Uncoupling endocardial bundles coupled by an epicardial bundle in the left atrium and pulmonary veins

Uncoupling of the endocardial bundles in the left atrium was suggested during modified posterior wall isolation. Although this fact may not be observed because of the possible bridging conduction by epicardial bundles in humans, partially failed transmural ablation in the atrial roof may have iatrogenically unveiled this fact. The left atrium and pulmonary veins exhibit three-dimensional complex structures composed of endocardial and epicardial bundles.1-5 This may also characterize the unique pathological conduction properties associated with development of atrial arrhythmias.1 Although electrical dissociation between the endo- and epicardium could be favored by the bilayer muscular architecture, electrical coupling and uncoupling of these bundles remain incompletely evaluated electrophysiologically. A 72-year-old man was referred to our institution for catheter ablation of paroxysmal atrial fibrillation (AF). Computed tomography scan revealed the presence of inferior common pulmonary veins (PVs) (Figure 1A). An electrophysiological study was performed under CARTO guidance (Biosense Webster, Inc., Diamond Bar, CA). A 6F 20-pole dual-site mapping catheter (BeeAT, Japan Lifeline Co., Ltd., Tokyo, Japan) was inserted through the subclavian vein and positioned in the coronary sinus, right atrium, and superior vena cava. Neither low-voltage areas nor slow conduction was present on left atrial mapping using a PENTARAY catheter (Biosense Webster) (Figure 1B,C; Supplementary movie 1). After isolation of the left-sided PVs by circumferential antral ablation at power settings of 25 to 35 W, we decided not to ablate the posterior wall for right-sided isolation because the left and right inferior PVs were close each other. Instead, linear ablations in the atrial roof and posterior bottom in addition to ablation of the anterior PV antrum were performed to encircle the right-sided PVs and posterior wall at one time (Figure 2A). After failure of first-pass isolation, additional anterior carina ablation immediately eliminated the potentials recorded by a Lasso catheter deep inside the right superior PV (Figure 2B), and dissociated activities were frequently observed, indicating bidirectional block (Figure 2C). In re-mapping of the left atrium and PV, however, discrete potentials were still observed in the posterior wall (Figure 3A), and pacing by the ablation catheter captured this region, suggesting that this localized area of the posterior wall was not coupled to the right PV musculature in the endocardial aspect despite no ablation performed between them. Notably, the activation map during right atrial pacing showed fractionated and high-frequency potentials at only one point in the mid posterior antrum (white arrow in Figure 3A,B,C). Although this point did not appear to depolarize earlier in this endocardial mapping (Figure 3B; Supplementary movie 2), a single ablation at this point eliminated all potentials in the posterior wall (entrance block) (Figure 3D), and pacing confirmed exit block. We considered that completion of transmural conduction block in the left atrial roof was unnecessary for this patient and that further excessive ablation would increase the risk for complications such as cardiac tamponade. Actually, pacing with and without isoproterenol infusion did not induce any atrial tachycardias. The patient has remained recurrence free from AF during a 12-month follow-up period. Pulmonary venous carina ablation is not infrequently required for PV isolation, and posterior wall isolation is commonly difficult to achieve without ablating sites inside the posterior wall and away from the ablation lines.2-4 This fact suggests that epicardial bundles preclude transmural ablation lesion sets and achievement of isolation. At baseline mapping in this patient, sinus impulses were smoothly conducted in the left atrium without conduction delay or block, indicating that left atrial activation used both endocardial and epicardial bundles electrically coupled to each other. After one round of ablation of the right-sided PVs and posterior wall, anterior carina ablation eliminated the right-sided PV sleeve potentials. This suggested that ablation eliminated conduction to the right-sided PV muscle through the intercaval bundle3 but did not affect conduction to the posterior wall (Figure 4). We thought that these two musculatures, the right-sided PV musculature and posterior wall musculature (septoatrial bundle), were originally not coupled to each other in the endocardial aspect but were bridged by an epicardial bundle (septopulmonary bundle). Because ablation at the narrow portion in the roof line failed to form a transmural lesion (open red circles in Figure 4), epicardial conduction could epicardially reach the mid posterior antrum (entrance conduction) where an interesting high-frequency potential was recorded. As a limitation of this report, an activation map of the whole atrium during posterior wall pacing was not created to evaluate exit conduction through the epicardial bundle. Roof line ablation can be difficult to achieve due to the multilayer architecture and the epicardial heat sink created by arterial blood flow.5 Two causes for this observation can be considered. First, the dissociated activities frequently occurred in the right-sided PVs, allowing us early recognition of bidirectional block and isolation in the right-sided PV sleeve alone. Second, we did not create a posterior line for right-sided PV isolation in this patient, thus allowing us to evaluate naïve conduction in the posterior antrum. The activation map for identification of the final ablation site at the mid-posterior antrum was not helpful in this patient. Endocardial mapping alone may not always be reliable because of contamination of the farfield electrograms from the epicardial bundles (Supplementary movie 2). We believe that the interesting high-frequency potentials observed in this patient can be a marker for an endocardial breakout site, that is, a connection site between endo- and epicardial musculatures at which an abrupt change in orientation of the impulse propagation can occur. Rao et al. reported an endocardial breakout site in the posterior wall by direct percutaneous epicardial mapping.4 The breakout site clearly exhibited high-frequency and fractionated potentials similar to those observed in our patient, although the authors did not refer to such potentials in their report, thus supporting our interpretation. When roof line block for posterior wall isolation is difficult to achieve, characteristic potentials of this type should be searched for in the posterior wall with a multi-electrode mapping catheter. In conclusion, uncoupling of the endocardial bundles in this patient was suggested during modified posterior wall isolation. Although this fact may not be observed because of the possible bridging conduction by epicardial bundles in humans, partially failed transmural ablation in the atrial roof may have iatrogenically unveiled this fact. Fractionated and high-frequency potentials may be a marker for an endocardial breakout site in posterior wall isolation. We thank Mari Ebine and Yoshiko Uehara for their technical support during the ablation procedure. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. The authors declare no conflict of interest. N/A. The patient signed the informed consent statement. N/A. The data that support the findings of this study are available from the corresponding author upon reasonable request. Supplementary movie 1. Propagation map of the left atrium during sinus rhythm at baseline, corresponding to Figure 1(C). Supplementary movie 2. Propagation map of the left atrium after isolation of the right PV sleeve, corresponding to Figure 3(B). Impulse propagation was directed supero-inferiorly at the posterior wall, suggesting the inclusion of epicardial activation in the endocardial map. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.