Article ID: CJ-22-0769
Background: An epicardial connection (EC) between the right-sided pulmonary venous (RtPV) carina and right atrium (RA) may preclude PV isolation, but its electrophysiological role during atrial fibrillation (AF) remains unknown.
Methods and Results: This prospective observational study included 98 consecutive patients undergoing catheter ablation for AF, subdivided into the EC group (n=17) and non-EC group (n=80) based on observation of RA posterior wall breakthrough during RtPV pacing. Mean left atrial (LA) dominant frequency (mean DFLA) was defined as the averaged DFs at the right and left PVs and LA appendage. The regional DF was higher in the EC group vs. the non-EC group except at the left PV antrum. The DF at the RA appendage (RAA) and mean DFLA were equivocal (6.5±0.7 vs. 6.6±0.7 Hz) in the EC group, but the mean DFLA was significantly higher than that at the RAA (5.8±0.6 vs. 6.1±0.5 Hz, P=0.001) in the non-EC group, suggesting an LA-to-RA DF gradient. A significant correlation of DF between the RtPV antrum and RAA was observed in the EC group (P<0.001, r=0.84) but not in the non-EC group.
Conclusions: An electrophysiological link via interatrial ECs might attenuate the hierarchical nature of activation frequencies of AF, leading to advanced electrical remodeling of the atria.
The drivers of atrial fibrillation (AF) are predominantly located in the posterior left atrial (LA) wall and include the pulmonary veins (PVs). PV isolation (PVI) is the cornerstone of catheter ablation for AF,1–6 but recently, the existence of an epicardial connection (EC) between the right-sided PVs (RtPVs) and right atrium (RA) has been reported as a mechanism for the failure of first-pass isolation and is associated with poor ablation outcomes.7–13 Although an EC can be a critical part of a macro-reentrant circuit of atrial tachycardia,9 and play a role in the right-to-left interatrial conduction during sinus rhythm,14 the electrophysiological role of an EC on activation frequency during AF has not been well studied. We hypothesized that direct conduction between the RtPVs and the RA may affect activation frequencies during AF, as well as the spatial dispersion that characterizes AF.
Frequency domain analysis is a well-established methodology for quantifying the characteristics of AF,1–3,15 so in the present study, the dominant frequencies (DFs) at various sites in the right and left atria and their gradients in patients with ECs were compared with those in patients without ECs. Also, the correlations of DFs between each pair of sites were analyzed.
This prospective observational study included 98 consecutive patients who were referred to Shizuoka Saiseikai General Hospital for initial catheter ablation of AF between November 2018 and August 2022 (paroxysmal AF in 50, persistent AF in 48 patients). The study protocol was approved by the local Institutional Review Board (approval no. 2-20-03) and complied with the Declaration of Helsinki. Written informed consent was given by all patients.
Electrophysiological StudyAll antiarrhythmic drugs were discontinued for at least 5 half-lives before the procedure, except for amiodarone, which was discontinued 8 weeks beforehand. The electrophysiological study and catheter ablation were performed under conscious sedation with dexmedetomidine and fentanyl. Surface electrocardiograms and intracardiac electrograms were continuously monitored and stored on a LabSystem PRO (Bard Electrophysiology, Lowell, MA, USA) recording system. A 6F 20-pole dual-site mapping catheter (BeeAT; Japan Lifeline Co., Ltd., Tokyo, Japan) was inserted through the subclavian or jugular vein and positioned in the coronary sinus (CS), RA, and superior vena cava throughout the procedure. After transseptal puncture, 2 long sheaths (SL0; AF Division, Abbott) were then advanced into the LA. All patients underwent bipolar recordings for frequency analysis during AF and the activation map in the LA during sinus rhythm before ablation, as follows.
Recording for Frequency Analysis If the patient presented in sustained AF, the recordings for frequency analysis were made immediately after transseptal puncture. If the patient presented in sinus rhythm, burst atrial pacing was used to induce AF, and recordings were obtained after AF was sustained for >5 min. Electrograms at the left and right atrial appendages (LAA and RAA, respectively), and the left and right PV antrum (LtPVA and RtPVA, respectively) were recorded by a decapolar circular catheter (Lasso, Biosense Webster Inc., Diamond Bar, CA, USA), and electrograms at the CS were recorded by the BeeAT.
