Effect of Mitral Isthmus Ablation on Recurrence Rates in High-Burden Paroxysmal Atrial Fibrillation

Introduction

Atrial fibrillation (AF) is a globally prevalent disorder that manifests with symptoms such as palpitations and chest tightness, predisposing individuals to complications including heart failure and cardioembolic stroke. Since Haissaguerre et al reported that pulmonary vein ectopy triggered and initiated AF [1], catheter ablation for AF has achieved substantial advancements. Current guidelines recommend pulmonary vein isolation (PVI) as the cornerstone first-line therapy for AF [2,3]. Paroxysmal AF represents the most common AF subtype and varies in the severity of AF burden. As the disease progresses, the AF burden in paroxysmal AF gradually increases, resulting in electrophysiological changes, such as enlargement of the left atrium (LA), reduced atrial voltage, and expanded distribution of scar tissue [4–6]. Clinically, these characteristics also begin to resemble those of persistent AF. Studies on PVI in patients with paroxysmal AF with different AF burdens have also found that those with a higher AF burden have a higher rate of AF recurrence after the initial ablation procedure than do those with a lower AF burden [7,8]. Consequently, a uniform approach, performing only PVI for all patients with paroxysmal AF, is suboptimal. Ablation strategy selection should be tailored according to the AF burden.

The mitral isthmus (MI) bridges the left inferior pulmonary vein (PV) and mitral annulus, incorporating the left atrial appendage, coronary sinus, and ligament of Marshall. Its anatomically complex architecture provides the substrate for arrhythmogenesis [9,10]. In addition, it serves as an important anatomical substrate for significant macro reentrant circuits within the LA, particularly following PVI. Literature reports that 76% of patients with persistent AF who undergo PVI alone will experience a recurrence of LA-dependent macro reentrant atrial tachycardia [11]. Therefore, adjunctive MI ablation may theoretically reduce the recurrence of AF. However, the technical challenge of achieving bidirectional blockade in the MI has discouraged many electrophysiologists for an extended period. The advent of ethanol infusion of the vein of Marshall (EI-VOM) has increased the success rate of MI block. The Vein of Marshall Ethanol for Unablated Persistent Atrial Fibrillation (VENUS) trial showed that PVI combined with MI ablation based on EI-VOM for treating persistent AF not only increased the MI block rate from 51.3% to 80.6%, but also improved freedom from AF/atrial tachycardia at 6 and 12 months [12].

The electropathological substrate of patients with high-burden paroxysmal AF tends to approximate that observed in patients with persistent AF, and MI ablation based on EI-VOM reduces recurrence in persistent AF. Therefore, in this study, we aimed to compare outcomes from PVI with and without MI ablation for high-burden paroxysmal AF.

Material and Methods

STUDY PROTOCOL:

Patients who received a diagnosis of high-burden paroxysmal AF by electrocardiograph (ECG) recording and underwent catheter ablation for the first time were recruited from January 2023 to September 2023. Patients undergoing the ablation strategy of bilateral PVI alone were included in the “2C” group, while patients who underwent bilateral PVI plus linear ablation of the MI were included in the “2C1L” group (Figure 1). High-burden paroxysmal AF was defined as ≥4 self-terminating episodes of AF in the preceding 6 months, with ≥1 episode lasting over 6 h, confirmed by ECG recording [7,8,13]. To mitigate selection bias, a propensity score matching technique was applied using a 2: 1 matching ratio between the 2C group and 2C1L group. This approach incorporated variables such as age, sex, duration of AF, CHA2DS2-VASc score (Congestive heart failure, Hypertension, Age ≥75 years, Diabetes mellitus, Stroke/transient ischemic attack/thromboembolism, Vascular disease, Age 65 to 74 years, Sex Category [female]), and left atrium diameter. The matching process was conducted within a propensity score caliper of 0.05. The inclusion criteria were as follows: (1) age between 18 and 80 years, and (2) symptomatic high-burden AF intolerant to ≥1 antiarrhythmic medication. Patients were excluded if they had severe structural heart disease or left atrial diameter exceeding 60 mm or volume ≥200 mL. The research adhered to the principles outlined in the Declaration of Helsinki and received approval from the Ethics Committee of Henan Provincial People’s Hospital (number 18 [2025], 2025.01.16). Prior to undergoing the procedures, all patients provided their written informed consent.

