Influence of Preoperative Atrial Fibrillation on Late Results of Mitral Repair: Is Concomitant Ablation Justified?

ORIGINAL ARTICLES: CARDIOVASCULAR

CARDIOVASCULAR

Influence of Preoperative Atrial Fibrillation on Late Results of Mitral Repair: Is Concomitant Ablation Justified? Dumbor L. Ngaage, MB, FRCS(C-Th), Hartzell V. Schaff, MD, Charles J. Mullany, MB, MS, Sunni Barnes, PhD, Joseph A. Dearani, MD, Richard C. Daly, MD, Thomas A. Orszulak, MD, and Thoralf M. Sundt III, MD Division of Cardiovascular Surgery and Department of Biostatistics, Mayo Clinic College of Medicine, Rochester, Minnesota

Background. There is considerable interest in atrial fibrillation (AF) ablation during cardiac operations, but there are few studies addressing the impact of preoperative AF on late outcome of surgery. We therefore investigated AF prevalence in nonrheumatic mitral regurgitation and its effect on late survival and morbidity after repair. Methods. From 1993 through 2002, 36% of 2,821patients with mitral regurgitation had preexisting AF. A cohort of these was matched with controls in sinus rhythm (SR) for age, gender, and ejection fraction. Follow-up was by questionnaire. Outcomes were compared between 231AF and 229 SR patients, and patients with different types of preoperative AF. Results. Patients with preoperative AF were more symptomatic and frequently had cardiomegaly, heart failure, and higher mean pulmonary artery systolic pressure. Operative mortality was higher for AF patients (2% vs 0, p ⴝ 0.05).

More AF patients had late adverse cardiac events and stroke (63% vs 31%, p < 0.0001). Five- and ten-year survival was, respectively, 95% and 88% for SR patients compared with 90% and 70% (p ⴝ 0.01) for the AF group. By multivariate analysis, preoperative AF was not a predictor of long-term survival but was an independent risk factor for late adverse cardiac events and stroke. Conclusions. Preoperative AF is a marker for increased surgical risk of mitral regurgitation repair, and a risk factor for late adverse cardiac events and stroke. Although the independent contribution of AF to late survival is uncertain, preoperative AF increases postoperative morbidity independently; therefore, corrective intervention would be expected to benefit patients in this regard.

A

tive AF. In order to identify the multiple impacts of preoperative AF on clinical outcomes after cardiac surgery, we performed a matched cohort comparison between patients with preoperative AF and those in preoperative sinus rhythm (SR) undergoing various cardiac operations. We have previously reported the adverse effects of preexisting AF on clinical results after aortic valve replacement [8], and coronary artery bypass grafting [9]. The present study compares clinical outcomes after repair of mitral valve regurgitation between patients with preoperative AF and matched controls in SR. Our objectives were to determine the prevalence of AF in patients undergoing repair of nonrheumatic mitral regurgitation, and to assess the influence of preoperative AF on early morbidity and mortality, late adverse cardiac and cerebrovascular events, and long-term survival.

trial fibrillation (AF) is common in patients with mitral valve disease [1] and has been identified as a risk factor for poor outcome after mitral valve replacement [2]. Because of the availability of radiofrequency devices and cryoprobes, many surgeons routinely ablate AF at the time of mitral valve surgery [1, 3]. There are, however, relatively few studies addressing the impact of preoperative AF on late outcome after mitral valve surgery, and the potential benefit of concomitant AF ablation is, therefore, difficult to judge. Further, the prevalence of AF in nonrheumatic mitral disease, which is most commonly encountered in current surgical practice, is not known, nor is its influence on early and late outcomes of mitral valve repair. Atrial fibrillation has been identified as a marker of severe cardiac disease and a risk factor for decreased long-term survival in nonsurgical series [4 – 6]. Quader and associates [7] have also reported excess late mortality after coronary artery bypass grafting in patients with preoperaAccepted for publication April 11, 2007. Presented at the Fifty-third Annual Meeting of the Southern Thoracic Surgical Association, Tucson, AZ, Nov 8 –11, 2006. Address correspondence to Dr Ngaage, Department of Cardiothoracic Surgery, Cardiothoracic Centre, Castle Hill Hospital, Cottingham, East Yorkshire, HU15 6JQ, United Kingdom; e-mail: [email protected].

© 2007 by The Society of Thoracic Surgeons Published by Elsevier Inc

(Ann Thorac Surg 2007;84:434 – 43) © 2007 by The Society of Thoracic Surgeons

Patients and Methods Study Population The Institutional Review Board approved the use of prospectively collected clinical data and the administration of a survey questionnaire to patients for this study, in May 2004. The cardiovascular surgery database was interrogated and data of all patients who underwent 0003-4975/07/$32.00 doi:10.1016/j.athoracsur.2007.04.036

Ann Thorac Surg 2007;84:434 – 43

NGAAGE ET AL PREOP AF AND REPAIR OF MITRAL REGURGITATION

435

Table 1. Baseline Characteristics and In-Hospital Outcome of Patients Undergoing Repair of Mitral Valve Regurgitation

Variables

Fig 1. Flow chart of the study patient selection process. The exclusion criteria were history of cardiac arrhythmia other than atrial fibrillation (AF), previous permanent pacemaker and (or) automatic implantable cardioverter defibrillator implantation, permanent neurological deficit from stroke, and residence outside the United States. (SR ⫽ sinus rhythm).

isolated mitral valve surgery with or without tricuspid valve repair from January 1993 through December 2002 were retrieved. Patients with rheumatic valvular heart disease, infective endocarditis, previous cardiac surgery, and those who declined research authorization were not included. Of 2,821 patients who had MR, 1,020 (36.2%) had AF before surgery. Other exclusion criteria were the following: history of cardiac arrhythmia other than AF, previous permanent pacemaker and (or) automatic implantable cardioverter defibrillator implantation, permanent neurological deficit from stroke, mitral valve replacement, and residence outside the United States. From published reports, we calculated a sample size of 200 per group to detect a 10% difference at 90% power. We selected 250 patients with preoperative AF using the random option of the survey select procedure of the SAS statistical analysis system (SAS Institute Inc, Cary, NC), and identified the best matched control cohort in preoperative SR based on age, gender, and left ventricular ejection fraction using the Greedy method [10]. The selection process is represented in Figure 1. All deaths were identified using the Social Security Death Index. We investigated causes of all deaths, and tracked late complications and reinterventions by direct patient follow-up with a survey questionnaire, telephone calls, and interrogation of the Mayo Integrated Clinical Information System for return visits. The response rate to follow-up was 92%.

