International Journal of Cardiology 170 (2014) 331–337
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Maze procedure in patients with left ventricular dysfunction Joon Bum Kim 1, Byung Kwon Chong 1, Sung-Ho Jung 1, Suk Jung Choo 1, Cheol Hyun Chung 1, Jae Won Lee ⁎,1 Department of Thoracic and Cardiovascular Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea
a r t i c l e
i n f o
Article history: Received 21 February 2013 Received in revised form 13 September 2013 Accepted 2 November 2013 Available online 12 November 2013 Keywords: Atrial fibrillation Ablation Surgery Heart failure Survival
a b s t r a c t Background: The risks and benefits of a concomitant Maze procedure for patients with LV dysfunction undergoing major cardiac surgery have not yet been elucidated. This study aimed to evaluate the clinical impacts of the Maze procedure in patients with atrial fibrillation and left ventricular (LV) dysfunction. Methods: Between January 1999 and March 2011, a total of 139 patients (mean age 52.7 ± 12.3 years, 54 females) with valvular atrial fibrillation (AF) and an LV ejection fraction (EF) of 40% or less underwent open heart surgery with (n = 77) or without (n = 62) a concomitant Maze procedure. We compared adverse outcomes (death and composite of death, thromboembolic events and congestive heart failure [CHF]) during a median follow-up period of 66.0 months (inter-quartile range, 27.5–106.9 months). Results: Adverse events occurred in 41 patients, including 36 deaths, seven thromboembolic events and eight hospitalizations due to CHF. After adjustment for baseline profiles with the use of propensity scores and inverse probability weighting, patients who had the Maze procedure were at lower risks of death (hazard ratio, 0.39; 95% confidence interval, 0.16–0.93; P = 0.033) and composite adverse outcomes (hazard ratio, 0.28; 95% confidence interval, 0.14–0.57; P = 0.017) than those not undergoing the Maze procedure. Furthermore, the Maze procedure resulted in superior functional status (P b 0.001) and reduced the need for long-term anticoagulation therapy (67.1% vs. 91.2%, P = 0.001). Conclusion: Performing the Maze procedure on patients with valvular AF and LV dysfunction reduced serious adverse outcomes and the need for long-term anticoagulation therapy when compared to cardiac surgery alone without the Maze procedure. Crown Copyright © 2013 Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Since the introduction of the Cox–Maze procedure, its technical evolution has included using alternative energy sources, replacing cutand-sew methods and creating simpler ablation lesion sets rather than complex “maze” lesions [1–5]. Consequently, the use of surgical atrial fibrillation (AF) ablation has been prompted in real clinical practice to play a substantial role in selected patients. The efficacy of the Maze procedure in the eliminating AF is well proven in many published studies with the use of the classic Cox–Maze III procedure or equivalent AF ablation procedures using cryothermia or radiofrequency ablation [6–9]. Aside from sinus rhythm restoration, clinical advantages of the Maze procedure include a reduction in thromboembolic risks and improvements in hemodynamic profiles [10–12]. Despite the benefits of the Maze procedure, its performance in patients with marginal left ventricular (LV) function undergoing major cardiac surgery has been controversial. This is because the procedure
⁎ Corresponding author at: Department of Thoracic and Cardiovascular Surgery, Asan Medical Center, University of Ulsan College of Medicine, 388-1 Pungnap-dong Songpa-gu, Seoul 138-736, South Korea. Tel.: +82 2 3010 3584; fax: +82 2 3010 6966. E-mail address: [email protected] (J.W. Lee). 1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
may increase surgical risks as a consequence of increased cardiac ischemic time and more extensive procedures for these vulnerable patients. These concerns well appear in recent consensus and guidelines on catheter and surgical ablation of AF, in that in the setting of major cardiac surgery, concomitant surgical AF ablation can be indicated only in selected patients in whom the ablation can be performed with “minimal risk” if symptoms of AF do not exist [13]. In this regard, there are no relevant data in the literature that assess the risks and benefits of the Maze procedure in patients with significant LV dysfunction undergoing major cardiac surgery. We therefore evaluated the long-term clinical outcomes of patients with valvular AF and LV dysfunction, undergoing major cardiac surgery with or without a concomitant Maze procedure. 2. Methods 2.1. Patients We used our institution's database, which prospectively registers all patients undergoing major cardiac surgery. Out of a total of 3,123 patients who underwent mitral valve (MV) surgery between January 1999 and March 2011 at the Asan Medical Center, Seoul, Korea, we identified 139 patients who had AF and an LV ejection fraction (EF) of 40% or less. These patients formed the subject population of this study. Of these patients, 77 concomitantly underwent the Maze procedure (Maze group) while 62 underwent cardiac surgery without the Maze procedure (control group). The decision to perform the concomitant Maze procedure was influenced by each patient's risk profiles, such as
0167-5273/$ – see front matter. Crown Copyright © 2013 Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijcard.2013.11.008
332
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
the severity of LV dysfunction, the types of valve surgery and implanted prosthesis and overall estimated surgical risks but was finally at the discretion of the attending surgeon. This study was approved by our institutional ethics committee/review board, which waived the requirement for informed consent due to the retrospective nature of the study. 2.2. Surgical procedure In the Maze group, the AF ablation procedure was performed either by cryoablation (n = 66) or by microwave ablation (n = 11), with the lesion sets for the Maze procedure being created as those described previously [4,14]. For the right side ablation, oblique right atriotomy was made typically under beating heart cardiopulmonary bypass (CPB) support. Then, the cavo-tricuspid isthmus isolation was achieved by two or three ablation lines. Another linear lesion from posterior end of atriotomy to superior vena cava was made. The left-side procedure began after aortic cross clamping (ACC) through a longitudinal right-sided left atriotomy. Left-side ablation was composed of pulmonary vein (PV) isolation lesions and two posterior linear lesions (1) one from PV isolation line to left atrial (LA) appendage orifice and (2) one from PV isolation line to MV annulus. Additional epicardial coronary sinus ablation was performed at the opposite side of the MV annular lesion. The LA size was reduced by resecting the redundant atrial tissue off the posterior LA wall parallel to the posterior mitral annulus, with a target width less than 4 cm between the inferior pulmonary vein openings and the MV annulus. 2.3. Postoperative management Patients who underwent valve repair or bioprosthetic valve implantation were routinely anticoagulated with warfarin for 3 to 6 months postoperatively with a target international normalized ratio (INR) of 1.5 to 2.5. The maintenance of the anticoagulation therapy thereafter was determined according to individual risk factors of thromboembolism. Postoperative AF or the absence of effective atrial contraction resulted in prolonging the anticoagulation therapy until both sinus rhythm and atrial contractility were restored. For patients with mechanical valve implantation, achieving an INR between 2.0 and 3.0 was the aim, regardless of cardiac rhythm status. Anticoagulation therapy was adjusted during outpatient visits at 3-month intervals. Generally, follow-up electrocardiography assessments were performed at 3, 6, 12, 18 and 24 months after surgery, and annually thereafter. For patients with restored sinus rhythm, 24-h Holter monitoring was done to validate “free-of-AF”. Postoperative rhythm management strategies differed according to the performance of the Maze procedure. In the Maze group, for instance, postoperative atrial tachyarrhythmia was managed with class I/III anti-arrhythmic medications or electrical cardioversion in an effort to restore sinus rhythm (“rhythm control strategy”). If the patients failed to achieve “AF elimination” despite these treatments, they were switched to a “rate control strategy” to control ventricular rate (with the use of digitalis, β-blockers or calcium channel blockers) in combination with anticoagulation therapy. The principal rhythm management for the control group was the “rate control strategy” with the only exception of those who showed spontaneous recovery of sinus rhythm followed by AF recurrence. These patients were managed with the “rhythm control strategy” as the initial treatment intention. 2.4. Definitions and Follow-up The primary end point was defined as all-cause death. All-cause death rather than cardiac death was chosen as the primary end point because it is the most robust and unbiased index, requiring no adjudication to avoid inaccurate or biased documentation and clinical assessments [15]. The secondary end point was defined as a composite of all-cause death, thromboembolic events and congestive heart failure (CHF) requiring hospitalization. Thromboembolic events, which may manifest as neurologic events (non-hemorrhagic stroke or transient ischemic attack) or non-cerebral (peripheral) embolic events were clinically diagnosed [16]. Hospitalization due to CHF was defined as an unplanned admission for the management of CHF in cases with resting dyspnea and pulmonary edema requiring intravenous diuretics therapy. Clinical follow-ups were performed at 3- to 6-month intervals as outpatient clinic visits, according to the requirements for anticoagulation therapy. To validate all information regarding mortality, data on vital status and the dates and causes of death were obtained up to March 31, 2012, from the Korean National Registry of Vital Statistics. All deaths were considered of cardiac origin unless a non-cardiac origin was established clinically. 2.5. Statistical analysis Categorical variables, presented as frequencies and percentages, were compared using the chi-square test or Fisher's exact test. Continuous variables, expressed as mean ± SD or median with range, were compared using the Student's unpaired t-test or the Mann–Whitney U-test, as appropriate. To reduce the effects of treatment selection bias and potential confounding factors, we adjusted for differences in baseline characteristics by weighted Cox regression models and inverse probability of treatment weighting (IPTW) [17,18]. With this technique, weights for patients receiving the concomitant Maze procedure were the inverse of propensity scores, and weights for patients receiving cardiac surgery alone, without the Maze procedure, were the inverse of “1 minus propensity score.” Propensity scores were estimated without considering the outcome variables, using multiple logistic regression models.
