Postoperative atrial fibrillation is not pulmonary vein dependent: Results from a randomized trial Bob Kiaii, MD,* Stephanie Fox, BA, RRT,* Lindsay Chase, BHSc, RRT,* Michaela Fernandes,* Larry W. Stitt, MSc,† Ray Guo, MD,* Mackenzie Quantz, MD,* Michael W. Chu, MD,* Pavan Koka, MD,* R. Scott McClure, MD,* F. Neil McKenzie, MD,* George J. Klein, MD,‡ Richard J. Novick, MD,* Allan C. Skanes, MD‡ From the *Departments of Surgery, †Biostatistics, and ‡Medicine, Western University, London Health Sciences Centre, University Hospital, Ivey Cardiac Centre, London, Ontario, Canada. BACKGROUND Although often short-lived and self-limiting, postoperative atrial fibrillation (POAF) is a well-recognized postoperative complication of cardiac surgery and is associated with a 2-fold increase in cardiovascular mortality and morbidity.
postoperative inotropic support, antiarrhythmic drug use, need for oral anticoagulation, and complication rates. The mean length of postoperative hospital stay was 8.2 ⫾ 6.5 days in the ablation group and 6.7 ⫾ 4.6 days in the control group (P o .001).
OBJECTIVE Our aim was to determine whether intraoperative bilateral pulmonary vein radiofrequency ablation decreases the incidence of POAF in patients undergoing coronary artery bypass grafting (CABG).
CONCLUSION Adjuvant pulmonary vein isolation does not decrease the incidence of POAF or its clinical impact but increases the mean length of stay in the hospital. The mechanism of POAF does not appear to depend on the pulmonary veins.
METHODS A total of 175 patients undergoing CABG was prospectively randomized to undergo adjuvant bilateral radiofrequency pulmonary vein ablation in addition to CABG (group A; n ¼ 89) or CABG alone (group B; n ¼ 86). Intraoperative pulmonary vein isolation was confirmed by the inability to pace the heart via the pulmonary veins after ablation. All patients received postoperative β-blocker.
KEYWORDS Ablation; Bypass; Electrophysiology; Atrial fibrillation; Surgery
RESULTS There was no difference in the incidence of POAF in the treatment group who underwent adjuvant pulmonary vein ablation (group A; 37.1%) compared with the control group who did not (group B; 36.1%) (P ¼ .887). There were no differences in
Introduction Atrial fibrillation (AF) is a well-recognized postoperative complication of cardiac surgery, occurring in approximately 30%–40% of patients.1–5 AF occurs most frequently within 24–72 hours after surgery.3,4,6 Although often short-lived and self-limiting, postoperative AF (POAF) is associated with a 2-fold increase in cardiovascular mortality and morbidity,6 including congestive heart failure, stroke, and bleeding complications from the resulting anticoagulation. It is associated with increased total hospital stay and has an This study was sponsored by Medtronic, which supplied the Cardioblate BP2 radiofrequency bipolar device but the study conception, data analysis, and manuscript preparation were performed only by the authors without any contribution of the sponsor. The study was also sponsored by the Department of Surgery, Western University through an internal research grant. Dr Kiaii, Dr Chu, and Dr Klein are consultants for Medtronic. Address reprint requests and correspondence: Dr Bob Kiaii, London Health Sciences Centre, University Hospital, 339 Windermere Rd, London, Ontario, Canada N6A 5A5. E-mail address: [email protected].
1547-5271/$-see front matter B 2015 Heart Rhythm Society. All rights reserved.
ABBREVIATIONS AF ¼ atrial Fibrillation; CABG ¼ coronary artery bypass grafting; CPB ¼ cardiopulmonary bypass; ECG ¼ electrocardiogram; ICU ¼ intensive care unit; LA ¼ left atrium; NSR ¼ normal sinus rhythm; POAF ¼ postoperative atrial fibrillation; PV ¼ pulmonary vein; PVI ¼ pulmonary vein isolation (Heart Rhythm 2015;12:699–705) I 2015 Heart Rhythm Society. All rights reserved.
immense overall impact on patient care and hospital resources.7 A large number of clinical trials have evaluated the effectiveness of various pharmacological modalities in decreasing rates of AF after cardiac surgery.1–6,8 In the largest available β-blocker prevention trial, metoprolol decreased the prevalence of POAF from 39% to 31%.9 Oral amiodarone reduced the incidence of POAF from 30% to 16% in 600 patients undergoing elective coronary artery bypass grafting (CABG).6 No agent completely eliminated the occurrence of POAF. Because of the inconvenience in starting a preoperative agent such as amiodarone at least 6 days before the surgery and affecting the timing and scheduling of the surgery, surgeons have been reluctant to use these agents in all cases. A simple intraoperative procedure to minimize POAF that will not require preoperative planning would be ideal. The mechanism of AF after cardiac surgery is not well understood, but likely includes rapid firing from one or more http://dx.doi.org/10.1016/j.hrthm.2015.01.014
700 ectopic foci and/or unstable reentrant circuits seeking within the atria.7,10 Recent work11,12 has highlighted the role of the left atrium (LA) in POAF where there is higher frequency of activity and ectopics initiating POAF. While not specifically studied in POAF, the pulmonary veins (PVs) and the surrounding ostial areas are critical to the initiation and maintenance of AF13,14 in patients with ambulatory AF in other contexts. We therefore hypothesized that the PVs might play a similar role in POAF. Although it has never been used in the prevention of POAF, circumferential PV catheter ablation, otherwise known as pulmonary vein isolation (PVI), has become the mainstay of ablative management in other AF contexts.15–19 The current prophylactic pharmacological agents are not completely effective at eliminating the occurrence of POAF.6,8,19,20 Since PVI is effective in patients with AF in other settings,13–15 we wished to determine the contribution of PVs toward POAF. The objective of this study was to determine whether performing bilateral PVI as an adjunct to CABG in patients without a history of AF but at high risk of developing POAF would significantly reduce the occurrence of this complication and its impact on patient care.
Methods Patients and study design This prospective randomized trial with blinded evaluation of all outcomes was performed at a single center. Patients were found to be eligible to be enrolled in the study with the following inclusion criteria: patients Z18 years with CAD undergoing elective CABG with cardiopulmonary bypass (CPB) who were at high risk of developing POAF; patients were required to meet one or more of the following criteria: age Z65 years, ejection fraction r40%, hypertension, and diabetes.7 Exclusion criteria were any documented history of AF (paroxysmal/chronic), any antiarrhythmic drug use, including amiodarone, concomitant valve procedure, planned offpump CABG, patients enrolled in another study, an inability to obtain consent, contraindications to β-blockers, and patients who would be unable to attend follow-up visits.
Study protocol Between July 2006 and March 2010, 193 patients were enrolled and randomized in a 1:1 ratio to either the ablation (group A) or the control (group B) group. None of the patients were taking antiarrhythmic drugs, including amiodarone, other than β-blockers preoperatively. The flowchart for patients participating in the trial is given in Figure 1. Group A included patients undergoing CABG with CPB with PVI. Group B included patients undergoing CABG with CPB alone. The study was approved by the local institutional ethics review board.
Anesthesia and surgery protocol All patients underwent monitoring with a Swan-Ganz catheter for measurements of cardiac index and pulmonary
Heart Rhythm, Vol 12, No 4, April 2015 arterial pressures. Anesthesia was induced with intravenous propofol combined with midazolam (0.05 mg/kg) and sufentanil. Cisatracurium besylate was used for neuromuscular paralysis. Maintenance of anesthesia was with isoflurane and sufentanil. In all patients, heparin was administered at a dose of 400 IU/kg to maintain activated clotting time 4450 seconds. All surgeries were conducted via a standard median sternotomy approach. Cannulation for CPB included the ascending thoracic aorta for arterial cannulation and the right atrium for venous cannulation. In all patients, the management of CPB was similar. The flow rate was maintained at 2.4 L/(min m2) and mean arterial pressure greater than 60 mm Hg. The patient’s temperature was allowed to drift down to a lowest threshold of 35.5ºC. The CPB circuit included a Terumo Capiox SX18 membrane oxygenator (Terumo Cardiovascular Systems Corporation, Ann Arbor, MI) with a cardiotomy reservoir and a roller pump. The circuit was coated with heparin and primed with 1000 mL of Ringer lactate solution, 500 mL of pentaspan, and 250 mL of 20% mannitol.
PVI PVI was achieved in the ablation group after the initiation of CPB on the beating heart. Patients were commenced on CPB and on the beating heart in the ablation group; initially, the right PVs were ablated using the Cardioblate BP2 radiofrequency bipolar device (Medtronic, Minneapolis, MN). Before the application of radiofrequency, temporary pacing wires were used to pace and capture the right PVs. Attention was paid to deliver radiofrequency energy proximal to the pacing site (between the pacing site and the LA). A total of 3 applications of the device were performed or until isolation was confirmed by demonstrating inability to capture the LA using temporary pacing wires at the previously documented PV site. On the beating heart, the left PVs were next ablated and isolation was confirmed in a similar fashion. After PVI in the ablation group and after initiation of CPB in the control group, all patients underwent coronary artery revascularization under cardioplegic arrest by using the left internal thoracic artery, radial artery, right internal thoracic artery, and saphenous vein, as indicated for conduits. After the completion of surgery, patients were transferred to the cardiac surgery intensive care unit (ICU).
