- Email: [email protected]
Clinical Determinants of Sinus Conversion by Radiofrequency Maze Procedure for Persistent Atrial Fibrillation in Patients Undergoing Concomitant Mitral Valvular Surgery
Clinical Determinants of Sinus Conversion by Radiofrequency Maze Procedure for Persistent Atrial Fibrillation in Patients Undergoing Concomitant Mitral Valvular Surgery Mien-Cheng Chen, MDa,*, Jen-Ping Chang, MDb, Hsueh-Wen Chang, PhDc, Chien-Jen Chen, MDa, Cheng-Hsu Yang, MDa, Yen-Hsun Chen, MDa, and Morgan Fu, MDa The radiofrequency Maze procedure can effectively restore sinus rhythm in most patients with atrial fibrillation (AF) and mitral valve disease. AF after cardiac surgery is associated with increased morbidity and mortality. However, clinical determinants of long-term postoperative AF after the radiofrequency Maze procedure and concomitant mitral valve surgery are poorly defined. This study comprised 99 consecutive patients with persistent AF and mitral valve disease who underwent radiofrequency Maze procedures and concomitant mitral valvular operations. The predictive values of clinical variables for postoperative AF were examined. After a mean follow-up period of 46.1 ⴞ 24.6 months, 83 patients (83.8%) had sinus conversion after the Maze procedure, and 16 patients remained in persistent or paroxysmal AF. Multiple logistic regression analysis determined that predictors of sinus conversion were preoperative left atrial diameter (odds ratio [OR] 1.127 per 1-mm increment in left atrial diameter, 95% confidence interval [CI] 1.045 to 1.215, p <0.002) and the duration of AF (OR 1.022 per 1-month increment in duration of AF, 95% CI 1.009 to 1.035, p <0.001). Discriminant analysis showed that the sinus conversion rate was significantly lower in patients with preoperative left atrial diameters >56.8 mm (p <0.001) or AF duration >66 months (p <0.001) than in patients with preoperative left atrial diameters <56.8 mm or AF duration <66 months. In conclusion, the preoperative left atrial size and duration of AF are primary predictors of sinus conversion by the radiofrequency Maze procedure for patients with persistent AF and mitral valve disease. © 2005 Elsevier Inc. All rights reserved. (Am J Cardiol 2005; 96:1553–1557)
Atrial fibrillation (AF) is the most frequent sustained cardiac arrhythmia in patients with mitral valve disease, contributing to increased risk for systemic embolization and mortality.1–3 The saline-irrigated radiofrequency Maze procedure successfully restores sinus rhythm and atrial transport function in ⬎80% of patients with persistent AF and mitral valve disease.4,5 The development of AF after cardiac surgery is associated with increased morbidity and mortality.6 However, clinical determinants of AF after the radiofrequency Maze procedure and concomitant mitral valve surgery are poorly defined. This study investigated the clinical determinants of long-term postoperative AF after radiofrequency Maze procedures for persistent AF in patients
a
Division of Cardiology, Department of Internal Medicine, and bDivision of Cardiovascular Surgery, Department of Surgery, Chang Gung Memorial Hospital; and cDepartment of Biological Sciences, National Sun Yat-Sen University, Kaohsiung, Taiwan, Republic of China. Manuscript received May 23, 2005; revised manuscript received and accepted July 7, 2005. *Corresponding author: Tel: 886-7-731-7123 ext. 8300; fax: 886-7732-2402. E-mail address: [email protected] (M.-C. Chen). 0002-9149/05/$ – see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2005.07.063
with mitral valve disease who underwent concomitant mitral valvular operations. Methods Patient population: A total of 99 consecutive patients with persistent AF and mitral valve disease who underwent radiofrequency Maze procedures and concomitant valvular operations composed the study population. There were 44 men and 55 women, with a mean ⫾ SD age of 51 ⫾ 11 years (range 21 to 77). All patients were in persistent AF before surgery, which was defined as documented AF lasting for ⬎4 weeks without spontaneous conversion. All patients had mitral valve disease: 19 patients had mitral stenosis, 38 patients had mitral regurgitation, and 42 patients had combined mitral stenosis and mitral regurgitation. Twenty patients had concomitant aortic valve disease, and 72 also had tricuspid valve disease. Preoperative left atrial dimensions ranged from 31 to 85 mm (mean 53.4 ⫾ 9.5). No patients received any class I or III antiarrhythmic agents at the time of the study. The radiofrequency Maze procedure was performed simultaneously with the following procedures: mitral valve replacement (mechanical valve, n ⫽ 46; www.AJConline.org
1554
The American Journal of Cardiology (www.AJConline.org)
nus rhythm. No other patients received any class I or III antiarrhythmic agents postoperatively.
