- Email: [email protected]
Left Atrial Appendage Resection Versus Preservation During the Surgical Ablation of Atrial Fibrillation
Left Atrial Appendage Resection Versus Preservation During the Surgical Ablation of Atrial Fibrillation Chee-Hoon Lee, MD, Joon Bum Kim, MD, PhD, Sung-Ho Jung, MD, PhD, Suk Jung Choo, MD, PhD, Cheol Hyun Chung, MD, PhD, and Jae Won Lee, MD, PhD Department of Thoracic and Cardiovascular Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
Background. Left atrial appendage (LAA) resection during the Maze procedure may decrease thromboembolic risks, but its preservation may improve left atrial contractile function. This study compared the clinical effects of LAA resection and preservation after the Maze procedure. Methods. A retrospective review was made of 379 patients (mean age 53.3 ± 12.6 years, 244 females) who underwent the cryo-Maze procedure in conjunction with mitral surgery from 1999 to 2011. The LAA was resected in 187 patients (resection group) but preserved in 192 patients (preservation group). Outcomes were compared using a propensity score study design based on 20 baseline characteristics to obtain well-matched patient pairs. Results. Propensity score matching yielded 119 pairs of patients in whom there were no significant differences in baseline profiles between the two groups. During a mean follow-up of 3.1 ± 2.8 years, there were 16 deaths, 6 cases
of stroke, and 39 cases of atrial fibrillation recurrence. There were no significant differences in stroke-free survival (p [ 0.88) and freedom from AF while off antiarrhythmic drugs (p [ 0.46) between the two groups. On serial echocardiographic assessments, patients in the preservation group showed a higher transmitral A-wave velocity (peak atrial contraction wave velocity; p [ 0.47, 0.020, and 0.001 at 3, 6, and 12 months, respectively) and lower E/A ratio (peak early filling wave [E-wave] velocity / A-wave velocity; p [ 0.34, 0.065, and 0.001 at 3, 6, and 12 months, respectively) at each timepoint compared with the resection group. Conclusions. Preservation of the LAA during the Maze procedure resulted in similar clinical and rhythm outcomes, but LA contractile function superior to that of LAA resection.
T
involved in atrial natriuretic peptide secretion and plays an important role in neurohormonal regulation [7, 8]. In these regards, few clinical studies exist that compare LAA resection with LAA preservation in clinical outcomes. In this study, we sought to compare the outcomes between LAA preservation and resection in terms of clinical and rhythm outcomes as well as atrial transport functions.
he Maze procedure is widely accepted as a way to effectively restore sinus rhythm in patients with atrial fibrillation (AF) who are undergoing cardiac surgery for other indications. Since the introduction of the Maze procedure, efforts have been made to reduce the complexity of the procedure and procedural complications [1, 2]. Accordingly, various techniques, including the use of alternative energy sources for creation of ablation lesions and modified simpler lesion sets, were developed to replace the conventional complex cut-and-sew procedure. One of these modifications involves left atrial appendage (LAA) preservation during the Maze procedure. The LAA is reported to be a major source of thrombus formation and its embolization, and can contribute as an ectopic focus of AF trigger [3–5]. However, the LAA can contribute greatly to left atrial mechanical contraction after the Maze procedure [6]. Moreover, some researchers assert that the LAA is largely Accepted for publication July 22, 2013. Address correspondence to Dr Jae Won Lee, Department of Thoracic and Cardiovascular Surgery, Asan Medical Center, University of Ulsan Collage of Medicine, 88, Olympic-ro 43-gil, Songpa-gu, Seoul 138-736, South Korea; e-mail: [email protected].
Ó 2013 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2013;-:-–-) Ó 2013 by The Society of Thoracic Surgeons
Patients and Methods Study Population Between January 1999 and January 2011, a total of 525 patients underwent mitral valve (MV) operations and concomitant biatrial cryo-Maze procedures at the Asan Medical Center, Seoul, Korea. Patients were excluded if they underwent concomitant multivessel coronary artery bypass graft surgery (n ¼ 22) or aortic replacement (n ¼ 8), or had a preoperative LAA thrombus (n ¼ 116). In the end, 379 patients were enrolled. The decision to resect or preserve LAA was influenced by several demographic and procedural factors. For instance, patients with high thromboembolic risks such as old age, coexisting diabetes mellitus, and advanced degree of tricuspid regurgitation 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2013.07.073
2
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
or pulmonary hypertension were more likely to undergo LAA resection. Nevertheless, the decision was finally at the discretion of the attending surgeon referencing those baseline profiles. This study was approved by the Institutional Ethics Committee/Review Board at the Asan Medical Center. The requirement to obtain informed consent was waived by the Board owing to the retrospective nature of the study.
Surgical Techniques A median sternotomy approach (n ¼ 280, 73.9%), a minithoracotomy approach using the AESOP 3000 system (Computer Motion, Santa Barbara, CA [n ¼ 90, 23.7%]), or port-assisted minithoracotomy approach using the da Vinci robotic system (Intuitive Surgical, Irvine, CA [n ¼ 9, 2.4%]) were used. The biatrial cryo-Maze procedure was performed as previously described [2, 9]. After snaring down the superior and inferior vena cavae, oblique right atriotomy was made on beating heart. Then, the cavotricuspid isthmus isolation was achieved by two cryoablation lines, and a linear lesion from posterior end of atriotomy to superior vena cava was made (Fig 1A). After aortic cross-clamping, the left side procedure began with a longitudinal right-sided left atriotomy, which was performed endocardially before the MV procedure. Left atrium ablation consisted of a single box lesion for pulmonary vein isolation and two posterior linear lesions, one from the pulmonary vein isolation line to the LAA orifice and one from the pulmonary vein isolation line to the MV annulus. Additional epicardial coronary sinus ablation was performed at the opposite side of the linear lesion to MV annulus. The left atrium size was reduced by resection of redundant atrial tissue between inferior pulmonary vein and posterior MV annulus (Fig 1B). An argon-based flexible cryoablation system (SurgiFrost; Medtronic, Minneapolis, MN) was used for ablation. The surgeons attempted to remain as much subvalvular tissue as possible in a chordae-sparing manner during MV replacement.
Ann Thorac Surg 2013;-:-–-
Andover, MA). During the postoperative hospitalization period, typical 12-channel surface electrocardiography was performed every day. For patients who had normal sinus rhythm, Holter monitoring was performed for adjunctive evaluation. Follow-up echocardiography was performed at 1, 3, 6, and 12 months after surgery and electrocardiography was generally performed at 1, 3, 6, 12, 24, and 36 months after surgery. Early AF events were defined as any AF episodes, including atrial tachycardia and atrial flutter, less than 3 months after the operation.
Postoperative Management Patients who underwent valve repair or bioprosthetic valve implantation were routinely administered warfarin for 3 to 6 months postoperatively, with a target international normalized ratio of 1.5 to 2.5 at the discretion of the attending surgeon. Maintenance of anticoagulation therapy thereafter was determined according to the presence of thromboembolic risks and cardiac rhythm status in each patient. And then, to discontinue warfarin medication, the patient should have shown sinus rhythm at least for more than 2 consecutive visits and be free of symptoms. For patients with mechanical valve implantation, an international normalized ratio of 2.0 to 3.0 was the aim, regardless of cardiac rhythm status. Postoperative atrial tachyarrhythmias (AF, atrial flutter, or atrial tachycardia) were managed with class I or III antiarrhythmic drugs including amiodarone, sotalol, flecainide, and pilsicainide in an effort to restore sinus rhythm. For patients who achieved “AF elimination,” those drugs were withdrawn within 1 or 2 weeks. Patients who failed to achieve AF elimination despite antiarrhythmic medications for sufficient duration were taken off these drugs and switched to a “rate control strategy” involving digitalis, beta-blockers, or calcium-channel blockers to control ventricular rate in combination with anticoagulation therapy. Any symptoms suggestive of neurologic deficit were evaluated by neurologists with adequate imaging studies.