Hospital course Cardiac rhythm was monitored continuously with telemetry in the ICU and on the ward until discharge home. Full disclosure of the telemetry was reviewed every 24 hours. All arrhythmias were documented with regard to onset and immediate management. The cardiac rhythm strips were automatically printed by the telemetry monitoring unit for any episode of AF or other arrhythmia. Strips were deidentified and analyzed by an electrophysiologist blinded to the treatment group. Any treatment of POAF (antiarrhythmic agents and anticoagulation) was recorded. Patients who were
Kiaii et al
Mechanism of Postoperative Atrial Fibrillation
701
Prophylactic Pulmonary Vein Isolation as Adjunct to CABG Total # of Patients approached 220 and enrolled 193 From 2006 to 2010 - 6 month follow-up timeframe
193 Patients Enrolled 14 Done OFF PUMP and taken out of the analysis 4 CABG + Valve taken out Leaving 175 Randomized and On PUMP (ITT)
89 ITT Ablation Patients, (7 incomplete isolation, 3 did NOT have ablation (pt. safety issues) *3 intraop complications related to ablation*
86 Control Patients (ITT)
OF 89 ITT 33 had Aib (37.1%)
OF 86 ITT 31 had Aib (36.1%)
FOLLOW UP
30 Day Clinic Visit – ECG 171 pts Completed (99.4%) (171/172 able to return) ECG – All Sinus Rhythm 3 pts expired before follow up
3 early deaths: 1 stroke, 1 cardiac arrest, 1 Respiratory
6 Month FUP -Holter Monitoring 129 pts Completed (75.9%) (129/170) Holter - All Sinus Rhythm 2 additional, Total 5 pts expired
2 late deaths: 1 stroke, 1 colon cancer
Figure 1 Flowchart for patients participating in the trial. AFib. ¼ atrial fibrillation; CABG ¼ coronary artery bypass grafting; ECG ¼ electrocardiogram; FUP ¼ follow-up period; ITT ¼ intention to treat.
on β-blockers preoperatively began taking their β-blockers postoperatively. Patients who were not on β-blockers also began taking a β-blocker on their first postoperative day such that all patients in the study were on β-blockers.
Discharge After discharge, all the complications and readmissions were noted through direct contact with the patient or their family doctors specifically looking for evidence of AF. Patients were seen in follow-up 4–6 weeks after discharge from hospital postsurgery. Patients were also seen in follow-up with an electrocardiogram (ECG) at 6 months. At 6 months, patients also had a 48-hour Holter monitor. Patients who were unable to attend follow-up visits were contacted either directly or their primary care physicians/specialists were
contacted via telephone and reports (ECGs and Holter monitor) were obtained from them via mail or fax.
Clinical outcomes The primary outcome was the incidence of continuous POAF as documented on telemetry/ECG for Z5 minutes that prompted treatment6 or asymptomatic AF lasting for 430 minutes while the patients were in the hospital. This definition was chosen to detect primarily problematic POAF and clinically unimportant short-lived (r30 seconds) episodes of POAF that were asymptomatic episodes were decided to be of less clinical importance. The secondary outcome included the incidence of AF during the 6-month follow-up period after discharge from the hospital. In addition, anticoagulation use and complications related to anticoagulation use, length of stay in the ICU and
702 hospital, and mortality during the 6-month follow-up period were noted. This was based on ECGs and Holter monitors performed at each follow-up visit as well as any ECGs from other hospital visits for either AF or other reasons. If patients required anticoagulation for persistent AF, the duration of the use of anticoagulation was documented at each follow-up visit. Complications (bleeding) associated with anticoagulation use at the follow-up visits, and records from hospital admissions, were also recorded.
Statistical analysis Our local rate of AF after CABG was 40%. The sample size was calculated to detect a reduction in the incidence of POAF from 40% to 20% (relative risk reduction 50%) using a 2-sided α of .05 and a power of 80%. The calculated sample size was approximately 78 patients in each group. Descriptive statistics for patient demographic and risk characteristics by treatment allocation were presented. Between-group comparisons of postoperative complications were made using χ2 tests for AF and for the composite of death or one or more of 10 complications. For individual complications, between-group comparisons were made using the Fisher exact 2-tailed test. Wilcoxon 2-sample tests were used to compare the 2 groups with respect to hospital, length of stay in the stay, and CHADS2 scores. Statistical comparisons with probability values o.05 were considered statistically significant. All comparisons were based on the intention-to-treat principle. The statistical analysis was performed using SAS 9.2 (SAS Institute Inc, Cary, NC).
Results A total of 193 patients were enrolled. Fourteen patients (7%) underwent CABG using off-pump techniques because of intraoperative findings: 10 in the ablation group and 4 in the control group. In addition, 4 patients required a concomitant valve procedure, leaving a total of 175 patients who were randomized, 89 patients to the ablation group (group A) and 86 patients to the control group (group B). Seven patients in the ablation group had undergone incomplete ablation procedures and 3 did not undergo ablation procedures because of safety issues. The 2 groups were well balanced with respect to preoperative demographic characteristics (Table 1).
Heart Rhythm, Vol 12, No 4, April 2015 Table 1 Demographic and baseline characteristics stratified by treatment group Characteristic Age (y) BMI (kg/m2) Preoperative hemoglobin Sex: male Ventricular grade 1 2 3 4 Not done Urgency Elective Emergent Urgent COPD Recent MI Hypertension PVD CCS class 1 2 3 4 Diabetes None Diet control Oral agents Insulin Cerebrovascular disease None CVA TIA Preoperative IABP Left main involvement Preoperative renal failure o120 120–180 4180 Preoperative dialysis CHF Left atrial size (cm)
Ablation group (n ¼ 89)
Control group (n ¼ 86)
69.6 ⫾ 7.2 30.0 ⫾ 4.8 140.4 ⫾ 16.0 79 (88.8)
68.3 ⫾ 8.6 30.2 ⫾ 5.6 136.6 ⫾ 14.2 77 (89.5)
50 (56.2) 31 (34.8) 7 (7.9) 1 (1.1) 0 (0.0)
41 31 11 3 0
(47.7) (36.1) (12.8) (3.5) (0.0)
53 (59.6) 1 (1.1) 35 (39.3) 10 (11.2) 22 (24.7) 73 (82) 7 (7.9)
52 (60.5) 1 (1.2) 33 (38.4) 8 (9.3) 18 (20.9) 64 (74.4) 10 (11.6)
2 (2.3) 12 (13.5) 33 (37.1) 42 (47.2)
4 8 42 32
(4.7) (9.3) (48.8) (37.2)
59 (66.3) 2 (2.3) 16 (18.0) 12 (13.5)
53 1 19 13
(61.6) (1.2) (22.1) (15.1)
81 (91.0) 5 (5.6) 3 (3.4) 1 (1.1) 21 (23.6)
75 (87.2) 5 (5.8) 6 (7.0) 2 (2.3) 29 (33.7)
84 (94.4) 5 (5.6) 0 (0.0) 0 (0.0) 5 (5.6) 3.7
76 (88.4) 10 (11.6) 0 (0.0) 1 (1.2) 2 (2.3) 3.9
Values are presented as mean ⫾ SD or as n (%). BMI ¼ body mass index; CCS ¼ Canadian Cardiovascular Society; CHF ¼ congestive heart failure; COPD ¼ chronic obstructive pulmonary disease; CVA ¼ cerebrovascular accident; IABP ¼ intra-aortic balloon pump; MI ¼ myocardial infarction; PVD ¼ peripheral vascular disease; TIA ¼ transient ischemic attack.