Pre-specified covariates listed in Table 1 were included in the models for the Maze versus the control groups. The discrimination and calibration abilities of the propensity score model were assessed by means of C statistics and the Hosmer–Lemeshow test. The model had a C statistic of 0.843 and a Hosmer–Lemeshow goodness of fit P value of 0.561, indicating that this model was well calibrated with reasonable discrimination. All reported P values were two-sided, and a value of P b 0.05 was considered statistically significant. SAS software, version 9.1 (SAS Institute, Cary, NC, USA) was used for the statistical analysis. The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
3. Results 3.1. Baseline characteristics The mean age at surgery was 52.7 ± 12.3 years, and 54 patients (38.8%) were female. Table 1 summarizes the baseline demographic and clinical characteristics of all patients. In summary, patients in the control group were more likely to have a history of previous cardiac surgery and were more likely to undergo valve replacement surgery, particularly mechanical valve replacement, than those in the Maze group. LV EF was lower and LA size was larger in the control group. An estimated risk of surgical mortality, represented by the EuroSCORE II [19], was significantly higher in the control group than in the Maze group. As expected, patients in the Maze group had a higher probability of undergoing the Maze procedure than did those in the control group (Fig. 1A). After adjustment using IPTW, both groups were fairly balanced in the distribution of the propensity scores (Fig. 1B). Consequently, all baseline profiles were well balanced, without a significant difference between the two groups after the adjustment (Table 1, rightmost column). 3.2. Early operative results and postoperative rhythm status The ACC (110.5 ± 51.3 min versus 104.7 ± 38.0 min; P = 0.54) and CPB times (177.1 ± 94.6 versus 159.0 ± 62.8 min; P = 0.088) tended to be longer in the Maze group than in the control group, but without a statistical significance. There were six (4.3%) early deaths, and 28 patients (20.1%) experienced major complications during postoperative hospitalization. There were no significant differences in early mortality or major morbidity rates between the two groups (Table 2). During the follow-up, patients in the Maze group showed superior rhythm outcomes in terms of freedom from AF and maintenance of sinus rhythm compared with those in the control group (Fig. 2; P b 0.001 for all postoperative time points). 3.3. Clinical end points Follow-up was complete in all patients (100%), with a median duration of 66.0 months (interquartile range 27.5–106.9 months, 805.2 patient-years). Throughout the study period, there were 30 late deaths (45.6%) including 27 cardiac deaths. Non-cardiac causes of death were malignancy in two patients, and suicide in one. Seven patients (10.7%) experienced thromboembolic events including non-hemorrhagic stroke in six patients, and transient ischemic attack in one patient. Eight patients (12.3%) were readmitted due to CHF. Table 3 summarizes the clinical end points according to the two treatment groups. The linearized rates of death, thromboembolic events and CHF were 4.4%, 0.9% and 1.0% per patient year, respectively, in the entire cohort. In the Maze group, these rates were 2.3%, 0.4% and 0.4%, respectively, and in the control group, they were 7.1%, 1.4% and 1.7%, respectively (P = 0.005, 0.13 and 0.073, respectively). Fig. 3 depicts the unadjusted rates of death and the composite of secondary end points in the two groups. The control group showed significantly higher rates of death and composite end points compared to the Maze group (P = 0.005 and b0.001, respectively).
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
333
Table 1 Baseline characteristics of all patients. Unadjusted
Age, years Male gender (%) NYHA functional class I II III IV Diabetes mellitus, n (%) Hypertension, n (%) History of thromboembolism, n (%) Previous cardiac surgery, n (%) CHADS2 score 1 2 ≥3 Type of AF, n (%) Paroxysmal Persistent (≤1 year) Longstanding persistent Echocardiographic data LV ejection fraction, % LV end-systolic dimension, mm LV end-diastolic dimension, mm Left atrial diameter, mm Peak PAP, mmHg Tricuspid regurgitation, n (%) None -trace Mild Moderate Severe MV operation type Repair Mechanical prosthetic replacement Bioprosthetic replacement Combined surgery AV replacement Mechanical prosthesis Bioprosthesis Coronary artery bypassing Tricuspid valve repair Estimated GFR, mL/min/1.73 m2 EuroSCORE II, %†
Adjusted by IPTW*
Maze group (n = 77)
Control group (n = 62)
P value
Maze group (n = 77)
Control group (n = 62)
P value
51.6 ± 12.5 45 (58.4)
54.0 ± 11.9 40 (64.5)
0.25 0.49 0.084
52.7 ± 12.1 46 (59.7)
52.2 ± 12.8 37 (59.7)
0.80 N0.99 0.39
8 (10.4) 24 (31.2%) 37 (48.1) 8 (10.4) 8 (10.4) 10 (13.0) 14 (18.2) 2 (2.6)
1 (1.6) 23 (37.1) 26 (41.9) 12 (19.4) 7 (11.3) 4 (6.5) 8 (12.9) 7 (11.3)
6 (7.8) 26 (33.8) 37 (48.1) 8 (10.4) 11 (14.3) 10 (13.0) 14 (18.2) 2 (2.6)
1 (1.6) 25 (40.3) 29 (46.8) 7 (11.3) 14 (22.6) 12 (19.4) 11 (11.7) 4 (6.5)
56 (72.7) 15 (19.5) 6 (7.8)
48 (77.4) 10 (16.1) 4 (6.5)
53 (67.9) 17 (21.8) 8 (10.3)
38 (61.3) 12 (19.4) 12 (19.4)
2 (2.6) 47 (61.0) 28 (36.4)
1 (1.6) 34 (54.8) 27 (43.5)
1 (1.3) 47 (61.8) 28 (36.8)
1 (1.6) 34 (55.7) 27 (44.3)
36.0 ± 48.0 ± 61.3 ± 57.3 ± 45.9 ±
33.5 ± 6.3 50.89 ± 11.0 64.0 ± 10.9 62.8 ± 10.3 50.2 ± 16.4
35.5 48.2 61.2 58.5 48.1
34.3 ± 49.5 ± 63.0 ± 60.1 ± 52.8 ±
0.87 0.20 0.40 0.038 0.77
0.66
3.6 8.6 10.2 8.4 17.7
32 (41.6) 19 (24.7) 14 (18.2) 12 (15.6)
21 (33.9) 11 (17.7) 13 (21.0) 17 (27.4)
34 (44.2) 36 (46.8) 7 (9.1)
12 (19.4) 41 (66.1) 9 (14.5)
16 (20.8) 14 (18.2) 2 (2.6) 14 (18.2) 35 (45.5) 73.3 ± 19.5 13.1 ± 8.0
29 (46.8) 21 (33.9) 8 (12.9) 7 (11.3) 24 (38.7) 71.6 ± 25.4 18.4 ± 13.7
0.006 0.096 0.14 0.001 0.15 0.38
0.48
± ± ± ± ±
3.8 8.6 10.1 8.2 19.7
5.1 10.7 11.7 8.6 15.6
30 (39.0) 17 (22.1) 15 (19.5) 15 (19.5)
22 (35.4) 19 (30.6) 9 (14.5) 12 (19.4)
27 (35.1) 41 (53.2) 9 (11.7)
23 (37.1) 33 (53.2) 6 (9.7)
18 (23.7) 16 (21.1) 2 (2.6) 12 (15.6) 33 (42.9) 73.0 ± 19.9 14.3 ± 9.1
18 (29.0) 13 (21.0) 5 (8.1) 10 (16.1) 29 (46.8) 67.6 ± 37.6 15.9 ± 11.2
0.008
0.003
0.26 0.42 0.66 0.008
0.27 0.35 0.95 0.41 0.28
0.11 0.45 0.35 0.25 0.12 0.80
0.92
0.35
N0.99 0.73 0.31 0.35
*Numbers of patients for categorical variables were rounded off to the nearest whole number. †In calculating the EuroSCORE II, the concomitant Maze procedure was not considered in the “weight of procedure.” IPTW indicates inverse probability treatment weighting; NYHA, New York Heart Association; AF, atrial fibrillation; LV, left ventricle; PAP, pulmonary artery pressure; MV, mitral valve; AV, aortic valve; GFR, glomerular filtration rate.
Table 4 summarizes the adjusted hazard ratios for clinical end points in the Maze group compared to the control group. Patients undergoing the concomitant Maze procedure were at significantly lower risks of
death, and the composite of death, thromboembolism and CHF, than those not receiving the procedure. Adjusted rates of clinical end points are illustrated in Fig. 4.
Fig. 1. Box plots for propensity scores in the Maze and the control groups. The propensity score is the probability given baseline variables that any patient in either group would be selected for a concomitant Maze procedure. (A) Patients in the Maze group have higher propensity scores than those in the control group. (B) After inverse-probability treatment weighting, the distribution of the propensity scores is fairly balanced between the two groups.
334
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
Table 2 Early postoperative outcomes.