Echocardiography and Rhythm Follow-Up Preoperative transthoracic echocardiography and transesophageal echocardiography were performed in all patients within 2 months before surgery using a Hewlett-Packard Sonos 2500 or 5500 imaging system equipped with a 2.5-MHz transducer (Hewlett-Packard, Fig 1. Schema of surgical ablation of atrial fibrillation. (CS ¼ coronary sinus; IVC ¼ inferior vena cava; LAA ¼ left atrial appendage; MV ¼ mitral valve; PAs ¼ pulmonary veins; RAA ¼ right atrial appendage; SVC ¼ superior vena cava; TV ¼ tricuspid valve.)
Statistical Analysis Categorical variables are presented as frequencies and percentages, and continuous variables are expressed as a mean with standard deviation or a median with range. To reduce the effect of treatment selection bias and
Ann Thorac Surg 2013;-:-–-
3
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
Table 1. Baseline Characteristics of the Entire Cohort Characteristics Number of patients Age, years Male Diabetes mellitus Hypertension Chronic renal failure Congestive heart failure NYHA functional classification I II III IV History of stroke Prior cardiac surgery AF profiles Fine (<1 mm) AF wave AF type Paroxysmal Persistent Mitral diagnosis Rheumatic Degenerative Others Echocardiographic data Mitral regurgitation grade 0 1 2 3 4 Mitral valve area, cm2 Tricuspid regurgitation grade 0 1 2 3 4 Left ventricular ejection fraction, % Left atrial dimension, mm Transtricuspid peak pressure gradient, mm Hg Mitral valve operation Repair Replacement, mechanical Replacement, bioprosthetic Minimally invasive approach Concomitant cardiac surgery Aortic valve repair or replacement Tricuspid valve repair Coronary artery bypass a
p < 0.05.
Data are n (%) or mean SD. AF ¼ atrial fibrillation;
NYHA ¼ New York Heart Association.
Resection Group
Preservation Group
187 55.98 12.2 61 (32.6) 22 (11.8) 37 (19.8) 1 (0.5) 107 (57.2)
192 50.7 12.4 74 (38.5) 11 (5.7) 28 (14.5) 0 119 (62.0)
30 100 48 9 12 8
(16.0) (53.5) (25.7) (4.8) (6.4) (4.3)
30 86 65 11 10 14
(15.6) (44.8) (33.9) (5.7) (5.2) (7.3)
p Value <0.001a 0.23 0.037 0.18 0.31 0.35 0.29
0.62 0.21
79 (42.2)
75 (39.6)
0.53 0.20
11 (5.9) 176 (94.1)
18 (9.4) 174 (90.6)
119 (63.6) 64 (34.2) 4 (2.1)
121 (63.0) 58 (30.2) 13 (6.8)
22 (11.8) 25 (13.4) 18 (9.6) 20 (10.7) 102 (54.5) 2.62 1.37
20 (10.4) 20 (10.4) 14 (7.3) 32 (16.7) 106 (55.2) 2.62 1.33
0.42
16 (8.6) 29 (15.5) 41 (21.9) 44 (23.5) 57 (30.5) 56.3 7.1 60.6 10.5 38.2 13.2
25 (13.0) 56 (29.2) 39 (20.3) 35 (18.2) 37 (19.3) 56.1 9.0 59.7 10.1 34.9 10.1
0.003a
0.082
83 87 17 56
(44.4) (46.5) (9.1) (29.9)
39 (20.9) 122 (65.2) 10 (5.3)
105 85 2 43
(54.7) (44.3) (1.0) (22.4)
26 (13.5) 98 (51.0) 5 (2.6)
0.99
0.81 0.40 0.007a 0.094
0.10 0.076 0.007a 0.20
4
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
Ann Thorac Surg 2013;-:-–-
Table 2. Baseline Characteristics of Propensity Matched Patients Characteristics Number of patients Age, years Male Diabetes mellitus Hypertension Chronic renal failure Congestive heart failure NYHA functional classification I II III IV History of stroke Prior cardiac surgery AF profiles Fine (<1 mm) AF wave AF type Paroxysmal Persistent Mitral diagnosis Rheumatic Degenerative Others Echocardiographic data Mitral regurgitation grade 0 1 2 3 4 Mitral valve area, cm2 Mitral valve diagnosis Predominant mitral regurgitation Predominant mitral stenosis Mixed lesion Tricuspid regurgitation grade 0 1 2 3 4 Left ventricular ejection fraction, % Left atrial dimension, mm 50 mm, <60 mm 60 mm Transtricuspid peak pressure gradient, mm Hg Mitral valve operation Repair Replacement, mechanical Replacement, bioprosthetic
Resection Group
Preservation Group
119 53.4 12.5 45 (37.8) 9 (7.6) 22 (18.5) 0 (0) 66 (55.5)
119 54.1 11.9 49 (41.2) 9 (7.6) 23 (19.3) 0 (0) 64 (53.8)
22 61 31 5 6 8
(18.5) (51.3) (26.1) (4.2) (5.0) (6.7)
23 60 32 4 7 5
(19.3) (50.4) (26.9) (3.4) (5.9) (4.2)
71 (59.7)
79 (66.4)
11 (9.2) 108 (90.8)
11 (9.2) 108 (90.8)
73 (61.3) 42 (35.3) 4 (3.4)
66 (55.5) 46 (38.7) 7 (5.9)
15 (12.6) 13 (10.9) 11 (9.2) 14 (11.8) 66 (55.5) 2.68 1.36
12 (10.1) 13 (10.9) 8 (6.7) 17 (14.3) 69 (58.0) 2.84 1.34
76 (63.9) 35 (29.4) 8 (6.7)
82 (68.9) 24 (20.2) 13 (10.9)
13 (10.9) 25 (21.0) 25 (21.0) 26 (21.8) 30 (25.2) 56.5 7.1 59.0 10.9 46 (38.7) 51 (42.9) 35.7 12.4
13 (10.9) 26 (21.8) 27 (22.7) 24 (20.2) 29 (24.4) 56.9 8.7 59.7 9.9 47 (39.5) 50 (42.0) 36.3 10.0
58 (48.7) 60 (50.4) 1 (0.8)
64 (53.8) 53 (44.5) 2 (1.7)
p Value 0.57 0.70 >0.99 >0.99 >0.99 0.89 0.86
>0.99 0.55 0.78 >0.99
0.91
0.44
0.36 0.92
0.80
0.67 0.59 >0.99 >0.99 0.66 0.54
(Continued)
Ann Thorac Surg 2013;-:-–-
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
5
Table 2. Continued Characteristics
Resection Group
Preservation Group
22 (18.5) 68 (57.1) 3 (2.5)
20 (16.8) 65 (54.6) 5 (4.2)
Concomitant cardiac surgery Aortic valve repair or replacement Tricuspid valve repair Coronary artery bypass
p Value
0.85 0.79 0.73
Data are n (%) or mean SD. AF ¼ atrial fibrillation;
NYHA ¼ New York Heart Association.
potential confounding, we performed adjustments for the differences in the baseline characteristics by use of propensity score matching. The propensity scores were estimated with multiple logistic regression analysis. Prespecified covariates, listed in Table 1, were included for the calculation of propensity scores. The discrimination and calibration abilities of the propensity score model were assessed by means of C statistics and the HosmerLemeshow test. The model had a C statistic of 0.76 and a Hosmer-Lemeshow goodness-of-fit p value of 0.44, indicating that this model was well calibrated with reasonable discrimination. For the development of the propensity score-matched pairs (a 1:1 match), the greedy 5-to-1 digit matching algorithm was used. After propensity score matching, the data of the two groups were compared using the paired Student’s t test or the Wilcoxon signed rank test for continuous variables, and with McNemar’s test or the marginal homogeneity test for categorical variables. All reported p values were two-sided, and a value of p less than 0.05 was considered statistically significant. For statistical analysis, SAS software, version 9.1 (SAS Institute, Cary, NC), was used.
