Disclosure : Dr. Gillinov is a consultant to Edwards Lifesciences, LLC, St. Jude Medical Inc. and Medtronic, Inc. The Cleveland Clinic has an indirect equity interest in AtriCure, Inc. through its interest in a private fund that has an equity interest in AtriCure, Inc. This work was supported by the Atrial Fibrillation Innovation Center. The Atrial Fibrillation Innovation Center is a Wright Center of Innovation funded by a grant from the State of Ohio.

Corresponding Author : Marc A. Gillinov, M.D., Department of Thoracic and Cardiovascular Surgery, The Cleveland Clinic Foundation/F24, 9500 Euclid Avenue, Cleveland, OH 44195. Tel: 216-445-8841,Fax: 216-444-0777, Email: gillinom@ccf.org

Submitted : 2008-03-05Accepted : 2008-05-11.

Copyright: Copyright belongs to Dr. Marc A Gillinov under Open Access License details available online at http://creativecommons.org

doi : 10.4022/jafib.v1i1.408

Abstract

Atrial fibrillation (AF) is now commonly treated at the time of valvular heart surgery or coronary artery bypass grafting. Surgical ablation of AF, which is predicated upon the Maze procedure, includes creation of lines of conduction block and excision of the left atrial appendage. A full bi-atrial lesion set is associated with success in 80% to 95% of patients and virtually eliminates the risk of late stroke. A complex but safe operation, the classic cut-and-sew Maze procedure has been applied by relatively few surgeons. However, recent advances in understanding of the pathogenesis of AF and development of new ablation technologies enable surgeons to perform pulmonary vein isolation, create linear left and right atrial lesions, and remove the left atrial appendage rapidly and safely. Lesions are created under direct vision, minimizing the risk of damage to the pulmonary veins and adjacent mediastinal structures. Recently developed instrumentation now enables thoracoscopic and keyhole approaches, facilitating extension of epicardial AF ablation and excision of the left atrial appendage to patients with isolated AF and no other indication for cardiac surgery. In addition, novel devices designed specifically for minimally invasive epicardial exclusion of the left atrial appendage will broaden the range of treatment options for patients with AF, possibly eliminating the need for anticoagulation in selected patients.

Background

Atrial fibrillation (AF) is common in patients presenting for management of valvular heart disease or coronary artery disease. Developed by Dr. James Cox, the classic Maze procedure is the predicate operation for ablation of AF; long-term data suggest that the Maze procedure eliminates AF in more than 90% of patients (^1-4).  In spite of these unmatched results, the complexity and time associated with the Maze procedure have prevented widespread application by surgeons. 
Recently, however, there has been increased interest in surgical ablation of AF, fueled by technological advances and demonstration that the pulmonary veins and left atrium are the drivers of AF in most patients (⁵).  New ablation technologies enable surgeons to perform pulmonary vein isolation, create strategically-placed, linear left and right atrial lesions, and excise the left atrial appendage rapidly and safely (^6-12).  Alternate energy sources used to create lines of conduction block and replace surgical incisions of the Maze procedure include radiofrequency, microwave, ultrasound, cryothermy, and laser (^6-12).  Applied primarily in patients with valvular heart disease, procedures using these ablation technologies add 15 to 20 minutes to operative time and appear to cure AF in 70% to 85% of patients.  In addition, recent adaptation of these ablation technologies for thoracoscopic and keyhole approaches now enables minimally invasive surgery in selected patients having stand-alone ablation.
In this report, we will 1) Review surgical technique and results of the Maze procedure, 2) Describe new technologies and approaches for surgical AF ablation and,  3) Discuss the development of technology for minimally invasive epicardial ablation and ligation of the left atrial appendage.

