The first description of the autonomic nervous system’s importance in initiating AF was provided by Coumel in the 1990s [43,44,45]. In 2005, Wolf and colleagues [46] demonstrated the feasibility of minimally invasive video-assisted left atrial appendage exclusion and bilateral pulmonary vein isolation for AF. In 2010, Edgerton and associates [47] reported prospective, nonrandomized study results from consecutive patients with symptomatic paroxysmal AF who underwent video-assisted, minimally invasive surgical ablation. The procedure reported included pulmonary vein isolation with bipolar RF application, intraoperative confirmation of transmurality, mapping for and ablation of ganglionated plexi (GP), division of the ligament of Marshall and excision or exclusion of the left atrial appendage [47].

Between March 2005 and January 2008, the procedure was performed in 52 patients with paroxysmal AF. Freedom from symptoms attributed to atrial tachyarrhythmias was 78.0% at 6 months and 63.8% at 12 months. Long-term monitoring, by either a 24-h Holter monitor, 2–3-week event monitoring, or interrogation of an implanted pacemaker, obtained at 6 and 12 months revealed freedom from atrial tachyarrhythmias was 86.3% and 80.8% at 6 and 12 months respectively. Although the authors concluded that the use of clinical symptoms underestimated clinical success as compared with long-term monitoring, it seems equally possible that 24-h Holter monitor and 2–3-week event monitoring overestimated clinical success [47].

Autonomic ganglia are known to play an important role in AF initiation and maintenance [48]. As seen in Fig. 3, the five major left atrial ganglionic plexi (GPs) include the Marshall tract ganglionic plexus (GP), the superior left GP, the anterior right GP, the inferior left GP, and inferior right GP. The GP are located consistently in areas where highly fractionated atrial potentials (FAPs), also known as CFAEs (complex fractionated atrial electrograms). GP may be localized via high-frequency pacing (e.g., cycle length 50 ms, at 12–15 V, 10-ms pulse width) delivered through mapping or ablation catheters to left atrial locations with FAPs to identify sites that exhibit transient AV block during AF. Occasionally, different signs of GP activation (such as activation of ectopic excitation from a PV other than the one adjacent to the GP stimulated) may occur [48].

Adapted from reference 48, with permission

Major left atrial ganglionic plexi. The five major left atrial automatic ganglionic plexi (GP) and axons (superior left GP, inferior left GP, anterior right GP, inferior right GP, and ligament of Marshall) are shown in yellow. The coronary sinus, which is enveloped is muscular fibers that have connections to the atrium is shown in blue. The vein and ligament of Marshall, which travels from the coronary sinus to the region between the left superior PV and the left atrial appendage is also shown in blue. IVC inferior vena cava, LIPV left inferior pulmonary vein, LSPV left superior pulmonary vein, RIPV right inferior pulmonary vein, RSPV right superior pulmonary vein, SVC superior vena cava

Despite the role of autonomic ganglia in initiating and maintaining AF, surgical ablation of AF does not seem to be enhanced by targeting the autonomic nervous system. In 2012, Fragakis et al. [49] noted that ganglionated plexi ablation was becoming popular among groups adopting surgical ablation. However, they noted that randomized data did not exist to clarify its potential benefit [49]. In 2015, Gelsomino and colleagues [50] reported 519 subjects with persistent or longstanding persistent AF who underwent RF maze IV procedures during open heart surgery from January 2006 to July 2013 at three institutions without (Group 1) or with (Group 2) ganglionated plexi (GP) ablation. The primary outcome was AF recurrence off antiarrhythmic drugs. The percentage of patients in sinus rhythm off-antiarrhythmic drugs did not differ between groups. Thus, GP ablation was not demonstrated to be beneficial to maintain stable postoperative sinus rhythm [50]. In 2016, the AFACT Study (Atrial Fibrillation Ablation and Autonomic Modulation Via Thorascopic Surgery), GP ablation for advanced AF during thoracoscopic surgery had no detectable impact on AF recurrence but caused a greater number of major adverse events such as sinoatrial node dysfunction, pacemaker implantation and major bleeding [51]. The 2017 HRS/EHRA/ECAS/APHRS/SOLAECE expert consensus statement on catheter and surgical ablation of AF noted that ablation of autonomic ganglia as an initial or repeat ablation strategy for paroxysmal, persistent, and longstanding persistent AF was not well established [48].

