For rheumatic mitral stenosis, percutaneous balloon

CARDIOVASCULAR Consequence of Atrial Fibrillation and the Risk of Embolism After Percutaneous Mitral Commissurotomy: The Necessity of the Maze Procedure Hiroyuki Nakajima, MD, Junjiro Kobayashi, MD, Ko Bando, MD, Yoshio Yasumura, MD, Satoshi Nakatani, MD, Kohji Kimura, MD, Kazuo Niwaya, MD, Osamu Tagusari, MD, and Soichiro Kitamura, MD Department of Cardiovascular Surgery, National Cardiovascular Center, Osaka, Japan Background. We sought to investigate the incidence and risk of embolism after percutaneous balloon mitral commissurotomy (PMC) and delineated the necessity of the maze procedure in patients with atrial fibrillation (AF). Methods. Clinical records of 326 patients, who underwent PMC between April 1988 and March 2003, and 162 patients, who underwent mitral valve surgery (MVS) combined with the maze procedure (MVS/maze) in the same period, were reviewed. We examined the consequence of cardiac rhythm and the incidence and risk of embolism after PMC. Next our current criteria for indication of the maze procedure, which included duration of AF less than 10 years, voltage of f wave on electrocardiography greater than 0.1 mv, cardiothoracic ratio less than 70%, and left atrial dimension less than 70 mm, were applied to all patients. These criteria were satisfied by 190 patients of PMC (PMC group) and 114 patients of MVS/maze (MVS/maze group) and the early and late results of the two groups were compared. Results. Thirty-eight patients suffered from embolic complications after PMC. Of these 33 patients experienced AF before intervention. The actuarial embolismfree rate at 10 years was 81.9% in patients with AF before PMC and 92.9% in patients with sinus rhythm before PMC (p 0.01). Univariate predictors of embolism after PMC included previous embolic history (p 0.01), AF (p 0.01), pressure gradient (p 0.01), age (p 0.03), and mitral valve area (p 0.04). Multivariate analysis identified AF as the independent predictor of embolism in the late follow-up period (p 0.03). In a comparative study of the selected patients the actuarial AF-free rates after MVS/maze were 89.9% at 1 year and 85.7% at 5 years and were significantly higher than those of 17.3% at 1 year and 4.2% at 5 years after PMC (p < 0.0001). The actuarial embolism-free rate in the MVS/maze group was 98.0% at 5 years and was significantly higher than that of 84.8% at 5 years and 82.4% at 10 years in the PMC group (p 0.01). Conclusions. PMC alone is not sufficient with regard to sinus rhythm recovery and prevention of embolism. Surgical treatment concomitant with the maze procedure may be beneficial for patients with AF. (Ann Thorac Surg 2004;78:800 6) 2004 by The Society of Thoracic Surgeons For rheumatic mitral stenosis, percutaneous balloon mitral commissurotomy (PMC) is now considered as the first line of therapy for asymptomatic and symptomatic patients [1]. The valve morphology is considered as the most important predictive factor of the immediate and long-term outcomes after PMC. Several prospective studies demonstrated that the efficacy of PMC was comparable with surgical commissurotomy in terms of improvement and its durability of the valve function with less procedural risk [2]. Atrial fibrillation (AF) is a common cardiac arrhythmia in rheumatic valvular disease. Initiation of AF may be closely related to the symptoms caused by a decrease of Accepted for publication April 1, 2004. Address reprint requests to Dr Nakajima, Department of Cardiovascular Surgery, National Cardiovascular Center, 5-7-1 Fujishirodai, Suita, Osaka 565-8565, Japan; e-mail: hnakajim@hsp.ncvc.go.jp. the cardiac output, elevation of the atrial pressure, and increase of the heart rate. AF is also closely associated with thrombus formation in the atrium and systemic arterial embolism. Long-term follow-up studies demonstrated that the development of AF markedly increased the risk of cardiovascular complications. Moreover it is generally accepted that AF concomitant with rheumatic valve disease exhibits a considerably higher relative risk of stroke than AF without valvular disease [3 5]. However the investigation about the effect of PMC on the cardiac rhythm and the incidence and risk of thromboembolic complications after PMC are not conclusive. In open-heart surgery AF and mitral valve disease can be treated simultaneously by performing the maze procedure. The maze procedure is reportedly effective for the prevention of thromboembolic events by elimination of AF [6]. In the last decade the technique of the maze 2004 by The Society of Thoracic Surgeons 0003-4975/04/$30.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2004.04.019

