- Email: [email protected]
Twelve-lead ECG features to identify ventricular tachycardia arising from the epicardial right ventricle
Twelve-lead ECG features to identify ventricular tachycardia arising from the epicardial right ventricle Victor Bazan, MD, Rupa Bala, MD, Fermin C. Garcia, MD, Jonathan S. Sussman, MD, Edward P. Gerstenfeld, MD, Sanjay Dixit, MD, David J. Callans, MD, Erica Zado, PA, Francis E. Marchlinski, MD From the Cardiovascular Division, Department of Medicine, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania. BACKGROUND Usefulness of 12-lead ECG for predicting an epicardial origin for ventricular tachycardia (VT) arising from the right ventricle (RV) has not been assessed. An epicardial approach is sometimes warranted to eliminate RV VT. OBJECTIVES The purpose of this study was investigate the hypothesis that specific ECG features identify an epicardial origin for RV VT. METHODS To mimic an endocardial or epicardial origin, we paced representative sites in 13 patients undergoing RV endocardial/ epicardial mapping (134/180 pace map sites). RESULTS QRS duration from epicardial vs endocardial sites was not different (183 ⫾ 27 ms vs 185 ⫾ 28 ms, P ⫽ .3). Reported cut-off values for identifying epicardial left ventricular origin, pseudo-delta wave (ⱖ34 ms), intrinsicoid deflection time (ⱖ85 ms), and RS complex (ⱖ121 ms) did not apply to the RV. A Q wave in lead II, III, or aVF was more likely noted from inferior epicardial
vs endocardial sites (53/73 vs 16/43, P ⬍.01). A Q wave in lead I was more frequently present from epicardial vs endocardial anterior RV sites (30/82 vs 5/52, P ⬍.001). QS in lead V2 was noted from anatomically matched epicardial anterior RV sites (22/33 vs 13/33, P ⬍.05). In the RV outflow tract, no ECG feature distinguishing epicardial/endocardial origin reached statistical significance. CONCLUSION A Q wave or QS in leads that best reflect local activation suggest an epicardial origin for RV depolarization and may help in identifying a probable epicardial site of origin for RV VT. QRS duration and reported criteria for epicardial origin of VT in the left ventricle do not identify a probable epicardial origin in the RV. KEYWORDS Epicardium; Ventricular tachycardia; Pace mapping; Right ventricle (Heart Rhythm 2006;3:1132–1139) © 2006 Heart Rhythm Society. All rights reserved.
Introduction
Methods
Successful catheter ablation for both left ventricular (LV) and right ventricular (RV) tachycardia (VT) sometimes requires an epicardial approach.1– 6 Specific 12-lead ECG patterns have been described as helpful for identifying an epicardial VT origin in the LV, including pseudo-delta wave ⱖ34 ms, intrinsicoid deflection time ⱖ85 ms, RS complex duration ⱖ121 ms, and delayed precordial maximum deflection index ⱖ0.55.7,8 The 12-lead ECG is useful for identifying ventricular arrhythmias from specific regions of the RV endocardium, especially from the RV outflow tract (RVOT).9 –11 However, its usefulness in predicting an epicardial site of origin in the RV has not been established. Thus, characterizing ECG patterns to predict an epicardial origin in the RV is of interest. We sought to describe ECG features that predict an epicardial site of origin for RV arrhythmias by mimicking the origin of VT with RV endocardial and epicardial pace mapping.
Patient population
Address reprint requests and correspondence: Dr. Francis E. Marchlinski, Hospital of the University of Pennsylvania, 9 Founders Pavilion, 3400 Spruce Street, Philadelphia, Pennsylvania 19104. E-mail address: [email protected]. (Received April 27, 2006; accepted June 23, 2006.)
Twenty-five patients underwent endocardial and epicardial catheter mapping and ablation for drug-refractory ventricular arrhythmias. In each patient, the risks of mapping/ablation were discussed in detail, and all patients gave written informed consent in accordance with institutional guidelines of the University of Pennsylvania Health System. Each patient underwent detailed voltage mapping, pace mapping, and activation mapping in an attempt to characterize the VT substrate and target/ablate their arrhythmia.
Study protocol To be included in this study analysis, each patient was required to undergo detailed pace mapping from both endocardial and epicardial free-wall surfaces of the RV. Each pace mapping site was tagged on an electroanatomic map (CARTO, Biosense Webster, Diamond Bar, CA, USA) of the RV chamber of interest for detailed offline analysis. A detailed electroanatomic map in sinus rhythm was performed using a 4-mm-tip catheter maintaining a fill threshold of 20 mm, with the entire surface of the RV endocardium and epicardium documented to be sampled. Pace mapping was performed using bipolar pacing (distance between poles 1 and 2 ⫽ 1 mm) at threshold from both
1547-5271/$ -see front matter © 2006 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2006.06.024
Bazan et al
ECG Features for Epicardial Right Ventricle VT
epicardial and endocardial free-wall surfaces at cycle lengths of 400 to 600 ms. This allowed for analysis of the QRS obtained as a consequence of local and not distant capture. Data acquisition created an anatomic shell of the RV that allowed for identification of pace mapping sites in specific regions of the RV endocardium and epicardium. The 12-lead ECG QRS complex acquired from each pace mapping site was recorded and subsequently analyzed using the Prucka CardioLab recording system (Houston, TX, USA). The following ECG features were assessed: (1) QRS duration; (2) initial Q wave or QS complex in leads V2, I, II, III, and aVF; (3) QRS polarity amplitude ratio between II/III and between aVL/aVR (if the QRS vector was more positive or less negative in lead II or aVL, respectively, the ratio was considered ⬎1); and (4) pseudo-delta wave, intrinsicoid deflection time, and shortest RS complex values, as described in reported data that successfully identified an LV epicardial origin.7 We hypothesized that, comparing epicardial to endocardial pace mapping sites, (1) QRS duration would be increased from the epicardium; (2) initial Q waves would more likely be present from the epicardium with an initial negative QRS vector in leads I and V2 for anterior (leftward) sites and in lead II, III, or aVF for inferior RV sites; (3) epicardial sites in the RVOT region would be more leftward and thus would produce polarity ratio ⬍1 for leads II/III and aVL/aVR; and (4) reported criteria for predicting epicardial origin in the LV do not apply to the RV. Each pace mapping site had to be separated by at least 5 mm as confirmed by electroanatomic mapping in order to be included as a distinct pace mapping site. QRS duration was assessed from the pacing stimulus to the latest QRS in any of the 12 recorded leads. The reviewer was blinded to the location of the pacing site at the time of all measurements.
Epicardial access Epicardial access for activation and pace mapping was obtained using the techniques described by Sosa et al.12 After pericardial puncture, which was guided by use of contrast, a wire was deployed into the pericardial space and the needle withdrawn. A standard 8Fr sheath was introduced. A 4-mmtip CARTO was advanced through the sheath for mapping and ablation on the epicardial surface.
RV free-wall sites To facilitate the identification of region-specific ECG features that might suggest an epicardial origin, pace mapping sites in the RV were grouped into distinct anatomic segments: from the pulmonic valve to the top of the tricuspid valve (outflow tract region), top of the tricuspid valve to mid tricuspid valve (mid RV region), and mid tricuspid valve to bottom of the RV (inferior RV region). The mid and inferior RV regions were further divided into two equal portions— basal and apical—in the vertical direction. Thus, a total of five distinct anatomic segments were defined from either the endocardium or the epicardium
1133
Figure 1 Schematic right anterior oblique view of the right ventricle (RV) divided into regions to identify the site-specific influences on ECG criteria for distinguishing endocardial vs epicardial site of origin (see text for details).
for each patient (Figure 1). Of note, the anterior margin of the RV was always established from the endocardium to avoid epicardial pace mapping sites obtained from the interventricular septum. This segmentation allowed us to group regions to assess the impact of an outflow tract location (RVOT), anterior location (apical superior and apical inferior), basal location (basal superior and basal inferior), and inferior location (basal inferior and apical inferior) on ECG characteristics. We also measured the QRS duration from RV septal sites with the goal of comparing these values to the QRS duration of free-wall endocardial and epicardial pace mapping sites.
Anatomically matched RV epicardial/endocardial site analysis In order to avoid unequal weighting of certain ECG patterns related to the heterogeneous sampling of pace mapping for the RV free wall among patients and between endocardium and epicardium, a more restrictive analysis was performed comparing individually epicardial pace mapping sites with the corresponding (just opposite) endocardial pace mapping sites. The average anatomic distance between endocardial and epicardial matched sites was 17 ⫾ 6 mm, and no site was more than 25 mm apart.