Clinical outcomes Course in the hospital Thirty-three patients in the ablation group (37.1%) and 31 patients in the control group (36.1%) developed POAF (P ¼ .887). Patients in the ablation group required longer time on CPB (105.8 ⫾ 24.5 minutes vs 89.8 ⫾ 23.8 minutes; P o 0.001) and had longer operating times (259.2 ⫾ 49.3 minutes vs 233.7 ⫾ 40.1 minutes; P o .001). However, the length of cross-clamp time was not different (58.5 ⫾ 15.8 minutes in the ablation group vs 59.3 ⫾ 18.8 minutes in the control group; P ¼ .79; Table 2). Inotropic support was required in
the first 24 hours in 42 patients (47%) in the ablation group and 53 patients (61%) in the control group (P ¼ .06). There were no significant differences in clinical impact or other postoperative outcomes including antiarrhythmic use (26 [29%] in the ablation group and 29 [34%] in the control group; P ¼ .09) or need for oral anticoagulation (12 [13%] in the ablation group and 16 [19%] in the control group; P ¼ .14) (Table 3). There was also no difference in the median length of stay in the ICU or hospital, although there was statistically significant increase in the mean length of stay in
Kiaii et al Table 2
Mechanism of Postoperative Atrial Fibrillation Surgical times—intention to treat
Variable Inotropic use, n (%) Operating time Mean ⫾ SD Minimum, maximum Median Total pump time Mean ⫾ SD Minimum, maximum Median Cross-clamp time n Mean ⫾ SD Minimum, maximum Median
703 Table 3
Ablation group (n ¼ 89)
Control group (n ¼ 86)
42 (47.2)
53 (61.6)
P
End-point comparisons: intention to treat Ablation group Control group (n ¼ 89) (n ¼ 86)
Variable .06
259.2 ⫾ 49.3 180, 482 252.0
233.7 ⫾ 40.1 o.001 136, 314 235.5
105.8 ⫾ 24.5 53, 165 106.0
89.8 ⫾ 23.8 36, 141 91.0
79 58.5 ⫾ 15.8 27, 102 57.0
74 59.3 ⫾ 18.8 22, 111 58.5
o.001
.785
Statistical comparisons were made using unpaired t tests. There is some concern that a t test may not be appropriate. The underlying distribution of these variables may not be normally distributed, and for operating time the variances may not be the same. For nonparametric tests, the conclusions are the same.
the ICU and in the hospital in the ablation group (P ¼ .014 and P o .001, respectively). The CHADS2 score of the patients in the 2 groups was similar (Table 4). There were no major complications from the ablation procedure. Three patients sustained small tears in their atrial wall that were repaired without adverse effect. No reoperation for bleeding was related to the PVI. Thirty-day follow-up Of the 175 patients, 171 patients were seen in follow-up (99.4%). All were in normal sinus rhythm (NSR). There were no complications from anticoagulation. There were 3 deaths within the first 30 days after discharge from hospital, all in the control group. All 3 patients developed transient AF, and 2 had contraindications to anticoagulation and did not began taking warfarin. These 2 were both discharged in NSR. One patient was readmitted to the hospital with an acute stroke. This was found to be secondary to extensive thrombus in his carotid artery. The other patient was readmitted to his home hospital with respiratory failure from pulmonary edema. The family decided for compassionate care in view of his age, and the patient died in the hospital. The third patient began receiving anticoagulation after he developed POAF. While in the hospital, he sustained a cardiac arrest and died of anoxic brain injury. Six-month follow-up During the 6-month period after discharge, there were 2 additional deaths, resulting in a total of 5 mortalities at 6 months. One of the 2 additional deaths at 6 months was due to another stroke and the second due to metastatic colon cancer. The patient who sustained the stroke was in the ablation group but did not undergo ablation. He sustained a stroke intraoperatively. Postoperatively, he did not have AF. The patient was transferred for rehabilitation and subsequently
Atrial fibrillation Inotropes Coumadin Amiodarone Death Postoperative IABP Neurological Reoperative for bleeding Arrest/arrhythmia Renal failure Septicemia Mediastinitis Sternal dehiscence Respiratory failure Reintervention Postoperative MI One or more of 10 major complications LOS in the hospital n Mean Median Maximum LOS in the ICU n Mean Median Maximum
33 (37.1) 22 (66.7) 12 (36) 26 (79) 1 (1.1) 1 (1.1) 1 (1.1) 2 (2.3) 0 (0.0) 0 (0.0) 1 (1.1) 0 (0.0) 0 (0.0) 1 (1.1) 0 (0.0) 3 (3.4) 9 (10.1)
31 (36.1) 21 (67.7) 17 (55) 29 (94) 3 (3.5) 1 (1.2) 2 (2.3) 1 (1.2) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (1.2) 0 (0.0) 0 (0.0) 5 (5.8)
88 8.2 ⫾ 6.5 6 58
83 6.7 ⫾ 4.6 6 41
89 2.2 ⫾ 4.5 1 43
85 1.5 ⫾ 1.2 1 10
P .887 .055 .138 .090 .362 4.999 .616 4.999 — — 4.999 — — 4.999 — .246 .295
o.001
.014
Values are presented as n (%) unless otherwise indicated. Statistical comparisons were made using the Fisher exact test for individual complications and χ2 tests for atrial fibrillation and one or more of 10 major complications. For LOS variables, the Wilcoxon 2-sample test was used. IABP ¼ intra-aortic balloon pump; ICU ¼ intensive care unit; LOS ¼ length of stay; MI ¼ myocardial infarction.
died during the 6-month period. Of the remaining 170 patients, 129 patients completed their follow-up (75.9%). All were in NSR verified by 48-hour Holter monitoring. None were on anticoagulants. There were no complications secondary to anticoagulation.
Table 4
CHADS2 scores (N ¼ 175)
CHADS2 score
Ablation group (n ¼ 89)
0 23 (25.8) 1 30 (33.7) 2 27 (30.3) 3 6 (6.7) 4 3 (3.4) 5 0 (0.0) Median (Q1, Q3) 1.0 (0, 2) P value, t test .428 P value, Wilcoxon 2-sample test .567 (nonparametric)
Control group (n ¼ 86) 13 (15.1) 41 (47.7) 23 (26.7) 4 (4.7) 3 (3.5) 2 (2.3) 1.0 (1, 2)
Values are presented as n (%) unless otherwise indicated. Statistical comparisons were made using unpaired t tests, and the Wilcoxon 2-sample test was used.
704
Discussion In this randomized trial of patients undergoing CABG, the incidence of POAF and its clinical impact as measured by antiarrhythmic drug use was not reduced with adjuvant bilateral intraoperative PVI even when the data was analyzed as per protocol (Table 5). Importantly, patients who underwent concomitant PVI were exposed to longer time on CPB and longer operating time. This likely translated to longer stay in the ICU and overall increased length of stay in the hospital. Despite standardized surgical methodology and demonstration of intraoperative exit block, adjunctive bilateral PVI appears to play no role in preventing POAF. The PVs have been shown to trigger AF in almost all conditions in which this mechanism has been sought. The LA has been shown to be the source of POAF in studies using atrial activation frequency and atrial epicardial wires to record the triggers of POAF.11,12 It was therefore reasonable to test the hypothesis that POAF was triggered from the PVs and that PVI would reduce its incidence. Adjunctive bilateral PVI did not reduce the incidence of POAF when analyzed using intention to or on treatment analysis. Unlike AF in other ambulatory contexts, this suggests that the mechanism of POAF is unique and does not primarily involve the PVs, despite studies showing LA ectopic activity as the source of POAF. Biatrial pacing has been shown to reduce the incidence of POAF and length of stay, although the authors proposed that a reduction in atrial activation time was an important factor.21 A number of other potential contributors have been implicated, including atrial ischemia, inflammation, hypoxia, acidosis, oxidative stress, electrolyte abnormalities, atrial trauma, and electrophysiological changes.7,10 Pharmacological treatment with amiodarone and βblockers including sotalol1–3,6,8 has reduced POAF. Amiodarone has been found to be more effective than β-blockers.8 Amiodarone given 6 days before CABG and continued 6 days after reduced POAF from 30% to 16%.6 This protocol has not been widely adopted, which is likely related to the added inconvenience of preoperative administration. In addition, anti-inflammatory agents such as statins have not been shown to be effective.18 Typical of POAF, all patients who had developed POAF in the present study returned to NSR by 30 days. In both groups, there were no complications related to anticoagulation for POAF and the rates of AF at 30 days and 6 months were not different. In patients who were anticoagulated because of POAF, there were no complications related to anticoagulation. Five patients died in the first 6 months after discharge. Two of these patients had developed POAF, but were not anticoagulated due to perceived contraindication. One of these patients suffered a stroke, possibly as a consequence of AF highlighting the impact of AF even if transient.6 As presented in Table 4, there was no difference in the median CHADS2 score of the ablation and the control group of patients (P ¼ .43).
Heart Rhythm, Vol 12, No 4, April 2015 This study is the first to assess bilateral radiofrequency PVI as a prophylaxis for POAF in a randomized controlled trial. This study was able to specifically analyze the effect of PVI as an intervention to prevent AF without influence from other factors such as pharmacological agents,6,19,20 since none of the patients were on any antiarrhythmic drugs, including amiodarone, other than β-blockers preoperatively and all patients were on β-blockers postoperatively. The technique of PVI was standardized by multiple surgeons in the study, and PVI was always confirmed by exit block from the PVs.