Ann Thorac Surg NAKAJIMA ET AL 2004;78:800 6 PMC VERSUS MVS/MAZE IN MITRAL STENOSIS AND AF Table 1. Characteristics of Patients Undergoing PMC Total SR AF p Value No. of patients 326 84 (26%) 242 (74%) Male 93 (29%) 16 (19%) 77 (32%) 0.02 Age at intervention 53 9 48 9 55 8 0.002 NYHA class 1.98 0.41 1.93 0.43 1.99 0.60 0.04 History of embolism 61 (5%) 4 (5%) 57 (24%) 0.0001 Surgical commissurotomy 10 (3%) 2 (2%) 8 (3%) 0.5 Mitral valve Orifice area (cm 2 ) 1.07 0.27 1.05 0.28 1.08 0.26 0.89 Regurgitation 2 55 (17%) 6 (7%) 49 (20%) 0.003 Mean PG (mm Hg) 7.6 3.5 8.8 3.9 7.2 3.3 0.0001 Tricuspid regurgitation 3 31 (10%) 4 (5%) 27 (11%) 0.06 Dimension of left atrium 49.9 8.0 46.1 7.3 51.3 7.9 0.001 Follow-up period (mean SD) 96 48 109 50 91 47 0.03 801 CARDIOVASCULAR NYHA New York Heart Association; PG pressure gradient; PMC percutaneous balloon mitral commissurotomy; SR sinus rhythm. procedure was modified with the use of ablation devices allowing it to be less invasive [7, 8]. In the present study we assessed the early and late results after PMC to delineate the incidence and predictive factors of embolic complication. Moreover we examined the efficacy of simultaneous treatment for mitral stenosis and AF in terms of thromboembolism prevention by comparing the results of PMC with those of mitral valve surgery combined with the maze procedure (MVS/maze). Material and Methods At the National Cardiovascular Center (Osaka, Japan) between April 1988 and March 2003, 326 patients who underwent PMC to treat severe mitral stenosis and 162 patients who underwent MVS/maze for the simultaneous treatment of mitral stenosis and AF were included in this study. Two-dimensional echocardiography and Doppler study were performed before and after the procedures. The clinical and echocardiographic data of 488 patients that were prospectively collected were reviewed. At first we analyzed the data to determine the incidence and risk of embolic complications after PMC. The patients profiles are summarized in Table 1. The mean age at intervention was 53 9 years. Most candidates of PMC were symptomatic: 232 patients (71.1%) were designated as New York Heart Association (NYHA) functional class II and 23 patients (7.1%) were designated as class III or IV. Upon admission 84 (25.8%) patients exhibited sinus rhythm, whereas 242 patients (74.2%) exhibited AF. PMC Management and Procedure At our institution PMC was initially considered as the treatment for rheumatic mitral stenosis. PMC was indicated mostly for symptomatic patients who did not exhibit moderate mitral regurgitation and extensive calcification. When the thrombus in the left atrium was detected by transesophageal echocardiography before PMC, it was repeatedly assessed after warfarin therapy for several weeks. In the event that the left atrial thrombus was intractable, surgical treatment was chosen. PMC was performed by the transvenous approach. An Inoue balloon was transseptally inserted into the left ventricle using the Brockenbrough technique. During the procedure transesophageal echocardiography was routinely monitored for guidance of the catheter delivery and assessment of the left atrial thrombus, commissural splitting, and changes of mitral valve function. The size of the balloon was determined by the valve orifice area before intervention. Comparative Study of Selected Patients Next we examined the efficacy of simultaneous treatment for mitral stenosis and AF using the comparative study of selected patients. We have established and have been applying the patients selection criteria for the maze procedure since 1998 [9]. The patients who satisfied all the new criteria, which included voltage of f-wave in V1 lead greater than 0.1 mv, cardiothoracic ratio less than 70%, left atrial dimension in systole less than 70 mm, and prior duration of AF less than 10 years, were expected