Original Article
Pulmonary vein isolation for persistent atrial fibrillation. Long-term results
Asian Cardiovascular & Thoracic Annals 2015, Vol. 23(6) 665–669 ß The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0218492315581265 aan.sagepub.com
Ovidio A Garcia-Villarreal
Abstract Background: Pulmonary vein isolation has been employed to eliminate paroxysmal atrial fibrillation. However, the longterm outcome in terms of long-standing persistent atrial fibrillation is unclear. Methods: One hundred patients with rheumatic mitral valve disease and long-standing (>1 year) persistent atrial fibrillation were operated on between July 1998 and June 2007. Mitral valve surgery and surgical isolation (cut-andsew) of the pulmonary veins were performed in all cases. Transthoracic echocardiography and 24-h Holter monitoring were obtained after 3 and 6 months and yearly thereafter. Results: Early and late follow-up was 99% and 92% complete, respectively. The endpoint was freedom from atrial fibrillation. There was one (1%) hospital death. Atrial fibrillation was present in 39%, 47%, 63% and 68% of patients at 3 months, 1 year, 3 years, and 5 years after surgery, respectively. The odds ratios for recurrence of atrial fibrillation postoperatively were 1.41 (95% confidence interval 1.14–1.74), 2.17 (95% confidence interval 1.63–2.90), and 3.62 (95% confidence interval 2.44–5.38) at 1 week, 3 years, and 5 years, respectively. Actuarial freedom from atrial fibrillation was 35% at 3 years, and 30% at 5 years. A direct relationship was observed between preoperative left atrial size >6 cm and atrial fibrillation recurrence at 5 years (p < 0.05 odds ratio ¼ 2.5, 95% confidence interval 1.15 – 5.4). Conclusions: No beneficial effects of simple pulmonary vein isolation for long-standing persistent atrial fibrillation concomitant with rheumatic mitral valve disease were observed. Atrial fibrillation cannot be fully treated using only pulmonary vein isolation.
Keywords Atrial fibrillation, Cardiac surgical procedures, Heart atria, Mitral valve, Pulmonary veins
Introduction
Patients and methods
Long-standing persistent atrial fibrillation (AF) is a common finding in patients undergoing mitral valve (MV) surgery. In our experience, this arrhythmia is present in 72% of cases of MV disease arriving in the operating room. Haı¨ ssaguerre and colleagues1 demonstrated that the triggers initiating AF are principally located in and around the pulmonary veins (PV). In the same year as the initial work of Haı¨ ssaguerre and colleagues,1 we started to isolate the PV in all cases of MV surgery with concomitant AF.2 Simple surgical isolation of the PV by means of a cut-and-sew technique is an excellent opportunity to test the effectiveness of Haı¨ ssaguerre’s statement in cases of long-standing persistent AF concomitant with rheumatic MV disease.
Between July 1998 and June 2007, 100 consecutive patients underwent PV isolation and MV surgery in our institution. All patients had rheumatic MV disease concomitant with long-standing persistent AF of more than 12-months duration. The study was approved by the institution’s ethics committee. All patients gave signed informed consent. Preoperative demographic
Department of Cardiac Surgery. Hospital of Cardiology UMAE 34, IMSS, Monterrey, Nuevo Leon, Mexico Corresponding author: Ovidio A Garcia-Villarreal, MD, Sierra Nayarita 143, Col. Virginia Tafich, 66374 Santa Catarina, Nuevo Leo´n, Me´xico. Email: [email protected]
Downloaded from aan.sagepub.com at NORTHERN KENTUCKY UNIV on June 5, 2016
666
Asian Cardiovascular & Thoracic Annals 23(6)
Table 1. Demographic, clinical, and surgical characteristics of 100 patients with atrial fibrillation. Variable
No. of patients
Age (years) Male Female Median NYHA functional class [range] Preoperative duration of atrial fibrillation (years) Preoperative left atrial size* (mm) Preoperative left ventricular ejection fraction Left atrial thrombus Mitral regurgitation and stenosis Mitral regurgitation Mitral stenosis Mitral bioprosthesis failure Preoperative PSAP (mm Hg) Preoperative antiarrhythmic therapy Aortic crossclamp time (min) Extracorporeal circulation time (min) Mechanical mitral valve replacement Biological mitral valve replacement Mitral valve repair Associated procedures Intensive care unit stay (days) Hospital stay (days)
52.8 12.6 30 70 2 [2–3] 3.6 6.5 7.4 1.08 47.6% 7.2% 20 68 22 8 2 56.3 18.5 1 78.2 23 104 37.6 64 5 31 27 2.1 0.7 7.5 1.7
*Obtained by echocardiographic study and expressed as the major bidimensional diameter of the left atrium. NYHA: New York Heart Association; PSAP: pulmonary systolic arterial pressure.
