Cardiac Rhythm Management
Articles Articles 2013 September

Combined Endocardial and Epicardial Ablation for Symptomatic Atrial Fibrillation: Single Center Experience in 100+ Consecutive Patients


Our Lady of the Lake Hospital, Baton Rouge, LA

PDF Download PDF

ABSTRACT.  Our aim was to document and evaluate the outcomes of the first 100+ patients who underwent the convergent procedure at Our Lady of the Lake Hospital. Between May 2010 and December 2011, 104 symptomatic atrial fibrillation (AF) patients underwent the convergent procedure combining surgical epicardial radiofrequency ablation and endocardial ablation. Antiarrhythmic drugs were discontinued at 8 weeks. Arrhythmia episodes were detected by electrocardiogram at 1 month, 3 months, 6 months, and 1 year. A≥72-h patient monitoring or interrogation of Permanent Pacemaker/Implantable cardioverter-defibrillator was performed at 6 and 12 months. Of the 104 patients (age 60.9 years, 77% males, body mass index 32.7, ejection fraction 56.1%, left atrial diameter 4.1 cm) paroxysmal AF was present in 27% and persistent/longstanding persistent AF in 73%. AF duration was 5.2 years. At 12 month follow up 87.5% (63/72) of patients were in sinus rhythm (SR)±antiarrhythmic drugs (AADs). At last follow-up 89.0% (92/104) of patients were in SR±AADs. Three patients underwent a repeat catheter ablation for atypical atrial flutter. No complications <7 days, no atrio-esophageal fistulas, myocardial infarction, or death were reported. The convergent procedure, bringing together the strengths of the endoscopic epicardial ablation and endocardial catheter ablation, provides a viable and promising treatment option for patients with symptomatic AF.

This work was supported by a restricted grant from nContact Surgical, Inc., Morrisville, North Carolina. Dr. Civello and Dr. Boedefeld are consultants for nContact Surgical, Inc. None of the authors have a financial interest in nContact Surgical, Inc. The manuscript was prepared independent of industrial editing.
Manuscript received July 1, 2013, final manuscript received August 19, 2013.

Address correspondence to: Kenneth Civello, MD, MPH, FACC, Louisiana Cardiology Associates, 7777 Hennessey Blvd, Suite 1000, Baton Rouge, LA 70808. E-mail:


Much debate has been made between cardiothoracic surgeons and electrophysiologists regarding the best way to treat patients who have drug-resistant atrial fibrillation (AF). Although surgical procedures, such as the “cut and sew” Cox maze have been successful in the treatment of AF, it requires cardiopulmonary bypass and has a higher rate of bleeding and need for permanent pacing. 1 The invasive nature of this procedure and number of surgeons who perform the procedure have also limited its use. Minimally invasive surgical ablation procedures, such as the “mini-maze” have not resulted in the success rates of open chest procedures. 2,3

Percutaneous pulmonary vein (PV) antrum isolation has also been challenging for electrophysiologists, and has not been demonstrated to result in the efficacy that has been seen in the Cox maze for patients with persistent and permanent AF. 4–6

The convergent procedure was developed to create a complete and comprehensive pattern of linear lesions under direct endoscopic visualization, while avoiding chest incisions, deflation of the lungs, and invasive heart dissections that have hindered electrophysiologist adoption of surgical ablation. It takes a multidisciplinary treatment to produce a complete comprehensive lesion pattern as both epicardial and endocardial ablation are required. Following epicardial ablation, standard electrophysiology (EP) testing is used to confirm isolation of the PVs and the posterior left atrium. 7 The purpose of this paper is to present our outcomes of the multidisciplinary convergent procedure for the treatment of paroxysmal, persistent and longstanding persistent AF patients.



Between May 2010 and December 2011, the first 104 AF patients received the convergent procedure at Our Lady of the Lake Regional Medical Center, Baton Rouge, LA. Patients were between the ages of 37 and 79 years and had documented AF based on the Heart Rhythm Society definitions. 8 Paroxysmal, persistent, and longstanding persistent AF patients were selected to undergo the convergent procedure. Patients were excluded if they had prior open heart surgery or a major abdominal procedure.

