Building a surgical atrial fibrillation program: Programmatic evolution, safety, and three-year rhythm outcomes of hybrid ablation

Abstract

Objective

Surgical ablation is a guideline-supported treatment for symptomatic persistent and long standing persistent atrial fibrillation (AF) refractory to catheter ablation, but how a program is built, refined, and sustained has not been described. We sought to describe the development of a de novo surgical ablation program as well as its procedural evolution, safety profile, and long-term rhythm outcomes.

Methods

We reviewed the programmatic evolution of our program and growth of procedural volume and complexity. We also analyzed the long-term electrophysiologic and safety outcomes of all patients undergoing minimally invasive surgical ablation from 2020 to 2024. Rhythm assessment was performed using 12 lead ECG, wearable monitor, implantable loop recorder, or ICD/PPM interrogation following a 3-month blanking period after endocardial ablation.

Results

One hundred and twenty-five consecutive patients undergoing hybrid surgical ablation between 2020 and 2024 at our institution were analyzed. Of which, one hundred and twelve had sufficient follow-up data. At 1, 2, and 3 years of follow-up, 94%, 85%, and 70% of patients were in normal sinus rhythm with 67% of those patients remaining off class I/III antiarrhythmic therapy at three years. There were no peri-procedural deaths, conversions to sternotomy, tamponade, strokes, or atrio-esophageal fistulae. Thirty-day adverse events occurred in 4 patients (3.6%) and included three patients with pericarditis and one patient with transient phrenic nerve palsy. The volume at our institution increased fourfold during the study period paralleled by increasing complexity and scope of operations.

Conclusions

This experience provides a roadmap for developing a multifaceted hybrid ablation program with excellent safety and durable rhythm outcomes.

Graphical abstract

Highlights

• •A practical blueprint for building a surgical hybrid atrial fibrillation ablation program
• •70% of patients remained free from atrial fibrillation 3 years after the Convergent procedure
• •Single-center hybrid ablation experience shows excellent safety and long-term freedom from atrial fibrillation

Introduction

Atrial fibrillation (AF) is the most common sustained arrhythmia in adults and carries substantial risk of stroke, heart failure, and decreased quality of life. AF is classified by duration: paroxysmal AF terminates spontaneously or with intervention within seven days, persistent AF is continuously sustained beyond seven days, and long-standing persistent AF is continuous for more than twelve months [1]. This distinction matters to the surgeon because the underlying electrophysiologic substrate differs fundamentally between phenotypes. In paroxysmal AF, arrhythmogenic triggers originate predominantly from the pulmonary veins, and pulmonary vein isolation (PVI) alone results in excellent freedom from atrial fibrillation [2]. In persistent and long-standing persistent AF, progressive atrial dilatation and interstitial fibrosis create a diffuse, self-sustaining substrate—particularly across the posterior left atrial wall, with non-pulmonary vein triggers arising from the left atrial appendage (LAA), the Ligament of Marshall (LoM), and the coronary sinus. Ablation confined to the pulmonary veins therefore addresses only part of the disease in this cohort of patients. Endocardial ablation alone in these patients has consistently shown high rates of recurrence of atrial fibrillation [3].

Contemporary guidelines reflect this gradient of difficulty. The 2023 ACC/AHA/ACCP/HRS guideline assigns catheter ablation a Class 1 recommendation for symptomatic AF in patients in whom class I or III antiarrhythmic drug (AAD) therapy has failed or is not tolerated, and as first-line rhythm control in selected patients with paroxysmal AF [1]. The 2024 ESC/EACTS guideline retains a Class I recommendation for catheter ablation as first-line therapy in paroxysmal AF but assigns only a Class IIb recommendation for first-line catheter ablation in persistent AF, and states that endoscopic or hybrid ablation should be considered (Class IIa) when catheter ablation has failed [4]. The Society of Thoracic Surgeons 2023 guideline correspondingly assigns a Class IIa recommendation to stand-alone surgical ablation for symptomatic AF refractory to class I/III AADs, catheter-based therapy, or both [5]. The patient with symptomatic persistent or long-standing persistent AF who has already failed one or more catheter ablations therefore occupies precisely the space in which support for a purely endocardial strategy is weakest and in which surgical ablation is endorsed by societal guidelines. Procedural options for these patients have traditionally relied on further endocardial ablations targeting the posterior wall of the left atrium. Outcomes of additional endocardial ablations targeting the left atrium have failed to demonstrate improved freedom from atrial fibrillation, likely due to an inability to produce transmural lesions [6].

