Variable character of epicardial connections during ablation of atrial fibrillation

Introduction

Key Teaching Points.

• •It may be important to recognize the presence of epicardial connections (ECs) as a possible cause for failure of first-pass isolation.
• •ECs may play an important role in initiating and maintaining atrial fibrillation (AF), since AF was easily induced until ECs were ablated.
• •Thus, physicians should consider performing an additional targeted ablation of the EC, which may have the possibility of changing its character during AF ablation, to achieve a complete pulmonary vein antrum isolation.

Pulmonary vein (PV) antrum isolation (PVAI) is a useful treatment strategy for atrial fibrillation (AF) worldwide.¹ Moreover, a complete PVAI is essential to prevent AF initiation and recurrence.¹ However, recent reports have demonstrated that approximately 10% of patients with AF² have bidirectional or unidirectional^3,4 epicardial connections (ECs) between the atrium and PV inside the PVAI lines, that play an important role in initiating and maintaining AF.^4,5 These connections are one of the mechanisms for failure to achieve a complete PVAI, contributing to AF recurrence.2, 3, 4, 5, 6

Case report

An 80-year-old male patient was admitted to our hospital to undergo ablation of persistent AF, with a background of hypertension, diabetes mellitus, heart failure, chronic kidney disease, and coronary artery disease. He had received an optimized medical therapy for those conditions. On admission, his blood pressure was 132/66 mm Hg and his heart rate was 72 beats/min and irregular. Precordial auscultation revealed normal cardiac and respiratory sounds. His body mass index was 25.1 kg/m², serum creatinine 1.66 mg/dL (estimated glomerular filtration rate 29.7 mL/min), B-type natriuretic peptide 721 pg/mL, and HbA1c 7.0%. A 12-lead electrocardiogram revealed AF (Figure 1). Echocardiography revealed left ventricular hypertrophy, a normal left ventricular ejection fraction (62%) with no evidence of structural heart disease, and left atrium (LA) enlargement of 48.2 mm. The patient’s CHADS₂/CHA₂DS₂-VASc score was 4/5.

Figure 1.

The admission 12-lead electrocardiogram.

Before ablation, electrical cardioversion was performed because the cardiac rhythm was AF, and then sinus rhythm was restored. However, AF was easily induced during mapping, and electrical cardioversion once again was performed to restore sinus rhythm. Following a double transseptal puncture, we performed voltage mapping using a 3-dimensional (3D) mapping system (EnSite™; Abbott) with a high-density mapping catheter (Advisor™ HD grid catheter, Abbott, Plymouth, MN), and confirmed that the LA had an almost normal voltage (Figure 2A). Then, a circumferential PVAI was performed under electroanatomic guidance with an EnSite mapping system using a TactiCath SE™ open irrigated ablation catheter (Abbott).

Figure 2.

The EnSite (Abbott) (A–D, F) and fluoroscopic (E, G) images during the procedure. The EnSite images demonstrate the voltage maps before ablation (A), pulmonary vein (PV) antrum isolation (B), post ablation of the epicardial connections (D), activation map (C), during pacing by a high-density mapping catheter inside the right-sided PV (yellow arrow in F), and conduction to His bundle (white arrow in F). The fluoroscopic images demonstrate the high-density mapping catheter (E) or ablation catheter (G) placed at the earliest activation site in the right-sided PVs. CS = coronary sinus; His = His bundle; IVC = inferior vena cava; LA = left atrium; LIPV = left inferior PV; LSPV = left superior PV; RA = right atrium; RSPV = right superior PV; RV = right ventricle; SVC = superior vena cava.