Results
valvuloplasty. Propensity score matching yielded 119 pairs of patients (resection group and preservation group). After propensity score matching, there were no significant intergroup differences in their baseline characteristics between the two groups (Table 2).
Early Outcomes Early postoperative complications are presented in Table 3. There was 1 early death (0.4%) in the LAA preservation group, of a patient who underwent mechanical MV implantation and myectomy of the left ventricular outflow tract. The patient died of panperitonitis followed by septic shock during mechanical circulatory support 13 days after surgery. There were 2 early cerebrovascular accidents (CVA [0.8%]), 1 in each group. In the LAA preservation group, the CVA occurred in a patient who had undergone mitral valvuloplasty and tricuspid valvuloplasty by the minimally invasive approach. The other CVA occurred in a patient of the LAA resection group who had a preoperative history of CVA and presented with altered mental status and motor weakness immediately after surgery. There were no statistically significant differences in early complications and mortalities between the two groups.
Baseline Patient Profiles
Late Clinical Outcomes
The baseline characteristics of the patients are listed in Table 1. The patients who underwent LAA resection were older than those who underwent LAA preservation. The LAA-preserved patients had a higher grade of tricuspid regurgitation and more frequently underwent tricuspid
Clinical follow-up was complete in all patients with a mean follow-up duration of 62.6 44.0 months (median 60.5; range, 0.03 to 161.4). The rates of anticoagulation therapy in the late period (more than 6 months) was similar for both groups (74 of 119 in the resection group
Table 3. Early Operative Complications Variables Early mortality Early complications Early stroke Low cardiac output syndrome Reoperation due to bleeding Requirement for dialysis Permanent pacemaker insertion Mediastinitis Wound revision Pericardial effusion
Resection Group, n (%) 0 (0) 1 0 13 9 4 0 1 8
(0.8) (0) (10.9) (7.6) (3.4) (0) (0.8) (6.7)
Preservation Group, n (%)
p Value
1 (0.8)
>0.99
1 1 6 10 2 1 2 6
(0.8) (0.8) (5.0) (8.4) (1.7) (0.8) (1.7) (5.0)
>0.99 >0.99 0.17 >0.99 0.63 >0.99 >0.99 0.77
6
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
Ann Thorac Surg 2013;-:-–-
Fig 2. Kaplan-Meier curves for (A) overall survival rate and (B) stroke-free survival. Solid line indicates left atrial appendage (LAA); broken line indicates LAA resection.
versus 66 of 119 in the preservation group, p ¼ 0.29), for which the most frequent indication was the implantation of mechanical valves (n ¼ 60 and n ¼ 53 in the resection group and preservation group, respectively) followed by the recurrence of AF or high risks of thromboembolism (more than CHADS2 [acronym for congestive heart failure, hypertension, age 75, diabetes mellitus, stroke or transient ischemic attack history] score of 2; n ¼ 14 and n ¼ 13 in the resection group and preservation group, respectively). During follow-up, there were 16 late deaths, 6 in the resection group and 10 in the preservation group, with no intergroup difference. Three patients were affected by late thromboembolic events (1 in the resection group [CVA] and 2 in the preservation group [1 transient ischemic attack and 1 CVA], p ¼ 0.63). All of these patients were receiving warfarin anticoagulation therapy owing to the insertion of mechanical valve (n ¼ 1 in the resection group), recurrence of AF (n ¼ 1 in the preservation group who underwent MV repair), and history of previous CVA (n ¼ 1 in the preservation group who underwent MV repair). There were no significant differences in the incidence of late stroke (p ¼ 0.63), reoperation of previous surgery site (2 in the resection group and 3 in the preservation group; p > 0.99), anticoagulation-related hemorrhages (4 in the resection group only; p ¼ 0.13), and admission due to congestive heart failure (2 in each group; p > 0.99). There were no
significant differences in overall survival or CVA-free survival between the two groups (Fig 2).
Rhythm Outcomes Early and late AF events (more than 3 months after surgery) and latest follow-up rhythms are summarized in Table 4; there were no significant differences in the rhythm status. Rhythm follow-up over 3 months after surgery was possible in 224 patients (94.1%), with a median follow-up duration of 26.3 months (range, 0.03 to 139.4). There was no significant difference in freedom from AF recurrence off antiarrhythmic medication between the two groups (Fig 3). The sequential follow-up of rhythm status during the 3-month, 6-month, 1-year, and 2-year points after surgery are shown in Figure 4. There were no significant differences at these timepoints after surgery between the two groups. There were 65 patients (33 in the resection group and 32 in the preservation group) who underwent electrical cardioversion during hospitalization (p > 0.99). During follow-up, electrical cardioversion was performed for 5 patients in each group.
Atrial Transport Functions Figure 5 depicts echocardiographic parameters of the left atrium in propensity score matched patients at 1-, 3-,
Table 4. Early and Late Atrial Fibrillation Events and Latest Follow-Up Cardiac Rhythm Status Variables Early (3 months) AF events Late (>3 months) AF events Latest follow-up rhythm (>3 months) Normal sinus rhythm Atrial fibrillation/flutter/tachycardia PPMa Junctional rhythm a
Resection Group, n (%)
Preservation Group, n (%)
p Value
27 (22.7) 21 (17.6)
28 (23.5) 18 (15.1)
0.878 0.600 0.514
98 11 4 1
(82.4) (9.2) (3.4) (0.8)
Permanent pacemaker (PPM) rhythm or sick sinus syndrome or complete atrioventricular block.
AF ¼ atrial fibrillation.
104 6 2 2
(87.4) (5.0) (1.7) (1.7)
Ann Thorac Surg 2013;-:-–-
Fig 3. Kaplan-Meier curves for freedom from atrial fibrillation (AF) off antiarrhythmic medication. (AAD ¼ antiarrhythmic drugs class I or III; LAA ¼ left atrial appendage.)
6-, and 12-month after surgery. Compared with the LAA resection group, LA diameter was smaller and the transmitral A-wave velocity was higher in the LAA preservation group at the 6- and 12-month points after surgery, with statistically significance (p < 0.05). The E/A ratio tended to be lower in the LAA preservation group throughout the follow-up period, and it was significantly lower at 12 months after surgery compared with that in the LAA resection group (p < 0.001; Fig 3C).