to be highly successful regarding sinus rhythm recovery after the maze procedure. In the present study these criteria were applied to all patients with AF. Onehundred ninety-nine patients (PMC group) who underwent PMC, and 114 patients (MVS/maze group) who underwent MVS/maze were selected. The profiles of the selected patients in both groups are indicated in Table 2. There were significant differences regarding the age at procedure (p 0.0001), duration of AF (p 0.0001), mitral valve area (p 0.01), the presence of moderate mitral regurgitation (p 0.0001), and the dimension of the left atrium (p 0.0001) between the two groups. Surgical Procedure Through a standard left atriotomy cardiopulmonary bypass commenced with ascending aortic cannulation and bicaval drainage approaching the mitral valve via the right-sided left atriotomy in all patients. Open mitral commissurotomy was performed in 14 patients (12.3%)

CARDIOVASCULAR 802 NAKAJIMA ET AL Ann Thorac Surg PMC VERSUS MVS/MAZE IN MITRAL STENOSIS AND AF 2004;78:800 6 Table 2. Characteristics of Selected Patients PMC MVS/maze p Value No. of selected patients 199 114... Age at procedure 54 9 58 9 0.0001 History of embolism 40 (20%) 18 (16%) 0.21 Coronary artery disease 1 (0.5%) 2 (2%) 0.30 NYHA class 1.97 0.39 2.04 0.35 0.30 Duration of AF (yrs) 3.3 3.7 4.7 4.0 0.0001 Amplitude of f wave (mv) 0.18 0.07 0.18 0.10 0.79 Cardiothoracic ratio (%) 55 7 57 7 0.05 Dimension of LA (mm) 51 7 56 7 0.0001 Tricuspid regurgitation 3 22 14 0.70 Mitral valve area (cm 2 ) 1.07 0.26 1.21 0.48 0.01 Mitral regurgitation 2 43 (22%) 60 (53%) 0.0001 Mean PG (mm Hg) 7.3 3.3 7.3 3.4 0.60 History of LA thrombus 3 (2%) 16 (14%) 0.0001 AF atrial fibrillation; LA left atrium; MVS/maze mitral valve surgery combined with the maze procedure; NYHA New York Heart Association; PG pressure gradient; PMC percutaneous balloon mitral commissurotomy. and a prosthetic valve was implanted in 100 patients (87.7%) (mechanical:bioprosthesis 98:2). The diagnosis of rheumatic mitral disease was confirmed by the histologic examination in all patients who underwent valve replacement. The details of the configuration of cryoablation and cut-and-sew lesions in the maze procedure were described previously [7]. Although entire pulmonary venous isolation was achieved by the cut-and-saw technique between May 1992 and August 1998, since September 1998 we have introduced the cryo-maze procedure, which consists of using cryoablation around the pulmonary veins to decrease perioperative complications. Administration of Anticoagulation and Antiarrhythmic Drugs Electrocardiography was continuously monitored after PMC and MVS/maze until the cardiac rhythm became stable. Immediately after the PMC procedure, cardioversion was usually performed in patients with AF. When AF was sustained despite administration of group Ia or Ic antiarrhythmic drugs, cardioversion was performed and done so before discharge. Irrespective of the cardiac rhythm, digoxin or verapamil were used to control heart rate. The antiarrhythmic drugs were withdrawn when the cardiac rhythm stabilized, which typically occurred 3 months after the procedure was performed (PMC or maze). Warfarin was routinely administered to patients who exhibited AF or had mechanical valves. Prothrombin time was usually monitored every month. The target range was between 1.8 and 2.8 of the international normalized ratio. Follow-Up Patients were followed-up with periodical electrocardiography, chest roentgenography, and echocardiography. Data was collected from the clinical records and by Table 3. Early and Late Results of PMC Events SR n 84 AF n 242 p Value Early results ( ); mortality Embolism 0 4 (0) 0.30 Mitral valve surgery 0 1 (0) 0.74 Endocarditis 0 1 (0) 0.74 Late results ( ); mortality Embolism 5 (0) 33 (5) 0.04 stroke 5 30... peripheral embolism 0 3... Bleeding 0 2 (0) 0.55 Mitral valve surgery 65 (0) 196 (2) 0.81 Endocarditis 1 (0) 3 (0) 0.73 Others 0 9 (9)... AF atrial fibrillation; PMC percutaneous balloon mitral commissurotomy; SR sinus rhythm. correspondence with physicians who applied these guidelines to patients after valvular operation, PMC, and MVS/maze. Embolism included stroke that was defined as a neurologic deficit concomitant with ischemic findings on computed