Efficacy of Intraoperative Mapping to Optimize the Surgical Ablation of Atrial Fibrillation in Cardiac Surgery

Transcription

1 Efficacy of Intraoperative Mapping to Optimize the Surgical Ablation of Atrial Fibrillation in Cardiac Surgery Shigeo Yamauchi, MD, Hidetugu Ogasawara, MD, Yoshiaki Saji, MD, Ryuzo Bessho, MD, Yasuo Miyagi, MD, and Masahiro Fujii, MD Department of Thoracic and Cardiovascular Surgery, Chiba Hokusoh Hospital, Nippon Medical School, Tokyo, Japan Background. Observation during open heart surgery in patients with chronic atrial fibrillation (AF) showed that the activation sequence of the left atrium was regular and that of the right atrium chaotic in most patients. We speculate that the left atrium plays a role as an important electrical driving chamber for AF and by mapping preoperatively, optimal sites for the cryoablation can be determined to minimize the extensiveness of the cryolesions. Methods. Forty patients who underwent cardiac surgery and cryoablation guided by epicardial mapping data to eliminate AF originating from the left atrium were included in this study. Results. Sustained reentrant movement or repetitive firing from foci located in the right atrium was never observed. Foci or reentry circuits located in the left atrium were clearly identified in 11 cases. Nine of the 11 cases resumed sinus rhythm by placing cryolesions at these sites. Two cases needed a pacemaker implantation. The exact site had not been identified in the 29 remaining cases. In these 29 cases a left atrial posterior longitudinal linear cryoablation was placed. Sinus rhythm resumed in 22 cases. Six cases still remained in AF and a pacemaker was implanted in 1 case. Ultimately, in this series of operations sinus rhythm was resumed in 31 of 40 cases; AF remained in 6 of them and pacemaker implantation was required in 3 cases. Conclusions. Mapping was useful to distinguish the two etiologies of the AF to facilitate optimal placement of the cryolesions. Sustained reentrant movement or repetitive firing from foci located in the right atrium was never observed and the left atrium played an important role as the electrical driving chamber for AF. (Ann Thorac Surg 2002;74:450 7) 2002 by The Society of Thoracic Surgeons Much attention has been drawn to reports of surgically treating atrial fibrillation (AF) with the Maze procedure. The Maze procedure is successful in ablating AF clinically [1 3]. The concept of the Maze procedure depends on a spectrum of arrhythmias consisting of multiple small macroreentrant circuits located throughout the atria tissue [4]. In our previous study of the surgical ablation of atrial flutter (AFL), activation sequence mapping before and after the operation demonstrated that there were multiple potential reentrant pathways. Therefore all potential pathways must be surgically interrupted to prevent the inducibility of AFL [5]. Further we also reported dissimilar rhythms occurring simultaneously in the right and left atria. Short cycle length tachycardias due to either circus movement or automatic foci in the left atrium did not propagate uniformly throughout the right atrium at equivalent cycle lengths. The right atrial activation pattern was so-called chaotic [6, 7]. Observation during open heart surgery in patients with chronic AF showed that the activation sequence of the left atrium was regular and that of the right atrium Accepted for publication April 16, Address reprint requests to Dr Yamauchi, Associate Professor of Surgery, Thoracic and Cardiovascular Surgery, Chiba Hokusoh Hospital, 1715 Kamagari, Inba, Chiba , Japan; shigeo@nms.ac.jp. was chaotic in most patients [8, 9]. We speculate that the left atrium plays an important role as the electrical driving chamber for AF. Based on the evidence from preoperative epicardial mapping data we simultaneously performed a surgical procedure on the left atrium to ablate the AF during open heart surgery, thus optimizing the location for the cryolesions. Material and Methods The patients consisted of 22 women and 18 men who ranged in age from 44 to 76 years (average 60). The underlying cardiac disease, etiology, AF duration, and surgical procedures are summarized in Table 1. The heart