Comment To date, few studies exist that compare the atrial mechanical function in the LAA resection and preservation
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
7
procedures. For instance, Yamanaka and colleagues [6] performed a clinical analysis to evaluate the mechanical function of the LAA. They quantitatively examined the geometry of the LAA and left atrium by multidetector computed tomography in patients who underwent a modified, simplified Maze procedure. In the cited study, patients with preservation of LAA during the Maze procedure showed a higher LA ejection fraction than the control group, a finding that corresponds well with our present study. Cox and colleagues [10, 11] showed that the LAA may play a role as a potential source of AF; however, the present study showed no significant differences in AF recurrence according to the resection versus preservation of the LAA. Several studies have shown that the LAA plays an important role in the formation of atrial thrombi [3, 4]. Conversely, other studies have shown that LAA contractile function contributes greatly to LA transport function [1, 4–6, 12, 13]; therefore, thrombus formation in LAA preservation group may not be a clinically relevant problem if the atrial contractile functions recover [6, 13]. Although the present study showed no significant difference in the CVA rates between the two groups, the relatively small number of patients affected by CVA may be insufficient to obtain an adequate statistical power, necessitating further studies regarding this issue. The LAA is considered as a neuroendocrine organ having a role of body fluids regulation [7, 8]. Yoshihara and colleagues [7] studied atrial natriuretic peptide secretion and body fluid balance after bilateral atrial appendage resection during the Maze procedure and found the plasma atrial natriuretic peptide level was significantly lower and there was a greater requirement for diuretics to maintain the body fluid balance than in the control group. However, our study examined neither the
Fig 4. Serial follow-up rhythm status at 3, 6, 12, and 24 months after surgery. Black areas indicate atrial fibrillation (AF); light gray areas, permanent pacemaker rhythm (PPM); hatched areas, junctional; dark gray, normal sinus rhythm (NSR).
8
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
Ann Thorac Surg 2013;-:-–-
Fig 5. The results of sequential follow-up echocardiography at 1, 3, 6, and 12 months after surgery. (A) Left atrial dimension. (B) Transmitral A-wave velocity. (C) E/A ratio, derived from peak velocities of the early filling wave (E wave) and the A wave.
body fluid balance nor the requirement of diuretics in the immediate postoperative period because of retrospective nature of the study. Another issue regarding the preservation or resection of LAA is surgical bleeding. Although the rate of reexploration due to surgical bleeding did not significantly differ between the two groups in this study (p ¼ 0.17), the rate of reexploration was almost twofold more frequent in the LAA resection group compared with the LAA preservation group. Because LAA resection can cause troublesome bleeding, especially when the LA chamber is dilated with wall thinning [14, 15], this complication needs to be compared in studies with larger number of patients. The measurement of peak velocities of the atrial contraction wave (A wave) is a simple way of evaluating the LA booster contribution to ventricular filling. The A wave, however, is influenced by many factors such as compliance of the LV; therefore, the E/A ratio, which is derived from peak velocities of the early filling wave (E wave) and the A wave, is regarded as a more reliable index in the assessment of LA booster functions [16, 17]. Nevertheless, the E/A ratio of the mitral valve are also known to be influenced by several factors other than LA function. In these regards, the assessment by twodimensional doppler echocardiography may have an intrinsic limitation in evaluating LA function [17].
limitation is that only a relatively small number of patients were available for statistical analysis for the echocardiography outcomes over 12 months after operation. Moreover, the number of patients affected by CVA is insufficient for calculating statistically significant differences between the two groups. That requires studies with prospective designs and completeness of follow-up evaluation. In this study, measurements of atrial transport functions were made solely by conventional two-dimensional echocardiography. For further comprehensive evaluations, modern sophisticated imaging modalities such as three-dimensional echocardiography, multidetector computed tomography and cardiac magnetic resonance imaging may offer more comprehensive threedimensional evaluations of atrial functions. Furthermore, postoperative MV hemodynamics (either regurgitation or pressure gradient) might have affected the postoperative LA size or E/A ratio as confounding factors. In conclusion, compared with LAA resection, LAA preservation conserves superior LA mechanical function after surgery without significant differences in postoperative adverse events among patients with AF who are undergoing concomitant cardiac surgery. The study results need to be verified by further prospective randomized studies involving reasonably sized populations.
Study Limitations
References
This study is subject to the limitations inherent in retrospective work with observational data. The nonrandomized design may have affected the results owing to unmeasured confounders, procedures, or detection bias, despite the use of propensity score matching. Another
1. Isobe F, Kumano H, Ishikawa T, et al. A new procedure for chronic atrial fibrillation: bilateral appendage-preserving maze procedure. Ann Thorac Surg 2001;72:1473–8. 2. Lee JW, Choo SJ, Kim KI, et al. Atrial fibrillation surgery simplified with cryoablation to improve left atrial function. Ann Thorac Surg 2001;72:1479–83.
Ann Thorac Surg 2013;-:-–-
3. Blackshear JL, Odell JA. Appendage obliteration to reduce stroke in cardiac surgical patients with atrial fibrillation. Ann Thorac Surg 1996;61:755–9. 4. Garcia-Fernandez MA, Perez-David E, Peralta J, GarciaRojas I, Bermejo J. Role of left atrial appendage obliteration in stroke reduction in patients with mitral valve prosthesis. J Am Coll Cardiol 2003;42:1253–8. 5. Stollberger C, Schneider B, Finsterer J. Elimination of left atrial appendage to prevent stroke or embolism. Chest 2003;124:2356–62. 6. Yamanaka K, Sekine Y, Nonaka M, et al. Left atrial appendage contributes to left atrial booster function after the Maze procedure: quantitative assessment with multidetector computed tomography. Eur J Cardiothorac Surg 2010;38: 361–5. 7. Fumiki Y, Nishikimi T, Kosakai Y, et al. Atrial natriuretic peptide secretion and body fluid balance after bilateral atrial appendectomy by the maze procedure. J Thorac Cardiovasc Surg 1998;116:213–9. 8. Albåge A, Kenneb€ ack G, van der Linden J, Berglund H. Improved neurohormonal markers of ventricular function after restoring sinus rhythm by the maze procedure. Ann Thorac Surg 2003;75:790–5. 9. Kim JB, Bang JH, Jung SH, Choo SJ, Chung CH, Lee JW. Left atrial ablation versus biatrial ablation in the surgical treatment of atrial fibrillation. Ann Thorac Surg 2011;92: 1397–405.
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
9
10. Cox JL. The central controversy surrounding the interventional-surgical treatment of atrial fibrillation. J Thorac Cardiovasc Surg 2005;129:1–4. 11. Cox JL, Bolneau JP, Schuessler RB, et al. Operations for atrial fibrillation. Clin Cardiol 1991;14:827–34. 12. Schneider B, Stollberger C, Sievers HH. Surgical closure of the left atrial appendage—a beneficial procedure? Cardiology 2005;104:127–32. 13. Buber J, Luria D, Sternik L, et al. Left atrial contractile function following a successful modified maze procedure at surgery and the risk for subsequent thromboembolic stroke. J Am Coll Cardiol 2011;58:1614–21. 14. Johnson WD, Ganjoo AL, Stone CD, Srivyas RC, Howard M. The left atrial appendage: our most lethal human attachment! Surgical implications. Eur J Cardiothorac Surg 2000;17:718–22. 15. Roth P, Rahimi A, Boening A. The pericardium-reinforced technique of amputation of the left atrial appendage: quick, safe, and simple. Ann Thorac Surg 2010;90:e11–3. 16. 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): II285–92. 17. Yashima N, Nasu M, Kawazoe K, Hiramori K. Serial evaluation of atrial function by Doppler echocardiography after the Maze procedure for chronic atrial fibrillation. Eur Heart J 1997;18:496–502.