Activation Map in the LA During sinus rhythm, each PV and the LA were mapped with a duo-decapolar mapping catheter (PENTARAY; Biosense Webster) to evaluate the breakthrough site in the LA. If the patient presented in sinus rhythm, the map was made immediately after transseptal puncture, but if the patient presented in sustained AF, internal cardioversion using a BeeAT catheter was performed after completion of the recordings for frequency analysis. Points were acquired using the CARTO ConfiDENSE module in the auto-freeze mode, as described previously.7
Catheter AblationThe ipsilateral PV antrum was circumferentially ablated with a ThermoCool STSF catheter under CARTO guidance (Biosense Webster) using real-time automated display of radiofrequency (RF) applications (Visitag; Biosense Webster). RF energy was delivered at 25–35 W at the posterior aspect, at 15–20 W at the sites adjacent to the esophagus, and at 30–40 W at the anterior aspect of the PV antrum. Target values of the ablation index (Biosense Webster) were 500 for the anterior aspect and 400 for the posterior aspect of the PV antrum, and the interlesional distance was 4 mm. PVI with entrance and exit blocks was confirmed with the Lasso catheter and a pacing maneuver within the isolation line with an ablation catheter.
As described in a previous study,8 if the RtPVs were either not isolated by circumferential ablation or reconnected during the procedure, an activation map was created during pacing from the mid carina of the RtPVs with a Lasso catheter at a pacing output <5.0 mV. If the earliest site was the RA, the patient was classified into the EC group and the earliest sites at the RA were first ablated at a power of 35 W. If the right-sided PVs were not isolated, additional ablation at the carina in reference to the carina breakthrough site in the baseline activation map was repeated. If the earliest site was the antrum of the RtPVs, the cause of failure in the first-pass isolation was considered to be a gap in the antral line, and careful remapping to identify gaps in reference to the activation map during pacing from the RtPV and additional ablation at the same power setting on or just inside the initial line were repeated until the RtPVs were isolated. Patients other than those in the EC group were classified into the non-EC group. Representative activation maps are shown in Figure 1.
Representative 3-dimensional maps. (A) In a patient in the EC group, activation mapping during sinus rhythm showed carina breakthrough (yellow arrow). Activation mapping during pacing from the RtPV carina after circumferential antral ablation revealed that the earliest site (white arrow) was the posterior RA. (B) In a patient in the non-EC group, activation mapping during sinus rhythm showed Bachmann bundle breakthrough only (light blue arrow). First-pass isolation of the RtPV was achieved by circumferential antral ablation. EC, epicardial connection; LS, left superior pulmonary vein (PV); RA, right atrium; RI, right inferior PV; RS, right superior PV; RtPV, right-sided PV.
Electrograms recorded during AF underwent frequency analysis. The details of spectral analysis have been reported previously.1,2,15 Briefly, bipolar electrograms recorded for 5 s during AF were processed off-line in the MATLAB environment (Math Works, Inc., Natick, MA, USA). In the spectral analysis, the preprocessing steps included band-pass filtering with cutoffs at 40 and 250 Hz, rectification, and low-pass filtering with a 20-Hz cutoff. The DF was defined as the frequency of the highest peak of the periodogram in the interval of 0.5–20 Hz using fast Fourier transformation. All bipolar electrograms of the multipolar catheter were analyzed, and the highest DF was chosen. The mean DF of the LA (mean DFLA) was defined as the averaged DF of the 3 measurement sites in the LA (right and left PV and LAA). A DF gradient was evaluated by comparing the mean DFLA with the DF at the RAA.
Scan Protocol and Data Acquisition of Multidetector Computed Tomography (CT)CT examinations were performed around 1 week before the catheter ablation using a Discovery 750HD 64-slice multidetector CT system (General Electric Medical Systems, Waukesha, WI, USA). For the contrast-enhanced scans, 1 mL/kg of nonionic contrast media (Iopamiron, 370 mg iodine/mL; Bayer Pharma AG, Berlin, Germany) was injected into the antecubital vein. Image reconstruction was retrospectively gated on the electrocardiogram, and electrocardiogram-gated datasets were reconstructed with a slice thickness of 0.625 mm at the point of the cardiac cycle that corresponded to atrial end-diastole (45% of the R-R interval after the QRS complex). The volumes of the RA, LA, and the cross-sectional areas of thoracic veins were manually measured using AW Server software (GE Healthcare, Little Chalfont, UK) for Windows (version 2.0-5.5). All measurements were performed by an investigator without knowledge of the patients’ clinical information. All of the atrial volumes and thoracic veins sizes were indexed to the body surface area.
