Midmyocardial and epicardial ventricular tachycardia. Original Article

Transcription

1 Original Article Characterization of Trans-septal Activation During Septal Pacing Criteria for Identification of Intramural Ventricular Tachycardia Substrate in Nonischemic Cardiomyopathy Brian P. Betensky, MD; Suraj Kapa, MD; Benoit Desjardins, MD, PhD; Fermin C. Garcia, MD; David J. Callans, MD; Sanjay Dixit, MD; David S. Frankel, MD; Mathew D. Hutchinson, MD; Gregory E. Supple, MD; Erica S. Zado, PA-C; Francis E. Marchlinski, MD Background Identification of intramural basal-septal ventricular tachycardia (VT) substrate is challenging in nonischemic cardiomyopathy. We sought to (1) characterize normal/abnormal trans-septal right ventricular (RV) to left ventricular activation; (2) assess the effect of opposite RV pacing on left ventricular septal bipolar electrograms (EGMs); and (3) establish criteria for the identification of intramural septal VT substrate. Methods and Results Endocardial activation mapping and local EGM assessment of the left interventricular septum was performed during RV basal septal pacing in 40 patients undergoing VT ablation with no evidence of septal scar (group 1, n=14) and with septal scar (group 2, n=26) defined by low septal unipolar voltage (<8.3 mv) and delayed enhancement on cardiac MRI with/without abnormal bipolar voltage (<1.5 mv) in sinus rhythm. Left ventricular trans-septal activation time was prolonged in Group 2 compared with Group 1 (55.3±33.0 versus 25.7±8.8 ms; P=0.003). In 6 group 2 patients, left ventricular septal breakthrough was displaced to the scar border. During RV pacing, group 2 had fractionated (8.8%), late (2.8%), and split (5.7%) EGMs not seen in group 1. Trans-septal activation >40 ms (sensitivity 60%, specificity 100%; P<0.001) and EGM duration >95 ms during pacing (sensitivity 22%, specificity 91%; P<0.001) identified septal scar (13/26 pts). Conclusions In patients with nonischemic cardiomyopathy, VT and septal scar, delayed transmural conduction time (>40 ms) and fractionated, late, split, and wide (>95 ms) bipolar EGMs during RV basal pacing identify intramural VT substrate. In select cases, the basal septum appears compartmentalized as the stimulated wavefront is rerouted to the scar border. (Circ Arrhythm Electrophysiol. 2013;6: ) Midmyocardial and epicardial ventricular tachycardia (VT) substrate in nonischemic cardiomyopathy (NICM) often exceeds that on the endocardium. 1 6 The interventricular septum (IVS) may be variably involved, occurring as an isolated abnormality or in conjunction with perivalvular and free wall scar. 7 Although the precise reason for these scar patterns remains unclear, characterization of the arrhythmogenic substrate is both essential for successful ablation and challenging in this setting. A mapping technique that enables one to delineate the boundaries of intramural scar and that unveils midmyocardial zones of slow conduction, typically obscured during sinus rhythm (SR) bipolar mapping by neighboring healthy myocardium, can improve our ability to identify and potentially target critical components of the VT circuit. Clinical Perspective on p 1130 Key Words: mapping tachycardia, ventricular We hypothesized that layered and patchy intramural fibrosis in the IVS region can cause varying degrees of transmural conduction delay and, in extreme cases, result in intramural conduction block. Such a finding could in theory support the concept of myocardial compartmentalization or separation of the ventricular myocardium into 2 electrically distinct chambers. To test this hypothesis, we developed a novel pacing maneuver involving catheters positioned opposite one another along the IVS (right ventricular [RV] and left ventricular [LV] aspect) in patients undergoing VT ablation. The primary objective of this study was to analyze the effect of nonischemic intramural scar on the timing and pattern of trans-septal propagation of a stimulated wavefront. In addition, we sought to determine whether a nonphysiological wavefront unmasks abnormal electrograms (EGMs) consistent with midmyocardial arrhythmogenic substrate not otherwise observed during endocardial SR mapping. Received March 29, 2013; accepted October 23, From the Division of Cardiac Electrophysiology (B.P.B., S.K., F.C.G., D.J.C., S.D., D.S.F., M.D.H., G.E.S., E.S.Z., F.E.M.) and Department of Radiology (B.D.), Hospital of the University of Pennsylvania, Philadelphia. The online-only Data Supplement is available at /CIRCEP /-/DC1. Correspondence to Francis E. Marchlinski, MD, Hospital of the University of Pennsylvania, 3400 Spruce St, Founder 9, Philadelphia, PA francis.marchlinski@uphs.upenn.edu 2013 American Heart Association, Inc. Circ Arrhythm Electrophysiol is available at DOI: /CIRCEP

