Author + information
- Received September 7, 2017
- Revision received November 2, 2017
- Accepted November 2, 2017
- Published online December 20, 2017.
- Masaharu Masuda, MD, PhD∗ (, )
- Mitsutoshi Asai, MD, PhD,
- Osamu Iida, MD,
- Shin Okamoto, MD,
- Takayuki Ishihara, MD,
- Kiyonori Nanto, MD,
- Takashi Kanda, MD,
- Takuya Tsujimura, MD,
- Yasuhiro Matsuda, MD,
- Shota Okuno, MD,
- Takuya Ohashi, MD,
- Aki Tsuji, MD and
- Toshiaki Mano, MD, PhD
- ↵∗Address for correspondence:
Dr. Masaharu Masuda, Kansai Rosai Hospital Cardiovascular Center, 3-1-69 Inabaso, Amagasaki-shi, Hyogo 660-8511, Japan.
Objectives The aim of this study was to assess the use of wave front propagation speed on a right ventricular map for determining the earliest activation site as the origin of outflow tract ventricular arrhythmias (VAs).
Background VAs with centrifugal right ventricular outflow tract (RVOT) activation can be from an RVOT focus or a focus outside the RVOT.
Methods This prospective observational study included 23 patients with idiopathic outflow tract VAs. Mapping of the RVOT was performed using a new ultra-high-resolution electroanatomic mapping system. The wave front propagation speed was estimated from the area surrounded by a propagated wave front at 5, 10, 15, and 20 ms after the earliest activation.
Results VAs disappeared following ablations in the RVOT in 15 patients (RVOT origin). The remaining 8 patients had VAs of non-RVOT origin determined by ablation success at another site or ablation failure. The areas surrounded by a propagated wave front were significantly smaller in VAs of RVOT origin than non-RVOT VAs at 5 ms (1.0 [0.7 to 1.1] cm2 vs. 2.2 [1.6 to 4.4] cm2), 10 ms (1.9 [1.4 to 2.2] cm2 vs. 4.5 [3.2 to 5.8] cm2), 15 ms (3.2 [2.3 to 4.4] cm2 vs. 7.1 [6.3 to 9.8] cm2), and 20 ms (5.0 [3.0 to 6.6] cm2 vs. 9.8 [9.3 to 14.8] cm2). A propagated area of <5.0 cm2 at 15 ms predicted RVOT VAs with 87% sensitivity, 100% specificity, and 91% predictive accuracy.
Conclusions VAs with slow wave front propagation speed on the right ventricular map indicate an RVOT origin.
Catheter ablation is an established therapy for idiopathic ventricular arrhythmias (VAs), including premature ventricular complexes and ventricular tachycardia. The ventricular outflow tracts (OTs) are common origins of idiopathic VAs (1,2). Specifically, the right ventricular outflow tract (RVOT) is the source of approximately 70% to 80% of idiopathic OT VAs, which are successfully treated by ablation in 90% to 95% of cases (3). The left ventricular OT (LVOT) is a less common source, and its ablation is relatively complex (4–6).
Given the low risk of catheter manipulation and the relatively high prevalence of RVOT VAs, mapping of the RVOT is usually performed first to identify the optimal ablation site. The origin of centrifugally propagated VAs is suggested by the earliest activation site (EAS) that demonstrates the following electrophysiological characteristics: a premature electrogram relative to QRS onset, QS configuration on the unipolar electrogram, and concordance of paced QRS configuration with that of clinical VA (7). However, even in the presence of these features, it is sometimes challenging to differentiate the EAS between the true origin of the VAs and a site resulting from passive activation from adjacent tissue.