PROCEDURAL STEPS OF 2C GROUP: Anesthesiologists administered general anesthesia to all patients. After puncturing the femoral veins, a steerable decapolar catheter (DecaNAV; Biosense Webster, Irvine, CA, USA) was used to create the matrix and was subsequently advanced into the coronary sinus. Concurrently, an intracardiac echocardiography catheter (Soundstar; Biosense Webster) was positioned within the right atrium to facilitate navigation of catheters and needles during the transeptal puncture procedure. Two transeptal punctures were performed at an optimally selected site on the interatrial septum using a Swartz sheath (Abbott, Chicago, IL, USA), and these procedures were conducted under the guidance of intracardiac echocardiography and X-ray. Subsequently, using a multi-electrode mapping Pentaray catheter (Biosense Webster), an electroanatomical map of the PVs and LA was reconstructed, with detailed documentation of left atrial voltage and the distribution of low-voltage areas (sinus rhythm voltage <0.5 mV) [14]. A 56-porous irrigated-tip contact force sensing catheter (Thermocool SmartTouch SF; Biosense Webster) was then used to localize and ablate the PV antrum. No additional ablation was performed unless an associated arrhythmia was identified.

PROCEDURAL STEPS OF 2C1L GROUP: In addition to the bilateral PVI procedure described above, MI ablation was performed. The steps of MI ablation have been previously described [15]. In step 1, the EI-VOM was performed if the VOM was present, which was identified by the coronary sinus venograms. The conduction block of MI was evaluated after that. If the endpoint was not achieved or there was absence of the VOM, step 2 was initiated. In step 2, endocardial “V-shape” linear ablation of the MI was performed. The upper ablation line was performed from 1 to 2 o’clock of the mitral annulus toward the left inferior PV; the lower corresponded to the ostium of the VOM from the left inferior PV to the lower mitral annulus. In addition, reinforced ablation was performed for those points on which conduction time and/or sequence changed during ablation. If the endpoint was not reached, step 3 was initiated. In step 3, the earliest activation sites around the V-shape ablation lines were identified and intensively ablated. If epicardial activation was earlier than the endocardium, or if the endpoint was still not achieved, step 4 was initiated. In step 4, precise epicardial ablations were performed at the ostium of the VOM and those positions directing to the endocardial points on which the activation was earliest or conduction time and/or consequence changed during endocardial ablation.

Radiofrequency applications were displayed automatically by Predefined VisiTag (Biosense Webster) settings. For PVI, the target ablation index values were set between 450 and 550, using 10 to 15 g of contact force for the anterior walls. For the posterior walls, target ablation index values were between 350 and 450, with a contact force of 5 to 10 g. Radiofrequency energy was applied at a high-power setting of 50 W (in accordance with the specified ablation index), except for the coronary sinus ablation, which used a reduced power of 25 W.

The immediate endpoint of PVI was the absence or dissociation of electrical activity within the PVs, along with a bidirectional conduction block. If these endpoints were not achieved, immediate touch-up ablations were performed at the earliest activation sites. After a waiting period of more than 20 min, repeat electrophysiological examinations were conducted using the Pentaray catheter during sinus rhythm and pacing along the lesions, to detect any acute pulmonary vein reconnection (PVR). Additional touch-up ablations were conducted if such reconnections were found. The final ablation endpoint was the establishment of the bidirectional blockade and the absence or dissociation of electrical activity within the PVs throughout the procedure. For MI ablation, the endpoint was defined as achieving a bidirectional conduction block according to a specific protocol. This involved observing a distal-to-proximal activation sequence on one side of the block line when pacing on the other side, indicating an activation detour.

FOLLOW-UP:

The postoperative period from 0 to 3 months was defined as the blanking phase, during which patients could use antiarrhythmic medications or undergo cardioversion as needed. After the blanking phase, outpatient follow-up visits were scheduled at 3, 6, 9, and 12 months for clinical assessments and 12-lead ECGs. At 6, 9, and 12 months, 7-day Holter monitoring was used for continuous 7-day cardiac monitoring. If patients exhibited symptoms of recurrence, an ECG was promptly performed if feasible, and 7-day Holter monitoring was recommended if symptoms were frequent. AF recurrence was defined as any symptomatic or asymptomatic AF/atrial flutter/atrial tachycardia occurring from 3 to 12 months after ablation, with episodes lasting ≥30 s. Recurrent AF was managed with antiarrhythmic therapy, cardioversion, or additional catheter ablation. This could include repeat PVI or targeted ablation based on atrial tachycardia mapping at the physician’s discretion.

The primary outcome was the maintenance of sinus rhythm after the initial ablation procedure, without the use of antiarrhythmic drugs, cardioversion, or repeat ablation, and occurring after the 3-month blanking period. The primary safety outcome included acute procedure-related complications. The secondary outcome included measures of acute effectiveness, such as first-pass isolation of PVI circles and the absence of acute PVRs. Procedural parameters were recorded as relevant metrics, including detailed documentation of left atrial voltage, the bidirectional conduction blockade rate of the MI, procedural time, fluoroscopy time, and more.

