Characterization of myocardial scars: Electrophysiological imaging correlates in a porcine infarct model

Characterization of myocardial scars: Electrophysiological imaging correlates in a porcine infarct model Shiro Nakahara, MD, PhD,* Marmar Vaseghi, MD, MS,* Rafael J. Ramirez, PhD,* Carissa G. Fonseca, PhD,† Chi K. Lai, MD,‡ J. Paul Finn, MD,*† Aman Mahajan, MD, PhD,* Noel G. Boyle, MD, PhD, FHRS,* Kalyanam Shivkumar, MD, PhD, FHRS*† From the *UCLA Cardiac Arrhythmia Center, †Department of Radiological Sciences, and ‡Department of Pathology, David Geffen School of Medicine at UCLA, Los Angeles, California. BACKGROUND Definition of myocardial scars as identified by electroanatomic mapping is integral to catheter ablation of ventricular tachycardia (VT). Myocardial imaging can also identify scars prior to ablation. However, the relationship between imaging and voltage mapping is not well characterized. OBJECTIVE The purpose of this study was to verify the anatomic location and heterogeneity of scars as obtained by electroanatomic mapping with contrast-enhanced MRI (CeMRI) and histopathology, and to characterize the distribution of late potentials in a chronic porcine infarct model. METHODS In vivo 3-dimensional cardiac CeMRI was performed in 5 infarcted porcine hearts. High-density electroanatomic mapping was used to generate epicardial and endocardial voltage maps. Scar surface area and position on CeMRI were then correlated with voltage maps. Locations of late potentials were subsequently identified. These were classified according to their duration and fractionation. All hearts underwent histopathological examination after mapping. RESULTS The total dense scar surface area and location on CeMRI correlated to the total epicardial and endocardial surface scar on electroanatomic maps. Electroanatomic mapping (average of 1,532 ⫾

Introduction Characterization of the myocardial substrate is crucial for defining therapeutic strategy for catheter ablation of ventricular tachycardia (VT), and will be important for delivery of biologic therapies for tissue engineering. Catheter ablation of VT is a well-established option for patients with recurrent symptomatic arrhythmias. However, patients referred for catheter ablation of VT often have hemodynamically unstable arrhythmias, resulting in the use of substrate-based mapping and ablation approaches. Electroanatomic mapping of the myocardial substrate generates voltage maps that are commonly used Funded by National Institutes of Health RO1-HL084261 and HL067647 grants to K.S. Drs. Nakahara and Vaseghi contributed equally to this work. Address reprint requests and correspondence: Dr. Kalyanam Shivkumar, UCLA Cardiac Arrhythmia Center, Ronald Reagan UCLA Medical Center, 47-139 CHS, David Geffen School of Medicine at UCLA, 10833 Le Conte Avenue, Los Angeles CA 90095-1679. E-mail address: [email protected]. (Received January 1, 2011; accepted February 19, 2011.)

480 points per infarcted heart) showed that fractionated late potentials were more common in dense scars (⬍0.50 mV) as compared with border zone regions (0.51 to 1.5 mV), and were more commonly observed on the epicardium. CONCLUSION In vivo, CeMRI can identify areas of transmural and nontransmural dense scars. Fractionated late diastolic potentials are more common on the epicardium than the endocardium in dense scar. These findings have implications for catheter ablation of VT and for targeting the delivery of future therapies to scarred regions. KEYWORDS Late potentials; Ventricular tachycardia; Contrast-enhanced magnetic resonance imaging ABBREVIATIONS 3D ⫽ 3-dimensional; CeMRI ⫽ contrast-enhanced magnetic resonance imaging; EGM ⫽ electrogram; f-LDP ⫽ fractionated late diastolic potentials; LDP ⫽ late diastolic potentials; LV ⫽ left ventricle; MDP ⫽ middiastolic potential; MI ⫽ myocardial infarction; VT ⫽ ventricular tachycardia; LP ⫽ late potential; LVA ⫽ low voltage area (Heart Rhythm 2011;8:1060 –1067) © 2011 Heart Rhythm Society. All rights reserved.