Figure 1. The radiofrequency Maze procedure. Left, epicardial view of the atria. Right, endocardial view of the atria. Circle, cryoablation; solid line with cross bars, incision and suturing; dotted line, radiofrequency ablation. FO ⫽ fossa ovalis; IVC ⫽ inferior vena cava; LAA ⫽ left atrial appendage; MV ⫽ mitral valve; PV ⫽ pulmonary vein; RAA ⫽ right atrial appendage; SVC ⫽ superior vena cava; TV ⫽ tricuspid valve.
tissue valve, n ⫽ 10), mitral valve repair (n ⫽ 42), the Ross II operation (n ⫽ 1), aortic valve replacement (mechanical valve, n ⫽ 15; tissue valve, n ⫽ 1), aortic valve repair (n ⫽ 4), tricuspid tissue valve replacement (mechanical valve, n ⫽ 1; tissue valve, n ⫽ 1), tricuspid annuloplasty (n ⫽ 70), and coronary artery bypass grafting (n ⫽ 2). A total of 47 of these patients (47.5%) underwent mechanical valve replacement. Informed consent was obtained from each patient. Echocardiography: Transthoracic echocardiographic examinations were performed in all of the patients with a 2.5-MHz transducer attached to a commercially available Doppler echocardiography machine (Sonos 5500, HewlettPackard Company, Palo Alto, California) before the radiofrequency Maze procedure in all of the patients. M-mode measurements were performed in accordance with the recommendations of the American Society of Echocardiography. Left and right atrial areas were measured by planimetry in a 4-chamber view, and maximum areas were the average of measurements (at the end of the T wave on the electrocardiogram) over 5 beats. Radiofrequency Maze procedure: The Maze procedure was conducted with handmade saline-irrigated radiofrequency catheter ablation guided by direct visualization. This technique was described in detail by Chen et al4 and Sie et al5 and has been demonstrated to produce transmural and continuous lesions. The radiofrequency ablation segment of the surgical operation was performed in accordance with the procedure developed by Chen et al4 (Figure 1). Postoperative management: Each patient was administered heparin intravenously. Cardiac rhythm was continuously monitored after surgery until a stable rhythm returned. Three patients received quinidine for 5 to 7 days postoperatively at the surgeon’s discretion; however, no patients regained si-
Electrocardiography and medications at follow-up: After hospital discharge, patients received monthly follow-up for medication adjustment and cardiac rhythm evaluation with 12-lead surface electrocardiograms. Anticoagulation therapy with warfarin was administered to all patients and was discontinued only when patients achieved stable sinus rhythm and atrial contraction documented by Doppler echocardiography after reparative surgery or tissue valve replacement. In contrast to other reports, no class I or III antiarrhythmic agents were prescribed to patients at follow-up. This decision excluded the possibility of late success, which would be secondary to the combined effect of surgical and pharmacologic sinus conversion. Statistical analysis: Unless otherwise specified, data are presented as means ⫾ SDs or percentages. Categorical variables between the sinus conversion group and the AF group (the failure group) were compared using the chi-square test or Fisher’s exact test (2 tailed). Continuous variables in the sinus conversion group and the AF group were compared using Wilcoxon’s rank-sum test. The significance of multiple variables found significant on univariate analysis was determined using multiple logistic regression analysis. Discriminant analysis was performed to determine the best cut-off value for significant preoperative variables that predicts the sinus conversion by the radiofrequency Maze procedure. The different cut-off values for the preoperative variables were assessed for their ability to discriminate between sinus converters and nonconverters. Sensitivity was defined as the proportion of patients not in sinus rhythm whose predictor values were larger than the cut-off values, which had the highest sensitivity and specificity in discriminating between sinus converters and nonconverters. Specificity was defined as the proportion of patients in sinus rhythm whose predictor values were less than the cut-off values. Positive predictive value was defined as the proportion of patients with predictor values larger than the cut-off values who did not regain sinus rhythm. Negative predictive value was defined as the proportion of patients with predictor values less than the cut-off values who regained sinus rhythm. The end point was persistent AF after surgery, and data were censored at the time of AF during follow-up. Event rates of AF during follow-up between patients dichotomized into 2 groups according to the predictor cut-off values determined by the discriminant analysis were calculated according to the Kaplan-Meier method and compared by means of the log-rank test. Statistical analysis was performed using a statistical software program (SAS for Windows version 8.02, SAS Institute Inc., Cary, North Carolina). All p values were 2 sided, and values ⬍0.05 were considered to indicate statistical significance.
