Characterization of Endocardial Electrophysiological Substrate in Patients With Nonischemic Cardiomyopathy and Monomorphic Ventricular Tachycardia
Background— Although catheter mapping has been used to define the endocardial electrogram characteristics in patients with ventricular tachycardia (VT) and coronary disease, characterization of the electrophysiological substrate in patients with VT and nonischemic cardiomyopathy is limited.
Methods and Results— Left ventricular endocardial electroanatomical mapping was performed in 19 patients with nonischemic cardiomyopathy and monomorphic VT with an average of 178±83 sites per chamber mapped. Abnormal bipolar electrogram was defined as endocardial voltage signal amplitude of <1.8 mV. The extent and location of abnormal endocardium was estimated by measuring areas of abnormal electrogram recordings from 3D voltage maps. The origin of VT was approximated by identifying sites of entrainment with concealed fusion or early presystolic activity and/or by pace mapping. Abnormal electrograms were recorded over a 41±28 cm2 area that represented 20±12% of total endocardial surface. The majority of patients (14/19 patients) had only a modest area (<25%) of endocardial abnormality. All patients had abnormal low-voltage endocardial areas located near the ventricular base in the perivalvular region. There were 3±1 VT morphologies per patient. The majority (88%) of the 57 mapped VTs originated from the ventricular base, corresponding to regions with abnormal endocardial electrograms.
Conclusions— Electroanatomical mapping in patients with monomorphic VT and nonischemic cardiomyopathy typically demonstrates a modest-sized basal area of endocardial electrogram abnormalities. The VT site of origin corresponds to these basal electrogram abnormalities. These findings have important implications regarding strategies for VT ablation in this setting.
Received March 21, 2002; de novo received December 31, 2002; revision received May 15, 2003; accepted May 16, 2003.
Much of what is known about the substrate for sustained ventricular tachycardia (VT) is based on studies of patients with coronary artery disease and animal models of myocardial infarction.1–3 Results of prior mapping studies indicated that reentrant VT in this setting often originate from the subendocardial surface of infarcted myocardium, adjacent to dense scar or aneurysm.3–5 Nonischemic cardiomyopathy is characterized by myocardial dysfunction in the absence of significant coronary artery disease. In contrast to ischemic heart disease, little is known about the electrophysiological substrate for uniform VT in patients with nonischemic left ventricular (LV) cardiomyopathy. Earlier studies suffered from limited sampling and provided only limited information.6,7
Magnetic electroanatomical mapping couples high-density electrogram recordings to a spatial display of anatomy.8 The purpose of this study was (1) to characterize the endocardial substrate in patients with nonischemic LV cardiomyopathy and monomorphic VT using catheter based voltage electroanatomical mapping, and (2) to correlate the endocardial abnormalities to the site of origin of VT.
Nineteen consecutive patients with the diagnosis of nonischemic LV cardiomyopathy were included in the study. All patients were referred to the Hospital of University of Pennsylvania for electrophysiology evaluation and catheter ablation. The diagnosis of nonischemic cardiomyopathy was established by the absence of significant (>75% stenosis) coronary artery disease, prior myocardial infarction, or primary valvular abnormalities. Patients who had sarcoidosis identified from pulmonary involvement or myocardial biopsy were excluded. Sixteen patients had implanted cardioverter-defibrillators (ICD) for prior cardiac arrest, sustained VT, or syncope. All patients presented with spontaneous monomorphic VT that was documented by either electrocardiography or stored electrogram data from the ICD (Table 1). All procedures were performed following the institutional guidelines of the University of Pennsylvania Health System.
All patients underwent programmed stimulation at right and left ventricular endocardial sites. The stimulation protocol included delivery of up to triple extrastimuli from multiple ventricular sites at multiple drive cycle lengths without isoproterenol infusion.9
All patients underwent magnetic electroanatomical voltage mapping as previously described.8,10 Access to the LV was via a retrograde, transaortic approach. Bipolar LV endocardial electrograms were recorded from the Navistar catheter that consisted of a 4-mm distal tip electrode and a 2-mm ring electrode with an interelectrode distance of 1 mm. The bipolar signals were filtered at 10 to 400 Hz and were displayed at 100 mm/sec speeds on the CARTO (Biosense, Inc) system. The peak-to-peak signal amplitude of the bipolar electrogram was measured automatically.