Statistical analysis Continuous variables are expressed as mean ⫾ SD and were compared using a paired or unpaired Student t-test as appropriate. Categorical variables were compared using Chisquare or Fisher exact test as appropriate. P ⱕ.05 was considered significant. Sensitivity and specificity values were determined for each ECG feature that reached statistical significance in anatomically matched pace mapping analysis.
1134 Table 1
Heart Rhythm, Vol 3, No 10, October 2006 Clinical characteristics of the patient population 50 ⫾ 18 10/3 CAD: 2 DCM: 4 RV CM: 2 No SHD: 5
Age (yr) Male/female Structural heart disease
Clinical arrhythmia: Premature ventricular complex/ nonsustained VT Sustained VT Pace mapping sites performed (epicardium/endocardium) Right ventricular outflow tract Basal superior Apical superior Basal inferior Apical inferior Total
4 9
A total of 180 epicardial and 134 endocardial free-wall pace mapping sites were analyzed. The proportion of epicardial/endocardial pace mapping sites was as follows: anterior RV 84/52; inferior RV 76/43; basal RV 59/62; and RVOT 37/20. Anatomically matched epicardial and endocardial sites were identified for 94 of the 134 endocardial sites, or an average of 7.3 ⫾ 4 sites per patient. In addition, the QRS duration was assessed for 51 RV septal sites.
QRS duration
37/20 24/40 43/31 35/22 41/21 180/134
Values are presented as number or mean ⫾ SD. CAD ⫽ coronary artery disease; DCM ⫽ dilated cardiomyopathy; RV CM ⫽ right ventricular cardiomyopathy; SHD ⫽ structural heart disease; VT ⫽ ventricular tachycardia.
The QRS duration for endocardial septal sites was significantly shorter than both epicardial (154 ⫾ 23 ms vs 183 ⫾ 27 ms, P ⬍.001) and endocardial free-wall (154 ⫾ 23 ms vs 185 ⫾ 28 ms, P ⬍.001) surfaces (Table 2). However, there was no difference in QRS duration between epicardial and endocardial free-wall sites (P ⫽ .3). This remained true when only the anatomically matched RV endocardial and epicardial sites were compared (172 ⫾ 27 ms vs 176 ⫾ 24 ms, respectively P ⫽ .3). No region-specific differences in QRS duration were found.
Anterior RV
Results Baseline characteristics of the patient population are given in Table 1. From a total of 25 patients undergoing epicardial VT procedures, 13 patients (10 men and 3 women; age 50 ⫾ 18 years) underwent detailed RV endocardial and epicardial mapping and were included in the analysis. Eight of the 13 patients had structural heart disease, and the remaining seven patients presented with idiopathic premature ventricular complexes or sustained VT. Table 2
The presence of a Q wave in lead I was noted to be more frequent when pacing from the epicardial surface (30/84 [36%]) than from the endocardial free wall (5/52 [9.6%]) from RV anterior segments (P ⬍.01; Figure 2). In addition, the presence of a QS complex in lead V2 was noted from 46 (55%) of 84 epicardial vs 21 (40%) of 52 endocardial pace mapping sites (P ⫽ .1). For anatomically matched sites from the anterior RV region, the presence of Q wave in lead I was still significantly more frequent when pacing from the
Results of region-specific ECG features distinguishing RV epicardial vs endocardial pace mapping sites
No. of pace mapping sites Epi Endo QRSd (ms) Epi Endo Q lead II, III, or aVF Epi (%) Endo (%) Q lead I Epi (%) Endo (%) QS lead V2 Epi (%) Endo (%) aVL/aVR ⬍1 Epi (%) Endo (%) II/III ⬍1 Epi (%) Endo (%)
Anterior RV
Inferior RV
Basal RV
RVOT
RV all sites
84 52
76 43
59 62
37 20
180 134
178 ⫾ 24 176 ⫾ 25
190 ⫾ 30 192 ⫾ 31
197 ⫾ 28 197 ⫾ 27
173 ⫾ 18 169 ⫾ 18
183 ⫾ 26 185 ⫾ 28
42 29
78* 40
61* 32
0 5
38* 27
36* 10
18 9
5 3
24 10
23* 12
55 40
28 33
36 27
59 20
46* 31
25 23
8 9
2 10
59 40
24 19
18 21
3 7
2 5
51 30
19 15
QRS duration is presented as mean ⫾ SD. Values are presented as percentage of pace map sites from the corresponding specific RV region and surface that meet a particular criteria. *P ⱕ.05 comparing epicardial to corresponding free-wall endocardial pace mapping sites. Endo ⫽ endocardium; Epi ⫽ epicardium; QRSd ⫽ QRS complex duration; RV ⫽ right ventricle; RVOT ⫽ right ventricular outflow tract.
Bazan et al
ECG Features for Epicardial Right Ventricle VT
1135
Figure 2 Site-specific ECG features for identifying an epicardial site of origin. “Distance” ⫽ distance between epicardial and just opposite matched endocardial pace mapping site for each anatomically matched pace mapping; RV ⫽ right ventricle; RVOT ⫽ right ventricular outflow tract.
epicardium (P ⬍.001), as was a QS complex in lead V2 (P ⬍.05; Figure 3). Sensitivity and specificity values for Q wave in lead I in this region were 52% and 94%, respectively, whereas those of a QS complex in lead V2 were 67% and 61%, respectively. Both a Q wave in lead I and a QS complex in lead V2 were present in 24 (29%) of 84 epicardial pace mapping sites but in only 5 (9.6%) of 52 endocardial pace mapping sites (P ⬍.01) and did not enhance the predictive value of the observed with the individual leads.
Inferior RV The presence of a Q wave in inferior leads was noted to be more frequent when pacing from the epicardial surface than from the endocardium from the RV inferior segments. For lead II, initial Q wave was present when pacing from the epicardial surface in 53 (70%) of 76 pace mapping sites vs 16 (37%) of 43 sites from the endocardium (P ⬍.001). For lead III, the proportion was 59 (78%) of 76 vs 16 (37%) of 43, respectively (P ⬍.001). For lead aVF, an initial Q wave was present for 53 (70%) of 76 epicardial pace mapping vs 17 (40%) of 43 endocardial pace mapping sites (P ⬍.01). This remained significant for anatomically matched sites obtained from the inferior region in the RV (Figure 3). A Q wave in lead II, III, or aVF was present for 71% of epicardial inferior pace mapping sites vs 26% of pace mapping sites from the endocardium (P ⬍.001). The sensitivity and specificity values for Q wave in lead II, III, or aVF when pacing from the epicardium were 71% and 74%, respectively, from this inferior RV region.
region of the RV (P ⬍.01). However, excluding inferior sites in the basal RV (basal inferior), a Q wave in inferior leads no longer distinguished an epicardial from an endocardial origin. For basal superior RV sites, the presence of a Q wave in lead V2 was significantly more frequent when pacing from the epicardial surface: 17 (71%) of 24 from the epicardium vs 10 (25%) of 40 endocardial pace mapping sites (P ⬍.01). However, these results did not remain statistically significant after anatomically matched site analysis was performed in this RV segment (P ⫽ .25).
RV outflow tract None of the described criteria were noted to be significantly more prevalent in the RVOT region when pacing from the epicardial surface. However, leads III and aVR tended to be more positive and less negative, respectively, when pacing from the epicardium compared with endocardial pace mapping sites in this region. As a result, there was a trend for lead II/III and aVL/aVR polarity amplitude ratios to be ⬍1 from the epicardium. For aVL/aVR, 22 (59%) of 37 epicardial pace mapping sites had a ratio ⬍1 vs 8 (40%) of 20 in the free-wall endocardium (P ⫽ .1). Likewise, II/III ratio was ⬍1 if pacing was performed from the epicardium in this area: 19 (51%) of 37 from epicardium vs 6 (30%) of 20 from the endocardial surface (P ⫽ .1). When anatomically matched pace mapping sites analysis was performed in the RVOT region, these differences did not reach statistical significance.