The convergent procedure

The procedure combines the expertise of both a cardiothoracic surgeon and electrophysiologist for the treatment of AF. In all patients, epicardial ablation was performed before endocardial mapping and ablation. Epicardial ablation is performed through a transdiaphragmatic access without chest incisions or ports, deflation of the lungs, or heart dissections. 7 Because the pericardial reflections between the left atrium and pericardium are not dissected, there are gaps between the continuous linear epicardial lesions. These known breakthrough locations are mapped and connected endocardially to complete the comprehensive pattern, as shown in Figure 1 . After endocardially completing the pattern, EP testing is performed to assure lesion transmurality, pattern completeness, and PV isolation.


Figure 1: Ablation lesions and pericardial reflections.

Description of convergent procedure

By definition, the convergent procedure is the completion of both epicardial and endocardial components of the procedure. The procedure was performed under general anesthesia. Both epicardial ablation and endocardial ablation were performed in the EP laboratory. Perioperative anticoagulation consisted of cessation of warfarin 3 days before the procedure and initiation of aspirin 325 mg daily. Patients with a CHADS score >2 were bridged with Enoxaparin prior to the procedure. Transesophageal echocardiography was performed on all patients to exclude left atrial thrombus the morning of the ablation.

An esophageal temperature probe was inserted to monitor esophageal temperature during both the epicardial and endocardial ablation.

The patient was prepped and draped for both abdominal epicardial access and catheterization through the femoral veins. Endocardial ablation was performed immediately after epicardial ablation and closing of abdominal access incisions.

Transdiaphragmatic access and epicardial ablation procedure

Access to the posterior surface of the heart was achieved via a transdiaphragmatic pericardial window created endoscopically under CO2 insufflation using a 5-mm Optiview trocar inserted in the left upper quadrant of the abdomen for laparoscopic exploration ( Figure 2a,b ). An additional trocar (10/12 mm) is inserted in the subxyphoid region and a 5-mm trocar in the right upper quadrant. A pericardioscopic cannula (nContact Inc., Morrisville, NC) is inserted through the pericardial window and manipulated throughout the pericardial space to allow for positioning the epicardial ablation electrode ( Figure 2c,d ). After access to the posterior of the heart is achieved, long linear lesions of the comprehensive biatrial pattern ( Figure 2e–g ) are created under direct endoscopic visualization without dissecting the pericardial reflections. 9 Epicardial ablation is performed along the posterior left atrium and around the antrum of the PVs. Each lesion was created utilizing the preset radiofrequency (RF) generator settings for power (30 W) and time (90 s). After completing all epicardial lesions and interrogating the pericardial space to ensure hemostasis, a pericardial drain was inserted through the cannula and into the pericardial space. The free end of the drain was pulled through one of the abdominal punctures. Abdominal incisions were then closed in standard fashion utilizing standard general surgical techniques to reduce the potential for incisional hernias.


Figure 2: Photograph sequence from endoscope during access and epicardial ablation. (a) Using an electrocautery tool to create an incision in the central tendon of the diaphragm. (b) Endograsper sizing the width of the incision created in the central tendon of the diaphragm. (c) View from inside the cannula as it is being introduced through the incision created in the diaphragm. (d) Visualization of the posterior left atrium with the left pulmonary veins visible in the bottom right corner. (e) Positioning the epicardial coagulation device with the exposed radiofrequency coil adjacent to the target posterior left atrial tissue. (f) Visualization of the lesion created and tissue discoloration resulting from the application of radiofrequency energy using the epicardial coagulation device. (g) Anatomical drawing illustrating the location of lesions created by the epicardial coagulation device.

All epicardial lesions were created using the Numeris® Coagulation System (nContact, Inc., Morrisville, NC), which utilizes a vacuum irrigated unipolar RF electrosurgical device with a 3-cm-long directional ablation electrode. Suction applied to the probe ensures consistent contact with epicardial tissue. Irrigation prevents spread of thermal injury and cools the tissue surface, which allows for concentrated, focused energy. The system provides audible (cessation of the suction hiss sound when complete device to epicardial tissue vacuum seal is achieved) and visible confirmation of contact (initiation of saline perfusion upon completion of the vacuum seal). The RF generator uses an algorithm based on the electrical resistance of the tissue to modulate power delivery to the lesion to avoid popping and over-desiccation.