The Maze Procedure

            The Cox-Maze III operation, or Maze procedure, is the gold standard for surgical treatment of AF.  In fact, it is the most effective curative therapy for AF yet devised, and it sets the standard for new surgical approaches to AF (^1-4,13,14).  Cox and colleagues designed the procedure based on early experimental and clinical evidence concerning the pathophysiology of AF (^1-4).  To improve results and simplify the operation, they modified the procedure twice, culminating in the Cox-Maze III.
            The Maze procedure includes a complex, bi-atrial lesion set that is applied to all patients having the operation; therefore, it does not include an ablation strategy that is specifically tailored to individual patients (^15-19) In the Maze procedure, right and left atrial incisions and cryolesions are constructed to interrupt the multiple, disorganized reentrant circuits that characterize AF (Figure 1).  This is termed “cut-and-sew” technique.  These lesions direct the sinus impulse from the sinoatrial node to the atrioventricular node along a specified route.  Multiple “blind alleys” off this main conduction pathway (the Maze analogy) facilitate coordinated electrical activation of the atrial myocardium.  Key components of the Maze procedure include isolation of the pulmonary veins and excision of the left atrial appendage.  These features are maintained in most of the newer operations designed to ablate AF.

Figure 1 : Left atrial lesion set of the Cox-Maze III procedure. Schematic illustration of the posterior left atrium. White ovals represent mitral valve, and sets of 4 black ovals represent pulmonary veins. Dashed lines indicate surgical incisions. The pulmonary veins are encircled by a surgical incision, and there is a connecting incision to the mitral valve annulus. The left atrial appendage is excised, and this incision is connected to the pulmonary vein encircling incision.

            Although the Maze procedure can be completed minimally invasively through a small chest wall incision, the operation entails cardiopulmonary bypass and cardiac arrest.  In experienced centers, the Maze procedure requires 45 to 60 minutes of cardiopulmonary bypass and cardiac arrest (^4,13,14).  The operation may be performed alone or in conjunction with other cardiac surgical procedures, such as mitral valve repair or coronary artery bypass grafting.
            Cox and colleagues (⁴) have reported the largest series of patients undergoing the Maze procedure.  Among 346 patients, operative mortality was 2%.  Reported AF cure rate was 99%, and only 2% required long-term postoperative anti-arrhythmic medication.  Of note, results were tabulated by calculating freedom from “symptomatic AF”; long-term rhythm monitoring was not employed.  Therefore, these results likely over-state the effectiveness of the cut-and-sew Maze procedure. Successful ablation of AF was unaffected by presence of mitral valve disease, left atrial size, and type of AF (paroxysmal, persistent or permanent).  Temporary postoperative AF, attributed to a shortened atrial refractory period, was common and did not diminish long-term results.  Fifteen percent of patients required new pacemakers after surgery, and this was generally necessary in patients with underlying sinus node dysfunction.  In spite of multiple right and left atrial incisions, right atrial transport function was demonstrated in 98% and left atrial transport in 93%.  However, there is some controversy concerning the extent to which atrial mechanical function returns after surgical ablation.  Perhaps most importantly, in addition to restoring sinus rhythm the Maze procedure virtually eliminated the risk of stroke or other thromboembolism (^20,21).
Other centers have documented excellent results with the Maze procedure, with restoration of sinus rhythm in 75% to 95% of patients, low risk of late stroke, and very low operative morbidity and mortality (^13,14,21).  In recent series, the need for a new pacemaker has decreased to 5% to 10% (²²).  These results confirm the safety of the Maze procedure, its efficacy at restoring sinus rhythm, and its prevention of late strokes.  In spite of these findings, the Maze procedure has been relatively underutilized.  Today, few patients are referred for a surgical Maze procedure for stand-alone AF ablation, and, even in patients requiring cardiac surgery for other reasons, surgeons are reluctant to add a Maze procedure.  The perceived surgical complexity and magnitude of the operation account for these practices. 

New Approaches for Surgical AF Ablation

           The development of new surgical approaches to AF has been predicated upon 2 factors: 1) Recognition that the pulmonary veins and left atrium are critical to the initiation and maintenance of AF and 2) Development of ablation technologies that use alternate energy sources to facilitate rapid and safe creation of lines of conduction block under direct vision.