data are summarized in Table 1. Atrial fibrillation was classified as long-standing persistent AF in all cases. On the basis of the 2012 HRS/EHRA/ECAS Expert Consensus Statement on Catheter and Surgical Ablation of Atrial Fibrillation, long-standing persistent AF was defined as continuous AF of more than 12 months’ duration.3 Worthy of note is that AF was related to underlying structural rheumatic MV disease in all cases. All patients were monitored with daily 12lead electrocardiograms from intensive care unit arrival to hospital discharge. Transthoracic echocardiography and 24-h Holter monitoring were obtained after 3 and 6 months, and every year thereafter. All operations were performed by the same surgeon. Cardiopulmonary bypass with bicaval cannulation was conducted through a median sternotomy with moderate hypothermia at 28 C and multidose anterograde cold blood cardioplegia. After aortic crossclamping, the left atrium (LA) was opened parallel to the interatrial sulcus in front of the right PV. The posterior antrum of the LA containing all the PV was fully isolated. To achieve this,
Figure 1. Operative photograph showing the entire chamber containing all the pulmonary veins fully isolated from the left atrium. PV: pulmonary veins.
the left atriotomy was extended superiorly through the LA roof up to the left PV. The lower end of the left atriotomy was extended through the LA floor between the MV and the LA antrum until reaching the left PV. Thus all the PV were totally isolated. The LA appendage was resected from outside the heart. MV surgery was performed, and the LA was circumferentially sutured with double running 3/0 polypropylene sutures. No additional incisions were made in the LA (Figure 1). Immediately before aortic crossclamp removal, intravenous amiodarone infusion was started. A total dose of 1000 mg was given over the first 24 h of therapy, delivered by the following infusion regimen: first rapid: 150 mg over the first 10 min (15 mg min1); followed by slow: 360 mg over the next 6 h (1 mg min1); and finally a maintenance infusion: 540 mg over the reminding 18 h (0.5 mg min1). Postoperatively, amiodarone 200 mg was given once daily to all patients for up to 3 months. This was interrupted only if the patient was free of AF or if induced-amiodarone toxicity was observed. Data are expressed as mean and standard deviation or median and range. In the statistical analysis, the primary endpoint was freedom from AF after surgery. All surgical interventions and results were determined by reviewing the medical records. Statistical analyses were performed with SPSS version 15.0 software (SPSS, Inc., Chicago, IL, USA). The chi-square test was used to analyze data of patients who had recurrent AF and those who remained in sinus rhythm in the immediate postoperative period, at 3 months, 6 months, and every year after surgery. Student’s t test was used to compare continuous variables between patients with and without AF at the endpoint of the study (5 years). A p value < 0.05 was considered to be statistically significant.
