aortic sinus cusp in 3. The origins of 3 PVCs, one each in 3 patients, were not identified. Six of 38 VTs were sustained and the remaining 32 VTs were non-sustained. RFA eliminated all but 3 focal PVCs, and all macroreentrant VTs at a critical conducting pathway, which was identified by the combined use of contact voltage and virtual activation maps. There were 11±9 applications, and the radiofrequency energy and fluoroscopy time were 11,354±13,360 J and 30±21 min, respectively. All patients with acute success were free of any symptoms during a follow up of 21±11 months. Conclusions: EA-guided RFA is safe and effective for VT/PVC, irrespective of its origin, mechanism, sustainability, hemodynamic condition, and underlying heart disease. Key Words: Catheter ablation; EnSite array; Premature ventricular contraction; Ventricular tachycardia Most idiopathic ventricular tachycardia (VT) and hemodynamically tolerable sustained monomorphic VT associated with organic heart disease are not usually difficult to eliminate, but elimination of some VTs by radiofrequency catheter ablation (RFA) is still challenging. The difficulties in VT ablation are mostly due to their non-sustained, multifocal, easily changed nature in the reentrant circuit (especially on pacing maneuver), and hemodynamic instability, or the difficulty in producing a transmural lesion. Until now, many efforts have been made to overcome these difficulties, but there are still a few shortcomings, even though an electroanatomical mapping system has been used. 1 6 EnSite Array (EA) is a 3-dimensional (D) mapping tool that provides virtual but precise activation of any tachycardia on a beat-to-beat basis. 7,8 We recently showed that virtual beat-to-beat elucidation of activation provides accurate iden- Table 1. Patient Characteristics No. patients 55 Age (years) 52±16 Male, n (%) 27 (49) Organic heart disease, n (%) 7 (13) VT/PVC, n (per patient) 79 (1.4±0.9) RV/LV VT/PVC 66/10 VT/PVC, n 38/41 Sustained VT/Non-sustained VT, n 6/32 The origins of three PVCs from 3 patients could not be classified into RV or LV because these PVCs could not be terminated by RF catheter ablation from both of the ventricles. VT, ventricular tachycardia; PVC, premature ventricular contraction; RV, right ventricle; LV, left ventricle; RF, radiofrequency. Received December 17, 2009; revised manuscript received February 22, 2010; accepted March 8, 2010; released online May 14, 2010 Time for primary review: 46 days EP Expert Doctors-Team Tsuchiya, Kumamoto, *Division of Cardiology, Saiseikai Futsukaichi Hospital, Fukuoka, **Division of Cardiology, Saga Prefectural Hospital Koseikan, Saga, Division of Cardiology, Omura Municipal Hospital Cardiovascular Center, Omura and Department of Internal Medicine 1, Faculty of Medical Oita University, Oita, Japan Mailing address: Takeshi Tsuchiya, MD, EP Expert Doctors-Team Tsuchiya, 3-14-28 Koto, Kumamoto 862-0909, Japan. E-mail: tsuchiya@s1.kcn-tv.ne.jp ISSN-1346-9843 doi: 10.1253/circj.CJ-09-1008 All rights are reserved to the Japanese Circulation Society. For permissions, please e-mail: cj@j-circ.or.jp

MIYAMOTO K et al. Figure 1. MEA positioning. Adequate placement of MEA is required to obtain the best virtual activation map. We made efforts to locate MEA as close as possible to the VT/PVC focus for focal VT/PVC and at the exit site for macroreentrant VT estimated by 12-lead ECG during palpitation. Examples of MEA placement for VT/PVC originating from (A) right ventricular outflow tract, (B) aortic sinus cusp, (C) anterior or lateral left ventricle, and (D) septal, posterior or inferior left ventricle. MEA was placed via the antegrade trans-septal approach in (B,C), and via the retrograde transaortic approach in (D). ABL, ablation catheter; LAO, left anterior oblique; MEA, multielectrode array; MV, mitral valve; PVC, premature ventricular contraction; RAO, right anterior oblique; VT, ventricular tachycardia. tification of the location of the atrial tachycardia (AT) focus and elucidation of the entire reentrant circuit of reentrant AT, even if the AT is non-sustained, multifocal, or easily changed in the reentrant circuit especially by a pacing maneuver. 