Basal RV
Previously reported criteria for assessing epicardial origin
The presence of Q wave in lead II, III, or aVF was also more frequently observed with epicardial pace mapping in this
Region-specific differences between epicardial and endocardial surfaces with respect to the presence of a pseudo-
1136
Heart Rhythm, Vol 3, No 10, October 2006
Figure 3 Anatomically matched pace mapping sites analysis. Site-specific ECG features for predicting an epicardial ventricular tachycardia origin for anterior right ventricle (RV) (A), inferior RV (B), right ventricular outflow tract (RVOT) (C), and basal RV (D). P ⱕ.05 is considered statistically significant. Values are presented as proportion of number of pace mapping (PM) sites that meet particular criteria in that region and as mean ⫾ SD. Mean distance between endocardial and epicardial matched pace mapping is shown for every represented RV region.
delta wave, intrinsicoid deflection time, and shortest RS complex in the precordial leads were assessed. Although some significant differences between endocardial and epicardial pace map sites were seen (Table 3), none of the reported cut-off values for predicting an epicardial origin in the LV (pseudo-delta wave ⱖ34 ms, intrinsicoid deflection time ⱖ85 ms, and shortest QRS complex throughout precordial leads ⱖ121 ms) significantly applied to the RV to differentiate an epicardial origin.
Clinical outcome All 13 patients had prior unsuccessful endocardial ablation attempts. After detailed evaluation of both endocardium and epicardium, a total of six VTs were targeted for ablative
therapy from the epicardial surface in the RV in three patients. Of these six VTs, 5 (83%) were successfully ablated, with no clinical recurrence over a mean follow-up of 10 ⫾ 2 months. The remaining VT could not be abolished despite endocardial and epicardial ablation, and indeed epicardial pace mapping did not present a good QRS match of the VT. Importantly, each of the five VTs that were successfully ablated fulfilled the site-specific ECG criteria described for identifying an epicardial origin of the arrhythmia. Two of these patients presented with RV cardiomyopathy. Figure 4 shows an example of VT QRS match from both the epicardial and endocardial best pace mapping sites, fulfilling the described ECG feature for identifying an epicardial RV VT origin.
Bazan et al Table 3
ECG Features for Epicardial Right Ventricle VT
1137
Reported criteria for predicting epicardial VT origin in the LV applied to the RV
Pseudo-delta wave Epi Endo P⬍ IDT Epi Endo P⬍ RS complex Epi Endo P⬍
Anterior RV
Inferior RV
Basal RV
RVOT
RV all sites
44 ⫾ 16 42 ⫾ 14 .6
46 ⫾ 18 44 ⫾ 15 .6
54 ⫾ 17 53 ⫾ 18 .8
59 ⫾ 13 51 ⫾ 13 .03
49 ⫾ 18 49 ⫾ 17 .9
16 ⫾ 20 13 ⫾ 14 .3
36 ⫾ 30 44 ⫾ 33 .02
36 ⫾ 32 20 ⫾ 19 .01
26 ⫾ 22 17 ⫾ 14 .06
25 ⫾ 26 17 ⫾ 16 .01
85 ⫾ 17 77 ⫾ 13 .01
95 ⫾ 23 88 ⫾ 20 .1
106 ⫾ 21 99 ⫾ 14 .03
108 ⫾ 21 97 ⫾ 16 .03
95 ⫾ 25 90 ⫾ 20 .06
Values are expressed as mean ⫾ SD (in milliseconds). P ⱕ.05 is considered statistically significant. Endo ⫽ endocardium; Epi ⫽ epicardium; IDT ⫽ intrinsicoid deflection time, distance from stimulation artifact to peak of R wave in lead V2; LV ⫽ left ventricle; pseudo-delta wave ⫽ interval from stimulation artifact to earliest fast deflection in any precordial lead; RS complex ⫽ shortest distance between stimulation artifact to nadir of the first S wave in any precordial lead; RV ⫽ right ventricle; RVOT ⫽ right ventricular outflow tract. See text for details.
Discussion The present study shows that when a stimulus rises from the epicardial surface of the RV, the QRS is more likely to demonstrate initial negative forces (Q waves) in inferior leads, lead I, and/or lead V2, depending upon the region of origin. More specifically, the anterior epicardial sites
present with Q wave or QS complex in lead I and V2, and inferior RV epicardial sites show an initial Q wave in inferior leads. These observations were consistently observed overall and when comparative analysis for anatomically matched epicardial and endocardial pace mapping sites was performed. ECG characteristics that predict epicardial site of origin in the LV and aortic cusps have been defined.7,8 These features result from the slurring initial portion of the QRS complex from the epicardium. No ECG features have been described to predict an epicardial site of origin in the RV. Pacing from either endocardial and epicardial surfaces should reproduce the activation pattern present in focal arrhythmias and from exit sites in reentrant tachycardias.13,14 We previously used this technique to successfully develop ECG criteria that can accurately localize RVOT and basal LV VT.10,15 The ECG features described in the present study appear to be a manifestation of the vector of the initial wavefront of the RV activation rising from either the epicardium or the endocardium. Thus, they should apply for defining a site of origin or exit site near the endocardial or the epicardial surface for RV tachycardias.
QRS duration
Figure 4 Comparison between epicardial ventricular tachycardia (VT) QRS and epicardial vs endocardial best pace mapping (PM) sites. Sitespecific ECG feature for predicting an epicardial origin for right ventricular VT is included.
According to our results, differences in QRS width between RV epicardial and endocardial free-wall sites of origin do not exist, as has been reported for the LV.7 We hypothesize that two contributing factors may play a role in this observation. The first is related to RV wall thickness. Wall thickness in the RV is less than in the LV; thus, muscle mass might not play a significant role in delaying stimulus conduction from the epicardium. Second, the extent of the Purkinje network over the RV free wall is less than that noted from over the LV free wall. Thus, slowing of conduction may be more similar from an endocardial and epicardial RV free wall site than noted for the LV. This hy-
1138 pothesis is further supported by the significant difference in QRS duration that was observed when comparing pace mapping from the endocardium of RV free wall vs septal locations. Our study population included a subgroup of patients with RV cardiomyopathy, and, in this specific subgroup as well, QRS duration was not noted to be significantly longer when pacing from epicardial vs endocardial locations. Thus, the presence of scar in the RV does not appear to increase QRS duration from the epicardium.
Site-specific ECG criteria Ventricular arrhythmias from the basal and inferior free wall in the RV are common in the setting of RV cardiomyopathy, and radiofrequency ablative therapy is frequently indicated.16,17 Identification of ECG features that predict an epicardial site of origin is important for defining the optimum ablation strategy in this RV region. The present study shows that initial Q wave in inferior leads but not a longer QRS duration may be helpful in determining the need for an epicardial approach in order to achieve successful ablative therapy in this region of the RV. Although the anterior RV is an infrequent site of origin for ventricular arrhythmias, it also is worthwhile for identifying criteria that may suggest the need for an epicardial approach. The presence of a Q wave in lead I and QS complex in lead V2 for epicardial pace mapping from this region is more likely than from the endocardium. Epicardial sites are more anterior (leftward) than are endocardial freewall opposite sites in this region. Thus, the initial net vector is predominantly traveling away from lead I and lead V2 posteriorly or rightward, indicating the initial wavefront traveling from the epicardium to the endocardium in the anterior RV. The superior part of the RV base corresponds to the His-bundle region and its vicinity. ECG features have been defined to predict a site of origin for ventricular arrhythmias in this specific region.18 We believe that the assessment of an epicardial site of origin based on surface ECG criteria in basal superior RV requires further investigation.
RVOT region Ventricular arrhythmias frequently originate from the RVOT.19,20 A late transition in precordial leads, a broader QRS, and a less prominent and more notched R wave in lead II were shown to accurately identify septal vs free-wall sites of origin of RVOT VT.10 Joshi and Wilber21 reported similar criteria that predict a free-wall origin in the RVOT. Other ECG features predict the site of origin from the LV outflow tract and from the aortic cusps, as well as from above the pulmonary valve.22–24 No previous studies have described ECG patterns that might suggest an epicardial origin in the RV outflow tract and its vicinity, although an epicardial approach is sometimes warranted in this region.25 When pacing from the epicardium in the RVOT region, we noted a trend toward aVL/aVR and II/III QRS polarity amplitude ratios ⬍1. Without overstating the significance of this trend, the anat-
Heart Rhythm, Vol 3, No 10, October 2006 omy of the RVOT would suggest that when a stimulus arises from the epicardium, it will tend to be more leftward in origin and support our observations.
Ablation strategy for RV arrhythmias Epicardial ablation will improve the outcome of ablative therapy for ventricular arrhythmias when the endocardial approach fails to suppress the arrhythmia.20,25,26 Of note, we previously demonstrated the usefulness of irrigated-tip catheter ablation in the RV when RV cardiomyopathy is present.17 Thus, this alternative ablation strategy has to be considered before an epicardial ablation is deemed necessary. Nevertheless, sometimes epicardial ablation may be warranted, and thus the described ECG criteria are useful for predicting those cases in which epicardial ablation might be needed to guarantee a successful outcome.
