Imaging and epicardial substrate ablation of ventricular tachycardia in patients late after myocarditis

Transcription

1 Europace (014) 16, doi: /europace/euu017 CLINICAL RESEARCH Electrophysiology and ablation Imaging and epicardial substrate ablation of ventricular tachycardia in patients late after myocarditis Giuseppe Maccabelli 1 *, Dimitris Tsiachris 1, John Silberbauer 1, Antonio Esposito, Caterina Bisceglia 1, Francesca Baratto 1, Caterina Colantoni, Nicola Trevisi 1, Anna Palmisano, Pasquale Vergara 1, Francesco De Cobelli, Alessandro Del Maschio, and Paolo Della Bella 1 1 Arrhythmia Unit and Electrophysiology Laboratories, San Raffaele University Hospital, Milan, Italy; and Department of Radiology, San Raffaele University Hospital, Milan, Italy Received 15 September 013; accepted after revision 9 January 014; online publish-ahead-of-print 0 February 014 Aims We present clinical, electroanatomical mapping (EAM), imaging, and catheter ablation (CA) strategies in patients with myocarditis-related ventricular tachycardia (VT).... Methods Between January 010 and July 01, 6 consecutive patients underwent imaging-guided CA of myocarditis-related ventricular arrhythmias, 3 of 6 using a combined endo epicardial approach. Segment per segment correspondence of late and results enhanced (LE) scar localization with EAM scar was assessed in all patients with available uni/bipolar maps (n ¼ 19). Induced VTs were targeted prior to substrate modification. Late potentials (LPs) abolition constituted a procedural endpoint independently from VT inducibility. Clinical monomorphic VT was induced in 15 of 6 patients (57.7%) and was associated with epicardial LPs in 10 of 15, completely abolished in 7 of 10 patients. Of the 10 patients rendered non-inducible VTs were ablated epicardially in 7. Late potentials were also detected in 7 of 11 initially non-inducible patients and completely abolished in 4. After a median follow-up of 3 (15 31), 0 of 6 patients (76.9%) remained free from VT recurrence. Bipolar mapping revealed low-voltage scar (,1.5 mv) in 1 patient endocardially and in 14 of 19 epicardially. Unipolar mapping revealed low-voltage scar (,8 mv) in 1 of 19 patients endocardially and in 18 of 19 epicardially. Correspondence of LE scar localization with endocardial bipolar scar was 1%, with endocardial unipolar scar 3.7%, with epicardial bipolar scar 39.8%, and with epicardial unipolar scar 66.%.... Conclusion Pre-procedural scar imaging and EAM findings support the necessity of an epicardial approach in patients with prior myocarditis. Epicardial unipolar mapping (,8 mv) is superior in scar identification and CA based on substrate modification is safe and effective in this setting Keywords Catheter ablation Electroanatomical mapping Late enhancement imaging Myocarditis Ventricular tachycardia Introduction Current guidelines on ventricular arrhythmias and sudden cardiac death recommend drug therapy for life-threatening ventricular arrhythmias in myocarditis and possible implantable cardioverterdefibrillator (ICD) implantation if arrhythmias persist after the acute phase despite pharmacotherapy. 1 Limited data suggest that catheter ablation (CA) of ventricular tachycardia (VT) in patients with chronic active myocarditis is effective. Specific data regarding the association of electroanatomical and imaging characteristics of patients with prior myocarditis are lacking. In the current study, clinical, electroanatomical mapping (EAM), imaging, and CA strategies in a series of patients presenting with VT and a history of myocarditis are presented. * Corresponding author. Arrhythmia Unit and Electrophysiology Laboratories, Ospedale San Raffaele, via Olgettina 60, Milan, Italy.Tel: ; Fax: , maccabelli.giuseppe@hsr.it Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 014. For permissions please journals.permissions@oup.com.