9 In the present study we examined how EA works during VT ablation. Methods Study Population Written informed consent was obtained from all patients before the study. Fifty-five consecutive patients (age 52±16 years, 27 men) undergoing VT/premature ventricular contraction (PVC) ablation under navigation using EA (EnSite version 3.0 in 20 patients and version 6.0J in 35 patients) between May 2006 and November 2009 were included (Table 1). Seven patients (13%) had organic heart disease, including old myocardial infarction in 1, dilated cardiomyopathy in 1, aortic regurgitation in 1, isolated non-compaction of the ventricular myocardium in 1, tachycardia-induced cardiomyopathy in 2 and Rastelli s operation for complete transposition of the great arteries in 1. All anti-arrhythmic medications were discontinued more than 5 days before the study. No patients were being treated with amiodarone. Electrophysiology and Ablation Electrophysiology investigation and subsequent RFA were performed under fasting conditions and conscious sedation, using iv hydroxyzine pamoate (25 50 mg). In all patients, a quadripolar electrode catheter was placed at the His bundle region and 20-pole multielectrode catheters were placed within the coronary sinus (St Jude Medical, Minnetonka, MN, USA) and the ventricle of interest (Ten-Ten, St Jude Medical). These catheters were used to record the bipolar electrogram and pace. Throughout the procedure activated clotting time was monitored every 30 min to maintain it between 200 and 300 s for the right ventricular (RV) VT/PVC, and 300 and 400 s for the left ventricular (LV) VT/PVC. If the activated clotting time was below the limits specified here, an adequate amount of heparin was injected to keep the activated clotting time up to the adequate level. Non-Contact Mapping Details of EA have been described previously. 7 In short, EA (St Jude Medical) consists of a multielectrode array (MEA) catheter and a custom-designed amplifier system connected to a Silicon Graphics workstation used to run specially designed system software. The system simultaneously reconstructs more than 3300 virtual unipolar electrograms (VUE) and displays them in 3-D images of the chamber of interest in isochronal or isopotential mode. Mapping Procedure The MEA was introduced into RV or LV from the right femoral vein or femoral artery depending on the patients (Figure 1). Best efforts were made to locate MEA as close as possible to the VT/PVC origin, as estimated from an ECG

Ablation of VT Using EnSite Array Table 2. VT/PVC: Idiopathic or Organic Idiopathic Organic P value VT/PVC, n (per patient) 63 (1.4±0.7) 13 (1.9±1.6) VT/PVC 27/36 11/2 0.0061 Sustained VT/Non-sustained VT, n 3/24 3/8 0.22 Origin, RV/LV 57/6 9/4 0.039 No. RF pulses 7±5 9±8 0.17 RF energy (J) 5,933±5,748 10,054±14,563 0.09 Fluoroscopy time (min) 24±12 58±32 <0.0001 Abbreviations see in Table 1. during a palpitation attack, including the VT/PVC focus for focal VT/PVC and an exit site for the macroreentrant VT, while the stability of MEA was ensured. In cases of VT/PVC originating from the RV outflow tract (RVOT), MEA was introduced into RVOT through a Mullins long sheath to locate the tip of the MEA just beneath the pulmonary valve (Figure 1A). In those originating from the RV body, MEA was placed at the middle of the RV body to locate the MEA tip at the RV apex. In cases of VT/PVC originating from the aortic sinus cusp, MEA was introduced into the left atrium through a Mullins long sheath in some, which was advanced into the left atrium using the standard Brockenbrough technique, and placed within the left atrium to locate the MEA tip at the mitral annulus (Figure 1B). In others, MEA was placed at RVOT while navigating the mapping catheter from MEA within the aortic sinus cusp. In cases of VT/PVC originating from anterior or lateral LV, MEA was introduced into the LV through a Mullins long sheath, which was introduced by the standard Brockenbrough technique to the left atrium, and the MEA tip was located at the LV apex (Figure 1C). In cases of VT/PVC originating from inferior, posterior, or septal LV, the MEA was introduced into the LV through the aorta and the MEA tip was placed at the LV apex (Figure 1D). At the beginning of the electrophysiology measurement, virtual activation mapping using EA was performed during sinus rhythm (SR), sometimes in conjunction with a contact voltage map during SR in an attempt to elucidate the VT/PVC substrate. In the substrate map, the low voltage zone (LVZ) was defined as a contact bipolar electrogram amplitude <1.2 mv and encircled with a black line within the ventricle of interest. Then VT/PVC was induced with programmed stimulation and/or isoproterenol injection. After induction the VT/PVC activation was analyzed using a virtual activation map of the EA. The VT/PVC activation was further confirmed by the morphology of VUE. The VT/PVC activation was analyzed separately for focal VTs/PVCs and macroreentrant VTs. For focal VTs/PVCs, the focal discharge pattern was recognized in a centrifugal propagating manner from a VT/PVC focus where VUE exhibited a QS or q-rs pattern. For macroreentrant VT, a critical conducting pathway was elucidated using the virtual