Magnetic electroanatomical mapping was performed during sinus rhythm in 11 patients, during right ventricular paced rhythm in 6 patients, and during hemodynamically tolerated monomorphic VT in 2 patients (Table 2). A three-dimensional anatomical shell of the cardiac chamber was constructed and the electrogram signals were coupled and displayed as color gradients on a voltage map.8 Valvular locations were tagged and excluded from the analysis. Valvular sites were identified by fluoroscopic catheter tip positions at the ventricular base, with simultaneous bipolar recordings that demonstrated both atrial and ventricular signals of approximately equal amplitude. The voltage maps were then edited and intracavitary points were eliminated. Adequate endocardial catheter contact was confirmed by concordant catheter tip motion with the cardiac silhouettes on fluoroscopy. Intracavitary points were identified as sites located at abrupt indentations on the endocardial contour, with associated sudden reduction in electrogram amplitudes compared with signals from surrounding sites.
None of our patients had inducible bundle branch reentry VT. The diagnosis of bundle branch reentry VT was excluded based on intracardiac recordings and the response to programmed stimulation. Because electrocardiographic documentation of the presenting arrhythmia was often not available, “clinical” versus “nonclinical” VT was difficult to define with certainty. Mapping was thus directed at all uniform VTs induced by programmed stimulation. An endocardial voltage map was first constructed during a baseline rhythm. Ventricular tachycardia was then induced, and the twelve lead ECG recordings of induced VT were analyzed to regionalize the site of origin.11 The VT mapping techniques included activation mapping, entrainment mapping, and pace-mapping. All mapping techniques were always performed in combination as clinically indicated. An early endocardial activation was defined as presystolic activity that was at least 40 ms pre-QRS onset during VT. Entrainment mapping was performed using the standard criteria.12,13 For hemodynamically tolerated monomorphic VT, activation mapping identified sites with early endocardial activation and entrainment responses to pacing at these sites were analyzed. The site of origin for hemodynamically tolerated VT was defined as the site of entrainment with concealed fusion and the return cycle length equaled to the VT cycle length. “Unmappable” VT was defined as VT associated with hemodynamic intolerance, inconsistent induction, altering QRS morphology, and/or nonsustained duration. For the unmappable VTs, pace-mapping was the predominant mapping technique. The catheter was first placed along the border of the endocardial low-voltage areas guided by analysis of the 12-lead ECG recordings of the VT. The site of origin of the VT was then determined by the site of pace-mapping with the paced QRS morphology mimicking that of VT on the 12 lead ECG.11,14,15 To further characterize the VT substrate, selective epicardial recordings were performed via coronary sinus and epicardial veins in 3 patients.
Reference Values for Bipolar Voltage Mapping
The reference value for distinguishing normal and abnormal bipolar electrogram amplitude in the left ventricle was previously established at 1.55 mV using the Navistar catheter and the CARTO system.10 The “normal” signal amplitude was defined as the value above which 95% of all bipolar signal voltages from the endocardium of normal left ventricles were included. In an attempt to adjust for factors such as hypertrophy in our patient population and to avoid underestimating areas of abnormal electrograms, the reference value for electrogram amplitude used to define normal LV endocardium was arbitrarily set at 1.8 mV in this study. “Dense scar” was defined as areas with signal amplitude less than 0.5 mV based on our catheter mapping experience in patients with coronary artery disease.10 The border zone was defined as a transition zone between dense scar and normal tissue (0.5 to 1.8 mV).
Estimation of Abnormal Endocardium
Abnormal endocardium was defined as areas of contiguous recordings with electrogram amplitude of less than 1.8 mV. The color display of voltage maps was set to outline the abnormal endocardium and dense scar with a color range of 0.5 to 1.8 mV. The extent of abnormal endocardium and dense scar was estimated by measuring contiguous areas of appropriate electrogram signals from computer-generated color images (Figure 1). By using an area calculation computer program, the endocardial surface area was outlined and divided into multiple, nonoverlapping triangular segments. The extent of abnormal endocardium demonstrating low signal voltage (<1.8 mV) or area of dense scar (<0.5 mV) can then be determined as the sum of all contiguous segments demonstrating the corresponding voltage characteristics. By using the same algorithm, the entire LV endocardial surface area can also be measured. The aortic and mitral valvular annuli were first outlined and excluded, and the remaining endocardial surface was divided into multiple nonoverlapping segments. The total LV endocardial surface area was then calculated as the sum of all segments (Figure 1).
All results were presented as mean±SD. Comparisons were analyzed using the ANOVA test. A value of P<0.05 was considered statistically significant.