Definitions The international consensus on nomenclature and classification [11] recently classified AF into initial event, paroxysmal, persistent, and permanent types. Data col-

Demographics Mean age (years)a Female gendera Mean BMI (kg/m2) Symptoms NYHA class III/IV Cardiac Morbidity: Ejection fractiona ⬍0.35 0.35 to 0.50 ⬎0.50 Cardiomegaly Myocardial infarction Mean PAP systolic Tricuspid regurgitation Trivial/mild Moderate/severe History of heart failure Comorbidities: Diabetes mellitus Renal insufficiency COPD Past history Stroke Operative details: Concomitant TV repair Maze procedure Surgery priority Elective Urgent Mean X-C time(mins) Postoperative data: Inotropic support IABP use Reoperation for bleeding Stroke Renal failure Operative death Hospital length of stay (days) 30-day readmission a

AF Group (n ⫽ 231) No. (%)

SR Group (n ⫽ 229) No. (%)

p Value

65 ⫾ 12 83 (36) 26 ⫾ 6

65 ⫾ 12 82 (36) 26 ⫾ 4

0.73 0.99 0.32

152 (66)

104 (45)

⬍0.0001

3 (1) 22 (10) 206 (89) 149 (65) 11 (5) 48 ⫾ 14

4 (2) 23 (10) 202 (88) 81 (36) 3 (1) 43 ⫾ 16

156 (68) 75 (32) 83 (36)

195 (85) 34 (15) 31 (14)

⬍0.0001

12 (5) 2 (1) 16 (7)

6 (3) 2 (1) 14 (6)

0.15 0.99 0.72

7 (3)

1 (0)

0.03

18 (8) 90 (39)

8 (3) 0

223 (97) 8 (3) 48 ⫾ 22

223 (97) 6 (3) 39 ⫾ 18

⬍0.0001

94 (41) 3 (1) 17 (7) 6 (3) 5 (2) 4 (2) 9⫾7

54 (24) 4 (2) 4 (2) 3 (1) 2 (1) 0 7⫾4

0.0001 0.69 0.004 0.32 0.26 0.05 ⬍0.0001

21 (9)

13 (6)

0.16

0.91

⬍0.0001 0.03 0.003

⬍0.0001

0.05 ⬍.0001 0.60

Matching variables.

AF ⫽ atrial fibrillation; BMI ⫽ body mass index; CHF ⫽ congestive heart failure; COPD ⫽ chronic obstructive pulmonary disease; EF ⫽ ejection fraction; IABP ⫽ Intraaortic balloon pump; ITA ⫽ internal thoracic artery; MI ⫽ myocardial infarction; NYHA ⫽ New York Heart Association functional class; PAP ⫽ pulmonary artery pressure; Periop ⫽ perioperative; SR ⫽ sinus rhythm; TR ⫽ tricuspid regurgitation; X-C ⫽ cross clamp.

CARDIOVASCULAR

lection during the study period utilized the simpler schema proposed by Cox [12] and widely used in the surgical literature [13]. However, the similarity in the

436

NGAAGE ET AL PREOP AF AND REPAIR OF MITRAL REGURGITATION

Ann Thorac Surg 2007;84:434 – 43

CARDIOVASCULAR

arrhythmia patterns that define the different types of AF in both Cox’s and the international classification, enabled reclassification. In this study, we used the following classifications: (1) paroxysmal or persistent atrial fibrillation is preoperative AF that is recurrent; (2) permanent atrial fibrillation is preoperative AF that is present at all times.

Statistical Analysis The primary endpoints were (1) late mortality and (2) adverse cardiac events and stroke. Adverse cardiac events were defined as cardiac death, myocardial infarction, heart failure, and heart rhythm-related intervention. Secondary endpoints included postoperative length of hospital stay, cardiac-related hospital readmission, and noncerebral bleeding-thromboembolic complications. Categoric variables are expressed as percentages and continuous variables as mean ⫾ SD. Univariate analysis was performed using the ␹2 test for categoric variables and the Wilcoxon rank test for continuous variables. The Cox proportional hazards models and logistic regression were utilized for multivariable analysis, and long-term outcomes were analyzed using the Kaplan-Meier method. Risk factors associated with each endpoint of interest were identified by the stepwise model selection procedure. If the primary variable of interest (preoperative AF versus preoperative SR) was selected, then this model is reported. Otherwise, if the preoperative AF versus preoperative SR grouping variable was not chosen, it was forced into the final model provided by the stepwise procedure. The variables included in the stepwise logistic regression and multivariable models are listed in Appendix 1. The SAS statistical analysis system (SAS Institute Inc) was used for data analysis. Statistical significance was defined as p less than 0.05 with a 2-tailed test.

Fig 2. Kaplan-Meier survival curves of patients with preoperative atrial fibrillation (AFib) and those with preoperative sinus rhythm (SR) after repair of mitral regurgitation.

Fig 3. Survival regression model showing predictors of all-cause late mortality by multivariate analysis. (AF ⫽ atrial fibrillation; CHF ⫽ congestive heart failure.)