In 2011, Krul and colleagues [52] described a “hybrid” AF ablation approach. The authors used extensive periprocedural electrophysiological testing during thoracoscopic GP ablation, and pulmonary vein antrum isolation. Additional left atrial ablation lines (ALAL) were created in patients with persistent AF and longstanding persistent AF [52].

GPs were located with high-output, high-frequency pacing (18 V, 1-ms pulse width, 1000 Hz) which was confirmed by pacing induced AV block or an increase in the R-R interval > 50%. Pulmonary vein antrum isolation was achieved with bipolar RF energy applied to clamps placed around the PV antrum. The PV antrum was considered isolated if either no bipolar potentials were recorded distal to the scar or if potentials of a slow automatic rhythm, dissociated from the atrial depolarizations were recorded. When ALAL were created, a superior line connecting the encircling lesions around the left and right PVs was added to prevent reentry around both scars. An additional line, similar to a mitral isthmus line, was created during ablation between the superior line and the left fibrous trigone. In selected patients an additional inferior line was created. If an inferior line was created, electrical isolation of the posterior wall box was demonstrated by entry and exit block. At the procedure’s conclusion, an endoscopic stapling and cutting device (Endo Gia stapler, Tyco Healthcare Group, North Haven, CT) was used to remove the left atrial appendage [52]. Thirty-one patients with AF (16 paroxysmal, 13 persistent, two longstanding persistent) were treated. Thirteen of 15 patients with nonparoxysmal AF received ALAL. After a year of follow up, 19 of 22 patients (86%) had no recurrent atrial tachyarrhythmias and were not using antiarrhythmic drugs. Three patients required a sternotomy because bleeding could not be controlled with thoracoscopic surgery. One patient developed a hemothorax due to bleeding from a thoracoscopic port entrance. There were three minor events; during admission, one patient developed a pneumothorax after the chest drains were removed. Two patients had pneumonia treated successfully with antibiotics. No deaths or thromboembolic events occurred [52].

In the same year (and month), Mahapatra and associates [53] compared results from 15 patients who sequentially underwent surgical epicardial-catheter endocardial ablation for persistent or longstanding persistent AF versus 30 who were treated only with catheter ablation. All patients had failed prior catheter ablation. Sequential catheter ablation was performed 4.3 ± 1.3 days after their surgical ablation. After a mean follow-up of 20.7 ± 4.5 months, 86.7% (13/15) of sequential patients were atrial tachyarrhythmia free off antiarrhythmic drugs, compared to 53.3% (16/30) of patients that had repeat catheter ablation alone (p = 0.04) [5, 53].

In 2012, Pison et al. [54] reported a cohort of 26 consecutive patients with AF who underwent simultaneous hybrid thoracoscopic surgical plus transvenous catheter ablation and were followed for up to 1 year. The epicardial lesions were not transmural in 23% of the patients, and endocardial touch-up was needed. Ten had persistent AF, 15 had paroxysmal AF, and 1 had longstanding persistent AF [54, 55]. Endocardial catheter ablation allowed completion of lesion sets that could (or were) not be completed surgically, especially the mitral isthmus and cavotricuspid isthmus lines. Endocardial ablation, guided by epicardial mapping, was used to fill gaps in the lesions applied surgically [54, 55]. The authors deemed the techniques feasible and safe. The single-procedure success rate at 1 year was 83% [54, 55]. Thus, the current form of hybrid ablation was shown to be effective [5, 54, 55].