Valvular Heart Disease/Maze Procedure for Atrial Fibrillation
1555
Table 1 Baseline clinical characteristics of patients with persistent AF Characteristic
Sinus Converters (n ⫽ 83)
Sinus Nonconverters (n ⫽ 16)
Age (yrs) Men Duration of AF (mo) Previous embolism New York Heart Association classification II III IV Diabetes mellitus Hypertension Coronary artery disease Hyperthyroidism Rheumatic heart disease Digitalis Beta blockade Calcium blockade Angiotensin-converting enzyme inhibitor Aortic valve disease Tricuspid valve disease Left atrial diameter (mm) Left atrial area (cm2) Right atrial area (cm2) Left ventricular end-diastolic diameter (mm) Left ventricular end-systolic diameter (mm) Ejection fraction (%)
51.8 ⫾ 11.7 38 (46%) 36 ⫾ 45 10 (12%) 83 (100%) 19 (23%) 54 (65%) 10 (12%) 11 (13%) 11 (13%) 4 (5%) 3 (4%) 52 (63%) 79 (95%) 13 (16%) 11 (13%) 49 (59%) 18 (22%) 58 (70%) 51.7 ⫾ 8.5 44.2 ⫾ 15.4 24.7 ⫾ 8.6 54.6 ⫾ 9.8 35.5 ⫾ 8.5 62.1 ⫾ 12.3
48.2 ⫾ 8.2 6 (38%) 96 ⫾ 51 0 (0%) 16 (100%) 1 (6%) 13 (81%) 2 (13%) 2 (13%) 3 (19%) 0 (0%) 0 (0%) 14 (88%) 16 (100%) 2 (13%) 4 (25%) 8 (50%) 2 (13%) 14 (88%) 61.9 ⫾ 10.2 66.9 ⫾ 28.8 26.6 ⫾ 8.8 54.4 ⫾ 9.2 36.1 ⫾ 7.9 60.9 ⫾ 10.8
Results Sinus conversion by the radiofrequency Maze procedure: All patients received follow-up until January 2005 or the time of death (mean duration 46.1 ⫾ 24.6 months, up to 91 months). Radiofrequency Maze procedure success was defined as persistent sinus rhythm achieved without antiarrhythmic drug therapy or electrical cardioversion after 2-month follow-up. Eighty-three patients (83.8%) had persistent sinus conversion after the radiofrequency Maze procedure (group 1), and 16 patients (persistent AF, 11 patients; paroxysmal AF, 5 patients) did not regain sinus rhythm (group 2). There was no atrial flutter by electrocardiographic criteria during follow-up in this study. Table 1 lists the clinical characteristics of the study patients. Predictors of sinus conversion by the radiofrequency Maze procedure: Table 1 lists the baseline characteristics of patients with and without sinus conversion. The duration of documented AF in group 1 patients was significantly shorter than that in group 2 patients. The preoperative left atrial area and left atrial diameter were significantly smaller in group 1 patients than in group 2 patients. There was no significant difference between the 2 groups for the following characteristics: age, gender, preoperative New York Heart Association functional class, the use of angiotensinconverting enzyme inhibitors, digitalis,  blockade and Ca channel blockade, the presence of coronary artery disease, the presence of rheumatic heart disease, hypertension, hyperthyroidism, diabetes mellitus, associated aortic valve dis-
p Value 0.169 0.5415 ⬍0.0001 0.3586 0.3656
1.00 0.6942 1.00 1.00 0.0535 1.00 1.00 0.2569 0.5031 0.5144 0.2218 ⬍0.001 ⬍0.001 0.248 0.805 0.794 0.524
ease, preoperative left ventricular size, and the preoperative left ventricular ejection fraction. There was no significant difference in terms of mitral valve repair (45.8% vs 25.0%, respectively, p ⬎0.1) between group 1 and group 2 patients. There was no significant difference in terms of tricuspid valve repair (67.5% vs 87.5%, respectively, p ⬎0.2) between group 1 and group 2 patients. After adjustments for age, gender, the presence of rheumatic heart disease, and the type of surgery in multiple logistic regression analysis, independent predictors of sinus conversion by radiofrequency Maze procedure were preoperative left atrial diameter, with an odds ratio (OR) for persistent AF after surgery of 1.127 for each 1-mm increment in left atrial diameter (95% confidence interval [CI] 1.045 to 1.215, p ⬍0.002), and AF duration, with an OR for persistent AF after surgery of 1.022 for each 1-month increment in AF duration (95% CI 1.009 to 1.035, p ⬍0.001). Preoperative left atrial size predicting the success of sinus conversion by the radiofrequency Maze procedure: Discriminant analysis was performed to discriminate the preoperative left atrial diameters in predicting sinus rhythm restoration by the radiofrequency Maze procedure. The sensitivity and specificity of the cut-off value of 56.8 mm were 68.8% and 77.1%, respectively. The positive and negative predictive values of the cut-off value of 56.8 mm were 36.7% and 92.8%, respectively. The sinus conversion rate was significantly lower in patients with preoperative left atrial diameters ⬎56.8 mm than in patients with preop-
1556
The American Journal of Cardiology (www.AJConline.org)
Figure 2. Kaplan-Meier estimates of the mean rates of freedom from AF in patients with persistent AF who underwent the radiofrequency Maze procedure, according to preoperative left atrial (LA) diameter. Note that the incidence of AF was significantly greater in patients with preoperative LA diameters ⬎56.8 mm than in patients with preoperative left atrial diameters ⬍56.8 mm.
Figure 3. Kaplan-Meier estimates of the mean rates of freedom from AF in patients with persistent AF who underwent the radiofrequency Maze procedure, according to the preoperative duration of AF. Note that the incidence of AF was significantly greater in patients with duration of AF ⬎66 months than in patients with duration of AF ⬍66 months.
erative left atrial diameters ⬍56.8 mm (OR 7.41, 95% CI 2.29 to 23.98, p ⬍0.001). The incidence of AF at 3 and 5 years was significantly greater in patients with left atrial diameters ⬎56.8 mm than in those with left atrial diameters ⬍56.8 mm (16.4% vs 2.9% and 38.7% vs 11.5%, respectively, p ⬍0.001; Figure 2).
renal failure 35 months after surgery, 1 patient died of multiorgan failure secondary to an herbal drug overdose 16 months after surgery, and 1 patient died of sudden death probably related to coronary artery restenosis after percutaneous coronary intervention at 19-month follow-up.
