Predictors of Sinus Rhythm Restoration After Cox Maze Procedure Concomitant With Other Cardiac Operations

Transcription

1 Predictors of Sinus Rhythm Restoration After Cox Maze Procedure Concomitant With Other Cardiac Operations Junya Kamata, MD, Kohei Kawazoe, MD, Hiroshi Izumoto, MD, Hiroto Kitahara, MD, Yoshitaka Shiina, MD, Yoshihiro Sato, MD, Kenji Nakai, MD, Takayoshi Ohkubo, MD, Ichiro Tsuji, MD, and Katsuhiko Hiramori, MD Third Department of Surgery,, Second Department of Internal Medicine, Iwate Medical University,, Iwate, and Department of Public Health, Tohoku University School of Medicine, Miyagi, Japan Background. There have been sporadic cases of persistent atrial fibrillation and sick sinus syndrome after the maze procedure. The purpose of this study was to identify the predictors of sinus rhythm restoration after operation. Methods. Between March 1993 and June 1995, we evaluated retrospectively 96 consecutive patients who underwent the maze procedure (maze III) in combination with another type of cardiac operation. Four patients who died and 6 patients who required permanent pacemaker implantation because of sick sinus syndrome were excluded. Ambulatory electrocardiographic monitoring was evaluated 1 year after operation. Multiple logistic regression analysis was applied to identify the predictors of sinus rhythm restoration. Results. The final population comprised 86 patients (mean age, 59.8 years; 67 patients with mitral valve disease). Overall, sinus rhythm was restored in 68 of 86 patients (79.1%). The magnitude of the atrial fibriuatory wave positively predicted postoperative sinus rhythm restoration. Conversely, left atrial diameter was inversely related to postoperative sinus rhythm restoration. The odds ratio of having both a fine atrial fibrillatory wave (<1.0 mm) and enlarged left atrial diameter (>_65 mm) for patients with sinus rhythm restoration was 0.04 (95% confidence interval, 0.01 to 0.28). Conclusions. Atrial fibrillatory wave and left atrial diameter were independent predictors of sinus rhythm restoration after the maze procedure in patients with chronic atrial fibrillation and organic heart disease. (Ann Thorac Surg 1997;64:394-8) 1997 by The Society of Thoracic Surgeons I n 1991, Cox and colleagues [1-5] described a new procedure for the radical surgical treatment of atrial fibrillation, the maze procedure, and presented their clinical experience. Initially, the maze procedure was performed as an isolated cardiac procedure [3]. However, as the experience of this procedure has expanded, it has been performed concomitantly with other cardiac procedures [6, 7]. It has been noted that operation for the underlying cardiac pathology alone usually fails to abolish atrial fibrillation [8]. Kosakai and colleagues [7, 9l reported the advantage of the maze procedure for atrial fibrillation in patients undergoing simultaneous open heart operation. They noted that patients who remained in atrial fibrillation after the maze procedure had a long duration of atrial fibrillation before operation and a large preoperative left atrial diameter [7]. However, definitive indications for the maze procedure in patients undergoing concomitant cardiac surgical treatment have not been developed. In this study, we use multiple logistic regression analysis to identify predictors of sinus rhythm restoration after the maze procedure. Accepted for publication Jan 22, Address reprint requests to Dr Kamata, Third Department of Surgery, Iwate Medical University, 19-1 Uchimaru, Morioka, lwate 020, Japan. Patients and Methods Patients Between March 1993 and June 1995, we evaluated retrospectively the charts of 96 consecutive patients undergoing the maze procedure combined with another type of cardiac operation. Four deaths and 6 patients who required permanent pacemaker implantation because of sick sinus syndrome [10] were excluded. This study was approved by Iwate Medical University Hospital Ethics Committee, and informed consent was obtained from all patients. Maze Procedure The operative procedure is fundamentally the same as that initially described by Cox [4], and basic atriotomies are performed according to Cox's second modification of his original procedure (maze III) [11]. After total cardiopulmonary bypass is established and cardioplegic arrest is obtained, left atriotomy is performed. The standard left atriotomy is extended inferiorly and superiorly around the left superior and inferior pulmonary veins. The left atrial appendage is excised. A cryoprobe is applied to tissue inferior to the excised atrial appendage and connecting the left atriotomy. A line of endocardial tissue, extending from the pulmonary vein isolation incision to 1997 by The Society of Thoracic Surgeons /97/$17.00 Published by Elsevier Science Inc PII S (97)