2 1124 Circ Arrhythm Electrophysiol December 2013 Methods Patient Selection Patients >18 years who underwent endocardial voltage mapping during ventricular premature depolarization (VPD)/VT ablation at the University of Pennsylvania from August 2011 to June 2012 were eligible. Patients with a history of myocardial infarction or significant coronary artery disease, congenital heart disease, primary valvular heart disease, or hypertrophic cardiomyopathy were excluded. Patients were classified into 2 groups: group 1 Controls: normal bipolar and unipolar septal voltage maps (>1 cm from the valvular annulus for unipolar EGMs) and no septal delayed enhancement (DE) on cardiac MRI when performed, and group 2 Septal scar: abnormally low unipolar septal voltage (>1 cm from the valvular annulus) with/without low bipolar voltage and abnormal septal DE on MRI (Figure 1). The study was approved by the University of Pennsylvania Institutional Review Board. All patients provided written informed consent. Medical records were reviewed for baseline characteristics, clinical, and demographic data. Transthoracic echocardiograms performed in proximity to the ablation were used for LV measurements (median 1.0 day before, range 198 days before to 1 day after the procedure). Antiarrhythmic drugs were discontinued a minimum of 3 half-lives before ablation, with the exception of amiodarone in 18 patients (median 2.0 days before, range 1 14 days before the procedure; Table 1). MRI Acquisition and Data Analysis ECG-gated images were acquired on a 1.5-Tesla MRI scanner (Siemens Magnetom Avanto, Erlangen, Germany). Fifteen minutes after administration of 0.20 mmol/kg of IV gadolinium DTPA (diethylenetriaminepentaacetic acid; Magnevist, Berlex Pharmaceuticals, Wayne, NJ), 2-dimensional DE imaging was performed using an inversion-recovery sequence in the LV short and long axes. Inversion time ( ms) was optimized to null the normal myocardium. Nonischemic scar was defined as focal or diffuse areas of DE in a distribution inconsistent with scar attributable to previous infarction. MRIs were reviewed by 2 independent observers blinded to the results of electroanatomic mapping. Discrepancies were resolved by consensus. LV endocardial, epicardial, septum and scar contours were manually traced using 3-dimensional postprocessing software (Osirix, Geneva, Switzerland). Scar distribution was refined using a full width at half-maximum approach. All measurements were performed automatically using customized software on Matlab (Mathworks, Natick, MA). Endocardial, epicardial, and scar 3-dimensional surface contours were processed by the Visualization Toolkit (Kitware, New York) to more precisely correlate scar architecture with electroanatomic maps. Prespecified variables included septal scar volume and thickness, wall thickness, and distance from the endocardium to scar border. For scar distribution, the septum was divided into thirds from base to apex (basal, midwall, apex) and from LV to RV (LV endocardium, intramural, and RV endocardium). Mapping Protocol Catheter mapping was performed in the postabsorptive state under conscious sedation or general anesthesia. A phased-array intracardiac echocardiography catheter (AcuNav, Siemens, Mountain View, CA) was deployed in the RV to assess catheter contact and monitor for complications. Electroanatomic mapping was performed in the baseline rhythm through the CARTO system (Biosense Webster, Diamond Bar, CA) and a 3.5-mm tip open irrigated catheter (NaviStar ThermoCool, Biosense Webster) or a nonirrigated 4-mm catheter (NaviStar, Biosense Webster). Electrograms (EGMs) were displayed at 200 mm/s sweep speed and stored for offline analysis. Standardized filtering was used (Bipolar Hz; Unipolar Hz), except for 2 group 2 patients who required preprocedure modification to improve signal:noise ratio (1 patient bipolar filtered Hz; 1 patient bipolar filtered Hz; and unipolar filtered Hz). Acquired EGMs were reconstructed into 3-dimensional voltage maps and scar surface area measured with the incorporated CARTO software. Mapping density was sufficient to allow complete surface reconstruction. Previously established cutoffs were used to define normal RV/LV bipolar and unipolar voltage. 4,8,9 Pacing Protocol and Activation Mapping A confluent area of septal unipolar EGM abnormality and area of DE on MR defined the area of interest for the pacing protocol. An RV quadripolar catheter was positioned at the center of the unipolar/mr abnormality at the RV basal septum, ensuring stable catheter position and pacing capture. In controls, the RV catheter was standardized to the basal RV septum, 2 to 3 cm apical to the anatomic location of the His-bundle region. The LV mapping catheter was then positioned along the LV side of the IVS directly opposite the RV catheter as seen by orthogonal fluoroscopy, intracardiac echocardiography, and on the CARTO mapping system (3.0 system version 2.3; Figure 2). Bipolar pacing at a fixed cycle length of 600 ms was performed from RV tip to ring just above threshold. When the subject s underlying heart rate was faster than 600 ms, pacing at 500 ms was used. Local activation time during RV pacing was determined by measuring the recorded EGM signal on the distal bipole of the LV mapping Figure 1. Representative substrate maps and cardiac MRIs that defined study groups. Left, Absence of delayed enhancement on MRI, normal bipolar and unipolar septal voltage in a patient with isolated ventricular premature depolarizations ablated in the left coronary cusp. Right, Midmyocardial septal scar by MRI, preserved left ventricular (LV) septal bipolar voltage but extensive low LV unipolar voltage. RAO indicates right anterior oblique projection; and RV, right ventricular.