Duration of AF predicting the success of sinus conversion by the radiofrequency Maze procedure: Discriminant analysis was performed to discriminate the duration of AF in predicting sinus rhythm restoration by the radiofrequency Maze procedure. The sensitivity and specificity of the cut-off value of 66 months were 75.0% and 75.9%, respectively. The positive and negative predictive values of the cut-off value of 66 months were 37.5% and 94.0%, respectively. The sinus conversion rate was significantly lower in patients with AF duration ⬎66 months than in patients with AF duration ⬍66 months (OR 9.45, 95% CI 2.74 to 32.60, p ⬍0.001). The incidence of AF at 3 and 5 years was significantly greater in patients with AF duration ⬎66 months than in those with AF duration ⬍66 months (22.6% vs 0% and 37.9% vs 12.5%, respectively, p ⬍0.0001; Figure 3).
Discussion
Perioperative morbidity and mortality: There were no surgical deaths, delayed cardiac tamponades, or stroke complications in this study. No patient required postoperative pacemaker placement. There were 4 perioperative complications: 2 patients had postoperative small Q-wave inferior wall myocardial infarctions, 1 had reversible renal failure, and 1 had postoperative pneumonia complicated by respiratory failure. Mortality in long-term follow-up: There were 6 late deaths: 1 patient died of staphylococcus sepsis 54 months after surgery, 1 patient died of acute necrotizing pancreatitis 13 months after surgery, 1 patient died of sudden death 63 months after surgery, 1 patient died of pneumonia and acute
This study examined the clinical determinants of sinus conversion by the radiofrequency Maze procedure for patients with persistent AF and mitral valve disease. Several important conclusions are drawn from this study. First, preoperative left atrial diameter is an independent predictor of sinus conversion by the radiofrequency Maze procedure: every 1-mm increase in left atrial diameter corresponds to a 12.7% risk for persistent AF after surgery. The risk for persistent AF after surgery is 7.41-fold when the preoperative left atrial diameter is ⬎56.8 mm. Second, the duration of AF before surgery is an independent predictor of sinus conversion by the radiofrequency Maze procedure: every 1-month increase in the duration of AF corresponds to a 2.2% risk for persistent AF after surgery. The risk for persistent AF after surgery is 9.45-fold when the preoperative duration of AF is ⬎66 months. This study therefore identified the predictive values of left atrial enlargement and long duration of AF in patients with mitral valve disease and persistent AF. This study clearly demonstrated that the probability of AF after surgery in patients with preoperative left atrial diameters ⬎56.8 mm was 36.7%, compared with only 7.2% in patients with preoperative left atrial diameters ⬍56.8 mm. The absence of AF after surgery is reassuring when the preoperative left atrial diameter is ⬍56.8 mm; 92.8% of these patients remained free of AF after a mean follow-up period of 46.1 ⫾ 24.6 months. The role of left atrial enlargement as a harbinger of subsequent AF has been demonstrated in patients without valve disease and patients with
Valvular Heart Disease/Maze Procedure for Atrial Fibrillation
mitral valve disease.3,7 Atrial size as an important factor in the development of AF has been investigated in human and animal studies.8,9 This study clearly demonstrated that the probability of AF after surgery in patients with AF duration ⬎66 months was 37.5%, compared with only 6.0% in patients with AF duration ⬍66 months. The absence of AF after surgery is reassuring when the preoperative duration of AF is ⬍66 months; 94.0% of patients remained free of AF after a mean follow-up period of 46.1 ⫾ 24.6 months. The Kaplan-Meier curves for duration of AF separated very early (Figure 3), indicating that patients with longer preoperative AF duration should revert to AF or continue to be in AF earlier than patients with shorter preoperative AF duration. In this study, patients were followed monthly for cardiac rhythm evaluation by 12-lead surface electrocardiograms after hospital discharge. Therefore, asymptomatic episodes of paroxysmal AF may not have been detected. However, it is currently impossible to constantly monitor for “silent” AF occurrences. Hence, it is critical to define the incidence of detectable AF, the only form of this arrhythmia that currently may lead to therapeutic intervention. This study suggests that left atrial size and the duration of AF should be integrated into the decision-making process for the treatment of patients with significant mitral valve disease and persistent AF and that surgery should be considered early in the course of the disease to improve sinus conversion rates after radiofrequency Maze procedure.
1557
1. Chesebro JH, Fuster V, Halperin JL. Atrial fibrillation—risk marker for stroke. N Engl J Med 1990;323:1556 –1558. 2. Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham heart study. Circulation 1998;98:946 –952. 3. Grigioni F, Avierinos JF, Ling LH, Scott CG, Bailey KR, Tajik AJ, Frye RL, Enriquez-Sarano M. Atrial fibrillation complicating the course of degenerative mitral regurgitation: determinants and long-term outcome. J Am Coll Cardiol 2002;40:84 –92. 4. Chen MC, Guo BF, Chang JP, Yeh KH, Fu M. Radiofrequency and cryoablation of atrial fibrillation in patients undergoing valvular operations. Ann Thorac Surg 1998;65:1666 –1672. 5. Sie HT, Beukema WP, Elvan A, Ramdat Misier AR. Long-term results of irrigated radiofrequency modified maze procedure in 200 patients with concomitant cardiac surgery: six years experience. Ann Thorac Surg 2004;77:512–516. 6. Almassi GH, Schowalter T, Nicolosi AC, Aggarwal A, Moritz TE, Henderson WG, Tarazi R, Shroyer AL, Sethi GK, Grover FL. Atrial fibrillation after cardiac surgery: a major morbid event? Ann Surg 1997;226:501–511. 7. Tsang TS, Barnes ME, Bailey KR, Leibson CL, Montgomery SC, Takemoto Y, Diamond PM, Marra MA, Gersh BJ, Wiebers DO, et al. Left atrial volume: important risk marker of incident atrial fibrillation in 1655 older men and women. Mayo Clin Proc 2001;76:467– 475. 8. Henry WL, Morganroth J, Pearlman AS, Clark CE, Redwood DR, Itscoitz SB, Epstein SE. Relation between echocardiographically determined left atrial size and atrial fibrillation. Circulation 1976;53:273– 279. 9. Rensma PL, Allessie MA, Lammers WJ, Bonke FI, Schalij MJ. Length of excitation wave and susceptibility to reentrant atrial arrhythmias in normal conscious dogs. Circ Res 1988;62:395– 410.
Atrial fibrillation (AF) is the most frequent sustained cardiac arrhythmia in patients with mitral valve disease, contributing to increased risk for systemic embolization and mortality.1–3 The saline-irrigated radiofrequency Maze procedure successfully restores sinus rhythm and atrial transport function in ⬎80% of patients with persistent AF and mitral valve disease.4,5 The development of AF after cardiac surgery is associated with increased morbidity and mortality.6 However, clinical determinants of AF after the radiofrequency Maze procedure and concomitant mitral valve surgery are poorly defined. This study investigated the clinical determinants of long-term postoperative AF after radiofrequency Maze procedures for persistent AF in patients
a
Division of Cardiology, Department of Internal Medicine, and bDivision of Cardiovascular Surgery, Department of Surgery, Chang Gung Memorial Hospital; and cDepartment of Biological Sciences, National Sun Yat-Sen University, Kaohsiung, Taiwan, Republic of China. Manuscript received May 23, 2005; revised manuscript received and accepted July 7, 2005. *Corresponding author: Tel: 886-7-731-7123 ext. 8300; fax: 886-7732-2402. E-mail address: [email protected] (M.-C. Chen). 0002-9149/05/$ – see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2005.07.063
with mitral valve disease who underwent concomitant mitral valvular operations. Methods Patient population: A total of 99 consecutive patients with persistent AF and mitral valve disease who underwent radiofrequency Maze procedures and concomitant valvular operations composed the study population. There were 44 men and 55 women, with a mean ⫾ SD age of 51 ⫾ 11 years (range 21 to 77). All patients were in persistent AF before surgery, which was defined as documented AF lasting for ⬎4 weeks without spontaneous conversion. All patients had mitral valve disease: 19 patients had mitral stenosis, 38 patients had mitral regurgitation, and 42 patients had combined mitral stenosis and mitral regurgitation. Twenty patients had concomitant aortic valve disease, and 72 also had tricuspid valve disease. Preoperative left atrial dimensions ranged from 31 to 85 mm (mean 53.4 ⫾ 9.5). No patients received any class I or III antiarrhythmic agents at the time of the study. The radiofrequency Maze procedure was performed simultaneously with the following procedures: mitral valve replacement (mechanical valve, n ⫽ 46; www.AJConline.org
1554
The American Journal of Cardiology (www.AJConline.org)
nus rhythm. No other patients received any class I or III antiarrhythmic agents postoperatively.
Figure 1. The radiofrequency Maze procedure. Left, epicardial view of the atria. Right, endocardial view of the atria. Circle, cryoablation; solid line with cross bars, incision and suturing; dotted line, radiofrequency ablation. FO ⫽ fossa ovalis; IVC ⫽ inferior vena cava; LAA ⫽ left atrial appendage; MV ⫽ mitral valve; PV ⫽ pulmonary vein; RAA ⫽ right atrial appendage; SVC ⫽ superior vena cava; TV ⫽ tricuspid valve.