Lesion Sets

           While the Maze procedure was designed to interrupt the multiple macro- reentrant circuits that characterize AF, new approaches are more precisely anatomically focused.  There is general agreement that AF requires a substrate and a trigger, and that these substrates and triggers are usually located in the pulmonary veins and left atrium (^23,24).  Haissaguerre and colleagues (⁵) demonstrated that paroxysmal AF originates from ectopic beats in the pulmonary veins in 94% of cases.  In addition, autonomic innervation of these regions may contribute to the pathogenesis of AF (²⁵).  Catheter ablation of the posterior left atrium, including the antra surrounding the pulmonary veins, has proven effective at ablating both paroxysmal and permanent AF (^25,26).  These data suggest that modification of the left atrial substrate, in combination with pulmonary vein isolation, is an effective therapy for all forms of AF.             Surgeons have used alternate energy sources to create a variety of left atrial lesions sets, ranging from wide pulmonary vein isolation with excision of the left atrial appendage to a lesion set that resembles that of the Maze procedure (Figure 2).  There is growing consensus that patients with persistent or long-standing persistent AF should receive a lesion set that incorporates more than simple pulmonary vein isolation (^27,28). In such patients, the lesion from the right pulmonary veins to the mitral annulus may be particularly important in prevention of post-procedure left atrial flutter (²⁹).  The addition of right atrial lesions appears to increase freedom from recurrent AF and atrial flutter, but their importance is controversial (³⁰).  However, creation of right atrial lesions is simple and safe, and we currently favor their incorporation at the time of surgical ablation.   

Figure 2 : Left atrial lesion sets created with alternate energy sources. Schematic illustration of the posterior left atrium as above. These require approximately 10 to 20 minutes of operative time. Dashed lines indicate lesions created with heat-based energy sources or cryothermy. (A) Bilateral pulmonary vein isolation with excision of the left atrial appendage. (B) Maze-like lesion set created with bipolar radiofrequency and cryothermy (for lesion to the mitral annulus).

Ablation Technology

Based upon these advances in understanding of the pathophysiology of AF, a variety of new ablation tools have been developed to facilitate surgical ablation of AF.  These probes and catheters rely on alternate energy sources to create long, continuous, linear lesions that block conduction.  Energy sources that have been used clinically include radiofrequency, laser, ultrasound, microwave, and cryothermy (^10,11).  Radiofrequency, laser, ultrasound and microwave are heat-based energy sources that create lines of conduction block through thermal injury.  Each of these modalities can be employed for ablation during concomitant open heart surgery or, alternatively, for stand-alone, minimally invasive epicardial ablation. 
           Because the surgeon has the advantages of 1) direct visualization of cardiac structures and 2) catheters that facilitate rapid creation of transmural lesions, there is great interest in ablating AF in patients presenting for other cardiac surgical procedures.  Completion of left atrial lesion sets requires only 10 to 20 minutes.  This amount of time contrasts with the 1 hour required to perform the traditional cut-and-sew Maze procedure.  In addition, because incisions are replaced by heat- or cryo-based lesions, the risk of bleeding is virtually eliminated when alternate energy sources are employed.
           Although approaches vary somewhat, results are similar with a variety of energy sources.  AF is ablated in 70% to 80% of patients having concomitant heart surgery (^10,11).  Thus far, most treated patients have had organic heart disease and have received a mitral valve procedure in addition to AF ablation.  For them, results with alternate energy sources fall just short of those reported for the classic cut-and-sew Maze procedure (^31,32). 
           After surgical ablation, perioperative AF is common, occurring in approximately 50% of patients.  Although 30% to 40% of patients leave the hospital in AF, many return to sinus rhythm over the ensuing 3 months.  Thus, discharge in AF is not an indication of procedure failure (³³).  Given the high incidence of perioperative AF, a strategy that includes 3 months of routine postoperative anti-arrhythmic therapy and anticoagulation in all patients is recommended.  Because heart rhythm varies in the first 3 months after surgery, we recommend aggressive attempts to restore sinus rhythm during this time frame when patients develop AF or atrial flutter.  Heart rhythm generally stabilizes by 3 to 6 months after surgery.  Factors that influence procedure success include larger left atrial size, longer duration of AF, and choice of lesion set in permanent AF (^16,34,35). 