A 2015 systematic review [56] compared the safety and efficacy of the hybrid procedure to the Cox-maze procedure (with cardiopulmonary bypass support [CPB]) and beating-heart epicardial ablation. At 1 year, sinus rhythm restoration rates were 70, 93, and 80% and sinus restoration without anti-arrhythmic medication was 71, 87, and 72%, for, the hybrid procedures, the Cox-maze, and epicardial surgical ablation, respectively. The minimally invasive Cox-maze procedure with CPB had the lowest incidence of reoperation for major bleeding as well as important safety advantages in conversion to sternotomy [56]. Minimally invasive thoracoscopic surgery may exceed AF catheter ablation results, but limitations in creation of transmural floor and roof lines on the posterior left atrial wall can reduce efficacy compared to open CPB surgery [5].

Despite use of a variety of catheter-based strategies for persistent AF, success rates after a single procedure have ranged from ~ 20–60% [5, 57, 58]. For the combination of persistent and longstanding persistent AF, efficacy rates are ~ 30–40%. Efficacy seems somewhat dependent on the approach used. Longstanding persistent AF may be effectively treated with a composite of extensive index catheter ablation, repeat procedures, and/or pharmaceuticals [5, 57,58,59]. Rostock and colleagues reported that after 2.3 ± 0.6 ablation procedures in 395 patients, 312 (79%) were arrhythmia free (including concomitant antiarrhythmic drug treatment in 38%) at mean follow-up of 15 ± 9 months after their last procedure [59]. However, a 2017 report (not confined to persistent and longstanding persistent AF) found that individuals who underwent an additional ablation had $39,409 greater costs during the subsequent year [60]. These results, as well as the fact that thoracoscopic surgical and catheter AF ablation may result in incomplete isolation of the PVs and the posterior left atrial wall, spurred hybrid strategy development combining these approaches [5, 57,58,59, 61].

Currently, hybrid AF ablation uses subsets of the Cox-maze IV minimally invasive epicardial lesion followed by endocardial catheter ablation to close non-transmural gaps between lesions as well as to address additional reentrant atrial circuits. Unfortunately, like surgical approaches [5, 18, 29, 35, 62], lack of uniform procedural approaches remain including (but not limited to) thoracoscopic vs. pericardioscopic approaches, energy sources used, lesion sets applied, timing between the respective surgical and catheter components, left atrial appendage (LAA) management as well as medical management of these patients [5, 62].

Superior outcomes have been achieved with a bilateral thoracoscopic ablative approach compared to using a pericardioscopic approach which include lower morbidity and mortality rates. The bilateral thoracoscopic approach excludes (e.g., ligation, clipping, or excision) the left atrial appendage in patients with persistent or longstanding persistent AF. Left atrial appendage exclusion has the potential to reduce lifetime risk of stroke as well as electrically isolating/eliminating AF triggers originating from the appendage [62, 63].

In 2019, Al-Jazairi et al. [64] reported the results of single-stage hybrid ablation in 50 consecutive patients. The PVs and superior vena cava were isolated, and a posterior left atrial box was created via thoracoscopic epicardial ablation. Isolation was assessed endocardially, and additional endocardial ablation was performed as needed. The pulmonary veins were isolated via the epicardial approach. Additional endocardial ablation was required to complete box isolation in 21 patients [64].

Five (10%) patients had paroxysmal, 34 (68%) patients had persistent, and 11 patients (22%) had longstanding persistent AF. Twenty-five (50%) individuals had unsuccessful prior catheter ablation(s) [64].

At 1 year, 76% of patients maintained sinus rhythm off antiarrhythmic drugs. Patients with paroxysmal AF had the greatest success, and patients with longstanding persistent AF had the poorest results. The procedure was successful in 100% of patients with paroxysmal AF versus 79% in those with persistent AF versus 55% in those with longstanding persistent AF, (p = 0.039) [64].