Results Baseline Characteristics and Early Outcome Patient demographics were similar and comorbidities were evenly distributed between the two groups. Patients with preoperative AF were more symptomatic, however, and had more cardiac morbidity in the form of cardiomegaly, moderate to severe tricuspid valve regurgitation, higher pulmonary artery pressure, and prior myocardial infarction (see Table 1). Patients with preoperative AF had longer postoperative hospital stays (9 ⫾ 7 vs 7 ⫾ 4 days, p ⬍ 0.0001) and higher in-hospital mortality (2% vs 0, p ⫽ 0.05). The determinants of prolonged hospital stay by multivariate analysis were history of heart failure (p ⫽ 0.003), preoperative AF (p ⫽ 0.0009), and postoperative renal failure (p ⬍ 0.0001).

Principal Findings: Late Morbidity and Mortality In univariate analysis, there was a 120% increase in the risk of late all-cause mortality (risk ratio ⫽ 2.2, p ⫽ 0.01) among patients with preoperative AF compared with SR patients. Patients with preoperative AF were 3.5 times more likely to die from cardiac causes (risk ratio ⫽ 3.5, p ⫽ 0.04). The five-year and ten-year overall survival for preoperative AF patients was thereby substantially reduced at 90% and 70% compared with 95% and 88% (p ⫽ 0.01) for preoperative SR patients (Fig 2). Correspondingly, the freedom from cardiac death at five and ten years was, respectively, 98% and 84% for the preoperative AF group compared with 99% and 96% for the preoperative SR group (p ⫽ 0.04). The independent contribution of preoperative AF to decreased survival was evaluated by multivariate analysis. The independent risk factors for late all-cause mortality (Fig 3) were cardiomegaly (hazard ratio ⫽ 2.82, 95% confidence interval 1.32 to 6.04, p ⫽ 0.0008) and preoperative heart failure (hazard ratio ⫽ 2.56, 95% confidence interval 1.35 to 4.83, p ⫽ 0.004); in this model, preoperative AF was not an independent predictor of late mortality. The rates of nonfatal cardiac events, stroke, and cardiacrelated hospital readmissions were also significantly

Ann Thorac Surg 2007;84:434 – 43

NGAAGE ET AL PREOP AF AND REPAIR OF MITRAL REGURGITATION

437

Table 2. Late Outcome After Repair of Mitral Valve Regurgitation SR Group

Variables

No.

Frequency (%)

No.

Frequency (%)

p Value

Mean follow-up duration (yrs) NYHA functional class III/IV Medications Coumadin Beta blocker ACE inhibitor Late hospital readmissions Cardiac rhythm Sinus rhythm Atrial fibrillation Paced rhythm Adverse events Cardiac Cardiac death Myocardial infarction Congestive heart failure Reintervention Stroke Bleeding disorder EF change

227 144

4.9 ⫾ 3.1 6 (4)

229 156

5.0 ⫾ 3.0 11 (7)

ns 0.28

155 148 148 156 172

74 (48) 73 (49) 59 (40) 71 (46)

159 161 161 161 167

21 (13) 52 (32) 55 (34) 28 (17)

⬍0.0001 0.002 0.30 0.0001 ⬍0.0001

ACE ⫽ angiotensin-converting enzyme; rhythm.

189

44 (26) 122 (71) 6 (3) 120 (63)

193 142 151 129 152 140 166

9 (5) 1 (1) 26 (17) 29 (24) 18 (12) 23 (16) ⫺3 ⫾ 11

AF ⫽ atrial fibrillation;

EF ⫽ ejection fraction;

higher for preoperative AF patients (Table 2). Late postoperatively, heart failure (17% vs 3%, p ⱕ 0.0001), subsequent heart rhythm-related intervention (Fig 4) in the form of cardioversion and (or) permanent pacemaker insertion (24% vs 9%, p ⫽ 0.002), and stroke (12% vs 5%, p ⫽ 0.03) were more frequently observed in the AF group. Notably, preoperative AF (odds ratio ⫽ 3.12, 95% confidence interval 1.99 to 4.97, p ⬍ 0.0001) was an independent risk factor for adverse cardiac events and stroke, as

173

130 (78) 33 (20) 4 (2) 54 (31)

⬍0.0001

229 156 159 155 159 155 174

3 (1) 3 (2) 4 (3) 13 (9) 8 (5) 11 (7) ⫺4 ⫾ 12

0.05 0.36 ⬍0.0001 0.002 0.03 0.01 0.59

NYHA ⫽ New York Heart Association;

SR ⫽ sinus

was cardiomegaly (odds ratio ⫽ 2.09, 95% confidence interval 1.32 to 3.30, p ⫽ 0.002). Cardiac-related hospital readmissions were more common for the AF group. Documented criteria for rehospitalization include rate-rhythm control of AF, regulation of anticoagulation, or treatment of heart failure. The use of anticoagulation (Coumadin; Bristol-Myers Squibb Co, Princeton, NJ) and beta blockers was more common in the AF group, as was the incidence of noncerebral bleeding and (or) thromboembolism. Improvement in symptoms and change in left ventricular ejection fraction were similar between both groups of patients at late follow-up (mean 5 ⫾ 3 years). Both groups reported a significant improvement in the New York Heart Association (NYHA) functional class. At the Table 3. The Clinical Type of Preoperative Atrial Fibrillation and Clinical Outcome Preoperative Atrial Fibrillation Type Variables Mean length of hospital stay (days) Late all-cause mortality

Fig 4. Kaplan-Meier curves showing intervention for heart rhythmrelated complications in patients with preoperative atrial fibrillation (Preop AFib) and those in sinus rhythm (Preop SR) after repair of mitral regurgitation.