After the PVAI, the LA and PVs were reconstructed in detail by a 3D mapping system using the high-density mapping catheter during sinus rhythm. However, AF remained easily induced during mapping, and electrical cardioversion was again performed to restore sinus rhythm. Then, we confirmed whether there were remaining potentials inside the PVAI line of the right-sided PVs (Figure 2B), but there were no concealed low-voltage signals⁷ along the PVAI lines. The activation map demonstrated that the earliest activation site was located at a posterior site of the right-sided PV carina (yellow arrow in Figure 2C). The distance between that site and the PVAI line was 12 mm. Further, we also confirmed no PV capture or conduction to the atrium by pacing from the tip of the ablation catheter along the right-sided PVAI lines with an output of 10 V. Thus, no conduction gaps were suspected.

Then, local activation time isochronal mapping was drawn during pacing with the high-density mapping catheter at the earliest activation site inside the right-sided PVs (yellow arrow in Figure 2E and 2F), demonstrating that the earliest activation site was near the His bundle in the right atrium (RA) (white arrows in Figures 2F and 3A) without any electrical endocardial conduction; the electrical signal spread in a centrifugal pattern from this breakout site. The activation near the His bundle in the RA happened before any other LA activation, including the activation of the circular mapping catheter placed beside the LA appendage (Figures 2F and 3A). The conduction time from the pacing site to the earliest activation site near the His bundle was 176 ms. Accordingly, we considered that this conduction ran through an EC but was not a gap conduction. Radiofrequency energy was delivered to the site of the EC in the right-sided PVs (yellow arrow in Figure 2C) during pacing with a maximum power of 25 W from the tip of the ablation catheter with an output of 5 V (Figure 2G). Then, the conduction from the right-sided PVs to near the His bundle in the RA was steadily abolished (exit block) (white arrow in Figure 3B); however, the EC potential still existed inside the right-sided PVs (incomplete entrance block) (yellow arrows in Figure 3B and 3C).

Figure 3.

Intracardiac electrocardiograms after the complete pulmonary vein (PV) antrum isolation. The pacing from a high-density mapping catheter (yellow arrows in A) placed within the right-sided PVs shows PV capture and conduction to near the His bundle. The activation near the His bundle in the right atrium happened before the activation of the circular mapping catheter placed beside the left atrial appendage (white arrows in A). Radiofrequency energy delivered to the epicardial connection (EC) inside the right-sided PVs steadily abolished the conduction from the right-sided PVs to near the His bundle (white arrows in B), but the EC potential inside the right-sided PV still existed (yellow arrows in B and C). A continuous radiofrequency energy delivery completely abolished the EC potential (blue arrow in C). However, the EC potential recurred (yellow arrows in D and E). Pacing from the tip of the ablation catheter with an output of 10 V inside the right-sided PVs could not conduct to the atria despite the presence of EC potentials inside the right-sided PVs (E). After delivery of radiofrequency energy to the EC potentials inside the right-sided PVs, the EC potentials again were steadily abolished (yellow arrow in F). Then, spontaneous activity within the right-sided PVs without conduction to the atrium was observed (white arrow in G).

A continuous radiofrequency energy delivered at that site completely abolished the EC potential (complete entrance block) (blue arrow in Figure 3C). Next, we only confirmed the potential from the superior vena cava (SVC) at the site where the right-sided PV electrode pairs 13–14 to 17–18 of the circular mapping catheter were placed (white arrows in Figure 3C). However, the EC potential recurred (entrance conduction) (yellow arrows in Figure 3D) during pacing at output of 10 V from the circular mapping catheter to confirm the exit block. Thus, the LA and right-sided PVs were reconstructed in detail by a 3D mapping system using the high-density mapping catheter during sinus rhythm. The activation map demonstrated that the earliest activation site inside the right-sided PVs was the same as the first ablation site. Interestingly, the pacing from the high-density mapping catheter or tip of the ablation catheter at output of 10 V inside the right-sided PVs could not establish conduction to the atrium (exit block) despite the presence of EC potential inside the right-sided PVs (entrance conduction) (yellow arrows in Figure 3E) (unidirectional conduction from the atrium to the right-sided PVs).