Study limitations The number of pace mapping sites obtained for our study analysis may be limited, and the results may not apply for ventricular arrhythmias in the setting of RV abnormalities other than those included in our patient population, such as congenital heart disease and RV remodeling after pulmonary hypertension. Data regarding successful VT ablation from the epicardial RV surface reported here, although providing valuable preliminary confirmation of our results, are limited and require additional evaluation. Clinical arrhythmia can occur at faster rates than the pacing cycle length used in this study. As previously noted, the QRS occasionally can vary depending on the cycle length during ventricular tachycardia or ventricular pacing.10 However, differences in QRS morphology related to the rate are subtle, especially in the absence of structural heart disease. Thus, in our opinion, the described ECG features still are applicable. Inability to capture from the epicardial surface has been reported.1 However, in our series, ventricular capture was consistently observed, albeit at a higher output threshold than from the endocardial surface. Our results may not apply to patients who have more extensive LV disease, especially in the setting of coronary disease with prior infarction, which can manifest on the ECG with an initial Q wave. Furthermore, the shape of the RV can vary importantly among individuals.27 This may affect the initial vector of the wavefront during the ventricular depolarization on which we based our criteria.
Conclusion For arrhythmias that arise from the RV, the presence of an initial Q wave in lead I and QS in lead V2 for anterior sites in the RV strongly predicts an epicardial origin. Similarly, an initial Q wave in leads II, III, and aVF is observed with pace mapping from the inferior epicardial locations in the RV. Further investigation is required to determine reliable ECG criteria for suggesting an epicardial origin in the RVOT. QRS duration does not distinguish an epicardial
Bazan et al
ECG Features for Epicardial Right Ventricle VT
from an endocardial origin, nor do reported criteria for predicting epicardial origin in the LV. These data have important implications for facilitating the optimum ablation strategy for managing ventricular arrhythmias from the RV.
References 1. Schweikert RA, Saliva WI, Tomassoni G, Marrouche NF, Cole CR, Dresing TJ, Tchou PJ, Bash D, Beheiry S, Lam C, Kanagaratnam L, Natale A. Percutaneous pericardial instrumentation for endo-epicardial mapping of previously failed ablations. Circulation 2003;108:1329 –1335. 2. Garcia F, Sussman J, Bala R, Toca M, Lin D, Dixit S, Marchlinski FE. Epicardial ablation to eliminate ventricular tachycardia associated with right ventricular cardiomyopathy: characterization of the substrate (abstr). Heart Rhythm Society 27th Annual Scientific Sessions, May 2006. 3. Morwood J, Triedman J, Berul C, Khairy P, Alexander M, Cecchin F, Walsh E. Radiofrequency catheter ablation of ventricular tachycardia in children and young adults with congenital heart disease. Heart Rhythm 2004;1:301–308. 4. Sosa E, Scanavacca M, d’Avila A, Piccioni J, Sánchez O, Velarde JL, Silva M, Reolao B. Endocardial and epicardial ablation guided by nonsurgical transthoracic epicardial mapping to treat recurrent ventricular tachycardia. J Cardiovasc Electrophysiol 1998;9:229 –239. 5. Marchlinski FE, Garcia F, Siadatan A, Sauer W, Beldner S, Zado E, Hsia H, Lin D, Cooper J, Verdino R, Gerstenfeld E, Dixit S, Russo A, Callans D. Ventricular tachycardia/ventricular fibrillation in the setting of ischemic heart disease. J Cardiovasc Electrophysiol 2005;16:S59 –S70. 6. Swarup V, Morton J, Arruda M, Wilber D. Ablation of epicardial macroreentrant ventricular tachycardia associated with idiopathic nonischemic dilated cardiomyopathy by a percutaneous transthoracic approach. J Cardiovasc Electrophysiol 2002;13:1164 –1168. 7. Berruezo A, Mont L, Nava S, Chueca E, Bartholomay E, Brugada J. Electrocardiographic recognition of the epicardial origin of ventricular tachycardias. Circulation 2004;109:1842–1847. 8. Daniels DV, Lu Y-Y, Morton JB, Santucci PA, Akar JG, Green A, Wilber DJ. Idiopathic epicardial left ventricular tachycardia originating remote from the sinus of Valsalva. Electrophysiologic characteristics, catheter ablation, and identification from the 12-lead electrocardiogram. Circulation 2006;113:1659 –1666. 9. Josephson ME, Callans D. Using the twelve-lead electrocardiogram to localize the site of origin of ventricular tachycardia. Heart Rhythm 2005;2:443– 446. 10. Dixit S, Gerstenfeld E, Callans D, Marchlinski FE. Electrocardiographic patterns of superior right ventricular outflow tract tachycardias: distinguishing septal and free-wall sites of origin. J Cardiovasc Electrophysiol 2003;14:1–7. 11. Tada H, Ito S, Naito S, Kurosaki K, Ueda M, Shinbo G, Hoshizaki H, Oshima S, Nogami A, Taniguchi. Prevalence and electrocardiographic characteristics of idiopathic arrhythmia originating in the free wall of the right ventricular outflow tract. Circ J 2004;68:909 –914. 12. Sosa E, Scanavacca M, d’Avila A, Pillegi F. A new technique to perform epicardial mapping in the electrophysiology laboratory. J Cardiovasc Electrophysiol 1996;7:531–539.
1139 13. Macchi E, Baruffi S, Rossi S, Miragoli M, Bertuzzi A, Musso E, Corradi D, Di Gregorio F. Does cardiac pacing reproduce the mechanism of focal impulse initiation? J Electrocardiol 2004;37(S):135–143. 14. Gerstenfeld E, Dixit S, Callans D, Rajawat Y, Rho R, Marchlinski FE. Quantitative comparison of spontaneous and paced 12-lead electrocardiogram during right ventricular outflow tract ventricular tachycardia. J Am Coll Cardiol 2003; 41:2046 –2053. 15. Dixit S, Gerstenfeld E, Lin D, Callans DJ, Hsia HH, Nayak HM, Zado E, Marchlinski, FE. Identification of distinct electrocardiographic patterns from the basal left ventricle: distinguishing medial and lateral sites in patients with idiopathic ventricular tachycardia. Heart Rhythm 2005;2:485– 491. 16. O’Donell D, Cox D, Bourke J, Mitchell L, Furniss S. Clinical and electrophysiological differences between patients with arrhythmogenic right ventricular dysplasia and right ventricular outflow tract tachycardia. Eur Heart J 2003;24: 801– 810. 17. Marchlinski FE, Zado E, Dixit S, Gerstenfeld E, Callans D, Hsia H, Lin D, Nayak H, Russo A, Pulliam W. Electroanatomic substrate and outcome of catheter ablative therapy for ventricular tachycardia in setting of right ventricular cardiomyopathy. Circulation 2004;110:2293–2298. 18. Yamauchi Y, Aonuma K, Takahashi A, Seikiguchi Y, Hachiya H, Yokoyama Y, Kumagai K, Nogami A, Iesaka Y, Isobe M. Electrocardiographic characteristics of repetitive monomorphic right ventricular tachycardia originated near the His-bundle. J Cardiovasc Electrophysiol 2005;16:1041–1048. 19. Buxton A, Waxman HL, Marchlinski FE, Simson MB, Cassidy D, Josephson ME. Right ventricular tachycardia: clinical and electrophysiologic characteristics. Circulation 1983;68:917–927. 20. Tsai C-F, Chen S-A, Tai C-T, Chiang C-E, Lee S-H, Wen Z-C, Huang J-L, Ding Y-A, Chang M-S. Idiopathic monomorphic ventricular tachycardia: clinical outcome, electrophysiologic characteristics and long-term results of catheter ablation. Int J Cardiol 1997;62:143–150. 21. Joshi S, Wilber D. Ablation of right ventricular outflow tract tachycardia. Current perspectives. J Cardiovasc Electrophysiol 2005;16(S1):S52–S58. 22. Callans D, Menz V, Schwartzman D, Gottlieb C, Marchlinski FE. Repetitive monomorphic tachycardia from the left ventricular outflow tract: electrocardiographic patterns consistent with a left ventricular site of origin. J Am Coll Cardiol 1997;29:1023–1027. 23. Hachiya H, Aonuma K, Yamauchi Y, Igawa M, Nogami A, Iesaka Y. How to diagnose, locate and ablate coronary cusp ventricular tachycardia. J Cardiovasc Electrophysiol 2002;13:551–556. 24. Timmermans C, Rodriguez L-M, Crijns H, Moorman A, Wellens H. Idiopathic left bundle-branch block-shaped ventricular tachycardia may originate above the pulmonary valve. Circulation 2003;108:1960 –1967. 25. Tanner H, Hindricks G, Schirdewahn P, Kobza R, Dorszewski A, Piorkorwski C, Gerds-Li J-H, Kottkamp H. Outflow tract tachycardia with R/S transition in lead V3. J Am Coll Cardiol 2005;45:418 – 423. 26. Zei P, Stevenson W. Epicardial catheter mapping and ablation of ventricular tachycardia. Heart Rhythm 2006;3:360 –363. 27. Fritz J, Solaiyappan M, Tandri H, Bomma C, Genc A, Clausen C, Lima J, Bluemke D. Right ventricle shape and contraction patterns and relation to magnetic resonance imaging findings. J Comput Assist Tomogr 2005;29: 725–733.