3 Betensky et al Septal NICM VT Substrate 1125 Table 1. Patient Characteristics Control (n=14) Scar (n=26) P Value Age, y 54.4± ± Females, n LVEF, % 41.6± ± IVS diastolic thickness, cm 1.1± ± Previous ablations, no. of patients Medications, no. of patients β-blockers Amiodarone Mexilitine Digoxin No AADs, no. of patients Underlying rhythm, no. of patients NSR Atrial paced RV-apical paced Biventricular paced Underlying conduction abnormality, no. of patients Right bundle branch block Left bundle branch block Left anterior fascicular block First degree AV delay Complete heart block Baseline intervals, ms QRS duration 120±35 129± PR 190±70 224± AH 99±38 123± HV 55±12 61± ICDs, no. of patients Cardiac MRIs, no. of patients Clinical arrhythmia, no. of patients VPD/sustained VT 11/3 8/ AAD indicates antiarrhythmic drugs; AV, atrioventricular; ICD, implantable cardioverter defibrillator; IVS, interventricular septum; LVEF, left ventricular ejection fraction; NSR, normal sinus rhythm; RV, right ventricular; VPD, ventricular premature depolarization; and VT, ventricular tachycardia. catheter. For convenience, local activation time was annotated from the RV stimulation artifact to the onset of the first high-frequency deflection of the bipolar EGM (Figure 2). Where a clear determination of precise local activation was precluded by very low signal amplitude (<0.2 mv), long duration (>133 ms), and multicomponent EGMs, these points were excluded from activation maps and mapped as location only. For split EGMs, the early, lower frequency deflection was considered far-field and the late higher-frequency deflection was taken to represent local activation. 10 Each point was analyzed by 2 independent reviewers. The earliest activation point was taken to be the LV breakthrough. Multiple breakthrough sites were defined by 2 discrete anatomic sites demonstrating <10 ms difference in time, separated by 2 cm. 11 The LV septal EGM directly opposite the RV pacing site was used to determine the trans-septal activation time. Activation patterns were classified into 2 groups. (1) Direct: LV breakthrough directly opposite the RV pacing site and (2) Disrupted or compartmentalized: LV breakthrough not directly opposite the RV pacing site and separated by 4 cm along the plane of the septum. EGM Analysis Offline, bipolar EGMs recorded from the LV septum during SR and RV pacing were quantitatively and qualitatively assessed. Prespecified variables included amplitude, duration, amplitude/duration ratio, wide, split, late, and fractionated EGMs. EGMs were measured using electronic calipers. Patients with ventricular paced rhythms at baseline (3 group 1, 9 group 2) were excluded from SR analysis but were included in the pacing analysis. Peak-to-peak EGM amplitude was measured automatically within the annotation window. EGM duration was measured from EGM onset of the first deflection to the EGM offset. Comparable septal regions were analyzed during pacing and SR. Previously published EGM definitions were used. (1) normal: >1.5 mv amplitude, <80 ms duration; (2) wide: >80 ms duration; (3) split: EGMs with 2 distinct components separated by >20 ms isoelectric segment between peaks of individual components; (4)isolated late: discrete high-frequency potential occurring after the surface QRS and separated from the initial ventricular EGM (when present) by >20 ms; and (5) fractionated: multicomponent signals <0.5 mv amplitude, 133 ms duration. 3,12 14 Late EGMs consistent with retrograde His or far-field atrial EGMs were excluded from our analysis. Statistical Analysis Continuous variables are reported as mean±sd. Categorical data are presented as number of cases or percentages. The Kolmogorov Smirnov test was used to test for normality. For comparison of continuous variables between 2 groups, an unpaired Student t test was performed for normally distributed variables and a Mann Whitney U test for non-normal data. Categorical variables were analyzed using Fisher exact or χ 2 tests. For EGM analysis, a linear mixed effects model was used to account for repeated measurements performed on each subject. The Sidak method was used to correct for multiple testing for pairwise comparisons made between groups. Weighting to the relative number of points obtained within individual subjects was performed. For analysis of data with unequal sample sizes or unequal variances between groups, a Welch t test was used. A prespecified subgroup analysis was performed by dividing the sample into 3 groups: (1) controls (n=14); (2) unipolar abnormality suggesting midmyocardial scar but normal bipolar voltage opposite the pacing site (bipolar voltage [BP] /unipolar voltage [UNI]+, n=11); and (3) abnormal UNI and BP opposite the pacing site (BP+/ UNI+, n=15). ANOVA was used to assess between group differences with a post hoc Bonferroni correction. Pearson correlations were estimated between selected continuous variables. A P 0.05 was considered statistically significant. Statistics were performed on SPSSv16.0 statistical software (SPSS, Chicago, IL). Results Ninety-four