Adverse events (%)

Paroxysmal or Persistent

Permanent

p Value

n ⫽ 134 9⫾7 n ⫽ 130 11 (8%) n ⫽ 107 66 (58)

n ⫽ 48 11 ⫾ 10 n ⫽ 48 10 (21%) n ⫽ 39 25 (66)

0.12 0.02 0.79

CARDIOVASCULAR

AF Group

438

NGAAGE ET AL PREOP AF AND REPAIR OF MITRAL REGURGITATION

Ann Thorac Surg 2007;84:434 – 43

Table 4. Baseline Characteristics and In-Hospital Outcome of Patients With Preoperative Atrial Fibrillation (Excluding Cox-Maze Patients) Compared to Sinus Rhythm Patients CARDIOVASCULAR

Variables

Fig 5. Long-term survival after repair of mitral regurgitation in patients with preoperative permanent and paroxysmal atrial fibrillation (AF).

time of surgery, 66% of preoperative AF and 45% of preoperative SR patients were in NYHA functional class III/IV compared with 4% and 7%, respectively, at followup. Similarly, the mean ejection fraction of both groups decreased over time (see Table 2).

Principal Findings: Impact of Clinical Type and Duration of Preoperative Atrial Fibrillation The clinical type of preoperative AF was known in 182 patients (79%). The clinical outcomes for the different clinical types of preoperative AF are shown in Table 3. Patients with permanent AF had a 170% increased risk of late mortality (risk ratio ⫽ 2.7, p ⫽ 0.02) compared with those with paroxysmal or persistent AF. The three-year and six-year survival were, respectively, 93% and 79% for patients with permanent AF compared with 97% and 91% for those with paroxysmal or persistent AF (Fig 5).

Secondary Analysis In order to eliminate the potential confounding effect of surgical ablation of AF, we excluded 90 AF patients who underwent the Cox-maze operation from this analysis (Table 4). Again the findings, shown on Table 5, were consistent with the previous comparison. The overall survival was remarkably reduced for AF patients (fiveyear and ten-year survival 88% and 68% for AF vs 95% and 88% for SR patients, p ⫽ 0.006), but in the multivariate analysis preoperative AF was not a statistically significant determinant of late mortality. We were unable to determine any favorable impact of the maze operation on outcomes in this study even though there was a trend toward reduced incidence of late stroke and cardiac death, as the baseline variables were vastly different from the other groups given the selection bias (see Appendices 2 and 3).

Comment In the current era, patients undergoing surgery for nonrheumatic mitral regurgitation have a high prevalence of

Demographics Mean age (years)a Female gendera Mean BMI (kg/m2) Symptoms Angina NYHA class III/IV Cardiac morbidity Ejection fractiona ⬍0.35 0.35 to 0.50 ⬎0.50 Cardiomegaly Myocardial infarction Mean PAP systolic Tricuspid regurgitation Moderate Severe History of heart failure Comorbidities Diabetes mellitus Renal insufficiency COPD Past history Stroke Operative details Concomitant TV repair Surgery priority Elective Urgent Mean X-C time(mins) Postoperative data Inotropic support IABP use Reoperation for bleeding Stroke Renal failure Operative death Hospital length of stay (days) 30-day readmission a

AF Group (n ⫽ 141) No. (%)

SR Group (n ⫽ 229) No. (%)

p Value

67 ⫾ 11 46 (33) 25 ⫾ 5

65 ⫾ 12 82 (36) 26 ⫾ 4

0.11 0.53 0.09

7 (5) 95 (68)

12 (5) 104 (45)

0.91 ⬍0.0001

2 (1) 12 (9) 127 (90) 87 (62) 9 (6) 50 ⫾ 14

4 (2) 23 (10) 202 (88) 81 (36) 3 (1) 43 ⫾ 16

32 (23) 13 (9) 56 (40)

29 (13) 5 (2) 31 (14)

⬍0.0001

6 (4) 2 (1) 10 (7)

6 (3) 2 (1) 14 (6)

0.39 0.62 0.71

4 (3)

1 (0)

0.05

14 (10)

8 (3)

0.01 0.61

136 (96) 5 (4) 40 ⫾ 18

223 (97) 6 (3) 39 ⫾ 18

0.52

52 (37) 2 (1) 11 (11) 4 (3) 5 (4) 3 (2) 8.6 ⫾ 6.6

54 (24) 4 (2) 4 (2) 3 (1) 2 (1) 0 6.8 ⫾ 3.6

0.006 0.81 0.004 0.30 0.07 0.03 ⬍0.0001

12 (9)

13 (6)

0.29

0.86

⬍0.0001 0.008 0.0004

⬍0.0001

Matching variables.

AF ⫽ atrial fibrillation; BMI ⫽ body mass index; CHF ⫽ congestive heart failure; COPD ⫽ chronic obstructive pulmonary disease; EF ⫽ ejection fraction; IABP ⫽ Intraaortic balloon pump; ITA ⫽ internal thoracic artery; MI ⫽ myocardial infarction; NYHA ⫽ New York Heart Association functional class; PAP ⫽ pulmonary artery pressure; Periop ⫽ perioperative; SR ⫽ sinus rhythm; TR ⫽ tricuspid regurgitation; X-C ⫽ cross clamp.

Ann Thorac Surg 2007;84:434 – 43

439

NGAAGE ET AL PREOP AF AND REPAIR OF MITRAL REGURGITATION

Table 5. Comparison of Late Outcome After Repair of Mitral Valve Regurgitation in 141 Patients With Preoperative Atrial Fibrillation and 229 With Sinus Rhythm Sinus Rhythm Group

Variables

No.

Frequency (%)

No.