2 Ann Thorac Surg KAMATA ET AL ;64:394-8 SINUS RHYTHM RESTORATION AFTER MAZE the mitral valve annulus is cryoprobed. The left atriotomy is closed halfway and the mitral valve procedure is performed at this time if needed. After the closure of left atriotomy is completed, caval snares are tightened and the right atrial appendage is excised. A lateral incision, parallel to the right atrioventricular groove, is made. In our institution a line of incision for posterior longitudinal right atriotomy is replaced by longitudinal cryolesion. Some right atrial incisions of the maze III procedure are replaced by cryoablation. We use cryoablation over the atrial septal area between the coronary sinus and the inferior vena cava. Electrocardiography Standard 12-lead electrocardiography was performed in each patient before operation. All electrocardiograms were standardized to normal speed (25 mm/min) and sensitivity (1 mv input produced a 10-mm deflection). The atrial fibrillatory wave with the greatest size was measured in lead V 1 for at least 10 cardiac cycles. It was measured from the upper edge of the peak to the upper edge of the trough and was expressed in millimeters according to the method described by Peter and colleagues [12]. Coarse atrial fibrillatory wave was defined as any fibrillatory wave in lead V 1 with an amplitude 1.0 mm or more, whereas those with all fibrillatory waves in V1 less than 1.0 mm were designated as fine atrial fibrillatory wave. A single coarse fibrillatory wave in lead V~ was considered sufficient to classify the patient as having coarse atrial fibrillatory wave. The influences of the artifact on the baseline and the T or U waves were carefully excluded. All measurements were performed by two independent observers. Echocardiography Echocardiographic examinations were performed in all patients with a cardiac ultrasound imaging system (77035A; Hewlett Packard Co, MA) before operation. Left atrial diameter, left ventricular end-diastolic diameter, left ventricular ejection fraction, and percent fractional shortening were measured in a standard manner by M-mode tracing taken from two-dimensional parasternal long-axis views. Cardiac Catheterization Serial hemodynamics were measured in all patients with a clinical polygraph system (RMC-2000; Nihon Kohden Co, Ltd, Tokyo, Japan) and a cardiac output computer (MTC-6210; Nihon Kohden) before operation. The parameters measured were left ventricular end-diastolic pressure, mean pulmonary artery wedge pressure, mean pulmonary artery pressure, mean right atrial pressure, and cardiac index. Ambulatory Electrocardiographic Monitoring Ambulatory electrocardiographic monitoring (SCM-280; Fukuda Denshi Co, Ltd" Tokyo, Japan) and standard 12-lead electrocardiography were performed in each patient approximately 1 year after operation [10]. In the present study the patients were classified into two groups according to postoperative sinus rhythm restoration: successful restoration (group A) and unsuccessful restoration (group B). Group B included persistent atrial fibrillation and paroxysmal atrial fibrillation. Study Variables We examined 15 preoperative parameters as possible predictors for sinus rhythm restoration after the maze procedure (age, sex, duration of atrial fibrillation, previous cardiac operation, New York Heart Association functional class, atrial fibrillatory wave, left atrial diameter, left ventricular end-diastolic diameter, left ventricular ejection fraction, percent fractional shortening, left ventricular end-diastolic pressure, mean pulmonary artery wedge pressure, mean pulmonary artery pressure, mean right atrial pressure, and cardiac index). We measured all parameters before operation. We emphasize that there is no measurement bias. Statistical Analysis The association between each study variable and sinus rhythm restoration was examined first by bivariate analysis. For continuous variables, the difference between mean values was examined by Student's unpaired t test. For categoric variables, difference in the distribution of the variables was examined by the )(2 test. The significant variables determined by bivariate analysis were then included in the model for multiple logistic regression analysis. In this analysis, these continuous variables were converted into dummy variables of dichotomy or trichotomy. In calculating the odds ratio for postoperative sinus rhythm restoration for each study variable, we treated the group of patients with better preoperative status as the reference group. The odds ratio, along with the 95% confidence interval, was derived from the coefficient and the standard error. For all statistical analyses, p values of less than 0.05 were considered significant. SAS system [13] was used for all statistical calculations. Results Patient Characteristics Preoperative characteristics are shown in Table 1. The final population comprised 86 patients, 41 men and 45 women, with ages ranging from 40 to 77 years (mean, years). The mean duration of atrial fibrillation before operation was 10.2 _+ 7.8 years (range, 0.4 to 47.0; median (interquartile range), 9.6 (4.3 to 14.6)). All patients had organic heart disease (67 with mitral valve disease, 8 with congenital heart disease, 7 with aortic valve disease, and 4 others). Five patients had had previous cardiac operations (5.8%). The operations performed in conjunction with the maze procedure are as follows: mitral valve repair, 31 patients; open mitral commissurotomy, 10; mitral valve replacement, 26; aortic valve repair, 1; aortic valve replacement, 6; atrial septal defect closure, 6; ventricular septal defect closure, 1; patent ductus arteriosus division, 1; coronary artery bypass grafting, 1; myxoma resection, 1; Valsalva closure, 1; and pericardiectomy, 1.