patients of 234 patients undergoing VPD/VT ablation between August 18, 2011, and June 31, 2012, met eligibility criteria. Of those, 40 patients were recruited. Twenty-six patients had an NICM with a septal voltage abnormality, comprising group 2. Six other patients with NICM and no septal scar were included in the control arm (group 1). The remaining group 1 patients included 8 patients with structurally normal hearts. Of the 54 patients who met eligibility criteria but who were not included in this article, 3 patients had the pacing protocol performed incorrectly at the apical aspect of the septum, and they were thus eliminated from analysis. One patient with RV cardiomyopathy was studied but removed from analysis because of the unique pathophysiology of scar formation in this disease. The remaining 50 patients either declined to participate or the operating electrophysiologist was reluctant to perform the detailed protocol. Of the 140 patients excluded, 62 patients had coronary artery disease, 37 patients had RV mapping only, 27 patients underwent LV activation mapping

4 1126 Circ Arrhythm Electrophysiol December 2013 Figure 2. Catheter configuration during pacing maneuver: kissing catheters. Top left, Fluoroscopy images. Bottom left, Catheter positions as seen in the electroanatomic mapping system. Right, Representative pacemap in a control subject. Red star, Right ventricular (RV) catheter used for pacing protocol. Red arrow, Annotation of local left ventricular septal activation. LAO indicates left anterior oblique projection; and RAO, right anterior oblique. without detailed voltage mapping, 12 patients had a diagnosis of RV cardiomyopathy, 1 patient had LV noncompaction, and 1 patient had Tetralogy of Fallot. Electroanatomic Mapping and Transmural Activation Results of voltage and activation mapping are listed in Table 2. Transmural conduction in group 1 was uniformly rapid (25.7±8.8 ms), with a direct transmural activation pattern and LV breakthrough opposite the RV pacing stimulus. In group 2, earliest LV breakthrough time (42.9±18.1 ms; P=0.002) and LV activation time directly opposite the RV pacing stimulus were prolonged (55.3±33.9 ms; P=0.003). Directly opposite LV septal activation time >40 ms (16/26 group 2 pts) was 100% specific for intramural septal scar (sensitivity, 60%; area under the curve=0.85; confidence interval, ; P<0.001; Figure 3). In a subgroup analysis of patients with/without IVS bipolar scar, transmural activation times were longest in BP+/UNI+, intermediate in BP /UNI+, and shortest in controls (70.9±36.2 versus 34.2±13.4 versus 25.7±8.8 ms, ANOVA; P<0.001). We observed a compartmentalized septal activation pattern in 6 out of 26 (23%) group 2 patients. In 5 patients, breakthrough occurred at the apical edge of the scar border (Figure 4). One of these patients had a second breakthrough site at the inferoseptal scar border with wavefronts coalescing at the basal septum. In 1 patient, simultaneous breakthrough occurred at the inferoseptal and anteroseptal borders, which merged before spreading toward the apical septum. Patients with disrupted pattern of trans-septal activation possessed larger MRI-defined septal scar volume (17.5±5.0 versus 9.5±5.5 cm 3 ; P=0.02) and thicker intramural scar (5.3±1.2 versus 4.1±0.6 mm; P=0.02) than those with delayed but direct transmural activation during pacing. RV voltage mapping was performed in 15 patients (6 group 1, 9 group 2). In the group 2 patients, low RV septal bipolar voltage was present in 5 out of 9 patients. Low RV septal unipolar abnormalities were observed in all 9 patients (Table 2). RV bipolar (P=0.02) and unipolar (P=0.002) voltage at the pacing site was significantly lower in group 2 than group 1. RV bipolar voltage at the pacing site did not correlate with trans-septal conduction time (r 2 = 0.31; P=0.24). RV unipolar voltage at the pacing site was inversely correlated with trans-septal conduction time (r 2 = 0.54; P=0.03). In the LV, 15 out of 26 group 2 patients had areas of low septal bipolar voltage, and all 26 patients had regions of low septal unipolar voltage (Table 2). LV bipolar (P=0.02) and unipolar voltage (P=0.002) opposite the RV pacing site was significantly lower in group 2 than in group 1. LV activation time directly Table 2. Voltage Mapping and Activation Data Control Scar P Value LV ENDO septal bipolar scar, cm ±14.5 n/a LV ENDO septal unipolar scar, cm ±28.6 n/a RV ENDO septal bipolar scar, cm ±11.2 n/a RV ENDO septal unipolar scar, cm ±21.9 n/a RV pacing site bipolar voltage, mv 7.2± ± RV pacing site unipolar voltage, mv 11.5± ± LV bipolar voltage opposite 6.8± ± pacing site, mv LV unipolar voltage opposite 12.4± ± pacing site, mv Sampled area, cm ± ± Points acquired during pacing 33.1± ± LV breakthrough time, ms 25.7± ± Directly opposite activation time, ms 25.7± ± LV breakthrough sites/pt (n) 1.0± ± Transmural activation pattern Direct/disrupted (n) 14/0 20/ ENDO indicates endocardial; EPI, epicardial; LV, left ventricle; n/a, not applicable; RV, right ventricle; and SA, surface area.