2 1364 G. Maccabelli et al. What s new? The high rate of epicardial ablation in our study population supports a first-line epicardial approach in these patients when performed in a dedicated unit. Epicardial bipolar and endocardial unipolar mapping were able to detect a thicker scar extending from the subepicardium to the mid-wall layers of the left ventricle. Epicardial unipolar mapping is superior in scar identification as it is able to additionally detect the thinner subepicardial scars. Late potentials were present exclusively in the epicardium in 65% of patients. Catheter ablation is a safe and effective procedure in the short term with 76.9% of patients remaining free of ventricular tachycardia recurrences after a median follow-up period of 3. Only 1 of 11 patients in whom complete abolition of late potentials was accomplished, recurred during follow-up. Serial programmed ventricular stimulation allowed us to treat one-third of patients using catheter ablation alone without the need for implantable cardioverter-defibrillator implantation after a follow-up period of 3. Methods Patient population Patients who underwent CA for drug-refractory scar-related VT at the San Raffaele Hospital VT unit, a national referral centre, between January 010 and July 01, constituted the initial study population (n ¼ 353). We selected from this cohort all the patients who were referred with preserved left ventricular (LV) function and a history of myocarditis (n ¼ 6, 7.4%) established in the referring unit. Twenty-four patients presented with sustained ventricular arrhythmias [ with VT degenerating to ventricular fibrillation (VF)] and the remaining patients had frequent ventricular extrasystoles and symptomatic episodes of nonsustained VT. Clinical VT morphology was defined as monomorphic or polymorphic and VT tolerance was based on the presence or not of a haemodynamic compromise. 3 An ICD had been implanted before referral to our centre in eight patients for secondary prevention of sudden cardiac death and of these, three experienced an electrical storm ( 3 episodes of VT separated by.5 min during a 4 h period). We undertook coronary angiography that excluded coronary artery disease in all cases. Pre-ablation echocardiography showed a median LV ejection fraction (LVEF) of 54.5% (range 40 65%). Imaging acquisition and analysis Imaging focused on scar identification was performed prior to the ablation. Cardiac magnetic resonance (CMR) was performed in 19 patients without an ICD and multidetector cardiac computed tomography (CCT) in 7 patients with previously implanted ICDs. In one patient, we performed both CMR and CCT. Cardiac magnetic resonance Cardiac magnetic resonance was performed on a 1.5 T whole-body scanner (Achieva; Philips Medical Systems) using a five-element cardiac phased-array coil (SENSE Cardiac, Philips Medical Systems) (voxel size mm, slice thickness 3.5 mm). Left ventricular volumes and LVEF were calculated by semiautomatic segmentation (ViewForum release 4., Philips Medical System) of a stack of 1 contiguous short-axis cine steady-state free precession slices. T-weighted Short Tau Inversion Recovery (Tw-STIR) images were acquired in the cardiac long- and short axis to detect the areas of myocardial oedema. Late enhancement (LE) short- and long-axis images were acquired min after peripheral bolus injection of gadobutrol (Gd-DO3A-butrol; Gadovist 1.0; Bayer HealthCare Pharmaceuticals) at the dose of 0.15 mmol/kg of body weight, using a three-dimensional TFE-inversion recovery T1-weighted sequences, with inversion time individually optimized to null the signal of normal myocardium. Computed tomography scan Cardiac computed tomography can be alternatively used for scar detection in patients with ICDs. 4 Cardiac computed tomography exams were performed on a 64-slice scanner (Brilliance 64, Philips Medical System) with mm collimation. The CCT protocol included a non-enhanced prospectively electrocardiogram (ECG)-gated scan (Calcium Score) followed by two ECG-gated contrast-enhanced scans: an angiographic 10 kvp scan performed during the first pass of contrast bolus in the LV/ascending aorta (cardiac computed tomography angiography) and a low energy (80 kvp) delayed scan performed 10 min later (CCT-LE). Cardiac computed tomography angiography and CCT-LE scans were prospectively ECG-gated in the case of heart rate 65 b.p.m. and retrospectively ECG-gated, with tube current modulation, in patients with higher heart rates. The LV location (using a 17-segment model) and the pattern (subendocardial, mid-wall, subepicardial, or transmural) of LE were assessed by the consensus of two expert observers blinded to the EAM findings. Electroanatomical mapping Endocardial (3 patients) and first-line endo epicardial (3 patients) EAM during sinus rhythm (SR) was performed using CARTO 3 (Biosense Webster Inc.) in 19 endo-epi patients and NavX Ensite Velocity (St Jude Medical TM Inc.) in 4 endo-epi patients. All procedures were performed under general anaesthesia. Access to the pericardial space was obtained as previously described. 5 A 3.5 mm distal-tip irrigated catheter (Navistar or Celsius, Biosense Webster Inc.) was used as the mapping catheter. A value of 1.5 mv defined normal LV endocardial bipolar electrogram amplitude and a voltage,0.5 mv defined dense scar. Bipolar epicardial voltages 1 and 1.5 mv were consecutively used to define normal LV epicardium. 3,6 A value of 8 mv defined normal LV endocardial and epicardial unipolar electrogram amplitude. Areas with,5 mv defined dense scar and intermediate values border zone. 7,8 Late potentials (LPs) were defined as previously described. 9 In addition to the voltage map, a colour-coded map of the activation delay was made online by manually marking the latest electrogram component of all LPs. The areas of reduced voltage and LPs were measured. A 17-segment model was used for anatomical description of the LV. 10 Ablation procedure We implemented an ablation strategy based on substrate modification, independent of VT non-inducibility, focused on complete abolition of late and fragmented activity. After the achievement of the substrate map, programmed ventricular stimulation (PVS) with up to four extrastimuli from the right ventricular apex and multiple LV sites was performed in the basal state and after isoproterenol infusion in all the cases. The primary procedural endpoint was LP abolition where feasible and the secondary endpoint was ventricular arrhythmia non-inducibility.