vs endocardial sites (53/73 vs 16/43, P ⬍.01). A Q wave in lead I was more frequently present from epicardial vs endocardial anterior RV sites (30/82 vs 5/52, P ⬍.001). QS in lead V2 was noted from anatomically matched epicardial anterior RV sites (22/33 vs 13/33, P ⬍.05). In the RV outflow tract, no ECG feature distinguishing epicardial/endocardial origin reached statistical significance. CONCLUSION A Q wave or QS in leads that best reflect local activation suggest an epicardial origin for RV depolarization and may help in identifying a probable epicardial site of origin for RV VT. QRS duration and reported criteria for epicardial origin of VT in the left ventricle do not identify a probable epicardial origin in the RV. KEYWORDS Epicardium; Ventricular tachycardia; Pace mapping; Right ventricle (Heart Rhythm 2006;3:1132–1139) © 2006 Heart Rhythm Society. All rights reserved.
Introduction
Methods
Successful catheter ablation for both left ventricular (LV) and right ventricular (RV) tachycardia (VT) sometimes requires an epicardial approach.1– 6 Specific 12-lead ECG patterns have been described as helpful for identifying an epicardial VT origin in the LV, including pseudo-delta wave ⱖ34 ms, intrinsicoid deflection time ⱖ85 ms, RS complex duration ⱖ121 ms, and delayed precordial maximum deflection index ⱖ0.55.7,8 The 12-lead ECG is useful for identifying ventricular arrhythmias from specific regions of the RV endocardium, especially from the RV outflow tract (RVOT).9 –11 However, its usefulness in predicting an epicardial site of origin in the RV has not been established. Thus, characterizing ECG patterns to predict an epicardial origin in the RV is of interest. We sought to describe ECG features that predict an epicardial site of origin for RV arrhythmias by mimicking the origin of VT with RV endocardial and epicardial pace mapping.
Patient population
Address reprint requests and correspondence: Dr. Francis E. Marchlinski, Hospital of the University of Pennsylvania, 9 Founders Pavilion, 3400 Spruce Street, Philadelphia, Pennsylvania 19104. E-mail address: [email protected]. (Received April 27, 2006; accepted June 23, 2006.)
Twenty-five patients underwent endocardial and epicardial catheter mapping and ablation for drug-refractory ventricular arrhythmias. In each patient, the risks of mapping/ablation were discussed in detail, and all patients gave written informed consent in accordance with institutional guidelines of the University of Pennsylvania Health System. Each patient underwent detailed voltage mapping, pace mapping, and activation mapping in an attempt to characterize the VT substrate and target/ablate their arrhythmia.
Study protocol To be included in this study analysis, each patient was required to undergo detailed pace mapping from both endocardial and epicardial free-wall surfaces of the RV. Each pace mapping site was tagged on an electroanatomic map (CARTO, Biosense Webster, Diamond Bar, CA, USA) of the RV chamber of interest for detailed offline analysis. A detailed electroanatomic map in sinus rhythm was performed using a 4-mm-tip catheter maintaining a fill threshold of 20 mm, with the entire surface of the RV endocardium and epicardium documented to be sampled. Pace mapping was performed using bipolar pacing (distance between poles 1 and 2 ⫽ 1 mm) at threshold from both
1547-5271/$ -see front matter © 2006 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2006.06.024
Bazan et al
ECG Features for Epicardial Right Ventricle VT
epicardial and endocardial free-wall surfaces at cycle lengths of 400 to 600 ms. This allowed for analysis of the QRS obtained as a consequence of local and not distant capture. Data acquisition created an anatomic shell of the RV that allowed for identification of pace mapping sites in specific regions of the RV endocardium and epicardium. The 12-lead ECG QRS complex acquired from each pace mapping site was recorded and subsequently analyzed using the Prucka CardioLab recording system (Houston, TX, USA). The following ECG features were assessed: (1) QRS duration; (2) initial Q wave or QS complex in leads V2, I, II, III, and aVF; (3) QRS polarity amplitude ratio between II/III and between aVL/aVR (if the QRS vector was more positive or less negative in lead II or aVL, respectively, the ratio was considered ⬎1); and (4) pseudo-delta wave, intrinsicoid deflection time, and shortest RS complex values, as described in reported data that successfully identified an LV epicardial origin.7 We hypothesized that, comparing epicardial to endocardial pace mapping sites, (1) QRS duration would be increased from the epicardium; (2) initial Q waves would more likely be present from the epicardium with an initial negative QRS vector in leads I and V2 for anterior (leftward) sites and in lead II, III, or aVF for inferior RV sites; (3) epicardial sites in the RVOT region would be more leftward and thus would produce polarity ratio ⬍1 for leads II/III and aVL/aVR; and (4) reported criteria for predicting epicardial origin in the LV do not apply to the RV. Each pace mapping site had to be separated by at least 5 mm as confirmed by electroanatomic mapping in order to be included as a distinct pace mapping site. QRS duration was assessed from the pacing stimulus to the latest QRS in any of the 12 recorded leads. The reviewer was blinded to the location of the pacing site at the time of all measurements.
Epicardial access Epicardial access for activation and pace mapping was obtained using the techniques described by Sosa et al.12 After pericardial puncture, which was guided by use of contrast, a wire was deployed into the pericardial space and the needle withdrawn. A standard 8Fr sheath was introduced. A 4-mmtip CARTO was advanced through the sheath for mapping and ablation on the epicardial surface.
RV free-wall sites To facilitate the identification of region-specific ECG features that might suggest an epicardial origin, pace mapping sites in the RV were grouped into distinct anatomic segments: from the pulmonic valve to the top of the tricuspid valve (outflow tract region), top of the tricuspid valve to mid tricuspid valve (mid RV region), and mid tricuspid valve to bottom of the RV (inferior RV region). The mid and inferior RV regions were further divided into two equal portions— basal and apical—in the vertical direction. Thus, a total of five distinct anatomic segments were defined from either the endocardium or the epicardium
1133
Figure 1 Schematic right anterior oblique view of the right ventricle (RV) divided into regions to identify the site-specific influences on ECG criteria for distinguishing endocardial vs epicardial site of origin (see text for details).
for each patient (Figure 1). Of note, the anterior margin of the RV was always established from the endocardium to avoid epicardial pace mapping sites obtained from the interventricular septum. This segmentation allowed us to group regions to assess the impact of an outflow tract location (RVOT), anterior location (apical superior and apical inferior), basal location (basal superior and basal inferior), and inferior location (basal inferior and apical inferior) on ECG characteristics. We also measured the QRS duration from RV septal sites with the goal of comparing these values to the QRS duration of free-wall endocardial and epicardial pace mapping sites.
Anatomically matched RV epicardial/endocardial site analysis In order to avoid unequal weighting of certain ECG patterns related to the heterogeneous sampling of pace mapping for the RV free wall among patients and between endocardium and epicardium, a more restrictive analysis was performed comparing individually epicardial pace mapping sites with the corresponding (just opposite) endocardial pace mapping sites. The average anatomic distance between endocardial and epicardial matched sites was 17 ⫾ 6 mm, and no site was more than 25 mm apart.
Statistical analysis Continuous variables are expressed as mean ⫾ SD and were compared using a paired or unpaired Student t-test as appropriate. Categorical variables were compared using Chisquare or Fisher exact test as appropriate. P ⱕ.05 was considered significant. Sensitivity and specificity values were determined for each ECG feature that reached statistical significance in anatomically matched pace mapping analysis.
1134 Table 1
Heart Rhythm, Vol 3, No 10, October 2006 Clinical characteristics of the patient population 50 ⫾ 18 10/3 CAD: 2 DCM: 4 RV CM: 2 No SHD: 5
Age (yr) Male/female Structural heart disease
Clinical arrhythmia: Premature ventricular complex/ nonsustained VT Sustained VT Pace mapping sites performed (epicardium/endocardium) Right ventricular outflow tract Basal superior Apical superior Basal inferior Apical inferior Total
4 9
A total of 180 epicardial and 134 endocardial free-wall pace mapping sites were analyzed. The proportion of epicardial/endocardial pace mapping sites was as follows: anterior RV 84/52; inferior RV 76/43; basal RV 59/62; and RVOT 37/20. Anatomically matched epicardial and endocardial sites were identified for 94 of the 134 endocardial sites, or an average of 7.3 ⫾ 4 sites per patient. In addition, the QRS duration was assessed for 51 RV septal sites.
QRS duration
37/20 24/40 43/31 35/22 41/21 180/134
Values are presented as number or mean ⫾ SD. CAD ⫽ coronary artery disease; DCM ⫽ dilated cardiomyopathy; RV CM ⫽ right ventricular cardiomyopathy; SHD ⫽ structural heart disease; VT ⫽ ventricular tachycardia.
The QRS duration for endocardial septal sites was significantly shorter than both epicardial (154 ⫾ 23 ms vs 183 ⫾ 27 ms, P ⬍.001) and endocardial free-wall (154 ⫾ 23 ms vs 185 ⫾ 28 ms, P ⬍.001) surfaces (Table 2). However, there was no difference in QRS duration between epicardial and endocardial free-wall sites (P ⫽ .3). This remained true when only the anatomically matched RV endocardial and epicardial sites were compared (172 ⫾ 27 ms vs 176 ⫾ 24 ms, respectively P ⫽ .3). No region-specific differences in QRS duration were found.
Anterior RV
Results Baseline characteristics of the patient population are given in Table 1. From a total of 25 patients undergoing epicardial VT procedures, 13 patients (10 men and 3 women; age 50 ⫾ 18 years) underwent detailed RV endocardial and epicardial mapping and were included in the analysis. Eight of the 13 patients had structural heart disease, and the remaining seven patients presented with idiopathic premature ventricular complexes or sustained VT. Table 2
The presence of a Q wave in lead I was noted to be more frequent when pacing from the epicardial surface (30/84 [36%]) than from the endocardial free wall (5/52 [9.6%]) from RV anterior segments (P ⬍.01; Figure 2). In addition, the presence of a QS complex in lead V2 was noted from 46 (55%) of 84 epicardial vs 21 (40%) of 52 endocardial pace mapping sites (P ⫽ .1). For anatomically matched sites from the anterior RV region, the presence of Q wave in lead I was still significantly more frequent when pacing from the
Results of region-specific ECG features distinguishing RV epicardial vs endocardial pace mapping sites
No. of pace mapping sites Epi Endo QRSd (ms) Epi Endo Q lead II, III, or aVF Epi (%) Endo (%) Q lead I Epi (%) Endo (%) QS lead V2 Epi (%) Endo (%) aVL/aVR ⬍1 Epi (%) Endo (%) II/III ⬍1 Epi (%) Endo (%)
Anterior RV
Inferior RV
Basal RV
RVOT
RV all sites
84 52
76 43
59 62
37 20
180 134
178 ⫾ 24 176 ⫾ 25
190 ⫾ 30 192 ⫾ 31
197 ⫾ 28 197 ⫾ 27
173 ⫾ 18 169 ⫾ 18
183 ⫾ 26 185 ⫾ 28
42 29
78* 40
61* 32
0 5
38* 27
36* 10
18 9
5 3
24 10
23* 12
55 40
28 33
36 27
59 20
46* 31
25 23
8 9
2 10
59 40
24 19
18 21
3 7
2 5
51 30
19 15
QRS duration is presented as mean ⫾ SD. Values are presented as percentage of pace map sites from the corresponding specific RV region and surface that meet a particular criteria. *P ⱕ.05 comparing epicardial to corresponding free-wall endocardial pace mapping sites. Endo ⫽ endocardium; Epi ⫽ epicardium; QRSd ⫽ QRS complex duration; RV ⫽ right ventricle; RVOT ⫽ right ventricular outflow tract.
Bazan et al
ECG Features for Epicardial Right Ventricle VT
1135
Figure 2 Site-specific ECG features for identifying an epicardial site of origin. “Distance” ⫽ distance between epicardial and just opposite matched endocardial pace mapping site for each anatomically matched pace mapping; RV ⫽ right ventricle; RVOT ⫽ right ventricular outflow tract.
epicardium (P ⬍.001), as was a QS complex in lead V2 (P ⬍.05; Figure 3). Sensitivity and specificity values for Q wave in lead I in this region were 52% and 94%, respectively, whereas those of a QS complex in lead V2 were 67% and 61%, respectively. Both a Q wave in lead I and a QS complex in lead V2 were present in 24 (29%) of 84 epicardial pace mapping sites but in only 5 (9.6%) of 52 endocardial pace mapping sites (P ⬍.01) and did not enhance the predictive value of the observed with the individual leads.
Inferior RV The presence of a Q wave in inferior leads was noted to be more frequent when pacing from the epicardial surface than from the endocardium from the RV inferior segments. For lead II, initial Q wave was present when pacing from the epicardial surface in 53 (70%) of 76 pace mapping sites vs 16 (37%) of 43 sites from the endocardium (P ⬍.001). For lead III, the proportion was 59 (78%) of 76 vs 16 (37%) of 43, respectively (P ⬍.001). For lead aVF, an initial Q wave was present for 53 (70%) of 76 epicardial pace mapping vs 17 (40%) of 43 endocardial pace mapping sites (P ⬍.01). This remained significant for anatomically matched sites obtained from the inferior region in the RV (Figure 3). A Q wave in lead II, III, or aVF was present for 71% of epicardial inferior pace mapping sites vs 26% of pace mapping sites from the endocardium (P ⬍.001). The sensitivity and specificity values for Q wave in lead II, III, or aVF when pacing from the epicardium were 71% and 74%, respectively, from this inferior RV region.
region of the RV (P ⬍.01). However, excluding inferior sites in the basal RV (basal inferior), a Q wave in inferior leads no longer distinguished an epicardial from an endocardial origin. For basal superior RV sites, the presence of a Q wave in lead V2 was significantly more frequent when pacing from the epicardial surface: 17 (71%) of 24 from the epicardium vs 10 (25%) of 40 endocardial pace mapping sites (P ⬍.01). However, these results did not remain statistically significant after anatomically matched site analysis was performed in this RV segment (P ⫽ .25).
RV outflow tract None of the described criteria were noted to be significantly more prevalent in the RVOT region when pacing from the epicardial surface. However, leads III and aVR tended to be more positive and less negative, respectively, when pacing from the epicardium compared with endocardial pace mapping sites in this region. As a result, there was a trend for lead II/III and aVL/aVR polarity amplitude ratios to be ⬍1 from the epicardium. For aVL/aVR, 22 (59%) of 37 epicardial pace mapping sites had a ratio ⬍1 vs 8 (40%) of 20 in the free-wall endocardium (P ⫽ .1). Likewise, II/III ratio was ⬍1 if pacing was performed from the epicardium in this area: 19 (51%) of 37 from epicardium vs 6 (30%) of 20 from the endocardial surface (P ⫽ .1). When anatomically matched pace mapping sites analysis was performed in the RVOT region, these differences did not reach statistical significance.
Basal RV
Previously reported criteria for assessing epicardial origin
The presence of Q wave in lead II, III, or aVF was also more frequently observed with epicardial pace mapping in this
Region-specific differences between epicardial and endocardial surfaces with respect to the presence of a pseudo-
1136
Heart Rhythm, Vol 3, No 10, October 2006
Figure 3 Anatomically matched pace mapping sites analysis. Site-specific ECG features for predicting an epicardial ventricular tachycardia origin for anterior right ventricle (RV) (A), inferior RV (B), right ventricular outflow tract (RVOT) (C), and basal RV (D). P ⱕ.05 is considered statistically significant. Values are presented as proportion of number of pace mapping (PM) sites that meet particular criteria in that region and as mean ⫾ SD. Mean distance between endocardial and epicardial matched pace mapping is shown for every represented RV region.
delta wave, intrinsicoid deflection time, and shortest RS complex in the precordial leads were assessed. Although some significant differences between endocardial and epicardial pace map sites were seen (Table 3), none of the reported cut-off values for predicting an epicardial origin in the LV (pseudo-delta wave ⱖ34 ms, intrinsicoid deflection time ⱖ85 ms, and shortest QRS complex throughout precordial leads ⱖ121 ms) significantly applied to the RV to differentiate an epicardial origin.