Future directions: Minimally invasive stand-alone ablation

 Open surgical treatment of AF has a long track record of success.  The surgeon has the advantage of direct visualization of the left atrium and pulmonary veins, either from the epicardial or endocardial surface of the heart, and this factor, coupled with new ablation technology, enables rapid and safe ablation.  Because the surgeon can see cardiac structures, ablation lines can be placed safely on the left atrial cuffs adjacent to the pulmonary vein orifices, thereby avoiding the dreaded complication of pulmonary vein stenosis.  Epicardial ablation eliminates the risk of esophageal injury; however, there is uncertainty concerning the ability of unipolar energy sources to create continuous, transmural lesions from the epicardium of the beating heart.  With surgical approaches, the left atrial appendage is excised, and this is likely important in decreasing the risk of late stroke.  Finally, the simplicity of these techniques makes them generally applicable; all cardiac surgeons can now ablate AF.  In current clinical practice, almost all patients with AF who present for cardiac surgery should have both AF ablation and the intended cardiac procedure.
   Using this experience as a spring-board, surgeons have developed epicardially-based, minimally invasive and thoracoscopic approaches to offer stand-alone AF ablation. While these procedures are currently in early stages of development and application, results are promising. Pulmonary vein isolation and excision of the left atrial appendage can be performed using a minimally invasive “keyhole approach” or thoracoscopically; neither approach requires cardiopulmonary bypass (^36-42).  These minimally invasive procedures enable wide, circumferential pulmonary vein isolation, either with a single “box” lesion or separate oval-shaped ablations on the right and left.  Connecting lesions across the dome of the left atrium and to the mitral annulus can now be created using specially-designed unipolar energy sources (⁴²). Procedure times are generally 2 to 4 hours, and median length of hospital stay is 3 days.  
Early results, obtained primarily in patients with paroxysmal AF, demonstrate 80% to 90% freedom from AF 6 months after ablation.  With continued experience and further advances in instrumentation, procedure time is expected to fall to less than 2 hours, and hospital length of stay to decline to a single day.  Application of these minimally invasive, thoracoscopic and keyhole procedures will offer the potential for cure of AF, with improved quality of life and freedom from anticoagulation and anti-arrhythmic medications, to large numbers of patients. 

Future directions: the left atrial appendage

Currently there is great interest in development and assessment of endocardial and epicardial procedures for exclusion of the left atrial appendage (^43,44). It is widely believed that formation and embolism of left atrial appendage thrombi are responsible for the increased risk of stroke in AF patients (^43,44).  In AF patients, warfarin inhibits formation of atrial appendage thrombi and reduces cardioembolic strokes, while aspirin prevents smaller, noncardioembolic strokes.  Based upon data from the Stroke Prevention in Atrial Fibrillation (SPAF) trials, one-third of AF patients are at high risk for stroke and should be treated with warfarin (^4,45,46).  In spite of these observations and recommendations, warfarin is under-prescribed in AF patients.  Furthermore, many patients cannot or will not take warfarin.  Therefore, interventional therapies that specifically address the left atrial appendage in AF patients are being investigated as potential alternatives to warfarin therapy.
           Excision or exclusion of the left atrial appendage is currently performed during surgical ablation of AF and is recommended in ACC/AHA guidelines for patients undergoing mitral valve surgery.  However, standard surgical exclusion by suture closure is incomplete in 30% of cases, and stapled closure or excision has been associated with bleeding complications (⁴⁷).  Thus, there is a need for new surgical approaches to the left atrial appendage.  Several devices for epicardial exclusion of the left atrial appendage are under development (Figure 3).  Pre-clinical studies suggest that device-based epicardial exclusion of the left atrial appendage is rapid and safe and has the advantage of avoiding placement of a foreign body in the fibrillating atrium (⁵⁰).  Early clinical results with this epicardially-based left atrial appendage clip are promising.  This and other new technologies for minimally invasive epicardial ablation of the left atrial appendage will soon be available for widespread clinical use. Clinical trials are being designed to test the hypothesis that epicardial, device-based exclusion of the left atrial appendage will reduce the risks of stroke and other thromboembolism in patients with AF.  Should this hypothesis prove correct, physicians will be able to offer a new strategy for primary or secondary stroke prevention in AF patients. 