3 Imaging and epicardial substrate ablation of VT 1365 The course of the phrenic nerve was mapped with high-output pacing (0 ma). Epicardial ablation was performed.5 mm from the phrenic nerve and coronary arteries. 5 Coronary angiography was performed in all the cases prior to radiofrequency (RF) delivery if we suspected the presence of a coronary artery. Computed tomography image integration was used alternatively in patients with CT. Radiofrequency current was delivered with an irrigated-tip catheter using a power setting of W with a temperature limit of 438C (irrigation flow was set at 17 and 30 ml/min in accordance with power setting). Ventricular tachycardia ablation strategies are outlined: (1) In patients with tolerated or haemodynamically supportable VT, VT was ablated by using activation and entrainment mapping. The ablation continued in SR aiming at the complete abolition of LPs. 9 () In patients with non-inducible or haemodynamically unsupportable VT, the ablation was performed during SR, targeting the LP areas when present. (3) In patients without LPs or inducible VT, ablation was undertaken using pacemapping (assessing for 1 of 1 match with a clinical VT if available). Statistical analysis Categorical variables are presented as absolute frequencies and percentages, while continuous variables are shown as means + standard deviations. Characteristics between groups were compared using Student s t-test and x tests for continuous and categorical variables, respectively. All tests were considered to be significant at the level of P, SPSS statistical package, release 15.0 (SPSS Inc.) was used for all statistical analyses. Results Clinical characteristics Baseline clinical, EAM, procedural, and follow-up data for each patient are depicted in Table 1. Mean age of the 6 patients (3 males) was 47.5 years and the median time from clinical diagnosis of myocarditis to the referral to our Unit for the management of ventricular arrhythmias was 13 (range ). In a previous presentation at the referral centre, biopsy-proven myocarditis was reported in seven cases (in three biopsies, viral detection was confirmed by polymerase chain reaction coxsackievirus B in two patients and adenovirus in the third). In none of these seven patients, giant cell myocarditis was present. All the remaining 19 patients had a previous clinical diagnosis of myocarditis and CMR findings suggestive of myocarditis during the acute phase of the disease. None of the patients had symptoms of heart failure or acute myocarditis and accordingly an indication for endomyocardial biopsy (EMB) at the time of presentation to our Unit. 11 Moreover, the evidence of active myocarditis was not present in patients with available CMR data and none of the patients had clinical evidence of extracardiac sarcoidosis. Of the 4 patients with VT, this was haemodynamically tolerated in 13 of 4 patients. The median number of VT episodes prior to ablation was two. All patients were treated with b-blockers. Class I or III antiarrhythmic agents had previously been administered to all the patients while amiodarone was used only in five patients. Three patients had undergone a previous failed endocardial CA. In three patients, a combined endo epicardial approach was not undertaken because of presentation with VT indicative of septal morphology. Electrophysiological features Clinical monomorphic VT could be induced with PVS in 15 of 6 patients (57.7%). All induced VTs were targeted for ablation. Catheter ablation of the tolerated VTs was performed during VT guided by entrainment and activation mapping in 9 of the 15 patients. In the remaining six patients, pacemapping assisted CA. Thirteen out of these 15 patients underwent a combined endo epi procedure and induced VTs were associated with an area of epicardial late activity in 10 patients. Complete epicardial LP abolition was achieved in seven of them rendering six non-inducible (in the remaining patient VF was induced). In two patients, epicardial LP abolition was incomplete due to the presence of major coronary vessels leading to persistent index VT inducibility. Catheter ablation was also limited by the presence of a coronary vessel and phrenic nerve capture in the LP area in a third case. Since this patient was rendered