5 Betensky et al Septal NICM VT Substrate 1127 Figure 3. Comparison of transmural activation times during right opposite septal pacing. Left, Scatter plot of transmural activation times. Black diamonds, Cases of direct transmural activation. Asterisks, Cases with disrupted transmural activation. Cross bars, Group means. Right, Receiver-operating characteristic curve demonstrating an optimal threshold of 40 ms. AUC indicates area under the curve. opposite the RV pacing site inversely correlated with SR unipolar voltage at that site (r 2 = 0.56; P<0.001), but only weakly correlated with bipolar voltage (r 2 = 0.38; P=0.02). Additional Subgroup Analyses There was no significant difference in trans-septal time between patients on or off antiarrhythmic drugs in group 1 or group 2 patients (P=0.86 and P=0.73, respectively). Trans-septal conduction time and IVS diastolic wall thickness were not significantly correlated (r 2 =0.12; P=0.49). A subgroup analysis of patients with MRIs of sufficient quality for quantitative analysis (7 group 1, 15 group 2) revealed prolonged trans-septal conduction time in patients with MRI-defined scar compared with those without MRI-defined scar (59.7±41 versus 25.9±7.4 ms; P=0.05). In a subgroup analysis of cases with an RV pacing site 2.0 to 3.5 cm from the His-bundle recording (14 group, 17 group 2), there was a significantly increased trans-septal conduction time in group 2 versus group 1 (61.5±39.5 versus 25.7±8.8 ms; P=0.002). Finally, when analyzing patients undergoing their first procedure (10 group 1, 12 group 2), there was a significantly longer trans-septal time in group 2 compared with group 1 (44.6±14.9 versus 24.5±10.2 ms; P=0.002). MRI Quantification of Scar Cardiac MRIs from 15 group 2 patients with high quality scans were available for quantitative analysis (Table 3). Notably, intramural scar was more common than RV or LV endocardial scar, and in the long axis, the basal septum was the most commonly affected location. No patients in group 2 had MRI-defined scar that extended from base to apex, although 5 apices were excluded because of implantable cardioverter defibrillator lead artifact. Patients with a disrupted trans-septal pattern were more likely to exhibit scar extending from the base to mid-lv on MRI (Figure 5). EGM Analysis In group 1, there were no significant differences in EGM amplitude (P=0.82), duration (P=0.07), or amplitude/duration Figure 4. Disrupted transmural activation during opposite pacing. Left, Control subjects uniformly manifested rapid (<40 ms) left ventricular (LV) septal breakthrough directly opposite the pacing site. Right, A subset of patients with extensive nonischemic scarring manifested delayed trans-septal conduction and altered pattern. In the illustrated example, pacing revealed effective myocardial compartmentalization, as activation extended to the scar border rather than directly transmurally. White dashed line, Inferred path of impulse propagation. ACT indicates activation map; BP, bipolar voltage; NICM, nonischemic cardiomyopathy; RV, right ventricular; and UNI, unipolar voltage.

6 1128 Circ Arrhythm Electrophysiol December 2013 Table 3. Quantification of Septal Nonischemic Scar by Cardiac MRI Scar Parameter Value Septal scar volume, cm ±6.5 ( ) Septal scar LV ENDO surface, cm ±10.1 ( ) Mean septal wall thickness, mm 11.8±1.7 ( ) Mean septal scar thickness, mm 4.5±1.0 ( ) Mean septal scar thickness (% wall) 38±7 (29 51) Maximum septal scar thickness, mm 8.2±2.3 ( ) Distance from LV ENDO surface to septal scar, mm 4.4±1.2 ( ) Distance from RV ENDO surface to septal scar, mm 2.9±0.7 ( ) Scar distribution across the septal wall thickness % of LV ENDO third 9±7 (0 23) % of midwall third 63±11 (38 81) % of RV ENDO third 28±12 (13 62) Scar distribution Basal septum, no. of patients 14/15 (93%) Midchamber septum, no. of patients 10/15 (67%) Apical septum, no. of patients 2/10* (20%) Transmural, no. of patients 4/15 (27%) Mean±SDs (limits). ENDO indicates endocardial; ICD, implantable cardioverter defibrillator; LV, left ventricle; and RV, right ventricle. *Five apices not visualized because of ICD artifact. ratios (P=0.92) during pacing (n=309) compared with SR (n=171; Table I in the online-only Data Supplement). There was a significant increase in wide EGMs (6% versus 16%; P=0.007) during pacing but no fractionated, late, or split EGMs were recorded (Figure IA in the online-only Data Supplement). Within group 2, there was no significant difference in EGM amplitude during pacing (n=910) compared with SR (n=418), but there was a significant increase in EGM duration (64.2±19.3 versus 82.8±36.6 ms; P=0.004) and a decrease in amplitude/duration ratio (0.08±0.06 versus 0.05±0.05 mv/ ms; P=0.01) during pacing compared with SR (Table I in the online-only Data Supplement). Bipolar EGM duration >95 ms during pacing had high specificity for septal scar (sensitivity, 22%; specificity, 91%; area under the curve=0.63; confidence interval, ; P<0.001). In group 2, there was a significant increase in fractionated (0.9% versus 8.8%; P=0.003) and split EGMs (0.2% versus 5.7%; P<0.001) during pacing compared with SR (Figure IA in the