Study limitations The study is limited by being from a single center. However, as a result, a unified technique and methodology was performed, resulting in greater methodological rigor including PVI confirmation. Postoperative electrophysiological testing to confirm PVI in patients with POAF in the ablation group was not done. Clearly, isolation of the PVs was demonstrated intraoperatively; however, early reconnection of PVs could have occurred within minutes to hours after PVI was completed. Although possible, it seems unlikely that large rates of recovery occurred given the method of PVI. In addition, when considered as a “strategy” to reduce POAF, PVI appears to have no impact. Moreover, the primary outcome was chosen to identify those for whom problematic Table 5 Variable
End-point comparisons: per protocol Ablation group (n ¼ 79)
Atrial fibrillation 29 (36.7) Death 1 (1.3) Postoperative IABP 0 (0.0) Neurological 1 (1.3) Reoperation for bleeding 2 (2.5) Arrest/arrhythmia 0 (0.0) Renal failure 0 (0.0) Septicemia 1 (1.3) Mediastinitis 0 (0.0) Sternal dehiscence 0 (0.0) Respiratory failure 1 (1.3) Reintervention 0 (0.0) Postoperative MI 3 (3.8) One or more of 10 major 8 (10.1) complications LOS in the hospital n 78 Mean 8.3 ⫾ 6.9 Median 6 Maximum 58 LOS in the ICU n 79 Mean 2.3 ⫾ 4.8 Median 1 Maximum 43
Control group (n ¼ 86) P 31 (36.1) 3 (3.5) 1 (1.2) 2 (2.3) 1 (1.2) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (1.2) 0 (0.0) 0 (0.0) 5 (5.8) 83 6.7 ⫾ 4.6 6 41 85 1.5 ⫾ 1.2 1 10
.93 .622 4.999 4.999 .607 — — .479 — — 4.999 — .108 .304
0
.028
Values are presented as n (%) unless otherwise indicated. Statistical comparisons were made using the Fisher exact test for individual complications and χ2 tests for atrial fibrillation and one or more of 10 major complications. For LOS variables, the Wilcoxon 2-sample test was used. IABP ¼ intra-aortic balloon pump; ICU ¼ intensive care unit; LOS ¼ length of stay; MI ¼ myocardial infarction.
Kiaii et al
Mechanism of Postoperative Atrial Fibrillation
POAF was found, recognizing that short-lived, asymptomatic episodes were unlikely to be of clinical importance. The number or duration of AF episodes was not recorded in the study; thus, we cannot exclude the possibility that PVI resulted in fewer and shortening of clinically unimportant episodes of POAF. Furthermore, at the 6-month follow-up, 40 patients were lost to follow-up; this could affect the accurate number of people in NSR at 6 months. All patients who were successfully followed were in NSR.
705
9.
10.
11.
12.
Conclusion Prophylactic PVI in patients undergoing CABG does not decrease the incidence of POAF or its clinical impact, but does prolong the length of operation and mean length of stay in the hospital. This study does not address the efficacy of this procedure in preventing future clinical AF; this will require longer follow-up. However, concomitant PVI for the reduction of POAF alone is not a viable strategy.
13.
14.
15.
References 1. Crystal E, Connolly SJ, Sleik K, Ginger TJ, Yusuf S. Interventions on prevention of postoperative atrial fibrillation in patients undergoing heart surgery: a metaanalysis. Circulation 2002;106:75–80. 2. Crystal E, Garfinkle MS, Connolly S, Ginger T, Sleik K, Yusuf S. Interventions for preventing post-operative atrial fibrillation in patients undergoing heart surgery. Cochrane Database Syst Rev 2009;4:CD003611-16. 3. DiDomenico RJ, Massad MG. Pharmacologic strategies for prevention of atrial fibrillation after open heart surgery. Ann Thorac Surg 2005;79:728–740. 4. Onalan O, Lashevsky I, Crystal E. Prophylaxis and management of postoperative atrial fibrillation. Curr Cardiol Rep 2005;7:382–390. 5. Crystal E, Healey J, Connolly SJ. Atrial fibrillation after cardiac surgery: update on the evidence on the available prophylactic interventions. Card Electrophysiol Rev 2003;7:189–192. 6. Mitchell LB, Exner DV, Wyse DG, Connolly CJ, Prystai GD, Bayes AJ, Kidd WT, Kieser T, Burgess JJ, Ferland A, MacAdams CL, Maitland A. Prophylactic Oral Amiodarone for the Prevention of Arrhythmias that Begin Early After Revascularization, Valve Replacement, or Repair: PAPABEAR: a randomized controlled trial. JAMA 2005;294:3093–3100. 7. Echahidi N, Pibarot P, O’Hara G, Mathieu P. Mechanisms, prevention, and treatment of atrial fibrillation after cardiac surgery. J Am Coll Cardiol 2008;51: 793–801. 8. Sleilaty G, Madi-Jebara S, Yazigi A, Haddad F, Hayeck G, El Rassi I, Ashoush R, Jebara V. Postoperative oral amiodarone versus oral bisoprolol as prophylaxis
16.
17.
18.
19.
20.
21.
against atrial fibrillation after coronary artery bypass graft surgery: a prospective randomized trial. Int J Cardiol 2009;137:116–122. Connolly SJ, Cybulsky I, Lamy A, Roberts RS, O’Brien B, Carroll S, Crystal E, Thorpe KE, Gent M. Double-blind, placebo-controlled, randomized trial of prophylactic metoprolol for reduction of hospital length of stay after heart surgery: the beta-Blocker Length Of Stay (BLOS) study. Am Heart J 2003;145: 226–232. Ishii Y, Schuessler RB, Gaynor SL, Yamada K, Fu AS, Boineau JP, Damiano RJ. Inflammation of atrium after cardiac surgery is associated with inhomogeneity of atrial conduction and atrial fibrillation. Circulation 2005;111:2881–2888. Pagé P, Vinet A, Yin Y, Cardinal R. The site of origin of atrial fibrillation occurring after cardiac surgery is located in the left atrium. Can J Cardiol 2000;16:226F–227F. Swartz MF, Fink GW, Lutz CJ, Taffet SM, Berenfeld O, Vikstrom KL, Kasprowicz K, Bhatta l, Puskas F, Kalifa J, Jalife J. Left versus right atrial differences in dominant frequency, Kþ channel transcripts, and fibrosis in patients developing atrial fibrillation after cardiac surgery. Heart Rhythm 2009;6: 1415–1422. Vicol C, Eifert S, Kur F, Reichart B. Minimally invasive off-pump pulmonary vein isolation to treat paroxysmal atrial fibrillation. Thorac Cardiovasc Surg 2005;53:176–178. Haissaguerre M, Jais P, Shah DC, Takahashi A, Hocini M, Quiniou G, Garrigue S, Le Mouroux A, Le Metayer P, Clementy J. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 1998;339:659–666. Pappone C, Rosanio S, Oreto G, Tocchi M, Giugliotta F, Vicedomni G, Salvati A, Dicandia C, Mazzone P, Santinelli V, Gulletta S, Chierchia S. Circumferential radiofrequency ablation of pulmonary vein ostia: a new anatomic approach for curing atrial fibrillation. Circulation 2000;102:2619–2628. Pappone C, Augello G, Sala S, et al. A randomized trial of circumferential pulmonary vein ablation versus antiarrhythmic drug therapy in paroxysmal atrial fibrillation: the APAF Study. J Am Coll Cardiol 2006;48: 2340–2347. Piccini JP LR, Kong MH. Pulmonary vein isolation for the maintenance of sinus rhythm in patients with atrial fibrillation: a meta-analysis of randomized, controlled trials. Circ Arrhythm Electrophysiol 2009;2:626–633. Parkash R, Tang AS, Sapp JL. Approach to the catheter ablation technique of paroxysmal and persistent atrial fibrillation: a meta-analysis of the randomized controlled trials. J Cardiovasc Electrophysiol 2011;22:729–738. Patti G, Chello M, Candura D, Pasceri V, D’Ambrosio A, Covino E, Di Sciascio G. Randomized trial of atorvastatin for reduction of postoperative atrial fibrillation in patients undergoing cardiac surgery: results of the ARMYDA-3 (Atorvastatin for Reduction of MYocardial Dysrhythmia After cardiac surgery) study. Circulation 2006;114:1455–1461. Hamid M, Kamal RS, Sami SA, Atiq F, Shafquat A, Naqvi HI, Khan FH. Effect of single dose magnesium on arrhythmias in patients undergoing coronary artery bypass surgery. J Pak Med Assoc 2008;58:22–27. Fan K, Lee KL, Chiu CS, Lee JW, He GW, Cheung D, Sun MP, Lau CP. Effects of biatrial pacing in prevention of postoperative atrial fibrillation after coronary artery bypass surgery. Circulation 2000;102:755–760.