Background. Left atrial appendage (LAA) resection during the Maze procedure may decrease thromboembolic risks, but its preservation may improve left atrial contractile function. This study compared the clinical effects of LAA resection and preservation after the Maze procedure. Methods. A retrospective review was made of 379 patients (mean age 53.3 ± 12.6 years, 244 females) who underwent the cryo-Maze procedure in conjunction with mitral surgery from 1999 to 2011. The LAA was resected in 187 patients (resection group) but preserved in 192 patients (preservation group). Outcomes were compared using a propensity score study design based on 20 baseline characteristics to obtain well-matched patient pairs. Results. Propensity score matching yielded 119 pairs of patients in whom there were no significant differences in baseline profiles between the two groups. During a mean follow-up of 3.1 ± 2.8 years, there were 16 deaths, 6 cases
of stroke, and 39 cases of atrial fibrillation recurrence. There were no significant differences in stroke-free survival (p [ 0.88) and freedom from AF while off antiarrhythmic drugs (p [ 0.46) between the two groups. On serial echocardiographic assessments, patients in the preservation group showed a higher transmitral A-wave velocity (peak atrial contraction wave velocity; p [ 0.47, 0.020, and 0.001 at 3, 6, and 12 months, respectively) and lower E/A ratio (peak early filling wave [E-wave] velocity / A-wave velocity; p [ 0.34, 0.065, and 0.001 at 3, 6, and 12 months, respectively) at each timepoint compared with the resection group. Conclusions. Preservation of the LAA during the Maze procedure resulted in similar clinical and rhythm outcomes, but LA contractile function superior to that of LAA resection.
T
involved in atrial natriuretic peptide secretion and plays an important role in neurohormonal regulation [7, 8]. In these regards, few clinical studies exist that compare LAA resection with LAA preservation in clinical outcomes. In this study, we sought to compare the outcomes between LAA preservation and resection in terms of clinical and rhythm outcomes as well as atrial transport functions.
he Maze procedure is widely accepted as a way to effectively restore sinus rhythm in patients with atrial fibrillation (AF) who are undergoing cardiac surgery for other indications. Since the introduction of the Maze procedure, efforts have been made to reduce the complexity of the procedure and procedural complications [1, 2]. Accordingly, various techniques, including the use of alternative energy sources for creation of ablation lesions and modified simpler lesion sets, were developed to replace the conventional complex cut-and-sew procedure. One of these modifications involves left atrial appendage (LAA) preservation during the Maze procedure. The LAA is reported to be a major source of thrombus formation and its embolization, and can contribute as an ectopic focus of AF trigger [3–5]. However, the LAA can contribute greatly to left atrial mechanical contraction after the Maze procedure [6]. Moreover, some researchers assert that the LAA is largely Accepted for publication July 22, 2013. Address correspondence to Dr Jae Won Lee, Department of Thoracic and Cardiovascular Surgery, Asan Medical Center, University of Ulsan Collage of Medicine, 88, Olympic-ro 43-gil, Songpa-gu, Seoul 138-736, South Korea; e-mail: [email protected].
Ó 2013 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2013;-:-–-) Ó 2013 by The Society of Thoracic Surgeons
Patients and Methods Study Population Between January 1999 and January 2011, a total of 525 patients underwent mitral valve (MV) operations and concomitant biatrial cryo-Maze procedures at the Asan Medical Center, Seoul, Korea. Patients were excluded if they underwent concomitant multivessel coronary artery bypass graft surgery (n ¼ 22) or aortic replacement (n ¼ 8), or had a preoperative LAA thrombus (n ¼ 116). In the end, 379 patients were enrolled. The decision to resect or preserve LAA was influenced by several demographic and procedural factors. For instance, patients with high thromboembolic risks such as old age, coexisting diabetes mellitus, and advanced degree of tricuspid regurgitation 0003-4975/$36.00 http://dx.doi.org/10.1016/j.athoracsur.2013.07.073
2
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
or pulmonary hypertension were more likely to undergo LAA resection. Nevertheless, the decision was finally at the discretion of the attending surgeon referencing those baseline profiles. This study was approved by the Institutional Ethics Committee/Review Board at the Asan Medical Center. The requirement to obtain informed consent was waived by the Board owing to the retrospective nature of the study.
Surgical Techniques A median sternotomy approach (n ¼ 280, 73.9%), a minithoracotomy approach using the AESOP 3000 system (Computer Motion, Santa Barbara, CA [n ¼ 90, 23.7%]), or port-assisted minithoracotomy approach using the da Vinci robotic system (Intuitive Surgical, Irvine, CA [n ¼ 9, 2.4%]) were used. The biatrial cryo-Maze procedure was performed as previously described [2, 9]. After snaring down the superior and inferior vena cavae, oblique right atriotomy was made on beating heart. Then, the cavotricuspid isthmus isolation was achieved by two cryoablation lines, and a linear lesion from posterior end of atriotomy to superior vena cava was made (Fig 1A). After aortic cross-clamping, the left side procedure began with a longitudinal right-sided left atriotomy, which was performed endocardially before the MV procedure. Left atrium ablation consisted of a single box lesion for pulmonary vein isolation and two posterior linear lesions, one from the pulmonary vein isolation line to the LAA orifice and one from the pulmonary vein isolation line to the MV annulus. Additional epicardial coronary sinus ablation was performed at the opposite side of the linear lesion to MV annulus. The left atrium size was reduced by resection of redundant atrial tissue between inferior pulmonary vein and posterior MV annulus (Fig 1B). An argon-based flexible cryoablation system (SurgiFrost; Medtronic, Minneapolis, MN) was used for ablation. The surgeons attempted to remain as much subvalvular tissue as possible in a chordae-sparing manner during MV replacement.
Ann Thorac Surg 2013;-:-–-
Andover, MA). During the postoperative hospitalization period, typical 12-channel surface electrocardiography was performed every day. For patients who had normal sinus rhythm, Holter monitoring was performed for adjunctive evaluation. Follow-up echocardiography was performed at 1, 3, 6, and 12 months after surgery and electrocardiography was generally performed at 1, 3, 6, 12, 24, and 36 months after surgery. Early AF events were defined as any AF episodes, including atrial tachycardia and atrial flutter, less than 3 months after the operation.
Postoperative Management Patients who underwent valve repair or bioprosthetic valve implantation were routinely administered warfarin for 3 to 6 months postoperatively, with a target international normalized ratio of 1.5 to 2.5 at the discretion of the attending surgeon. Maintenance of anticoagulation therapy thereafter was determined according to the presence of thromboembolic risks and cardiac rhythm status in each patient. And then, to discontinue warfarin medication, the patient should have shown sinus rhythm at least for more than 2 consecutive visits and be free of symptoms. For patients with mechanical valve implantation, an international normalized ratio of 2.0 to 3.0 was the aim, regardless of cardiac rhythm status. Postoperative atrial tachyarrhythmias (AF, atrial flutter, or atrial tachycardia) were managed with class I or III antiarrhythmic drugs including amiodarone, sotalol, flecainide, and pilsicainide in an effort to restore sinus rhythm. For patients who achieved “AF elimination,” those drugs were withdrawn within 1 or 2 weeks. Patients who failed to achieve AF elimination despite antiarrhythmic medications for sufficient duration were taken off these drugs and switched to a “rate control strategy” involving digitalis, beta-blockers, or calcium-channel blockers to control ventricular rate in combination with anticoagulation therapy. Any symptoms suggestive of neurologic deficit were evaluated by neurologists with adequate imaging studies.