online-only Data Supplement). On subgroup analysis (Figure IB in the online-only Data Supplement), even in the absence of bipolar defined scar (BP /UNI+), there was a significant increase in fractionated (0% versus 2.3%; P=0.04), late (0% versus 3.1%; P=0.02), and wide EGMs (16.5% versus 22.8%; P=0.01) recorded during pacing compared with SR. During pacing, there was also an increase in fractionated (0% versus 2.3%; P=0.003), late (0% versus 3.1%; P=0.001), and split (0% versus 1.2%; P=0.001) EGMs in patients with isolated intramural scar (BP /UNI+) compared with controls (BP /UNI ; Figure IB in the online-only Data Supplement). Clinical Arrhythmias In group 1, 8 out of 14 patients had isolated VPDs targeted (Table 1). Three patients had sustained VT, and 3 patients had no spontaneous or inducible clinical VPDs/VT. The basal septum was not targeted in group 1. In group 2, the clinical arrhythmia was more commonly sustained VT rather than isolated VPDs (P=0.007; Table 1), and there were 86 induced or spontaneous VPDs/VTs identified. Fifty-two VTs were unmappable because of hemodynamic intolerance. Of the 34 VPDs/VTs that were mapped using conventional techniques, 17 (50%) VPD/VTs were localized to the IVS. Of those 17 Figure 5. MRI 3-dimensional (3-D) septal scar reconstruction. Two patients (A and B) who demonstrated disrupted transseptal activation are shown with earliest left ventricular activation at the apical region. Left, Short-axis MRIs showing intramural midwall delayed enhancement of the interventricular septum (red arrows). Right, Three-dimensional reconstruction of MRI-defined intramural scar and registration with corresponding CARTO activation maps during basal right ventricular (RV) septal pacing. A simulated RV pacing catheter is shown to indicate typical catheter position. LAT indicates local activation time.

7 Betensky et al Septal NICM VT Substrate 1129 septal VPD/VTs, 15 (88%) were targeted at the basal septum and the remainder from a more apical LV septal site. Clinical VPD/VTs arising from the IVS were more common in patients with delayed trans-septal conduction time during pacing (13 out of 16 versus 4 out of 10; P=0.009). Discussion We present activation and EGM criteria for identifying and characterizing intramural nonischemic VT substrate in the IVS region. During basal RV septal pacing, our findings demonstrate that (1) LV trans-septal activation time can be delayed in patients with septal nonischemic scar when compared with patients without septal scar; (2) in some cases with extensive septal scar extending from the base to midwall, trans-septal activation patterns can be disrupted, appearing effectively compartmentalized, with the stimulated wavefront activating the basal LV septal endocardium from the scar border rather than breaking through directly opposite the RV pacing site; and (3) trans-septal pacing unmasks abnormal LV septal endocardial EGMs even when bipolar EGMs in SR are typically normal. These observations have important implications for catheter ablation, which (1) improve our understanding of the electrophysiological properties/behavior of intramural scar, (2) provide a conceptual framework for VT circuits within the IVS, (3) help to define criteria that identify midseptal scar, which may require more aggressive radiofrequency delivery to effectively target, and (4) complement MRI and unipolar EGM data. This easily performed mapping technique may also be useful when MRI is unavailable. Transmural Time/Patterns Transmural activation has been investigated in patients with coronary disease; however, little is known about transmural activation in NICM. 15 Vassallo et al 11 found a transmural time of 26±14 ms during RV pacing in healthy controls similar to our results. Contrary to the shorter LV breakthrough time in patients with anterior infarction, we found trans-septal conduction was delayed in patients with NICM. More recently, in arrhythmogenic RV cardiomyopathy, intramural/subepicardial fibrosis was shown to produce disruption of direct transmural activation in the RV free wall and result in independent layered sinus activation of the epicardium from the scar border. 16 Our observations may represent an analogous finding in the IVS, such that the left IVS is electrically uncoupled or dissociated from the right IVS in selected cases. We speculate that the delayed transmural activation in the present study may be explained by (1) replacement of the extracellular matrix by interstitial fibrosis altering wavefront direction or slowing propagation; (2) myocyte disarray or changes in myocyte dimensions; and (3) changes at the cellular or molecular level Correlation of MRI and activation data indicates that patients with disrupted activation patterns had larger scar volume and thickness on DE-MRI than those with direct transseptal activation patterns. Results from our subgroup analysis, with/without low bipolar voltage, suggest that while the presence of LV endocardial scar extension prolongs transmural conduction time and increases the proportion of abnormal EGMs during pacing, this phenomenon also occurs in those with isolated unipolar abnormalities. Cardiac