CLINICAL PERSPECTIVES Postoperative atrial fibrillation (POAF) still occurs in 30%–40% of patients undergoing cardiac surgery. Despite being short-lived and self-limiting in most cases, studies have repeatedly demonstrated increased lengths of stay and increased morbidity and mortality. Atrial fibrillation (AF) outside the postoperative period appears to be triggered from the pulmonary veins in almost all circumstances where it has been investigated. Preliminary evidence suggests that this might be true for postoperative AF also. In this randomized controlled trial, bilateral surgical pulmonary vein isolation, using a radiofrequency bipolar clamp catheter, failed to reduce the incidence of POAF, but lengthened stay in the ICU and hospital. Despite preliminary evidence, it does not appear that POAF is pulmonary vein dependent. Pulmonary vein isolation should not be performed to prevent POAF.
postoperative inotropic support, antiarrhythmic drug use, need for oral anticoagulation, and complication rates. The mean length of postoperative hospital stay was 8.2 ⫾ 6.5 days in the ablation group and 6.7 ⫾ 4.6 days in the control group (P o .001).
OBJECTIVE Our aim was to determine whether intraoperative bilateral pulmonary vein radiofrequency ablation decreases the incidence of POAF in patients undergoing coronary artery bypass grafting (CABG).
CONCLUSION Adjuvant pulmonary vein isolation does not decrease the incidence of POAF or its clinical impact but increases the mean length of stay in the hospital. The mechanism of POAF does not appear to depend on the pulmonary veins.
METHODS A total of 175 patients undergoing CABG was prospectively randomized to undergo adjuvant bilateral radiofrequency pulmonary vein ablation in addition to CABG (group A; n ¼ 89) or CABG alone (group B; n ¼ 86). Intraoperative pulmonary vein isolation was confirmed by the inability to pace the heart via the pulmonary veins after ablation. All patients received postoperative β-blocker.
KEYWORDS Ablation; Bypass; Electrophysiology; Atrial fibrillation; Surgery
RESULTS There was no difference in the incidence of POAF in the treatment group who underwent adjuvant pulmonary vein ablation (group A; 37.1%) compared with the control group who did not (group B; 36.1%) (P ¼ .887). There were no differences in
Introduction Atrial fibrillation (AF) is a well-recognized postoperative complication of cardiac surgery, occurring in approximately 30%–40% of patients.1–5 AF occurs most frequently within 24–72 hours after surgery.3,4,6 Although often short-lived and self-limiting, postoperative AF (POAF) is associated with a 2-fold increase in cardiovascular mortality and morbidity,6 including congestive heart failure, stroke, and bleeding complications from the resulting anticoagulation. It is associated with increased total hospital stay and has an This study was sponsored by Medtronic, which supplied the Cardioblate BP2 radiofrequency bipolar device but the study conception, data analysis, and manuscript preparation were performed only by the authors without any contribution of the sponsor. The study was also sponsored by the Department of Surgery, Western University through an internal research grant. Dr Kiaii, Dr Chu, and Dr Klein are consultants for Medtronic. Address reprint requests and correspondence: Dr Bob Kiaii, London Health Sciences Centre, University Hospital, 339 Windermere Rd, London, Ontario, Canada N6A 5A5. E-mail address: [email protected].
1547-5271/$-see front matter B 2015 Heart Rhythm Society. All rights reserved.
ABBREVIATIONS AF ¼ atrial Fibrillation; CABG ¼ coronary artery bypass grafting; CPB ¼ cardiopulmonary bypass; ECG ¼ electrocardiogram; ICU ¼ intensive care unit; LA ¼ left atrium; NSR ¼ normal sinus rhythm; POAF ¼ postoperative atrial fibrillation; PV ¼ pulmonary vein; PVI ¼ pulmonary vein isolation (Heart Rhythm 2015;12:699–705) I 2015 Heart Rhythm Society. All rights reserved.
immense overall impact on patient care and hospital resources.7 A large number of clinical trials have evaluated the effectiveness of various pharmacological modalities in decreasing rates of AF after cardiac surgery.1–6,8 In the largest available β-blocker prevention trial, metoprolol decreased the prevalence of POAF from 39% to 31%.9 Oral amiodarone reduced the incidence of POAF from 30% to 16% in 600 patients undergoing elective coronary artery bypass grafting (CABG).6 No agent completely eliminated the occurrence of POAF. Because of the inconvenience in starting a preoperative agent such as amiodarone at least 6 days before the surgery and affecting the timing and scheduling of the surgery, surgeons have been reluctant to use these agents in all cases. A simple intraoperative procedure to minimize POAF that will not require preoperative planning would be ideal. The mechanism of AF after cardiac surgery is not well understood, but likely includes rapid firing from one or more http://dx.doi.org/10.1016/j.hrthm.2015.01.014
700 ectopic foci and/or unstable reentrant circuits seeking within the atria.7,10 Recent work11,12 has highlighted the role of the left atrium (LA) in POAF where there is higher frequency of activity and ectopics initiating POAF. While not specifically studied in POAF, the pulmonary veins (PVs) and the surrounding ostial areas are critical to the initiation and maintenance of AF13,14 in patients with ambulatory AF in other contexts. We therefore hypothesized that the PVs might play a similar role in POAF. Although it has never been used in the prevention of POAF, circumferential PV catheter ablation, otherwise known as pulmonary vein isolation (PVI), has become the mainstay of ablative management in other AF contexts.15–19 The current prophylactic pharmacological agents are not completely effective at eliminating the occurrence of POAF.6,8,19,20 Since PVI is effective in patients with AF in other settings,13–15 we wished to determine the contribution of PVs toward POAF. The objective of this study was to determine whether performing bilateral PVI as an adjunct to CABG in patients without a history of AF but at high risk of developing POAF would significantly reduce the occurrence of this complication and its impact on patient care.
Methods Patients and study design This prospective randomized trial with blinded evaluation of all outcomes was performed at a single center. Patients were found to be eligible to be enrolled in the study with the following inclusion criteria: patients Z18 years with CAD undergoing elective CABG with cardiopulmonary bypass (CPB) who were at high risk of developing POAF; patients were required to meet one or more of the following criteria: age Z65 years, ejection fraction r40%, hypertension, and diabetes.7 Exclusion criteria were any documented history of AF (paroxysmal/chronic), any antiarrhythmic drug use, including amiodarone, concomitant valve procedure, planned offpump CABG, patients enrolled in another study, an inability to obtain consent, contraindications to β-blockers, and patients who would be unable to attend follow-up visits.
Study protocol Between July 2006 and March 2010, 193 patients were enrolled and randomized in a 1:1 ratio to either the ablation (group A) or the control (group B) group. None of the patients were taking antiarrhythmic drugs, including amiodarone, other than β-blockers preoperatively. The flowchart for patients participating in the trial is given in Figure 1. Group A included patients undergoing CABG with CPB with PVI. Group B included patients undergoing CABG with CPB alone. The study was approved by the local institutional ethics review board.
Anesthesia and surgery protocol All patients underwent monitoring with a Swan-Ganz catheter for measurements of cardiac index and pulmonary
Heart Rhythm, Vol 12, No 4, April 2015 arterial pressures. Anesthesia was induced with intravenous propofol combined with midazolam (0.05 mg/kg) and sufentanil. Cisatracurium besylate was used for neuromuscular paralysis. Maintenance of anesthesia was with isoflurane and sufentanil. In all patients, heparin was administered at a dose of 400 IU/kg to maintain activated clotting time 4450 seconds. All surgeries were conducted via a standard median sternotomy approach. Cannulation for CPB included the ascending thoracic aorta for arterial cannulation and the right atrium for venous cannulation. In all patients, the management of CPB was similar. The flow rate was maintained at 2.4 L/(min m2) and mean arterial pressure greater than 60 mm Hg. The patient’s temperature was allowed to drift down to a lowest threshold of 35.5ºC. The CPB circuit included a Terumo Capiox SX18 membrane oxygenator (Terumo Cardiovascular Systems Corporation, Ann Arbor, MI) with a cardiotomy reservoir and a roller pump. The circuit was coated with heparin and primed with 1000 mL of Ringer lactate solution, 500 mL of pentaspan, and 250 mL of 20% mannitol.
PVI PVI was achieved in the ablation group after the initiation of CPB on the beating heart. Patients were commenced on CPB and on the beating heart in the ablation group; initially, the right PVs were ablated using the Cardioblate BP2 radiofrequency bipolar device (Medtronic, Minneapolis, MN). Before the application of radiofrequency, temporary pacing wires were used to pace and capture the right PVs. Attention was paid to deliver radiofrequency energy proximal to the pacing site (between the pacing site and the LA). A total of 3 applications of the device were performed or until isolation was confirmed by demonstrating inability to capture the LA using temporary pacing wires at the previously documented PV site. On the beating heart, the left PVs were next ablated and isolation was confirmed in a similar fashion. After PVI in the ablation group and after initiation of CPB in the control group, all patients underwent coronary artery revascularization under cardioplegic arrest by using the left internal thoracic artery, radial artery, right internal thoracic artery, and saphenous vein, as indicated for conduits. After the completion of surgery, patients were transferred to the cardiac surgery intensive care unit (ICU).