The EC potential inside the right-sided PVs was again steadily abolished after commencing radiofrequency energy delivery (complete entrance block) (yellow arrows in Figure 3F). Thereafter, no EC conduction recurred. Then, we confirmed that no potential remained inside the right-sided PVs by detailed mapping using the high-density mapping catheter (complete entrance block) (Figure 2D). We also clarified that electric impulse was not conducted to the atrium by detailed pacing inside the right-sided PVs at output of 5–10 V using the high-density mapping catheter and spontaneous activity within the right-sided PVs without conduction to the atrium (complete exit block) (white arrow in Figure 3G).

Repeated bolus injections of isoproterenol (5 μg) could not induce any extra-PV triggers, including from the SVC and/or AF initiation. During a 30-minute observation period, no recurrence of PV-to-atrium conduction was confirmed. Finally, programmed stimulation induced no arrhythmias following an isoproterenol administration.

Discussion

Discriminating between ECs and gap conduction on the PVAI line is important. Notably, as the earliest activation site after the complete PVAI was located at a site more than 5 mm (12 mm) away from the PVAI line, gap conduction was unlikely.^2,4, 5, 6 Additionally, local activation map isochronal mapping during pacing with a high-density mapping catheter at the earliest activation site inside the right-sided PVs (yellow arrow in Figure 2E and 2F) demonstrated that the earliest activation site was near the His bundle in the RA (white arrows in Figures 2F and 3A) without electrical endocardial conduction. Moreover, pacing at output of 10 V inside the right-sided PVs at the tip of the ablation catheter revealed the far-field SVC signals of the circular mapping catheter (blue arrows in Figure 3B) to be earlier than the atrial signals of the His bundle. These findings possibly indicate that the true earliest site was at the anterior side of the His bundle in the mid-septal RA free wall.

The pattern of electrical spreading at the breakout site and the earlier activation near the His bundle in the RA (white arrow in Figure 2F) support the notion of EC conduction, rather than gap conduction. The activation pattern of the coronary sinus (CS) demonstrated that CS distal activation occurred earlier than CS proximal, suggesting that lateral LA was activated earlier. Activation near the His bundle extending to the LA would have allowed the CS to activate from proximal to distal. Electrical spreading was considered to occur near the His bundle going through the Bachmann bundle. As a result, CS distal activation occurred earlier than CS proximal.

Discriminating between EC and SVC potentials is also important. Notably, after ablation of the EC in the right-sided PVs, we confirmed the SVC potentials that preceded the potentials of coronary sinus at a right superior PV site using the electrode pairs 13–14 to 17–18 of the circular mapping catheter placed at that site (Figure 3F). Moreover, the pacing with a low output of 5 V could capture the right-sided PV carina and conduct to the atrium before ablation of EC (Figure 3A and 3B) (but not with a high output of even 10 V after ablation of EC [Figure 3C–3E]); this does not suggest a far-field capture and local electromyography. Thus, the potentials detected at the right superior PV site from electrode pairs 5–6 to 19–20 of the circular mapping catheter placed in the right-sided PVs (Figure 3B–3F) were definitely EC potentials, not far-field SVC potentials. This was further supported by the noninduction of SVC firing and/or AF initiation by intravenous administration of isoproterenol (5 μg).

After EC ablation inside the right-sided PVs, we confirmed the complete absence of potential by detailed mapping (complete entrance block) (Figure 2D) and no electrical conduction to the atrium by detailed pacing with an output of 10 V (complete exit block) using the high-density mapping catheter inside the right-sided PVs. After EC ablation, we also confirmed the lack of conduction to the atrium of the spontaneous activities within the right-sided PVs (complete exit block) (white arrow in Figure 3G), supporting the first-pass isolation of the PVAI lines.