tomography and/or magnetic resonance imaging and transient ischemia due to cardiogenic embolism certainly diagnosed by a neurologist, and peripheral thromboembolism. Permanent and transient ischemia related to atherosclerotic disease were not included as endpoints of this study. Follow-up was performed by the outpatient clinic and biannual mail interview. The mean follow-up period was 49 36 months in patients post-mvs/maze and 96 48 months in those post-pmc (p 0.0001). Statistical Analysis The continuous variables are expressed as the mean values SD. The Kaplan Meier method was used to determine the actuarial embolism-free and AF-free rates. Cox regression analysis was used to examine the significance of the clinical and echocardiographic variables for predicting the embolism-free time. The clinical and echocardiographic data of the two groups were compared using Fisher s exact test or Wilcoxon s log-rank sum test. The differences were considered statistically significant at p less than 0.05. Results Early and Late Outcome of PMC Early and late complications post-pmc are listed in Table 3. During the follow-up period 38 patients experienced an embolic complication. Of these, 33 patients exhibited AF and 5 patients exhibited sinus rhythm before intervention. In addition the second embolic attack occurred in 6 patients. The actuarial AF-free rates were 15.0% at 1 year and 3.4% at 5 years in patients with AF before PMC and were 93.5% at 1 year and 75.6% at 5 years in patients with sinus rhythm before PMC (p 0.0001). The actuarial embolism-free rates in patients with sinus rhythm before

Ann Thorac Surg NAKAJIMA ET AL 2004;78:800 6 PMC VERSUS MVS/MAZE IN MITRAL STENOSIS AND AF Table 5. Early and Late Results in Selected Patients PMC n 199 MVS/maze n 114 803 p Value CARDIOVASCULAR Fig 1. The actuarial embolism-free rate after percutaneous balloon mitral commissurotomy. AF atrial fibrillation; SR sinus rhythm. PMC were 97.3% at 5 years and 92.9% at 10 years and in patients with AF before PMC the rates were 85.7% at 5 years and 81.9% at 10 years, which is significantly lower (p 0.01) (Fig 1). The univariate predictors of embolism after PMC included age (p 0.03), history of embolism (p 0.01), AF (p 0.01), mitral valve area (p 0.04), and pressure gradient (p 0.01). Multivariate analysis identified AF as the independent predictor of embolism in the late follow-up period (p 0.03) (Table 4). Comparative Study for Selected Patients The early and late complications are summarized in Table 5. The early mortality rate of MVS/maze was 0.9% in the selected patients. One patient who underwent the conventional maze procedure suffered from myocardial infarction and expired. In late results embolism occurred in 25 patients after PMC and in 2 patients after MVS/ maze (p 0.0004). There was no significant difference in the incidence of anticoagulant-related bleeding between the groups (p 0.74). The actuarial AF-free rates after MVS/maze were 89.9% at 1 year and 85.7% at 5 years and were significantly higher than those of 17.3% at 1 year and 4.2% at 5 years after PMC (p 0.0001) (Fig 2). The actuarial embolism-free rates were 98.0% at 5 years and 10 years after MVS/maze as compared with 84.8% at 5 years and 82.4% at 10 years after PMC (p 0.01) (Fig 3). The Early results Mortality Operative death ( 30 days) 0 0 1.00 Hospital death ( 30 days) 0 1* (0.9%) 0.36 Morbidity ( 30 days) Embolism (stroke) 2 (1.0%) 0 0.40 Mitral valve surgery 1 (0.5%) 0 0.64 Mediastinitis 0 2 (1.9%) 0.13 Late results ( ); mortality Embolism 25 (4) 2 (0) 0.001 Bleeding 3 (0) 2 (1) 0.74 Mitral valve surgery 38 (2) 4 (0) 0.0001 Endocarditis 3 (0) 0 0.25 Others 6 (6) 1 (1)... MVS/maze mitral valve surgery combined with the maze procedure; PMC percutaneous balloon mitral commissurotomy. actuarial embolism-free survival rate in the MVS/maze group was 95.5% at 5 years and 91.3% at 10 years and was significantly higher that that of 85.3% at 5 years and 79.3% at 10 years in the MVS/maze group (p 0.04). Comment As valvular fibrosis and calcification progresses, AF usually occurs in patients with rheumatic heart disease. In fact about 40% 75% of the patients with mitral stenosis exhibit concomitant AF at the time of intervention [10 13]. It was reported that there was a clear correlation between the development of mitral stenosis and the initiation of AF [14] and that AF closely correlated with the initiation of symptoms caused by a loss of atrial contraction and tachycardia. In addition the risk of the associated embolic complications is considerably high because of increased platelet function, coagulation status, and dilated atrium [3 5]. Therefore thromboembolism can occur as the initial event in some patients with mitral stenosis and AF. The administration of anticoagulants is feasible but not sufficient for the prevention of a stroke [14, 15]. Consequently AF has been accepted as an independent risk factor for morbidity and mortality [3]. Regarding the initiation of AF it had been previously Table 4. Cox Regression Analysis of Predictors of Embolism After PMC Variables Univariate Multivariate Hazard Ratio (95% CI) p Value Hazard Ratio (95% CI) p Value Age (yrs) 1.40 (1.00 1.08) 0.03 1.02 (0.97 1.06) 0.5 History of embolism 2.33 (1.21 4.98) 0.01 1.20 (0.51 2.79) 0.67 Atrial fibrillation 3.11 (1.26 8.00) 0.01 3.39 (1.08 10.60) 0.03 Mitral valve area (cm 2 ) 3.26 (1.04 10.24) 0.04 3.47 (0.85 14.18) 0.08 Pressure gradient (mm Hg) 0.99 (0.90 1.10) 0.01 1.10 (0.98 1.23) 0.09 Dimension of LA (mm) 1.01 (0.96 1.05) 0.97 0.98 (0.93 1.03) 0.39 CI confidence interval; LA left atrium; PMC percutaneous balloon mitral commissurotomy.

CARDIOVASCULAR 804 NAKAJIMA ET AL Ann Thorac Surg PMC VERSUS MVS/MAZE IN MITRAL STENOSIS AND AF 2004;78:800 6 Fig 2. The actuarial atrial fibrillation free rate in selected patients. (MVS/maze mitral valve surgery combined with the maze procedure; PMC percutaneous balloon mitral commissurotomy.) believed that AF was primarily related to the elevation of left atrial pressure and enlargement of the left atrium. However recent investigations have proposed that the presence of ectopic foci in pulmonary veins is the trigger initiating AF [16]. The dilated left atrium, degeneration of the left atrial myocardium caused by rheumatic inflammation, and the elevated atrial pressure also contribute to perpetuation of AF. Additionally the persistence of AF induces some structural and electrophysiological remodeling in the atrial myocardium [17]. Therefore it has been considered that recovery of valve function alone is not sufficient for sinus rhythm recovery [18]. PMC provides a hemodynamic improvement via enlargement of the valve orifice and is regarded as an effective and less invasive alternative to surgical commissurotomy. PMC is widely applicable for various patients including the young, aged, pregnant, and those unsuitable for surgery [19]. Several randomized studies demonstrated that the improvement and durability of the valve orifice area after PMC are similar to those after conventional surgical commissurotomy [3]. The morphology of the mitral valvular structure is considered as one of the most important predictive factors of successful Fig 3. The actuarial embolism-free rate in selected patients. (MVS/ maze mitral valve surgery combined with the maze procedure; PMC percutaneous balloon mitral commissurotomy.) PMC. Echocardiographic assessment has been widely used [10, 20, 21]. In addition the age [20, 22], NYHA functional class [10, 20], and anatomical form [22] are also long-term predictive factors regarding survival rate and repeated intervention. Concerning the freedom from valve replacement and survival after PMC the influence of AF is still controversial. Tarka and associates mentioned that although patients with sinus rhythm exhibited a greater cardiac output resulting in a larger final valve area than patients with AF, there was no significant difference in the longterm event-free survival rate between sinus rhythm and AF patients [12, 23]. In contrast there are some reports which state that the presence of AF is an independent or notable predictor of suboptimal early results, repeated intervention, and late mortality after PMC [10, 11, 22]. The investigation about the incidence and risk of embolism after PMC are very few, because embolism was not included among the end points in most of the clinical follow-up studies and prospective randomized studies. Although it has been believed that PMC exhibits favorable effects with regard to the prevention of thrombus formation