Hospital course Cardiac rhythm was monitored continuously with telemetry in the ICU and on the ward until discharge home. Full disclosure of the telemetry was reviewed every 24 hours. All arrhythmias were documented with regard to onset and immediate management. The cardiac rhythm strips were automatically printed by the telemetry monitoring unit for any episode of AF or other arrhythmia. Strips were deidentified and analyzed by an electrophysiologist blinded to the treatment group. Any treatment of POAF (antiarrhythmic agents and anticoagulation) was recorded. Patients who were
Kiaii et al
Mechanism of Postoperative Atrial Fibrillation
701
Prophylactic Pulmonary Vein Isolation as Adjunct to CABG Total # of Patients approached 220 and enrolled 193 From 2006 to 2010 - 6 month follow-up timeframe
193 Patients Enrolled 14 Done OFF PUMP and taken out of the analysis 4 CABG + Valve taken out Leaving 175 Randomized and On PUMP (ITT)
89 ITT Ablation Patients, (7 incomplete isolation, 3 did NOT have ablation (pt. safety issues) *3 intraop complications related to ablation*
86 Control Patients (ITT)
OF 89 ITT 33 had Aib (37.1%)
OF 86 ITT 31 had Aib (36.1%)
FOLLOW UP
30 Day Clinic Visit – ECG 171 pts Completed (99.4%) (171/172 able to return) ECG – All Sinus Rhythm 3 pts expired before follow up
3 early deaths: 1 stroke, 1 cardiac arrest, 1 Respiratory
6 Month FUP -Holter Monitoring 129 pts Completed (75.9%) (129/170) Holter - All Sinus Rhythm 2 additional, Total 5 pts expired
2 late deaths: 1 stroke, 1 colon cancer
Figure 1 Flowchart for patients participating in the trial. AFib. ¼ atrial fibrillation; CABG ¼ coronary artery bypass grafting; ECG ¼ electrocardiogram; FUP ¼ follow-up period; ITT ¼ intention to treat.
on β-blockers preoperatively began taking their β-blockers postoperatively. Patients who were not on β-blockers also began taking a β-blocker on their first postoperative day such that all patients in the study were on β-blockers.
Discharge After discharge, all the complications and readmissions were noted through direct contact with the patient or their family doctors specifically looking for evidence of AF. Patients were seen in follow-up 4–6 weeks after discharge from hospital postsurgery. Patients were also seen in follow-up with an electrocardiogram (ECG) at 6 months. At 6 months, patients also had a 48-hour Holter monitor. Patients who were unable to attend follow-up visits were contacted either directly or their primary care physicians/specialists were
contacted via telephone and reports (ECGs and Holter monitor) were obtained from them via mail or fax.
Clinical outcomes The primary outcome was the incidence of continuous POAF as documented on telemetry/ECG for Z5 minutes that prompted treatment6 or asymptomatic AF lasting for 430 minutes while the patients were in the hospital. This definition was chosen to detect primarily problematic POAF and clinically unimportant short-lived (r30 seconds) episodes of POAF that were asymptomatic episodes were decided to be of less clinical importance. The secondary outcome included the incidence of AF during the 6-month follow-up period after discharge from the hospital. In addition, anticoagulation use and complications related to anticoagulation use, length of stay in the ICU and
702 hospital, and mortality during the 6-month follow-up period were noted. This was based on ECGs and Holter monitors performed at each follow-up visit as well as any ECGs from other hospital visits for either AF or other reasons. If patients required anticoagulation for persistent AF, the duration of the use of anticoagulation was documented at each follow-up visit. Complications (bleeding) associated with anticoagulation use at the follow-up visits, and records from hospital admissions, were also recorded.
Statistical analysis Our local rate of AF after CABG was 40%. The sample size was calculated to detect a reduction in the incidence of POAF from 40% to 20% (relative risk reduction 50%) using a 2-sided α of .05 and a power of 80%. The calculated sample size was approximately 78 patients in each group. Descriptive statistics for patient demographic and risk characteristics by treatment allocation were presented. Between-group comparisons of postoperative complications were made using χ2 tests for AF and for the composite of death or one or more of 10 complications. For individual complications, between-group comparisons were made using the Fisher exact 2-tailed test. Wilcoxon 2-sample tests were used to compare the 2 groups with respect to hospital, length of stay in the stay, and CHADS2 scores. Statistical comparisons with probability values o.05 were considered statistically significant. All comparisons were based on the intention-to-treat principle. The statistical analysis was performed using SAS 9.2 (SAS Institute Inc, Cary, NC).
Results A total of 193 patients were enrolled. Fourteen patients (7%) underwent CABG using off-pump techniques because of intraoperative findings: 10 in the ablation group and 4 in the control group. In addition, 4 patients required a concomitant valve procedure, leaving a total of 175 patients who were randomized, 89 patients to the ablation group (group A) and 86 patients to the control group (group B). Seven patients in the ablation group had undergone incomplete ablation procedures and 3 did not undergo ablation procedures because of safety issues. The 2 groups were well balanced with respect to preoperative demographic characteristics (Table 1).
Heart Rhythm, Vol 12, No 4, April 2015 Table 1 Demographic and baseline characteristics stratified by treatment group Characteristic Age (y) BMI (kg/m2) Preoperative hemoglobin Sex: male Ventricular grade 1 2 3 4 Not done Urgency Elective Emergent Urgent COPD Recent MI Hypertension PVD CCS class 1 2 3 4 Diabetes None Diet control Oral agents Insulin Cerebrovascular disease None CVA TIA Preoperative IABP Left main involvement Preoperative renal failure o120 120–180 4180 Preoperative dialysis CHF Left atrial size (cm)
Ablation group (n ¼ 89)
Control group (n ¼ 86)
69.6 ⫾ 7.2 30.0 ⫾ 4.8 140.4 ⫾ 16.0 79 (88.8)
68.3 ⫾ 8.6 30.2 ⫾ 5.6 136.6 ⫾ 14.2 77 (89.5)
50 (56.2) 31 (34.8) 7 (7.9) 1 (1.1) 0 (0.0)
41 31 11 3 0
(47.7) (36.1) (12.8) (3.5) (0.0)
53 (59.6) 1 (1.1) 35 (39.3) 10 (11.2) 22 (24.7) 73 (82) 7 (7.9)
52 (60.5) 1 (1.2) 33 (38.4) 8 (9.3) 18 (20.9) 64 (74.4) 10 (11.6)
2 (2.3) 12 (13.5) 33 (37.1) 42 (47.2)
4 8 42 32
(4.7) (9.3) (48.8) (37.2)
59 (66.3) 2 (2.3) 16 (18.0) 12 (13.5)
53 1 19 13
(61.6) (1.2) (22.1) (15.1)
81 (91.0) 5 (5.6) 3 (3.4) 1 (1.1) 21 (23.6)
75 (87.2) 5 (5.8) 6 (7.0) 2 (2.3) 29 (33.7)
84 (94.4) 5 (5.6) 0 (0.0) 0 (0.0) 5 (5.6) 3.7
76 (88.4) 10 (11.6) 0 (0.0) 1 (1.2) 2 (2.3) 3.9
Values are presented as mean ⫾ SD or as n (%). BMI ¼ body mass index; CCS ¼ Canadian Cardiovascular Society; CHF ¼ congestive heart failure; COPD ¼ chronic obstructive pulmonary disease; CVA ¼ cerebrovascular accident; IABP ¼ intra-aortic balloon pump; MI ¼ myocardial infarction; PVD ¼ peripheral vascular disease; TIA ¼ transient ischemic attack.
Clinical outcomes Course in the hospital Thirty-three patients in the ablation group (37.1%) and 31 patients in the control group (36.1%) developed POAF (P ¼ .887). Patients in the ablation group required longer time on CPB (105.8 ⫾ 24.5 minutes vs 89.8 ⫾ 23.8 minutes; P o 0.001) and had longer operating times (259.2 ⫾ 49.3 minutes vs 233.7 ⫾ 40.1 minutes; P o .001). However, the length of cross-clamp time was not different (58.5 ⫾ 15.8 minutes in the ablation group vs 59.3 ⫾ 18.8 minutes in the control group; P ¼ .79; Table 2). Inotropic support was required in
the first 24 hours in 42 patients (47%) in the ablation group and 53 patients (61%) in the control group (P ¼ .06). There were no significant differences in clinical impact or other postoperative outcomes including antiarrhythmic use (26 [29%] in the ablation group and 29 [34%] in the control group; P ¼ .09) or need for oral anticoagulation (12 [13%] in the ablation group and 16 [19%] in the control group; P ¼ .14) (Table 3). There was also no difference in the median length of stay in the ICU or hospital, although there was statistically significant increase in the mean length of stay in
Kiaii et al Table 2
Mechanism of Postoperative Atrial Fibrillation Surgical times—intention to treat
Variable Inotropic use, n (%) Operating time Mean ⫾ SD Minimum, maximum Median Total pump time Mean ⫾ SD Minimum, maximum Median Cross-clamp time n Mean ⫾ SD Minimum, maximum Median
703 Table 3
Ablation group (n ¼ 89)
Control group (n ¼ 86)
42 (47.2)
53 (61.6)
P
End-point comparisons: intention to treat Ablation group Control group (n ¼ 89) (n ¼ 86)
Variable .06
259.2 ⫾ 49.3 180, 482 252.0
233.7 ⫾ 40.1 o.001 136, 314 235.5
105.8 ⫾ 24.5 53, 165 106.0
89.8 ⫾ 23.8 36, 141 91.0
79 58.5 ⫾ 15.8 27, 102 57.0
74 59.3 ⫾ 18.8 22, 111 58.5
o.001
.785
Statistical comparisons were made using unpaired t tests. There is some concern that a t test may not be appropriate. The underlying distribution of these variables may not be normally distributed, and for operating time the variances may not be the same. For nonparametric tests, the conclusions are the same.