The study population consisted of 15 men and 4 women, with an average age of 61±16 years (range 37 to 91 years old). The mean LV ejection fraction of the group was 34±11%. All but 2 patients were on antiarrhythmic drugs at the time of electrophysiological procedure, and 16 of the 19 patients have implantable defibrillators (Table 1).
Voltage Mapping Results
The number of sites sampled per chamber was 178±83 (range 79 to 354). The total LV endocardial surface area in this population was 190±35 cm2 (range 143 to 265 cm2). The extent of abnormal endocardial electrogram recordings averaged 41±28 cm2 in size with a wide range of 8 to 94 cm2. The abnormal low-voltage area (<1.8 mV) involved only 20±12% (range 6% to 48%) of the total LV endocardial surface. Modest endocardial abnormality, arbitrarily defined as less than 25% involvement of the total LV endocardial surface, was observed in 14 out of 19 patients. Furthermore, the amount of dense scar (<0.5 mV) accounted for approximately 27±20% (range 0% to 64%) of the overall abnormal low-voltage endocardial substrate (Table 2).
The location of abnormal electrogram recordings was described as either basal or apical, determined by dividing the ventricular shell into equal halves perpendicular to its longitudinal axis. All 19 patients had abnormal low-voltage endocardial areas located near the ventricular base, frequently surrounding the mitral valve annulus, whereas only 1 patient had abnormal endocardial electrograms recorded from an apical region (Figure 2).
VT Induction and Localization
Multiple morphologies of VT were induced in patients with nonischemic cardiomyopathy. The average number of induced VT morphologies was 3±1 per patient with a range of 1 to 6 VTs. In this study, all episodes of VT were induced by programmed stimulation without isoproterenol infusion.
A total of 57 VTs were mapped in 19 patients with an average cycle length of 373±91 ms (range 245 to 570 ms). Overdrive pacing during sustained monomorphic tachycardia resulted in an entrainment response with a constant return cycle length in 14 out of 19 patients, suggesting stimulation at or near the reentrant circuit. Pace-mapping with activation mapping was performed in 18 VTs, and pace-mapping alone was performed in the remaining 27 VTs.
Overall, 50 out of 57 VTs (88%) had site of origin located near the basal LV region, whereas only 7 VTs (12%) were determined to originate from a nonbasal area (Table 3). Entrainment mapping identified sites (9 exit and 3 isthmus) within the reentrant circuits in 12 monomorphic VTs, and all 12 VT circuit components were located near the left ventricular base. All 19 patients had at least one VT originating from the LV base (Figure 3). This prevalence of VT originating from the ventricular basal region correlated to the predominant basal distribution of abnormal endocardial voltage recordings. Therefore, in most patients with nonischemic cardiomyopathy, it would appear that the VT sites-of-origin is consistent with the location of the abnormal anatomic substrate demonstrating low endocardial voltage.
In 3 patients, a mapping catheter was positioned in an epicardial vein via the coronary sinus (patients no. 1, 2, and 12). Although detailed epicardial mapping was not performed, abnormal epicardial electrograms were recorded along the vein. Presystolic, multicomponent epicardial electrogram recordings appear to extend beyond the abnormal endocardial involvement, and epicardial electrogram recordings were always earlier than the endocardial recordings in each of these 3 patients. Furthermore, during sustained monomorphic VT, pacing from epicardial sites resulted in entrainment with concealed fusion in all 3 patients. This is consistent with recordings from within the reentrant circuit, located at the basal LV epicardium (Figure 4). Pacing from endocardial sites opposite to the epicardial catheter resulted in QRS fusion with a return cycle length equal to the VT cycle length, consistent with recordings from an outer loop site of the circuit.
Outcome of VT Ablation and Followup
Thirty-nine out of 57 VTs were targeted with radiofrequency ablation based on mapping information. Three VTs were determined to be epicardial in origin and thus were not targeted for ablation. Radiofrequency energy delivery resulted in noninducibility in 30 targeted VTs using focal and/or linear ablation strategies. Additional VTs became noninducible after ablating different arrhythmia morphologies. Fourteen of the 19 patients had no VT inducible at the end of procedure. After a follow-up of 22±12 months (range 1 to 42), 5 patients are alive without VT recurrence, and 8 patients are alive with only rare VT recurrences (less than one event every 6 months). Two patients died of progressive heart failure with frequent recurrent arrhythmias, and two patients died of nonarrhythmic causes. Two patients underwent cardiac transplantation; one for end-stage heart failure and one for incessant arrhythmia.