A hybrid approach allows for a minimally invasive beating heart approach to achieve more durable transmural lesions. In hybrid ablation, a surgeon performs an epicardial ablation of the posterior left atrium through a subxiphoid pericardial window, and the electrophysiologist subsequently completes PVI and documents entrance and exit block endocardially using high-density electroanatomic mapping. Hybrid ablation, combining surgical and transcatheter substrate modification, has consistently yielded superior freedom from AF compared with catheter ablation alone [7], [8], [9], [10], [11]. This evidence has fueled the growth of hybrid ablation programs across the United States and elsewhere.

Superimposed on this landscape is the rapid adoption of pulsed field ablation (PFA). This method of endocardial ablation is a non-thermal catheter-based energy source that applies high-voltage, microsecond-duration electrical fields to produce irreversible electroporation of cardiomyocyte membranes [12]. Adoption has been extremely rapid, including at our own institution, and has altered referral patterns for surgical AF therapy. This is despite no published data showing superior efficacy [12], [13]. How surgical AF programs should position themselves in this environment is unresolved and is a central motivation for the present report.

The landmark CONVERGE trial showed superiority in freedom from atrial fibrillation in patients treated with a hybrid approach compared to endocardial ablation alone. In this trial, 68% of patients in the hybrid arm were free from atrial fibrillation at 18 months, compared to only 30% in the endocardial ablation arm. Since the publication of CONVERGE, additional components have been incorporated into the hybrid procedure in an attempt to improve durability and efficacy, including concomitant management of the left atrial appendage, ablation of the epicardial aspect of the left atrial dome, and division of the Ligament of Marshall. Few studies have reported on the long-term results of hybrid epicardial ablation. Among them, only one has examined the impact of iterative modifications to the original lesion set in follow-up [14], and none has described how a de novo hybrid program is built— a question of practical importance to any surgeon or institution considering adoption.

In the present work, we describe the development, procedural evolution, and 30-day safety of a de novo multidisciplinary hybrid ablation program, and report rhythm outcomes with follow-up to three years in a consecutive cohort of 112 patients. Our aim is to provide a practical, surgically oriented account of what building such a program requires and what it delivers.

Patients and methods

This was a single-center retrospective study of 125 adults who underwent a staged hybrid epicardial-endocardial ablation at our institution between January 2020 and April 2024. These procedures were performed by a single cardiac surgeon and nine electrophysiologists during the study period. We performed a retrospective review of our prospectively collected data with approval from the Institutional Review Board at Trinity Health Ann Arbor Hospital (E-24-1133, 10/29/2024). Preoperative characteristics, intraoperative data, and postoperative outcomes were analyzed. The primary end point was freedom from atrial fibrillation following a 3-month blanking period that began on the date of the endocardial ablation. Primary effectiveness failures were defined by any AF episode of at least 30 s by Holter monitor/Zio patch or any AF recorded on a full 10-s standard 12‑lead electrocardiogram, direct current cardioversion for AF, or subsequent left-sided catheter ablation for AF. Rhythm follow-up was performed using standard 12‑lead ECG (58%), 7-day Zio patch monitoring (21%), implantable loop recorder or permanent pacemaker interrogation (14%), and 24-h Holter monitoring (7%). A subgroup analysis of freedom from AF was performed in patients assessed exclusively by continuous or prolonged monitoring (ILR, PPM, or ≥ 7-day patch), and results are presented in Supplemental Table 1. Secondary outcomes included evaluation for potential benefit with the addition of left atrial appendage ligation, ablation of the Ligament of Marshall, or addition of an epicardial roof line. Additional outcomes included procedure-related complications, hospital length of stay, chest tube removal timeframe, and anticoagulation or anti-arrhythmic drug (AAD) use. Additionally, we report outcomes on five patients who underwent stand-alone robotic-assisted Cox Maze III, analyzed separately from the primary cohort.