Figure 3 : Left atrial appendage clip. A cloth-covered, nitinol and titanium clip has been applied to the epicardial aspect of a canine left atrial appendage.

Acknowledgements

I thank Laura Roberts for expert editorial assistance.

References

1. Cox JL, Schuessler RB, D’Agostino HJ Jr, Stone CM, Chang BC, Cain ME, Corr PB, Boineau JP.  The surgical treatment of atrial fibrillation.  III.  Development of a definitive surgical procedure.  J Thorac and Cardiovasc Surg.  1991;101:569-583.
2. Cox JL, Schuessler RB, Boineau JP.  The development of the Maze procedure for the treatment of atrial fibrillation.  Semin Thorac Cardiovasc Surg.  2000;12:2-14.
3. Cox JL, Boineau JP, Schuessler RB, Kater KM, Ferguson TB, Jr, Cain ME, Lindsay BD, Smith JM, Corr PB, Hogue CB, Lappas DG.  Electrophysiologic basis, surgical development, and clinical results of the Maze procedure for atrial flutter and atrial fibrillation.  Adv Card Surg.  1995;6:1-67.
4. Cox JL, Ad N, Palazzo T, Fitzpatrick S, Suyderhoud JP, DeGroot KW, Pirovic EA, Lou HC, Duvall WZ, Kim YD.  Current status of the Maze procedure for the treatment of atrial fibrillation.  Semin Thorac Cardiovasc Surg.   2000;12:15-19.
5. Haissaguerre M, Jais P, Shah DC, Takahashi A, Hocini M, Quiniou G, Garrigue S, LeMouroux A, LeMetayer P, Clementy J.  Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins.  N Eng J Med.  1998:339:659-666.
6. Gillinov AM, Blackstone EH, McCarthy PM.  Atrial fibrillation: Current surgical options and their assessment.  Ann Thorac Surg.  2002;74:2210-2217.
7. Gillinov AM, McCarthy PM.  Atricure bipolar radiofrequency clamp for intraoperative ablation of atrial fibrillation.  Ann Thorac Surg.  2002;74:2165-2168.
8. Gillinov AM, McCarthy PM, Marrouche N, Natale A.  Contemporary surgical treatment for atrial fibrillation.  PACE.  2003;26:1-4.
9. Gillinov AM, McCarthy PM.  Advances in the surgical treatment of atrial fibrillation.  Cardiol Clin. 2004;147-57.
10. Gillinov AM, Saltman AE.  Ablation of atrial fibrillation with concomitant cardiac surgery.  Semin Thorac Cardiovasc Surg.  2007;19:25-32.
11. Comas GM, Imren Y, Williams MR.  An overview of energy sources in clinical use for the ablation of atrial fibrillation.  Semin Thorac Cardiovasc Surg. 2007;19:16-24.
12. Groh MA, Binns OA, Burton HG III, Ely SW, Johnson AM.  Ultrasonic cardiac ablation for atrial fibrillation during concomitant cardiac surgery: Long-tern clinical outcomes.  Ann Thorac Surg.  2007;84:1978-83.
13. McCarthy PM, Gillinov AM, Castle L, Chung M, Cosgrove DM 3^rd.  The Cox-maze procedure: the Cleveland Clinic experience.  Semin Thorac Cardiovasc Surg.  2000;12:25-29.
14. Schaff HV, Dearani JA, Daly RC.  Cox-maze procedure for atrial fibrillation: Mayo Clinic experience.  Semin Thorac Cardiovasc Surg. 2002;12:30-37.
15. Gillinov AM.  Advances in the surgical treatment of atrial fibrillation.  Stroke.  2007;38:618-23.
16. Gillinov AM.  Choice of surgical lesion set: Answers from data.  Ann Thorac Surg. 2007;84:1786-92.