The current study using electroanatomical bipolar voltage mapping provides a unique description of the endocardial electrophysiological substrate for uniform VT in the setting of nonischemic cardiomyopathy. The voltage maps constructed from several hundred sampled sites, and provided insight into the size and location of endocardial abnormalities that were not previously attainable with conventional mapping technologies.6,7 The endocardial substrate in this setting is marked by modest distribution of abnormal electrogram recordings with low-voltage signals, ranging from 6% to 48% of the LV endocardial surface, but uncommonly involving more than 25% of the total endocardial surface area. Furthermore, in patients with nonischemic cardiomyopathy, the predominant distribution of abnormal endocardial electrogram recordings is located at the left ventricular base, frequently involving the perivalvular regions. Ventricular tachycardias in these patients typically originate from the basal region of the left ventricle, corresponding to the locations of the anatomic endocardial substrate.
In prior endocardial mapping studies by Cassidy et al,7 the extent of abnormal endocardial sites in patients with nonischemic cardiomyopathy and VT was noted to be larger (37±27% versus 20±12%). This difference is most likely due to differences in sampling density (12 versus 178 sites per chamber). Our detailed sampling helped to define the large extent of normal endocardium and thus the proportionately smaller area of electrogram abnormalities. Because of the large standard deviation noted in the former series, differences in patient selection may have also played a role. The unique propensity for abnormal basal endocardial voltage and VT site of origin in patients with nonischemic cardiomyopathy has not been previously observed and remains unexplained. Similar distributions of endocardial electrogram abnormalities and perfusion defects have also been reported in patients with cardiac sarcoidosis.16
Epicardial versus Endocardial Abnormalities
Although our study did not attempt to evaluate all patients for the presence of epicardial electrogram abnormalities, the results of epicardial venous mapping during VT in selected patients provided some insight with respect to the possibility of epicardial arrhythmogenic site of origin in some patients. With epicardial venous mapping, abnormal epicardial electrograms were observed in sinus rhythm and entrainment with concealed fusion can be demonstrated during VT. Admittedly, only 3 patients underwent epicardial venous mapping. These patients were selected because of the absence of appropriate endocardial VT site of origin; additional investigation using epicardial mapping via a pericardial puncture technique will be required to further define the epicardial substrate in this population.17 Based on results of prior intraoperative sinus rhythm mapping during ICD patch lead placement, the overall extent of epicardial versus endocardial electrogram abnormalities was similar in patients with nonischemic cardiomyopathy and inducible monomorphic VT.18 Importantly, epicardial electrogram abnormalities predominated in some patients, whereas endocardial abnormalities were prevalent in others. The relatively equal distributions of electrogram abnormalities suggest some VT circuits may be located in the myocardial or epicardial layers, whereas other tachycardia may originate from the subendocardium. These findings are consistent with our current observations.
Mechanism of VT in Nonischemic Cardiomyopathy
Based on present observations, the potential mechanisms of uniform VT in patients with nonischemic cardiomyopathy include reentry or focal activity.19–21 Most of our patients demonstrated entrainment in responses to overdrive pacing, with some demonstrating progressive fusion, either from remote sites,22 or from sites within a reentrant circuit.12,13 Based on our findings and similar observations by other investigators,20 reentry appears to play a major role in the mechanism of sustained monomorphic VT in this patient population.
The influence of rhythm on voltage maps was considered. Although individual electrogram characteristic may change with different rhythms, the overall distribution of electrogram pattern determined by detailed mapping should not be significantly affected. Importantly, right ventricular pacing does not create abnormal electrogram recordings in the left ventricle (unpublished observation, 2003). Based on our previous observations, voltage maps constructed during sinus rhythm and sustained monomorphic VT showed similar locations of abnormal endocardium with nearly identical distribution of dense scar and border zone. With high-density recordings (178±83 points per chamber), we believe the electroanatomical maps accurately reflect the underlying voltage distribution regardless of the underlying rhythm.
The effect of antiarrhythmic drugs on the endocardial voltage characteristics is unknown. Only 2 patients were not on either Class IA or III antiarrhythmic drugs at the time of electrophysiology study. Twelve out of 19 patients were on amiodarone, either alone or in combination with other drugs for control of recurrent arrhythmias. A separate analysis was therefore performed to compare patients on or off amiodarone. The extent of abnormal endocardial electrogram distribution was similar between the two patient groups (21±12 versus 15±6% of total LV endocardial surface; P=NS).