3 396 KAMATA ET AL Ann Thorac Surg SINUS RHYTHM RESTORATION AFTER MAZE 1997;64:394-8 Predictors of Sinus Rhythm Restoration Sinus rhythm was restored in 68 of 86 patients (79.1%; group A) and was not restored in 18 patients (20.9%; group B). The atrial fibrillatory wave in group A (2.1 ± 1.0 ram) was greater than in group B (1.3 ± 0.9 ram; p < 0.01). The left atrial diameter in group A (55 ± 11 ram) was smaller than in group B (69 _+ 16 mm; p < 0.001). The mean pulmonary artery wedge pressure (13.2 ± 7.2 mm Hg) in group A was lower than in group B ( mm Hg; p < 0.05). The mean pulmonary artery pressure in group A (21.7 ± 8.3 mm Hg) was lower than in group B (27.7 ± 7.4 mm Hg; p < 0.01). The mean right atrial pressure in group A (4.4 ± 2.9 mm Hg) was lower than in group B (6.8 _+ 3.6 mm Hg; p < 0.01). The above variables plus age were included in a model of logistic regression analysis. First, the association between each preoperative state and sinus rhythm restoration was examined quantitatively. In this analysis, the patients were classified into three groups according to Table 1. Preoperative Characterish'cs of Patients: Comparison Between Successful Sinus Rhythm Restoration and Unsuccessful Sinus Rhythm Restoration" Group A Group B Characteristic (n = 68) (n = 18) p Value Age (y) NS Sex (male/female) 31/37 10/8 NS Duration of AF (y) ± 6.7 NS Previous cardiac operation 4 (5.9) 1 (5.6) NS NYHA (class) ~ 0.9 NS AFW (mm) 2.1 ± <0.01 LAD (ram) ~ 16 <0.001 LVEDD (mm) 55 _ ± 14 NS LVEF 0.64 _ NS %FS (%) 35 _ ± 9 NS LVEDP (ram Hg) 8.7 _ ± 4.3 NS Mean PAWP (mm Hg) ± 4.7 <0.05 Mean PAP (mm Hg) 21.7 _ ± 7.4 <0.01 Mean RAP (mm Hg) <0.01 CI(L.min 1.m 2) ±0.5 NS Mitral valve disease 51 (75.0) 16 (88.9) MS 14 4 MR 26 8 MSR 11 4 Rheumatic disease 27 8 Degenerative disease 22 7 Ischemic disease 2 1 Aortic valvular disease 7 (10.3) 0 Congenital heart disease 6 (8.8) 2 (11.1) Other 4 (5.9) 0 " Data are presented as mean _+ standard deviation or number (%). AF = atrial fibrillation; AFW - atrial fibrillatory wave; CI = cardiac index; %FS = percent fractional shortening; LAD = left atrial diameter; LVEDD = left ventricular end-diastolic diameter; LVEDP = left ventricular end-diastolic pressure; LVEF = left ventricular ejection fraction; MS - mitral stenosis; MSR mitral stenosis and regurgitation; MR = mitral regurgitation; NS = not significant; NYHA - New York Heart Association function class; PAP = RAP = right atrial pressure, Table 2. Odds Ratios and 95% Confidence Intervals for Postoperative Sinus Rhythm Restoration for Selected Preoperative Characteristics ~ Characteristic Confidence Interval) p Value Age (y) < >56, < ( ) 0.39 > ( ) AFW (mm) > <2.5, > ( ) < ( ) LAD (mm) < >51, < ( ) 0.03 > ( ) Mean PAWP (mm Hg) < >-10, < ( ) 0.04 > { ) Mean PAP (mm Hg) < >19, < ( ) > ( ) Mean RAP (mm Hg) < >3, < ( ) > ( ) a The odds ratio for each variable was calculated with a logistic regression model. AFW = atrial fibrillatory wave; LAD = left atrial diameter; PAP = RAP = right atrial pressure. the intertrisection range for each variable. As referenced to the patients with better preoperative state, odds ratio for postoperative sinus rhythm restoration decreased gradually in accordance with worse preoperative state for each variable (atrial fibrillatory wave, left atrial diameter, mean pulmonary artery wedge pressure, mean pulmonary artery pressure, and mean right atrial pressure). The )(2 test for linear trend was significant in these variables, suggesting that there were dose-response relationship between the magnitude of the above preoperative parameters and the odds for postoperative sinus rhythm restoration (Table 2). Second, all the above variables were converted into dummy variables of dichotomy, then included into the multiple logistic regression model. The reference category for each variable was defined as follows: age less than 65 years, atrial fibrillatory wave 1.0 mm or more [12], left atrial diameter less than 65 mm [10], mean pulmonary artery wedge pressure less than 15 mm Hg, mean pulmonary artery pressure less than 25 mm Hg, and mean right atrial pressure less than 5 mm Hg. Because the definite criteria about age, mean pulmonary artery wedge pressure, mean pulmonary artery pressure, and mean right atrial pressure were not published, we chose