tissue valve, n ⫽ 10), mitral valve repair (n ⫽ 42), the Ross II operation (n ⫽ 1), aortic valve replacement (mechanical valve, n ⫽ 15; tissue valve, n ⫽ 1), aortic valve repair (n ⫽ 4), tricuspid tissue valve replacement (mechanical valve, n ⫽ 1; tissue valve, n ⫽ 1), tricuspid annuloplasty (n ⫽ 70), and coronary artery bypass grafting (n ⫽ 2). A total of 47 of these patients (47.5%) underwent mechanical valve replacement. Informed consent was obtained from each patient. Echocardiography: Transthoracic echocardiographic examinations were performed in all of the patients with a 2.5-MHz transducer attached to a commercially available Doppler echocardiography machine (Sonos 5500, HewlettPackard Company, Palo Alto, California) before the radiofrequency Maze procedure in all of the patients. M-mode measurements were performed in accordance with the recommendations of the American Society of Echocardiography. Left and right atrial areas were measured by planimetry in a 4-chamber view, and maximum areas were the average of measurements (at the end of the T wave on the electrocardiogram) over 5 beats. Radiofrequency Maze procedure: The Maze procedure was conducted with handmade saline-irrigated radiofrequency catheter ablation guided by direct visualization. This technique was described in detail by Chen et al4 and Sie et al5 and has been demonstrated to produce transmural and continuous lesions. The radiofrequency ablation segment of the surgical operation was performed in accordance with the procedure developed by Chen et al4 (Figure 1). Postoperative management: Each patient was administered heparin intravenously. Cardiac rhythm was continuously monitored after surgery until a stable rhythm returned. Three patients received quinidine for 5 to 7 days postoperatively at the surgeon’s discretion; however, no patients regained si-
Electrocardiography and medications at follow-up: After hospital discharge, patients received monthly follow-up for medication adjustment and cardiac rhythm evaluation with 12-lead surface electrocardiograms. Anticoagulation therapy with warfarin was administered to all patients and was discontinued only when patients achieved stable sinus rhythm and atrial contraction documented by Doppler echocardiography after reparative surgery or tissue valve replacement. In contrast to other reports, no class I or III antiarrhythmic agents were prescribed to patients at follow-up. This decision excluded the possibility of late success, which would be secondary to the combined effect of surgical and pharmacologic sinus conversion. Statistical analysis: Unless otherwise specified, data are presented as means ⫾ SDs or percentages. Categorical variables between the sinus conversion group and the AF group (the failure group) were compared using the chi-square test or Fisher’s exact test (2 tailed). Continuous variables in the sinus conversion group and the AF group were compared using Wilcoxon’s rank-sum test. The significance of multiple variables found significant on univariate analysis was determined using multiple logistic regression analysis. Discriminant analysis was performed to determine the best cut-off value for significant preoperative variables that predicts the sinus conversion by the radiofrequency Maze procedure. The different cut-off values for the preoperative variables were assessed for their ability to discriminate between sinus converters and nonconverters. Sensitivity was defined as the proportion of patients not in sinus rhythm whose predictor values were larger than the cut-off values, which had the highest sensitivity and specificity in discriminating between sinus converters and nonconverters. Specificity was defined as the proportion of patients in sinus rhythm whose predictor values were less than the cut-off values. Positive predictive value was defined as the proportion of patients with predictor values larger than the cut-off values who did not regain sinus rhythm. Negative predictive value was defined as the proportion of patients with predictor values less than the cut-off values who regained sinus rhythm. The end point was persistent AF after surgery, and data were censored at the time of AF during follow-up. Event rates of AF during follow-up between patients dichotomized into 2 groups according to the predictor cut-off values determined by the discriminant analysis were calculated according to the Kaplan-Meier method and compared by means of the log-rank test. Statistical analysis was performed using a statistical software program (SAS for Windows version 8.02, SAS Institute Inc., Cary, North Carolina). All p values were 2 sided, and values ⬍0.05 were considered to indicate statistical significance.
Valvular Heart Disease/Maze Procedure for Atrial Fibrillation
1555
Table 1 Baseline clinical characteristics of patients with persistent AF Characteristic
Sinus Converters (n ⫽ 83)
Sinus Nonconverters (n ⫽ 16)
Age (yrs) Men Duration of AF (mo) Previous embolism New York Heart Association classification II III IV Diabetes mellitus Hypertension Coronary artery disease Hyperthyroidism Rheumatic heart disease Digitalis Beta blockade Calcium blockade Angiotensin-converting enzyme inhibitor Aortic valve disease Tricuspid valve disease Left atrial diameter (mm) Left atrial area (cm2) Right atrial area (cm2) Left ventricular end-diastolic diameter (mm) Left ventricular end-systolic diameter (mm) Ejection fraction (%)
51.8 ⫾ 11.7 38 (46%) 36 ⫾ 45 10 (12%) 83 (100%) 19 (23%) 54 (65%) 10 (12%) 11 (13%) 11 (13%) 4 (5%) 3 (4%) 52 (63%) 79 (95%) 13 (16%) 11 (13%) 49 (59%) 18 (22%) 58 (70%) 51.7 ⫾ 8.5 44.2 ⫾ 15.4 24.7 ⫾ 8.6 54.6 ⫾ 9.8 35.5 ⫾ 8.5 62.1 ⫾ 12.3