Results The clinical, surgical, and demographical characteristics of the patient population are summarized in Table 1. In all cases, MV disease was the underlying
Downloaded from aan.sagepub.com at NORTHERN KENTUCKY UNIV on June 5, 2016
Garcia-Villarreal
667
cause of AF. Tricuspid annuloplasty was performed in 27 (27%) cases as the only associated procedure. There was only one (1%) hospital death due to persistent low cardiac output. There were 2 (2%) reoperations for postoperative bleeding unrelated to the LA anastomotic line, renal failure in 5 (5%), and persistent atrioventricular block needing definitive pacemaker implantation in 2 (2%). There was no immediate postoperative stroke. Long-term follow-up was possible in 92% of cases. Two patients died due to major stroke complications in the first year, 3 died from persistent cardiac failure in the second year, and 2 from noncardiac causes in the 3rd and 5th years. There was a trend towards AF recurrence with longer follow-up. AF was present in 39%, 47%, 63%, and 68% of patients at 3 months, 1 year, 3 years, and 5 years, respectively. Statistical analysis indicated that in the immediate postoperative period the odds ratio (OR) for recurrence of AF was 1.41 with a 95% confidence interval (CI) of 1.14–1.74. At the 3-year and 5-year follow-up, the OR for AF recurrence was 2.17 (95%CI 1.63–2.90) and 3.62
(95%CI 2.44–5.38), respectively (Table 2); the longer the follow-up, the higher the OR for recurrence of AF. The actuarial freedom from AF was 35% during the first 3 years, and 30% at 5 years (Figure 2). No significant association was found between age, sex, preoperative history of AF, left ventricular ejection fraction, or functional class, and the postoperative recurrence of AF. Preoperative LA size was 7.6 1.1 cm in patients with recurrent AF vs. 6.8 0.5 cm in those who maintained normal sinus rhythm 5 years after surgery (p < 0.05). A direct relationship was observed between preoperative LA size > 6 cm and AF recurrence at the 5-year follow-up (p < 0.05, OR ¼ 2.5, 95%CI 1.15 – 5.4). Only two patients required a permanent pacemaker. Considering that all the tricuspid valve repairs in this series were performed by partial annuloplasty (insertion of a Cosgrove band) with a no-touch technique in the septal annular area, this could be related to preexisting underlying sick sinus disease.
Discussion Table 2. Recurrent atrial fibrillation during long-term follow-up. Time
OR
95%CI
p value
3 12 24 36 48 60
1.41 1.49 1.77 2.17 3.30 3.62
1.14–1.74 1.19–1.86 1.37–2.22 1.63–2.90 2.27–4.81 2.44–5.38
0.001
Pulmonary vein isolation for persistent atrial fibrillation. Long-term results
Asian Cardiovascular & Thoracic Annals 2015, Vol. 23(6) 665–669 ß The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0218492315581265 aan.sagepub.com
Ovidio A Garcia-Villarreal
Abstract Background: Pulmonary vein isolation has been employed to eliminate paroxysmal atrial fibrillation. However, the longterm outcome in terms of long-standing persistent atrial fibrillation is unclear. Methods: One hundred patients with rheumatic mitral valve disease and long-standing (>1 year) persistent atrial fibrillation were operated on between July 1998 and June 2007. Mitral valve surgery and surgical isolation (cut-andsew) of the pulmonary veins were performed in all cases. Transthoracic echocardiography and 24-h Holter monitoring were obtained after 3 and 6 months and yearly thereafter. Results: Early and late follow-up was 99% and 92% complete, respectively. The endpoint was freedom from atrial fibrillation. There was one (1%) hospital death. Atrial fibrillation was present in 39%, 47%, 63% and 68% of patients at 3 months, 1 year, 3 years, and 5 years after surgery, respectively. The odds ratios for recurrence of atrial fibrillation postoperatively were 1.41 (95% confidence interval 1.14–1.74), 2.17 (95% confidence interval 1.63–2.90), and 3.62 (95% confidence interval 2.44–5.38) at 1 week, 3 years, and 5 years, respectively. Actuarial freedom from atrial fibrillation was 35% at 3 years, and 30% at 5 years. A direct relationship was observed between preoperative left atrial size >6 cm and atrial fibrillation recurrence at 5 years (p < 0.05 odds ratio ¼ 2.5, 95% confidence interval 1.15 – 5.4). Conclusions: No beneficial effects of simple pulmonary vein isolation for long-standing persistent atrial fibrillation concomitant with rheumatic mitral valve disease were observed. Atrial fibrillation cannot be fully treated using only pulmonary vein isolation.