Follow-upThe patients were scheduled for an outpatient clinic visit around 4 weeks after the procedure and every 2–6 months thereafter for at least 2 years after the ablation session. Documented episodes of atrial tachyarrhythmia by 12-lead ECG, 24-h Holter monitoring, and/or event recorder after the 3-month blanking period were identified as recurrences. A repeat procedure was recommended for patients with recurrences.
Statistical AnalysisContinuous variables are expressed as mean±SD or median (interquartile range [IQR]), and categorical variables are reported as number and percentage. Statistical differences were tested only between the EC group and gap group using an unpaired Student t-test, Mann-Whitney U test, or Chi-square analysis, as appropriate. One-way analysis of variance was used to compare continuous variables among multiple factors. Pairwise DF comparisons were made after Bonferroni correction for multiple comparisons. To determine the correlation of DFs between sites, an independent Pearson’s correlation coefficient was calculated. Statistical significance was considered when the P value was <0.05. The proportion of AF-free patients was determined using Kaplan-Meier analysis with a log-rank test. Statistical analyses were carried out using SPSS version 24 (SPSS Inc., Chicago, IL, USA) or GraphPad Prism 9 (GraphPad Prism Software Inc., San Diego, CA, USA).
All PVs were successfully isolated, and 17 (17%) and 80 (83%) patients were classified into the EC and non-EC groups, respectively (Figure 2). One patient initially classified into the non-EC group was excluded from the study because additional RF applications inside the initial line were needed. The baseline patient characteristics are summarized in Table 1. The prevalence of structural heart disease was significantly higher in the EC group (35% vs. 6%; P<0.05). The results of anatomic measurements by 3D CT are summarized in Table 2. The LA and RA volume indexes and thoracic vein area indexes were compatible between groups. No major complications such as stroke, cardiac tamponade or atrial-esophageal fistula occurred.
Study protocol. EC, epicardial connection; PVI, pulmonary vein isolation. Other abbreviations as in Figure 1.
| EC group (n=17) |
Non-EC group (n=80) |
|
|---|---|---|
| Male sex (N) | 13 (77%) | 65 (81%) |
| Age (years) | 58±9 | 61±9 |
| Age at AF diagnosis (years) | 53±8 | 58±8 |
| Type of AF (paroxysmal, N) | 9 (53%) | 41 (51%) |
| Body mass index (kg/m2) | 25±4 | 25±3 |
| Body surface area (m2) | 1.8±0.2 | 1.9±0.2 |
| Hypertension (N) | 11 (65%) | 48 (60%) |
| Diabetes mellitus (N) | 5 (29%) | 20 (25%) |
| Structural heart disease | 6 (35%) | 5 (6%)† |
| AF-induced cardiomyopathy | 4 (26%) | 3 (4%) |
| Dilated cardiomyopathy | 0 | 1 (1%) |
| Hypertrophic cardiomyopathy | 1 (6%) | 0 |
| Ischemic heart disease | 1 (6%) | 1 (1%) |
| Sick sinus syndrome (N) | 3 (18%) | 15 (19%) |
| eGFR (mL/min/1.73 m2) | 61±13 | 60±11 |
| A-type natriuretic peptide (pg/mL) | 85±49 | 69±47 |
| B-type natriuretic peptide (pg/mL) | 82±38 | 72±62 |
| Echocardiographic findings | ||
| LA diameter (mm) | 45±3 | 44±3 |
| LA volume index (mL/m2) | 40±7 | 38±6 |
| LVEF (%) | 64±10 | 65±8 |
| IVC diameter (mm) | 13±3 | 12±3 |
| E/e’ | 8.1±2.2 | 9.0±3.3 |
†P<0.05 vs. EC group. AF, atrial fibrillation; E, early diastolic transmitral flow velocity; e’, early diastolic mitral annular velocity; EC, epicardial connection; eGFR, estimated glomerular filtration rate; IVC, inferior vena cava; LA , left atrium; LVEF, left ventricular ejection fraction.
| EC group (n=17) |
Non-EC group (n=80) |
|
|---|---|---|
| LA volume index (mL/m2) | 70.9±25.6 | 70.6±11.9 |
| RA volume index (mL/m2) | 70.4±24.7 | 68.8±12.1 |
| Thoracic vein area index | ||
| Left superior PV (cm2/m2) | 1.89±0.48 | 1.85±0.37 |
| Left inferior PV (cm2/m2) | 1.38±0.35 | 1.36±0.25 |
| Right superior PV (cm2/m2) | 2.01±0.41 | 2.05±0.50 |
| Right inferior PV (cm2/m2) | 1.72±0.32 | 1.69±0.46 |
| Superior vena cava (cm2/m2) | 2.00±0.34 | 1.98±0.32 |
LA, left atrium; PV, pulmonary vein; RA, right atrium.