the ICU and in the hospital in the ablation group (P ¼ .014 and P o .001, respectively). The CHADS2 score of the patients in the 2 groups was similar (Table 4). There were no major complications from the ablation procedure. Three patients sustained small tears in their atrial wall that were repaired without adverse effect. No reoperation for bleeding was related to the PVI. Thirty-day follow-up Of the 175 patients, 171 patients were seen in follow-up (99.4%). All were in normal sinus rhythm (NSR). There were no complications from anticoagulation. There were 3 deaths within the first 30 days after discharge from hospital, all in the control group. All 3 patients developed transient AF, and 2 had contraindications to anticoagulation and did not began taking warfarin. These 2 were both discharged in NSR. One patient was readmitted to the hospital with an acute stroke. This was found to be secondary to extensive thrombus in his carotid artery. The other patient was readmitted to his home hospital with respiratory failure from pulmonary edema. The family decided for compassionate care in view of his age, and the patient died in the hospital. The third patient began receiving anticoagulation after he developed POAF. While in the hospital, he sustained a cardiac arrest and died of anoxic brain injury. Six-month follow-up During the 6-month period after discharge, there were 2 additional deaths, resulting in a total of 5 mortalities at 6 months. One of the 2 additional deaths at 6 months was due to another stroke and the second due to metastatic colon cancer. The patient who sustained the stroke was in the ablation group but did not undergo ablation. He sustained a stroke intraoperatively. Postoperatively, he did not have AF. The patient was transferred for rehabilitation and subsequently
Atrial fibrillation Inotropes Coumadin Amiodarone Death Postoperative IABP Neurological Reoperative for bleeding Arrest/arrhythmia Renal failure Septicemia Mediastinitis Sternal dehiscence Respiratory failure Reintervention Postoperative MI One or more of 10 major complications LOS in the hospital n Mean Median Maximum LOS in the ICU n Mean Median Maximum
33 (37.1) 22 (66.7) 12 (36) 26 (79) 1 (1.1) 1 (1.1) 1 (1.1) 2 (2.3) 0 (0.0) 0 (0.0) 1 (1.1) 0 (0.0) 0 (0.0) 1 (1.1) 0 (0.0) 3 (3.4) 9 (10.1)
31 (36.1) 21 (67.7) 17 (55) 29 (94) 3 (3.5) 1 (1.2) 2 (2.3) 1 (1.2) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (1.2) 0 (0.0) 0 (0.0) 5 (5.8)
88 8.2 ⫾ 6.5 6 58
83 6.7 ⫾ 4.6 6 41
89 2.2 ⫾ 4.5 1 43
85 1.5 ⫾ 1.2 1 10
P .887 .055 .138 .090 .362 4.999 .616 4.999 — — 4.999 — — 4.999 — .246 .295
o.001
.014
Values are presented as n (%) unless otherwise indicated. Statistical comparisons were made using the Fisher exact test for individual complications and χ2 tests for atrial fibrillation and one or more of 10 major complications. For LOS variables, the Wilcoxon 2-sample test was used. IABP ¼ intra-aortic balloon pump; ICU ¼ intensive care unit; LOS ¼ length of stay; MI ¼ myocardial infarction.
died during the 6-month period. Of the remaining 170 patients, 129 patients completed their follow-up (75.9%). All were in NSR verified by 48-hour Holter monitoring. None were on anticoagulants. There were no complications secondary to anticoagulation.
Table 4
CHADS2 scores (N ¼ 175)
CHADS2 score
Ablation group (n ¼ 89)
0 23 (25.8) 1 30 (33.7) 2 27 (30.3) 3 6 (6.7) 4 3 (3.4) 5 0 (0.0) Median (Q1, Q3) 1.0 (0, 2) P value, t test .428 P value, Wilcoxon 2-sample test .567 (nonparametric)
Control group (n ¼ 86) 13 (15.1) 41 (47.7) 23 (26.7) 4 (4.7) 3 (3.5) 2 (2.3) 1.0 (1, 2)
Values are presented as n (%) unless otherwise indicated. Statistical comparisons were made using unpaired t tests, and the Wilcoxon 2-sample test was used.
704
Discussion In this randomized trial of patients undergoing CABG, the incidence of POAF and its clinical impact as measured by antiarrhythmic drug use was not reduced with adjuvant bilateral intraoperative PVI even when the data was analyzed as per protocol (Table 5). Importantly, patients who underwent concomitant PVI were exposed to longer time on CPB and longer operating time. This likely translated to longer stay in the ICU and overall increased length of stay in the hospital. Despite standardized surgical methodology and demonstration of intraoperative exit block, adjunctive bilateral PVI appears to play no role in preventing POAF. The PVs have been shown to trigger AF in almost all conditions in which this mechanism has been sought. The LA has been shown to be the source of POAF in studies using atrial activation frequency and atrial epicardial wires to record the triggers of POAF.11,12 It was therefore reasonable to test the hypothesis that POAF was triggered from the PVs and that PVI would reduce its incidence. Adjunctive bilateral PVI did not reduce the incidence of POAF when analyzed using intention to or on treatment analysis. Unlike AF in other ambulatory contexts, this suggests that the mechanism of POAF is unique and does not primarily involve the PVs, despite studies showing LA ectopic activity as the source of POAF. Biatrial pacing has been shown to reduce the incidence of POAF and length of stay, although the authors proposed that a reduction in atrial activation time was an important factor.21 A number of other potential contributors have been implicated, including atrial ischemia, inflammation, hypoxia, acidosis, oxidative stress, electrolyte abnormalities, atrial trauma, and electrophysiological changes.7,10 Pharmacological treatment with amiodarone and βblockers including sotalol1–3,6,8 has reduced POAF. Amiodarone has been found to be more effective than β-blockers.8 Amiodarone given 6 days before CABG and continued 6 days after reduced POAF from 30% to 16%.6 This protocol has not been widely adopted, which is likely related to the added inconvenience of preoperative administration. In addition, anti-inflammatory agents such as statins have not been shown to be effective.18 Typical of POAF, all patients who had developed POAF in the present study returned to NSR by 30 days. In both groups, there were no complications related to anticoagulation for POAF and the rates of AF at 30 days and 6 months were not different. In patients who were anticoagulated because of POAF, there were no complications related to anticoagulation. Five patients died in the first 6 months after discharge. Two of these patients had developed POAF, but were not anticoagulated due to perceived contraindication. One of these patients suffered a stroke, possibly as a consequence of AF highlighting the impact of AF even if transient.6 As presented in Table 4, there was no difference in the median CHADS2 score of the ablation and the control group of patients (P ¼ .43).
Heart Rhythm, Vol 12, No 4, April 2015 This study is the first to assess bilateral radiofrequency PVI as a prophylaxis for POAF in a randomized controlled trial. This study was able to specifically analyze the effect of PVI as an intervention to prevent AF without influence from other factors such as pharmacological agents,6,19,20 since none of the patients were on any antiarrhythmic drugs, including amiodarone, other than β-blockers preoperatively and all patients were on β-blockers postoperatively. The technique of PVI was standardized by multiple surgeons in the study, and PVI was always confirmed by exit block from the PVs.