Keywords Atrial fibrillation, Cardiac surgical procedures, Heart atria, Mitral valve, Pulmonary veins
Introduction
Patients and methods
Long-standing persistent atrial fibrillation (AF) is a common finding in patients undergoing mitral valve (MV) surgery. In our experience, this arrhythmia is present in 72% of cases of MV disease arriving in the operating room. Haı¨ ssaguerre and colleagues1 demonstrated that the triggers initiating AF are principally located in and around the pulmonary veins (PV). In the same year as the initial work of Haı¨ ssaguerre and colleagues,1 we started to isolate the PV in all cases of MV surgery with concomitant AF.2 Simple surgical isolation of the PV by means of a cut-and-sew technique is an excellent opportunity to test the effectiveness of Haı¨ ssaguerre’s statement in cases of long-standing persistent AF concomitant with rheumatic MV disease.
Between July 1998 and June 2007, 100 consecutive patients underwent PV isolation and MV surgery in our institution. All patients had rheumatic MV disease concomitant with long-standing persistent AF of more than 12-months duration. The study was approved by the institution’s ethics committee. All patients gave signed informed consent. Preoperative demographic
Department of Cardiac Surgery. Hospital of Cardiology UMAE 34, IMSS, Monterrey, Nuevo Leon, Mexico Corresponding author: Ovidio A Garcia-Villarreal, MD, Sierra Nayarita 143, Col. Virginia Tafich, 66374 Santa Catarina, Nuevo Leo´n, Me´xico. Email: [email protected]
Downloaded from aan.sagepub.com at NORTHERN KENTUCKY UNIV on June 5, 2016
666
Asian Cardiovascular & Thoracic Annals 23(6)
Table 1. Demographic, clinical, and surgical characteristics of 100 patients with atrial fibrillation. Variable
No. of patients
Age (years) Male Female Median NYHA functional class [range] Preoperative duration of atrial fibrillation (years) Preoperative left atrial size* (mm) Preoperative left ventricular ejection fraction Left atrial thrombus Mitral regurgitation and stenosis Mitral regurgitation Mitral stenosis Mitral bioprosthesis failure Preoperative PSAP (mm Hg) Preoperative antiarrhythmic therapy Aortic crossclamp time (min) Extracorporeal circulation time (min) Mechanical mitral valve replacement Biological mitral valve replacement Mitral valve repair Associated procedures Intensive care unit stay (days) Hospital stay (days)
52.8 12.6 30 70 2 [2–3] 3.6 6.5 7.4 1.08 47.6% 7.2% 20 68 22 8 2 56.3 18.5 1 78.2 23 104 37.6 64 5 31 27 2.1 0.7 7.5 1.7
*Obtained by echocardiographic study and expressed as the major bidimensional diameter of the left atrium. NYHA: New York Heart Association; PSAP: pulmonary systolic arterial pressure.
data are summarized in Table 1. Atrial fibrillation was classified as long-standing persistent AF in all cases. On the basis of the 2012 HRS/EHRA/ECAS Expert Consensus Statement on Catheter and Surgical Ablation of Atrial Fibrillation, long-standing persistent AF was defined as continuous AF of more than 12 months’ duration.3 Worthy of note is that AF was related to underlying structural rheumatic MV disease in all cases. All patients were monitored with daily 12lead electrocardiograms from intensive care unit arrival to hospital discharge. Transthoracic echocardiography and 24-h Holter monitoring were obtained after 3 and 6 months, and every year thereafter. All operations were performed by the same surgeon. Cardiopulmonary bypass with bicaval cannulation was conducted through a median sternotomy with moderate hypothermia at 28 C and multidose anterograde cold blood cardioplegia. After aortic crossclamping, the left atrium (LA) was opened parallel to the interatrial sulcus in front of the right PV. The posterior antrum of the LA containing all the PV was fully isolated. To achieve this,
Figure 1. Operative photograph showing the entire chamber containing all the pulmonary veins fully isolated from the left atrium. PV: pulmonary veins.