Table 3 Overall adverse outcomes during follow-up. Maze
Control
P
N (%)
(n = 77) (n = 62) Unadjusted Adjusted
Early death Major complications LCOS requiring mechanical support New dialysis Bleeding reoperation Pneumonia Multi-organ failure Permanent pacemaker insertion Wound revision Prolonged ventilator support (N24 hr)
2 (2.6) 4 (6.5) 0.41 15 (19.5) 13 (21.0) 0.83 4 (5.2) 3 (4.8) 0.92
0.66 0.84 N0.99
6 (7.8) 5 (6.5) 2 (2.6) 2 (2.6) 2 (2.6) 3 (3.9) 13 (16.9)
0.19 0.73 0.59 N0.99 0.14 0.13 0.39
3 (4.8) 4 (6.5) 2 (3.2) 0 (0) 2 (3.2) 1 (1.6) 12 (19.4)
0.48 0.99 0.83 0.20 0.83 0.42 0.83
LCOS indicates low cardiac output syndrome requiring a mechanical support.
3.4. Other adverse outcomes Nineteen cases of other adverse clinical events occurred in 15 patients (10.8%): nine patients (11.7%) in the Maze group and six patients (9.7%) in the control group. These adverse outcomes included hemorrhagic complications related to anticoagulation therapy in eight patients, heart valve reoperation in five patients, infective endocarditis in two patients, ventricular arrhythmia in one patient, acute myocardial infarction in one patient and a requirement for percutaneous coronary stenting in one patient. Cumulative event rates at 5 years were 18.9 ± 6.7% in the Maze group and 13.1 ± 5.1% in the control group (unadjusted and adjusted P = 0.91 and 0.93, respectively).
Patients, n (%)
Maze group
Control group
Death Early (≤30 days) death Late death Thromboembolism CHF Death, thromboembolism and CHF
11 (14.3) 2 (2.6) 9 (11.7) 2 (2.6) 2 (2.6) 12 (15.6)
25 (40.3) 4 (6.5) 21 (33.9) 5 (8.1) 6 (9.7) 29 (46.8)
CHF indicates congestive heart failure.
any evidence of atrial contraction. In the control group, the indication of long-term anticoagulation for patients who had undergone bioprosthetic valve replacement or valve repair (n = 9) was persistent AF in all patients. For the other five patients in the control group, anticoagulation therapy was not maintained because they showed spontaneous recovery of sinus rhythm persistently (n = 3), or the risk of stroke was expected to be less than moderate despite the presence of AF (n = 2). In summary, requirements for long-term anticoagulation therapy were less frequent in the Maze group for the entire cohort (67.1% vs. 91.2%, P = 0.001) and for a subgroup of patients who underwent bioprosthetic valve replacement or valve repair (30.6% vs. 64.3%, P = 0.052), compared to the control group. Although both groups showed significant improvements in New York Heart Association functional class after surgery (P b 0.001), patients in the Maze group showed superior functional status than those in the control group after adjustment using IPTW (P b 0.001, Fig. 5). 4. Discussion
3.5. Long-term anticoagulation therapy and functional status The anticoagulation status of 130 patients who survived beyond the late (N 6 months) postoperative period is detailed in Table 5. All patients who underwent mechanical valve replacement (n = 80) received lifelong anticoagulation therapy, regardless of treatment group. In the Maze group, indications of long-term warfarin anticoagulation for patients who had undergone bioprosthetic valve replacement or valve repair (n = 12) were as follows: (1) six patients had persistent (n = 4) or paroxysmal (n = 2) AF; (2) three patients had moderate– to-high risk of developing a stroke in the absence of AF; (3) two patients had reoperative mechanical valve replacement due to bioprosthetic valve failure; and (4) one patient showed junctional rhythm (a rhythm originating from AV junction without a P wave) persistently, without
In patients with limited ventricular function undergoing major cardiac surgery, postoperative recovery in the early period can be hampered by reduced cardiac function resulting from ischemic injuries and surgical manipulations. Consequently, LV dysfunction is a well-known risk factor for surgical mortality and morbidity in patients undergoing cardiac surgery [20–22]. In the same context, current risk calculation systems, both STS and EuroSCORE II, developed for predicting operative mortality after adult cardiac surgery, incorporate LV EF as one of the key covariates in the calculation of the projected mortality [19,23]. In patients with poorly functioning ventricles, overall hemodynamic performance can be further deteriorated by the presence of AF because atrial contraction contributes to overall cardiac output by up to 30% [24,25]. In this regard, performing the Maze procedure can contribute
Fig. 2. Postoperative cardiac rhythm status at each postoperative year in the Maze (A) and the control (B) groups. PPM, AF and NSR indicate permanent pacemaker rhythm, atrial fibrillation and normal sinus rhythm, respectively.
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
Fig. 3. Unadjusted curves for death (A) and the composite of clinical end points (B).
to improved hemodynamic performance and, hence, to overall clinical benefits by restoring sinus rhythm and synchronous contraction of the atrio-ventricular system [12]. Furthermore, the Maze procedure may help to circumvent the unstable early postoperative period in these high-risk patients, during which LV function tends to deteriorate due to cardiac insults caused by the surgical procedure. However, concerns exist with regard to performing the Maze procedure in patients with significant LV dysfunction. Adding the procedure inevitably increases cardiac ischemic and cardiopulmonary bypass Table 4 Adjusted hazard ratios for clinical outcomes of the maze procedure compared with controls.
Death
Thromboembolism
CHF
Death, thromboembolism and CHF
Crude Propensity score IPTW Crude Propensity score IPTW Crude Propensity score IPTW Crude Propensity score IPTW
HR
95% CI
P
0.33 0.49 0.39 0.28 0.33 1.25 0.24 0.15 0.19 0.30 0.36 0.28
0.15–0.72 0.19–1.26 0.16–0.93 0.05–1.44 0.05–2.28 0.48–3.26 0.05–1.16 0.02–0.98 0.05–0.70 0.15–0.59 0.16–0.81 0.14–0.57
0.005 0.14 0.033 0.13 0.26 0.64 0.076 0.047 0.013 b0.001 0.014 0.0011
CHF indicates congestive heart failure; IPTW, inverse probability of treatment weighting.
335
Fig. 4. Adjusted curves for death (A) and the composite of clinical end points (B).
times, which may further deteriorate cardiac and end-organ function in the early postoperative periods. Increased procedural complexity by adding the Maze procedure may put patients at a greater risk in the early postoperative period. For these reasons, performing the Maze procedure is regarded as more challenging than undertaking structural heart surgery alone for patients with LV dysfunction. In the updated guidelines on catheter and surgical ablation of AF, concomitant surgical AF ablation in the setting of cardiac surgery is indicated only in “symptomatic AF” patients, but there are no guidelines for patients “without symptoms” [13]. Given that a significant proportion of patients with AF have no symptoms relevant to AF, recent guidelines need to be further refined with regard to undertaking AF ablation in high-risk patients not showing symptoms of AF. Furthermore, the recent increase in surgical risk in patients undergoing cardiac surgery necessitates better definitions and indications about these issues.
Table 5 Long-term anticoagulation status of patients who survived longer than 6 months postoperatively.
Mechanical valve implantation Bioprosthetic valve implantation or repair Entire cohort
Maze group
Control group
n = 73
n = 57
37/37 (100%) 11/36 (30.6%) 48/73 (65.8%)
43/43 (100%) 9/14 (64.3%) 52/57 (91.2%)
P
NA 0.052 0.001
336
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
Fig. 5. New York Heart Association (NYHA) functional class groups before surgery and at the end of follow-up (N6 months). Pre- and postoperative statuses were adjusted by inverse-probability of treatment weighting.