non-inducible, implementation of an advanced technique for the avoidance of phrenic nerve damage was not considered necessary (Figure 1). In the remaining three inducible patients without epicardial LP activity, ablation was performed endocardially in two of them guided by activation mapping rendering index VT non-inducible in both cases. In the third patient, epicardial ablation guided by VT activation mapping was performed; however, this was limited due to the presence of major coronary vessels and the index VT remained inducible. Regarding the two induced patients who underwent an endocardial procedure, no LPs were identified and the endocardial ablation rendered the index VT non-inducible only in one (Figure 1). Considering the non-inducible cohort at baseline PVS, 10 of 11 underwent a combined endo epi procedure and 1 patient had an endocardial procedure without identification of LPs and was not ablated. Late potentials were detected in seven patients and incomplete abolition was achieved in three due to phrenic nerve capture (coronary vasculature was also limiting in one of these). Regarding the three patients without identifiable LPs, CA was finally not performed in two cases and the third underwent endocardial ventricular extrasystole activation mapping and CA (Figure 1). There were no major procedural complications and one minor complication (femoral pseudoaneurysm) treated conservatively. The mean total RF time in those patients who underwent ablation was min. Of the 18 patients without pre-existing ICD, 8 underwent ICD implantation prior to discharge. The reasons for ICD implantation included VT/VF inducibility in five patients, no ablation in one non-inducible patient without LPs, and the presence of a significant mesocardial scar area in non-inducible patients. At discharge, amiodarone (n ¼ 1) or sotalol (n ¼ 8) was administered in patients with either post-procedural VT/VF inducibility and in those with baseline non-inducibility in whom complete LP abolition was not achieved. Long-term outcome After a median follow-up of 3 (15 31), 0 of 6 patients (76.9%) remained free from VT recurrence, 4 patients had appropriate ICD activations (antitachycardia pacing in 3 cases), 1 patient experienced an electrical storm with amiodarone-related thyrotoxicosis, treated with steroids, while 1 patient required a second

4 Table 1 Baseline clinical, EAM, procedural, and follow-up data individually for each patient Clinical Months since Age Gender EF Imaging Baseline Endo Endo Epi Epi LPs Final VT RF ablation ICD VT in FU FU presentation diagnosis VT BIP scar UNI BIP UNI induction duration induction scar scar scar... 1 NSVT, PVCs M 55 CMR () () (+) (+) (+) Epi na Endo epi () () 36 Tolerated VT M 55 CMR (+) () () (+) (+) Epi (+) Endo epi ICD post NSVT in VT-ATP 3 3 M 60 CT () () () (+) (+) Epi na Endo epi, PN and CV limited ICD pre ATP in 9 and 10 4 VT/VF 1 3 M 65 CMR () () na a (+) na a () na No RF ICD post () 36 5 VT shock 8 53 F 43 CT () () (+) (+) (+) () na No RF ICD pre () 6 6 Tolerated VT M 51 CMR, CT (+) () (+) (+) (+) Epi () Epi, epi VT ablated serial EPS () 36 7 NSVT, PVCs 18 5 M 55 CMR (+) () () (+) (+) Epi (+) Endo epi ICD post () 3 CV limited 8 Tolerated VT 8 57 F 60 CMR () () na a na b na b () na NoRF () 9 Tolerated VT 1 50 M 50 CMR (+) () (+) (+) (+) Epi () Endo epi, epi serial EPS () 31 VT ablated 10 Tolerated VT 4 4 M 65 CMR () () () () (+) Epi na Endo epi serial EPS () 0 11 Tolerated VT 1 3 F 55 CMR (+) () na a na b na b () (+) Endo ICD post Shock Non-tolerated VT M 64 CMR (+) () (+) (+) (+) Epi () Epi, epi VT ablated, CV limited 4 ICD post () 6 13 Non-tolerated 1 53 M 50 No (+) () na a (+) na a Epi () Epi, epi VT ablated ICD pre () 4 VT 14 Tolerated VT 1 33 M 55 CMR (+) () (+) () (+) Epi () Epi, epi VT ablated () () 4 15 ES-ICD M 55 CT () () na a (+) na a Epi na Epi ICD pre () 1 16 Tolerated VT M 55 CMR (+) () na a (+) na a () (+) Endo epi, CV limited ICD post ATP 1 17 Tolerated VT 1 9 M 55 CMR () () () (+) (+) Epi na Epi PN limited serial EPS () 6 18 Tolerated VT 1 7 M 60 CMR () () (+) (+) (+) () na PVC map ICD post serial EPS () 0 19 Shocks s-icd 6 55 M 60 CMR () () () (+) (+) Epi na Epi S-ICD pre () 9 0 ES-ICD M 50 CT () (+) (+) (+) (+) Epi na Epi PN limited ICD pre ES 5 15 Continued G. Maccabelli et al.