4 Ann Thorac Surg KAMATA ET AL ;64:394-8 SINUS RHYTHM RESTORATION AFTER MAZE Table 3. Odds Ratios and 95% Confidence Intervals for Postoperative Sinus Rhythm Restoration for Selected Preoperative Characteristics a Characteristic Confidence Interval) Age (y) < > ( ) AFW (mm) -> < ( ) LAD (mm) < > ( ) Mean PAWP (ram Hg) < > ( ) Mean PAP (ram Hg) < > ( ) Mean RAP (mm Hg) < ~ ( ) The odds ratio for each variable was calculated with a logistic regression model that included all the above variables. AFW = atrial fibrillatory wave; LAD = left atrial diameter; PAP = RAP = right atrial pressure. the upper points of the intertrisection range as the cutoff points. As shown in Table 3, the results indicated that the atrial fibrillatory wave and left atrial diameter were the significant predictors of sinus rhythm restoration. Mean pulmonary artery wedge pressure, mean pulmonary artery pressure, and mean right atrial pressure did not become significant predictors of sinus rhythm restoration when the effect of other variables was controlled for. These findings were attributable to the fact that these parameters had high correlation with atrial fibrillatory wave and left atrial diameter. The magnitude of the atrial fibrillatory wave positively predicted postoperative sinus rhythm restoration. Conversely, left atrial diameter was inversely related to postoperative sinus rhythm restoration. The odds ratio of a fine atrial fibrillatory wave (<1.0 rnm) for sinus rhythm restoration was 0.14 (95% confidence interval, 0.03 to 0.57). The odds ratio of an enlarged left atrium (->65 mm) for sinus rhythm restoration was 0.21 (95% confidence interval, 0.05 to 0.87) (Table 3). A different model also was examined (Table 4). In this model, four categories were constructed according to the atrial fibrillatory wave and the left atrial diameter. Patients with both a coarse atrial fibrillatory wave (->1.0 ram) and a medium-sized left atrial diameter (<65 mm) were treated as the reference group). The odds ratio was calculated with a logistic regression model adjusted for the effects of age, mean pulmonary artery wedge pressure, mean pulmonary artery pressure, and mean right atrial pressure. The model revealed that the odds ratio of having both a fine atrial fibrillatory wave (<1.0 ram) and enlarged left atrial diameter (->65 ram) for sinus rhythm restoration was 0.04 (95% confidence interval, 0.01 to 0.28). Comment The present study demonstrates that the atrial fibrillatory wave and left atrial diameter were independent predictors of sinus rhythm restoration after the maze procedure in patients undergoing concomitant surgical treatment, and that the odds ratio of having both a fine atrial fibrillatory wave (<1.0 ram) and an enlarged left atrial diameter (->65 ram) for patients with sinus rhythm restoration was 0.04 (95% confidence interval, 0.01 to 0.28). Cox and colleagues [14] have reported a high rate of sinus rhythm restoration. However, 50% of their patients had paroxysmal atrial fibrillation, and 75% had idiopathic lone atrial fibrillation. In Japan in general the maze procedure is not performed on patients with lone atrial fibrillation. Therefore, we applied the Cox maze procedure to patients with organic heart diseases and atrial fibrillation. The baseline characteristics of the patients in this study differ greatly from the patients studied by Cox. There has been substantial agreement that direct current cardioversion after open commissurotomy for mitral stenosis is effective in patients who have had atrial fibrillation for 1 year or less and patients whose cardiothoracic ratio is less than 60% [8]. However, in our 86 patients, the mean duration of atrial