MRI analysis revealed that septal scar distribution in this sample was predominantly intramural, with only a small rim of preserved RV/LV endocardium. Such an unusual pattern of scar distribution can result in normal bipolar EGMs in SR and failure to identify intramural arrhythmogenic substrate. It is in this setting that our pacing technique may be particularly useful for identifying midseptal scar and, in the future, potentially help to define substrate-based ablation targets in cases with hemodynamically untolerated septal VT. Electrogram Interpretation During Pacing A pacing wavefront directly adjacent to a suspected intramural/ transmural abnormality unmasked slow conduction obscured during SR bipolar voltage mapping. These findings are concordant with observations made in ischemic cardiomyopathy. 20 Not surprisingly, there was an increase in the number of wide EGMs in the control group during pacing compared with SR, likely because of the paced wavefront spreading obliquely via cell-to-cell conduction across the IVS. However, the emergence of abnormal late and split EGMs in the scar group likely reflects a more profound slowing and conduction block. Limitations This study is limited by small sample size and the opportunity for a type I error may have been inflated because of the number of statistical comparisons performed. Selection and referral bias may have been introduced by performing this study at a single center. This study constitutes a derivation cohort only, and further research is warranted to validate the diagnostic thresholds identified in receiver-operating characteristic analyses. Although activation times on the LV septum were used to infer transmural conduction during RV pacing, a lack of intramural electrodes precluded a more definitive determination of intramural paths of conduction. Twelve patients had underlying ventricular paced rhythms, limiting the number of SR EGMs that could be interpreted. Investigation of bidirectional trans-septal conduction with stable LV septal pacing was not pursued, primarily for reasons of patient safety. This study was not designed to evaluate the effect of cycle length, coupling interval and current strength dependent delays. Instead, we used hemodynamically tolerated, slow continuous, and low output RV pacing in an effort to establish reference values. In patients who underwent previous catheter ablation, it is possible that previous lesion sets altered MRI findings, low voltage areas, and the number of abnormal EGMs recorded during SR and pacing. However, the number of patients with previous ablations was not significantly different between the groups, and the septum was not routinely targeted during previous procedures. Although a statistically significant increase in abnormal EGMs was identified during RV opposite pacing, the overall prevalence of abnormal EGMs was low. Use of the criteria described in this article may therefore require meticulous signal analysis to recognize uncommon but dramatic electrophysiological phenomena. Finally, our study was not designed to correlate EGM criteria with the likelihood of successful ablation of VT arising from the IVS. Conclusions Delayed transmural activation time (>40 ms), disrupted transmural activation patterns, long duration (>95 ms) electrograms

8 1130 Circ Arrhythm Electrophysiol December 2013 and manifest fractionated/late/split EGMs during pacing adjacent to a midmyocardial septal abnormality identify intramural nonischemic VT substrate that may not be apparent on SR bipolar voltage mapping. Use of the septal activation mapping technique described in our study may demonstrate disrupted transmural activation or effective myocardial compartmentalization in a subset of patients with extensive nonischemic septal VT substrate. Acknowledgments This work was performed during Dr Brian Betensky s Research Fellowship with the Heart Rhythm Society. Sources of Funding This work was supported by the F. Harlan Batrus Research Fund and the Murray and Susan Bloom Research Fund. Disclosures Drs Marchlinski and Callans received research grant support and honoraria from Biosense Webster on topics unrelated to the content of this study. The other authors report no conflicts. References 1. Hsia HH, Callans DJ, Marchlinski FE. Characterization of endocardial electrophysiological substrate in patients with nonischemic cardiomyopathy and monomorphic ventricular tachycardia. Circulation. 2003;108: Marchlinski FE. Perivalvular fibrosis and monomorphic ventricular tachycardia: toward a unifying hypothesis in nonischemic cardiomyopathy. Circulation. 2007;116: Cano O, Hutchinson M, Lin D, Garcia F, Zado E, Bala R, Riley M, Cooper J, Dixit S, Gerstenfeld E, Callans D, Marchlinski FE. Electroanatomic substrate and ablation outcome for suspected epicardial ventricular tachycardia in left ventricular nonischemic cardiomyopathy. J Am Coll Cardiol. 2009;54: Hutchinson MD, Gerstenfeld EP, Desjardins B, Bala R, Riley MP, Garcia FC, Dixit S, Lin D, Tzou WS, Cooper JM, Verdino RJ, Callans DJ, Marchlinski FE. Endocardial unipolar voltage mapping to detect epicardial ventricular tachycardia substrate in patients with nonischemic left ventricular cardiomyopathy. Circ Arrhythm Electrophysiol. 