Study limitations The study is limited by being from a single center. However, as a result, a unified technique and methodology was performed, resulting in greater methodological rigor including PVI confirmation. Postoperative electrophysiological testing to confirm PVI in patients with POAF in the ablation group was not done. Clearly, isolation of the PVs was demonstrated intraoperatively; however, early reconnection of PVs could have occurred within minutes to hours after PVI was completed. Although possible, it seems unlikely that large rates of recovery occurred given the method of PVI. In addition, when considered as a “strategy” to reduce POAF, PVI appears to have no impact. Moreover, the primary outcome was chosen to identify those for whom problematic Table 5 Variable
End-point comparisons: per protocol Ablation group (n ¼ 79)
Atrial fibrillation 29 (36.7) Death 1 (1.3) Postoperative IABP 0 (0.0) Neurological 1 (1.3) Reoperation for bleeding 2 (2.5) Arrest/arrhythmia 0 (0.0) Renal failure 0 (0.0) Septicemia 1 (1.3) Mediastinitis 0 (0.0) Sternal dehiscence 0 (0.0) Respiratory failure 1 (1.3) Reintervention 0 (0.0) Postoperative MI 3 (3.8) One or more of 10 major 8 (10.1) complications LOS in the hospital n 78 Mean 8.3 ⫾ 6.9 Median 6 Maximum 58 LOS in the ICU n 79 Mean 2.3 ⫾ 4.8 Median 1 Maximum 43
Control group (n ¼ 86) P 31 (36.1) 3 (3.5) 1 (1.2) 2 (2.3) 1 (1.2) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (1.2) 0 (0.0) 0 (0.0) 5 (5.8) 83 6.7 ⫾ 4.6 6 41 85 1.5 ⫾ 1.2 1 10
.93 .622 4.999 4.999 .607 — — .479 — — 4.999 — .108 .304
0
.028
Values are presented as n (%) unless otherwise indicated. Statistical comparisons were made using the Fisher exact test for individual complications and χ2 tests for atrial fibrillation and one or more of 10 major complications. For LOS variables, the Wilcoxon 2-sample test was used. IABP ¼ intra-aortic balloon pump; ICU ¼ intensive care unit; LOS ¼ length of stay; MI ¼ myocardial infarction.
Kiaii et al
Mechanism of Postoperative Atrial Fibrillation
POAF was found, recognizing that short-lived, asymptomatic episodes were unlikely to be of clinical importance. The number or duration of AF episodes was not recorded in the study; thus, we cannot exclude the possibility that PVI resulted in fewer and shortening of clinically unimportant episodes of POAF. Furthermore, at the 6-month follow-up, 40 patients were lost to follow-up; this could affect the accurate number of people in NSR at 6 months. All patients who were successfully followed were in NSR.
705
9.
10.
11.
12.
Conclusion Prophylactic PVI in patients undergoing CABG does not decrease the incidence of POAF or its clinical impact, but does prolong the length of operation and mean length of stay in the hospital. This study does not address the efficacy of this procedure in preventing future clinical AF; this will require longer follow-up. However, concomitant PVI for the reduction of POAF alone is not a viable strategy.
13.
14.
15.
References 1. Crystal E, Connolly SJ, Sleik K, Ginger TJ, Yusuf S. Interventions on prevention of postoperative atrial fibrillation in patients undergoing heart surgery: a metaanalysis. Circulation 2002;106:75–80. 2. Crystal E, Garfinkle MS, Connolly S, Ginger T, Sleik K, Yusuf S. Interventions for preventing post-operative atrial fibrillation in patients undergoing heart surgery. Cochrane Database Syst Rev 2009;4:CD003611-16. 3. DiDomenico RJ, Massad MG. Pharmacologic strategies for prevention of atrial fibrillation after open heart surgery. Ann Thorac Surg 2005;79:728–740. 4. Onalan O, Lashevsky I, Crystal E. Prophylaxis and management of postoperative atrial fibrillation. Curr Cardiol Rep 2005;7:382–390. 5. Crystal E, Healey J, Connolly SJ. Atrial fibrillation after cardiac surgery: update on the evidence on the available prophylactic interventions. Card Electrophysiol Rev 2003;7:189–192. 6. Mitchell LB, Exner DV, Wyse DG, Connolly CJ, Prystai GD, Bayes AJ, Kidd WT, Kieser T, Burgess JJ, Ferland A, MacAdams CL, Maitland A. Prophylactic Oral Amiodarone for the Prevention of Arrhythmias that Begin Early After Revascularization, Valve Replacement, or Repair: PAPABEAR: a randomized controlled trial. JAMA 2005;294:3093–3100. 7. Echahidi N, Pibarot P, O’Hara G, Mathieu P. Mechanisms, prevention, and treatment of atrial fibrillation after cardiac surgery. J Am Coll Cardiol 2008;51: 793–801. 8. Sleilaty G, Madi-Jebara S, Yazigi A, Haddad F, Hayeck G, El Rassi I, Ashoush R, Jebara V. Postoperative oral amiodarone versus oral bisoprolol as prophylaxis
16.
17.
18.
19.
20.
21.
against atrial fibrillation after coronary artery bypass graft surgery: a prospective randomized trial. Int J Cardiol 2009;137:116–122. Connolly SJ, Cybulsky I, Lamy A, Roberts RS, O’Brien B, Carroll S, Crystal E, Thorpe KE, Gent M. Double-blind, placebo-controlled, randomized trial of prophylactic metoprolol for reduction of hospital length of stay after heart surgery: the beta-Blocker Length Of Stay (BLOS) study. Am Heart J 2003;145: 226–232. Ishii Y, Schuessler RB, Gaynor SL, Yamada K, Fu AS, Boineau JP, Damiano RJ. Inflammation of atrium after cardiac surgery is associated with inhomogeneity of atrial conduction and atrial fibrillation. Circulation 2005;111:2881–2888. Pagé P, Vinet A, Yin Y, Cardinal R. The site of origin of atrial fibrillation occurring after cardiac surgery is located in the left atrium. Can J Cardiol 2000;16:226F–227F. Swartz MF, Fink GW, Lutz CJ, Taffet SM, Berenfeld O, Vikstrom KL, Kasprowicz K, Bhatta l, Puskas F, Kalifa J, Jalife J. Left versus right atrial differences in dominant frequency, Kþ channel transcripts, and fibrosis in patients developing atrial fibrillation after cardiac surgery. Heart Rhythm 2009;6: 1415–1422. Vicol C, Eifert S, Kur F, Reichart B. Minimally invasive off-pump pulmonary vein isolation to treat paroxysmal atrial fibrillation. Thorac Cardiovasc Surg 2005;53:176–178. Haissaguerre M, Jais P, Shah DC, Takahashi A, Hocini M, Quiniou G, Garrigue S, Le Mouroux A, Le Metayer P, Clementy J. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 1998;339:659–666. Pappone C, Rosanio S, Oreto G, Tocchi M, Giugliotta F, Vicedomni G, Salvati A, Dicandia C, Mazzone P, Santinelli V, Gulletta S, Chierchia S. Circumferential radiofrequency ablation of pulmonary vein ostia: a new anatomic approach for curing atrial fibrillation. Circulation 2000;102:2619–2628. Pappone C, Augello G, Sala S, et al. A randomized trial of circumferential pulmonary vein ablation versus antiarrhythmic drug therapy in paroxysmal atrial fibrillation: the APAF Study. J Am Coll Cardiol 2006;48: 2340–2347. Piccini JP LR, Kong MH. Pulmonary vein isolation for the maintenance of sinus rhythm in patients with atrial fibrillation: a meta-analysis of randomized, controlled trials. Circ Arrhythm Electrophysiol 2009;2:626–633. Parkash R, Tang AS, Sapp JL. Approach to the catheter ablation technique of paroxysmal and persistent atrial fibrillation: a meta-analysis of the randomized controlled trials. J Cardiovasc Electrophysiol 2011;22:729–738. Patti G, Chello M, Candura D, Pasceri V, D’Ambrosio A, Covino E, Di Sciascio G. Randomized trial of atorvastatin for reduction of postoperative atrial fibrillation in patients undergoing cardiac surgery: results of the ARMYDA-3 (Atorvastatin for Reduction of MYocardial Dysrhythmia After cardiac surgery) study. Circulation 2006;114:1455–1461. Hamid M, Kamal RS, Sami SA, Atiq F, Shafquat A, Naqvi HI, Khan FH. Effect of single dose magnesium on arrhythmias in patients undergoing coronary artery bypass surgery. J Pak Med Assoc 2008;58:22–27. Fan K, Lee KL, Chiu CS, Lee JW, He GW, Cheung D, Sun MP, Lau CP. Effects of biatrial pacing in prevention of postoperative atrial fibrillation after coronary artery bypass surgery. Circulation 2000;102:755–760.
CLINICAL PERSPECTIVES Postoperative atrial fibrillation (POAF) still occurs in 30%–40% of patients undergoing cardiac surgery. Despite being short-lived and self-limiting in most cases, studies have repeatedly demonstrated increased lengths of stay and increased morbidity and mortality. Atrial fibrillation (AF) outside the postoperative period appears to be triggered from the pulmonary veins in almost all circumstances where it has been investigated. Preliminary evidence suggests that this might be true for postoperative AF also. In this randomized controlled trial, bilateral surgical pulmonary vein isolation, using a radiofrequency bipolar clamp catheter, failed to reduce the incidence of POAF, but lengthened stay in the ICU and hospital. Despite preliminary evidence, it does not appear that POAF is pulmonary vein dependent. Pulmonary vein isolation should not be performed to prevent POAF.