Pace-mapping was used to determine the “site of origin” for the majority of VTs in this study. Pace-mapping was used because the induced VTs were frequently poorly tolerated hemodynamically, self-terminating, or changed morphology frequently with catheter manipulation or pacing. Compared with entrainment mapping, the ability of pace-mapping to define the precise site of origin of VT is limited. Nevertheless, endocardial pacing has been shown to be an effective corroborative method to “regionalize” the site of origin of VT.14,15 Of note, pace-mapping and entrainment mapping were performed in 12 hemodynamically tolerated monomorphic VTs. Entrainment mapping identified sites within the reentrant circuits located near the LV basal region. Pace mapping in these 12 VTs also identified basal sites from where the VT originated. The data in these 12 VTs support the use of the pace-mapping information to regionalize the origin of the remaining VTs.
In conclusion, monomorphic ventricular tachycardia in the setting of nonischemic cardiomyopathy is associated with a predominant basal distribution of endocardial electrogram abnormalities and arrhythmia site of origin. This unique distribution of the electroanatomical substrate for VT in this setting is unexplained. Nevertheless, an awareness of this important electroanatomical information may facilitate strategies for effective ablation strategies.
Gardner P, Ursell P, Fenoglio J, et al. Electrophysiologic and anatomical basis for fractionated electrograms recorded from healed myocardial infarcts. Circulation. 1985; 72: 596–611.
Dillon S, Allessie M, Ursell P, et al. Influences of anisotropic tissue structure on reentrant circuits in the epicardial border zone of subacute canine model. Circ Res. 1988; 63: 182–206.
Fenoglio J, Pham T, Harken A, et al. Recurrent sustained ventricular tachycardia: structure and ultrastructure of subendocardial regions in which tachycardia originates. Circulation. 1983; 68: 518–533.
Josephson M, Horowitz L, Farshidi A, et al. Recurrent sustained ventricular tachycardia: 2. Endocardial mapping. Circulation. 1977; 57: 440–447.
Horowitz L, Josephson M, Harken A. Epicardial and endocardial activation during sustained ventricular tachycardia in man. Circulation. 1980; 61: 1227–1238.
Cassidy D, Vassallo J, Buxton A, et al. The value of catheter mapping during sinus rhythm to localize site of origin of ventricular tachycardia. Circulation. 1984; 69: 1003–1010.
Cassidy D, Vassallo J, Miller J, et al. Endocardial catheter mapping in patients in sinus rhythm: Relationship to underlying heart disease and ventricular arrhythmias. Circulation. 1986; 73: 645–652.
Gepstein L, Hayam G, Ben-Haim S. A novel method for nonfluoroscopic catheter-based electroanatomical mapping of the heart: in vitro and in vivo accuracy results. Circulation. 1997; 95: 1611–1622.
Buxton A, Waxman H, Marchlinski F, et al. Role of triple extrastimuli during electrophysiologic study of patients with documented sustained ventricular tachyarrhythmias. Circulation. 1984; 69: 532–540.
Marchlinski F, Callans D, Gottlieb C, et al. Linear ablation lesions for control of unmappable ventricular tachycardia in patients with ischemic and nonischemic cardiomyopathy. Circulation. 2000; 101: 1288–1296.
Miller J, Marchlinski F, Buxton A, et al. Relationship between the 12-lead electrocardiogram during ventricular tachycardia and endocardial site of origin in patients with coronary artery disease. Circulation. 1988; 77: 759–766.
Stevenson W, Khan H, Sager P, et al. Identification of reentry circuit sites during catheter mapping and radiofrequency ablation of VT late after myocardial infarction. Circulation. 1993; 88: 1647–1670.
Perlman R, Miller J, Kindwall K, et al. Abnormal epicardial and endocardial electrograms in patients with idiopathic dilated cardiomyopathy: relationship to arrhythmias. Circulation. 1990; 82: I–708.
Pogwizd S, McKenzie J, Cain M. Mechanisms underlying spontaneous and induced ventricular arrhythmias in patients with idiopathic dilated cardiomyopathy. Circulation. 1998; 98: 2404–2414.
Callans D, Hook B, Josephson M. Comparison of resetting and entrainment of uniform sustained ventricular tachycardia: further insights into the characteristics of the excitable gap. Circulation. 1993; 87: 1229–1238.