Considering these findings, it would be possible that the bidirectional PV-to-atrium conduction ran through the EC (exit and entrance conduction), but there was no gap conduction. Moreover, we confirmed that the spontaneous activity sites and EC were in close proximity within the right-sided PVs (white arrows in Figure 3F and 3G). Thus, the spontaneous activities might have originated from this EC site despite the abolishment of the right-sided PV-to-atrium conduction (Figure 3G). These findings suggest that ECs may play an important role in initiating AF. The mechanisms of ECs may contribute to an electrically incomplete PVAI, leading to AF recurrence.2, 3, 4, 5, 6

Interestingly, the EC conduction recurred as unidirectional conduction from the atrium to the right-sided PVs at the same site of the first ablation. Two potential mechanisms might explain the reasons behind the EC being bidirectional in the first ablation but unidirectional in the second ablation. Although the EC conduction from the right-sided PVs to the RA was steadily abolished (exit block) after delivery of radiofrequency energy, the EC conduction from the atrium to the right-sided PVs still existed inside those PVs (entrance conduction) (Figure 3B and 3E). This finding might indicate that the EC conduction from the atrium to right-sided PVs was more tenacious than from the right-sided PVs to the atrium. The next potential mechanism was that there might have been 2 ECs in close proximity inside the right-sided PVs. In the first ablation, the bidirectional EC might have been completely ablated, but not the other unidirectional EC. Since an unstable exit block was observed under the presence of entrance conduction, this connection may become substrate for AF recurrence.³ Thus, it may be reasonable to ablate this unidirectional EC.

A recent report demonstrated that PV mapping and pacing with a circular mapping catheter missed ECs after a PVAI line completion.^4,8 Thus, mapping and detailed pacing using the high-density mapping catheter⁴ after EC ablation allowed us to confirm the entrance and exit blocks of the 4 PVs, including their carinas. Since the high-density mapping catheter could evaluate the ECs as 2-dimensional surfaces with 4 × 4 electrodes and a 3 mm interelectrode distance, resulting in a simultaneous 1.6 cm² mapping field in conjunction with EnSite, this might have facilitated and improved the precision of EC localization as compared to evaluating along a line or at a point with a circular mapping catheter and ablation catheter.⁴ Considering our findings, mapping and pacing inside the PVs using the high-density mapping catheter may be one of the most feasible strategies to find ECs after the completion of a PVAI in patients with AF.

Recent reports demonstrated that carina ablation was required in approximately 40% of patients with AF to complete PVAI isolation.⁹ Furthermore, our recent report demonstrated that the ECs were most dominantly in the right-sided PV carina, and the rate of conduction of ECs from there to the RA was most dominant.² A series of cases in which PVAI required ablation in the RA described its mechanism as involving electrical conduction via intercaval bundle fibers connecting the right-sided PVs to the RA.¹⁰ Thus, this case may be reflected by conduction through fiber(s) connecting epicardially between the right-sided PV carina and the RA.

A previous report demonstrated that unidirectional entrance block with spontaneous activity in the PVs may prove an incomplete PVAI, while bidirectional block can decrease acute PV reconnections and AF recurrence in patients undergoing PVAI for AF.¹¹ Finally, because we achieved complete entrance and exit blocks between the PVs and atrium with the EC ablation in the right-sided PVs and confirmed the noninduction of AF, we hypothesized there would be a low AF recurrence rate. The patient remained well, without any AF recurrence, for 1 year after ablation of the ECs within the PVAI lines. Thus, regarding the outcomes of the present case, the possible appropriate endpoint for a successful ablation of AF may be a complete PVAI resulting from an EC ablation.

Conclusion

To the best of our knowledge, no reports have covered the variable character of ECs during ablation of AF. Thus, this is a clinically important and educational case wherein an interesting variation of ECs, which might contribute to initiating and maintaining AF, was observed during ablation. Finally, we should consider performing an additional targeted ablation of the EC, which may have the possibility of changing its character during AF ablation, to achieve a complete PVAI.