Recently, Ad et al. [12] reported their series of 42 patients with LVEF of less than 40% who underwent surgical AF ablation. Although the expected mortality rate, as assessed by the EuroSCORE, was 7.5%, surgical mortality occurred in only one patient (2.4%). Furthermore, the surgery resulted in significant improvements in LV systolic function as well as in quality of life. Consequently, authors concluded that surgical AF ablation should always be considered for this high-risk cohort, especially at centers with experienced surgeons. With regard to evaluating patients with LV dysfunction, the cited study may be the first to analyze the impact of the Maze procedure; however, it was not possible to determine whether the desirable clinical effects were a result of the Maze procedure or a consequence of correcting structural heart diseases because no control group was included in the study. To address this, we analysed clinical outcomes in a larger population according to a combination of the Maze procedure. Significant survival benefits were observed in those receiving the Maze procedure in the present study. Although several groups report superior survival rates in patients who maintained sinus rhythm compared with those who experienced AF [26], this considerable survival difference resulting from the Maze procedure is previously unreported. This finding was not even observed in our previous study involving 569 patients who underwent mechanical valve replacement in the presence of AF, in which a reduction in thromboembolic events, but not an improvement in overall survival, was observed in those receiving the Maze procedure [11]. Based on the findings of several studies, the differences in survival when restoring sinus rhythm may be attributable to discrepancies in subject populations and also the tools used for rhythm conversion [26,27]. First, surgical AF ablation, compared with catheter ablation or rhythm control strategy using various medications, is much more effective in restoring and maintaining sinus rhythm [7,27,28]. Second, patients with organic heart disease and severe LV dysfunction may be more affected by rhythm status than those without these factors since loss of atrial contraction can predispose patients to overt “low cardiac output” or increase the overall risk profiles in vulnerable patients [6,12]. In the present study, the risks of thromboembolism and CHF were lower in the Maze group than in the control group, and the two groups showed further marked discrepancies in eventfree survival. We believe that the incorporation of atrial kicking into the cardiac output by the Maze procedure improves hemodynamics with results in overall clinical benefits. These assumptions are strongly supported by our previous studies, in that around 80% of patients who restored sinus rhythm by the Maze procedure were found to restore effective atrial contraction as evidenced by (1) the presence of effective trans-mitral/trans-tricuspid atrial kicking wave (A wave) and (2) the
quality of “E/A ratio” (peak velocity of early diastolic filling [E wave] divided by peak velocity of A wave) [29,30]. Furthermore, performing the Maze procedure did not increase early mortality or morbidity, which indicates that the procedure can be carried out with an acceptable safety profile, even in this high-risk population. Long-term anticoagulation therapy is also an important issue for patients undergoing heart valve surgery. Even in the setting of bioprosthetic valve implantation or valve repair, about 30% of patients in this study required long-term anticoagulation therapy, even when receiving the Maze procedure. The most common indication of warfarin anticoagulation was the absence of atrial contraction either by AF or junctional rhythm (a rhythm originating from AV junction without a P wave), followed by the individual patient's comorbidity. Nevertheless, the need for anticoagulation was reduced by the Maze procedure. Although the rate of bleeding secondary to anticoagulation therapy could not be compared due to the small number of patients in this study, it is likely to show the benefits of the Maze procedure in larger populations. Given that maintaining lifelong anticoagulation therapy is directly related to quality of life, the Maze procedure will provide further benefits in avoiding warfarinization in patients undergoing bioprosthetic valve implantation and valve repair. 4.1. Limitations This study is subject to the limitations inherent to retrospective analyses of observational data. Although statistical models were used to adjust for treatment selection bias, our results may have been affected by unmeasured confounders, procedure bias or detection bias. Among patients with valvular AF, the prevalence of moderate-tosevere LV dysfunction was very low. Consequently, even though we extracted the subject population from a large population, the sample size was relatively small to draw robust conclusions. 5. Conclusions For patients with valvular AF and moderate-to-severe LV dysfunction undergoing major cardiac surgery, the combination of the Maze procedure was associated with an improvement in long-term survival and a reduction in serious adverse outcomes compared with cardiac surgery alone. Furthermore, the Maze procedure reduced the need for long-term anticoagulation therapy in this population. Therefore, surgical AF ablation should always be considered as a concomitant procedure during open cardiac surgery to improve long-term outcomes, even in a high-risk cohort. Conclusions regarding this issue need further verification through prospective randomized trials. Acknowledgement The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology. References [1] Cox JL, Boineau JP, Schuessler RB, et al. Successful surgical treatment of atrial fibrillation. Review and clinical update. JAMA 1991;266:1976–80. [2] Nakajima H, Kobayashi J, Bando K, et al. The effect of cryo-maze procedure on early and intermediate term outcome in mitral valve disease: case matched study. Circulation 2002;106:I46–50. [3] Bando K, Kobayashi J, Kosakai Y, et al. Impact of Cox maze procedure on outcome in patients with atrial fibrillation and mitral valve disease. J Thorac Cardiovasc Surg 2002;124:575–83. [4] 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. [5] Lall SC, Melby SJ, Voeller RK, et al. The effect of ablation technology on surgical outcomes after the Cox–Maze procedure: a propensity analysis. J Thorac Cardiovasc Surg 2007;133:389–96. [6] Gillinov AM, Sirak J, Blackstone EH, et al. The Cox maze procedure in mitral valve disease: predictors of recurrent atrial fibrillation. J Thorac Cardiovasc Surg 2005;130:1653–60.
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337 [7] Weimar T, Schena S, Bailey MS, et al. The Cox–Maze procedure for lone atrial fibrillation: a single-center experience over 2 decades. Circ Arrhythm Electrophysiol 2012;5:8–14. [8] Stulak JM, Dearani JA, Sundt III TM, Daly RC, Schaff HV. Ablation of atrial fibrillation: comparison of catheter-based techniques and the Cox–Maze III operation. Ann Thorac Surg 2011;91:1882–8. [9] Stulak JM, Schaff HV, Dearani JA, Orszulak TA, Daly RC, Sundt III TM. Restoration of sinus rhythm by the Maze procedure halts progression of tricuspid regurgitation after mitral surgery. Ann Thorac Surg 2008;86:40–4. [10] Cox JL, Ad N, Palazzo T. Impact of the maze procedure on the stroke rate in patients with atrial fibrillation. J Thorac Cardiovasc Surg 1999;118:833–40. [11] Kim JB, Moon JS, Yun SC, et al. Long-term outcomes of mechanical valve replacement in patients with atrial fibrillation: impact of the maze procedure. Circulation 2012;125:2071–80. [12] Ad N, Henry L, Hunt S. The impact of surgical ablation in patients with low ejection fraction, heart failure, and atrial fibrillation. Eur J Cardiothorac Surg 2012;40:70–6. [13] Calkins H, Kuck KH, Cappato R, et al. HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for patient selection, procedural techniques, patient management and follow-up, definitions, end points, and research trial design: a report of the Heart Rhythm Society (HRS) Task Force on Catheter and Surgical Ablation of Atrial Fibrillation. Developed in partnership with the European Heart Rhythm Association (EHRA), a registered branch of the European Society of Cardiology (ESC) and the European Cardiac Arrhythmia Society (ECAS); and in collaboration with the American College of Cardiology (ACC), American Heart Association (AHA), the Asia Pacific Heart Rhythm Society (APHRS), and the Society of Thoracic Surgeons (STS). Endorsed by the governing bodies of the American College of Cardiology Foundation, the American Heart Association, the European Cardiac Arrhythmia Society, the European Heart Rhythm Association, the Society of Thoracic Surgeons, the Asia Pacific Heart Rhythm Society, and the Heart Rhythm Society. Heart Rhythm 2012;9:632–96 [e21]. [14] 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;92:1397–404. [15] Lauer MS, Blackstone EH, Young JB, Topol EJ. Cause of death in clinical research: time for a reassessment? J Am Coll Cardiol 1999;34:618–20. [16] Akins CW, Miller DC, Turina MI, et al. Guidelines for reporting mortality and morbidity after cardiac valve interventions. J Thorac Cardiovasc Surg 2008;135:732–8. [17] Robins JM, Hernan MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology 2000;11:550–60.
337
[18] Normand SL. Evaluating the optimal timing of angiography: landmark or off the mark? Circulation 2007;116:2656–7. [19] Nashef SA, Roques F, Sharples LD, et al. EuroSCORE II. Eur J Cardiothorac Surg 2012;41:734–44. [20] Lee EM, Shapiro LM, Wells FC. Mortality and morbidity after mitral valve repair: the importance of left ventricular dysfunction. J Heart Valve Dis 1995;4:460–8. [21] Enriquez-Sarano M, Tajik AJ, Schaff HV, Orszulak TA, Bailey KR, Frye RL. Echocardiographic prediction of survival after surgical correction of organic mitral regurgitation. Circulation 1994;90:830–7. [22] Bonow RO, Carabello BA, Kanu C, et al. ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease): developed in collaboration with the Society of Cardiovascular Anesthesiologists: endorsed by the Society for Cardiovascular Angiography and Interventions and the Society of Thoracic Surgeons. Circulation 2006;114:e84-231. [23] O'Brien SM, Shahian DM, Filardo G, et al. The Society of Thoracic Surgeons 2008 cardiac surgery risk models: part 2–isolated valve surgery. Ann Thorac Surg 2009;88: S23–42. [24] Middlekauff HR, Stevenson WG, Stevenson LW. Prognostic significance of atrial fibrillation in advanced heart failure. A study of 390 patients. Circulation 1991;84:40–8. [25] Pritchett EL. Management of atrial fibrillation. N Engl J Med 1992;326:1264–71. [26] Pedersen OD, Bagger H, Keller N, Marchant B, Kober L, Torp-Pedersen C. Efficacy of dofetilide in the treatment of atrial fibrillation-flutter in patients with reduced left ventricular function: a Danish investigations of arrhythmia and mortality on dofetilide (diamond) substudy. Circulation 2001;104:292–6. [27] Wyse DG, Waldo AL, DiMarco JP, et al. A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med 2002;347:1825–33. [28] Boersma LV, Castella M, van Boven W, et al. Atrial fibrillation catheter ablation versus surgical ablation treatment (FAST): a 2-center randomized clinical trial. Circulation 2012;125:23–30. [29] Kim JB, Lee SH, Jung SH, et al. The influence of postoperative mitral valve function on the late recurrence of atrial fibrillation after the maze procedure combined with mitral valvuloplasty. J Thorac Cardiovasc Surg 2010;139:1170–6. [30] Kim JB, Yun TJ, Chung CH, Choo SJ, Song H, Lee JW. Long-term outcome of modified maze procedure combined with mitral valve surgery: analysis of outcomes according to type of mitral valve surgery. J Thorac Cardiovasc Surg 2010;139:111–7.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Maze procedure in patients with left ventricular dysfunction Joon Bum Kim 1, Byung Kwon Chong 1, Sung-Ho Jung 1, Suk Jung Choo 1, Cheol Hyun Chung 1, Jae Won Lee ⁎,1 Department of Thoracic and Cardiovascular Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea
a r t i c l e
i n f o
Article history: Received 21 February 2013 Received in revised form 13 September 2013 Accepted 2 November 2013 Available online 12 November 2013 Keywords: Atrial fibrillation Ablation Surgery Heart failure Survival
a b s t r a c t Background: The risks and benefits of a concomitant Maze procedure for patients with LV dysfunction undergoing major cardiac surgery have not yet been elucidated. This study aimed to evaluate the clinical impacts of the Maze procedure in patients with atrial fibrillation and left ventricular (LV) dysfunction. Methods: Between January 1999 and March 2011, a total of 139 patients (mean age 52.7 ± 12.3 years, 54 females) with valvular atrial fibrillation (AF) and an LV ejection fraction (EF) of 40% or less underwent open heart surgery with (n = 77) or without (n = 62) a concomitant Maze procedure. We compared adverse outcomes (death and composite of death, thromboembolic events and congestive heart failure [CHF]) during a median follow-up period of 66.0 months (inter-quartile range, 27.5–106.9 months). Results: Adverse events occurred in 41 patients, including 36 deaths, seven thromboembolic events and eight hospitalizations due to CHF. After adjustment for baseline profiles with the use of propensity scores and inverse probability weighting, patients who had the Maze procedure were at lower risks of death (hazard ratio, 0.39; 95% confidence interval, 0.16–0.93; P = 0.033) and composite adverse outcomes (hazard ratio, 0.28; 95% confidence interval, 0.14–0.57; P = 0.017) than those not undergoing the Maze procedure. Furthermore, the Maze procedure resulted in superior functional status (P b 0.001) and reduced the need for long-term anticoagulation therapy (67.1% vs. 91.2%, P = 0.001). Conclusion: Performing the Maze procedure on patients with valvular AF and LV dysfunction reduced serious adverse outcomes and the need for long-term anticoagulation therapy when compared to cardiac surgery alone without the Maze procedure. Crown Copyright © 2013 Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Since the introduction of the Cox–Maze procedure, its technical evolution has included using alternative energy sources, replacing cutand-sew methods and creating simpler ablation lesion sets rather than complex “maze” lesions [1–5]. Consequently, the use of surgical atrial fibrillation (AF) ablation has been prompted in real clinical practice to play a substantial role in selected patients. The efficacy of the Maze procedure in the eliminating AF is well proven in many published studies with the use of the classic Cox–Maze III procedure or equivalent AF ablation procedures using cryothermia or radiofrequency ablation [6–9]. Aside from sinus rhythm restoration, clinical advantages of the Maze procedure include a reduction in thromboembolic risks and improvements in hemodynamic profiles [10–12]. Despite the benefits of the Maze procedure, its performance in patients with marginal left ventricular (LV) function undergoing major cardiac surgery has been controversial. This is because the procedure
⁎ Corresponding author at: Department of Thoracic and Cardiovascular Surgery, Asan Medical Center, University of Ulsan College of Medicine, 388-1 Pungnap-dong Songpa-gu, Seoul 138-736, South Korea. Tel.: +82 2 3010 3584; fax: +82 2 3010 6966. E-mail address: [email protected] (J.W. Lee). 1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
may increase surgical risks as a consequence of increased cardiac ischemic time and more extensive procedures for these vulnerable patients. These concerns well appear in recent consensus and guidelines on catheter and surgical ablation of AF, in that in the setting of major cardiac surgery, concomitant surgical AF ablation can be indicated only in selected patients in whom the ablation can be performed with “minimal risk” if symptoms of AF do not exist [13]. In this regard, there are no relevant data in the literature that assess the risks and benefits of the Maze procedure in patients with significant LV dysfunction undergoing major cardiac surgery. We therefore evaluated the long-term clinical outcomes of patients with valvular AF and LV dysfunction, undergoing major cardiac surgery with or without a concomitant Maze procedure. 2. Methods 2.1. Patients We used our institution's database, which prospectively registers all patients undergoing major cardiac surgery. Out of a total of 3,123 patients who underwent mitral valve (MV) surgery between January 1999 and March 2011 at the Asan Medical Center, Seoul, Korea, we identified 139 patients who had AF and an LV ejection fraction (EF) of 40% or less. These patients formed the subject population of this study. Of these patients, 77 concomitantly underwent the Maze procedure (Maze group) while 62 underwent cardiac surgery without the Maze procedure (control group). The decision to perform the concomitant Maze procedure was influenced by each patient's risk profiles, such as
0167-5273/$ – see front matter. Crown Copyright © 2013 Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijcard.2013.11.008
332
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
the severity of LV dysfunction, the types of valve surgery and implanted prosthesis and overall estimated surgical risks but was finally at the discretion of the attending surgeon. This study was approved by our institutional ethics committee/review board, which waived the requirement for informed consent due to the retrospective nature of the study. 2.2. Surgical procedure In the Maze group, the AF ablation procedure was performed either by cryoablation (n = 66) or by microwave ablation (n = 11), with the lesion sets for the Maze procedure being created as those described previously [4,14]. For the right side ablation, oblique right atriotomy was made typically under beating heart cardiopulmonary bypass (CPB) support. Then, the cavo-tricuspid isthmus isolation was achieved by two or three ablation lines. Another linear lesion from posterior end of atriotomy to superior vena cava was made. The left-side procedure began after aortic cross clamping (ACC) through a longitudinal right-sided left atriotomy. Left-side ablation was composed of pulmonary vein (PV) isolation lesions and two posterior linear lesions (1) one from PV isolation line to left atrial (LA) appendage orifice and (2) one from PV isolation line to MV annulus. Additional epicardial coronary sinus ablation was performed at the opposite side of the MV annular lesion. The LA size was reduced by resecting the redundant atrial tissue off the posterior LA wall parallel to the posterior mitral annulus, with a target width less than 4 cm between the inferior pulmonary vein openings and the MV annulus. 2.3. Postoperative management Patients who underwent valve repair or bioprosthetic valve implantation were routinely anticoagulated with warfarin for 3 to 6 months postoperatively with a target international normalized ratio (INR) of 1.5 to 2.5. The maintenance of the anticoagulation therapy thereafter was determined according to individual risk factors of thromboembolism. Postoperative AF or the absence of effective atrial contraction resulted in prolonging the anticoagulation therapy until both sinus rhythm and atrial contractility were restored. For patients with mechanical valve implantation, achieving an INR between 2.0 and 3.0 was the aim, regardless of cardiac rhythm status. Anticoagulation therapy was adjusted during outpatient visits at 3-month intervals. Generally, follow-up electrocardiography assessments were performed at 3, 6, 12, 18 and 24 months after surgery, and annually thereafter. For patients with restored sinus rhythm, 24-h Holter monitoring was done to validate “free-of-AF”. Postoperative rhythm management strategies differed according to the performance of the Maze procedure. In the Maze group, for instance, postoperative atrial tachyarrhythmia was managed with class I/III anti-arrhythmic medications or electrical cardioversion in an effort to restore sinus rhythm (“rhythm control strategy”). If the patients failed to achieve “AF elimination” despite these treatments, they were switched to a “rate control strategy” to control ventricular rate (with the use of digitalis, β-blockers or calcium channel blockers) in combination with anticoagulation therapy. The principal rhythm management for the control group was the “rate control strategy” with the only exception of those who showed spontaneous recovery of sinus rhythm followed by AF recurrence. These patients were managed with the “rhythm control strategy” as the initial treatment intention. 2.4. Definitions and Follow-up The primary end point was defined as all-cause death. All-cause death rather than cardiac death was chosen as the primary end point because it is the most robust and unbiased index, requiring no adjudication to avoid inaccurate or biased documentation and clinical assessments [15]. The secondary end point was defined as a composite of all-cause death, thromboembolic events and congestive heart failure (CHF) requiring hospitalization. Thromboembolic events, which may manifest as neurologic events (non-hemorrhagic stroke or transient ischemic attack) or non-cerebral (peripheral) embolic events were clinically diagnosed [16]. Hospitalization due to CHF was defined as an unplanned admission for the management of CHF in cases with resting dyspnea and pulmonary edema requiring intravenous diuretics therapy. Clinical follow-ups were performed at 3- to 6-month intervals as outpatient clinic visits, according to the requirements for anticoagulation therapy. To validate all information regarding mortality, data on vital status and the dates and causes of death were obtained up to March 31, 2012, from the Korean National Registry of Vital Statistics. All deaths were considered of cardiac origin unless a non-cardiac origin was established clinically. 2.5. Statistical analysis Categorical variables, presented as frequencies and percentages, were compared using the chi-square test or Fisher's exact test. Continuous variables, expressed as mean ± SD or median with range, were compared using the Student's unpaired t-test or the Mann–Whitney U-test, as appropriate. To reduce the effects of treatment selection bias and potential confounding factors, we adjusted for differences in baseline characteristics by weighted Cox regression models and inverse probability of treatment weighting (IPTW) [17,18]. With this technique, weights for patients receiving the concomitant Maze procedure were the inverse of propensity scores, and weights for patients receiving cardiac surgery alone, without the Maze procedure, were the inverse of “1 minus propensity score.” Propensity scores were estimated without considering the outcome variables, using multiple logistic regression models.