5 Imaging and epicardial substrate ablation of VT 1367 ICD pre 6 1 VT shocks 7 56 M 55 CT (+) () (+) na b na b () () Endo, endo VT ablated 6 ICD post ATP F 50 CMR (+) () () (+) () () () Endo, endo VT ablated Non-tolerated VT 3 ES-ICD 1 5 M 55 CT (+) () (+) (+) (+) Epi () Epi, epi VT ablated ICD pre () 15 () 15 () Epi ICD post serial EPS 4 Tolerated VT M 48 CMR (+) () (+) (+) (+) Endo Epi ICD post () 15 (+) EndoPN and CV prevented 5 Tolerated VT M 51 CMR (+) () (+) (+) (+) Endo Epi serial EPS () M 40 CMR (+) () (+) (+) (+) () () Endo epi, epi VT ablated 6 Non-tolerated VT a na, unipolar mapping was not applicable; b na, epicardial mapping was not performed. EF, ejection fraction; VT, ventricular tachycardia; LPs, late potentials; RF, radiofrequency; ICD, implantable cardioverter-defibrillator; NSVT, non-sustained VT; PVCs, premature ventricular contractions; ATP, antitachycardia pacing; ES, electrical storm; CMR, cardiac magnetic resonance; CT, computed tomography; FU, follow-up; PN, phrenic nerve; CV, coronary vessel. endo epi procedure for the recurrence of non-sustained VT episodes (Figure ). Of the 10 patients discharged without ICD implantation, 3 declined serial PVS but remained arrhythmia-free. Of the seven who have undergone serial PVS, two underwent subsequent ICD implantation due to monomorphic VT inducibility (no VT recurrence) and the remaining five had a negative PVS result both at 3 and 1 (Figure ). Substrate analysis A detailed subset analysis focused on segment per segment correspondence of LE scar localization with EAM scar was undertaken in those 19 patients with endocardial and epicardial uni/bipolar voltage maps. Cardiac magnetic resonance imaging was used in 15 patients and multidetector CCT in 4 patients. Myocardial imaging No areas of regional oedema were observed on Tw-STIR images. The median number of LV segments demonstrating scar was four (interquartile range 3 5). As depicted in Figure 3, the majority of patients had scar localizing to the basal-mid inferolateral walls. Scar extension was subepicardial (SS) in 10 of 19 patients but extended to the mid-wall (MS-SS) in 9 and was transmural in. Endocardial electroanatomical mapping Bipolar mapping revealed low-voltage scar in only one segment in one patient. There was no dense scar detected. Unipolar mapping revealed scar (,8 mv) in 1 of 19 patients (63.%). The median number of LV segments demonstrating scar was one (range 1 3). Segment per segment correspondence of LE scar localization with endocardial bipolar mapping was only 1% and with endocardial unipolar map was 3.7% (Figure 3). No patients with endocardial LPs were identified. Epicardial electroanatomical mapping Bipolar mapping with a cut-off value of 1 mv revealed border-zone scar in 14 of 19 patients and in 1 (63.%), there was a correspondence with LE imaging. Based on a cut-off value of 1.5 mv, border-zone scar was revealed in an additional two patients increasing sensitivity to 73.7%. The median number of LV segments demonstrating scar was two (range 1 5). Unipolar mapping revealed scar (,8 mv) corresponding to imaging in 18 of 19 patients (94.7%). The median number of LV segments demonstrating scar was three (range 1 5) and the scar area was cm. Segment per segment correspondence of LE scar localization with epicardial bipolar mapping (cut-off value of 1.5 mv) was 39.8% and with epicardial unipolar mapping was 66.% (Figures 3 and 4). In the two patients with bipolar scar but no correspondence with imaging, there was also no corresponding scar on unipolar imaging and the reduced voltage may have been related to epicardial fat. Epicardial LPs were identified in 15 of 19 (78.9%) patients.

6 1368 G. Maccabelli et al. n = 6 Baseline VT inducibility Inducible patients 5 Non-inducible patients 1 Endo vs. Endo-epi procedure Endo n = Endo epi 3 Endo epi 0 Endo Late potential detection No LP n = No LP n = 3 Epi LP 0 Epi LP n = 7 No LP n = 3 No LP Ablation strategy VT map n = VT map n = 3 Complete n = 7 Limited n = 3 Complete n = 4 Limited n = 3 PVC map No abl n = No abl Post-ablation inducibility No Yes No n = Yes No n = 6 VF No Yes n = ICD implantation Pre = Pre = 1 Post = 1 Post = 1 Post = 1 Post Post = 1 Post = 1 Post = Pre = Pre = EPS = 1 Post EPS = 1 Pre =1 Post = 1 No Figure 1 Flow chart of the study patients according to procedural characteristics. LP, late potential, VT, ventricular tachycardia, VF, ventricular fibrillation, PVC, premature ventricular contraction, ICD, implantable cardioverter-defibrillator. Scar extension and electroanatomical mapping Patients with greater scar extension (MS-SS vs. SS pattern) had more frequently identifiable epicardial bipolar dense scar (70 vs..