activation map, the exit from which an rs or QS pattern was demonstrated in VUE. Virtual activation of macroreentrant VT was further analyzed in relation to LVZ, which was examined in advance during SR. In some patients with reentrant VT, entrainment pacing was sometimes performed to confirm the reentrant circuit, taking care not to transform one VT either into another VT or a ventricular fibrillation. Ablation Procedures In each patient a 4- or 8-mm tip steerable catheter (Fantasista, Japan Lifeline; Blazer II, Boston Scientific) was used to record bipolar electrograms, pacing, and ablation. Radiofrequency (RF) energy was delivered in a temperature-controlled fashion with an upper limit of 50 C and 40 W. RF energy was applied to the VT/PVC focus for focal VT/PVC, and to the exit and/or critical central common pathway within the reentrant circuit for macroreentrant VT. The endpoint of RFA for sustained VT was VT termination, subsequent non-inducibility of VT for focal VT, and non-inducibility for non-sustained focal VT and PVC. In patients with macroreentrant VT, creation of a block line on the critical conducting pathway was confirmed in addition to VT termination and non-inducibility. Post-Ablation Management and Follow-up Immediately after VT/PVC ablation, ECG recording and transthoracic echocardiography were performed in all patients to examine procedure-related complications. Patients were followed up for 1 week, 1 month, and 3 months after the procedure, and every 6 months thereafter. Holter ECG monitoring was performed if the patient complained of any symptoms. Statistical Analysis Data are expressed as mean ± SD. Data were analyzed using unpaired t-test if they were normally distributed. The chisquare test was used to analyze the independence of the 2 classification criteria in the qualitative data. P<0.05 was considered statistically significant. Results Seventy-nine VTs/PVCs in 55 patients (1.4±0.9 VTs/PVCs per patient) were analyzed with EA, and EA-guided RFA was performed (Table 2). A virtual activation map showed that 66 VTs/PVCs originated from RV, in which 62 VTs/PVCs resulted from focal discharge and the remaining 4 VTs were due to macroreentry. Ten VTs/PVCs originated from LV, in which 5 VTs/PVCs resulted from focal discharge and the remaining 5 VTs were due to macroreentry. The origins of 3 PVCs, 1 each from 3 patients could not be classified into RV or LV because they could not be terminated by RFA from either ventricle. Six of 38 VTs were sustained and the remaining 32 VTs were non-sustained. RFA eliminated all but 3 focal VTs/PVCs by RF energy delivery to the focus, and all macroreentrant VTs by RF energy delivery to a critical conducting pathway that was identified by combined use of contact voltage and virtual activation maps. The number of RFA applications was 8±8, RF energy was 7,419±10,818 J, and fluoroscopy time was 30±20 min/patient. No complications were noted. EA-guided RFA resulted in acute success in 52 of 55 cases (95%). This result was independent of whether

MIYAMOTO K et al. Figure 2. Multifocal PVCs. (A) Twelve-lead ECG indicating multifocal PVCs originating from RVOT. Multifocal PVCs with 5 different QRS morphologies were induced after isoproterenol injection in this patient. All PVCs were left bundle branch blocks with an inferior axis pattern and subtle morphological change in leads I, II, III and avl. (B) Contact voltage map of RVOT during sinus rhythm in this patient from RAO projection. Voltage map indicates LVZ defined as <1.2 mv at the posterolateral wall of RVOT. Purple represents areas with voltage >1.2 mv. (C) Virtual activation maps at the QRS onset of each PVC. The black circle indicates LVZ identified by contact voltage map. Five different PVC exits are demonstrated. Virtual activation maps of PVCs superimposed on contact voltage maps show that PVCs 1, 2, 4, and 5 are originating from within or near LVZ at the posterolateral wall of RVOT, whereas PVC 3 originates relatively far from LVZ. The white arrow in each panel indicates the direction of virtual activation. LVZ, low voltage zone; PVC, premature ventricular contraction; RAO, right anterior oblique; RVOT, right ventricular outflow tract. the origin of VT/PVC was RV or LV, focal or reentrant mechanism, sustained or non-sustained, hemodynamically stable or unstable, and whether patients had organic heart disease or not. All patients with acute success were free of any symptoms during a follow-up of 21±11 months. VT/PVC Originating From RV A virtual activation map showed that 66 VTs/PVCs originated from RV, including RVOT in 57, the lateral wall of RV in 4, the His bundle region in 3, and the tricuspid annulus in 2. Sixty-two VTs/PVCs resulted from focal discharge and the remaining 4 VTs were due to macroreentry. It was noted that