Pre-specified covariates listed in Table 1 were included in the models for the Maze versus the control groups. The discrimination and calibration abilities of the propensity score model were assessed by means of C statistics and the Hosmer–Lemeshow test. The model had a C statistic of 0.843 and a Hosmer–Lemeshow goodness of fit P value of 0.561, indicating that this model was well calibrated with reasonable discrimination. All reported P values were two-sided, and a value of P b 0.05 was considered statistically significant. SAS software, version 9.1 (SAS Institute, Cary, NC, USA) was used for the statistical analysis. The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
3. Results 3.1. Baseline characteristics The mean age at surgery was 52.7 ± 12.3 years, and 54 patients (38.8%) were female. Table 1 summarizes the baseline demographic and clinical characteristics of all patients. In summary, patients in the control group were more likely to have a history of previous cardiac surgery and were more likely to undergo valve replacement surgery, particularly mechanical valve replacement, than those in the Maze group. LV EF was lower and LA size was larger in the control group. An estimated risk of surgical mortality, represented by the EuroSCORE II [19], was significantly higher in the control group than in the Maze group. As expected, patients in the Maze group had a higher probability of undergoing the Maze procedure than did those in the control group (Fig. 1A). After adjustment using IPTW, both groups were fairly balanced in the distribution of the propensity scores (Fig. 1B). Consequently, all baseline profiles were well balanced, without a significant difference between the two groups after the adjustment (Table 1, rightmost column). 3.2. Early operative results and postoperative rhythm status The ACC (110.5 ± 51.3 min versus 104.7 ± 38.0 min; P = 0.54) and CPB times (177.1 ± 94.6 versus 159.0 ± 62.8 min; P = 0.088) tended to be longer in the Maze group than in the control group, but without a statistical significance. There were six (4.3%) early deaths, and 28 patients (20.1%) experienced major complications during postoperative hospitalization. There were no significant differences in early mortality or major morbidity rates between the two groups (Table 2). During the follow-up, patients in the Maze group showed superior rhythm outcomes in terms of freedom from AF and maintenance of sinus rhythm compared with those in the control group (Fig. 2; P b 0.001 for all postoperative time points). 3.3. Clinical end points Follow-up was complete in all patients (100%), with a median duration of 66.0 months (interquartile range 27.5–106.9 months, 805.2 patient-years). Throughout the study period, there were 30 late deaths (45.6%) including 27 cardiac deaths. Non-cardiac causes of death were malignancy in two patients, and suicide in one. Seven patients (10.7%) experienced thromboembolic events including non-hemorrhagic stroke in six patients, and transient ischemic attack in one patient. Eight patients (12.3%) were readmitted due to CHF. Table 3 summarizes the clinical end points according to the two treatment groups. The linearized rates of death, thromboembolic events and CHF were 4.4%, 0.9% and 1.0% per patient year, respectively, in the entire cohort. In the Maze group, these rates were 2.3%, 0.4% and 0.4%, respectively, and in the control group, they were 7.1%, 1.4% and 1.7%, respectively (P = 0.005, 0.13 and 0.073, respectively). Fig. 3 depicts the unadjusted rates of death and the composite of secondary end points in the two groups. The control group showed significantly higher rates of death and composite end points compared to the Maze group (P = 0.005 and b0.001, respectively).
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
333
Table 1 Baseline characteristics of all patients. Unadjusted
Age, years Male gender (%) NYHA functional class I II III IV Diabetes mellitus, n (%) Hypertension, n (%) History of thromboembolism, n (%) Previous cardiac surgery, n (%) CHADS2 score 1 2 ≥3 Type of AF, n (%) Paroxysmal Persistent (≤1 year) Longstanding persistent Echocardiographic data LV ejection fraction, % LV end-systolic dimension, mm LV end-diastolic dimension, mm Left atrial diameter, mm Peak PAP, mmHg Tricuspid regurgitation, n (%) None -trace Mild Moderate Severe MV operation type Repair Mechanical prosthetic replacement Bioprosthetic replacement Combined surgery AV replacement Mechanical prosthesis Bioprosthesis Coronary artery bypassing Tricuspid valve repair Estimated GFR, mL/min/1.73 m2 EuroSCORE II, %†
Adjusted by IPTW*
Maze group (n = 77)
Control group (n = 62)
P value
Maze group (n = 77)
Control group (n = 62)
P value
51.6 ± 12.5 45 (58.4)
54.0 ± 11.9 40 (64.5)
0.25 0.49 0.084
52.7 ± 12.1 46 (59.7)
52.2 ± 12.8 37 (59.7)
0.80 N0.99 0.39
8 (10.4) 24 (31.2%) 37 (48.1) 8 (10.4) 8 (10.4) 10 (13.0) 14 (18.2) 2 (2.6)
1 (1.6) 23 (37.1) 26 (41.9) 12 (19.4) 7 (11.3) 4 (6.5) 8 (12.9) 7 (11.3)
6 (7.8) 26 (33.8) 37 (48.1) 8 (10.4) 11 (14.3) 10 (13.0) 14 (18.2) 2 (2.6)
1 (1.6) 25 (40.3) 29 (46.8) 7 (11.3) 14 (22.6) 12 (19.4) 11 (11.7) 4 (6.5)
56 (72.7) 15 (19.5) 6 (7.8)
48 (77.4) 10 (16.1) 4 (6.5)
53 (67.9) 17 (21.8) 8 (10.3)
38 (61.3) 12 (19.4) 12 (19.4)
2 (2.6) 47 (61.0) 28 (36.4)
1 (1.6) 34 (54.8) 27 (43.5)
1 (1.3) 47 (61.8) 28 (36.8)
1 (1.6) 34 (55.7) 27 (44.3)
36.0 ± 48.0 ± 61.3 ± 57.3 ± 45.9 ±
33.5 ± 6.3 50.89 ± 11.0 64.0 ± 10.9 62.8 ± 10.3 50.2 ± 16.4
35.5 48.2 61.2 58.5 48.1
34.3 ± 49.5 ± 63.0 ± 60.1 ± 52.8 ±
0.87 0.20 0.40 0.038 0.77
0.66
3.6 8.6 10.2 8.4 17.7
32 (41.6) 19 (24.7) 14 (18.2) 12 (15.6)
21 (33.9) 11 (17.7) 13 (21.0) 17 (27.4)
34 (44.2) 36 (46.8) 7 (9.1)
12 (19.4) 41 (66.1) 9 (14.5)
16 (20.8) 14 (18.2) 2 (2.6) 14 (18.2) 35 (45.5) 73.3 ± 19.5 13.1 ± 8.0
29 (46.8) 21 (33.9) 8 (12.9) 7 (11.3) 24 (38.7) 71.6 ± 25.4 18.4 ± 13.7
0.006 0.096 0.14 0.001 0.15 0.38
0.48
± ± ± ± ±
3.8 8.6 10.1 8.2 19.7
5.1 10.7 11.7 8.6 15.6
30 (39.0) 17 (22.1) 15 (19.5) 15 (19.5)
22 (35.4) 19 (30.6) 9 (14.5) 12 (19.4)
27 (35.1) 41 (53.2) 9 (11.7)
23 (37.1) 33 (53.2) 6 (9.7)
18 (23.7) 16 (21.1) 2 (2.6) 12 (15.6) 33 (42.9) 73.0 ± 19.9 14.3 ± 9.1
18 (29.0) 13 (21.0) 5 (8.1) 10 (16.1) 29 (46.8) 67.6 ± 37.6 15.9 ± 11.2
0.008
0.003
0.26 0.42 0.66 0.008
0.27 0.35 0.95 0.41 0.28
0.11 0.45 0.35 0.25 0.12 0.80
0.92
0.35
N0.99 0.73 0.31 0.35
*Numbers of patients for categorical variables were rounded off to the nearest whole number. †In calculating the EuroSCORE II, the concomitant Maze procedure was not considered in the “weight of procedure.” IPTW indicates inverse probability treatment weighting; NYHA, New York Heart Association; AF, atrial fibrillation; LV, left ventricle; PAP, pulmonary artery pressure; MV, mitral valve; AV, aortic valve; GFR, glomerular filtration rate.
Table 4 summarizes the adjusted hazard ratios for clinical end points in the Maze group compared to the control group. Patients undergoing the concomitant Maze procedure were at significantly lower risks of
death, and the composite of death, thromboembolism and CHF, than those not receiving the procedure. Adjusted rates of clinical end points are illustrated in Fig. 4.
Fig. 1. Box plots for propensity scores in the Maze and the control groups. The propensity score is the probability given baseline variables that any patient in either group would be selected for a concomitant Maze procedure. (A) Patients in the Maze group have higher propensity scores than those in the control group. (B) After inverse-probability treatment weighting, the distribution of the propensity scores is fairly balanced between the two groups.
334
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
Table 2 Early postoperative outcomes.