17. Van Wagoner DR.  Recent insights into the pathophysiology of atrial fibrillation.  Semin Thorac Cardiovasc Surg.  2007;19:9-15.
18. Morady F.  Patient-specific ablation strategy for atrial fibrillation: Promises and difficulties.  Heart Rhythm Society.  2007;4:1094-96.
19. Nitta T.  Surgery for atrial fibrillation: A worldwide review.  Semin Thorac Cardiovasc Surg.  2007;19:3-8.
20. Ad N, Cox JL.  Stroke prevention as an indication for the Maze procedure in the treatment of atrial fibrillation.  Semin Thorac Cardiovasc Surg.  2001;12:56-62.
21. Wong JWW.  Mak KH.  Impact of maze and concomitant mitral valve surgery on clinical outcomes.  Ann Thorac Surg.  2006;82:1938-47.
22. Gillinov AM, Sirak J, Blackstone EH, McCarthy Pm, Rajeswaren J, Pettersson G, Sabik J, III, Svensson LG, Navia JL, Cosgrove DM, Marrouche N, Natale A.  The Cox maze procedure in mitral valve disease: Predictors of recurrent atrial fibrillation.  J Thorac Cardiovasc Surg. 2005;130:1653-60.
23. Falk RH.  Atrial fibrillation.  N Engl J Med.  2001;344:1067.
24. Nattel S.  New ideas about atrial fibrillation 50 years on.  Nature.  2002;415:219-226.
25. Pappone C, Santinelli V, Manguso F, Vicedomini G, Gugliotta F, Augello G, Mazzone P, Tortoriello V, Landoni G, Zangrillo A, Lang C, Tomita T, Mesas C, Mastella E, Alfieri O.  Pulmonary vein denervation enhances long-term benefit after circumferential ablation for paroxysmal atrial fibrillation.  Circulation.  2004;109:327-34.
26. Oral H, Pappone C, Chugh A, Good E, Bogun F, Pelosi F, Bates ER, Lehmann MH, Vicedomini G, Auguello G, Agricola E, Sala S, Santinelli V, Morady F.  Circumferential pulmonary-vein ablation for chronic atrial fibrillation.  N Engl J Med.  2006;354:934-41.
27. Gillinov AM, Bhavani S, Blackstone EH, et al.  Surgery for permanent atrial fibrillation: Impact of patent factors and lesion set.  Ann Thorac Surg. 2006;85:502-14.
28. Wisser W, Seebacher G, Fleck TAigner C, Khaze C, Stix G, Hutschala D, Wolner E.  Permanent chronic atrial fibrillation: Is pulmonary vein isolation alone enough?  Ann Thorac Surg.  2007;84:1151-57.
29. Cox JL, Ad N.  The importance of cryoablation of the coronary sinus during the maze procedure.  Semin Thorac Cardiovasc Surg.  2000;12:20-24.
30. Barnett SD, Niv A.  Surgical ablation as treatment for the elimination of atrial fibrillation: A meta-analysis.  J Thorac Cardiovasc Surg.  2006;131:1029-35.
31. Stulak JM, Dearani JA, Sundt TM III, Daly RC, McGregor CGA, Zehr KJ, Schaff HV.  Superiority of cut-and-sew techniques for the Cox maze procedure: Comparison with radiofrequency ablation.  J Thorac Cardiovasc Surg.  2007;133:1022-27.
32. Lall Sc, Melby SJ, Voeller RK, Zierer A, Bailey MS, Guthrie TJ, Moon MR, Moazami N, Lawton JS, Damiano RJ Jr.  The effect of ablation technology on surgical outcomes after the Cox-maze procedure: A propensity analysis.  J Thorac Cardiovasc Surg.  2007;133:389-96.
33. Ishii Y, Gleva MJ, Gamache MC, Schuessler RB, Boineau JP, Bailey MS, Damiano RJ, Jr.  Atrial tachyarrhythmias after the maze procedure: Incidence and prognosis.  Circulation.  2004;110(suppl II):II-164-II-168.