Clinical outcome All 13 patients had prior unsuccessful endocardial ablation attempts. After detailed evaluation of both endocardium and epicardium, a total of six VTs were targeted for ablative
therapy from the epicardial surface in the RV in three patients. Of these six VTs, 5 (83%) were successfully ablated, with no clinical recurrence over a mean follow-up of 10 ⫾ 2 months. The remaining VT could not be abolished despite endocardial and epicardial ablation, and indeed epicardial pace mapping did not present a good QRS match of the VT. Importantly, each of the five VTs that were successfully ablated fulfilled the site-specific ECG criteria described for identifying an epicardial origin of the arrhythmia. Two of these patients presented with RV cardiomyopathy. Figure 4 shows an example of VT QRS match from both the epicardial and endocardial best pace mapping sites, fulfilling the described ECG feature for identifying an epicardial RV VT origin.
Bazan et al Table 3
ECG Features for Epicardial Right Ventricle VT
1137
Reported criteria for predicting epicardial VT origin in the LV applied to the RV
Pseudo-delta wave Epi Endo P⬍ IDT Epi Endo P⬍ RS complex Epi Endo P⬍
Anterior RV
Inferior RV
Basal RV
RVOT
RV all sites
44 ⫾ 16 42 ⫾ 14 .6
46 ⫾ 18 44 ⫾ 15 .6
54 ⫾ 17 53 ⫾ 18 .8
59 ⫾ 13 51 ⫾ 13 .03
49 ⫾ 18 49 ⫾ 17 .9
16 ⫾ 20 13 ⫾ 14 .3
36 ⫾ 30 44 ⫾ 33 .02
36 ⫾ 32 20 ⫾ 19 .01
26 ⫾ 22 17 ⫾ 14 .06
25 ⫾ 26 17 ⫾ 16 .01
85 ⫾ 17 77 ⫾ 13 .01
95 ⫾ 23 88 ⫾ 20 .1
106 ⫾ 21 99 ⫾ 14 .03
108 ⫾ 21 97 ⫾ 16 .03
95 ⫾ 25 90 ⫾ 20 .06
Values are expressed as mean ⫾ SD (in milliseconds). P ⱕ.05 is considered statistically significant. Endo ⫽ endocardium; Epi ⫽ epicardium; IDT ⫽ intrinsicoid deflection time, distance from stimulation artifact to peak of R wave in lead V2; LV ⫽ left ventricle; pseudo-delta wave ⫽ interval from stimulation artifact to earliest fast deflection in any precordial lead; RS complex ⫽ shortest distance between stimulation artifact to nadir of the first S wave in any precordial lead; RV ⫽ right ventricle; RVOT ⫽ right ventricular outflow tract. See text for details.
Discussion The present study shows that when a stimulus rises from the epicardial surface of the RV, the QRS is more likely to demonstrate initial negative forces (Q waves) in inferior leads, lead I, and/or lead V2, depending upon the region of origin. More specifically, the anterior epicardial sites
present with Q wave or QS complex in lead I and V2, and inferior RV epicardial sites show an initial Q wave in inferior leads. These observations were consistently observed overall and when comparative analysis for anatomically matched epicardial and endocardial pace mapping sites was performed. ECG characteristics that predict epicardial site of origin in the LV and aortic cusps have been defined.7,8 These features result from the slurring initial portion of the QRS complex from the epicardium. No ECG features have been described to predict an epicardial site of origin in the RV. Pacing from either endocardial and epicardial surfaces should reproduce the activation pattern present in focal arrhythmias and from exit sites in reentrant tachycardias.13,14 We previously used this technique to successfully develop ECG criteria that can accurately localize RVOT and basal LV VT.10,15 The ECG features described in the present study appear to be a manifestation of the vector of the initial wavefront of the RV activation rising from either the epicardium or the endocardium. Thus, they should apply for defining a site of origin or exit site near the endocardial or the epicardial surface for RV tachycardias.
QRS duration
Figure 4 Comparison between epicardial ventricular tachycardia (VT) QRS and epicardial vs endocardial best pace mapping (PM) sites. Sitespecific ECG feature for predicting an epicardial origin for right ventricular VT is included.
According to our results, differences in QRS width between RV epicardial and endocardial free-wall sites of origin do not exist, as has been reported for the LV.7 We hypothesize that two contributing factors may play a role in this observation. The first is related to RV wall thickness. Wall thickness in the RV is less than in the LV; thus, muscle mass might not play a significant role in delaying stimulus conduction from the epicardium. Second, the extent of the Purkinje network over the RV free wall is less than that noted from over the LV free wall. Thus, slowing of conduction may be more similar from an endocardial and epicardial RV free wall site than noted for the LV. This hy-
1138 pothesis is further supported by the significant difference in QRS duration that was observed when comparing pace mapping from the endocardium of RV free wall vs septal locations. Our study population included a subgroup of patients with RV cardiomyopathy, and, in this specific subgroup as well, QRS duration was not noted to be significantly longer when pacing from epicardial vs endocardial locations. Thus, the presence of scar in the RV does not appear to increase QRS duration from the epicardium.
Site-specific ECG criteria Ventricular arrhythmias from the basal and inferior free wall in the RV are common in the setting of RV cardiomyopathy, and radiofrequency ablative therapy is frequently indicated.16,17 Identification of ECG features that predict an epicardial site of origin is important for defining the optimum ablation strategy in this RV region. The present study shows that initial Q wave in inferior leads but not a longer QRS duration may be helpful in determining the need for an epicardial approach in order to achieve successful ablative therapy in this region of the RV. Although the anterior RV is an infrequent site of origin for ventricular arrhythmias, it also is worthwhile for identifying criteria that may suggest the need for an epicardial approach. The presence of a Q wave in lead I and QS complex in lead V2 for epicardial pace mapping from this region is more likely than from the endocardium. Epicardial sites are more anterior (leftward) than are endocardial freewall opposite sites in this region. Thus, the initial net vector is predominantly traveling away from lead I and lead V2 posteriorly or rightward, indicating the initial wavefront traveling from the epicardium to the endocardium in the anterior RV. The superior part of the RV base corresponds to the His-bundle region and its vicinity. ECG features have been defined to predict a site of origin for ventricular arrhythmias in this specific region.18 We believe that the assessment of an epicardial site of origin based on surface ECG criteria in basal superior RV requires further investigation.
RVOT region Ventricular arrhythmias frequently originate from the RVOT.19,20 A late transition in precordial leads, a broader QRS, and a less prominent and more notched R wave in lead II were shown to accurately identify septal vs free-wall sites of origin of RVOT VT.10 Joshi and Wilber21 reported similar criteria that predict a free-wall origin in the RVOT. Other ECG features predict the site of origin from the LV outflow tract and from the aortic cusps, as well as from above the pulmonary valve.22–24 No previous studies have described ECG patterns that might suggest an epicardial origin in the RV outflow tract and its vicinity, although an epicardial approach is sometimes warranted in this region.25 When pacing from the epicardium in the RVOT region, we noted a trend toward aVL/aVR and II/III QRS polarity amplitude ratios ⬍1. Without overstating the significance of this trend, the anat-
Heart Rhythm, Vol 3, No 10, October 2006 omy of the RVOT would suggest that when a stimulus arises from the epicardium, it will tend to be more leftward in origin and support our observations.
Ablation strategy for RV arrhythmias Epicardial ablation will improve the outcome of ablative therapy for ventricular arrhythmias when the endocardial approach fails to suppress the arrhythmia.20,25,26 Of note, we previously demonstrated the usefulness of irrigated-tip catheter ablation in the RV when RV cardiomyopathy is present.17 Thus, this alternative ablation strategy has to be considered before an epicardial ablation is deemed necessary. Nevertheless, sometimes epicardial ablation may be warranted, and thus the described ECG criteria are useful for predicting those cases in which epicardial ablation might be needed to guarantee a successful outcome.
Study limitations The number of pace mapping sites obtained for our study analysis may be limited, and the results may not apply for ventricular arrhythmias in the setting of RV abnormalities other than those included in our patient population, such as congenital heart disease and RV remodeling after pulmonary hypertension. Data regarding successful VT ablation from the epicardial RV surface reported here, although providing valuable preliminary confirmation of our results, are limited and require additional evaluation. Clinical arrhythmia can occur at faster rates than the pacing cycle length used in this study. As previously noted, the QRS occasionally can vary depending on the cycle length during ventricular tachycardia or ventricular pacing.10 However, differences in QRS morphology related to the rate are subtle, especially in the absence of structural heart disease. Thus, in our opinion, the described ECG features still are applicable. Inability to capture from the epicardial surface has been reported.1 However, in our series, ventricular capture was consistently observed, albeit at a higher output threshold than from the endocardial surface. Our results may not apply to patients who have more extensive LV disease, especially in the setting of coronary disease with prior infarction, which can manifest on the ECG with an initial Q wave. Furthermore, the shape of the RV can vary importantly among individuals.27 This may affect the initial vector of the wavefront during the ventricular depolarization on which we based our criteria.