2011;4: De Cobelli F, Pieroni M, Esposito A, Chimenti C, Belloni E, Mellone R, Canu T, Perseghin G, Gaudio C, Maseri A, Frustaci A, Del Maschio A. Delayed gadolinium-enhanced cardiac magnetic resonance in patients with chronic myocarditis presenting with heart failure or recurrent arrhythmias. J Am Coll Cardiol. 2006;47: McCrohon JA, Moon JC, Prasad SK, McKenna WJ, Lorenz CH, Coats AJ, Pennell DJ. Differentiation of heart failure related to dilated cardiomyopathy and coronary artery disease using gadolinium-enhanced cardiovascular magnetic resonance. Circulation. 2003;108: Haqqani HM, Tschabrunn CM, Tzou WS, Dixit S, Cooper JM, Riley MP, Lin D, Hutchinson MD, Garcia FC, Bala R, Verdino RJ, Callans DJ, Gerstenfeld EP, Zado ES, Marchlinski FE. Isolated septal substrate for ventricular tachycardia in nonischemic dilated cardiomyopathy: incidence, characterization, and implications. Heart Rhythm. 2011;8: 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: Polin GM, Haqqani H, Tzou W, Hutchinson MD, Garcia FC, Callans DJ, Zado ES, Marchlinski FE. Endocardial unipolar voltage mapping to identify epicardial substrate in arrhythmogenic right ventricular cardiomyopathy/dysplasia. Heart Rhythm. 2011;8: Tung S, Soejima K, Maisel WH, Suzuki M, Epstein L, Stevenson WG. Recognition of far-field electrograms during entrainment mapping of ventricular tachycardia. J Am Coll Cardiol. 2003;42: Vassallo JA, Cassidy DM, Miller JM, Buxton AE, Marchlinski FE, Josephson ME. Left ventricular endocardial activation during right ventricular pacing: effect of underlying heart disease. J Am Coll Cardiol. 1986;7: Cassidy DM, Vassallo JA, Marchlinski FE, Buxton AE, Untereker WJ, Josephson ME. Endocardial mapping in humans in sinus rhythm with normal left ventricles: activation patterns and characteristics of electrograms. Circulation. 1984;70: Cassidy DM, Vassallo JA, Buxton AE, Doherty JU, Marchlinski FE, Josephson ME. The value of catheter mapping during sinus rhythm to localize site of origin of ventricular tachycardia. Circulation. 1984;69: Bogun F, Good E, Reich S, Elmouchi D, Igic P, Lemola K, Tschopp D, Jongnarangsin K, Oral H, Chugh A, Pelosi F, Morady F. Isolated potentials during sinus rhythm and pace-mapping within scars as guides for ablation of post-infarction ventricular tachycardia. J Am Coll Cardiol. 2006;47: Taggart P, Sutton PM, Opthof T, Coronel R, Trimlett R, Pugsley W, Kallis P. Inhomogeneous transmural conduction during early ischaemia in patients with coronary artery disease. J Mol Cell Cardiol. 2000;32: Haqqani HM, Tschabrunn CM, Betensky BP, Lavi N, Tzou WS, Zado ES, Marchlinski FE. Layered activation of epicardial scar in arrhythmogenic right ventricular dysplasia: possible substrate for confined epicardial circuits. Circ Arrhythm Electrophysiol. 2012;5: Roberts WC, Siegel RJ, McManus BM. Idiopathic dilated cardiomyopathy: analysis of 152 necropsy patients. Am J Cardiol. 1987;60: Spach MS, Dolber PC. Relating extracellular potentials and their derivatives to anisotropic propagation at a microscopic level in human cardiac muscle. Evidence for electrical uncoupling of side-to-side fiber connections with increasing age. Circ Res. 1986;58: de Bakker JM, van Rijen HM. Continuous and discontinuous propagation in heart muscle. J Cardiovasc Electrophysiol. 2006;17: Arenal A, Glez-Torrecilla E, Ortiz M, Villacastín J, Fdez-Portales J, Sousa E, del Castillo S, Perez de Isla L, Jimenez J, Almendral J. 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: CLINICAL PERSPECTIVE Characterization of the arrhythmogenic substrate is challenging in patients with nonischemic cardiomyopathy undergoing ventricular tachycardia (VT) ablation. Isolated midmyocardial VT scar can be obscured by more healthy neighboring myocardium, resulting in normal bipolar electrograms during sinus mapping. We sought to establish activation and left ventricular septal bipolar electrogram criteria that identify intramural VT substrate using a novel opposite pacing maneuver. We hypothesized that intramural scarring can result in delayed trans-septal activation and in selected patients act as an effective barrier with myocardial compartmentalization occurring between the right and left aspects of the interventricular septum. In patients without septal scar, trans-septal conduction during right ventricular basal septal pacing was rapid, with left ventricular septal breakthrough directly opposite the pacing site. In contrast, septal scar was associated with prolonged transmural activation (>40 ms) in 62% of patients and disrupted or compartmentalized activation patterns in 23% of patients. Cardiac MRI scar reconstructions merged with electroanatomic maps were consistent with rerouting of activation wavefronts to the scar border in some cases. Manifest fractionated/split/late and wide electrograms >95 ms, which were unmasked by pacing, also helped to identify the location of septal VT substrate. This technique may be useful in patients requiring substrate-based VT ablation, in whom sinus rhythm endocardial bipolar recordings fail to identify the substrate location.