EC Group First-pass isolation was not achieved in 14 of the 17 (82%) patients. In the other 3 patients (18%), first-pass isolation was achieved once, but reconnection occurred. The earliest site in the activation map during pacing from the RtPVs was the posterior RA in all patients. Focal RF applications at the earliest site in the posterior RA resulted in complete elimination of PV potentials in 4 (24%) patients. In the remaining 13 (76%) patients, focal RF applications in the posterior RA resulted in significant prolongation of the interval between the local RA electrogram to PV potential (40 [IQR, 32–55] ms), and the RtPVs were subsequently isolated by 2 [IQR, 1–4.8] RF applications at the carina.
Non-EC Group First-pass isolation was not achieved in 9 (11%) patients, and reconnection occurred in 11 of the remaining 71 patients (16%). In all of these patients, conduction gaps were identified along the antral line, and the RtPVs were successfully isolated by 2 [IQR, 1–3] RF applications on or just inside the initial line.
Regional DFsIn the EC group, there was no significant difference in regional DFs (6.5±0.7 for RAA vs. 6.2±0.7 for CS vs. 6.6±0.6 for RtPVA vs. 6.7±0.9 for LAA vs. 6.6±0.8 Hz for LtPVA; ANOVA P=0.48; Figure 3A). In the non-EC group, there was a significant difference in regional DFs (5.8±0.6 vs. 5.6±0.5 vs. 6.0±0.7 vs. 6.0±0.6 vs. 6.2±0.7 Hz, respectively; ANOVA P<0.001). In the pairwise analysis, the DF at the LtPVA was significantly higher than that at the CS and RAA (P<0.001), that at the RtPVA was significantly higher than that at the CS (P<0.001), and that at the LAA was significantly higher than that at the CS (P<0.001). In the between-group comparison, the DF at the RAA (P<0.001), CS (P<0.001), RtPVA (P<0.01), LAA (P<0.01), and LtPVA (P=0.023), or the mean DFLA (P<0.001), in the EC group was significantly higher than at these sites in the non-EC group.
Dominant frequency (DF) at each site and the DF gradient. (A) No significant difference in regional DFs in the epicardial connection (EC) group, but a significant difference in the non-EC group. (B) No DF gradient in the EC group, but a left atrium to right atrium DF gradient in the non-EC group. *P<0.05 in the comparison between groups, †P<0.05 vs. DF at CS, ‡P<0.05 vs. DF at RAA. CS, coronary sinus; LAA, left atrial appendage; LtPVA, left pulmonary vein antrum; Non-EC, non-epicardial connection group; Mean DFLA, mean dominant frequency in the left atrium; RAA, right atrial appendage; RtPVA, right pulmonary vein antrum.
In the EC group, the DFs at the RAA and the mean DFLA were equivocal (6.5±0.7 vs. 6.6±0.7 Hz), suggesting no DF gradient (Figure 3B). In the non-EC group, however, the mean DFLA was significantly higher than that at the RAA (5.8±0.6 vs. 6.1±0.5, P=0.001), suggesting an LA-to-RA DF gradient.
Correlation of Regional DFsThere was a significant correlation of DFs between the RtPV and RAA in the EC group (P<0.001, r=0.84, slope=0.94), but no such significant correlation was observed in the non-EC group (Figure 4A). The correlations of DFs between other sites are shown in Figure 4B,C. Representative cases in each group are shown in Figure 5.
Correlation of the regional dominant frequencies (DFs). (A) Significant correlation of DFs between the RtPVA and RAA in the epicardial connection (EC) group but not in the non-EC group. (B) Correlations of the regional DFs between the other recording sites in the EC group and (C) non-EC group. Other abbreviations as in Figure 3.
Representative regional distributions of dominant frequencies (DFs). (A) The DF at the right-sided pulmonary antrum (RtPVA) and at the right atrial appendage (RAA) were equivalent in a patient with paroxysmal atrial fibrillation (PAF) in the EC group. (B) Apparent DF gradient between the RtPVA and RAA in a patient with PAF in the non-EC group. Other abbreviations as in Figure 3.