Echocardiography and Rhythm Follow-Up Preoperative transthoracic echocardiography and transesophageal echocardiography were performed in all patients within 2 months before surgery using a Hewlett-Packard Sonos 2500 or 5500 imaging system equipped with a 2.5-MHz transducer (Hewlett-Packard, Fig 1. Schema of surgical ablation of atrial fibrillation. (CS ¼ coronary sinus; IVC ¼ inferior vena cava; LAA ¼ left atrial appendage; MV ¼ mitral valve; PAs ¼ pulmonary veins; RAA ¼ right atrial appendage; SVC ¼ superior vena cava; TV ¼ tricuspid valve.)
Statistical Analysis Categorical variables are presented as frequencies and percentages, and continuous variables are expressed as a mean with standard deviation or a median with range. To reduce the effect of treatment selection bias and
Ann Thorac Surg 2013;-:-–-
3
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
Table 1. Baseline Characteristics of the Entire Cohort Characteristics Number of patients Age, years Male Diabetes mellitus Hypertension Chronic renal failure Congestive heart failure NYHA functional classification I II III IV History of stroke Prior cardiac surgery AF profiles Fine (<1 mm) AF wave AF type Paroxysmal Persistent Mitral diagnosis Rheumatic Degenerative Others Echocardiographic data Mitral regurgitation grade 0 1 2 3 4 Mitral valve area, cm2 Tricuspid regurgitation grade 0 1 2 3 4 Left ventricular ejection fraction, % Left atrial dimension, mm Transtricuspid peak pressure gradient, mm Hg Mitral valve operation Repair Replacement, mechanical Replacement, bioprosthetic Minimally invasive approach Concomitant cardiac surgery Aortic valve repair or replacement Tricuspid valve repair Coronary artery bypass a
p < 0.05.
Data are n (%) or mean SD. AF ¼ atrial fibrillation;
NYHA ¼ New York Heart Association.
Resection Group
Preservation Group
187 55.98 12.2 61 (32.6) 22 (11.8) 37 (19.8) 1 (0.5) 107 (57.2)
192 50.7 12.4 74 (38.5) 11 (5.7) 28 (14.5) 0 119 (62.0)
30 100 48 9 12 8
(16.0) (53.5) (25.7) (4.8) (6.4) (4.3)
30 86 65 11 10 14
(15.6) (44.8) (33.9) (5.7) (5.2) (7.3)
p Value <0.001a 0.23 0.037 0.18 0.31 0.35 0.29
0.62 0.21
79 (42.2)
75 (39.6)
0.53 0.20
11 (5.9) 176 (94.1)
18 (9.4) 174 (90.6)
119 (63.6) 64 (34.2) 4 (2.1)
121 (63.0) 58 (30.2) 13 (6.8)
22 (11.8) 25 (13.4) 18 (9.6) 20 (10.7) 102 (54.5) 2.62 1.37
20 (10.4) 20 (10.4) 14 (7.3) 32 (16.7) 106 (55.2) 2.62 1.33
0.42
16 (8.6) 29 (15.5) 41 (21.9) 44 (23.5) 57 (30.5) 56.3 7.1 60.6 10.5 38.2 13.2
25 (13.0) 56 (29.2) 39 (20.3) 35 (18.2) 37 (19.3) 56.1 9.0 59.7 10.1 34.9 10.1
0.003a
0.082
83 87 17 56
(44.4) (46.5) (9.1) (29.9)
39 (20.9) 122 (65.2) 10 (5.3)
105 85 2 43
(54.7) (44.3) (1.0) (22.4)
26 (13.5) 98 (51.0) 5 (2.6)
0.99
0.81 0.40 0.007a 0.094
0.10 0.076 0.007a 0.20
4
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
Ann Thorac Surg 2013;-:-–-
Table 2. Baseline Characteristics of Propensity Matched Patients Characteristics Number of patients Age, years Male Diabetes mellitus Hypertension Chronic renal failure Congestive heart failure NYHA functional classification I II III IV History of stroke Prior cardiac surgery AF profiles Fine (<1 mm) AF wave AF type Paroxysmal Persistent Mitral diagnosis Rheumatic Degenerative Others Echocardiographic data Mitral regurgitation grade 0 1 2 3 4 Mitral valve area, cm2 Mitral valve diagnosis Predominant mitral regurgitation Predominant mitral stenosis Mixed lesion Tricuspid regurgitation grade 0 1 2 3 4 Left ventricular ejection fraction, % Left atrial dimension, mm 50 mm, <60 mm 60 mm Transtricuspid peak pressure gradient, mm Hg Mitral valve operation Repair Replacement, mechanical Replacement, bioprosthetic
Resection Group
Preservation Group
119 53.4 12.5 45 (37.8) 9 (7.6) 22 (18.5) 0 (0) 66 (55.5)
119 54.1 11.9 49 (41.2) 9 (7.6) 23 (19.3) 0 (0) 64 (53.8)
22 61 31 5 6 8
(18.5) (51.3) (26.1) (4.2) (5.0) (6.7)
23 60 32 4 7 5
(19.3) (50.4) (26.9) (3.4) (5.9) (4.2)
71 (59.7)
79 (66.4)
11 (9.2) 108 (90.8)
11 (9.2) 108 (90.8)
73 (61.3) 42 (35.3) 4 (3.4)
66 (55.5) 46 (38.7) 7 (5.9)
15 (12.6) 13 (10.9) 11 (9.2) 14 (11.8) 66 (55.5) 2.68 1.36
12 (10.1) 13 (10.9) 8 (6.7) 17 (14.3) 69 (58.0) 2.84 1.34
76 (63.9) 35 (29.4) 8 (6.7)
82 (68.9) 24 (20.2) 13 (10.9)
13 (10.9) 25 (21.0) 25 (21.0) 26 (21.8) 30 (25.2) 56.5 7.1 59.0 10.9 46 (38.7) 51 (42.9) 35.7 12.4
13 (10.9) 26 (21.8) 27 (22.7) 24 (20.2) 29 (24.4) 56.9 8.7 59.7 9.9 47 (39.5) 50 (42.0) 36.3 10.0
58 (48.7) 60 (50.4) 1 (0.8)
64 (53.8) 53 (44.5) 2 (1.7)
p Value 0.57 0.70 >0.99 >0.99 >0.99 0.89 0.86
>0.99 0.55 0.78 >0.99
0.91
0.44
0.36 0.92
0.80
0.67 0.59 >0.99 >0.99 0.66 0.54
(Continued)
Ann Thorac Surg 2013;-:-–-
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
5
Table 2. Continued Characteristics
Resection Group
Preservation Group
22 (18.5) 68 (57.1) 3 (2.5)
20 (16.8) 65 (54.6) 5 (4.2)
Concomitant cardiac surgery Aortic valve repair or replacement Tricuspid valve repair Coronary artery bypass
p Value
0.85 0.79 0.73
Data are n (%) or mean SD. AF ¼ atrial fibrillation;
NYHA ¼ New York Heart Association.
potential confounding, we performed adjustments for the differences in the baseline characteristics by use of propensity score matching. The propensity scores were estimated with multiple logistic regression analysis. Prespecified covariates, listed in Table 1, were included for the calculation of propensity scores. The discrimination and calibration abilities of the propensity score model were assessed by means of C statistics and the HosmerLemeshow test. The model had a C statistic of 0.76 and a Hosmer-Lemeshow goodness-of-fit p value of 0.44, indicating that this model was well calibrated with reasonable discrimination. For the development of the propensity score-matched pairs (a 1:1 match), the greedy 5-to-1 digit matching algorithm was used. After propensity score matching, the data of the two groups were compared using the paired Student’s t test or the Wilcoxon signed rank test for continuous variables, and with McNemar’s test or the marginal homogeneity test for categorical variables. All reported p values were two-sided, and a value of p less than 0.05 was considered statistically significant. For statistical analysis, SAS software, version 9.1 (SAS Institute, Cary, NC), was used.