Referral pathway

Patients initially selected for hybrid ablation were those with paroxysmal, persistent or long-standing persistent atrial fibrillation who had at least one prior endocardial ablation and/or had failed a class I/III AAD. Towards the end of the study, we also included patients without prior endocardial ablation if they were felt to benefit from an upfront combined approach. Examples of these patients are those with persistent/longstanding persistent atrial fibrillation, arrhythmia-induced cardiomyopathy, or an enlarged left atrium. Absolute contraindications to a hybrid approach were an unacceptable surgical risk profile or a contraindication to anticoagulation. This included patients unable to tolerate single lung ventilation, those with a prior sternotomy, or those found to require a concomitant procedure during their preoperative workup, such as coronary artery bypass grafting.

Clinical pathway and procedural timeline

All patients were referred by an electrophysiologist for surgical evaluation. Our clinical pathway is shown in Supplemental Fig. 1. Preoperative assessment included transthoracic echocardiography, chemical stress test to exclude concomitant coronary artery disease, cross-sectional imaging to define left atrial and appendage anatomy, and pulmonary function testing to ensure patient tolerance to single lung ventilation during thoracoscopic left atrial appendage closure.

Stage 1 (epicardial)

Patients underwent subxiphoid pericardioscopic epicardial radiofrequency ablation of the posterior left atrium. Left-sided video-assisted thoracoscopy (VATS) for left atrial appendage ligation was added after our first ten patients. Later in our cohort, Ligament of Marshall division and an epicardial roof line were performed during the VATS portion of the procedure. All patients were cardioverted to sinus rhythm at the conclusion of the epicardial procedure and admitted post-operatively. Antiarrhythmic drugs and anticoagulation were resumed on POD# 1.

Stage 2 (endocardial)

Our first ten patients underwent the endocardial stage either immediately following the epicardial procedure or on the following day. After that early experience, the endocardial stage was deliberately deferred by six to eight weeks in all subsequent patients, to allow epicardial lesion maturation and resolution of post-procedural pericardial inflammation before endocardial mapping. High-density endocardial mapping identified any residual conduction into the posterior wall or pulmonary veins, and radiofrequency or cryoablation was used to complete pulmonary vein and posterior wall isolation with confirmation of entrance and exit block. Additional lines were added at the discretion of the performing electrophysiologist.

A 3-month blanking period was applied from the date of the endocardial stage. Oral anticoagulation was continued for a minimum of 3 months and thereafter individualized by CHA₂DS₂-VASc score and appendage closure status. Class I/III antiarrhythmic drugs were weaned at the discretion of the treating electrophysiologist. Rhythm surveillance was performed at least yearly with the modality determined by each referring physician.

Surgical technique

Hybrid ablation

Epicardial ablation was performed with the vacuum-assisted, unipolar radiofrequency device (EPi-Sense, AtriCure, OH) (Fig. 1). Pericardial access was obtained through a subxiphoid approach, and the radiofrequency device was positioned inside a pericardioscopic cannula with an endoscope. Three-dimensional electroanatomic mapping was performed epicardially with the Advisor HD Grid device and the Ensite Precision mapping system (Abbott, IL). Epicardial mapping was performed in the patient's presenting rhythm to the operating room. The cannula was placed over the diaphragm and inserted between both left and right inferior pulmonary veins. Sequential epicardial ablations were made from the right to the left inferior pulmonary veins. An additional row of ablations were made cranially from the left to right with overlap. Additional ablations were performed at the pericardial reflection as needed to completely treat the posterior wall. Epicardial mapping was then repeated to ensure electrical quiescence. If greater than 0.2 mV of activity remained, additional ablations were performed until electrical isolation was achieved.

Fig. 1.