by suppression of increased coagulability and disappearance of spontaneous echo contrast in the left atrium [4, 5, 13, 22, 24], it may be hypothesized that patients who underwent PMC exhibit a potential risk similar to those who underwent surgical commissurotomy. Hickey and associates reported that the thromboembolism-free rate after surgical commissurotomy was 94% at 5 years and 90% at 10 years, whereas AF was present in 38% of their patients [25]. In a review of the published literature it was noted that Smith and associates reported the incidence of embolism after open and closed valvotomy as 2.0 10.2% per patient per year in patients with AF, whereas it was reported to be 0.4 2.7% per patient per year in patients with sinus rhythm [26]. In the present study patients with AF before PMC could not resume sinus rhythm after PMC and the presence of AF before PMC significantly correlated with embolism in the late follow-up period after PMC. The incidence of embolic complication was 2.0% per patient per year in patients with AF before PMC. These results are in agreement with previous reports [25, 26]. In addition the maze procedure was effective and indispensable for sinus rhythm recovery and provided a significantly lower incidence of embolic complications as compared with PMC alone. These results strongly suggest that the presence of AF and the probability of sinus rhythm recovery should be taken into consideration when determining the indication of PMC and surgical treatment with regard to reducing embolic complications in the late follow-up period. The safety of including the maze procedure in the surgical treatment is considered acceptable. One patient died of myocardial infarction in the MVS/maze group, however this occurred before the introduction of our modified maze procedure. Although the cause remains unclear we speculate that it may be related to the operative maneuver of the conventional maze procedure that mainly employed the cut-and-sew technique. Re-

Ann Thorac Surg NAKAJIMA ET AL 2004;78:800 6 PMC VERSUS MVS/MAZE IN MITRAL STENOSIS AND AF cently the surgical technique of the maze procedure has been simplified. In our current maze procedure we use cryoablation for isolation of the pulmonary venous orifices instead of the cut-and-sew technique. It provides a sinus rhythm recovery rate comparable with that of the conventional maze procedure with a simple and safe surgical technique [7]. Because this study was retrospective and nonrandomized some differences regarding the characteristics of the PMC and MVS/maze groups were noted. These differences are the limitations of this study, however the influence of these differences on the late results before procedures were performed may not be important. This is because older age, lengthier history of AF, and larger left atriums in the MVS/maze group indicated that the patients in the MVS/maze group exhibited more advanced rheumatic valvular disease. Moreover although patients before MVS/maze commonly exhibited moderate mitral regurgitation, most of them underwent valve replacement and were diagnosed with rheumatic degeneration by the pathologic examination. NYHA class and mean pressure gradient across the mitral valve were similar in both groups. In conclusion because the presence of AF before PMC significantly correlated with thromboembolism in the late follow-up period after PMC, surgical treatment may only be beneficial for the patients who are considered as favorable candidates for the maze procedure. The prolonged duration of AF by deferment of surgical treatment may diminish the opportunity to restore sinus rhythm and lead to increased the risk of embolism. References 1. Iung B, Gohlke-Barwolf C, Tornos P, et al. Working Group on Valvular Heart Disease. Recommendation on the management of the asymptomatic patient with valvular heart disease. Eur Heart J 2002;23:1253 66. 2. Reyes VP, Raju BS, Wynne J, et al. Percutaneous balloon valvuloplasty compared with open surgical commissurotomy for mitral stenosis. N Engl J Med 1994;331:961 7. 3. Wolf PA, Dawber TR, Thomas HE, et al. Epidemiologic assessment of chronic atrial fibrillation and risk of stroke: The Framingham Study. Neurology 1978;28:973 7. 