was exposed by a median sternotomy. Before the institution of cardiopulmonary bypass, activation sequence maps were obtained during AF from unipolar potentials recorded simultaneously from electrodes arranged on an epicardial template. In an early series of 29 patients 30 silver unipolar electrodes were used [9], and in the remaining 11 patients atrial mapping was performed with a newly developed mapping system with 60 unipolar electrodes. The card-type electrode had 60 unipolar electrodes, 2-mm in diameter each, mounted in six rows of ten on a flexible plastic rectangular sheet 2002 by The Society of Thoracic Surgeons /02/$22.00 Published by Elsevier Science Inc PII S (02)

2 Ann Thorac Surg YAMAUCHI ET AL 2002;74:450 7 SURGICAL TREATMENT OF ATRIAL FIBRILLATION 451 Table 1. Underlying Cardiac Disease, Etiology, Atrial Fibrillation Duration, and Surgical Procedures Patient No. Sex (Male/Female) Age (y) Diagnosis Etiology Duration of AF (y) Surgical Procedures 1 F 54 MS, AR Rheumatic 10 OMC, AVR 2 F 63 MS, TR Rheumatic 20 MVR, TAP 3 F 62 MS, LA thrombus Rheumatic 15 MVR 4 F 59 MR Degenerative 5 MVR 5 M 52 MR Degenerative 4 MVR 6 M 61 ASD Congenital 5 ICR 7 M 49 ECD, MR, TR Congenital Unknown ICR 8 F 55 MR Rheumatic 2 MVR 9 M 64 MR Rheumatic 4 MVP 10 M 65 MR Rheumatic 4 MVR 11 M 75 MR, TR Degenerative 10 MVR, TAP 12 F 62 MS, TR, post PTMC Rheumatic 10 MVR, TAP 13 F 51 AS, MS, post CMC Rheumatic 10 AVR, MVR 14 F 59 MS, LA thrombus Rheumatic 10 MVR 15 M 52 MR Chorda rupture 1 MVP 16 F 58 MS, TR, LA thrombus Rheumatic 5 MVR, TAP 17 F 57 MS, LA thrombus Rheumatic 10 MVR 18 M 64 MR Degenerative 5 MVP 19 F 64 MS Rheumatic Unknown MVR 20 F 48 MS, LA thrombus Rheumatic 10 MVR 21 M 44 MR Degenerative 5 MVP 22 F 52 MS, TR, LA thrombus Rheumatic 10 MVR, TAP 23 F 51 MS, TR Rheumatic 6 MVR, TAP 24 F 72 MR Rheumatic 5 MVR 25 M 56 MR Degenerative 2 MVP 26 M 56 MS, TR Rheumatic Unknown MVR, TAP 27 M 70 MR, TR Rheumatic 4 MVR, TAP 28 M 66 MS Rheumatic 10 MVR 29 F 58 MR Degenerative 20 MVR 30 F 68 MS Rheumatic 20 MVR 31 F 75 MR Degenerative 4 MVR 32 F 57 MS, TR, LA thrombus Rheumatic 20 MVR, TAP 33 F 72 MS, AS Rheumatic 15 MVR, AVR 34 M 52 MR Degenerative 4 MVR 35 F 55 MS, TR Rheumatic Unknown MVR, TAP 36 F 69 MS, TR Rheumatic 7 MVR, TAP 37 M 67 MR IE 0.4 MVR 38 M 54 ASD, MR Congenital 10 ICR, MVP 39 M 76 MS Rheumatic 5 MVR 40 M 56 AS Rheumatic 3 AVR AS aortic valve stenosis; ASD atrial septal defect; AVR aortic valve replacement; CMC closed mitral commissurotomy; ECD endocardial cushion defect; ICR intracardiac repair; IE infective endocarditis; MR mitral valve regurgitation; MS mitral valve stenosis; MVP mitral valvuloplasty; MVR mitral valve replacement; OMC open mitral commissurotomy; PTMC percutaneous transluminal mitral commissurotomy; TAP tricuspid annuloplasty; TR tricuspid valve regurgitation. (card-type electrode, mm). Each unipolar site contained a fine silver wire electrode separated by a distance of 5 mm. The unipolar data were recorded at a frequency response of 50 to 1000Hz. A computer was used to store the digitalized unipolar data and display the waveforms. Our 60-channel mapping system was not able to perform simultaneous right and left atrial mapping. The card-type electrode was attached to the right atrium to record 60 local right atrial epicardial electrograms in 3 seconds and the electrode was switched to the left atrium to record the local left atrial epicardial electrograms. If we were able to detect the precise location of the focus or reentry circuit, cryoablation was placed at the point of the focus or reentry circuit. If we could not detect the precise location of the focus or reentry circuit, linear cryoablation and pulmonary vein cryoablation were performed as described below. After heparinization the ascending aorta and both vena cavae were cannulated directly for cardiopulmonary bypass. On complete bypass with normothernic extracorporeal circulation, the aorta was cross-clamped and bidirectional cardioplegia including terminal warm blood

3 452 YAMAUCHI ET AL Ann Thorac Surg SURGICAL TREATMENT OF ATRIAL FIBRILLATION 2002;74:450 7 cardioplegia was given to arrest the heart. The combined superior-transseptal approach was used to access the left atrium [10]. In the case of mitral valve replacement, before the attachment of the mitral mechanical valve cryoablation lesions were placed at a point near one third of the posteromedial commisure of the mitral annulus. A left atrial posterior longitudinal linear cryoablation was then performed by making a cryolesion that extended from the posterior mitral annulus to the incisional line of the roof of the transverse sinus of the pericardium including both the right inferior and superior pulmonary veins. Another posterior longitudinal linear cryolesion, which included both the left inferior and superior pulmonary veins, was made starting from a distance of 1 cm from the mitral annulus to the incisional line of the roof of the transverse sinus of the pericardium. The left atrial appendage was not resected Results After the aortic cross-clamping was released, the heart was beating spontaneously in all cases. We never used a cardiodefibrillator. The postoperative heart rhythm, left and right atrial activation patterns, and mean fibrillatory cycle length are summarized in Table 2. The mean fibrillatory cycle length was calculated within a window of 1200 ms. We never observed sustained reentrant movement or repetitive firing from foci originating from the right atrium. Foci or reentry circuits originating from the left atrium were clearly identified in 11 cases (patients 1 to 11). Cryoablation was placed precisely at the focus or reentry circuit in these cases. Nine of the 11 cases resumed sinus rhythm after the operation. Two cases (patients 8 and 11) needed a pacemaker implantation 1 month after the operation owing to sick sinus syndrome. Both had single-chamber pacemakers programmed to the VVI mode, implanted because of a very low amplitude of the atrial electrograms, which provided successful pacing therapy. Twenty-two cases exhibited a regular repetitive left atrial activation pattern and another 4 cases showed a chaotic activation pattern in the left atrium. However, we could not identify the precise location of the focus or reentry circuit in the area that was mapped in the left atrium. Either both or either the right or left atrial electrograms were unable to be recorded in 3 cases. One case (patient 13) had an extensive adhesion of both atria, which was caused by a former operation, and 2 cases (patients 14 and 33) had a remarkably low amplitude of the atrial voltage. Therefore linear cryoablation lesions were applied to the left atrium in 29 cases (patients 12 to 40). Sinus rhythm resumed in 22 cases and 6 cases still remained in AF. One case (patient 33) needed a pacemaker implantation for backup pacing due to sick sinus rhythm. Her pacemaker was programmed to the VVI mode and successfully supported her rhythm. In total, for this series of operations sinus rhythm was regained in 31 cases (31 of 40) and AF remained in 6 cases (6 of 40). A pacemaker implantation was required in 3 cases (3 of 40). The decision to implant a pacemaker was based on evidence of low electrocardiograpy (ECG) and atrial voltages. Owing to the presence of atrial thrombi detected preoperatively, electrophysiology studies were not possible and the low ECG and atrial voltages were the only available predictors to determine the need for a pacemaker implantation. Epicardial Mapping Data Figure 1 demonstrates an example of regular repetitive firing from the left atrial appendage (patient 3). Activation maps were made from 4 time frames corresponding to successive left atrial repetitive firings occurring from the left atrial appendage. The four time frames (A, B, C and D) in Figure 1 represent successive cycles with a window of analysis of 150 ms. On the isochrome maps the earliest activation is depicted by the silver lines and latest activation by the blue lines. These are isoactivation maps. The analysis window of each timeframe was equal to the cycle length of the tachycardia. The cycle length of this sustained tachycardia was 150 msec. In each frame the site of earliest activation always originated from the left atrial appendage and the wavefront propagated into the posterior wall in all these maps. The ectopic focus could be defined as the origin of the activation occurring from the same location (left atrial appendage) at a constant interval of 150 ms. Conversely the right atrium was activated passively and relatively regularly through Bachmann s bundle, the interatrial septum, or the posterior portion of the left atrium. The mean right atrial fibrillatory cycle length was 155 ms. The epicardial mapping proved that repetitive firing existed in the left atrial appendage and a cryoablation lesion (shown by the asterisk in Fig 1) was placed at this point, successfully eliminating the tachycardia and negating the need for linear lesions. Figure 2 demonstrates an example of circus movement reentry around an obstacle created by functional block at the junction between the left posterior atrial appendage and left atrial posterior wall (patient 5). Two consecutive left atrial maps were demonstrated. In this example, circus movement occurred around a small segment of functional block. We defined functional block (a functional barrier of conduction block) as decremental conduction leading to difference in propagation or conduction properties or both. The pulmonary vein orifices did not appear to play even a subsidiary role in this tachycardia. The mean fibrillatory cycle length of the left atrium was 191 ms. The direction of the circus wavefront was clockwise. Conversely the right atrium was activated passively through Bachmann s bundle or the posterior portion of the left atrium. The cycle length of the right atrium was 195 msec. Cryoablation from a site distant from the mitral annulus was performed and included placing the ablation line along an identifiable line, which passed along the left inferior and superior pulmonary veins. The sites of the cryolesions are shown as asterisks in Figure 2. Again this required fewer linear lesions than did that of the linear lesion method for chaotic mechanisms. Figure 3 shows repetitive firing from the left superior

4 Ann Thorac Surg YAMAUCHI ET AL 2002;74:450 7 SURGICAL TREATMENT OF ATRIAL FIBRILLATION 453 Table 2. Postoperative Heart Rhythm, Atrial Activation Pattern, and Mean Fibrillatory Cycle Length Patient No. Postoperative Rhythm LA Activation Pattern LA Mean Fibrillatory Cycle Length (ms) RA Activation Pattern RA Mean Fibrillatory Cycle Length (ms) 1 Sinus Focus (LSPV) 170 Regular Sinus Focus (appendage) 160 Regular Sinus Focus (appendage) 150 Regular Sinus Reentry (appendage) 150 Regular Sinus Reentry 191 Regular Sinus Focus (LSPV) 114 Chaotic Sinus Reentry (appendage) 160 Chaotic SSS (PMI) Reentry (appendage) 180 Chaotic Sinus Focus (LSPV) 200 Chaotic Sinus Reentry 180 Chaotic SSS (PMI) Reentry (appendage) Sinus Regular 150 Chaotic Sinus 14 AF 15 Sinus Regular 200 Chaotic Sinus Regular 165 Regular Sinus Regular 160 Regular Sinus Chaotic 158 Regular Sinus Chaotic 158 Regular Sinus Regular 130 Chaotic Sinus Regular 128 Regular AF Regular 190 Chaotic AF Regular 168 Chaotic Sinus Regular 182 Chaotic AF Regular 168 Chaotic Sinus Regular 170 Chaotic Sinus Regular 160 Regular AF Chaotic 155 Chaotic Sinus Regular 135 Chaotic Sinus Regular 155 Chaotic Sinus Regular 160 Chaotic Sinus Regular 163 Chaotic SSS (PMI) 34 Sinus Regular 140 Chaotic Sinus Regular 128 Chaotic Sinus Regular 125 Chaotic Sinus Regular 150 Regular Sinus Chaotic 170 Chaotic AF Regular 185 Chaotic Sinus Regular 130 Chaotic 144 AF atrial fibrillation; LSPV left superior pulmonary vein; PMI pacemaker implantation; SSS sick sinus syndrome; epicardial electrograms could not be recorded. pulmonary vein (patient 1). The tachycardia wavefront originated repetitively from a site in the left superior pulmonary vein. The cycle length of this sustained tachycardia was 170 ms. The right atrium was activated through Bachmann s bundle and the cycle length of the right atrium was 190 ms. On the isochrome maps the earliest activation is depicted by the purple lines and the latest activation by the red lines. A cryolesion was placed in the left superior pulmonary vein, as shown by the asterisk, that rendered the AF noninducible and eliminated the need for linear lesions. Comment The Maze procedure has been used in patients undergoing surgical treatment for mitral valve disease or atrial septal defects without first performing intraoperative atrial activation mapping. Activation mapping has repeatedly been used to better understand the mechanisms underlying AFL and AF. Owing to its complexity AFL has been defined and described adequately but not its mechanism. One question that can be asked is when does an apparent irregular and chaotic signal propagation represent a chaotic source and when does it represent singular

5 454 YAMAUCHI ET AL Ann Thorac Surg SURGICAL TREATMENT OF ATRIAL FIBRILLATION 2002;74:450 7 Fig 1. Isochrome maps and electrograms showing regular repetitive firing from the left atrial appendage. The four time frames show the isochrome activation maps corresponding to four successive left atrial repetitive ectopic firings occurring from the left atrial appendage. The wavefront propagated into the posterior wall in these maps. The cycle length of this sustained tachycardia and the window of analysis of each successive timeframe were 150 ms. In contrast, the right atrium was activated passively and relatively regularly through Bachmann s bundle, the interatrial septum, or the posterior portion of the left atrium. The mean right atrial fibrillatory cycle length was 155 ms. The silver lines on the isochrome map represent the earliest activation and the blue lines the latest activation. In the four frames, activation corresponding to the earliest site of activation seen in the electrograms is always from the left atrial appendage, suggesting a focal origin of this tachycardia. The stars represent the sites of the cryolesions. focal activation that, occurring at a high rate, results in complex dissociation of departing wavefronts? The exact mechanism of AF is still unknown but experimental studies have suggested that AF occurs through a mechanism of reentrant movement rather than as an ectopic foci. Recently Haissaguerre and coworkers [12] reported the spontaneous initiation of AF due to ectopic beats originating from the pulmonary veins and reported the successful application of radiofrequency ablation at these focal sources. Although tachycardias of focal origin were described in this study it seems likely that these arrhythmias also represent some form of reentry in which the circus pathway could not be properly defined owing either to the location of the circuit or to the resolution of the maps. Additionally the sites were always near the atrial margins where electrodes were either sparse or located adjacent to an anatomic edge in which the tissue folded on itself or joined with an orifice. Although in such cases a competing focal tachycardia could not be ruled out, logic suggested otherwise. We presume that the focality in such cases is the result of a change to a more convoluted three-dimentional reentrant circuit, which could not be resolved accurately by the mapping techniques limited to the epicardial surfaces. The effective refractory period of the left atrium is significantly shorter than that of the right atrium [13, 14]. The excitation due to circus movement and automatic foci in the left atrium does not demonstrate 1:1 conduction within the right atrium, which has a prolonged refractory period. As a result the excitation propagation in the right atrium clearly exhibited a chaotic state. The right atrial dissociation was characterized by some areas conducting 1:1 and other areas conducting with a progressive Wenckebach-like delay and ending in unidirectional block with activation from another or opposite direction by either a subsequent left atrial cycle or by the same right atrial wavefront arriving after a considerable delay over a longer pathway. This atrial discordance can

6 Ann Thorac Surg YAMAUCHI ET AL 2002;74:450 7 SURGICAL TREATMENT OF ATRIAL FIBRILLATION 455 Fig 2. Isochrome maps and electrograms showing circus movement located in the left atrial posterior wall. This shows an example of circus movement around an obstacle created by functional block at the junction between the left posterior atrial appendage and left atrial posterior wall. Two consecutive left atrial maps were demonstrated. In this example, circus movement occurred around a small segment of functional block. The orifices did not appear to play even a subsidiary role in this tachycardia. The mean fibrillatory cycle length of the left atrium was 191 ms and the direction of the circus wavefront was clockwise. In contrast, the right atrium was activated passively through Bachmann s bundle or the posterior portion of the left atrium. The cycle length of the right atrium was 195 ms. Note that in frames A and B, double potentials are recorded suggesting that the first potential represents the activation moving away and the second potential the activation returning as it goes around the area of functional block. Although there were no electrodes located exactly on the turnaround point, the electrogram recordings suggest that the turnaround point was located somewhere posterior to the atrial appendage. The stars represent the sites of the cryolesions. be explained by the difference in the refractory periods of the two atria [7, 11]. The mean fibrillatory cycle length of the left atrium was shorter than that of the right atrium. Sueda and coworkers [15] reported the same results. We never observed sustained reentrant movement or repetitive firing from a focus originating in the right atrium. We do not believe the alternate explanation that the intricate activation in the right atrium propagates to the left atrium and results in the repetitive activation in the left atrium. Vigano and coworkers [16] applied several surgical techniques to restore sinus rhythm in patients with mitral valve disease and an atrial septal defect. They recommended that different surgical options should be chosen for different cases of AF and underlying cardiac disease. We agree with their recommendation. Furthermore mapping is useful in distinguishing the two etiologies, ie, circus movement or repetitive firing from a focus. If the mechanism is clearly identified in each case, it would become a good index for choosing to use either the Maze procedure or cryoablation at sites of repetitive firing. Cox and coworkers [4] reported that multiple wavefronts, nonuniform conduction, bidirectional block, and macroreentrant circuits occur during AF. The left atrium demonstrated multiple wavefronts and conduction block but left atrial reentry could not be detected. The presence of macroreentrant circuits and the absence of either microreentrant circuits or evidence of atrial automaticity suggest that atrial fibrillation should be amenable to surgical ablation. Nevertheless we could clearly identify that the focus and reentrant circuit originated from the left atrium in 11 cases. Our present study suggests that there is a general tendency for reentrant circuits or ectopic foci to exist in the left atrium but not in the right atrium. The left atrium plays an important role as the

7 456 YAMAUCHI ET AL Ann Thorac Surg SURGICAL TREATMENT OF ATRIAL FIBRILLATION 2002;74:450 7 Fig 3. Isochrome maps and electrograms showing repetitive firing from the left superior pulmonary vein. This tachycardia wavefront originated repetitively from a site in the left superior pulmonary vein. The cycle length of this sustained tachycardia was 170 ms. The right atrium was activated through Bachmann s bundle. The cycle length of the right atrium was 190 ms. The star represents the site of the cryolesion. driving chamber for the AF in the majority of patients with chronic AF complicated by isolated mitral valve disease even if secondary to an ASD. We believe that the Maze procedure is the first choice of operation for AF caused by macroreentry because all potential pathways must be surgically interrupted to prevent the inducibility of AF even if the dominant reentry circuit is interrupted [5]. However, when the tachycardia can be determined as focal in origin or involving an isolated reentrant circuit, it can be cured with limited cryolesions. For the above reasons activation mapping is necessary in order to understand the mechanism of AF independent of the underlying cardiac disease. In our series, 3 patients needed pacemaker implantation. Two cases had a remarkably low atrial voltage and low magnitude of the fibrillatory wave. The magnitude of the fibrillatory voltage and wave is a good predictor for postoperative sinus restoration [3]. In summary, if epicardial mapping is performed preoperatively to determine the mechanism