Frequency (%)

p Value

Late all cause mortality NYHA class III/IV Medications Coumadin Beta blocker ACE inhibitor Late hospital readmissions Cardiac rhythm Sinus rhythm Atrial fibrillation Paced rhythm Adverse events Cardiac Cardiac death Myocardial infarction Heart failure Reintervention Stroke Bleeding disorder EF change

138 86

24 (17) 2 (2)

229 156

14 (6) 11 (7)

0.006 0.12

95 89 88 91 104

46 (48) 38 (43) 32 (36) 32 (35)

159 161 161 161 167

21 (13) 52 (32) 55 (34) 28 (17)

⬍.0001 0.10 0.73 0.002 ⬍.0001

EF ⫽ ejection fraction;

24 (23) 79 (76) 1 (1)

138 86 89 83 93 84 91

9 (7) 0 12 (13) 7 (8) 14 (15) 15 (18) ⫺5.0 ⫾ 12

130 (78) 33 (20) 4 (2)

229 156 159 155 159 155 174

3 (1) 3 (2) 4 (3) 13 (9) 8 (5) 11 (7) ⫺4 ⫾ 12

0.02 0.20 0.0007 0.96 0.007 0.01 0.75

NYHA ⫽ New York Heart Association.

AF (36%). These patients frequently present with severe symptoms and increased cardiac morbidity in the form of prior myocardial infarction, cardiomegaly, tricuspid valve regurgitation, and high pulmonary artery pressure. Previous studies [7, 14 –17], have identified preexisting AF as a marker of advanced cardiovascular disease, which can influence the early and late outcome of surgery. The regular association of AF with morphologic and functional changes indicative of increased cardiac morbidity, as observed in our study, lends credence to this argument. The higher requirement for hemodynamic support with inotropes, and more in-hospital deaths observed in the AF group compared with the SR group, highlights the increased surgical risk (2% vs 0).

Long-Term Survival and Late Adverse Events There was a substantial reduction in long-term survival for patients with preoperative AF. Studies investigating the impact of preoperative AF on late survival after mitral valve surgery have given conflicting results. Studies comparing small unmatched groups like the series of Jessurun and associates [18] (68 AF and 57 SR patients) and Obadia and colleagues [15], (96 AF vs 95 SR patients) did not find a significant survival difference between the two groups. Similarly, in a previous study of a different cohort, we [14] did not detect a significant difference in survival between 97 AF and 216 SR patients. In all these series, however, there was a trend toward decreased survival in patients with preoperative AF. These studies are hindered by sample size and limited follow-up interval,

hence the power to detect significant survival disparity is decreased. Studies with larger sample sizes have identified significant differences in long-term survival. Bando and colleagues [17] compared 363 patients with preoperative AF (163 had maze operation) with an unmatched cohort of 663 patients with SR and reported a survival advantage for the latter group. A similar result was reported by Lim [16] and Eguchi [19] and their associates in their large series. The present investigation differs from previous studies in the case-matched design, which minimized differences in preoperative characteristics. We observed a reduced late survival among patients with preoperative AF compared with patients with SR after repair of MR. In addition, and importantly, we found a difference in late survival of patients with preoperative AF on the basis of the clinical type of AF. There were more late deaths among patients with permanent AF compared with those with paroxysmal AF. This finding is in keeping with nonsurgical reports that describe permanent AF as an advanced form of the disease [20] associated with worse survival [21]. Despite matching patients for known risk factors of age, gender, and ejection fraction, we found that other variables influencing survival were not evenly distributed between the groups. In a multivariable model, these factors rather than presence of preoperative AF were predictive of late death. In clinical practice, AF is frequently observed in patients with mitral valve disease

CARDIOVASCULAR

Atrial Fibrillation Group

440

NGAAGE ET AL PREOP AF AND REPAIR OF MITRAL REGURGITATION

CARDIOVASCULAR

who have congestive heart failure and cardiomegaly. These markers of advanced valve disease appear to be more important determinants of postoperative mortality than the accompanying arrhythmia itself. Among patients with preoperative AF, cardiac-related hospital readmission was more common than among those in preoperative SR despite otherwise appropriate medical therapy. The rates of noncerebral bleeds and (or) thromboembolism, heart failure, subsequent heart rhythm-related interventions, and stroke were also higher. These data are consistent with our findings for patients undergoing aortic valve replacement [8] and other reports [18, 22]. Although preoperative AF was not a predictor of late death, it had an important impact on other nonfatal cardiac events and stroke. The risk of developing a stroke and (or) heart failure late postoperatively was increased twofold by preoperative AF. There is a strong association between AF and congestive heart failure, stroke, repeat hospital admissions, and cardiac death [17, 23]. The clinical impact of AF ablation at the time of mitral valve repair should therefore be put into perspective. Our data suggest that the potential benefit of concomitant AF ablation is reduction of late postoperative morbidity. The survival benefit, as has been suggested by others [24, 25] is not certain because of the other risk factors commonly present with AF. A recent metaanalysis by Wong and Mak [26] shows that the restoration of sinus rhythm by the maze procedure reduces the rate of adverse events, but did not improve survival. Avoidance of AF after operation might improve survival by decreasing fatal complications of stroke, but much larger numbers of patients would be necessary to study this. Understanding the influence of preoperative AF on late outcome is further complicated by the fact that late AF occurs in approximately one-quarter of patients who have sinus rhythm preoperatively.

Study Limitations The exclusive study of nonrheumatic MR and matching of patients on fundamental variables that affect prognosis makes this study relevant to contemporary practice. However, findings in this study should be interpreted in the context of its retrospective design and the difficulty in separating paroxysmal and persistent AF. Also, the inclusion of patients who had the Cox-maze procedure (a strategy used previously [17]) is a potential weakness, although analysis of the subgroup, excluding the Coxmaze patients, yielded essentially identical results as were found with the overall cohort of patients with preoperative AF; probably because of the small number of patients who had the maze procedure. This study was not designed and adequately powered to compare the outcome of the maze operation performed in a selective subset of patients with patients in sinus rhythm. Late postoperative heart rhythm was not confirmed by electrocardiography in all patients in this study. In 22% of patients who had electrocardiograms within 6 months of the survey, there was a 100% concordance between the self-reported heart rhythms and documented rhythms.