Seven patients suffered procedure related complications. Two had bleeding requiring thoracotomy and recovered without sequelae. One required a dual chamber pacemaker after restoration of sinus rhythm. Another one developed pleural and pericardial effusions that required drainage. Unfortunately, two suffered permanent phrenic nerve injury [64].

The authors emphasized that the likelihood of success and potential gains in patient quality of life had to be weighed against the risks of complications. They concluded that individuals with paroxysmal AF after unsuccessful catheter ablation, or those with relatively shorter durations of persistent AF appeared to be the best candidates for hybrid ablation [64].

The “European multicentre experience of staged hybrid atrial fibrillation ablation for the treatment of persistent and longstanding persistent atrial fibrillation” was reported by Haywood and associates in January of 2020 [65]. Patients from four European cardiothoracic centers, underwent a 1st stage video-assisted thoracoscopic (VATS) epicardial surgical AF ablative procedure. After ≥ 8 weeks, 166 of 175 (95%) patients underwent endocardial RF catheter ablation. The majority of gaps found during the endocardial procedure were found around the right PVs (28%) and in the roof line (36%) [65].

At a median follow up of 18 months 93/166 (56%) remained free of atrial tachyarrhythmia recurrence without antiarrhythmic drugs. At last clinic follow-up, 110/166 (62.9%) were in sinus rhythm without antiarrhythmic drugs and 132/175 (75.4%) were in sinus rhythm ± antiarrhythmic drugs. Implantable loop recorders were inserted in 56 patients. At a median 14 months of follow-up, 60% had remained fully arrhythmia free and 80% had an AF burden < 5%. Among all patients who underwent at least the 1st stage procedure, outcomes in persistent and longstanding persistent AF did not differ significantly [65].

Complications of first stage surgical ablation were relatively frequent and occurred in 35 of 175 patients (20%). These included phrenic nerve injuries (20; only four after 3–12 months), hemothorax (6), liver abrasion (1), pleural effusion (3), pneumonia (3), pericarditis (1), pericardial effusion (1), pericardial adhesions (1), gastrointestinal bleeding (1), obstructive ileus requiring hemicolectomy (1), transient ischemic attack (1), stroke (5), temporary acute kidney injury (3), and left atrial appendage thrombus (1; on preoperative transesophageal echo). Complications of second stage catheter ablation occurred in four of 166 patients (2.4%) and included pericarditis (1), bradycardia requiring permanent pacing (2), and damage to a permanent right atrial pacing lead (one; revision not required, pacing not needed). The authors speculated that patients severely affected by AF symptoms who have unfavorable characteristics for catheter ablation alone, may accept the procedural risks in order to gain the increased level of efficacy represented by this approach [65].

A multicenter randomized controlled trial, CONVERGE (Convergence of Epicardial and Endocardial Ablation for the Treatment of Symptomatic Persistent AF), evaluated the safety of hybrid ablation for treatment of persistent and longstanding persistent AF and compared its efficacy to catheter ablation [61]. Endocardial and epicardial procedures were performed in one setting [66].

This industry (AtriCure, Mason, OH, USA) sponsored trial, randomized (in a 2:1 ratio) patients (ages 18–80 years) with symptomatic persistent AF refractory to or intolerant of ≥ 1 class I or class III antiarrhythmic drugs with left atrial diameters ≤ 6 cm to endocardial catheter or hybrid convergent ablation. Longstanding persistent AF was present in 42% of the study participants [66].

Epicardial ablation, in contrast to seemingly preferable techniques described previously, was achieved using pericardioscopic access with vacuum-assisted delivery of unipolar RF energy (EPi-Sense, AtriCure, Mason, OH, USA). Subsequently, endocardial ablation was performed (using an irrigated RF catheter) aiming to assure that PV isolation was complete, linear lesion gaps were filled, and cavotricuspid isthmus block was achieved [66]. If AF termination did not result, complex fractionated atrial electrograms could, at the operator’s discretion, be targeted with/without intent to alter the GPs [66].