The activation pattern of VAs originating from the RVOT endocardium propagates along the RVOT endocardial surface in 2 dimensions, whereas VAs originating at an adjacent site tend to approach the RVOT from multiple directions and in 3 dimensions. We hypothesized that the centrifugal propagation speed on the right ventricular (RV) map would be slower in VAs originating from the RVOT endocardium compared with VAs arising from adjacent tissue. Herczku et al. (8) supported this hypothesis by measuring the propagation area 10 ms after the earliest RVOT activation using a conventional electroanatomic mapping system in patients with OT VAs. Recently, a new electroanatomic mapping system (Rhythmia, Boston Scientific, Marlborough, Massachusetts) using a mini-basket-array catheter with 64 mini-electrodes (IntellaMap Orion, Boston Scientific) has been introduced to rapidly obtain an ultra-high-resolution electroanatomic map (9–11). We sought to assess wave front propagation speeds by measuring the area surrounded by the wave front at 5, 10, 15, and 20 ms after the earliest activation on the RV propagation map and to compare the areas between VAs of RVOT origin with those of non-RVOT origin.
This prospective study included consecutive patients who underwent catheter ablation of idiopathic OT VAs using the new electroanatomic mapping system from September 2016 to May 2017. Patients were diagnosed with idiopathic OT VAs on the basis of a 12-lead electrocardiogram demonstrating the inferior axis without any abnormal findings suggestive of structural heart disease by symptoms, chest radiography, cardiac echocardiography, and 12-lead electrocardiography.
The following exclusion criteria were used: age <20 years, concomitant structural heart disease, prior ablation for VAs, and prior cardiac surgery. If the frequency of VAs was too low to draw a sufficient map in the RVOT even after isoproterenol infusion, the patient was also excluded.
This study complied with the Declaration of Helsinki. Written informed consent for ablation and participation in the study was obtained from all patients, and the protocol was approved by our Institutional Review Board.
Electrophysiological study and mapping
After withdrawal of antiarrhythmic drugs for ≥5 half-lives, an electrophysiological study and ablation were performed without any sedative drugs by 2 experienced operators. During the procedure, intravenous heparin was administered to maintain activated clotting times of 250 to 350 s. The prematurity index was calculated as the ratio of the coupling interval of the first ventricular tachycardia beat or isolated premature ventricular complexes to the preceding R-R interval. A 2.3-F micro-multielectrode catheter (Pathfinder, Cardima, Fremont, California) was advanced into the great cardiac vein (GCV) as far as possible through the lumen of a 5-F 10-pole electrode catheter (Luma Cath, St. Jude Medical, St. Paul, Minnesota) situated in the coronary sinus. A 6-F decapolar electrode catheter (TENTEN, St. Jude Medical) was placed in the right ventricle.
Mapping in the right ventricle was then performed under the guidance of the electroanatomic mapping system using the mini-basket catheter, which was inserted into the right ventricle through a steerable long sheath (Agilis S Curve, St. Jude Medical). The mini-basket catheter consists of 8 splines, each containing 8 small electrodes with a surface area of 0.4 mm2 and an interelectrode spacing of 2.5 mm as measured from center to center. Isoproterenol was administered to provoke VAs if necessary. During the mapping, the VA configuration concordant with the clinical one was automatically accepted using an electrocardiographic cutoff value of 0.05 by the beat acceptance criteria incorporated in the mapping system. Other criteria used for beat acceptance included respiratory gating, stable catheter location, and stable catheter signals compared with adjacent points. The confidence mask and the projection distance were set at 0.03 mV and 2 mm, respectively. The system annotates the largest bipolar signal and the point of steepest negative slope of the unipolar electrogram within the mapping window, taking into account the consistency of the activation timing with surrounding points.
The prematurity of the electrogram recorded by the electrodes in the GCV was assessed. The LVOT endocardium and aortic sinus cusps (ASCs) were also mapped if these regions were expected to be the VA origin by using 12-lead surface electrocardiography, less prematurity of the local electrogram at the RV EAS compared with QRS onset and/or other mapping electrodes, and/or ablation failure at the RV EAS. A retrograde transaortic approach was used to access the aortic root and the left ventricular (LV) endocardium. First, aortography, coronary angiography, and left ventriculography were performed. Then, mapping in the aortic root and the left ventricle was performed with the mini-basket catheter in conjunction with a 3.5-mm linear ablation catheter if necessary. In cases with severe atherosclerotic lesions in the aorta or basket catheter manipulation difficulties using the transaortic approach in the LVOT, the transseptal approach was used.