Ann Thorac Surg 2007;84:434 – 43

Electrocardiograms provide a “snapshot” assessment of heart rhythm, and may not detect paroxysmal AF.

Conclusion Atrial fibrillation is common in patients undergoing surgery for nonrheumatic mitral regurgitation and frequently coexists with morphologic and functional changes indicative of advanced mitral disease. It has multiple adverse effects on prognosis after repair of mitral regurgitation and, therefore, is a marker for poor clinical outcome. Patients with preoperative AF have high rates of adverse cardiac events and stroke, and a substantial reduction in survival. Preoperative AF is an important predictor of adverse cardiac events and stroke but is not an independent predictor of late mortality. We believe that ablation of AF at the time of mitral valve surgery for nonrheumatic regurgitation is warranted in order to reduce late morbidity, but the benefit of ablation of AF on late survival is uncertain.

References 1. Ad N, Cox JL. The significance of atrial fibrillation ablation in patients undergoing mitral valve surgery. Semin Thorac Cardiovasc Surg 2002;14:193–7. 2. Ruel M, Kulik A, Lam BK, et al. Long-term outcomes of valve replacement with modern prostheses in young adults. Eur J Cardiothorac Surg 2005;27:425–33. 3. Cox JL. The role of surgical intervention in the management of atrial fibrillation. Tex Heart Inst J 2004;31:257– 65. 4. Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation 1998;98:946 –52. 5. Stewart S, Hart CL, Hole DJ, McMurray JJ. A populationbased study of the long-term risks associated with atrial fibrillation: 20-year follow-up of the Renfrew/Paisley study. Am J Med 2002;113:359 – 64. 6. Vidaillet H, Granada JF, Chyou PH, et al. A populationbased study of mortality among patients with atrial fibrillation or flutter. Am J Med 2002;113:365–70. 7. Quader MA, McCarthy PM, Gillinov AM, et al. Does preoperative atrial fibrillation reduce survival after coronary artery bypass grafting? Ann Thorac Surg 2004;77:1514 –24. 8. Ngaage DL, Schaff HV, Barnes SA, et al. Prognostic implications of preoperative atrial fibrillation in patients undergoing aortic valve replacement: is there an argument for concomitant arrhythmia surgery? Ann Thorac Surg 2006;82: 1392–9. 9. Ngaage DL, Schaff HV, Mullany CJ, et al. Does preoperative atrial fibrillation influence early and late outcomes of coronary artery bypass grafting? J Thorac Cardiovasc Surg 2007; 133:182–9. 10. Aronson J, Dyer M, Frieze A, Suen S. Randomized greedy matching. II. Random Structures & Algorithms 1995;6:55–73. 11. Levy S, Camm AJ, Saksena S, et al. International consensus on nomenclature and classification of atrial fibrillation: a collaborative project of the Working Group on Arrhythmias and the Working Group of Cardiac Pacing of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. J Cardiovasc Electrophysiol 2003;14:443–5. 12. Cox JL. Atrial fibrillation I: a new classification system. J Thorac Cardiovasc Surg 2003;126:1686 –92. 13. Haissaguerre M, Jais P, Shah DC, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 1998;339:659 – 66.

14. Chua YL, Schaff HV, Orszulak TA, Morris JJ. Outcome of mitral valve repair in patients with preoperative atrial fibrillation. Should the maze procedure be combined with mitral valvuloplasty? J Thorac Cardiovasc Surg 1994;107: 408 –15. 15. Obadia JF, el Farra M, Bastien OH, Lievre M, Martelloni Y, Chassignolle JF. Outcome of atrial fibrillation after mitral valve repair. J Thorac Cardiovasc Surg 1997;114:179 – 85. 16. Lim E, Barlow CW, Hosseinpour AR, et al. Influence of atrial fibrillation on outcome following mitral valve repair. Circulation 2001;104:I59 – 63. 17. Bando K, Kasegawa H, Okada Y, et al. Impact of preoperative and postoperative atrial fibrillation on outcome after mitral valvuloplasty for nonischemic mitral regurgitation. J Thorac Cardiovasc Surg 2005;129:1032– 40. 18. Jessurun ER, van Hemel NM, Kelder JC, et al. Mitral valve surgery and atrial fibrillation: is atrial fibrillation surgery also needed? Eur J Cardiothorac Surg 2000;17:530 –7. 19. Eguchi K, Ohtaki E, Matsumura T, et al. Pre-operative atrial fibrillation as the key determinant of outcome of mitral valve repair for degenerative mitral regurgitation. Eur Heart J 2005;18:1866 –72. 20. Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA. Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation 1995;92:1954 – 68. 21. Gajewski J, Singer RB. Mortality in an insured population with atrial fibrillation. JAMA 1981;245:1540 – 4. 22. Kernis SJ, Nkomo VT, Messika-Zeitoun D, et al. Atrial fibrillation after surgical correction of mitral regurgitation in sinus rhythm: incidence, outcome, and determinants. Circulation 2004;110:2320 –5. 23. Enriquez-Sarano M, Schaff HV, Orszulak TA, Bailey KR, Tajik AJ, Frye RL. Congestive heart failure after surgical correction of mitral regurgitation. A long-term study. Circulation 1995;92:2496 –503. 24. Chaput M, Bouchard D, Demers P, et al. Conversion to sinus rhythm does not improve long-term survival after valve surgery: insights from a 20-year follow-up study. Eur J Cardiothorac Surg 2005;28:206 –10. 25. Raanani E, Albage A, David TE, Yau TM, Armstrong S. The efficacy of the Cox/maze procedure combined with mitral valve surgery: a matched control study. Eur J Cardiothorac Surg 2001;19:438 – 42. 26. Wong JW, Mak KH. Impact of maze and concomitant mitral valve surgery on clinical outcomes. Ann Thorac Surg 2006; 82:1938 – 47.