48.2 ⫾ 8.2 6 (38%) 96 ⫾ 51 0 (0%) 16 (100%) 1 (6%) 13 (81%) 2 (13%) 2 (13%) 3 (19%) 0 (0%) 0 (0%) 14 (88%) 16 (100%) 2 (13%) 4 (25%) 8 (50%) 2 (13%) 14 (88%) 61.9 ⫾ 10.2 66.9 ⫾ 28.8 26.6 ⫾ 8.8 54.4 ⫾ 9.2 36.1 ⫾ 7.9 60.9 ⫾ 10.8
Results Sinus conversion by the radiofrequency Maze procedure: All patients received follow-up until January 2005 or the time of death (mean duration 46.1 ⫾ 24.6 months, up to 91 months). Radiofrequency Maze procedure success was defined as persistent sinus rhythm achieved without antiarrhythmic drug therapy or electrical cardioversion after 2-month follow-up. Eighty-three patients (83.8%) had persistent sinus conversion after the radiofrequency Maze procedure (group 1), and 16 patients (persistent AF, 11 patients; paroxysmal AF, 5 patients) did not regain sinus rhythm (group 2). There was no atrial flutter by electrocardiographic criteria during follow-up in this study. Table 1 lists the clinical characteristics of the study patients. Predictors of sinus conversion by the radiofrequency Maze procedure: Table 1 lists the baseline characteristics of patients with and without sinus conversion. The duration of documented AF in group 1 patients was significantly shorter than that in group 2 patients. The preoperative left atrial area and left atrial diameter were significantly smaller in group 1 patients than in group 2 patients. There was no significant difference between the 2 groups for the following characteristics: age, gender, preoperative New York Heart Association functional class, the use of angiotensinconverting enzyme inhibitors, digitalis,  blockade and Ca channel blockade, the presence of coronary artery disease, the presence of rheumatic heart disease, hypertension, hyperthyroidism, diabetes mellitus, associated aortic valve dis-
p Value 0.169 0.5415 ⬍0.0001 0.3586 0.3656
1.00 0.6942 1.00 1.00 0.0535 1.00 1.00 0.2569 0.5031 0.5144 0.2218 ⬍0.001 ⬍0.001 0.248 0.805 0.794 0.524
ease, preoperative left ventricular size, and the preoperative left ventricular ejection fraction. There was no significant difference in terms of mitral valve repair (45.8% vs 25.0%, respectively, p ⬎0.1) between group 1 and group 2 patients. There was no significant difference in terms of tricuspid valve repair (67.5% vs 87.5%, respectively, p ⬎0.2) between group 1 and group 2 patients. After adjustments for age, gender, the presence of rheumatic heart disease, and the type of surgery in multiple logistic regression analysis, independent predictors of sinus conversion by radiofrequency Maze procedure were preoperative left atrial diameter, with an odds ratio (OR) for persistent AF after surgery of 1.127 for each 1-mm increment in left atrial diameter (95% confidence interval [CI] 1.045 to 1.215, p ⬍0.002), and AF duration, with an OR for persistent AF after surgery of 1.022 for each 1-month increment in AF duration (95% CI 1.009 to 1.035, p ⬍0.001). Preoperative left atrial size predicting the success of sinus conversion by the radiofrequency Maze procedure: Discriminant analysis was performed to discriminate the preoperative left atrial diameters in predicting sinus rhythm restoration by the radiofrequency Maze procedure. The sensitivity and specificity of the cut-off value of 56.8 mm were 68.8% and 77.1%, respectively. The positive and negative predictive values of the cut-off value of 56.8 mm were 36.7% and 92.8%, respectively. The sinus conversion rate was significantly lower in patients with preoperative left atrial diameters ⬎56.8 mm than in patients with preop-
1556
The American Journal of Cardiology (www.AJConline.org)
Figure 2. Kaplan-Meier estimates of the mean rates of freedom from AF in patients with persistent AF who underwent the radiofrequency Maze procedure, according to preoperative left atrial (LA) diameter. Note that the incidence of AF was significantly greater in patients with preoperative LA diameters ⬎56.8 mm than in patients with preoperative left atrial diameters ⬍56.8 mm.
Figure 3. Kaplan-Meier estimates of the mean rates of freedom from AF in patients with persistent AF who underwent the radiofrequency Maze procedure, according to the preoperative duration of AF. Note that the incidence of AF was significantly greater in patients with duration of AF ⬎66 months than in patients with duration of AF ⬍66 months.
erative left atrial diameters ⬍56.8 mm (OR 7.41, 95% CI 2.29 to 23.98, p ⬍0.001). The incidence of AF at 3 and 5 years was significantly greater in patients with left atrial diameters ⬎56.8 mm than in those with left atrial diameters ⬍56.8 mm (16.4% vs 2.9% and 38.7% vs 11.5%, respectively, p ⬍0.001; Figure 2).
renal failure 35 months after surgery, 1 patient died of multiorgan failure secondary to an herbal drug overdose 16 months after surgery, and 1 patient died of sudden death probably related to coronary artery restenosis after percutaneous coronary intervention at 19-month follow-up.
Duration of AF predicting the success of sinus conversion by the radiofrequency Maze procedure: Discriminant analysis was performed to discriminate the duration of AF in predicting sinus rhythm restoration by the radiofrequency Maze procedure. The sensitivity and specificity of the cut-off value of 66 months were 75.0% and 75.9%, respectively. The positive and negative predictive values of the cut-off value of 66 months were 37.5% and 94.0%, respectively. The sinus conversion rate was significantly lower in patients with AF duration ⬎66 months than in patients with AF duration ⬍66 months (OR 9.45, 95% CI 2.74 to 32.60, p ⬍0.001). The incidence of AF at 3 and 5 years was significantly greater in patients with AF duration ⬎66 months than in those with AF duration ⬍66 months (22.6% vs 0% and 37.9% vs 12.5%, respectively, p ⬍0.0001; Figure 3).