Conclusion For arrhythmias that arise from the RV, the presence of an initial Q wave in lead I and QS in lead V2 for anterior sites in the RV strongly predicts an epicardial origin. Similarly, an initial Q wave in leads II, III, and aVF is observed with pace mapping from the inferior epicardial locations in the RV. Further investigation is required to determine reliable ECG criteria for suggesting an epicardial origin in the RVOT. QRS duration does not distinguish an epicardial
Bazan et al
ECG Features for Epicardial Right Ventricle VT
from an endocardial origin, nor do reported criteria for predicting epicardial origin in the LV. These data have important implications for facilitating the optimum ablation strategy for managing ventricular arrhythmias from the RV.
References 1. Schweikert RA, Saliva WI, Tomassoni G, Marrouche NF, Cole CR, Dresing TJ, Tchou PJ, Bash D, Beheiry S, Lam C, Kanagaratnam L, Natale A. Percutaneous pericardial instrumentation for endo-epicardial mapping of previously failed ablations. Circulation 2003;108:1329 –1335. 2. Garcia F, Sussman J, Bala R, Toca M, Lin D, Dixit S, Marchlinski FE. Epicardial ablation to eliminate ventricular tachycardia associated with right ventricular cardiomyopathy: characterization of the substrate (abstr). Heart Rhythm Society 27th Annual Scientific Sessions, May 2006. 3. Morwood J, Triedman J, Berul C, Khairy P, Alexander M, Cecchin F, Walsh E. Radiofrequency catheter ablation of ventricular tachycardia in children and young adults with congenital heart disease. Heart Rhythm 2004;1:301–308. 4. Sosa E, Scanavacca M, d’Avila A, Piccioni J, Sánchez O, Velarde JL, Silva M, Reolao B. Endocardial and epicardial ablation guided by nonsurgical transthoracic epicardial mapping to treat recurrent ventricular tachycardia. J Cardiovasc Electrophysiol 1998;9:229 –239. 5. Marchlinski FE, Garcia F, Siadatan A, Sauer W, Beldner S, Zado E, Hsia H, Lin D, Cooper J, Verdino R, Gerstenfeld E, Dixit S, Russo A, Callans D. Ventricular tachycardia/ventricular fibrillation in the setting of ischemic heart disease. J Cardiovasc Electrophysiol 2005;16:S59 –S70. 6. Swarup V, Morton J, Arruda M, Wilber D. Ablation of epicardial macroreentrant ventricular tachycardia associated with idiopathic nonischemic dilated cardiomyopathy by a percutaneous transthoracic approach. J Cardiovasc Electrophysiol 2002;13:1164 –1168. 7. Berruezo A, Mont L, Nava S, Chueca E, Bartholomay E, Brugada J. Electrocardiographic recognition of the epicardial origin of ventricular tachycardias. Circulation 2004;109:1842–1847. 8. Daniels DV, Lu Y-Y, Morton JB, Santucci PA, Akar JG, Green A, Wilber DJ. Idiopathic epicardial left ventricular tachycardia originating remote from the sinus of Valsalva. Electrophysiologic characteristics, catheter ablation, and identification from the 12-lead electrocardiogram. Circulation 2006;113:1659 –1666. 9. Josephson ME, Callans D. Using the twelve-lead electrocardiogram to localize the site of origin of ventricular tachycardia. Heart Rhythm 2005;2:443– 446. 10. Dixit S, Gerstenfeld E, Callans D, Marchlinski FE. Electrocardiographic patterns of superior right ventricular outflow tract tachycardias: distinguishing septal and free-wall sites of origin. J Cardiovasc Electrophysiol 2003;14:1–7. 11. Tada H, Ito S, Naito S, Kurosaki K, Ueda M, Shinbo G, Hoshizaki H, Oshima S, Nogami A, Taniguchi. Prevalence and electrocardiographic characteristics of idiopathic arrhythmia originating in the free wall of the right ventricular outflow tract. Circ J 2004;68:909 –914. 12. Sosa E, Scanavacca M, d’Avila A, Pillegi F. A new technique to perform epicardial mapping in the electrophysiology laboratory. J Cardiovasc Electrophysiol 1996;7:531–539.
1139 13. Macchi E, Baruffi S, Rossi S, Miragoli M, Bertuzzi A, Musso E, Corradi D, Di Gregorio F. Does cardiac pacing reproduce the mechanism of focal impulse initiation? J Electrocardiol 2004;37(S):135–143. 14. Gerstenfeld E, Dixit S, Callans D, Rajawat Y, Rho R, Marchlinski FE. Quantitative comparison of spontaneous and paced 12-lead electrocardiogram during right ventricular outflow tract ventricular tachycardia. J Am Coll Cardiol 2003; 41:2046 –2053. 15. Dixit S, Gerstenfeld E, Lin D, Callans DJ, Hsia HH, Nayak HM, Zado E, Marchlinski, FE. Identification of distinct electrocardiographic patterns from the basal left ventricle: distinguishing medial and lateral sites in patients with idiopathic ventricular tachycardia. Heart Rhythm 2005;2:485– 491. 16. O’Donell D, Cox D, Bourke J, Mitchell L, Furniss S. Clinical and electrophysiological differences between patients with arrhythmogenic right ventricular dysplasia and right ventricular outflow tract tachycardia. Eur Heart J 2003;24: 801– 810. 17. Marchlinski FE, Zado E, Dixit S, Gerstenfeld E, Callans D, Hsia H, Lin D, Nayak H, Russo A, Pulliam W. Electroanatomic substrate and outcome of catheter ablative therapy for ventricular tachycardia in setting of right ventricular cardiomyopathy. Circulation 2004;110:2293–2298. 18. Yamauchi Y, Aonuma K, Takahashi A, Seikiguchi Y, Hachiya H, Yokoyama Y, Kumagai K, Nogami A, Iesaka Y, Isobe M. Electrocardiographic characteristics of repetitive monomorphic right ventricular tachycardia originated near the His-bundle. J Cardiovasc Electrophysiol 2005;16:1041–1048. 19. Buxton A, Waxman HL, Marchlinski FE, Simson MB, Cassidy D, Josephson ME. Right ventricular tachycardia: clinical and electrophysiologic characteristics. Circulation 1983;68:917–927. 20. Tsai C-F, Chen S-A, Tai C-T, Chiang C-E, Lee S-H, Wen Z-C, Huang J-L, Ding Y-A, Chang M-S. Idiopathic monomorphic ventricular tachycardia: clinical outcome, electrophysiologic characteristics and long-term results of catheter ablation. Int J Cardiol 1997;62:143–150. 21. Joshi S, Wilber D. Ablation of right ventricular outflow tract tachycardia. Current perspectives. J Cardiovasc Electrophysiol 2005;16(S1):S52–S58. 22. Callans D, Menz V, Schwartzman D, Gottlieb C, Marchlinski FE. Repetitive monomorphic tachycardia from the left ventricular outflow tract: electrocardiographic patterns consistent with a left ventricular site of origin. J Am Coll Cardiol 1997;29:1023–1027. 23. Hachiya H, Aonuma K, Yamauchi Y, Igawa M, Nogami A, Iesaka Y. How to diagnose, locate and ablate coronary cusp ventricular tachycardia. J Cardiovasc Electrophysiol 2002;13:551–556. 24. Timmermans C, Rodriguez L-M, Crijns H, Moorman A, Wellens H. Idiopathic left bundle-branch block-shaped ventricular tachycardia may originate above the pulmonary valve. Circulation 2003;108:1960 –1967. 25. Tanner H, Hindricks G, Schirdewahn P, Kobza R, Dorszewski A, Piorkorwski C, Gerds-Li J-H, Kottkamp H. Outflow tract tachycardia with R/S transition in lead V3. J Am Coll Cardiol 2005;45:418 – 423. 26. Zei P, Stevenson W. Epicardial catheter mapping and ablation of ventricular tachycardia. Heart Rhythm 2006;3:360 –363. 27. Fritz J, Solaiyappan M, Tandri H, Bomma C, Genc A, Clausen C, Lima J, Bluemke D. Right ventricle shape and contraction patterns and relation to magnetic resonance imaging findings. J Comput Assist Tomogr 2005;29: 725–733.