Table 3 Overall adverse outcomes during follow-up. Maze
Control
P
N (%)
(n = 77) (n = 62) Unadjusted Adjusted
Early death Major complications LCOS requiring mechanical support New dialysis Bleeding reoperation Pneumonia Multi-organ failure Permanent pacemaker insertion Wound revision Prolonged ventilator support (N24 hr)
2 (2.6) 4 (6.5) 0.41 15 (19.5) 13 (21.0) 0.83 4 (5.2) 3 (4.8) 0.92
0.66 0.84 N0.99
6 (7.8) 5 (6.5) 2 (2.6) 2 (2.6) 2 (2.6) 3 (3.9) 13 (16.9)
0.19 0.73 0.59 N0.99 0.14 0.13 0.39
3 (4.8) 4 (6.5) 2 (3.2) 0 (0) 2 (3.2) 1 (1.6) 12 (19.4)
0.48 0.99 0.83 0.20 0.83 0.42 0.83
LCOS indicates low cardiac output syndrome requiring a mechanical support.
3.4. Other adverse outcomes Nineteen cases of other adverse clinical events occurred in 15 patients (10.8%): nine patients (11.7%) in the Maze group and six patients (9.7%) in the control group. These adverse outcomes included hemorrhagic complications related to anticoagulation therapy in eight patients, heart valve reoperation in five patients, infective endocarditis in two patients, ventricular arrhythmia in one patient, acute myocardial infarction in one patient and a requirement for percutaneous coronary stenting in one patient. Cumulative event rates at 5 years were 18.9 ± 6.7% in the Maze group and 13.1 ± 5.1% in the control group (unadjusted and adjusted P = 0.91 and 0.93, respectively).
Patients, n (%)
Maze group
Control group
Death Early (≤30 days) death Late death Thromboembolism CHF Death, thromboembolism and CHF
11 (14.3) 2 (2.6) 9 (11.7) 2 (2.6) 2 (2.6) 12 (15.6)
25 (40.3) 4 (6.5) 21 (33.9) 5 (8.1) 6 (9.7) 29 (46.8)
CHF indicates congestive heart failure.
any evidence of atrial contraction. In the control group, the indication of long-term anticoagulation for patients who had undergone bioprosthetic valve replacement or valve repair (n = 9) was persistent AF in all patients. For the other five patients in the control group, anticoagulation therapy was not maintained because they showed spontaneous recovery of sinus rhythm persistently (n = 3), or the risk of stroke was expected to be less than moderate despite the presence of AF (n = 2). In summary, requirements for long-term anticoagulation therapy were less frequent in the Maze group for the entire cohort (67.1% vs. 91.2%, P = 0.001) and for a subgroup of patients who underwent bioprosthetic valve replacement or valve repair (30.6% vs. 64.3%, P = 0.052), compared to the control group. Although both groups showed significant improvements in New York Heart Association functional class after surgery (P b 0.001), patients in the Maze group showed superior functional status than those in the control group after adjustment using IPTW (P b 0.001, Fig. 5). 4. Discussion
3.5. Long-term anticoagulation therapy and functional status The anticoagulation status of 130 patients who survived beyond the late (N 6 months) postoperative period is detailed in Table 5. All patients who underwent mechanical valve replacement (n = 80) received lifelong anticoagulation therapy, regardless of treatment group. In the Maze group, indications of long-term warfarin anticoagulation for patients who had undergone bioprosthetic valve replacement or valve repair (n = 12) were as follows: (1) six patients had persistent (n = 4) or paroxysmal (n = 2) AF; (2) three patients had moderate– to-high risk of developing a stroke in the absence of AF; (3) two patients had reoperative mechanical valve replacement due to bioprosthetic valve failure; and (4) one patient showed junctional rhythm (a rhythm originating from AV junction without a P wave) persistently, without
In patients with limited ventricular function undergoing major cardiac surgery, postoperative recovery in the early period can be hampered by reduced cardiac function resulting from ischemic injuries and surgical manipulations. Consequently, LV dysfunction is a well-known risk factor for surgical mortality and morbidity in patients undergoing cardiac surgery [20–22]. In the same context, current risk calculation systems, both STS and EuroSCORE II, developed for predicting operative mortality after adult cardiac surgery, incorporate LV EF as one of the key covariates in the calculation of the projected mortality [19,23]. In patients with poorly functioning ventricles, overall hemodynamic performance can be further deteriorated by the presence of AF because atrial contraction contributes to overall cardiac output by up to 30% [24,25]. In this regard, performing the Maze procedure can contribute
Fig. 2. Postoperative cardiac rhythm status at each postoperative year in the Maze (A) and the control (B) groups. PPM, AF and NSR indicate permanent pacemaker rhythm, atrial fibrillation and normal sinus rhythm, respectively.
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
Fig. 3. Unadjusted curves for death (A) and the composite of clinical end points (B).
to improved hemodynamic performance and, hence, to overall clinical benefits by restoring sinus rhythm and synchronous contraction of the atrio-ventricular system [12]. Furthermore, the Maze procedure may help to circumvent the unstable early postoperative period in these high-risk patients, during which LV function tends to deteriorate due to cardiac insults caused by the surgical procedure. However, concerns exist with regard to performing the Maze procedure in patients with significant LV dysfunction. Adding the procedure inevitably increases cardiac ischemic and cardiopulmonary bypass Table 4 Adjusted hazard ratios for clinical outcomes of the maze procedure compared with controls.
Death
Thromboembolism
CHF
Death, thromboembolism and CHF
Crude Propensity score IPTW Crude Propensity score IPTW Crude Propensity score IPTW Crude Propensity score IPTW
HR
95% CI
P
0.33 0.49 0.39 0.28 0.33 1.25 0.24 0.15 0.19 0.30 0.36 0.28
0.15–0.72 0.19–1.26 0.16–0.93 0.05–1.44 0.05–2.28 0.48–3.26 0.05–1.16 0.02–0.98 0.05–0.70 0.15–0.59 0.16–0.81 0.14–0.57
0.005 0.14 0.033 0.13 0.26 0.64 0.076 0.047 0.013 b0.001 0.014 0.0011
CHF indicates congestive heart failure; IPTW, inverse probability of treatment weighting.
335
Fig. 4. Adjusted curves for death (A) and the composite of clinical end points (B).
times, which may further deteriorate cardiac and end-organ function in the early postoperative periods. Increased procedural complexity by adding the Maze procedure may put patients at a greater risk in the early postoperative period. For these reasons, performing the Maze procedure is regarded as more challenging than undertaking structural heart surgery alone for patients with LV dysfunction. In the updated guidelines on catheter and surgical ablation of AF, concomitant surgical AF ablation in the setting of cardiac surgery is indicated only in “symptomatic AF” patients, but there are no guidelines for patients “without symptoms” [13]. Given that a significant proportion of patients with AF have no symptoms relevant to AF, recent guidelines need to be further refined with regard to undertaking AF ablation in high-risk patients not showing symptoms of AF. Furthermore, the recent increase in surgical risk in patients undergoing cardiac surgery necessitates better definitions and indications about these issues.
Table 5 Long-term anticoagulation status of patients who survived longer than 6 months postoperatively.
Mechanical valve implantation Bioprosthetic valve implantation or repair Entire cohort
Maze group
Control group
n = 73
n = 57
37/37 (100%) 11/36 (30.6%) 48/73 (65.8%)
43/43 (100%) 9/14 (64.3%) 52/57 (91.2%)
P
NA 0.052 0.001
336
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337
Fig. 5. New York Heart Association (NYHA) functional class groups before surgery and at the end of follow-up (N6 months). Pre- and postoperative statuses were adjusted by inverse-probability of treatment weighting.