Discussion
Perioperative morbidity and mortality: There were no surgical deaths, delayed cardiac tamponades, or stroke complications in this study. No patient required postoperative pacemaker placement. There were 4 perioperative complications: 2 patients had postoperative small Q-wave inferior wall myocardial infarctions, 1 had reversible renal failure, and 1 had postoperative pneumonia complicated by respiratory failure. Mortality in long-term follow-up: There were 6 late deaths: 1 patient died of staphylococcus sepsis 54 months after surgery, 1 patient died of acute necrotizing pancreatitis 13 months after surgery, 1 patient died of sudden death 63 months after surgery, 1 patient died of pneumonia and acute
This study examined the clinical determinants of sinus conversion by the radiofrequency Maze procedure for patients with persistent AF and mitral valve disease. Several important conclusions are drawn from this study. First, preoperative left atrial diameter is an independent predictor of sinus conversion by the radiofrequency Maze procedure: every 1-mm increase in left atrial diameter corresponds to a 12.7% risk for persistent AF after surgery. The risk for persistent AF after surgery is 7.41-fold when the preoperative left atrial diameter is ⬎56.8 mm. Second, the duration of AF before surgery is an independent predictor of sinus conversion by the radiofrequency Maze procedure: every 1-month increase in the duration of AF corresponds to a 2.2% risk for persistent AF after surgery. The risk for persistent AF after surgery is 9.45-fold when the preoperative duration of AF is ⬎66 months. This study therefore identified the predictive values of left atrial enlargement and long duration of AF in patients with mitral valve disease and persistent AF. This study clearly demonstrated that the probability of AF after surgery in patients with preoperative left atrial diameters ⬎56.8 mm was 36.7%, compared with only 7.2% in patients with preoperative left atrial diameters ⬍56.8 mm. The absence of AF after surgery is reassuring when the preoperative left atrial diameter is ⬍56.8 mm; 92.8% of these patients remained free of AF after a mean follow-up period of 46.1 ⫾ 24.6 months. The role of left atrial enlargement as a harbinger of subsequent AF has been demonstrated in patients without valve disease and patients with
Valvular Heart Disease/Maze Procedure for Atrial Fibrillation
mitral valve disease.3,7 Atrial size as an important factor in the development of AF has been investigated in human and animal studies.8,9 This study clearly demonstrated that the probability of AF after surgery in patients with AF duration ⬎66 months was 37.5%, compared with only 6.0% in patients with AF duration ⬍66 months. The absence of AF after surgery is reassuring when the preoperative duration of AF is ⬍66 months; 94.0% of patients remained free of AF after a mean follow-up period of 46.1 ⫾ 24.6 months. The Kaplan-Meier curves for duration of AF separated very early (Figure 3), indicating that patients with longer preoperative AF duration should revert to AF or continue to be in AF earlier than patients with shorter preoperative AF duration. In this study, patients were followed monthly for cardiac rhythm evaluation by 12-lead surface electrocardiograms after hospital discharge. Therefore, asymptomatic episodes of paroxysmal AF may not have been detected. However, it is currently impossible to constantly monitor for “silent” AF occurrences. Hence, it is critical to define the incidence of detectable AF, the only form of this arrhythmia that currently may lead to therapeutic intervention. This study suggests that left atrial size and the duration of AF should be integrated into the decision-making process for the treatment of patients with significant mitral valve disease and persistent AF and that surgery should be considered early in the course of the disease to improve sinus conversion rates after radiofrequency Maze procedure.
1557
1. Chesebro JH, Fuster V, Halperin JL. Atrial fibrillation—risk marker for stroke. N Engl J Med 1990;323:1556 –1558. 2. Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham heart study. Circulation 1998;98:946 –952. 3. Grigioni F, Avierinos JF, Ling LH, Scott CG, Bailey KR, Tajik AJ, Frye RL, Enriquez-Sarano M. Atrial fibrillation complicating the course of degenerative mitral regurgitation: determinants and long-term outcome. J Am Coll Cardiol 2002;40:84 –92. 4. Chen MC, Guo BF, Chang JP, Yeh KH, Fu M. Radiofrequency and cryoablation of atrial fibrillation in patients undergoing valvular operations. Ann Thorac Surg 1998;65:1666 –1672. 5. Sie HT, Beukema WP, Elvan A, Ramdat Misier AR. Long-term results of irrigated radiofrequency modified maze procedure in 200 patients with concomitant cardiac surgery: six years experience. Ann Thorac Surg 2004;77:512–516. 6. Almassi GH, Schowalter T, Nicolosi AC, Aggarwal A, Moritz TE, Henderson WG, Tarazi R, Shroyer AL, Sethi GK, Grover FL. Atrial fibrillation after cardiac surgery: a major morbid event? Ann Surg 1997;226:501–511. 7. Tsang TS, Barnes ME, Bailey KR, Leibson CL, Montgomery SC, Takemoto Y, Diamond PM, Marra MA, Gersh BJ, Wiebers DO, et al. Left atrial volume: important risk marker of incident atrial fibrillation in 1655 older men and women. Mayo Clin Proc 2001;76:467– 475. 8. Henry WL, Morganroth J, Pearlman AS, Clark CE, Redwood DR, Itscoitz SB, Epstein SE. Relation between echocardiographically determined left atrial size and atrial fibrillation. Circulation 1976;53:273– 279. 9. Rensma PL, Allessie MA, Lammers WJ, Bonke FI, Schalij MJ. Length of excitation wave and susceptibility to reentrant atrial arrhythmias in normal conscious dogs. Circ Res 1988;62:395– 410.