To examine the consistency of the DFs over time, we reviewed the electrophysiological recordings for frequency analysis in all the enrolled patients. We found that the electrograms for frequency analysis at the RtPVA and RAA were recorded for >60 s in 5 of the 17 patients in the EC group and in 12 of 80 patients in the non-EC group. In these patients, we compared the mean DF at the RtPVA and RAA during the first 10 s with that of the last 10 s of the 60-s recordings. There was excellent agreement among DFs at the RtPVA (EC group: r=0.95; non-EC group: r=0.98) and RAA (EC group: r=0.92; non-EC group: r=0.96) during the 2 time periods.
Ablation Outcome and Redo ProceduresThe mean follow-up period was 20±15 months in the EC group and 27±12 months in the non-EC group. Kaplan-Meier analysis revealed that the survival rate of patients free from AF was lower in the EC group than in the non-EC group (71% vs. 89%; log-rank P=0.014; Figure 6).
Kaplan-Meier curves of AF-free survival in the 2 groups during follow-up after the index procedure. AF, atrial fibrillation; EC, epicardial connection.
In the EC group, a second procedure was performed in 4 of 5 patients with recurrence. The RtPVs were reconnected via the EC in all 4 patients. Focal RF applications at the posterior RA failed to re-isolate the RtPVs, and they were subsequently isolated by RF applications inside the PV. The LtPVs were reconnected and successfully re-isolated in 2 of the 4 patients. In the non-EC group, a second procedure was performed in all 9 patients with recurrence, and the LtPVs and RtPVs were reconnected via the conduction gap in 3 and 5 of these 9 patients, respectively. All reconnected PVs were successfully re-isolated. No PV reconnections were observed in 4 of the 9 patients.
To the best of our knowledge, this is the first study to have evaluated the effect of an EC on the regional DFs of AF. The main findings are given below.
1. An LA-to-RA DF gradient, suggesting the presence of an AF driver in the LA and fibrillary conduction from the LA to the RA, was observed in the non-EC group but not in the EC group.
2. The global DF was higher in the EC group than in the non-EC group, suggesting more advanced electrical remodeling in the EC group.
3. There was a significant correlation of the regional DFs between the RtPV and RAA in the EC group, but was not observed in the non-EC group.
4. Structural change of the atria and PVs as assessed with CT imaging was not associated with the presence or absence of ECs.
These findings suggested that ECs may attenuate the hierarchical nature of activation frequencies of AF via an electrophysiological link between the RtPV and RAA.
Impact of ECs on DFsInteratrial conduction via the Bachmann bundle, fossa ovalis, and CS plays an important role in the spatial dispersion of activation frequency during AF.3 In addition to these classical interatrial conduction routes, we previously reported a RtPV carina breakthrough via an EC between the RtPVs and the RA during sinus rhythm.7–12 In addition to altering the manner of conduction during sinus rhythm, direct conduction between the RtPVs and RA via ECs other than the classical pathways may affect the dynamics of AF.14 A comparison of the DFs between the EC group and non-EC group should provide clues to the role of the EC in the dynamics of AF. We placed especial importance on the LA-to-RA DF gradient to explore the role of the EC as an interatrial conduction pathway. In addition, we referred to the correlation of the DFs between the RtPV and RAA when considering this issue.
In general, acute AF is maintained by high-frequency reentrant sources (drivers) in the posterior LA wall, including the PVs, and therefore there is an LA-to-RA frequency gradient, suggesting that the high-frequency activities in the posterior LA wall decrement while traveling to the RA via interatrial conduction pathways.1–3,6 The LA-to-RA DF gradient in the non-EC group was consistent with this concept of a “breakdown frequency”. In agreement with previous reports,16,17 the DF at the RAA was <5 Hz in 8 of 80 patients in the non-EC group. In contrast, the ECs could directly conduct high-frequency activities in the RtPV to the RA without interactions with anatomic and/or functional obstacles. There was no DF gradient in the EC group, and the DF at the RAA was >5 Hz in all patients in this group, suggesting that the electrophysiological link via the ECs may contribute to the increase in DFs, advancement of electrical remodeling, and sustainability of AF. The significant correlation of the DFs between the RtPV and RAA supports this hypothesis (Figure 4A). During AF, high-frequency sources can drift and meander for short periods; however, it is unlikely that drifting and meandering over a small area affected our main results related to DF values, DF gradients and correlations of the DF between the RtPV and RAA. In fact, the consistency of the DFs over time was confirmed in the present study.