Local activation timing at the EAS compared with QRS onset and the signal configuration of the unipolar electrogram at the EAS were recorded. Pace mapping at a cycle length of 500 ms with a minimum stimulus amplitude required for consistent capture was performed, and the pace map score was calculated as the number of leads recording R/S wave configuration and fine notching identical to the pre-procedural 12-lead electrocardiogram, as previously reported (12).
Ablation was performed at the RV EAS without mapping other chambers at the discretion of the operator if the surface electrocardiogram suggested RVOT origin, the local electrogram at the RV EAS preceded surface QRS onset by >25 ms and was earlier than the electrogram recorded in the GCV, the unipolar electrogram had recorded the QS configuration, and/or the pace map score was ≥23/24. Otherwise, ablation was performed at the site with the earliest bipolar electrogram and QS pattern unipolar electrogram after mapping the ASCs, left ventricle, and GCV. An open-irrigated ablation catheter with a 3.5-mm tip (ThermoCool Celsius, Biosense Webster, Diamond Bar, California) through an Agilis sheath was used. Radiofrequency energy was applied for 30 s at each site in a temperature-controlled mode with a maximum temperature of 42°C. The initial application power settings were as follows: 20 W at the GCV and ASCs and 30 W at the RVOT and LV endocardium. If the VAs decreased but were still identified, the power levels were increased to a maximum of 25 W at the GCV, 35 W at the ASCs, 40 W at the RV endocardium, and 50 W at the LV endocardium. Radiofrequency application was immediately stopped if catheter dislodgement occurred or if clinical VAs were still present even after 20 s of radiofrequency application. After ablation, a provocation test was performed using isoproterenol infusion. Acute ablation outcomes were categorized as a success if the VAs had completely disappeared as a result of the radiofrequency application even after isoproterenol infusion, a partial success if the frequency of VAs was significantly reduced by the radiofrequency application but still continued to develop spontaneously or with isoproterenol infusion, and a failure if the frequency of VAs did not significantly change even after the radiofrequency application.
Measuring the area surrounded by the propagation wave
The centrifugal propagation speed was estimated from the areas surrounded by the propagated wave front on the RV propagation map (Figure 1) at 5, 10, 15, and 20 ms after the earliest RV activation. Each propagated area was depicted by setting the beginning of “early meets late” at the timing of the earliest RV activation using durations of 5, 10, 15, and 20 ms, respectively. Subsequently, the area was measured by manually tracing the border of “early meets late.” Interobserver variability of the area (6%) was calculated in 10 randomly selected patients as the difference between 2 measurements in the same patients by 2 different observers divided by the mean value.
Patients with successful elimination of VAs by ablation were followed up without any antiarrhythmic drugs. Patients were seen every 4 weeks at the arrhythmia clinic of our institution for a minimum of 3 months. Routine electrocardiograms were obtained at each outpatient visit, and 24-h ambulatory Holter monitoring was performed at 1 to 2 months post-ablation. When patients experienced symptoms suggestive of an arrhythmia, a surface electrocardiogram, ambulatory electrocardiogram, and/or cardiac event recording were also obtained.
We calculated the sample size on the basis of the power analysis, assuming that the propagation area of the RVOT and non-RVOT VAs demonstrated a normal distribution with an SD of 1.0 cm2 and a difference in the mean value of 2.0 cm2, estimating from a previous report (8). The inclusion of 5 patients in each group allowed statistical power of 80%, with a type I error of 0.05.