Transseptal access and endocardial catheter ablation procedure

Percutaneous access to the left atrium is obtained through a conventional double transseptal puncture. Four venous accesses were obtained: two 8-French sheaths in the right femoral vein, one 11-French sheath, and one 8-French sheath in the left femoral vein. Using standard electrophysiological ablation techniques, endocardial lesions are created to connect the epicardial linear lesions along the pericardial reflections. The technique of PV antrum isolation has been described extensively elsewhere. 10 Briefly, a 3.5-mm irrigated tip catheter (Navistar Thermocool®, Biosense Webster, Diamond Bar, CA) was used to create endocardial lesions. Intracardiac EchoICE was used to identify PV antra, to guide RF delivery, and to look for potential complications. Electroanatomical mapping was performed using the CARTO system. RF energy was set on 35 W up to 40 W and to no more than 41°C catheter tip temperature. When ablating on the posterior wall, if needed, power was reduced to 25 Watts. When targeting PV potentials the potentials located at the pericardial reflections were first targeted. If this resulted in entrance/exit block no further lesions were delivered. Following isolation of the veins, mitral annular lines were only performed if the patient had clinical evidence of left atrial flutter. Roof lines were not performed endocardially. EP testing was performed to confirm PV isolation.

Postoperative management

Antiarrhythmic drug (AAD) therapy was started immediately after the procedure and continued for 8 weeks (dofetilide 65%, flecainide 12%, amiodarone 8%, profafenone 5%, dronedarone 3%, sotalol 1%, no AADs 6%). Anticoagulation was reinitiated postoperatively and continued for at least 5 months.

Follow-up monitoring

Patients were seen by the surgeon for a postoperative visit including an incision inspection at 1 month. Post-procedure follow-up visits at 1, 3, 6, and 12 months were performed by the electrophysiologist. A physical examination, electrocardiogram AAD therapy, and any complications were recorded at each follow-up visit. All patients were monitored for at least 72 h at 6 months and at a year. Patients with PPM/ICD had their implanted devices interrogated at their scheduled follow-up visits.

Two months following, the convergent procedure was defined as a blanking period. During the blanking period, any treated or untreated AF/Aflutterepisodes were not considered treatment failures.

Statistical analysis

Continuous data were described as mean±standard deviation (SD) or as mean (range) for skewed distributions. Baseline and follow-up values are reported as mean±SD for numeric measures, and counts and percentages for categorical measures.


Patient characteristics

A total of 104 patients receiving the convergent procedure were included in the study; all 104 patients underwent the treatment in a single setting.

Table 1 summarizes the baseline clinical characteristics of the study population. The average patient age was 60.9 years, and the majority of the patients were male (77%). Twenty-eight percent of the patients (28/104) were paroxysmal, 30%(31/104) were persistent, and 43% (45/104) were longstanding persistent. The average left atrial size was 4.1 cm based on transthoracic echocardiography measurements.

Table 1: Demographic characteristics


Procedure details

Total procedure time including both the epicardial and endocardial procedure was 212.7 (±22.9) min. Average procedure time (skin to skin) for the epicardial ablation was 84.1 (±21.6) min and epicardial ablation time was 34.7 (±7.7) min. The total time of the endocardial ablation was 128.7 (±28.7) min. Acute procedural success defined as confirmation of entrance and/or exit block was achieved in 100% of the patients. Fluoroscopy time was 19.1 (±7.4) min. PVs were successfully isolated in all patients. Average intensive care unit stay was 1 (±0.4) day and average length of stay was 3 (±0.8) days.

Procedural complications

Safety data for the study is described in Table 2 . There were no major adverse events <7 days post procedure. After 7 days to last follow-up we observed one cerebral vascular accident in a patient with longstanding persistent AF and severe Left Ventricular dysfunction (EF 30%) who presented to the ER with stroke symptoms in sinus rhythm (SR). There was one pericardial effusion (>30 days after procedure) that did require drainage and two pleural effusions. There was one case of PV stenosis that required stenting.

Table 2: Major adverse cardiac event rates (<7 days post procedure) and at last follow-up (>7 days post procedure)


Clinical outcomes

Efficacy outcomes are detailed in Table 3 . At 12 months post procedure 72% (52/72) of patients were in SR off AADs, and 87.5% (63/72) were in SR with or without AADs. At last follow-up (215 days), 73% (76/104) were in sinus rhythm off AAD, and 89% of patients (93/104) were in SR with or without AADs. Any atrial arrhythmia episode >30 s flagged on Holter reports was manually reviewed for accurate identification. Patients were classified as in SR based on this adjudication review. At last follow-up 5% of patients underwent a re-ablation procedure; three patients (paroxysmal AF patients, 1; longstanding persistent AF patients, 2) underwent a repeat catheter ablation procedure for atypical atrial flutter and two patients (paroxysmal AF patients, 1; longstanding persistent AF patients, 1) underwent a repeat catheter ablation for right sided typical atrial flutter.