34. Gillinov AM, McCarthy PM, Blackstone EH, Rajeswaren J, Pettersson G, Sabik JF, III, Svensson LG, Cosgrove DM, Hill KM, Gonzalez-Stawinski GV, Marrouche N, Natale A.  Surgical ablation of atrial fibrillation with bipolar radiofrequency as the primary modality.  J Thorac Cardiovasc Surg 2005;129:1322-9.
35. Chen MC, Chang JP, Guo GB, Chang HW.  Atrial size reduction as a predictor of the success of radiofrequency maze procedure for chronic atrial fibrillation in patients undergoing concomitant valvular surgery.  J Cardiovasc Electrophysiol.   2001;12:867-874.
36. Wolf RK, Schneeberger EW, Osterday R, Miller D, Merrill W, Flege JB, Jr, Gillinov AM.  Video-assisted bilateral pulmonary vein isolation and left atrial appendage exclusion for atrial fibrillation.  J Thorac Cardiovasc Surg. 2005;13:797-802.
37. Pruitt JC, Lazzara RR, Dworkin GH, Badhwar V, Kuma C, Ebra G.  Totally endoscopic ablation of lone atrial fibrillation: Initial clinical experience.  J Ann Thorac Surg.  2006;81:1325-31.
38. Wudel JH.  Chaudhuri P, Hiller JJ.  Video-assisted epicardial ablation and left atrial appendage exclusion for atrial fibrillation: Extended follow-up.  Ann Thorac Surg.  2008;85:34-8.
39. McClelland JH, Duke D, Reddy R.  Preliminary results of a limited thoracotomy: New approach to treat atrial fibrillation.  J Cardiovasc Electrophysiol.  2007;18:1-7.
40. Puskas J.  Lin E, Bailey D, Guyton R.  Thoracoscopic radiofrequency pulmonary vein isolation and atrial appendage occlusion.  Ann Thorac Surg.  2007;83:1870-2.
41. Saltman AE.  Minimally invasive surgery for atrial fibrillation.  Semin Thorac Cardiovasc Surg.  2007;19:33-38.
42. Edgerton JR, Jackman WM, Mack MJ.  Minimally invasive pulmonary vein isolation and partial autonomic denervation for surgical treatment of atrial fibrillation.  J Interv Card Electrophysiol.  2007;20:1-5.
43. Gillinov AM, Pettersson G, Cosgrove DM, 3^rd.  Stapled excision of the left atrial appendage.  J Thorac Cardiovasc Surg.  2005;129:679-80.
44. Johnson WD, Ganjoo AK, Stone CD, Srivyas RC, Howard M.  The left atrial appendage: Our most lethal human attachment!  Surgical implications.  Eur J Cardiothorac Surg.  2000;17:718-22.
45. Hart RG, Halperin JL, Pearce LA, Anderson DC, Kronmal RA, McBride R, Nasco E, Sherman DG, Talbert RL, Marler JR.  Lessons from the stroke prevention in atrial fibrillation trials.  Ann Intern Med.  2003;138:831-838.
46. Savelieva I, Camm J.  Update on atrial fibrillation: Part I.  Clin Cardiol.  2008;31:55-62.
47. Katz ES, Tsiamtsiouris T, Applebaum RM, Schwartzbard A, Tunik PA, Kronzon I.  Surgical left atrial appendage ligation is frequently incomplete: a Transesophageal echocardiograhic study.  J Am Coll Cardiol. 2000;36:468-71.
48. Kamohara K, Fukamachi K, Ootaki Y, Akiyam M, Cingoz F, Ootaki C, Vince G, Popovic ZB, Kopak MW Jr, Dessoffy R, Liu J, Gillinov AM.  Evaluation of a novel device for left atrial appendage exclusion: The second-generation atrial exclusion device.  J Thorac Cardiovasc Surg.  2006;132:340-6.