Results
valvuloplasty. Propensity score matching yielded 119 pairs of patients (resection group and preservation group). After propensity score matching, there were no significant intergroup differences in their baseline characteristics between the two groups (Table 2).
Early Outcomes Early postoperative complications are presented in Table 3. There was 1 early death (0.4%) in the LAA preservation group, of a patient who underwent mechanical MV implantation and myectomy of the left ventricular outflow tract. The patient died of panperitonitis followed by septic shock during mechanical circulatory support 13 days after surgery. There were 2 early cerebrovascular accidents (CVA [0.8%]), 1 in each group. In the LAA preservation group, the CVA occurred in a patient who had undergone mitral valvuloplasty and tricuspid valvuloplasty by the minimally invasive approach. The other CVA occurred in a patient of the LAA resection group who had a preoperative history of CVA and presented with altered mental status and motor weakness immediately after surgery. There were no statistically significant differences in early complications and mortalities between the two groups.
Baseline Patient Profiles
Late Clinical Outcomes
The baseline characteristics of the patients are listed in Table 1. The patients who underwent LAA resection were older than those who underwent LAA preservation. The LAA-preserved patients had a higher grade of tricuspid regurgitation and more frequently underwent tricuspid
Clinical follow-up was complete in all patients with a mean follow-up duration of 62.6 44.0 months (median 60.5; range, 0.03 to 161.4). The rates of anticoagulation therapy in the late period (more than 6 months) was similar for both groups (74 of 119 in the resection group
Table 3. Early Operative Complications Variables Early mortality Early complications Early stroke Low cardiac output syndrome Reoperation due to bleeding Requirement for dialysis Permanent pacemaker insertion Mediastinitis Wound revision Pericardial effusion
Resection Group, n (%) 0 (0) 1 0 13 9 4 0 1 8
(0.8) (0) (10.9) (7.6) (3.4) (0) (0.8) (6.7)
Preservation Group, n (%)
p Value
1 (0.8)
>0.99
1 1 6 10 2 1 2 6
(0.8) (0.8) (5.0) (8.4) (1.7) (0.8) (1.7) (5.0)
>0.99 >0.99 0.17 >0.99 0.63 >0.99 >0.99 0.77
6
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
Ann Thorac Surg 2013;-:-–-
Fig 2. Kaplan-Meier curves for (A) overall survival rate and (B) stroke-free survival. Solid line indicates left atrial appendage (LAA); broken line indicates LAA resection.
versus 66 of 119 in the preservation group, p ¼ 0.29), for which the most frequent indication was the implantation of mechanical valves (n ¼ 60 and n ¼ 53 in the resection group and preservation group, respectively) followed by the recurrence of AF or high risks of thromboembolism (more than CHADS2 [acronym for congestive heart failure, hypertension, age 75, diabetes mellitus, stroke or transient ischemic attack history] score of 2; n ¼ 14 and n ¼ 13 in the resection group and preservation group, respectively). During follow-up, there were 16 late deaths, 6 in the resection group and 10 in the preservation group, with no intergroup difference. Three patients were affected by late thromboembolic events (1 in the resection group [CVA] and 2 in the preservation group [1 transient ischemic attack and 1 CVA], p ¼ 0.63). All of these patients were receiving warfarin anticoagulation therapy owing to the insertion of mechanical valve (n ¼ 1 in the resection group), recurrence of AF (n ¼ 1 in the preservation group who underwent MV repair), and history of previous CVA (n ¼ 1 in the preservation group who underwent MV repair). There were no significant differences in the incidence of late stroke (p ¼ 0.63), reoperation of previous surgery site (2 in the resection group and 3 in the preservation group; p > 0.99), anticoagulation-related hemorrhages (4 in the resection group only; p ¼ 0.13), and admission due to congestive heart failure (2 in each group; p > 0.99). There were no
significant differences in overall survival or CVA-free survival between the two groups (Fig 2).
Rhythm Outcomes Early and late AF events (more than 3 months after surgery) and latest follow-up rhythms are summarized in Table 4; there were no significant differences in the rhythm status. Rhythm follow-up over 3 months after surgery was possible in 224 patients (94.1%), with a median follow-up duration of 26.3 months (range, 0.03 to 139.4). There was no significant difference in freedom from AF recurrence off antiarrhythmic medication between the two groups (Fig 3). The sequential follow-up of rhythm status during the 3-month, 6-month, 1-year, and 2-year points after surgery are shown in Figure 4. There were no significant differences at these timepoints after surgery between the two groups. There were 65 patients (33 in the resection group and 32 in the preservation group) who underwent electrical cardioversion during hospitalization (p > 0.99). During follow-up, electrical cardioversion was performed for 5 patients in each group.
Atrial Transport Functions Figure 5 depicts echocardiographic parameters of the left atrium in propensity score matched patients at 1-, 3-,
Table 4. Early and Late Atrial Fibrillation Events and Latest Follow-Up Cardiac Rhythm Status Variables Early (3 months) AF events Late (>3 months) AF events Latest follow-up rhythm (>3 months) Normal sinus rhythm Atrial fibrillation/flutter/tachycardia PPMa Junctional rhythm a
Resection Group, n (%)
Preservation Group, n (%)
p Value
27 (22.7) 21 (17.6)
28 (23.5) 18 (15.1)
0.878 0.600 0.514
98 11 4 1
(82.4) (9.2) (3.4) (0.8)
Permanent pacemaker (PPM) rhythm or sick sinus syndrome or complete atrioventricular block.
AF ¼ atrial fibrillation.
104 6 2 2
(87.4) (5.0) (1.7) (1.7)
Ann Thorac Surg 2013;-:-–-
Fig 3. Kaplan-Meier curves for freedom from atrial fibrillation (AF) off antiarrhythmic medication. (AAD ¼ antiarrhythmic drugs class I or III; LAA ¼ left atrial appendage.)
6-, and 12-month after surgery. Compared with the LAA resection group, LA diameter was smaller and the transmitral A-wave velocity was higher in the LAA preservation group at the 6- and 12-month points after surgery, with statistically significance (p < 0.05). The E/A ratio tended to be lower in the LAA preservation group throughout the follow-up period, and it was significantly lower at 12 months after surgery compared with that in the LAA resection group (p < 0.001; Fig 3C).