Table 3: Outcomes of atrial fibrillation success for all atrial fibrillation types


Efficacy outcomes were also looked at based on AF type. Paroxysmal, persistent, and longstanding persistent AF patients at 12 months had success rates of 72%, 84%, and 62% in SR off of AAD and 89%, 88%, and 86% in SR with or without AADs. Table 4 shows outcomes based on AF type at 1 year and at time of last follow-up.

Table 4: Outcomes of atrial fibrillation success based on type of atrial fibrillation



Main finding

To date, this is the largest study assessing the safety and outcomes of the convergent procedure for the treatment of AF. We have demonstrated that the procedure is effective and safe. The percentage of patients free from AF at 1 year was 72% not using AADs and 87.5% with or without AADs.

Even though this was our early experience with a small number of patients, these results make this technique a promising approach to the treatment of AF. In our view, the collaboration between surgeon and electrophysiologist is beneficial to the outcome of the procedure, since it allows electrophysiologists and surgeons to work in a collaborative rather than a competitive fashion, and as a result it has allowed one to bridge the weaknesses of the other. The surgical approach allows for a complete lesion set on the posterior wall and around the PVs but is weak in mapping and is unable to reliably complete a cavotricuspid ablation line. The EP catheter-based approach allows for completion of the PV isolation at the pericardial reflections and meticulous testing of block. Using epicardial and endocardial ablation we can also overcome the challenges seen in achieving transmurality of ablation lesions.

Effect on AF type

In our patients we elected to treat paroxysmal, persistent, and longstanding persistent AF. Although PV isolation is now established as the cornerstone for ablation of paroxysmal AF, it is not effective in all patients. Since there is a continuum of arrhythmia burden in the paroxysmal AF population, we elected to choose patients who clinically or by pacemaker interrogation had a high pre-ablation AF burden since we felt additional substrate modification and complete posterior wall isolation would be required to improve outcomes.

In our experience, paroxysmal AF patients did not have higher success rates than persistent AF patients off of AADs. This may have been reflected by a higher incidence of both post ablation right atrial and left atrial flutter seen in the paroxysmal group. Later in our experience we included cavotricuspid lines empirically in all paroxysmal patients due to an increased finding of failures due to typical right atrial flutter as well as the common finding of induction of right atrial flutter during burst pacing after completion of the left sided lesions. Post ablation left atrial flutter was not a common finding, so we did not feel the need to perform a mitral isthmus ablation on all patients unless they had a documented left atrial flutter. Further study of the paroxysmal patient population will need to be performed to elucidate this issue.

Procedural safety

Single center experience previously reported using the convergent procedure demonstrated 87% of persistent and longstanding persistent patients had less than 3% AF burden verified by implantable loop recorders at 24 months. This success rate unfortunately carried a major complication rate of 10%, including mortality secondary to atrial esophageal fistula and CVA, pericardial effusion, and excessive bleeding. 11 Based on the early experience learned from this study on how to mitigate the risks of the procedure (esophageal temperature monitoring for both, epi- and endo procedure, use of saline for cooling during epicardial ablation, minimal posterior endocardial ablation, prescribing Proton Pump Inhibitor prophylactically), we were able to decrease the complication rate. In our series, we did not experience any atrial esophageal fistulas, but did report one CVA, one pericardial effusion without tamponade, and one case of PV stenosis. Further experience and development of a standardized lesion set, and continued vigilance to improve safety will hopefully help us work toward achieving a higher success rate while keeping complications equivalent to that seen with stand-alone percutaneous catheter ablation.