Ablation of VT Using EnSite Array Figure 3. Non-sustained VT originating from lateral wall of the RV. (A) Twelve-lead ECG during PVC showed a left bundle branch block and left axis deviation. (B) Contact voltage map of RV obtained during sinus rhythm in RAO and LAO projections. There is no low voltage zone in RV. (C) VT virtual activation map showing that VT is originating from the lateral wall of RV. The white arrow indicates the activation direction. The earliest activation site was identified at the lateral wall of RV, and the electrogram at the site preceded QRS by 40 ms. Virtual unipolar electrogram at the VT origin shows QS deflection. Radiofrequency ablation was performed at the VT focus, which resulted in abolition of VT. The brown and red circles show sites of successful ablation. (B,C) White circles, TV. LAO, left anterior oblique; PVC, premature ventricular contraction; RAO, right anterior oblique; RV, right ventricle; RVOT, right ventricular outflow tract; TV, tricuspid valve; VT, ventricular tachycardia. 3 kinds of VTs with similar but slightly different QRS morphologies originated from RVOT in the same patient who did not have organic heart disease, all of which were sustained VTs and were due to macroreentry because the classical entrainment phenomenon was observed by pacing from the inferior RVOT. All VTs were terminated to the earliest ventricular activation site by RFA. One of 31 VTs was sustained and the remaining 30 VTs were non-sustained. All VTs/PVCs were abolished by RFA with a mean number of 6±5 RF pulses and an RF energy of 6,044±7,735 J. Fluoroscopy time was 25±17 min/patient. Representative cases are shown in Figures 2,3. Figure 2 shows multifocal PVCs originating from RVOT in a 67- year-old female patient. Five kinds of PVCs with similar but slightly different QRS morphologies were recorded in 12-lead ECG (Figure 2A). All PVCs exhibited a left bundle branch block and inferior axis pattern, suggesting RVOT origins. MEA was introduced into RVOT, and the MEA tip was located just beneath the pulmonary valve, which was confirmed on RV angiography. To begin the procedure, a contact voltage map of RVOT was created with EA, and LVZ confined to the posterolateral RVOT was identified (Figure 2B). An EA virtual activation map identified the origins of 5 kinds of PVCs corresponding to the QRS morphologies. It was noted that 4 of 5 PVCs originated from or near the LVZ, while the remaining 1 was from the low posterolateral RVOT outside LVZ (Figure 2C). RF energy was delivered to each PVC origin, which eliminated all PVCs. Figure 3 shows non-sustained VT originating from lateral RV in a 58-year-old male patient without organic heart disease. ECG during PVCs demonstrated a left bundle branch block and left axis deviation (Figure 3A). MEA was placed at the middle of the RV body to locate the MEA tip at the RV apex. To begin the procedure, a contact voltage map was constructed during SR in an attempt to elucidate the VT substrate (Figure 3B), but LVZ was not found in RV. A virtual activation map showed that VT originated from the lateral RV (Figure 3C), and RFA to the VT focus eliminated VT. VT/PVC Originating From LV A virtual activation map showed that 10 VTs/PVCs originated from LV, including the LV endocardium in 7 and the aortic sinus cusp in 3, in which 5 VTs/PVCs resulted from focal discharge and the remaining 5 VTs were due to macroreentry. Macroreentrant VTs were associated with organic heart disease in 4 VTs from 2 patients. Five of 7 VTs were sustained and the remaining 2 were non-sustained. All VTs/PVCs were abolished by RFA with a mean number of 12±9 RF pulses and RF energy of 10,560±8,881 J. Fluoroscopy time was 47± 18 min/patient. Two of the 7 VTs were hemodynamically intolerant, RFA was thus performed in these VTs during SR, according to the virtual activation map obtained during VT. Representative cases are shown in Figures 4,5. Figure 4 shows focal VT originating from the aortic sinus cusp. This case was previously reported in a case report. 10 The patient was a 15-year-old girl whose VT exhibited an inferior axis

MIYAMOTO K et al. Figure 4. VT originating from the aortic sinus cusp. (A) Twelve-lead ECG of VT originating from the left coronary cusp. The VT exhibited an inferior axis pattern and a rightward deviation, and R > S in lead V1, suggesting that VT came from the aortic sinus cusp or the superior portion of the mitral annulus. (B) Virtual activation map during VT in RAO projection. VT virtual activation map shows that VT originates from the left coronary cusp. The white arrow indicates the direction of activation from the left coronary cusp. The activation occurred at the left coronary cusp and subsequent activation proceeded toward the LV septum and then to the entire LV. Radiofrequency ablation was performed at the VT focus, which resulted in abolition of VT. LV, left ventricle; RAO, right anterior oblique; VT, ventricular tachycardia. Figure 5. VT originating from the LV endocardium. (A) Twelve-lead electrocardiogram during induced non-sustained VT. The VT was a right bundle branch block with superior axis pattern and a ventricular rate of 198 beats/min. (B) MEA positioning in RAO and LAO projection. MEA was placed into LV via the retrograde transaortic approach to locate the MEA tip near the posteroinferior wall of LV. Two deflectable electrode catheters were introduced into LV by transaortic (Map 1) and trans-septal (Map 2) approaches to create LV geometry and construct a contact voltage map. (C) Contact voltage map during SR in RAO and PA projections. LVZ defined as <1.2 mv is present at the posteroseptal wall of LV. Purple, voltage >1.2 mv. A VT virtual activation map superimposed on the contact voltage map exhibits macroreentrant activation, and shows the entire reentrant circuit of VT, in which the activation slowly passes through LVZ. (C) Red arrow, direction of VT activation. The contact bipolar electrogram during SR shows fragmented electrograms along the slow conducting pathway within LVZ. LAO, left anterior oblique; LV, left ventricle; LVZ, low voltage zone; MEA, multielectrode array; PA, posteroanterior; RA, right atrium; RAO, right anterior oblique; RV, right ventricle; SR, sinus rhythm; VT, ventricular tachycardia; B-RV, B-LV, biventricular pacing leads located in RV and LV.

Ablation of VT Using EnSite Array Figure 6. Failure to resolve VT. (A) Twelve-lead electrocardiogram during PVC. QRS morphologies showed slow initial precordial QRS activation. (B) Virtual activation map at the QRS onset of the PVC. The virtual unipolar electrogram at the earliest activation site in RVOT showed a q-s pattern. (C) Ablation points and contact bipolar and unipolar electrograms in the left coronary cusp. LCC, left coronary cusp; LL, left lateral; LP, left posterior; PA, posteroanterior; PVC, premature ventricular contraction; RL, right lateral; RVOT, right ventricular outflow tract; VT, ventricular tachycardia. pattern and right axis deviation. The amplitude of R wave was greater than that of S wave on ECG lead V1, suggesting that VT came from the aortic sinus cusp or the superior portion of the mitral annulus (Figure 4A). MEA was introduced into the left atrium to locate the MEA tip across the mitral valve, as shown in Figure 1B. A virtual activation map of VT showed that it was originating from the left coronary cusp (Figure 4B). The activation occurred at the left coronary cusp and the subsequent activation propagated toward the LV septum and then to the entire LV. Pacing at the site during SR resulted in a perfect pace match, confirming that the site was the VT focus. Coronary angiography was performed before delivering RF energy to the VT focus, to confirm that the site was >1 cm away from the left coronary artery ostium. RFA was performed at the VT focus, which resulted in abolition of VT.

MIYAMOTO K et al. Figure 5 shows macroreentrant VT in a 74-year-old male patient who suffered from VT associated with dilated cardiomyopathy. CRT-D was implanted and the VT episodes terminated by shock deliveries. VT consisted of a right bundle branch block with a superior axis pattern (Figure 5A), suggesting that the focus or exit site was located at the inferior or posterior wall of LV. MEA was placed into the LV via the transaortic approach to locate MEA near the postero-inferior LV (Figure 5B). To begin the procedure, a contact voltage map was constructed during SR (Figure 5C), and LVZ defined as <1.2 mv was found at the posteroseptal LV. Only non-sustained VT was induced by ventricular pacing and isoproterenol injection. ECG during VT showed a right bundle block and superior axis pattern with a ventricular rate of 198 beats/min. A virtual activation map identified the entire reentrant circuit of VT, in which LVZ acted as a critical slow conduction zone. RF energy was delivered to create a block line along the border between LVZ and the healthy tissue, where the virtual activation map showed that the virtual activation exited from the slow conduction zone. In addition, another linear RFA was performed from the center of the LVZ border to the possible central portion of the slowly conducting pathway within LVZ. Subsequently, VT could no longer be re-induced by ventricular pacing and/or isoproterenol injection. Comparison of VTs/PVCs