the left atriotomy was extended superiorly through the LA roof up to the left PV. The lower end of the left atriotomy was extended through the LA floor between the MV and the LA antrum until reaching the left PV. Thus all the PV were totally isolated. The LA appendage was resected from outside the heart. MV surgery was performed, and the LA was circumferentially sutured with double running 3/0 polypropylene sutures. No additional incisions were made in the LA (Figure 1). Immediately before aortic crossclamp removal, intravenous amiodarone infusion was started. A total dose of 1000 mg was given over the first 24 h of therapy, delivered by the following infusion regimen: first rapid: 150 mg over the first 10 min (15 mg min1); followed by slow: 360 mg over the next 6 h (1 mg min1); and finally a maintenance infusion: 540 mg over the reminding 18 h (0.5 mg min1). Postoperatively, amiodarone 200 mg was given once daily to all patients for up to 3 months. This was interrupted only if the patient was free of AF or if induced-amiodarone toxicity was observed. Data are expressed as mean and standard deviation or median and range. In the statistical analysis, the primary endpoint was freedom from AF after surgery. All surgical interventions and results were determined by reviewing the medical records. Statistical analyses were performed with SPSS version 15.0 software (SPSS, Inc., Chicago, IL, USA). The chi-square test was used to analyze data of patients who had recurrent AF and those who remained in sinus rhythm in the immediate postoperative period, at 3 months, 6 months, and every year after surgery. Student’s t test was used to compare continuous variables between patients with and without AF at the endpoint of the study (5 years). A p value < 0.05 was considered to be statistically significant.
Results The clinical, surgical, and demographical characteristics of the patient population are summarized in Table 1. In all cases, MV disease was the underlying
Downloaded from aan.sagepub.com at NORTHERN KENTUCKY UNIV on June 5, 2016
Garcia-Villarreal
667
cause of AF. Tricuspid annuloplasty was performed in 27 (27%) cases as the only associated procedure. There was only one (1%) hospital death due to persistent low cardiac output. There were 2 (2%) reoperations for postoperative bleeding unrelated to the LA anastomotic line, renal failure in 5 (5%), and persistent atrioventricular block needing definitive pacemaker implantation in 2 (2%). There was no immediate postoperative stroke. Long-term follow-up was possible in 92% of cases. Two patients died due to major stroke complications in the first year, 3 died from persistent cardiac failure in the second year, and 2 from noncardiac causes in the 3rd and 5th years. There was a trend towards AF recurrence with longer follow-up. AF was present in 39%, 47%, 63%, and 68% of patients at 3 months, 1 year, 3 years, and 5 years, respectively. Statistical analysis indicated that in the immediate postoperative period the odds ratio (OR) for recurrence of AF was 1.41 with a 95% confidence interval (CI) of 1.14–1.74. At the 3-year and 5-year follow-up, the OR for AF recurrence was 2.17 (95%CI 1.63–2.90) and 3.62
(95%CI 2.44–5.38), respectively (Table 2); the longer the follow-up, the higher the OR for recurrence of AF. The actuarial freedom from AF was 35% during the first 3 years, and 30% at 5 years (Figure 2). No significant association was found between age, sex, preoperative history of AF, left ventricular ejection fraction, or functional class, and the postoperative recurrence of AF. Preoperative LA size was 7.6 1.1 cm in patients with recurrent AF vs. 6.8 0.5 cm in those who maintained normal sinus rhythm 5 years after surgery (p < 0.05). A direct relationship was observed between preoperative LA size > 6 cm and AF recurrence at the 5-year follow-up (p < 0.05, OR ¼ 2.5, 95%CI 1.15 – 5.4). Only two patients required a permanent pacemaker. Considering that all the tricuspid valve repairs in this series were performed by partial annuloplasty (insertion of a Cosgrove band) with a no-touch technique in the septal annular area, this could be related to preexisting underlying sick sinus disease.
Discussion Table 2. Recurrent atrial fibrillation during long-term follow-up. Time
OR
95%CI
p value
3 12 24 36 48 60
1.41 1.49 1.77 2.17 3.30 3.62
1.14–1.74 1.19–1.86 1.37–2.22 1.63–2.90 2.27–4.81 2.44–5.38
0.001