%, P ¼ 0.037) and epicardial bipolar scar (,1.0 mv) (100 vs. 44.4%, P ¼ 0.006). No significant difference was observed greater in patients with greater scar extension(ms-ss vs. SS pattern) in the identification of epicardial unipolar scar (90 vs. 100%, P ¼ 0.33), although there was a tendency towards significance with endocardial unipolar mapping (80 vs. 44.4%, P ¼ 0.10) (Table ). Unipolar compared with bipolar mapping detected scar more frequently endocardially (44.4 vs. 0%, P ¼ 0.03) and epicardially (,1 mv) (100 vs. 44.4%, P ¼ 0.009) in patients with only SS. In patients with MS-SS pattern, endocardial unipolar compared with bipolar mapping detected a scar more frequently (80 vs. 10%, P ¼ 0.00), while no difference was observed between epicardial unipolar and bipolar mapping. Discussion Myocarditis is an uncommon disease with an estimated incidence ranging from 0.1 to 1% based on autopsy data from young adults who died suddenly. 1,13 The true occurrence of myocarditis is probably underestimated due to its high variability of clinical presentation. Diagnosis is usually based on clinical data, non-invasive diagnostic tools like serum biomarkers, echocardiography, nuclear cardiology, or CMR and, when available, EMB findings. Ventricular arrhythmias, usually common in the acute phase and efficiently treated with antiarrhythmic drugs, may become clinically challenging in the chronic state. Catheter ablation is gaining an increasing role as an effective treatment for ventricular arrhythmias even if scarce data are available for ventricular arrhythmias in the setting of prior myocarditis.,14 The largest published series includes 0 patients who underwent CA with a history of chronic active myocarditis and refractory VT. Our study is the first to assess the impact of CA in patients with prior myocarditis and ventricular arrhythmias using an imagingguided substrate modification strategy. A first-line endo epicardial procedure was undertaken in 3 of 6 patients and 19 were ablated epicardially indicating the importance of this approach in these patients. According to our study results, CA is a safe and effective procedure in the short term with 76.9% of patients remaining free of VT recurrences after a median follow-up period of 3. Serial PVS allowed us to treat one-third of patients using CA alone without the need for ICD implantation.

7 Imaging and epicardial substrate ablation of VT 1369 Median follow-up of 3 in 6 patients ICD implantation ICD pre-ablation = 8 ICD post-ablation = 8 No ICD at discharge = 10 Serial PVS = 7 Negative = Positive = None = 6 None = 4 None = 3 Late ICD = Arrhythmia recurrence NSVT = 1 VT-ATP = 1 VT-ATP = ES = 1 Shock = 1 None = 5 None = Figure Flow chart of the study patients according to outcome results. ICD, implantable cardioverter-defibrillator, VT, ventricular tachycardia, ATP, antitachycardia pacing, ES, electrical storm, PVC, premature ventricular contraction. Endocardial 9 A AS AL Bipolar 7 Unipolar IS IL 8 I Epicardial Figure 3 Central panel shows segmental localization of LE scar. Outer panels depict colour-graded level of agreement of LE scar localization with endo epicardial bipolar/unipolar voltage mapping. (blue, no agreement, yellow, full agreement). AS, anteroseptal, AL, anterolateral, IS, inferoseptal, IL, inferolateral.

8 1370 G. Maccabelli et al. A B C D Basal Mid Apex Figure 4 (Representative figure)(a C) depict epicardial bipolar (cut-off value ¼ 1.5 mv), unipolar, and LP maps and (D) short-axis viewsof CMR LE scar located in the basal-mid inferolateral wall. Epicardial unipolar mapping better identifies the corresponding CMR-detected scar and encloses the LP area that was targeted for ablation. The substrate modification strategy has been fundamental in the management of our study patients. Considering our low inducibility rate (58%), most likely due to the low reproducibility of PVS and to the deep sedation necessary for an epicardial approach, 5 LPs abolition, a consistent endpoint for ablation, 9,15 was a target of all our procedures. Late potentials were present exclusively in the epicardium in 65% of patients, were co-located with the presence of EAM and LE scar, and completely abolished in 11 of 17 patients mostly due to anatomical restrictions. More importantly, only 1 of 11 patients in whom a complete abolition of LPs was accomplished, recurred during follow-up. The absence of LPs in the remaining cases might be attributed to either low-mapping resolution or to the deep substrate. In the cases of an extensive epicardial arrhythmogenic substrate, scar dechannelling could alternatively be used increasing the success rate of the procedure. 16 Magnetic resonance imaging (MRI) and CT imaging, in patients with pre-existing ICDs, was used for pre-procedural scar localization. Prior reports have demonstrated a subepicardial location in the lateral wall with PVB19 infection and septal location with HHV6 and HHV6/PVB19 myocarditis that frequently progresses towards chronic heart failure. 17 Scar was detected using LE imaging in all of our patients (usually in the lateral wall) suffering from ventricular arrhythmias and a prior history of myocarditis. Due to the localization of the arrhythmogenic substrate in the lateral LV wall, ablation was frequently limited by the presence of coronary arteries and the phrenic nerve. Even though techniques for avoiding phrenic nerve damage have been described, the proximity of a coronary artery to the ablation site remains an unresolved issue. 5 In the present study, we assessed the discriminating capacity of both bipolar and unipolar mapping for scar detection. Endocardial unipolar voltage reduction in the absence of endocardial voltage abnormalities is indicative of epicardial and/or MS due to the larger field of view of the unipolar electrogram, making it an indispensable EAM tool in the non-ischaemic cardiomyopathies. 7 Accordingly, the endocardial bipolar had no diagnostic value in this population, while the established epicardial bipolar cut-off values of 1 mv, identified the