Hybrid surgical technique. (A) Pericardioscopic view of the posterior left atrium. (B) EpiSense catheter performing ablation of the posterior left atrium. (C) Completed posterior left atrial ablation (D) Left-sided VATS for left atrial appendage ligation * posterior left atrium.

Three 5 mm ports were then placed in the left chest at the 2nd interspace mid-clavicular line, 4th interspace anterior axillary line, and 8th interspace anterior axillary line. A 10 mm port was placed in the 6th interspace posterior axillary line which was used for the Episense catheter and AtriClip (Fig. 1D). If present, the Ligament of Marshall was ablated at the time of the VATS portion of the procedure. An additional roof line ablation was performed beginning in December 2022. This line was performed along the dome of the left atrium from the right superior pulmonary vein across and onto the left superior pulmonary vein. The Episense catheter sensing function was utilized during this portion to ensure electrical quiescence. Of note, in patients that had a pericardiotomy for left atrial appendage ligation, this incision was carried inferiorly and posteriorly to facilitate drainage of any pericardial fluid post-procedure. All patients were cardioverted to normal sinus rhythm at the conclusion of the epicardial ablation.

Beginning after our first ten epicardial ablations, left atrial appendage exclusion was attempted in all patients. The left atrial appendage was excluded with an Atriclip device (AtriCure, OH) sized individually. All patients had their appendage excluded unless a previously placed endocardial closure device was present (eg. Watchman Boston Scientific, MA).

The second portion of the hybrid ablation was later performed by the referring electrophysiologist. Endocardial mapping was performed to identify any connection to the posterior wall or the pulmonary veins to guide ablation. The endocardial ablation was performed with an irrigated radiofrequency catheter or cryoablation catheter via standard approach to complete isolation of the pulmonary veins and the posterior wall. Isolation of the posterior wall was performed after detailed high density mapping. Roof and floor lines were performed to create a box lesion if needed, and targeted viable voltage substrate ablation was performed to address any breakthrough gaps based on electro-anatomical mapping. Standard entrance/exit block was performed to confirm PVI and posterior wall isolation after endocardial ablation. Additional ablation lines were created at the discretion of the performing electrophysiologist.

Robot assisted MAZE procedure

Patients were placed on cardiopulmonary bypass through the groin either totally percutaneous or after a femoral cutdown and graft placement, depending on the size of the femoral artery. The 8 mm robotic trocars were inserted via stab incisions at the following locations: third intercostal space anterior axillary line for the left arm, fourth intercostal space midclavicular line for the left atrial retractor, and fifth intercostal space mid-axillary line for the right arm. A stab incision in the second intercostal space posterior axillary line was made for the aortic cross clamp. After initiation of cardiopulmonary bypass, the cardioplegia cannula was inserted into the ascending aorta for the administration of antegrade del Nido cardioplegia (Antegrade 14F long MIS; Medtronic). The Da Vinci Xi robotic system (Intuitive Surgical Inc., Mountain View, Calif) was then docked. A biatrial Cox MAZE III was then performed utilizing cryotherapy using the cryoFORM probe (Atricure, OH). This lesion set consisted of epicardial coronary sinus lesion, mitral annular line ablation, and a posterior box ablation which included the LAA orifice for the left atrium. For the right atrium, the lesion set included a tricuspid line, right atrial appendage line, superior vena cava to inferior vena cava line, and right atrial completion T lesion. All left-sided ablations were performed for three minutes, and all right-sided ablations were performed for two minutes. The left atrial appendage was then closed with an Atriclip (Atricure, OH).

Results

Patient characteristics

Between January 2020 and April 2024, 125 patients underwent hybrid epicardial–endocardial ablation as part of the development of a new multidisciplinary surgical atrial fibrillation program. After exclusions for incomplete endocardial completion (n = 4), insufficient follow-up (n = 8), or death (n = 1), 112 patients were included in the final analysis. One patient death occurred from heart failure exacerbation 11 months after hybrid ablation. The mean age was 68 ± 6.9 years, and 63% were male. The mean duration of AF prior to surgery was 7.5 ± 6.0 years. Most patients (75%) had persistent or long-standing persistent AF, and nearly nine in ten (89%) had undergone ≥2 prior endocardial ablations. The mean left atrial diameter was 4.5 ± 0.7 cm, and the mean CHA₂DS₂-VASc score was 3.1 ± 1.7 (Table 1).