In AF with manifest WPW syndrome, rapid ventricular conduction over an accessory pathway can occur because, compared with the atrioventricular node, accessory pathways have different electrophysiological properties. Similarly, ECs may differ from the classical interatrial conductions in their electrophysiological properties. In fact, a rare type of EC with a unidirectional conduction property during PVI has been reported.11,18 In the present study, we showed a significant correlation of regional DFs between the RtPV and RAA and loss of the LA-to-RA DF gradient in the EC group. From these findings it is reasonable to speculate that high-frequency activities in the posterior LA would preferentially conduct to the RA via the EC than via the classical interatrial pathways. A previous study by Wong et al showed that conduction abnormalities of the atria were greater during CS pacing (left-to-right conduction) than during sinus rhythm (right-to-left conduction) in AF patients.19
Clinical ImplicationsThe EC group had worse ablation outcomes than the non-EC group, which was consistent with results from a previous study by Barrio-Lopez et al.13 Difficulty in PVI and associated PV reconnections are the likely reason for the poor ablation outcome. In fact, the RtPV was reconnected via the EC in all 4 patients who underwent a redo procedure in the present EC group. On the other hand, an additional explanation may be drawn from our results suggesting more extensive and advanced electrical remodeling in the presence of ECs. Our speculation is consistent with a report that the existence of ECs was associated with a progression of AF from paroxysmal to persistent.20 Also, PV reconnection via ECs was observed at redo procedures in only 7 of 16 (44%) patients with ECs in the study by Barrio-Lopez et al,13 implying that extra-PV factors such as advanced electrical remodeling of the atria contributed to the poor ablation outcome. Further studies with more data on redo procedures are needed to elucidate this issue.
For successful isolation of the RtPVs and better long-term outcomes, it is important to recognize the existence of an EC between the RtPVs and RA. The presence of a breakthrough at the RtPVs in the activation map during sinus rhythm or a simple pacing maneuver after the completion of continuous linear ablation at the RtPV antrum can predict the existence of ECs. Maintenance of stable sinus rhythm is necessary to perform these methods. In contrast, the correlation of DFs between the RtPVs and RA can be evaluated even during AF. Of note, the DF at the RAA was >7 Hz in 4 of 17 patients in the EC group, whereas the DF at the RAA was <7 Hz in all patients in the non-EC group. A DF of 7 Hz at the RAA could potentially be a cutoff value to discriminate the presence of an EC. Further studies are needed to validate this hypothesis.
Study LimitationsFirst, the study was small, and second, the recordings for frequency analysis were performed as part of clinical ablation procedures, so the recording sites were limited to the areas of greatest interest, namely, around the PVs, CS, LAA, and RAA. Recordings from other sites may have further validated the findings. Third, this prospective observational study did not provide any causal relationship between the presence of ECs and the dynamics of AF. Nonetheless, the main findings of the present study are hypothesis generating, and they merit reporting and further investigation. Fourth, we may have missed some ECs in the non-EC group. However, we found ECs in 17% of the enrolled patients, whereas the prevalence of ECs was <10% in 2 previous studies,13,20 so underdiagnosis seems unlikely. Finally, the sole measurement of a DF does not totally clarify the role of ECs in the genesis and persistence of AF. The fractionation and disorganization of the electrograms due to fibrillatory conduction and complex conduction disturbances in AF patients may have affected the frequency analysis in the present study and should be investigated further in future studies.
Patients with ECs between the RtPVs and RA showed a significant correlation of DFs between the RtPVs and RA, higher regional DFs, and a lower LA-to-RA DF gradient when compared with those without ECs. An electrophysiological link via ECs might attenuate the hierarchical nature of activation frequencies of AF, leading to advanced electrical remodeling of the atria.
We thank Mrs. Masataka Iida, Naoki Hatano, Gai Narita, and Ayana Sugizaki for their technical assistance with the electrophysiological studies in the cardiac catheterization laboratory.
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
A.N. has received honoraria from Johnson & Johnson and Boehringer-Ingelheim, and an endowment from Medtronic Japan and DVx. The other authors declare no conflicts of interest associated with this manuscript.
The study protocol was approved by the Institutional Review Board of Shizuoka Saiseikai General Hospital (approval no. 2-20-03).
M.I. is a member of Circulation Journal’s Editorial Team.
The deidentified participant data will not be shared.