Comment To date, few studies exist that compare the atrial mechanical function in the LAA resection and preservation
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
7
procedures. For instance, Yamanaka and colleagues [6] performed a clinical analysis to evaluate the mechanical function of the LAA. They quantitatively examined the geometry of the LAA and left atrium by multidetector computed tomography in patients who underwent a modified, simplified Maze procedure. In the cited study, patients with preservation of LAA during the Maze procedure showed a higher LA ejection fraction than the control group, a finding that corresponds well with our present study. Cox and colleagues [10, 11] showed that the LAA may play a role as a potential source of AF; however, the present study showed no significant differences in AF recurrence according to the resection versus preservation of the LAA. Several studies have shown that the LAA plays an important role in the formation of atrial thrombi [3, 4]. Conversely, other studies have shown that LAA contractile function contributes greatly to LA transport function [1, 4–6, 12, 13]; therefore, thrombus formation in LAA preservation group may not be a clinically relevant problem if the atrial contractile functions recover [6, 13]. Although the present study showed no significant difference in the CVA rates between the two groups, the relatively small number of patients affected by CVA may be insufficient to obtain an adequate statistical power, necessitating further studies regarding this issue. The LAA is considered as a neuroendocrine organ having a role of body fluids regulation [7, 8]. Yoshihara and colleagues [7] studied atrial natriuretic peptide secretion and body fluid balance after bilateral atrial appendage resection during the Maze procedure and found the plasma atrial natriuretic peptide level was significantly lower and there was a greater requirement for diuretics to maintain the body fluid balance than in the control group. However, our study examined neither the
Fig 4. Serial follow-up rhythm status at 3, 6, 12, and 24 months after surgery. Black areas indicate atrial fibrillation (AF); light gray areas, permanent pacemaker rhythm (PPM); hatched areas, junctional; dark gray, normal sinus rhythm (NSR).
8
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
Ann Thorac Surg 2013;-:-–-
Fig 5. The results of sequential follow-up echocardiography at 1, 3, 6, and 12 months after surgery. (A) Left atrial dimension. (B) Transmitral A-wave velocity. (C) E/A ratio, derived from peak velocities of the early filling wave (E wave) and the A wave.
body fluid balance nor the requirement of diuretics in the immediate postoperative period because of retrospective nature of the study. Another issue regarding the preservation or resection of LAA is surgical bleeding. Although the rate of reexploration due to surgical bleeding did not significantly differ between the two groups in this study (p ¼ 0.17), the rate of reexploration was almost twofold more frequent in the LAA resection group compared with the LAA preservation group. Because LAA resection can cause troublesome bleeding, especially when the LA chamber is dilated with wall thinning [14, 15], this complication needs to be compared in studies with larger number of patients. The measurement of peak velocities of the atrial contraction wave (A wave) is a simple way of evaluating the LA booster contribution to ventricular filling. The A wave, however, is influenced by many factors such as compliance of the LV; therefore, the E/A ratio, which is derived from peak velocities of the early filling wave (E wave) and the A wave, is regarded as a more reliable index in the assessment of LA booster functions [16, 17]. Nevertheless, the E/A ratio of the mitral valve are also known to be influenced by several factors other than LA function. In these regards, the assessment by twodimensional doppler echocardiography may have an intrinsic limitation in evaluating LA function [17].
limitation is that only a relatively small number of patients were available for statistical analysis for the echocardiography outcomes over 12 months after operation. Moreover, the number of patients affected by CVA is insufficient for calculating statistically significant differences between the two groups. That requires studies with prospective designs and completeness of follow-up evaluation. In this study, measurements of atrial transport functions were made solely by conventional two-dimensional echocardiography. For further comprehensive evaluations, modern sophisticated imaging modalities such as three-dimensional echocardiography, multidetector computed tomography and cardiac magnetic resonance imaging may offer more comprehensive threedimensional evaluations of atrial functions. Furthermore, postoperative MV hemodynamics (either regurgitation or pressure gradient) might have affected the postoperative LA size or E/A ratio as confounding factors. In conclusion, compared with LAA resection, LAA preservation conserves superior LA mechanical function after surgery without significant differences in postoperative adverse events among patients with AF who are undergoing concomitant cardiac surgery. The study results need to be verified by further prospective randomized studies involving reasonably sized populations.
Study Limitations
References
This study is subject to the limitations inherent in retrospective work with observational data. The nonrandomized design may have affected the results owing to unmeasured confounders, procedures, or detection bias, despite the use of propensity score matching. Another
1. Isobe F, Kumano H, Ishikawa T, et al. A new procedure for chronic atrial fibrillation: bilateral appendage-preserving maze procedure. Ann Thorac Surg 2001;72:1473–8. 2. Lee JW, Choo SJ, Kim KI, et al. Atrial fibrillation surgery simplified with cryoablation to improve left atrial function. Ann Thorac Surg 2001;72:1479–83.
Ann Thorac Surg 2013;-:-–-
3. Blackshear JL, Odell JA. Appendage obliteration to reduce stroke in cardiac surgical patients with atrial fibrillation. Ann Thorac Surg 1996;61:755–9. 4. Garcia-Fernandez MA, Perez-David E, Peralta J, GarciaRojas I, Bermejo J. Role of left atrial appendage obliteration in stroke reduction in patients with mitral valve prosthesis. J Am Coll Cardiol 2003;42:1253–8. 5. Stollberger C, Schneider B, Finsterer J. Elimination of left atrial appendage to prevent stroke or embolism. Chest 2003;124:2356–62. 6. Yamanaka K, Sekine Y, Nonaka M, et al. Left atrial appendage contributes to left atrial booster function after the Maze procedure: quantitative assessment with multidetector computed tomography. Eur J Cardiothorac Surg 2010;38: 361–5. 7. Fumiki Y, Nishikimi T, Kosakai Y, et al. Atrial natriuretic peptide secretion and body fluid balance after bilateral atrial appendectomy by the maze procedure. J Thorac Cardiovasc Surg 1998;116:213–9. 8. Albåge A, Kenneb€ ack G, van der Linden J, Berglund H. Improved neurohormonal markers of ventricular function after restoring sinus rhythm by the maze procedure. Ann Thorac Surg 2003;75:790–5. 9. Kim JB, Bang JH, Jung SH, Choo SJ, Chung CH, Lee JW. Left atrial ablation versus biatrial ablation in the surgical treatment of atrial fibrillation. Ann Thorac Surg 2011;92: 1397–405.
LEE ET AL LAA RESECTION VERSUS PRESERVATION FOR AF
9
10. Cox JL. The central controversy surrounding the interventional-surgical treatment of atrial fibrillation. J Thorac Cardiovasc Surg 2005;129:1–4. 11. Cox JL, Bolneau JP, Schuessler RB, et al. Operations for atrial fibrillation. Clin Cardiol 1991;14:827–34. 12. Schneider B, Stollberger C, Sievers HH. Surgical closure of the left atrial appendage—a beneficial procedure? Cardiology 2005;104:127–32. 13. Buber J, Luria D, Sternik L, et al. Left atrial contractile function following a successful modified maze procedure at surgery and the risk for subsequent thromboembolic stroke. J Am Coll Cardiol 2011;58:1614–21. 14. Johnson WD, Ganjoo AL, Stone CD, Srivyas RC, Howard M. The left atrial appendage: our most lethal human attachment! Surgical implications. Eur J Cardiothorac Surg 2000;17:718–22. 15. Roth P, Rahimi A, Boening A. The pericardium-reinforced technique of amputation of the left atrial appendage: quick, safe, and simple. Ann Thorac Surg 2010;90:e11–3. 16. 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): II285–92. 17. Yashima N, Nasu M, Kawazoe K, Hiramori K. Serial evaluation of atrial function by Doppler echocardiography after the Maze procedure for chronic atrial fibrillation. Eur Heart J 1997;18:496–502.