of the tachycardia it will not be necessary to perform the Maze procedure when the mechanism involves circus movement or repetitive firing from a focus. Further, in this series of patients no AF originated from the right atrium in any cases and the left atrium played an important role as the electrical driving chamber for AF. Limitations Our 60-channel mapping system was not able to perform simultaneous right and left atrial mapping or to cover the whole atrium. A more sophisticated mapping system capable of simultaneous mapping of both the right and left atria is required to analyze the detailed mechanism of AF. Further, it was not possible to clearly determine the focal origin of some of the tachycardias and there was the possibility that they involved a reentrant pathway in which the circus movement could not be clearly defined because of either the location of the circuit or the resolution of the maps. Often the sites were near atrial margins where electrodes were either sparse or located adjacent to an anatomic edge in which the tissue was folded upon itself or joined an orifice. Although we could not rule out a competing reentrant tachycardia, logic suggested otherwise. We presumed that the focality was the result of a change to a more convoluted threedimensional reentrant circuit that we could not deter-

8 Ann Thorac Surg YAMAUCHI ET AL 2002;74:450 7 SURGICAL TREATMENT OF ATRIAL FIBRILLATION 457 mine with the mapping techniques limited to the epicardial surface. References 1. Cox JL, Boineau JP, Schuessler RB, Kater KM, Lappas DG. Five-year experience with the maze procedure for atrial fibrillation. Ann Thorac Surg 1993;56: Gregori F Jr, Cordeiro CO, Couto WJ, da Silva SS, de Aquino WK, Nechar A Jr. Cox maze operation without cryoablation for the treatment of chronic atrial fibrillation. Ann Thorac Surg 1995;60: Kamata J, Kawazoe K, Izumoto H, et al. Predictors of sinus rhythm restoration after maze procedure concomitant with other cardiac operations. Ann Thorac Surg 1997;64: Cox JL, Canovan TE, Schuessler RB, et al. Surgical treatment of atrial fibrillation. II. Intraoperative electrophysiologic mapping and description of the electrophysiologic basis of atrial flutter and atrial fibrillation. J Thorac Cadiovasc Surg 1991;101: Yamauchi S, Schuessler RB, Kawamoto T, Shuman T, Boineau JP, Cox JL. Use of intraoperative mapping to optimize surgical ablation of atrial flutter. Ann Thorac Surg 1993; 56: Yamauchi S, Sato S, Schuessler RB, Boineau JP, Matsunaga Y, Cox JL. Induced atrial arrhythmias in a canine model of left atrial enlargement PACE 1990;13: Yamauchi S, Boineau JP, Schuessler RB, Cox JL. Varying types of circus movement re-entry with both normal and dissociated contralateral conduction causing different right and left atrial rhythms in canine flutter. Jpn Circ J 1998;62: Yamauchi S, Asano T, Harada A, Ikeshita M, Tanaka S, Shoji T. Simaltaneous surgical treatment for atrial fibrillation, and mitral valve disease. Jpn J Cardiovasc Surg 1994;23: Harada A, Sasaki K, Fukushima T, et al. Atrial activation during chronic atrial fibrillation in patients with isolated mitral valve disease. Ann Thorac Surg 1996;61: Eric Berreklouw, Ercan H, Schönberger JP. Combined superior-transseptal approach to the left atrium. Ann Thorac Surg 1991;51: Yamauchi S, Imura H, Bessho R, Yamada K, Tanaka S. Simultaneous surgical correction of a common atrium and impure flutter. Ann Thorac Surg 1997;64: Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 1998;339: Rensma PL, Allessie MA, Lammers WJ, Bonke FI, Schalij MJ. Length of excitation wave and susceptibility to reentrant atria arrhythmias in normal conscious dogs. Circ Res 1988; 62: Sato S, Yamauchi S, Schuessler RB, Boineau JP, Matsunaga Y, Cox JL. The effect of augmented atrial hypothermia on atrial refractory period, conduction, and atrial flutter/ fibrillation in the canine heart. J Thorac Cardiovasc Surg 1992;104: Sueda T, Nagata H, Shikata H, et al. Simple left atrial procedure for chronic atrial fibrillation associated with mitral valve disease. Ann Thorac Surg 1996;62: Vigano M, Gaffing A, Ressia L, Minzioni G, Aiello M, Gazzoli F. Surgery for atrial fibrillation. Eur J Cardiothorac Surg 1996;10:490 7.