to identify and target reentrant circuits within these scars. Substrate mapping utilizes electrogram (EGM) recordings from dense scars and border zones to visualize and identify critical isthmuses, or regions of slowed electrical conduction.1-5 Recently, endocardial sites demonstrating late potential EGMs, which are low-voltage recordings that occur 240 ms after the onset of local EGM, have been shown to correlate well with areas containing critical isthmuses in postinfarct ventricular tachycardias.1,2 However, conventional electroanatomic mapping and current approaches to VT ablation have several limitations: (1) far-field recording of EGMs and poor catheter contact can lead to mischaracterization of normal myocardium as low-voltage areas, (2) point-bypoint mapping of the left ventricle (LV) is time consuming and leads to prolonged procedure/anesthesia time, and (3) the location and characterization of late potentials is not well understood. Thus, a more selective strategy for identification of sites for mapping and ablation would be desirable.

1547-5271/$ -see front matter © 2011 Heart Rhythm Society. All rights reserved.

doi:10.1016/j.hrthm.2011.02.029

Nakahara et al

CeMRI and Characterization of Late Potentials

Figure 1 A timeline of the experimental protocol is shown. A myocardial infarct was followed by CeMRI 4 to 6 weeks later. Within 1 to 2 weeks of the CeMRI, all animals had electrophysiological study and electroanatomic mapping performed, after which the hearts were analyzed for histopathology. CeMRI ⫽ contrast-enhanced magnetic resonance imaging.

The purpose of this study was to use a porcine infarct model to (1) verify the identification and extent of scars as obtained by electroanatomic mapping with contrast-enhanced MRI (CeMRI) and histopathology; and (2) to characterize the spatial organization and voltage distribution of late potentials using high-resolution epicardial and endocardial mapping techniques. Defining image-substrate correlation has implications for imaging-guided catheter ablation and targeting of biologic therapies in vivo.

Methods A myocardial infarct (MI) was created in 5 pigs; 4 to 8 weeks were allowed for maturation of infarcts. The pigs subsequently underwent CeMRI, and within 10 days, electrophysiological study and histopathology was performed on the hearts (Fig. 1). This protocol was approved by the University of California at Los Angeles subcommittee of Animal Research Care.

Creation of myocardial infarcts Pigs underwent MI under general anesthesia using polystyrene microspheres injected through the lumen of an angioplasty balloon catheter inflated in the mid left anterior descending coronary artery (Online Appendix 1).

MRI acquisition and analysis A myocardial scar map was obtained from CeMRI delayed enhanced images (slice gap 6 mm).6,7,8 Three-dimensional models of the scar were constructed using VPT software (Siemens Healthcare, Princeton, New Jersey). This reconstruction provided a high-resolution 3-dimensional (3D) myocardial scar map for visualization of infarct location and extent.

Electrophysiological mapping and EGM analysis Electrophysiological studies were performed within 1 week of CeMRI under general anesthesia (Online Appendix 1). Endocardial and epicardial LV voltage maps were obtained (Online Appendix 1).9 Previous studies reported duration of fractionated EGMs in myocardial infarction exceeding 130 ms,10 and VT ablation targets with EGM durations of 150 to 240 ms.2,10 In this study, late potential EGMs were defined as any distinct EGM component that occurred after the end of the surface QRS. Late potentials were classified as: middiastolic poten-

1061 tial (MDP; duration of EGM onset to LP (late potential) ⱕ250 ms), late-diastolic potential (LDP; duration of EGM onset to LP ⬎250 ms). If the LDP contained more than 2 discrete deflections, it was further classified as fractionated LDP (f-LDP). Infarcted tissue was identified by low-voltage EGMs. Strict voltage criteria were used (dense scar: EGM amplitude ⱕ 0.5 mV; border zone: EGM amplitude 0.5 mV to 1.5 mV, normal tissue: EGM amplitude ⬎1.5 mV). On the epicardium, fractionation and late potentials were also used as markers of scar to differentiate coronary vasculature/ epicardial fat from scar. The American Heart Association 17-segment model was used to assess distribution of scar in CeMRI images and electroanatomic maps.11 Spatial distribution of transmural scars and scar surface areas were compared between CeMRI and electroanatomic map low-voltage areas (ⱕ1.5 mV) to determine positive or negative matching of scar tissues within individual segments (Online Appendix 1).

Histopathological analysis After completion of mapping, animals were euthanized. The heart was resected immediately. The gross specimens were photographed, and the epicardial and endocardial infarct areas were delineated. The hearts were then placed in 10% neutral-buffered formalin, sectioned, and routinely processed and embedded in paraffin. Serial 4-␮m-thick sections were prepared for routine hematoxylin-and-eosin and Masson trichrome/elastic-Van Gieson stains.

Statistical analysis Results are expressed as mean ⫾ standard deviation. Categorical variables were compared by ␹2 analysis and Fisher exact test. Continuous variables were compared using the unpaired Student t test. A value of P ⬍.05 was considered statistically significant. The StatView J-5.0 program (Abacus Concepts, Inc., Berkeley, California) was used for statistical analyses.

Results High-resolution mapping of MI Five pigs underwent electrophysiological study 45 ⫾ 12 days after creation of MI (Fig. 2). Voltage maps were constructed by recording EGMs from the epicardium (794 ⫾ 239 points/ heart; range 464 to 977) and endocardium (738 ⫾ 301 points/heart; range 462 to 1,176). On electroanatomic mapping, postinfarct endocardial dense scar tissue was seen in all 5 animals, whereas scar tissue on the surface of the epicardium was identified in 4 animals. The total lowvoltage area (border zone and dense scar combined) was 10.9 ⫾ 10.4 cm2 (recorded from 143 ⫾ 151 points/heart) on the epicardium and 17.9 ⫾ 10.4 cm2 (recorded from 457 ⫾ 151 points/heart) on the endocardium. On electroanatomic mapping, the dense scar in the epicardium and endocardium had an area of 2.9 ⫾ 3.4 cm2 (recorded from 80 ⫾ 104 points/heart) and 5.6 ⫾ 4.8 cm2 (239 ⫾ 80 points/heart). One animal did not have demonstrable dense scar or border

1062

Heart Rhythm, Vol 8, No 7, July 2011

Figure 2 High-resolution combined endocardial and epicardial electroanatomical mapping of infarcted porcine heart. A: Fluoroscopic images. The ENDO catheter is positioned in the LV via transaortic approach. The EPI catheter is positioned near the apex in the pericardial space. B: Voltage maps of ENDO and EPI surfaces. A large ENDO scar is observed in the anteroseptal region. The scar and border zone regions observed on the EPI surface spatially coincides with that on the ENDO surface, although the dense scar area is smaller on EPI than ENDO maps. Non–scar tissue (⬎1.5 mV) ⫽ purple, scar border zone areas ⫽ color range from 0.51 to 1.5 mV, dense scar (⬍0.5 mV) ⫽ gray. C: Activation maps showing late activation in areas of endocardial and epicardial scar. ENDO ⫽ endocardial; EPI ⫽ epicardial; LAO ⫽ left anterior oblique; LV ⫽ left ventricular; RAO ⫽ right anterior oblique.

zone regions on the epicardium and a second animal did not have any epicardial dense scar. The total procedure times for geometric construction and voltage mapping using a multielectrode catheter were 43.0 ⫾ 5.2 minutes and 24.4 ⫾ 5.5 minutes for epicardial and endocardial surfaces, respectively.

contact. In 1 infarcted porcine heart, scar was not detected by CeMRI, although a small area of endocardial dense scar (ⱕ0.5 mV, 0.7 cm2), and significant border zone regions (11.7 cm2) were detected by electroanatomic mapping. On histopathology in this heart, an area of patchy endocardial scar was observed in the corresponding region.

Comparison of scar between CeMRI and electroanatomic mapping

Distribution of late potentials in infarcted hearts

Examination of 5 infarcted hearts by CeMRI revealed LV anteroseptal scars in 4 hearts (Fig. 3), 2 of which also had apical scars. The total scar surface area measured by CeMRI was similar to the surface area of dense scar (⬍0.5 mV, endocardial ⫹ epicardial) on the electroanatomic maps (10.3 ⫾ 7.4 cm2 and 8.5 ⫾ 7.9 cm2, respectively, P ⫽ .76) (Fig. 4). The CeMRI scar area was significantly smaller (P ⬍ .05) than the total low-voltage area (⬍1.5 mV) on electroanatomic mapping (28.8 ⫾ 15.9 cm2), as CeMRI was not able to detect border zone regions (P ⬍ .05). The location of transmural scars on CeMRI corresponded with dense scars on electroanatomic mapping (Fig. 5). CeMRI and electroanatomic mapping displayed positive segment matching in an average of 13.0 ⫾ 1.6 of 17 segments (76%). The predominant mismatched areas (24%) were endocardial pseudoscars at the base of the LV seen on electroanatomic maps that did not represent true scars as seen by MRI or histopathology. These false-positive lowvoltage areas may be due to far-field recordings of EGM signals from distant ventricular or atrial tissue or from poor

In 5 animals, a total of 425 mapped sites/EGMs with late potentials were observed (epicardial: 228 points, 46 ⫾ 42 points/animal, range 0 to 104, endocardial: 197 points, 39 ⫾ 13 points/animal, range 26 to 57). The mismatched scar regions between CeMRI and electroanatomic maps in basal areas of the LV were excluded from analysis of the LP distribution. As expected, late potential EGMs were predominantly located in dense scar and border zone regions. Of note, there were significantly more late potential EGMs on the epicardium than endocardium (30% vs. 8.4%, P ⬍ .05) (Table 1, Figs. 5 and 6). Figure 4 shows typical examples of epicardial late potentials (MDP, LDP, and f-LDP) recorded from dense scar areas. Most epicardial LDPs, and more specifically, most epicardial f-LDPS, were positioned directly above endocardial low-voltage areas (scar and border zones, Table 1). Of interest, LDPs were prominent in epicardial border zone tissues but not as prevalent in endocardial border zones (P ⬍ .0001) (Fig. 6, Table 1). Figure 7 shows a histogram of recorded nonbasal late potential EGMs sorted by voltage amplitude in all swine. Fractionated LDP

Nakahara et al

CeMRI and Characterization of Late Potentials

1063

Figure 3 CeMRI and 3-dimensional scar rendering of infarcted porcine heart. A: Short-axis and 4-chamber views of CeMRI from the same porcine heart mapped in Figure 2. The white arrows indicate delayed enhancement of scar in the anteroseptal wall. B: Three- dimensional scar map rendered from multiple short-axis images showing an anteroseptal scar (blue). C. Representative histopathological sections (epicardial and endocardial) from infarct, border zone, and remote regions is shown, Masson trichrome/elastic-Van Gieson stain. CeMRI ⫽ contrast-enhanced magnetic resonance imaging; other abbreviations as in Figure 2.

were more likely to be observed in dense scar as compared with border zone regions (0.5 to 1.5 mV) (Fig. 6, Table 1). MDPs were predominantly located in the border zone regions of the endocardium and epicardium, and 40% were not located above endocardial low-voltage areas. Hence, the significance of MDPs as targets for VT ablation remains questionable.

Discussion Major findings In vivo CeMRI correctly identifies dense scar as defined by areas ⬍0.5 mV on electroanatomic maps. This allows for preprocedure planning of catheter ablation of VT and has implications for delivery of percutaneous biological therapies for heart disease. LDPs are found on both the epicardium and endocardium in and around dense scars

and are more prevalent on the epicardium. f-LDPs are much more common in dense scar than border zone regions in the endocardium. However, on the epicardium, border zone regions also harbor a significant number of f-LDPs.

Value of in vivo CeMRI In this study, we describe the first animal infarct model that can be used for scar characterization studies in which CeMRI was performed in vivo. Although qualitative studies have previously been done in humans, this is the first systematic assessment of imaging-electroanatomic correlates of scars and late potentials. The fact that in vivo MRI can be performed in an infarct experimental model has implications beyond catheter ablation. CeMRI can be used to guide the delivery of future therapies, such as stem cells, to ap-

1064

Heart Rhythm, Vol 8, No 7, July 2011 providing targets for catheter ablation prior to the procedure, decreasing mapping time. The correlation of electroanatomic map low-voltage ischemic areas (ⱕ0.5 mV) with MRI and histopathology validates the definition of dense scar. The high resolution and short duration of our electroanatomical mapping approach may help improve current substrate-based ablation approaches for ischemic VT.

Late potential distribution

Figure 4 Scar surface area measured by CeMRI and electroanatomic mapping. Total scar surface area was calculated from CeMRI and voltage maps. Scar surface area measured by CeMRI was not significantly different than the total DS on the EPI and ENDO. However, the combined EPI and ENDO low-voltage areas (LVA⫽ border zone and dense scar, ⬍1.5 mV) were significantly greater than scar areas measured by CeMRI, P ⬍ .05. DS ⫽ dense scar; LVA ⫽ Low voltage area; ns ⫽ not significant; other abbreviations as in Figure 3.

propriate targets using catheter-based intramyocardial administration. MRI-guided verification of scar regions recorded in a porcine infarct model using sock electrodes has been previously reported. Ashikaga et al12 showed that MRI identified scars with spatially complex structures that served as the substrate for multiple VT morphologies ex vivo. Although the feasibility for MRI-guided catheter ablation was previously demonstrated in porcine infarction model, no correlation of electroanatomic maps to CeMRI was made in these experiments.13 This is the first study to demonstrate MRI based visualization of complex endocardial and epicardial scar morphologies in vivo.

CeMRI-guided multipolar catheter electroanatomic mapping Current conventional approaches for VT ablation utilize a pair of bipolar catheters to map scars and identify late potentials. Depending on the clinical VT, this procedure can take several hours under general anesthesia in cardiomyopathic hearts. A further limitation of this approach is that electroanatomic mapping may characterize normal areas as scar due to far-field recording, poor catheter contact, or from interference by epicardial fat deposits.14, 15 Thus, MRI is a valuable adjunct to electroanatomic mapping by confirming location of subendocardial and transmural scars and

A better understanding of substrate characteristics could provide insights into VT mechanisms and facilitate catheter ablation strategies. Characterization of local EGM voltage and morphology are useful for identification of VT isthmuses or slow conduction channels.16,17 Although sinus rhythm late potentials are very sensitive in identifying critical isthmuses of VT,1,2,17-19 they are not very specific. Many late potentials are generated by sites representing areas of fixed conduction block within bystander or deadend pathways, which are not suitable targets for ablation.19,20 Ideally, late potential guided ablation targets can be identified using a pace map or entrainment techniques.4 Recently, ideal pace map sites have been shown to have significantly longer EGM durations and lower amplitudes than poor pace map sites.2 Endocardial pace map sites identified as perfect were shown to have average EGM durations of 241 ms and amplitudes of ⬍0.1 mV,2 corresponding to LDP in the present study. Further, earlier studies examining endocardial EGM morphologies suggested that late potentials with multiple discrete deflections (defined as f-LDP in the present study) were closely associated with ideal ablation targets with greater specificity than nonfractionated late potential EGMs.17 In this study, the profile of EGM late potential amplitudes and morphologies within dense scar and border zone areas of infarcted hearts showed that the majority of f-LDPs and LDPs are primarily situated in dense scars on both the epicardium and the endocardium. This represents a starting point for targeting VTs during substrate based mapping. Conventional VT ablation strategies have generally focused on identification of endocardial late potential EGMs as ideal sites for targeting slow conduction channels.1,2,17,21 However, about 15% of ischemic cardiomyopathic patients and up to 30% of nonischemic cardiomyopathy patients benefit from an epicardial approach.15,22 On the epicardium, critical isthmuses are more difficult to identify using entrainment or pace map techniques because the presence of epicardial fat deposits results in a high pacing threshold.9,23,24 A high-density and rapid substrate-based electroanatomical mapping approach, as outlined in the current study, might prove invaluable for epicardial VT ablation. Further, our data point to the novel finding that, although f-LDPs are common in dense scar regions on both the epicardium as well as the endocardium, they are in fact more prevalent on the epicardium (Table 1). Endocardial scars were compact and dense compared with the epicardium, consistent with previous findings.15 The patchy nature of dense scar patches (Figs. 4 and 5) surrounded by