Appendix 1 Variables Examined by Stepwise Logistic Regression 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Age Gender Body mass index Diabetes mellitus Preoperative renal insufficiency Previous stroke Prior myocardial infarction Preoperative cardiac rhythm: sinus rhythm versus atrial fibrillation Preoperative ejection fraction Pulmonary artery systolic pressure Cardiomegaly Congestive heart failure New York Heart Association functional class III/IV Canadian Cardiovascular Society Scoring class III/IV for angina Duration of aortic cross clamp Moderate and severe tricuspid regurgitation Surgery priority: elective versus urgent versus emergent

NGAAGE ET AL PREOP AF AND REPAIR OF MITRAL REGURGITATION

18. 19.

441

Concomitant tricuspid surgery Postoperative renal failure/dialysis

Variables Selected and Retained in the Multivariable Model 1. 2. 3. 4. 5.

Cardiomegaly Congestive heart failure Pulmonary artery systolic pressure Surgical priority Preoperative rhythm

Appendix 2 Baseline Characteristics and In-Hospital Outcome of Patients With Atrial Fibrillation Undergoing Mitral Valve Repair With and Without Cox-Maze Operation

Variable Demographics Mean age (years) Male gender Mean body mass index (kg/m2) NYHA class III/IV Cardiac morbidity Left ventricular ejection fraction ⬍0.35 0.35 to 0.50 ⬎0.50 Cardiomegaly Myocardial infarction Mean PA systolic pressure Moderate/severe TR Congestive heart failure Comorbidities Diabetes mellitus Renal insufficiency COPD Past history stroke Operative details: Surgery priority: Elective Urgent Tricuspid valve repair Mean cross-clamp time (minutes) Operative death Hospital length of stay (days) 30-day readmission

With Maze Without Maze (n ⫽ 90) (n ⫽ 141) No. (%) No. (%) p Value 63 ⫾ 12 53 (59) 27 ⫾ 7

67 ⫾ 11 95 (67) 25 ⫾ 5

0.008 0.19 0.09

57 (64)

95 (67)

0.60 0.79

1 (1) 10 (11) 79 (88) 62 (69) 2 (2) 45 ⫾ 14

2 (1) 12 (9) 127 (90) 87 (62) 9 (6) 50 ⫾ 14

0.30 0.15 0.03

30 (33) 27 (30)

45 (32) 56 (40)

0.82 0.13

6 (7) 0 6 (7)

6 (4) 2 (1) 10 (7)

0.42 0.26 0.90

3 (3)

4 (3)

0.83

87 (97) 3 (3) 4 (4) 59 ⫾ 4

136 (96) 5 (4) 14 (10) 40 ⫾ 18

0.93 0.13 ⬍0.0001

1 (1) 10 ⫾ 7

3 (2) 9⫾7

0.56 0.004

9 (10)

12 (9)

0.70

COPD ⫽ chronic obstructive pulmonary disease; NYHA ⫽ New York Heart Association; PA ⫽ pulmonary artery; TR ⫽ tricuspid regurgitation.

CARDIOVASCULAR

Ann Thorac Surg 2007;84:434 – 43

442

NGAAGE ET AL PREOP AF AND REPAIR OF MITRAL REGURGITATION

Ann Thorac Surg 2007;84:434 – 43

Appendix 3 Late Outcome of Mitral Regurgitation Repair With or Without Cox-Maze Operation in Patients With Preexisting Atrial Fibrillation CARDIOVASCULAR

With Maze

Late all cause mortality NYHA functional class III/IV Medications: Coumadin Beta blocker ACE inhibitor Adverse events: Cardiac death Myocardial infarction Congestive heart failure Stroke Bleeding disorders ACE ⫽ angiotensin converting enzyme;

Without Maze

No. of Patients

Frequency (%)

No. of patients

Frequency (%)

p Value

55 58

4 (7) 4 (7)

138 86

24 (17) 2 (2)

0.31 0.18

60 59 60

28 (47) 35 (59) 27 (45)

95 89 88

46 (48) 38 (43) 32 (36)

0.83 0.05 0.29

89 56 62 59 56

2 (2) 1 (2) 14 (22) 4 (7) 8 (14)

138 86 89 93 84

9 (7) 0 12 (13) 14 (15) 15 (18)

0.40 0.21 0.15 0.12 0.58

NYHA ⫽ New York Heart Association.

DISCUSSION DR W. RANDOLPH CHITWOOD (Greenville, NC): I rise to congratulate Dr Ngaage and his colleagues for this excellent presentation of what I think is seminal data regarding the ravages of atrial fibrillation when associated with mitral valve disease. First, I want to thank the authors for the opportunity and the pleasure of having read the manuscript in advance. In a case-matched cohort of patients having had modern mitral valve repairs at the Mayo Clinic, the authors have shown us that the operative mortality was increased and the operative morbidity was increased in patients who had prior atrial fibrillation. Moreover, the paper suggests that both five- and ten-year patient survival were shortened as well. However, atrial fibrillation in and of itself did not show up as an independent risk factor for late death. Those who had permanent atrial fibrillation had a significantly higher risk of death than those with intermittent episodes. Preoperative atrial fibrillation did remain an independent risk factor for late adverse cardiac events and strokes. This is really the first large surgical study to define clearly the direction in which surgeons should aspire when repairing degenerative mitral valves in the presence of preoperative atrial fibrillation. The present study differs from other author studies in that there was a case-matched design. Over the years surgeons, and especially cardiologists, largely have ignored the ravages of atrial fibrillation when combined with mitral disease. Their focus has been on replacing the valve, pharmacologic rate control, and degree of anticoagulation to provide hemodynamic stability and protect from strokes. As you know, Dr Jim Cox brought us an operation that was highly effective for surgically treating atrial fibrillation, either for standalone or combined with valve surgery. Nevertheless, the adoption rate of the combined operation remained very small, and your center is one of the very few to do the combined operation, especially when there is a complex repair needed. However, several large centers did show that the combination of the Cox-maze could be effective with a repair. In fact, Gillinov at the