Continuous data are expressed as median (interquartile range [IQR]). Categorical data are expressed as absolute values and percentages. Tests for significance were conducted using the Wilcoxon rank sum test to compare continuous variables between independent groups, and the Fisher exact test was used for categorical variables. For the prediction of VA origin, receiver-operating characteristic curves were constructed for the area surrounded by the propagated wave front. The area under the curve was determined and a 95% confidence interval for the area under the curve was calculated using the bootstrap method. To assess correlations between continuous variables, a Pearson correlation coefficient analysis was performed. All analyses were performed using SPSS version 22.0 (SPSS, Chicago, Illinois).
Patient characteristics and procedural overview
Twenty-three patients who underwent the initial catheter ablation for idiopathic OT VAs using the new mapping system were included. Baseline characteristics are presented in Table 1. At the beginning of the electrophysiological study, VAs spontaneously developed in 18 patients (78%). Isoproterenol infusion–provoked VAs were confirmed to have identical configuration to clinical VAs in the 5 remaining patients (22%). Total procedural time was 102 minutes (IQR: 69 to 150 min), and no severe complications developed.
RV mapping and ablation
Mapping in the RVOT was completed using the mini-basket catheter in all patients with mapping points numbering 1,617 (IQR: 1,014 to 2,827) and a mapping time of 11.5 min (IQR: 8.2 to 17.4 min). Representative cases are presented in Figure 2. Although temporary disappearance of VAs during RVOT mapping due to a mechanical bump by the mini-basket catheter occurred in 3 cases (13%), the VAs recovered within several minutes in all cases. All activation maps demonstrated a centrifugal activation pattern. Electrogram properties at the EAS on the RVOT map are presented in Table 2. A flowchart of the ablation procedures is presented in Figure 3. Radiofrequency applications were delivered in 19 patients (83%) who had sufficient prematurity at the RV EAS, local activation timing earlier than 25 ms relative to QRS onset, and earlier activation timing than that recorded in the GCV. Ablation at the RVOT was successful: complete VA disappearance in 15 patients, partially successful in 2 patients who did not have further mapping in the LVOT and ASCs performed because of significant reduction in the VA frequency (non-RVOT origin), and failure in 2 patients (non-RVOT origin). The remaining 4 patients (non-RVOT origin) did not undergo ablation at the RVOT given that their electrophysiological findings clearly suggested a non-RVOT origin.
Mapping and ablation in the LVOT endocardium and ASCs
Mapping of the LVOT endocardium and ASCs was performed using a transaortic approach in 4 of the aforementioned 6 patients (Figure 3). In the remaining 2 patients, only the LVOT endocardium was mapped using a transseptal approach to avoid intra-aortic catheter manipulation due to advanced atherosclerotic lesions in the aorta. Among the 4 patients undergoing the transaortic approach, mapping of the ASCs was completed with the mini-basket catheter in all patients; however, LVOT mapping with the mini-basket catheter could not generate a sufficient map for identifying the EAS in 3 of the 4 patients, necessitating concomitant use of a transseptal approach in 1 patient and ablation catheter use in 2 patients.
Examples of VAs originating at the LVOT and ASCs are presented in Figure 2. Ablation finally eliminated VAs at the ASCs in 2 of 23 patients and at the LVOT beneath the aortic valve in 3 patients. In 1 patient, VAs did not disappear even after the ablation at the EAS of the RVOT, LV endocardium, and ASCs, resulting in procedural failure.
At the follow-up clinic, 20 patients with immediate procedural success did not report any symptoms of VAs, and 24-h ambulatory Holter monitoring demonstrated that the percentage of VAs in the total number of heartbeats was <1%. In contrast, the 3 patients who had ablation procedures end in partial success or failure still experienced symptoms related to VAs, and the percentage of VAs in the total number of heartbeats remained at >10%.
Propagation areas on the RV map
Representative cases demonstrating the areas surrounded by the propagated wave front on the RVOT maps at 5, 10, 15, and 20 ms after the earliest RV activation are presented in Figure 4. The area at each time point was significantly smaller in the VAs of RVOT origin than those of non-RVOT origin (Figure 5). The RVOT-propagated wave front reached the pulmonary valve at 20 ms in 7 patients (23%), suggesting that the propagation area might not correctly reflect the propagation speed in some cases at times of ≥20 ms. Receiver-operating characteristic curve analysis using the propagation area for the differentiation of RVOT origin from non-RVOT origin at each time point demonstrated areas under the curve at 5 ms of 0.98, at 10 ms of 0.95, at 15 ms of 0.98, and at 20 ms of 0.98. The best cutoff value of the propagation area at 15 ms for the prediction of RVOT origin was 5.0 cm2, with sensitivity of 87%, specificity of 100%, and predictive accuracy of 91%.
The influence of VA prematurity on propagation area was assessed. The prematurity index did not correlate with propagation area at 15 ms (r = −0.26, p = 0.24). In addition, there was no difference in the prematurity index between VAs with an RVOT origin and those with a non-RVOT origin (0.75 [IQR: 0.62 to 0.83] vs. 0.69 [IQR: 0.62 to 0.79], p = 0.42).
Diagnostic accuracy of conventional parameters
The prematurity of the local electrogram at the RV EAS (preceding ≥25 ms relative to QRS onset) predicted the RVOT origin of VAs with diagnostic accuracy of 83% (19 of 23) (Table 2). A QS configuration on unipolar electrogram and a pace map score of ≥23/24 at the RV EAS demonstrated diagnostic accuracy of 91% (21 of 23) and 78% (18 of 23), respectively. Wave front propagation speed correctly differentiated VA origin in the majority of cases (9 of 11 [82%]), with an incorrect diagnosis made by at least 1 conventional parameter.
This prospective observational study included 23 patients who underwent initial ablation for idiopathic OT VAs. RVOT maps were successfully created using the new mapping system in conjunction with the mini-basket catheter in all patients. This study presents the new concept that a slow centrifugal wave front propagation speed on the RVOT map indicates VAs of RVOT origin. Conversely, a fast centrifugal wave front propagation speed on the RVOT map reflects VAs of non-RVOT origin coming toward the RVOT through multiple connections. The present study confirms the findings reported by Herczku et al. (8), but with a new ultra-high-resolution mapping system.
Mapping VAs using the new mapping system
RV mapping using the new mapping system in conjunction with the mini-basket catheter was performed without hindrance in all cases. The time to identify the EAS in the right ventricle was relatively short, and it did not require any manual reannotation of activation timing. The number of mapping points in the RVOT was sufficient to delineate the propagation area. ASCs were also mapped successfully using the transaortic approach with the mini-basket catheter. However, mapping in the LVOT required transaortic and transseptal approaches or ablation catheter use in some cases. The LVOT is a narrow and complex structure with numerous obstacles. In addition, a mapping catheter reaches the LVOT passing through the aortic or mitral valves. Therefore, catheter manipulation of the mini-basket catheter, which has a large deflection curve and rigid catheter tip, would be restricted.
Mechanism of high propagation speed when VAs originate at an adjacent site
Different propagation speeds between VAs originating from the right ventricle and those from adjacent tissue would be explained by the different propagation patterns as follows. VAs originating from the RV endocardium would radiate outward in concentric circles to cover the RV endocardial surface. In case of VAs originating from adjacent tissue, the activation wave front would propagate though some connections radially in a spherical pattern and be projected to the RV endocardium. Assuming that the 3-dimensional propagation speeds were identical, the 2-dimensional spreading speed on the RV endocardium becomes faster in the latter case.
Advantage of using wave front propagation speed for the differentiation of VAs of RVOT origin from VAs of adjacent tissue origin
The propagation speed analysis on the RV map using a specific cutoff value could differentiate VAs of RVOT origin from VAs with adjacent tissue origin even for some VAs for which the conventional methods were not able to determine RVOT or other origin.
Conventional parameters indicating that VAs originate from the EAS of a centrifugal pattern consisted of the prematurity of a local electrogram (7), a QS pattern configuration on a unipolar electrogram, and a paced QRS configuration identical to clinical VAs (7). However, even applying these methods, it is occasionally challenging to differentiate between the EAS on an RV map indicating true VA origin and a site reflecting a passive propagation wave coming from adjacent tissue.
The propagation speed of VAs is an observation from a panoramic viewpoint, in contrast to that of conventional parameters, which focus mainly on the local electrogram only at the EAS. Therefore, the propagation speed in conjunction with conventional parameters would enable a more accurate estimation of VA origin.
Several limitations of this study warrant mention. First, the results represent a single-center experience that might be dependent on specific operators’ skills.
Second, mapping using the mini-basket catheter may have failed to draw the entire endocardial surface in the RVOT, possibly resulting in an incorrect EAS location and/or propagation area. To confirm contact between the mini-basket catheter and myocardium, the sharpness of local electrograms, especially near the EAS, was cautiously assessed during the mapping procedure. Areas with possible insufficient catheter-tissue contact were remapped.
Third, VA origin estimated by the response to the radiofrequency application might not be necessarily correct, because conductive heating of radiofrequency application could injure the adjunctive tissue near the radiofrequency application site.
Finally, the statistical analyses were potentially influenced by the relatively small size of the study population. Multicenter studies including larger populations are needed to form reliable conclusions about the general usefulness of this method.
The new ultra-high-resolution mapping system using the mini-basket catheter is feasible for the ablation of idiopathic OT VAs. A slow propagation speed of the VA wave front on the RV map indicated that the origin was at the right ventricle.
COMPETENCY IN MEDICAL KNOWLEDGE: The new ultra-high-resolution mapping system is useful for the ablation of idiopathic OT VAs. In addition, wave front propagation speed in conjunction with conventional electrophysiological methods could more precisely estimate the origin of VAs.
TRANSLATIONAL OUTLOOK: To elucidate the usefulness of wave front propagation speed on the centrifugal propagation map for the determination of the origin being at the EAS in cases with other arrhythmias, clinical studies including patients with varieties of arrhythmias need to be performed.
The authors appreciate the devoted support on statistical issues of Mitsuyoshi Takahara, MD, PhD, of the Department of Metabolic Medicine, Osaka University Graduate School of Medicine. In addition, the authors thank Norimasa Abe and Naoya Kurata (medical engineers from Kansai Rosai Hospital) for their technical support with the electroanatomic mapping system.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- aortic sinus cusp
- earliest activation site
- great cardiac vein
- interquartile range
- left ventricular
- left ventricular outflow tract
- outflow tract
- right ventricular
- right ventricular outflow tract
- ventricular arrhythmia
- Received September 7, 2017.
- Revision received November 2, 2017.
- Accepted November 2, 2017.
- 2017 American College of Cardiology Foundation
- Priori S.G.,
- Blomström-Lundqvist C.,
- Mazzanti A.,
- et al.
- Latchamsetty R.,
- Yokokawa M.,
- Morady F.,
- et al.
- Kamioka M.,
- Mathew S.,
- Lin T.,
- et al.
- Yamada T.,
- Litovsky S.H.,
- Kay G.N.
- Ouyang F.,
- Mathew S.,
- Wu S.,
- et al.
- Hutchinson M.D.,
- Garcia F.C.
- Herczku C.,
- Berruezo A.,
- Andreu D.,
- et al.
- Nakagawa H.,
- Ikeda A.,
- Sharma T.,
- Lazzara R.,
- Jackman W.M.
- Sohns C.,
- Saguner A.M.,
- Lemes C.,
- et al.
- Coggins D.L.,
- Lee R.J.,
- Sweeney J.,
- et al.