4. Zaki A, Salama M, El Marsy M, et al. Immediate effect of balloon valvuloplasty on hemostatic changes in mitral stenosis. Am J Cardiol 2000;85:370 5. 5. Cormier B, Vahanian A, Iung B, et al. Influence of percutaneous mitral commissurotomy on left atrial spontaneous contrast of mitral stenosis. Am J Cardiol 1993;71:852 7. 6. Cox JL, Ad N, Palazzo T. Impact of the maze procedure on the stroke rate in patients with atrial fibrillation. J Thorac Cardiovasc Surg 1999;118:833 40. 7. Nakajima H, Kobayashi J, Bando K, et al. The effect of cryo-maze procedure on early and intermediate term outcome in mitral disease: case matched study. Circulation 2002;106(Suppl I):I46 50. 8. Sie HT, Beukema WP, et al. Radiofrequency modified maze 805 in patients with atrial fibrillation undergoing concomitant cardiac surgery. J Thorac Cardiovasc Surg 2001;122:249 56. 9. Kobayashi J, Kosakai Y, Nakano K, et al. Improved success rate of the maze procedure in mitral valve disease by new criteria for patients selection. Eur J Cardiothorac Surg 1998;13:247 52. 10. Palacios IF, Sanchez PL, Harrell LC, et al. Which patients benefit from percutaneous mitral balloon valvuloplasty? Prevalvuloplasty and post valvuloplasty variables that predict long-term outcome. Circulation 2002;105:1465 71. 11. Leon MN, Harrel LC, Simosa HF, et al. Mitral balloon valvotomy for patients with mitral stenosis in atrial fibrillation Immediate and long-term results. J Am Coll Cardiol 1999;34:1145 52. 12. Tarka EA, Blitz LR, Herrmann HC. Homodynamic Effects and long-term outcome of percutaneous balloon valvuloplasty in patients with mitral stenosis and atrial fibrillation. Clin Cardiol 2000;23:673 7. 13. Chiang CW, Lo SK, Ko YS, et al. Predictor of systemic embolism in patients with mitral stenosis. A prospective study. Ann Intern Med 1998;128:885 9. 14. Horskotte D, Niehues R, Stauer BE. Pathomorphological aspects, aetiology and natural history of acquired mitral valve stenosis. Eur Heart J 1991;12:55 60. 15. Stern S, Altkorn D, Levinson W. Anticoagulation for chronic atrial fibrillation. JAMA 2000;283:2901 3. 16. Haissaguerre M, Jais P, Shan DC, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary vein. N Engl J Med 1998;339:659 66. 17. Kamalvand K, Tan K, Lloyd G, et al. Alteration in atrial electrophysiology associated with chronic atrial fibrillation in man. Eur Heart J 1999;20:888 95. 18. Wang A, Pulsipher MW, Harrison JK, et al. Predictors and significance of atrial rhythm before and six months after percutaneous balloon mitral commissurotomy. Am J Cardiol 1999;83:125 8. 19. Sutaria N, Elder AT, Shaw TRD. Long term outcome of percutaneous mitral balloon valvotomy in patients aged 70 and over. Heart 2000;83:433 8. 20. Dean LS, Mickel M, Bonan R, et al. Four-year follow up of the patients under going percutaneous balloon mitral commissurotomy. A report from the National Heart, Lung, and Blood Institute Balloon Valvuloplasty Registry. J Am Coll Cardiol 1996;28:1452 7. 21. Palacios IF, Tuzcu ME, Weyman AE, et al. Clinical follow-up of patients undergoing percutaneous mitral balloon valvotomy. Circulation 1995;91:671 6. 22. Iung B, Garbarz E, Mauchaud P, et al. Late results of percutaneous mitral commissurotomy in series of 1024 patients. Analysis of late clinical deterioration: frequency, anatomic findings, and predictive factors. Circulation 1999; 99:3272 8. 23. Iung B, Vahanian A. The long-term outcome of balloon valvuloplasty for mitral stenosis. Curr Cardiol Rep 2002;4: 118 24. 24. Iung B, Cormier B, Farah B, et al. Percutaneous mitral commissurotomy in the elderly. Eur Heart J 1995;16:1092 9. 25. Hickey MSJ, Blackstone EH, Kirklin JW, et al. Outcome probabilities and life history after surgical mitral commissurotomy: implications for balloon commissurotomy. J Am Coll Cardiol 1991;17:29 42. 26. Smith WM, Neutze JM, Barratt-Boyes BG, et al. Open mitral valvotomy. Effect of the preoperative factors on results. J Thorac Cardiovasc Surg 1981;82:738 51.35 CARDIOVASCULAR INVITED COMMENTARY This study, although limited by its retrospective design, is I believe an important one. The data confirm the excellent results that can be achieved with percutaneous balloon commissurotomy (PBC) with respect to function 2004 by The Society of Thoracic Surgeons 0003-4975/04/$30.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2004.05.039