Table 1.

Patient and procedural characteristics.

	N = 112 (%)
Age (median, IQR)	69 (65–74)
Male sex	71 (64)
Caucasian race	111 (99)
CHA₂DS₂-VASc score	3 (2–4)
Age at AF Dx
Years since AF Dx	6.5 (3−11)
Cardioversions	68 (60.7)
Cardioversion within 12 months	2 (1–2)
Prior endocardial ablation	99 (88.4)
More than 3 endocardial ablations	21 (18.9)
Type of atrial fibrillation at time of ablation
Paroxysmal	28 (25.0)
Persistent or longstanding	84 (75.0)
Left atrial size	4.5 (4.2–5)
LVEF	60 (55–65)
BMI	35 (31–38)
HTN	96 (85.7)
COPD	14 (12.5)
DM	20 (17.9)
OSA	73 (65.2)
CAD	33 (29.5)
Heart failure	42 (37.5)
Failed AAD therapy	24 (21.4)
Preoperative medications
Amiodarone	22 (19.6)
Dofetilide	5 (4.5)
Dronedarone	2 (1.8)
Flecainide	9 (8.0)
Propafenone	3 (2.7)
Sotalol	3 (2.7)
Beta blocker	80 (71.4)
Calcium channel blocker	34 (30.4)
Coumadin	6 (5.4)
Direct oral anti-Xa inhibitors	103 (92.9)
Epicardial characteristics
Ablation access- subxiphoid	112 (100)
3D mapping	99 (88.4)
Roof-line	46 (41.1)
Vein of marshall	82 (73.2)
Left atrial appendage ligation	95 (84.8)
Intra-operative cardioversion	58 (51.8)
Length of stay (day)	2 (1–3)
Day chest tube removed (days)	1 (1–2)
Endocardial ablation characteristics (n = 111)
Median fluoroscopy time	8 (2–14)
Pulmonary vein isolation	110
Roof line	69 (62.2)
Cavotricuspid-isthmus line	31 (27.9)
Mitral flutter	9 (8.2)
Endocardial success	110 (99.1)
Length of stay	1 (1–1)

Programmatic growth and multidisciplinary integration

Institutional referrals for surgical AF interventions—including hybrid ablation, standalone left atrial appendage ligation, and robotic MAZE—increased sixfold, from 10 referrals in 2020 to 61 in 2023 (Fig. 2). The number of hybrid ablations performed mirrored this growth, peaking in 2023 before a modest decline in 2024 corresponding to the introduction of pulse field ablation (PFA) at our institution. This expansion was accompanied by progressive engagement from the electrophysiology service, growing from a single collaborating electrophysiologist in the first year to nine by the end of the study period, representing a fully integrated, multidisciplinary AF team.

Fig. 2.

Annual referrals for surgical atrial fibrillation intervention, hybrid ablation procedures performed, and stand-alone left atrial appendage clip procedures, 2020–2024.

Procedural evolution and efficiency

The surgical and endocardial protocols evolved over time. Epicardial mapping became standard practice after the first ten patients, with additional lesion sets introduced in later years: roof line ablation in 41% and Ligament of Marshall ablation in 73% of patients. Beginning in 2021, concomitant LAA ligation was performed in 88% of eligible patients, with successful closure confirmed in 94% of follow-up TEEs. This mirrors current published literature for complete closure [15], [16]. Incomplete closures occurred early in the experience and were eliminated by the final two years of the study. Perioperative drainage strategy also evolved: the first ten patients received a subxiphoid pericardial drain alone while two drains were placed once the thoracoscopic portion was added, and from 2023 onward a single left pleural drain was used.