The follow-up duration targeted was 1 year [5, 66]. Following a 3-month blanking period, the primary end point, was freedom from atrial tachyarrhythmias [AF/atrial flutter (AFL)/atrial tachycardia (AT)] without use of class I or class III antiarrhythmic agents (except previously intolerable or failed drugs without dosage increase). Secondary endpoints included an AF burden reduction of 90% as well as freedom from AF (only) absent dose increases or new class I or class III antiarrhythmic agents [5, 65].

Hybrid convergent ablation was superior to endocardial catheter ablation for persistent and longstanding persistent AF. Freedom from atrial arrhythmias in the absence of new or increased doses of previously failed (class I/III) antiarrhythmic drugs was 67.7 vs. 50.0% (p = 0.036). Without antiarrhythmic drugs success rates were 53.5 vs. 32.0% (p = 0.0128). Follow-up (via 7-day Holter monitor) at 18 months revealed that ≥ 90% AF burden reduction was achieved in 74% of hybrid convergent patients versus just 55% among those who only underwent endocardial catheter ablation. Major adverse events were more common in the hybrid convergent group (8/102, 7.8% vs. 0/51, 0%, p = 0.0525), however, the difference was not quite statistically significant [5, 67].

In a 2022 report from Bhatia and colleagues [68], 81 consecutive patients underwent stage 1 epicardial (VATs) surgical ablation. During the surgical procedure, acute isolation of the PVs and PW was achieved in all patients. At the end of this stage, the left atrial appendage (LAA) was excluded (AtriClip, AtriCure, Mason, Ohio) [67].

Sixty-four patients underwent endocardial catheter mapping and ablation (stage 2). Endocardial ablation was not performed after surgical ablation in 17 patients. Twelve had stage 1 complications including two intracranial hemorrhages, two instances of cardiac thrombosis, two phrenic nerve injuries, and one instance each of gastrointestinal bleeding, heparin-induced thrombocytopenia, decompensated heart failure, perioperative stroke, pulmonary embolism, and constrictive pericarditis [68].

Stage 2 included mapping of the pulmonary veins and/or posterior wall (PW) seeking reconnections, as well as selective examination of low-voltage zones and searches for potential localized AF drivers. Modes of AF recurrence after surgical ablation were identified by systematic high-density mapping looking for pulmonary vein and posterior wall reconnection as well as other mechanisms during the endocardial portion of hybrid ablation. Gaps in PV isolation or PW isolating box lesion sets were recorded, ablated, and related to post-surgical AF recurrence. Completion of the endocardial lesion set was defined by isolation of the PVs and the PW [68].

Patient-tailored additional lesions were also delivered. Left atrial voltage maps were used to identify low-voltage zones (bipolar voltage < 0.45 mV in sinus rhythm or < 0.31 mV during AF) and were selectively ablated. Low-voltage-zone ablation was accomplished in six patients by creating a line of block to the nearest non-conducting structure. In 21 patients with sustained AF a 64-pole basket was advanced to the left and right atria to identify (a) focal impulses, with centrifugal activation, or (b) rotational activation (reentrant drivers), in localized regions for > 50% of the duration of mapped AF [63, 64]. Ablation was abandoned at sites that increased esophageal temperatures or overlayed regions of phrenic nerve capture [68, 69].

During stage 2, reconnection of the PVs was noted in 18/64 patients (28.1%). Nevertheless, no relationship was found between PV reconnection and the presence or absence of recurrent AF. Acutely, after (stage 2) endocardial ablation, all patients had complete PV isolation. PW isolation was achieved in 57/64 (89.1%) but was not able to be completed in seven patients due to rising esophageal temperatures [68].

Inefficacy events were defined as AF > 30 s in duration. Safety endpoints included atrioesophageal fistula, phrenic nerve paralysis, unsuccessful appendage ligation, stroke or transient ischemic attack, and all-cause mortality [68].

Recurrence of AT/AF after stages 1 and 2 was more prevalent in patients when PW isolation could not be achieved, compared to those with a completed lesion set (p