Epicardial connections and hybrid ablation

Despite the adoption of catheter ablation for treatment of patients with AF, the success rate needs to improve. It has been reported that as much as 20% of electrical impulses arising from the PVs might be propagated through pathways other than the venoatrial continuity at the veins ostium, such as epicardial connections with a nearby PV or epicardial venoatrial connections at a distance from the ostium. 12 These connections may be responsible for some recurrence seen with catheter-based endocardial ablation and offer a clear rationale for the epicardial approach. Electrophysiologists and surgeons have shown that they can work together to improve the safety and efficacy of hybrid ablations, and we are learning that there are advantages of a combined endocardial and epicardial approach. 13

Despite some early promise, some would argue that, given the invasiveness of a surgical procedure, surgery should be limited to failed endocardial cases or for treatment of longstanding persistent AF. Our interpretation of the published literature shows that catheter ablation success is still in need of improvement and the invasiveness of a hybrid procedure has not resulted in an unacceptable complication rate. Furthermore, the perceived logistical problems of a hybrid strategy 14 were overcome at our institution by performing the ablation in the EP laboratory with a total procedure time of less than four hours.

Study limitations

This was a retrospective assessment of the safety and effectiveness of the convergent procedure, although all the data were collected and entered in a prospective database. Quality of life surveys were not administered to evaluate the impact of the convergent procedure on patients' improved activity levels. A control group of only endocardial catheter ablation procedures completed during the same period of time was not evaluated. The study was limited to patients with symptomatic AF.


The convergent procedure, a hybrid approach for AF, is a safe and successful procedure with a procedure success rate of 72% at 1 year without the use of AADs. The clinical implications of our findings are relevant in helping cardiologists, electrophysiologists, and surgeons determine the best procedure to choose when recommending AF ablation for their patients.


  1. Prasad SM, Maniar HS, Camillo CJ, et al. The Cox maze III procedure for atrial fibrillation: long-term efficacy in patients undergoing lone versus concomitant procedures. J Thorac Cardiovasc Surg 2003; 126:1822–1828.
  2. Krul SP, Driessen AH, Zwinderman AH, et al. Navigating the mini-maze: systematic review of the first results and progress of minimally-invasive surgery in the treatment of atrial fibrillation. Int J Cardiol 2013; 166:132–140.
  3. Han FT, Kasirajan V, Kowalski M, et al. Results of a minimally invasive surgical pulmonary vein isolation and ganglionic plexi ablation for atrial fibrillation: single-center experience with 12-month follow-up. Circ Arrhythm Electrophysiol 2009; 2:370–377.
  4. Bhargava M, Di Biase L, Mohanty P, et al. Impact of type of atrial fibrillation and repeat catheter ablation on long-term freedom from atrial fibrillation: results from a multicenter study. Heart Rhythm 2009; 6:1403–1412.
  5. Claude S Elayi, Atul Verma, Luigi Di Biase, et al. Ablation for longstanding permanent atrial fibrillation: results from a randomized study comparing three different strategies. Heart Rhythm 2008; 5:1658–1664.
  6. Hugh Calkins, Matthew R Reynolds, Peter Spector, et al. Treatment of atrial fibrillation with antiarrhythmic drugs or radiofrequency ablation. circulation: arrhythmia and electrophysiology. Circ Arrhythm Electrophysiol 2009; 2:349–361.
  7. Kiser AC, Landers M, Horton R, Hume A, Natale A, Gersak B. The Convergent procedure: a multidisciplinary atrial fibrillation treatment. Heart Surg Forum 2010; 13:E317–21.
  8. Calkins H, Brugada J, Packer DL, et al. HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for personnel, policy, procedures and follow-up. Heart Rhythm 2007; 4:1–46.
  9. Kiser AC. Paracardioscopy provides endoscopic visualization of the heart. Innovations 2009; 4:233–235.
  10. Kanj M, Wazni O, Natale A. Pulmonary vein antrum isolation. Heart Rhythm 2007; 4:S73–9.
  11. Gersak B, Pernat A, Robic B, Sinkovec M. Low rate of atrial fibrillation recurrence verified by implantable loop recorder monitoring following a convergent epicardial and endocardial ablation of atrial fibrillation. J Cardiovasc Electrophysiol 2012; 23:1059–1066.
  12. Pérez-Castellano N, Villacastín J, Salinas J, et al. J Cardiovasc Electrophysiol 2011; 22:149–159.
  13. Pison L, La Meir M, van Opstal J, Blaauw Y, Maessen J, Crijns HJ. Hybrid thoracoscopic surgical and transvenous catheter ablation of atrial fibrillation. J Am Coll Cardiol 2012; 60:54–61.
  14. Calkins H. Hybrid thorascopic and transvenous catheter ablation of atrial fibrillation. J Am Coll Cardiol 2012; 60:54–61.