Operative efficiency improved significantly across the study period (Supplemental 2). Mean operative time decreased from 151 ± 28 min in 2020 to 100 ± 17 min in 2024. Median length of stay decreased from 4.3 ± 1.0 to 1.4 ± 0.9 4), largely driven by earlier chest tube removal (postoperative day 4.1 → day 1.0) (Fig. 3). These improvements were achieved without an increase in complications.

Fig. 3.

Median hospital length of stay and postoperative day of chest tube removal, by year of the program.

Electrophysiologic findings and freedom from atrial fibrillation

During the endocardial procedure, complete posterior wall isolation was confirmed in all patients, and redo pulmonary vein isolation was performed in 99%. Median follow-up was 24 months (IQR 15–36 months), and 26.9% of patients were in atrial fibrillation at the last available time point. Using Kaplan-Meier modeling, 94%, 85%, and 70% of patients undergoing hybrid ablation were in normal sinus rhythm at 1, 2, and 3 years, respectively (Fig. 4). Approximately 67% of patients without recurrence of AF remained off class I/III antiarrhythmic therapy at three years (Fig. 5).

Fig. 4.

Kaplan-Meier estimate of freedom from atrial fibrillation following staged hybrid epicardial–endocardial ablation.

Fig. 5.

Antiarrhythmic drug use among patients in normal sinus rhythm after hybrid ablation, by year of follow-up.

On multivariate logistic regression, neither addition of an epicardial roof line nor ablation of Ligament of Marshall significantly improved freedom from AF at 1–3 years (p > 0.05 OR 2.36 CI 0.89–6.25 and OR 0.89 respectively CI 0.27–2.86). However, concomitant LAA ligation was independently associated with improved freedom from AF at 3 years (OR 7.30 CI 1.3–39.1 p = 0.020); among patients without LAA ligation [n = 7], 3 of 7 [43%] were free from AF at 3 years, compared to 19 of 26 [73%] who had ligation performed. The wide confidence interval reflects the limited number of patients who did not undergo LAA ligation late in the study period, suggesting a potential late benefit from elimination of non-pulmonary vein triggers. Of note, in patients that had repeat endocardial ablation after completion of the hybrid procedure, all patients continued to have complete posterior wall isolation (n = 17). The most common reason for additional repeat endocardial ablation was for mitral flutter (n = 8) (Supplemental Fig. 3).

Safety outcomes

All patients were seen in person after 30 days from their surgical procedure and then 30 days from their endocardial ablation. No deaths occurred within 30 days of either procedural stage, and there were no conversions to sternotomy or thoracotomy. Thirty-day adverse events occurred in 4 of 112 patients (3.6%) (Table 3). The most common event was post-procedural pericarditis manifesting as a chest pain with pleural effusion and requiring corticosteroid therapy in 3 patients (2.7%). All of the cases of pericarditis resolved without drainage, tamponade, or readmission. One patient (0.9%) sustained a transient left phrenic nerve palsy during the thoracoscopic portion of the procedure with eventual complete clinical and radiographic recovery. There were no cases of cardiac tamponade, pericardial effusion requiring drainage or pericardiocentesis, atrio-esophageal fistula, pulmonary vein stenosis, myocardial infarction, stroke, transient ischemic attack, or major bleeding, and no patient required an unplanned return to the operating room.

Table 3.

Thirty-day safety outcomes following hybrid epicardial–endocardial ablation (n = 112).

Post operative complications	All patients (n = 112)
Pericarditis requiring steroids	3
Phrenic nerve injury, n	1
Cardiac tamponade, n	0
Pericardiocentesis, n	0
Pericardial effusion requiring drainage, n	0
Severe pulmonary vein stenosis, n	0
Increased bleeding, n	0
Conversion to open procedure, n	0
Myocardial infarction, n	0
Cerebral vascular accident, n	0
Transient ischemic attack, n	0
Atrioesophageal fistula, n	0
Death, n	0

* Adverse Events (AE).

**All complications/adverse events recorded within 30 days of procedure.

Expansion to robotic and standalone procedures

Programmatic experience led to the introduction of a robotic-assisted Cox MAZE procedure in late 2022 for highly selected patients with persistent or long-standing persistent AF. Robotic assisted Cox MAZE was performed on five patients. These patients were analyzed separately and were not included in the primary 112-patient hybrid ablation cohort or the Kaplan-Meier freedom from AF analysis. This cohort demonstrated 100% freedom from AF off antiarrhythmic drugs and anticoagulation at up to 3 years of follow-up (Table 2). In parallel, standalone thoracoscopic LAA ligations increased annually from none in 2020 to five in 2024, primarily in patients unsuitable for percutaneous closure or unable to tolerate short-term anticoagulation.

Table 2.

Long-term follow-up of patients undergoing robot-assisted Cox Maze III.

Patient	Time from surgery	Type of AF	Monitor type	Class I/III AAD	Anticoagulation	Rhythm
1	36 months	LSPAF	ILR	No	No	NSR
2	31 months	Persistent	PPM	No	No	NSR
3	24 months	Persistent	7-day Zio	No	No	NSR
4	12 months	LSPAF	7-day Zio	No	No	NSR
5	6 months	Persistent	7-day Zio	No	No	NSR

LSPAF – longstanding persistent atrial fibrillation, ILR – implantable loop recorder, PPM – permanent pacemaker, NSR – normal sinus rhythm.

Discussion

This study presents one of the first detailed accounts of the creation and maturation of a real-world hybrid surgical AF program. Beyond reporting favorable long-term rhythm outcomes, our experience provides a practical framework for how a multidisciplinary surgical AF program can evolve—from inception, through iterative refinement of technique, to institutional integration—while maintaining excellent safety and efficacy.

While the CONVERGE trial and subsequent hybrid ablation studies have established clinical efficacy, few have addressed the successful implementation and scaling within an existing cardiovascular service line [14]. Our data show that structured collaboration between cardiac surgery and electrophysiology is essential for sustainable growth. Over four years, participation expanded from a single electrophysiologist to nine, with referrals increasing sixfold. This progressive integration was accompanied by diversification of procedural offerings, including robotic MAZE and standalone LAA ligation, reflecting the maturation of a comprehensive surgical AF program.

The decrease in hybrid ablation volume in 2024 corresponded to the introduction of pulse field ablation at our institution, underscoring the importance of program adaptability amid changing ablation technologies. As emerging energy sources are adopted, long-term comparative data such as ours will remain critical in determining the optimal procedural approach for durable rhythm control.

An important finding of our experience is that procedural efficiency improved significantly over time without compromising safety. Mean operative time decreased by one-third, and postoperative length of stay was reduced from four to one day, largely due to refinements in pericardial management and early drain removal. These improvements mirror the learning curve and standardization seen with other complex cardiac procedures, such as minimally invasive valve surgery.

Similarly, incorporation of adjunctive epicardial mapping, Ligament of Marshall ablation, and LAA ligation was an important step in our procedural evolution. Although additional lesion sets did not statistically improve freedom from AF, the improved transmurality and reduced need for endocardial touch-up ablations that we observed suggest a possible electrophysiologic benefit, though the present analysis is not powered to demonstrate such a statistical effect. Importantly, 100% of patients undergoing repeat ablation demonstrated a quiescent posterior wall, consistent with durable lesion formation. The association between LAA exclusion and improved long-term rhythm success in our cohort further supports the growing evidence that the LAA may serve as a source of non-pulmonary vein triggers in persistent AF.

Long-term rhythm success in this cohort compares favorably with both randomized and observational studies of hybrid ablation. Freedom from AF of 70% at three years is similar or superior to other hybrid trials, despite our inclusion of a higher-risk population with multiple prior ablations and longer AF duration [10], [11], [14]. The absence of peri-procedural mortality, major bleeding, or pericardial effusion requiring drainage attests to the safety of the hybrid approach when performed within a dedicated, experienced program.

The results from our early robotic MAZE cohort further emphasize the importance of program diversification. While small, this highly selected group achieved 100% freedom from AF off antiarrhythmic therapy and anticoagulation, reinforcing the role of minimally invasive surgical ablation in selected patients and illustrating how programmatic infrastructure can be leveraged to expand procedural capability. Given the very small sample size (n = 5), definitive conclusions regarding the efficacy of the robotic Cox MAZE III should not be drawn from this cohort; these results are presented as illustrative of programmatic evolution rather than a declaration of superior efficacy.

Catheter ablation using irreversible electroporation, or pulsed field ablation (PFA), has been adopted more rapidly than any preceding ablation technology. Pulsed field ablation introduction at our institution in 2024 coincided with the decline in hybrid volume during the study period. Its principal advantages—relative tissue selectivity and markedly shorter procedure times—are well established for paroxysmal AF [12], [13]. Its performance for patients with non-paroxysmal atrial fibrillation is far less certain. Registry and comparative data show that pulmonary vein isolation after PFA, while broadly comparable to radiofrequency, is not uniformly durable [17], [18], [19], and failure of transmural posterior wall isolation after PFA has been directly demonstrated by subsequent epicardial mapping [20]. Varying posterior wall thickness and presence of epicardial fat are plausible mechanisms, and these are precisely the constraints that epicardial ablation is designed to overcome. What our data does support is more modest but arguably more useful: as catheter technology evolves, an integrated surgical AF program retains a defined role for patients with advanced atrial substrate, prior failed catheter ablation, or an indication for concomitant appendage exclusion, and provides the institutional infrastructure within which new modalities can be evaluated rather than simply adopted.

This study is limited by its retrospective single-center, single surgeon design and inherent selection bias. Endocardial lesion sets and energy sources were not standardized and were left to the discretion of individual electrophysiologists, which introduces variability but also reflects real-world practice. Follow-up rhythm assessment was not uniform, ranging from spot electrocardiograms to continuous monitoring. Nonetheless, the consistency of rhythm outcomes across time intervals supports the robustness of our findings. Importantly, although 3-year freedom from AF outcomes are encouraging, the Kaplan-Meier curve demonstrates continued attrition in rhythm success beyond this point, and longer prospective follow-up will be required to determine the true durability of this hybrid approach. Additionally, approximately one-third of patients in normal sinus rhythm remain on class I/III AAD therapy, which constitutes failure by strict HRS criteria; the HRS-defined success rate of approximately 48% at 3 years is an important consideration when comparing these results to published benchmarks.

Future areas of investigation for hybrid programs should include prospective standardization of lesion sets, integration of new energy modalities (including PFA-assisted hybrid ablation), and formal cost-effectiveness analysis of the hybrid strategy. Additionally, long-term registry data will be critical to determine which combinations of lesion sets and LAA management strategies provide the greatest durability.

In summary, we demonstrate that a surgical AF program can be safely developed and scaled with excellent rhythm outcomes through multidisciplinary collaboration, procedural refinement, and systematic integration into institutional workflow. The durability of rhythm success and the safety profile achieved underscore the hybrid approach as a cornerstone of contemporary AF management. Beyond procedural outcomes, the true innovation of this work lies in illustrating how programmatic infrastructure—not just technology—drives success in surgical AF therapy.

CRediT authorship contribution statement

Kristin Pacholke: Writing – original draft, Visualization, Data curation, Conceptualization. Yas Sanaiha: Writing – review & editing, Visualization, Formal analysis, Data curation. Mohammed El Nayir: Writing – original draft, Data curation. Jihn Han: Writing – original draft, Conceptualization. Peter Downey: Writing – review & editing, Visualization. Peyman Benharash: Writing – review & editing, Supervision, Methodology, Formal analysis. Robert Lyons: Writing – review & editing, Writing – original draft, Visualization, Supervision, Methodology, Investigation, Data curation, Conceptualization.

Funding

None.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Robert Lyons reports a relationship with AtriCure Inc. that includes: speaking and lecture fees. Peyman Benharash reports a relationship with AtriCure Inc. that includes: speaking and lecture fees. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplemental Fig. 1.

Referral algorithm.

Supplemental Fig. 2.

Operative time.

Supplemental Fig. 3.

Repeat endocardial ablation findings.

Supplemental Table 1

Rhythm breakdown by monitoring method.