STATISTICAL ANALYSIS:

Considering the potential impact of baseline demographic differences on ablation outcomes in both cohorts, a 1: 2 propensity score matching approach was used to mitigate possible confounding factors in the selection of cases. Normally distributed continuous data are presented as the mean±standard deviation and analyzed using the t test. For non-normally distributed continuous variables, results are expressed as median (IQR) and evaluated through the Mann-Whitney U test. Categorical data are summarized as counts (n) and percentages (%) and analyzed using either the chi-square test or Fisher exact test. Kaplan-Meier analysis was conducted to compare the difference in freedom from AF, atrial flutter, or atrial tachycardia between the 2 groups, and the significance was determined using the log-rank test. This differential effect was further compared in a subgroup analysis of patients with low-voltage zones. RStudio (Rstudio, Inc) and Statistical Package for the Social Sciences (version 25.0; IBM Corp, Armonk, NY, USA) software were used for statistical analyses. A value of P<0.05 with 2-sided testing was considered statistically significant.

Results

PATIENT CHARACTERISTICS:

A total of 51 patients were recruited into the 2C1L group, matched by 102 patients in the 2C group. In the overall population, 26.8% were women, with an average age of 61.0±10.0 years. The baseline characteristics of the 2 groups are presented in Table 1. There were no significant differences between the 2 groups.

PROCEDURAL OUTCOMES:

As depicted in Table 2, all PVI circles successfully achieved bidirectional blockade in both groups. There was higher incidence of first-pass isolation (94.1% vs 79.5%, P=0.034) with comparable incidence of acute PVR (3.9% vs 4.9%, P=1.000) of left PVI circles in the 2C1L group, compared with the 2C group, while no significant difference was observed in the right PVI circles between the groups (first-pass isolation: 84.3% vs 82.4%, P=0.761; acute PVR: 11.8% vs 13.7%, P=0.734). The 2C1L group required fewer touch-up ablations in left PVI circles (P=0.001), with comparable sites in the right PVI circles (P=1.000), compared with the 2C group. Furthermore, the MI blockade rate was 98.0% in the 2C1L group. New distributions of endocardial low-voltage areas frequently occurred in the MI (47/51 [92.2%]) and at the ridge of the left PVs (42/51 [82.4%]) after EI-VOM (Figure 2). Additionally, the 2C1L group required longer total procedural time (137.6±18.5 vs 122.6±19.6 min, P<0.001), operation time (59.7±8.8 vs 51.8±8.8 min, P<0.001), and fluoroscopy time (14.1±2.7 vs 7.2±2.7 min, P<0.001).

FOLLOW-UP:

At 12-month follow-up after the initial procedure, 4 of 51 patients (7.8%) in the 2C1L group experienced AF recurrence, including 3 cases of AF and 1 case of atrial tachycardia detected by 7-day Holter monitoring. In contrast, the 2C group exhibited a higher AF recurrence rate, with 21 of 102 patients (20.6%) at 12 months, including 10 cases of AF, 5 of atrial tachycardia, and 6 of atrial flutter. Kaplan-Meier analysis demonstrated a higher rate of freedom from AF/atrial flutter/atrial tachycardia in the 2C1L group than in the 2C group (Figure 3, log-rank test, P=0.045). Subgroup analysis revealed that the reduction in AF recurrence was observed more in patients with low-voltage zones, but there was no significant difference (Figure 4, P=0.12). Six patients underwent a redo procedure. In the 2C1L group, 1 patient with AF recurrence was found to have reconnection of the right PVI circle. In the 2C group, 2 patients presented with MI-dependent atrial flutter, and the remaining 3 patients exhibited reconnections of the PVI circles, 2 on the left side and 1 on the right.

COMPLICATIONS:

The rate of major complications was low and comparable between the 2 groups (3.9% vs 4.9%, P=1.000). In the 2C1L group, 1 patient experienced pericardial tamponade, which was successfully managed with pericardiocentesis, and 1 patient developed a pseudoaneurysm. In the 2C group, there were 2 cases of pericardial effusion, 1 pseudoaneurysm, and 2 hematomas. Although additional EI-VOM and MI ablation did not appear to increase complications, further studies with larger populations are warranted to confirm this.

Discussion

STUDY LIMITATIONS:

This was a single-center, small-sample, non-randomized study. The follow-up period was limited to 12 months, and long-term effects beyond that period were not evaluated. It was not possible to assess the status of PVI circles and MI blockade in all patients at 12 months. The definition of high burden and the monitoring method used during follow-up may have underestimated the actual AF burden and overestimated success rates. To further confirm the benefit of MI ablation after PVI in patients with high-burden paroxysmal AF, larger multicenter randomized controlled trials are needed.

Conclusions

In patients with high-burden paroxysmal AF, PVI combined with MI blockade resulted in a higher rate of freedom from AF/atrial flutter/atrial tachycardia at 12 months after the procedure than did PVI alone.