Cleveland Clinic showed that at five years after the repair and a combined maze, 80% were out of atrial fibrillation. The length of preoperative atrial fibrillation, age, increased left atrial size, and higher left ventricular mass were negative factors. Thus, with the newer, quicker methods for intraoperative ablation, it seems incumbent that the treatment philosophy espoused this morning should be adopted widely. I have several questions for you, Dr Ngaage. My first question relates to a statistical matter. The study appears to have been powered sufficiently at over 200 patients per cohort for correct statistical analysis. However, all of your other information suggests that atrial fibrillation should have been a strong independent risk factor for mortality. There was a higher five- and ten-year death rate. Nevertheless, your data suggests that this was because of associated factors other than atrial fibrillation, especially in your secondary analysis. Despite your present analysis, do you think that if you had followed up all 1,020 patients in your repair group with mazes that death would have clearly risen as an independent risk factor? The second question. Can you comment on what percent of patients revert to nonfibrillating rhythms with repairs alone? I believe the Mayo Clinic has done some studies in that area. The third question. Have you analyzed the features of preoperative atrial fibrillation and tailored your maze operation that you are doing now based on those predictors for atrial fibrillation such as left atrial reduction, left-sided lesions only versus a full maze, or appendage closure? And the last question. Are you now using alternative energy sources? Since you have classically done a cut and sew maze, are you using sources like cryoablation to perform these operations quicker? I really think that this paper should be a guideline as to what we should do when repairing mitral valves in patients with atrial fibrillation. It is now left up to the rest of us to follow your lead. Thank you for the opportunity to discuss this paper.

DR NGAAGE: Thank you, Dr Chitwood. In answer to your first question about the power of the study: From previous reports we calculated a sample size of 200 patients in each group to detect a 10% difference in five-year survival between the groups at 90% power, and 149 patients in each group at 80% power, so this study was adequately powered. AF (atrial fibrillation) was not a risk factor, and I was surprised that this was the case. The association of atrial fibrillation with other cardiac morbidities like cardiomegaly and congestive cardiac failure, which were more prevalent in the AF group and exhibited profound impact on survival, may explain why atrial fibrillation was not a risk factor. When we did not include cardiomegaly and congestive cardiac failure in the analysis, preoperative AF was a risk factor for late mortality. Also, in another paper in the JTCVS (Journal of Thoracic and Cardiovascular Surgery) press for patients with coronary artery disease where cardiomegaly and congestive cardiac failure was not as prevalent, preoperative AF turned out to be a risk factor for late mortality. The classical maze,

NGAAGE ET AL PREOP AF AND REPAIR OF MITRAL REGURGITATION

443

that is the biatrial cut and sew maze, allows some reduction of the right and left atrial size and that may be of advantage in patients with cardiomegaly. Finally, at our institution now, we use the alternative energy sources like the radiofrequency ablating devices, and our early results are similar to that of classical maze. DR MICHAEL MACK (Dallas, TX): Did you compare your subgroup of 90 patients with AF who did have a maze procedure with either those that did not have a maze procedure or with sinus rhythm patients to see whether doing a maze changed the prognosis in these AF patients? DR NGAAGE: We did not compare the groups because at our institution the maze procedure at the time of this study was performed predominantly for patients who were younger, so there was a bias in the selection of patients. When we looked at the baseline variables there was great disparity between the groups, which precluded a meaningful comparison.

Requirements for Maintenance of Certification in 2008 Diplomates of the American Board of Thoracic Surgery (ABTS) who plan to participate in the Maintenance of Certification (MOC) process which will begin in 2008 must hold an unrestricted medical license in the locale of their practice and privileges in a hospital accredited by the JCAHO (or other organization recognized by the ABTS). In addition, a valid ABTS certificate is an absolute requirement for entrance into the Maintenance of Certification process. If your certificate has expired, the only pathway for renewal of a certificate is to take and pass the Part I (written) and the Part II (oral) certifying examinations. The names of individuals who have not maintained their certificate will no longer be published in the American Board of Medical Specialties directories. Diplomates’ names will be published upon successful completion of the Maintenance of Certification process. The CME requirements are 30 Category I credits earned during each year prior to application. At least half of these CME hours need to be in the broad area of thoracic surgery. Category II credits are not allowed. Interested individuals should refer to the Booklet of Information for Maintenance of Certification for a complete description of acceptable CME credits. Diplomates in the Maintenance of Certification process will be required to complete all sections of the SESATS

© 2007 by The Society of Thoracic Surgeons Published by Elsevier Inc

self-assessment examination. It is not necessary for Diplomates to purchase SESATS individually because it will be sent to them after their application has been approved. Diplomates may apply for Maintenance of Certification in the year their certificate expires, or if they wish to do so, they may apply up to two years before it expires. However, the new certificate will be dated 10 years from the date of expiration of their original certificate or most recent recertification certificate. In other words, going through the Maintenance of Certification process early does not alter the 10-year validation. Diplomates certified prior to 1976 (the year that time-limited certificates were initiated) are also required to participate in MOC if they wish to maintain valid certificates. The deadline for submission of application for the Maintenance of Certification is May 10 of each year. All ABTS diplomates will receive a letter from the Board outlining their individual timeline and MOC requirements. A brochure outlining the rules and requirements for Maintenance of Certification in thoracic surgery is available upon request from the American Board of Thoracic Surgery, 633 North St. Clair St, Suite 2320, Chicago, IL 60611; telephone (312) 202-5900; fax (312) 202-5960; e-mail: [email protected]. This booklet is also published on the website: www.abts.org.

Ann Thorac Surg 2007;84:443

•

0003-4975/07/$32.00

CARDIOVASCULAR

Ann Thorac Surg 2007;84:434 – 43