Recently, Ad et al. [12] reported their series of 42 patients with LVEF of less than 40% who underwent surgical AF ablation. Although the expected mortality rate, as assessed by the EuroSCORE, was 7.5%, surgical mortality occurred in only one patient (2.4%). Furthermore, the surgery resulted in significant improvements in LV systolic function as well as in quality of life. Consequently, authors concluded that surgical AF ablation should always be considered for this high-risk cohort, especially at centers with experienced surgeons. With regard to evaluating patients with LV dysfunction, the cited study may be the first to analyze the impact of the Maze procedure; however, it was not possible to determine whether the desirable clinical effects were a result of the Maze procedure or a consequence of correcting structural heart diseases because no control group was included in the study. To address this, we analysed clinical outcomes in a larger population according to a combination of the Maze procedure. Significant survival benefits were observed in those receiving the Maze procedure in the present study. Although several groups report superior survival rates in patients who maintained sinus rhythm compared with those who experienced AF [26], this considerable survival difference resulting from the Maze procedure is previously unreported. This finding was not even observed in our previous study involving 569 patients who underwent mechanical valve replacement in the presence of AF, in which a reduction in thromboembolic events, but not an improvement in overall survival, was observed in those receiving the Maze procedure [11]. Based on the findings of several studies, the differences in survival when restoring sinus rhythm may be attributable to discrepancies in subject populations and also the tools used for rhythm conversion [26,27]. First, surgical AF ablation, compared with catheter ablation or rhythm control strategy using various medications, is much more effective in restoring and maintaining sinus rhythm [7,27,28]. Second, patients with organic heart disease and severe LV dysfunction may be more affected by rhythm status than those without these factors since loss of atrial contraction can predispose patients to overt “low cardiac output” or increase the overall risk profiles in vulnerable patients [6,12]. In the present study, the risks of thromboembolism and CHF were lower in the Maze group than in the control group, and the two groups showed further marked discrepancies in eventfree survival. We believe that the incorporation of atrial kicking into the cardiac output by the Maze procedure improves hemodynamics with results in overall clinical benefits. These assumptions are strongly supported by our previous studies, in that around 80% of patients who restored sinus rhythm by the Maze procedure were found to restore effective atrial contraction as evidenced by (1) the presence of effective trans-mitral/trans-tricuspid atrial kicking wave (A wave) and (2) the
quality of “E/A ratio” (peak velocity of early diastolic filling [E wave] divided by peak velocity of A wave) [29,30]. Furthermore, performing the Maze procedure did not increase early mortality or morbidity, which indicates that the procedure can be carried out with an acceptable safety profile, even in this high-risk population. Long-term anticoagulation therapy is also an important issue for patients undergoing heart valve surgery. Even in the setting of bioprosthetic valve implantation or valve repair, about 30% of patients in this study required long-term anticoagulation therapy, even when receiving the Maze procedure. The most common indication of warfarin anticoagulation was the absence of atrial contraction either by AF or junctional rhythm (a rhythm originating from AV junction without a P wave), followed by the individual patient's comorbidity. Nevertheless, the need for anticoagulation was reduced by the Maze procedure. Although the rate of bleeding secondary to anticoagulation therapy could not be compared due to the small number of patients in this study, it is likely to show the benefits of the Maze procedure in larger populations. Given that maintaining lifelong anticoagulation therapy is directly related to quality of life, the Maze procedure will provide further benefits in avoiding warfarinization in patients undergoing bioprosthetic valve implantation and valve repair. 4.1. Limitations This study is subject to the limitations inherent to retrospective analyses of observational data. Although statistical models were used to adjust for treatment selection bias, our results may have been affected by unmeasured confounders, procedure bias or detection bias. Among patients with valvular AF, the prevalence of moderate-tosevere LV dysfunction was very low. Consequently, even though we extracted the subject population from a large population, the sample size was relatively small to draw robust conclusions. 5. Conclusions For patients with valvular AF and moderate-to-severe LV dysfunction undergoing major cardiac surgery, the combination of the Maze procedure was associated with an improvement in long-term survival and a reduction in serious adverse outcomes compared with cardiac surgery alone. Furthermore, the Maze procedure reduced the need for long-term anticoagulation therapy in this population. Therefore, surgical AF ablation should always be considered as a concomitant procedure during open cardiac surgery to improve long-term outcomes, even in a high-risk cohort. Conclusions regarding this issue need further verification through prospective randomized trials. Acknowledgement The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology. References [1] Cox JL, Boineau JP, Schuessler RB, et al. Successful surgical treatment of atrial fibrillation. Review and clinical update. JAMA 1991;266:1976–80. [2] Nakajima H, Kobayashi J, Bando K, et al. The effect of cryo-maze procedure on early and intermediate term outcome in mitral valve disease: case matched study. Circulation 2002;106:I46–50. [3] Bando K, Kobayashi J, Kosakai Y, et al. Impact of Cox maze procedure on outcome in patients with atrial fibrillation and mitral valve disease. J Thorac Cardiovasc Surg 2002;124:575–83. [4] 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. [5] Lall SC, Melby SJ, Voeller RK, et al. The effect of ablation technology on surgical outcomes after the Cox–Maze procedure: a propensity analysis. J Thorac Cardiovasc Surg 2007;133:389–96. [6] Gillinov AM, Sirak J, Blackstone EH, et al. The Cox maze procedure in mitral valve disease: predictors of recurrent atrial fibrillation. J Thorac Cardiovasc Surg 2005;130:1653–60.
J.B. Kim et al. / International Journal of Cardiology 170 (2014) 331–337 [7] Weimar T, Schena S, Bailey MS, et al. The Cox–Maze procedure for lone atrial fibrillation: a single-center experience over 2 decades. Circ Arrhythm Electrophysiol 2012;5:8–14. [8] Stulak JM, Dearani JA, Sundt III TM, Daly RC, Schaff HV. Ablation of atrial fibrillation: comparison of catheter-based techniques and the Cox–Maze III operation. Ann Thorac Surg 2011;91:1882–8. [9] Stulak JM, Schaff HV, Dearani JA, Orszulak TA, Daly RC, Sundt III TM. Restoration of sinus rhythm by the Maze procedure halts progression of tricuspid regurgitation after mitral surgery. Ann Thorac Surg 2008;86:40–4. [10] Cox JL, Ad N, Palazzo T. Impact of the maze procedure on the stroke rate in patients with atrial fibrillation. J Thorac Cardiovasc Surg 1999;118:833–40. [11] Kim JB, Moon JS, Yun SC, et al. Long-term outcomes of mechanical valve replacement in patients with atrial fibrillation: impact of the maze procedure. Circulation 2012;125:2071–80. [12] Ad N, Henry L, Hunt S. The impact of surgical ablation in patients with low ejection fraction, heart failure, and atrial fibrillation. Eur J Cardiothorac Surg 2012;40:70–6. [13] Calkins H, Kuck KH, Cappato R, et al. HRS/EHRA/ECAS expert consensus statement on catheter and surgical ablation of atrial fibrillation: recommendations for patient selection, procedural techniques, patient management and follow-up, definitions, end points, and research trial design: a report of the Heart Rhythm Society (HRS) Task Force on Catheter and Surgical Ablation of Atrial Fibrillation. Developed in partnership with the European Heart Rhythm Association (EHRA), a registered branch of the European Society of Cardiology (ESC) and the European Cardiac Arrhythmia Society (ECAS); and in collaboration with the American College of Cardiology (ACC), American Heart Association (AHA), the Asia Pacific Heart Rhythm Society (APHRS), and the Society of Thoracic Surgeons (STS). Endorsed by the governing bodies of the American College of Cardiology Foundation, the American Heart Association, the European Cardiac Arrhythmia Society, the European Heart Rhythm Association, the Society of Thoracic Surgeons, the Asia Pacific Heart Rhythm Society, and the Heart Rhythm Society. Heart Rhythm 2012;9:632–96 [e21]. [14] 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;92:1397–404. [15] Lauer MS, Blackstone EH, Young JB, Topol EJ. Cause of death in clinical research: time for a reassessment? J Am Coll Cardiol 1999;34:618–20. [16] Akins CW, Miller DC, Turina MI, et al. Guidelines for reporting mortality and morbidity after cardiac valve interventions. J Thorac Cardiovasc Surg 2008;135:732–8. [17] Robins JM, Hernan MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology 2000;11:550–60.
337
[18] Normand SL. Evaluating the optimal timing of angiography: landmark or off the mark? Circulation 2007;116:2656–7. [19] Nashef SA, Roques F, Sharples LD, et al. EuroSCORE II. Eur J Cardiothorac Surg 2012;41:734–44. [20] Lee EM, Shapiro LM, Wells FC. Mortality and morbidity after mitral valve repair: the importance of left ventricular dysfunction. J Heart Valve Dis 1995;4:460–8. [21] Enriquez-Sarano M, Tajik AJ, Schaff HV, Orszulak TA, Bailey KR, Frye RL. Echocardiographic prediction of survival after surgical correction of organic mitral regurgitation. Circulation 1994;90:830–7. [22] Bonow RO, Carabello BA, Kanu C, et al. ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (writing committee to revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease): developed in collaboration with the Society of Cardiovascular Anesthesiologists: endorsed by the Society for Cardiovascular Angiography and Interventions and the Society of Thoracic Surgeons. Circulation 2006;114:e84-231. [23] O'Brien SM, Shahian DM, Filardo G, et al. The Society of Thoracic Surgeons 2008 cardiac surgery risk models: part 2–isolated valve surgery. Ann Thorac Surg 2009;88: S23–42. [24] Middlekauff HR, Stevenson WG, Stevenson LW. Prognostic significance of atrial fibrillation in advanced heart failure. A study of 390 patients. Circulation 1991;84:40–8. [25] Pritchett EL. Management of atrial fibrillation. N Engl J Med 1992;326:1264–71. [26] Pedersen OD, Bagger H, Keller N, Marchant B, Kober L, Torp-Pedersen C. Efficacy of dofetilide in the treatment of atrial fibrillation-flutter in patients with reduced left ventricular function: a Danish investigations of arrhythmia and mortality on dofetilide (diamond) substudy. Circulation 2001;104:292–6. [27] Wyse DG, Waldo AL, DiMarco JP, et al. A comparison of rate control and rhythm control in patients with atrial fibrillation. N Engl J Med 2002;347:1825–33. [28] Boersma LV, Castella M, van Boven W, et al. Atrial fibrillation catheter ablation versus surgical ablation treatment (FAST): a 2-center randomized clinical trial. Circulation 2012;125:23–30. [29] Kim JB, Lee SH, Jung SH, et al. The influence of postoperative mitral valve function on the late recurrence of atrial fibrillation after the maze procedure combined with mitral valvuloplasty. J Thorac Cardiovasc Surg 2010;139:1170–6. [30] Kim JB, Yun TJ, Chung CH, Choo SJ, Song H, Lee JW. Long-term outcome of modified maze procedure combined with mitral valve surgery: analysis of outcomes according to type of mitral valve surgery. J Thorac Cardiovasc Surg 2010;139:111–7.