fibrillation before operation was more than 10 years. Also, the mean preoperative cardiothoracic ratio was more than 60%. The success rate of the maze procedure tended to be lower in patients with mitral valve disease. Among the patients with mitral valve disease, however, the probability of sinus rhythm restoration was not associated with its clinical type such as mitral stenosis and mitral regurgitation. The success rate of sinus rhythm restoration in patients with rheumatic disease was similar to the success rate in patients with degenerative disease. Table 4. Odds Ratios and 95% Confidence Intervals for Postoperative Sinus Rhythm Restoration for Atrial Fibrillatory Wave and Left Atrial Diameter ~ Characteristic Group 1: AFW -> 1.0 mm and LAD < 65 mm Group 2: AFW -> 1.0 mm and LAD -~ 65 rnm Group 3: AFW < 1.0 mm and LAD < 65 mm Group 4: AFW < 1.0 mm and LAD -> 65 mrn Confidence Interval) ( ) 0.12 ( ) 0.04 ( ) Having both a coarse atrial fibrillatory wave (AFW) (>1.0 mm) and medium-sized left atrial diameter (LAD) (<65 mm) was treated as reference group. The odds ratio was calculated with a logistic regression model that adjusted for the effects of age, mean pulmonary, artery wedge pressure, mean pulmonary artery pressure, and mean right atrial pressure.

5 398 KAMATA ET AL Ann Thorac Surg SINUS RHYTHM RESTORATION AFTER MAZE 1997;64:394-8 Electrocardiographically, atrial fibrillation is characterized by disorganized atrial electrical activity and is classified into coarse or fine according to the atrial fibrillatory wave amplitude. Various workers have studied the relation between the left atrial size and the atrial fibrillatory wave amplitude [15, 16]. However, the relation could not be established. The atrial fibrillatory wave amplitude is mainly affected by the state of the heart, especially the atrium, such as atrial fibrosis and degeneration as well as the atrial size. With prolongation of atrial fibrillation, the atrial electric motive force is considered to decrease as a result of progression of atrial dilatation [17], loss of atrial muscle mass, and atrial fibrosis [18, 19], as well as degeneration of the atrial myocardium due to underlying disorders. Previously, we reported that the intraoperative atrial epicardial mapping data [20] were useful for predicting sinus rhythm restoration after the maze procedure. Preoperative average peak-to-peak atrial amplitude during atrial fibrillation in patients with restored sinus rhythm were significantly higher than in patients with persistent postoperative atrial fibrillation. Segawa and colleagues [21] have reported that the atrium showed severe degeneration and fibrosis in patients who had persistent atrial fibrillation after the maze procedure by pathologic analysis. This finding suggests that sinus rhythm restoration cannot be expected in cases where the atrium is expanded by disease to the point of degeneration and fibrosis. In fact, we noticed in our patients that the walls of large left atria were very thin on inspection at operation. The present results suggest that the indication for the maze procedure at the same time as another cardiac operation requires caution when the atrial fibrillatory wave is less than 1.0 mm and the left atrial diameter exceeds 65 ram. Further follow-up studies of atrial mechanical function [22-24], the incidence of thromboembolic events, and the evaluation of quality of life are needed to assess the Cox maze procedure in patients undergoing concomitant surgical treatment. A prospective, randomized trial may help to better define the criteria for the maze procedure. We gratefully thank Dr Shigeru Hisamichi of the Department of Public Health, Tohoku University School of Medicine, for his statistical expertise and assistance with the preparation of the manuscript. References 1. Cox JL, Schuessler RB, Boineau JP. The surgical treatment of atrial fibrillation. I. Summary of the current concepts of the mechanisms of atrial flutter and atrial fibrillation. J