According to Whether They Were Idiopathic or Organic Parameters of VTs/PVCs are compared in Table 2 according to whether they were idiopathic or organic. The frequency of VTs was significantly higher in organic VTs than in idiopathic ones. There were no significant differences in number of RF pulse and RF energy. Fluoroscopy time was longer in organic VTs/PVCs than idiopathic ones. Discussion Main Findings We showed that a virtual activation map of EA identified VT/PVC activation clearly, irrespective of the chamber of origin, its mechanisms including focal and macroreentry, sustainability, hemodynamic condition, and whether it was associated with organic heart disease. We also showed that the virtual activation-guided RFA was safe and effective with an acute success rate of 95%, without any recurrence during a follow-up period of 21±11 months. Virtual VT/PVC Activation EA provides a detailed activation map of VT/PVC in a beatto-beat fashion with high resolution VUE of >3,000 endocardial points. In patients with non-sustained, pleomorphic, or hemodynamically intolerant VT, the power of one-beat analysis in EA is worth emphasizing. 11,12 Furthermore, EA can navigate any electrode catheter, clearly identifying virtual VT/PVC activation, irrespective of whether VT/PVC resulted from focal discharge or macroreentrant mechanism. 7 9,11,12 The origins of focal VTs/PVCs were distributed at RVOT, the lateral RV, His bundle region, tricuspid annulus, aortic sinus cusp, and LV endocardium, and those of macroreentrant VTs were distributed at RVOT, lateral RV, and LV endocardium. 13 15 In focal VTs/PVCs, activation occurred from the VT/PVC focus, where VUE exhibited a QS or qrs pattern depending on the case, and from which activation spread out toward the entire ventricle. RFA targeting the VT/PVC focus resulted in elimination of almost all VTs/PVCs, even if VT/PVC focus was multiple and changed in location after each RF application. It seems important to place MEA at the position nearest to the earliest activation because accurate analysis of the activation could be made when it was located near the crucial site of VT/PVC. 16 18 It has been reported that the reliability of VUE cannot be retained at remote points >4.0 cm from the center of MEA, and at LVZ. 16 18 Many patients with organic heart disease had macroreentrant VTs with a large ventricle, in which the reentrant circuit included a broad LVZ. To get the best virtual activation map, MEA should be placed at the possible exit site of reentrant VT. In cases of a large ventricle and/or reentrant circuit including LVZ, it was very useful to perform contact bipolar mapping during SR to localize the arrhythmogenic substrate in the ventricle. Thereafter, the VT/ PVC virtual activation map should be superimposed on the contact substrate map obtained during SR to understand the relationship between the arrhythmogenic substrate and the VT reentrant circuit to make virtual activation analysis easier. 11 Efficacy of EA in VT Associated With Organic Heart Disease The strategy of RFA for macroreentrant VT associated with organic heart disease consists of substrate modification, circuit modification, or both. 1 6 In the former, the RFA target is a possible reentrant circuit suggested from the substrate map. In the latter, the real reentrant circuit is targeted with entrainment mapping during VT. But this strategy cannot be applied for non-sustained, hemodynamically intolerant, or pleomorphic VT because it is a time-consuming process. EA, however, allows us to elucidate a real-time reentrant circuit, and target critical sites without taking too much time. A virtual activation map of VT using EA could be obtained even if the VT is unmappable on conventional fluoroscopyguided or electroanatomical mapping; and this map is useful and safe for ablation of VT associated with organic heart disease. 7 9,11,12 Elucidation of the entire VT reentrant circuit was achieved in all cases of macroreentrant VT, and successful RFA could be performed in all patients with organic heart disease. In cases of VT/PVC originating from the LV endocardium, MEA placement is more important. 19 It should be placed in accordance with the suspected critical region, such as sites near the infarcted areas in patients with old myocardial infarction, as shown in Figures 1C,D. We positioned MEA into LV via an antegrade trans-septal approach for anterior and/ or lateral infarction, and a retrograde transaortic approach for inferior, posterior, and/or septal infarction so that MEA would be closer to the areas of interest. 20 In a patient suspected to have a globally damaged ventricle as in dilated cardiomyopathy, the critical region such as LVZ should be expected due to QRS morphology and the axis on 12-lead ECG as shown in Figure 5. In the present study accurate elucidation of VT/PVC activation in LV was obtained by locating the MEA tip near the critical region, and a high success rate was achieved. In the present study, irrespective of detailed mapping of RV and LV, RFA could not eliminate PVCs in 3 patients. In these patients, QRS morphologies showed slow initial precordial QRS activation. Thus, PVC foci may be located at the epicardium in these patients. 21 PVCs have been controlled with β-blockers in these patients. Figure 6 shows the virtual activation map of PVC in the case of failure. Figure 6A shows the 12-lead ECG of PVC and the virtual activation map at the earliest activation site within RVOT. In this case,

Ablation of VT Using EnSite Array MEA was located within the RVOT just beneath the pulmonary valve. The VUE at the earliest activation site at the RVOT showed a low-frequency small q wave followed by S wave (q-s pattern), but the activation time was not early relative to the onset of the QRS complex (Figure 6B). In general, MEA located within the RVOT allows quick discrimination between right- and left-sided origins by observing either QS pattern or typical rs pattern, on VUE, on the earliest activation site. We should be attentive, however, in interpreting the morphology of unipolar electrogram because there are exceptions, such as this case. Even if the morphology of the VUE is q-s pattern at the earliest activation site, we can say that it does not originate from RVOT endocardium when the initial timing of the q-s wave is not early relative to the QRS onset of PVC/VT and the initial q wave is low frequency. 22 This phenomenon usually results from an epicardial focus and subsequent anisotropic delayed activation toward the endocardium. Figure 6C shows ablation points, and contact bipolar and unipolar electrograms at the left coronary cusp. The contact unipolar electrogram at the earliest activation site in the LV outflow tract (LVOT) including the left coronary cusp, right coronary cusp, and anterior portion around the mitral annulus showed the rsr pattern. The RF applications to the relatively early sites at both RVOT and LVOT failed to eliminate PVCs. Use of Mullins Sheath to Introduce MEA to the Most Adequate Site for VT Adequate MEA positioning is paramount to elucidate accurate and detailed virtual activation of any type of VT. For focal VT, the MEA should be placed at a site nearest to the VT focus. For macroreentrant VT, the MEA should be placed at a site nearest to the exit of a slow conduction zone. We frequently used a Mullins sheath to introduce and stabilize MEA at the proper position, that is, the Mullins sheath was placed at the right atrium through which MEA was advanced into the RVOT just beneath the pulmonary valve in those with VT originating from RVOT, or into the RV body in those with VT originating from the RV body or tricuspid annulus. The Mullins sheath was placed at the left atrium to introduce MEA into LV using a trans-septal approach, MEA was advanced into LV, and placed at the anterior lateral LV in those with VT originating from the lateral or anterior LV. Study Limitations The aim of the present study was to evaluate the feasibility of EA-guided RFA for VT/PVC, and therefore we did not compare this method with conventional fluoroscopy-guided or electroanatomical mapping. EA has intrinsic limitations with regard to the signal-to-noise ratio. This issue is particularly important for virtual activation at sites with LVZ. Furthermore, resolution of virtual activation is limited at remote sites >4.0 cm from the MEA center. These limitations, however, have been partially resolved by combined use of a virtual activation map and contact substrate and/or activation map, and optimal positioning of MEA. In the present study the number of patients with organic heart disease was relatively small, thus further investigation of the feasibility of EA-guided RFA for VT/PVC is required in these patients. Conclusion EA-guided RFA is safe and effective for VT/PVC irrespective of its origin, mechanism, sustainability, hemodynamic condition, and underlying heart disease. Acknowledgements Dr Tsuchiya has served as a speaker and consultant for Nihon Kohden and St Jude Medical. References 1. Marchlinski FE, Callans DJ, Gottlieb CD, Zado E. Linear ablation lesions for control of unmappable ventricular tachycardia in patients with ischemic and nonischemic cardiomyopathy. Circulation 2000; 101: 1288 1296. 2. Arenal A, Glez-Torrecilla E, Ortiz M, Villacastín J, Fdez-Portales J, Sousa E, et al. Ablation of electrograms with an isolated, delayed component as treatment of unmappable monomorphic ventricular tachycardias in patients with structural heart disease. J Am Coll Cardiol 2003; 41: 81 92. 