9 Imaging and epicardial substrate ablation of VT 1371 Table Prevalence of electroanatomical scar detection according to the extension pattern based on LE imaging Mid-wall Subepicardial P and subepicardial scar scar... Endocardial bipolar scar (1.5 mv), % Endocardial unipolar scar (8 mv), % Epicardial bipolar dense scar (0.5 mv), % Epicardial bipolar scar (1 mv), % Epicardial bipolar scar (1.5 mv), % Epicardial unipolar scar (8 mv), % presence of a scar in only 63% of the patients. Applying a cut-off of 1.5 mv in the epicardium increased the sensitivity by 11%. In contrast, unipolar mapping with a cut-off value of 8 mv identified scar in 95% of patients. Focusing on scar thickness, the epicardial bipolar and endocardial unipolar mapping were able to detect thicker scars extending from the subepicardium to the mid-wall layers of the LV. Epicardial unipolar mapping was able to additionally detect thinner SSs possibly due to the wider field of view compared with the bipolar mapping. 7 The correspondence with scar imaging using a segment-bysegment analysis was 66% with unipolar and 40% using the more sensitive 1.5 mv cut-off value of bipolar mapping. This is in accordance with recently published data by Piers et al., 8 which demonstrated the optimal values of epicardial bipolar and unipolar mapping for scar detection in patients with non-ischaemic cardiomyopathy. A 3D colour-coded shell map of the segmented LV has also been proposed to describe MRI-derived scars and compare them with EAM maps. 18 In the present study, we did not aim to obtain new cut-off values EAM data that correspond to the presence of LE scar, but only to examine the value of the established cut-offs in the setting of prior myocarditis. Clinical implications Complication rates associated with epicardial ablation procedures are acceptably low when undertaken in experienced high-volume centres. 19 The high rate of epicardial ablation in our study population (9 of 6 were ablated endo epicardially and 10 of 6 only epicardially) supports a first-line epicardial approach in these patients when performed in a dedicated VT unit. Presently, there are no recommendations for the post-ablation management of patients with preserved LVEF and structural heart disease with respect to ICD implantation. Furthermore, caution should be exercised when considering ablation curative based on the possibility of future disease progression. Our main criterion for ICD implantation was post-ablation VT inducibility either acutely or during serial PVS (performed at 3 and 1 ). Study limitations The lackof EMB data in most of our patients constitutes a limitation of the present study. However, EMB is only performed in highly specialized centres and requires specialized tests presently unavailable in commercial kits. In addition, considering that the current guidelines recommend EMB only in special clinical presentations, such as acute heart failure, 11 a majority of patients who are admitted in peripheral hospitals receive common medical therapy and are rarely transferred to tertiary centres. As a National Referral Centre for VT, all patients included in this study were referred for arrhythmia recurrence refractory to conventional therapy. In all, a previous diagnosis of myocarditis was made and none had documentation of worsening EF in the before the ablation or clinical or laboratory diagnostic criteria suggestive of an acute myocardial inflammatory process at the time of admission, thus limiting any potential clinical benefit from EMB. We should also acknowledge that substrate modification was finally completely achieved in only 11 patients either due to the absence of LPs or due to difficulties in their abolition. Conclusions Catheter ablation in patients with prior myocarditis and ventricular arrhythmias is safe and effective using a strategy based on imaging-guided substrate modification. This strategy allowed one-third of patients to be treated without an ICD. Pre-procedural scar imaging and EAM findings support the necessity of an epicardial approach as a first-line treatment. Established endocardial and epicardial bipolar cut-offs are not sensitive for scar identification in our patient population and novel approaches such as epicardial unipolar mapping appear superior in this setting. Conflict of interest: P.D.B. is a consultant for St Jude Medical and has received honoraria