Nakahara et al

CeMRI and Characterization of Late Potentials

1065

Figure 5 Transmural scars viewed in CeMRI correspond with dense scar areas from voltage maps. The position of transmural scars in 3D CeMRI images corresponded well with DS areas on electroanatomic maps. A: Rendering of 3D CeMRI image of the same porcine heart from Figures 2 and 3 showing endocardial (ENDO, dark pink) and epicardial (EPI, light pink) of the LV, with a cutout view of the right ventricle shown for orientation. Subendocardial scar tissue is depicted in light green, and transmural scar is shown in dark green. Upper image shows a short-axis view of the 3D CeMRI with an overlay of the American Heart Association 17-segment division lines. B: High-resolution voltage maps demonstrate spatial correspondence of dense scar areas with CeMRI (gray dotted arrows). C: Typical epicardial late potentials (lower tracings) and lead II electrocardiographic measurements (upper tracing) recorded simultaneously. 3D ⫽ three-dimensional; LDP ⫽ late diastolic potentials; f-LDP ⫽ fractionated late diastolic potentials, MDP ⫽ middiastolic potentials, BZ ⫽ border zone; other abbreviations as in Figure 4.

border zone regions, as seen by MRI and electroanatomic mapping, would be expected to support greater fractionation of LDPs due to nonhomogeneous or zigzag conduction.25 This may explain why up to 30% of clinical VT cases cannot be successfully ablated using solely an endocardial approach.14,22,26 In this study, the border zone of the epicardium also contained more f-LDPs than endocardium (Fig. 6, white bars). Most epicardial border zones lay directly above endocardial scars. Therefore, the unique positioning of epicardial border zone LDP sites likely gives rise to slowed endocardial-to-epicardial conduction properties. This findTable 1

ing suggests that ideal epicardial ablation targets should include EGM sites containing f-LDPs. Taken together, the late potential distribution data suggest that epicardial conduction may be slower and more anisotropic than endocardial conduction within low-voltage areas. Further, ideal endocardial ablation targets may be best identified in dense scar areas, whereas ideal epicardial ablation targets may be best identified in both dense scar and border zone regions. Finally, the presence of fractionated LDPs on the epicardium can also help confirm the presence of scar (as opposed to fat, which may cause low amplitude EGMs but does not cause fractionation).14,27

Distribution of late potential EGMs recorded in scar and border zone regions Endocardium

Epicardium

Swine (n ⫽ 5)

Dense scar

Border zone

P value (scar vs. border zone)

Dense scar

Border zone

P value (scar vs. border zone)

Total EGMs recorded Total late potentials LDP f-LDP Non f-LDP MDP

1,194 115 (9%)§ 91 (79%)* 23 (20%)† 68 (59%) 24 (21%)

1,093 76 (7%)§ 18 (23%)§ 1 (1%) 17 (22%)‡ 58 (77%)§

⬍.05 ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001

399 134 122 51 71 12

318 71 (22%) 43 (61%) 5 (7%) 38 (54%) 28 (39%)

⬍.05 ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001

(36%) (91%) (42%) (58%) (9%)

Values in parentheses indicate percentage of total late potentials in either scar or border zone. For comparison between epicardium vs. endocardium: EGMs ⫽ ; f-LDP ⫽ fractionated late diastolic potentials, LDP ⫽ late diastolic potentials; MDP ⫽ middiastolic potentials; EGM ⫽ electrograms. *P ⬍ .05, †P ⬍ .01, ‡P ⬍ .001, §P ⬍ .0001.

1066

Heart Rhythm, Vol 8, No 7, July 2011

Figure 6 Distribution of late potential recordings within scar and border zone regions of the endocardium and epicardium. EGM ⫽ ; other abbreviations as in Figure 4.

Study limitations In this study, the histogram of late potential EGMs showed a skewed distribution with a larger proportion of LDPs and f-LDPs in lower voltage areas. We have previously shown that the distribution of late potential EGMs is similarly skewed in ischemic cardiomyopathy patients.26 In humans, the peak of the low EGM voltages occurred in the 0.11 to 0.2 mV range.26 However, in infarcted porcine hearts, endocardial f-LDPs and LDPs occurred in the 0.21 to 0.3 mV range. It is possible that the peak of the late potential histogram shifts leftward and becomes narrower with infarct age, as human infarcts examined in previous studies are generally several years old. Further, although f-LDPs were more common on the epicardium, there exists the possibility that this is related to a sampling bias. However, detailed voltage maps were obtained in all animals, on epicardium and endocardium, making this less likely. Comparison of MRI and electroanatomic mapping reveals that although MRI sufficiently delineates dense scar

areas, MRI is not sensitive in detecting patchy scar regions. Thus, the MRI-guided substrate mapping might be less applicable in small scar-related nonischemic VT cases, in which patchy scar tissues are more common than in ischemic hearts. Further, the presence of endocardial pseudoscars correctly in porcine basal regions may be due to limitations of the retrograde aortic approach. Low-voltage potentials are, by definition, small in amplitude. Late potential EGM recordings were repeatedly validated in this study by comparing multiple recordings from the same position. Furthermore, the skewed distribution of LPs toward low-voltage areas is indicative of true positive recordings. Repeated false positive acquisition would give rise to an even distribution, independent of voltage amplitude. Use of multipolar catheters produces a large amount of data in a short time. Analysis of LP EGMs was tediously performed by hand in the present study. Future clinical applications of this method for high-resolution substrate mapping, based on LP analysis, would require rapid assess-

Figure 7 Late potential histogram in low-voltage areas of infarcted hearts. The distribution of late potentials is plotted against EGM voltage amplitudes grouped in 0.1-mV bin sizes. The dashed line delineates dense scar sites (⬍0.5 mV) from border zone tissue (0.51 to 1.5 mV). Abbreviations as in Figure 6.

Nakahara et al

CeMRI and Characterization of Late Potentials

ment of LP durations from a potentially large information pool. Furthermore, a limitation of this study remains that VT was not induced in the animals.

Conclusion The present study, to our knowledge, is the first systematic assessment of the myocardial substrate as defined by in vivo imaging and high-resolution electroanatomic (late potential) mapping in animal infarct model. The 3D CeMRI helps identify potential sites of dense scars that could harbor critical isthmuses of VT, as well as identify pseudoscars (areas of epicardial fat or far-field recording). Further, it can help guide catheter based intramyocardial therapies for heart disease. Thus, high-density late potential EGM mapping complements CeMRI, and together they have the potential to facilitate endocardial and epicardial substratebased VT ablation strategies.

Acknowledgments

1067

10.

11.

12.

13.

14.

15.

16.

The authors thank Michael S. Lee, MD, for assistance in creation of infarcts. 17.

Appendix Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.hrthm.2011. 02.029.

References 1. Arenal A, Glez-Torrecilla E, Ortiz M, et al. Ablation of electrograms with an isolated, delayed component as treatment of unmappable monomorphic ventricular tachycardias in patients with structural heart disease. J Am Coll Cardiol 2003;41:81–92. 2. Bogun F, Good E, Reich S, et al. Isolated potentials during sinus rhythm and pace mapping within scars as guides for ablation of post-infarction ventricular tachycardia. J Am Coll Cardiol 2006;47:2013–2019. 3. Marchlinski FE, Callans DJ, Gottlieb CD, Zado E. Linear ablation lesions for control of unmappable ventricular tachycardia in patients with ischemic and nonischemic cardiomyopathy. Circulation 2000;101:1288 –1296. 4. Stevenson WG, Soejima K. Catheter ablation for ventricular tachycardia. Circulation 2007;115:2750 –2760. 5. Zeppenfeld K, Stevenson WG. Ablation of ventricular tachycardia in patients with structural heart disease. Pacing Clin Electrophysiol 2008;3:358 –374. 6. Codreanu A, Odille F, Aliot E, et al. Electroanatomic characterization of post-infarct scars comparison with 3-dimensional myocardial scar reconstruction based on magnetic resonance imaging. J Am Coll Cardiol 2008; 52:839 – 842. 7. Carr JC, Simonetti O, Bundy J, Li D, Pereles S, Finn JP. Cine MR angiography of the heart with segmented true fast imaging with steady-state precession. Radiology 2001;219:828 – 834. 8. Simonetti OP, Kim RJ, Fieno DS, et al. An improved MR imaging technique for the visualization of myocardial infarction. Radiology 2001;218:215–223. 9. Sosa E, Scanavacca M, d’Avila A, Oliveira F, Ramires JA. Nonsurgical transthoracic epicardial catheter ablation to treat recurrent ventricular tachycardia

18.

19.

20.

21.

22.

23.

24.

25. 26.

27.

occurring late after myocardial infarction. J Am Coll Cardiol 2000;35:1442–1449. Kienzle MG, Miller J, Falcone RA, Harken A, Josephson ME. Intraoperative endocardial mapping during sinus rhythm: relationship to site of origin of ventricular tachycardia. Circulation 1984;70:957–965. Cerqueira MD, Weissman NJ, Dilsizian V, et al. 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. J Nucl Cardiol 2002;9:240 –245. Ashikaga H, Sasano T, Dong J, et al. Magnetic resonance-based anatomical analysis of scar-related ventricular tachycardia: implications for catheter ablation. Circ Res 2007;101:939 –947. Reddy VY, Malchano ZJ, Holmvang G, et al. Integration of cardiac magnetic resonance imaging with three-dimensional electroanatomic mapping to guide left ventricular catheter manipulation: feasibility in a porcine model of healed myocardial infarction. J Am Coll Cardiol 2004;44:2202–2213. Tung R, Nakahara S, Ramirez R, Lai C, Fishbein MC, Shivkumar K. Distinguishing epicardial fat from scar: analysis of electrograms using high-density electroanatomic mapping in a novel porcine infarct model. Heart Rhythm 2010; 7:389 –395. Cesario DA, Vaseghi M, Boyle NG, et al. Value of high-density endocardial and epicardial mapping for catheter ablation of hemodynamically unstable ventricular tachycardia. Heart Rhythm 2006;3:1–10. Arenal A, del Castillo S, Gonzalez-Torrecilla E, et al. Tachycardia-related channel in the scar tissue in patients with sustained monomorphic ventricular tachycardias: influence of the voltage scar definition. Circulation 2004;110: 2568 –2574. Brunckhorst CB, Stevenson WG, Jackman WM, et al. Ventricular mapping during atrial and ventricular pacing. Relationship of multipotential electrograms to ventricular tachycardia reentry circuits after myocardial infarction. Eur Heart J 2002;23:1131–1138. Bogun F, Bahu M, Knight BP, et al. Response to pacing at sites of isolated diastolic potentials during ventricular tachycardia in patients with previous myocardial infarction. J Am Coll Cardiol 1997;30:505–513. Bogun F, Bender B, Li YG, et al. Analysis during sinus rhythm of critical sites in reentry circuits of postinfarction ventricular tachycardia. J Interv Card Electrophysiol 2002;7:95–103. Harada T, Stevenson WG, Kocovic DZ, Friedman PL. Catheter ablation of ventricular tachycardia after myocardial infarction: relation of endocardial sinus rhythm late potentials to the reentry circuit. J Am Coll Cardiol 1997;30:1015– 1023. Brunckhorst CB, Delacretaz E, Soejima K, et al. Ventricular mapping during atrial and right ventricular pacing: relation of electrogram parameters to ventricular tachycardia reentry circuits after myocardial infarction. J Interv Card Electrophysiol 2004;11:183–191. Verma A, Marrouche NF, Schweikert RA, et al. Relationship between successful ablation sites and the scar border zone defined by substrate mapping for ventricular tachycardia post-myocardial infarction. J Cardiovasc Electrophysiol 2005;16:465– 471. Brugada J, Berruezo A, Cuesta A, et al. Nonsurgical transthoracic epicardial radiofrequency ablation: an alternative in incessant ventricular tachycardia. J Am Coll Cardiol 2003;41:2036 –2043. Schweikert RA, Saliba WI, Tomassoni G, et al. Percutaneous pericardial instrumentation for endo-epicardial mapping of previously failed ablations. Circulation 2003;108:1329 –1335. de Bakker JM, van Capelle FJ, Janse MJ, et al. Slow conduction in the infarcted human heart. ‘Zigzag’ course of activation. Circulation 1993;88:915–926. Nakahara S, Tung R, Ramirez RJ, et al. Characterization of the arrhythmogenic substrate in ischemic and nonischemic cardiomyopathy implications for catheter ablation of hemodynamically unstable ventricular tachycardia. J Am Coll Cardiol 2010;55:2355–2365. d’Avila A. Epicardial catheter ablation of ventricular tachycardia. Heart Rhythm 2008;5 Suppl:S73–75.