3. Stevenson WG, Soejima K. Catheter ablation for ventricular tachycardia. Circulation 2007; 115: 2750 2760. 4. El-Shalakany A, Hadjis T, Papageorgiou P, Monahan K, Epstein L, Josephson ME. Entrainment/mapping criteria for the prediction of termination of ventricular tachycardia by single radiofrequency lesion in patients with coronary artery disease. Circulation 1999; 99: 2283 2289. 5. Harada T, Stevenson WG, Kocovic DZ, Friedman PL. Catheter ablation of ventricular tachycardia after myocardial infarction: Relation of endocardial sinus rhythm late potentials to the reentry circuit. J Am Coll Cardiol 1997; 30: 1015 1023. 6. Soejima K, Suzuki M, Maisel WH, Brunckhorst CB, Delacretaz E, Blier L, et al. Catheter ablation in patients with multiple and unstable ventricular tachycardias after myocardial infarction: Short ablation lines guided by reentry circuit isthmuses and sinus rhythm mapping. Circulation 2001; 104: 664 669. 7. Schilling RJ, Peters NS, Davies DW. Feasibility of a noncontact catheter for endocardial mapping of human ventricular tachycardia. Circulation 1999; 99: 2543 2552. 8. Rajappan K, Schilling RJ. Non-contact mapping in the treatment of ventricular tachycardia after myocardial infarction. J Interv Card Electrophysiol 2007; 19: 9 18. 9. Narita S, Miyamoto K, Tsuchiya T, Nagamoto Y, Yamaguchi T. Radiofrequency catheter ablation of atrial tachycardia under navigation using EnSite array. Circ J 2010; 74: 59 65. 10. Narita S, Tsuchiya T, Ashikaga K, Ushinohama H, Tanaka Y, Yoshinaga M, et al. Radiofrequency catheter ablation with the use of a noncontact mapping system for ventricular tachycardia originating from the aortic sinus cusp: A case report. J Arrhythmia 2008; 24: 45 50. 11. Klemm HU, Ventura R, Steven D, Johnsen C, Rostock T, Lutomsky B, et al. Catheter ablation of multiple ventricular tachycardias after myocardial infarction guided by combined contact and noncontact mapping. Circulation 2007; 115: 2697 2704. 12. Strickberger SA, Knight BP, Michaud GF, Pelosi F, Morady F. Mapping and ablation of ventricular tachycardia guided by virtual electrograms using a noncontact, computerized mapping system. J Am Coll Cardiol 2000; 35: 414 421. 13. Hiramatsu S, Tada H, Sakamoto Y, Kaseno K, Sato C, Irie T, et al. Quantitative analysis and characteristics of the electrograms recorded within the non-coronary aortic sinus of Valsalva. Circ J 2009; 73: 838 845. 14. Nakano E, Harada T, Ikeda K, Nakazawa K, Wakimoto H, Miyake F, et al. Reentrant ventricular tachycardia originating in the right ventricular outflow tract: Slow conduction identified by right coronary artery ostium pacing. Circ J 2008; 72: 855 860. 15. Yoshida K, Sekiguchi Y, Tanoue K, Endo M, Suzuki A, Kanemoto M, et al. Feasibility of targeting catheter ablation to the markedly low-voltage area surrounding infarct scars in patients with postinfarction ventricular tachycardia. Circ J 2008; 72: 1112 1119. 16. Schilling RJ, Peters NS, Davies DW. Simultaneous endocardial mapping in the human left ventricle using a noncontact catheter: Comparison of contact and reconstructed electrograms during sinus rhythm. Circulation 1998; 98: 887 898. 17. Kadish A, Hauck J, Pederson B, Beatty G, Gornick C. Mapping of atrial activation with a noncontact multielectrode catheter in dogs. Circulation 1999; 99: 1906 1913. 18. Earley MJ, Abrams DJR, Sporton SC, Shilling RJ. Validation of the noncontact mapping system in the left atrium during permanent atrial fibrillation and sinus rhythm. J Am Coll Cardiol 2006; 48: 485 491. 19. Pratola C, Baldo E, Notarstefano P, Tiziano T, Ferrari R. Feasibility of the transseptal approach for fast and unstable left ventricular tachycardia mapping and ablation with a non-contact mapping sys-

MIYAMOTO K et al. tem. J Interv Card Electrophysiol 2006; 16: 111 116. 20. Miyamoto K, Tsuchiya T, Yasuoka C, Tanioka Y. A case of radiofrequency catheter ablation of ventricular tachycardia associated with an old myocardial infarction guided by a noncontact mapping system. J Arrhythmia 2009; 25: 36 41. 21. Daniels DV, Lu YY, Morton JB, Santucci PA, Akar JG, Green A, et al. Idiopathic epicardial left ventricular tachycardia originating remote from the sinus of Valsalva: Electrophysiological characteristics, catheter ablation, and identification from the 12-lead electrocardiogram. Circulation 2006; 113: 1659 1666. 22. MT de Bakker, RNW Hauer, TA Simmers. Activation mapping: Unipolar versus bipolar recording. In: Zipes DP, Jalife J, editors. Cardiac electrophysiology from cell to bedside, 2nd edn. Philadelphia: Saunders, 1068 1078.