for lectures from Biosense Webster, St Jude Medical, and Biotronik. All other authors have no conflicts to declare. Funding This research was partially supported by a grant from Italian Ministry of Health: Giovani Ricercatori Ricerca Finalizzata ; project number GR The funders had no role in this study. References 1. Zipes DP, Camm AJ, Borggrefe M, Buxton AE, Chaitman B, Fromer M et al. ACC/ AHA/ESC 006 Guidelines for Management of Patients with Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death: a report of the ACC/AAHA Task Force and the ESC Committee for Practice Guidelines (writing committee to develop Guidelines for Management of Patients with Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death): developed in collaboration with the EHRA and the HRS. Europace 006;8:

10 137 G. Maccabelli et al.. Dello Russo A, Casella M, Pieroni M, Pelargonio G, Bartoletti S, Santangeli P et al. Drug-refractory ventricular tachycardias after myocarditis: endocardial and epicardial radiofrequency catheter ablation. Circ Arrhythm Electrophysiol 01;5: Aliot EM, Stevenson WG, Almendral-GarroteJM, Bogun F, Calkins CH, DelacretazE et al. EHRA/HRS Expert Consensus on Catheter Ablation of Ventricular Arrhythmias Developed in a partnership with the EHRA, a Registered Branch of the ESC, and the HRS; in collaboration with the ACC and the AHA. Europace 009;11: Gerber BL, Belge B, Legros GJ, Lim P, Poncelet A, Pasquet A et al. Characterization of acute and chronic myocardial infarcts by multidetector computed tomography: comparison with contrast-enhanced magnetic resonance. Circulation 006;113: Maccabelli G, Mizuno H, Della Bella P. Epicardial ablation for ventricular tachycardia. Indian Pacing Electrophysiol J 01;1: Cano O, Hutchinson M, Lin D, Garcia F, Zado E, Bala R et al. Electroanatomic substrate and ablation outcome for suspected epicardial ventricular tachycardia in left ventricular nonischemic cardiomyopathy. J Am Coll Cardiol 009;54: Hutchinson MD, Gerstenfeld EP, Desjardins B, Bala R, Riley MP, Garcia FC et al. Endocardial unipolar voltage mapping to detect epicardial ventricular tachycardia substrate inpatients with nonischemic left ventricularcardiomyopathy. Circ Arrhythm Electrophysiol 011;4: PiersSR, van Huls vantaxiscf, TaoQ, vander GeestRJ, Askar SF, Siebelink HM et al. Epicardial substrate mapping for ventricular tachycardia ablation in patients with non-ischaemic cardiomyopathy: a new algorithm to differentiate between scar and viable myocardium developed by simultaneous integration of computed tomography and contrast-enhanced magnetic resonance imaging. Eur Heart J 01;34: Vergara P, Trevisi N, Ricco A, Petracca F, Baratto F, Cireddu M et al. Late potentials abolition as an additional technique for reduction of arrhythmia recurrence in scar related ventricular tachycardia ablation. J Cardiovasc Electrophysiol 01;3: Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK et al. American Heart Association Writing Group on Myocardial Segmentation and Registration for Cardiac Imaging. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation 00;105: CooperLT, Baughman KL, Feldman AM, Frustaci A, Jessup M, KuhlU et al. The roleof endomyocardial biopsy in the management of cardiovascular disease: a scientific statement from the American Heart Association, the American College of Cardiology, and the European Society of Cardiology. Endorsed by the Heart Failure Society of America and the Heart Failure Association of the European Society of Cardiology. Eur Heart J 007;8: Wakafuji S, Okada R. Twenty year autopsystatistics of myocarditis incidence in Japan. Jpn Circ J 1986;50: Gravanis MB, Sternby NH. Incidence of myocarditis. A 10-year autopsy study from Malmo, Sweden. Arch Pathol Lab Med 1991;115: ChauhanVS, HameedullahI, Nanthakumar K, DownarE. Epicardial catheterablation of incessant ventricular tachycardia in giant cell myocarditis. J Cardiovasc Electrophysiol 008;19: Hsia H, Lin D, Sauer W, Callans DJ, Marchlinski FE. Anatomic characterization of endocardial substrate for hemodynamically stable reentrant ventricular tachycardia: identification of endocardial conducting channels. Heart Rhythm 006;3: Berruezo A, Fernández-Armenta J, Mont L, Zeljko H, Andreu D, Herczku C et al. Combined endocardial and epicardial catheterablation in arrhythmogenic right ventricular dysplasia incorporating scar dechanneling technique. Circ Arrhythm Electrophysiol 01;5: Mahrholdt H, Wagner A, Deluigi CC, Kispert E, Hager S, Meinhardt G et al. Presentation, patterns of myocardial damage, and clinical course of viral myocarditis. Circulation 006;114: Fernández-ArmentaJ, BerruezoA, Andreu D, CamaraO, Silva E, Serra L et al. Threedimensional architecture of scar and conducting channels based on high resolution ce-cmr: insights for ventricular tachycardia ablation. Circ Arrhythm Electrophysiol 013;6: Della Bella P, Brugada J, Zeppenfeld K, Merino J, Neuzil P, Maury P et al. Epicardial ablationfor ventriculartachycardia: aeuropeanmulticenter study. CircArrhythm Electrophysiol 011;4:653 9.