Thorac Cardiovasc Surg 1991;101: Cox JL, Canavan TE, Schuessler RB, et al. The surgical treatment of atrial fibrillation. II. Intraoperative electrophysiologic mapping and description of the electrophysiologic basis of atrial flutter and atrial fibrillation. J Thorac Cardiovasc Surg 1991;101: Cox JL, Schuessler RB, D'Agostino HJ, et al. The surgical treatment of atrial fibrillation. III. Development of a definitive surgical procedure. J Thorac Cardiovasc Surg 1991;101: Cox JL. The surgical treatment of atrial fibrillation. IV. Surgical technique. J Thorac Cardiovasc Surg 1991;101: Cox JL, Boineau JP, Schuessler RB, et al. Successful surgical treatment of atrial fibrillation. Review and clinical update. JAMA 1991;266: Bonchek LI, Burlingame MW, Worley SJ, Vazales BE, Lundy EF. Cox/maze procedure for atrial septal defect with atrial fibrillation: management strategies. Ann Thorac Surg 1993; 55: Kosakai Y, Kawaguchi AT, Isobe F, et al Cox maze procedure for chronic atrial fibrillation associated with mitral valve disease. J Thorac Cardiovasc Surg 1994;108: Sato S, Kawashima Y, Hirose H, Nakano S, Matsuda H, Shirakawa R. Long-term results of direct current cardioversion after open commissurotomy for mitral stenosis. Am J Cardiol 1986;57: Kosakai Y, Kawaguchi AT, Isobe F, et al. Modified maze procedure for patients with atrial fibrillation undergoing simultaneous open heart surgery. Circulation 1995;92(Suppl 2): Kamata J, Nakai K, Chiba N, et al. Electrocardiographic nature of restored sinus rhythm after Cox maze procedure in patients with chronic atrial fibrillation who also had other cardiac surgery. Heart 1997;77: Cox JL. Evolving applications of the maze procedure for atrial fibrillation. Ann Thorac Surg 1993;55: Peter RH, Morris JJ Jr, McIntosh HD. Relationship of fibrillatory waves and P waves in the electrocardiogram. Circulation 1966;33: SAS Institute Inc. SAS/STAT user's guide, release 6.10 Edition. Cary, NC: SAS Institute Inc, Cox JL, Boineau JP, Schuessler RB, Kater KM, Lappas DG. Surgical interruption of atrial reentry as a cure for atrial fibrillation. In: Olsson SB, Allesie MA, Campbell RWF, eds. Atrial fibrillation: mechanisms and therapeutic strategies. New York: Futura, 1994: Aysha MH, Hassan AS. Diagnostic importance of fibrillatory wave amplitude: a clue to echocardiographic left atrial size and etiology of atrial fibrillation. J Electrocardio11988;21: Morganroth J, Horowitz IN, Josephson ME, Kaster JA. Relationship of atrial fibrillatory wave amplitude to left atrial size and etiology of heart disease. Am Heart J 1979;97: Henry WL, Morganroth J, Pearlman AS, et al. Relation between echocardiographically determined left atrial size and atrial fibrillation. Circulation 1976;53: Bailey GWH, Branitf BA, Hancock EW, Cohn KE. Relation of left atrial pathology to atrial fibrillation in mitral valvular disease. Ann Intern Med 1968;69: Davies MJ, Pomerance A. Pathology of atrial fibrillation in man. Br Heart J 1972;34: Yagi Y, Mukaida M, Chiba N, et al. Predictors of clinical outcome of maze (Cox III) procedure for chronic atrial fibrillation associated with underlying heart disease: the importance of atrial epicardial mapping [Abstract]. Circulation 1995;92(Suppl 1): Segawa I, Tashiro A, Sato M, Hiramori K, Yagi Y, Kawazoe K. Clinical experience of the tissue characterization of the atrium and clinical course after the maze procedure [Abstract]. Jpn Circ J 1995;59: Feinberg MS, Waggoner AD, Kater KM, Cox JL, Lindsay BD, Perez JE. Restoration of atrial function after the maze procedure for patients with atrial fibrillation. Assessment by Doppler echocardiography. Circulation 1994;90(Suppl 2): Feinberg MS, Waggoner AD, Kater KM, Cox JL, Perez JE. Echocardiographic automatic boundary detection to measure left atrial function after the maze procedure. J Am Soc Echocardiogr 1995;8: Itoh T, Okamoto H, Nimi T, et al. Left atrial function after Cox's maze operation concomitant with mitral valve operation. Ann Thorac Surg 1995;60: