Ann Thorac Surg 2006;81:1676-1682
© 2006 The Society of Thoracic Surgeons
Original article: Cardiovascular
Coronary Surgery in Patients With Preexisting Chronic Atrial Fibrillation: Early and Midterm Clinical Outcome
Chris A. Rogers, PhD,
Gianni D. Angelini, MD, MCh,
Lucy A. Culliford, PhD,
Radek Capoun, MRCS,
Raimondo Ascione, MD, MCh
*
Bristol Heart Institute, University of Bristol, Bristol, United Kingdom
Accepted for publication November 28, 2005.
* Address correspondence to Dr Ascione, Bristol Heart Institute, University of Bristol, Bristol Royal Infirmary, Bristol BS2 8HW, United Kingdom; (Email: r.ascione{at}bristol.ac.uk).
 |
Abstract
|
|---|
BACKGROUND: The purpose of this study was to investigate the effect of preexisting atrial fibrillation on early and midterm clinical outcome in patients undergoing coronary surgery.
METHODS: All elective patients undergoing coronary artery bypass grafting surgery between April 1996 and September 2002 were investigated. Patients were grouped according to their preoperative cardiac rhythm: sinus rhythm (SR) or preexisting atrial fibrillation (AF). In-hospital clinical outcomes and 5-year patient survival and cardiac-related event-free survival were compared using regression methods to adjust for differences between the groups. In all, 5,092 patients were identified, 175 (3.4%) with a history of preexisting AF. These patients were older (median, 64 versus 68 years) and had higher Parsonnet scores (median, 4 versus 8) than the SR group. Previous myocardial infarction, cerebrovascular accident, hypertension, diabetes mellitus, renal impairment, peripheral vascular disease, ejection fraction less than 50%, previous surgery, congestive heart failure, and use of angiotensin-converting enzyme inhibitors were also more common in the AF group.
RESULTS: There were 60 in-hospital deaths (1.2%), with no difference between the two groups (odds ratio 1.02, 95% CI: 0.35 to 2.94). Atrial fibrillation patients were more likely to need intraoperative inotropes (p = 0.044), postoperative intra-aortic balloon pump (p = 0.038), and were less likely to be discharged within 6 days (p = 0.017). The risk of death in the 5 years after surgery was higher in the AF group (relative risk 1.49, 95% CI: 1.06 to 2.08, p = 0.020). In the AF group, 109 (62.2%) patients were cardioverted spontaneously by surgery, but only 69 (39.4%) remained in SR until discharge. Longer-term rhythm follow-up data were available for 48 of these 69 patients, and only 36 remained in SR at a median follow-up of 1,483 days (interquartile range, 1,120 to 2,209). Spontaneous conversion to SR after surgery did not confer a midterm survival benefit (p = 0.91).
CONCLUSIONS: Preexisting AF in patients undergoing coronary artery bypass graft surgery is not associated with increased in-hospital mortality and major morbidity; however, it is a risk factor for reduced 5-year survival. Spontaneous cardioversion to SR during surgery is transient in the majority of patients and is not associated with midterm survival benefit.
|
Introduction
|
|---|
The loss of coordinated atrial contraction during atrial fibrillation (AF) creates a predisposition for thrombus formation contributing to the risk of embolic stroke [1, 2]. Long-lasting AF is generally accepted to be a cause of tachycardia-induced cardiomyopathy [3] and has been associated with a twofold increase in risk of premature death [1, 4]. A substantial number of patients undergoing coronary artery bypass graft surgery (CABG) also present with a history of preexisting chronic AF, and very little is known about their outcome after surgery. The aim of this study was to ascertain the incidence and effect of preexisting AF on early in-hospital outcomes, midterm survival, and cardiac-related event-free survival after isolated CABG when compared with patients in sinus rhythm (SR). We also investigated the likelihood of these patients of being spontaneously converted to SR during surgery and the related outcome.
|
Patients and Methods
|
|---|
Patient Selection and Data Collection
A consecutive series of patients undergoing isolated CABG, operated on between April 1996 and September 2002, were included in this retrospective analysis. Emergency and salvage procedures were excluded. Patients were grouped according to their preoperative cardiac rhythm: sinus rhythm (SR) or preexisting atrial fibrillation (AF). Atrial fibrillation patients were defined as those suffering from either permanent or multiple episodes of recurrent AF during the 6 months before admission. Patients with a complete heart block or with history of ventricular fibrillation/ventricular tachycardia were excluded. Surgical procedures were carried out either with cardiopulmonary bypass (on-pump) or without it (off-pump), according to the preference and expertise of the surgeon carrying out the operation. Baseline demographic and clinical data and perioperative clinical outcome data were collected prospectively at the time of operation and entered into the Patient Activity Tracking System database (Dendrite Systems, London, United Kingdom). Postdischarge deaths were identified from mortality data obtained from the National Health Service Strategic Tracing Service (NSTS) national database. Annual follow-up was by postal questionnaire to the patient and to the patient's family physician.
Management of Preoperative and Postoperative Medications
Beta-blockers, diuretics, antihypertensives, calcium-channel blockers, and antiarrhythmic drugs were routinely omitted on the day of surgery, and restarted during the postoperative in-hospital recovery period before discharge. Aspirin and warfarin were omitted 2 to 3 days before surgery. Angiotensin-converting enzyme inhibitors were withdrawn on the evening before the operation. To prevent thromboembolic events, and in absence of significant bleeding, 300 mg aspirin was administered 4 hours after surgery, and then daily thereafter. Patients with a history of preexisting AF were restarted on warfarin on the first postoperative day, with the aim of achieving an international normalized ratio (INR) between 2.0 and 3.0 before discharge. Atrial fibrillation patients spontaneously cardioverted to SR at surgery were started on amiodarone as per standard protocol.
Surgical Technique and Postoperative Management
Anesthetic and surgical techniques were standardized for all patients and have been reported previously [5–7]. During surgery, no attempt was made at direct current cardioversion, and there was no additional intervention to prevent postoperative AF, such as pulmonary vein isolation. At the end of surgery, patients were transferred to the intensive care unit and managed according to the unit protocol [5–7]. During the first 72 hours after surgery, the heart rate and rhythm were continuously monitored and displayed on a screen with an automated arrhythmia detector (Solar 8000 Patient Monitor; Marquette Medic Systems, Milwaukee, Wisconsin). Twelve-lead electrocardiographic recordings were performed preoperatively, 2 hours postoperatively, and then daily thereafter until discharge.
Clinical Data Collection and Definitions
In-hospital death was defined as any death occurring during the same hospital admission for surgery. Perioperative myocardial infarction (MI), ST-segment changes, pacing, arrhythmias, and inotropic requirement were recorded and defined as previously reported [5]. An independent intensivist was responsible for interpreting episodes of arrhythmia after surgery and for deciding if either temporary or permanent pacing was needed. Postoperative blood loss was defined as total chest tube drainage [8]. Stroke included permanent and transient stroke [9]. The clinical diagnosis of stroke was confirmed by a neurologist on the basis of the clinical findings.
Nonfatal cardiac related events included the need for a further coronary revascularization procedure (whether reoperation or percutaneous intervention), patient reported hospital attendance for MI or coronary angiography, and report of cardiac-related hospital admission by the family physician.
Statistical Analysis
The statistical analyses comparing outcomes for patients with and without preexisting AF were adjusted for baseline and operative factors known to influence outcome or which were imbalanced between the two groups. (Potential confounding variables included in the adjusted analyses are detailed in the footnote to Table 3.)
View this table:
[in this window]
[in a new window]
|
Table 3. Effect Sizes for Intraoperative and Postoperative Clinical Outcome (Atrial Fibrillation Versus Sinus Rhythm)
|
|
Baseline and operative characteristics were compared using the 
2 or Fisher's exact test (categorical variables) or the Wilcoxon rank sum test (continuous or discretely measured variables). The effect of preexisting AF on in-hospital outcomes and 5-year survival was quantified using standard regression methods: multiple linear regression for continuous variables, logistic regression for binary variables, and Cox proportional hazards regression for time to event variables. Analyses were carried out using STATA, version 8.2 (Stata Corporation, College Station, Texas).
Blood loss was transformed into natural logarithms to normalize the distribution. For the analysis of postoperative hospital stay, patients who died before discharge were treated as censored observations. Most, 99.6%, of the study cohort was successfully matched to the NSTS database. For the few patients who could not be traced through NSTS, the survival time was censored at hospital discharge. Patients free from cardiac related events were censored at last follow-up.
Standard errors were calculated (a) assuming independence and (b) taking account of the clustering of patients within surgeons, and the analysis resulting in the largest standard errors were chosen in each case.
Our interpretation of the findings is based on the consistency of the findings and their magnitude as well as their statistical significance. Results are reported with 95% confidence intervals and associated p values quoted relate to the adjusted analyses.
A post-hoc analysis of patient survival after hospital discharge was carried out to ascertain whether being discharged home in SR after spontaneous surgery-related cardioversion determined any midterm survival benefit in patients with preexisting AF.
|
Results
|
|---|
Incidence and Baseline Distribution of Variables
From the database of 7,340 cardiac procedures carried out in the study period, 5,092 eligible procedures were identified; 2,296 operations did not satisfy our inclusion criteria, and 42 were omitted because data were missing. Of the 5,092 patients undergoing CABG, 175 (3.4%, 95% CI: 2.9% to 4.0%) had history of preexisting AF.
The distributions of a range of demographic, comorbid, and cardiac factors are shown in Table 1. Compared with the SR group, the patients with preexisting AF were older, had higher Parsonnet scores, and were more likely to suffer from severe dyspnea, hypertension, diabetes mellitus, renal impairment, peripheral vascular disease, cerebrovascular disease, congestive heart failure, have an ejection fraction less than 50%, a history of previous MI, be taking angiotensin-converting enzyme inhibitors, and have had previous surgery. Seven patients had an intra-aortic balloon pump inserted before surgery (2 in the AF group). Fourteen patients had chronic renal failure (13 in SR group and 1 in the AF group).
In-Hospital Outcomes
Operative and postoperative clinical outcomes are shown in Table 2. Similar numbers were operated off-pump in the two groups (p = 0.59). The number of distal anastomoses was similar in the two groups (p = 0.86), and there was no difference in the distribution of target coronary vessels (p = 0.90). Coronary revascularization determined spontaneous intraoperative conversion to SR in 109 patients (62.6%) in the AF group. Two hundred fifty-eight patients (5.3%) in the SR cohort and 37 (21.3%) in the AF group required temporary intraoperative pacing (p < 0.001).
The overall in-hospital mortality rate was 1.2% (95% CI: 0.9% to 1.5%), with no difference between the two groups (p = 0.96, after adjustment for differences in preoperative risk factors and operative variables).
Sixty-nine of the 109 patients (63.3%) in the AF group spontaneously converted to SR by surgery remained in SR to discharge, whereas the remaining 40 returned to AF. Three patients (1.7%) with preexisting AF suffered ventricular tachycardia or ventricular fibrillation, or both, after surgery, as compared with 67 patients (1.4%) in the control group.
The quantified effect of preexisting AF for each outcome is given in Table 3. For the majority outcomes risk adjustment reduced the magnitude of the effect of AF compared with the unadjusted analysis; the differences were smaller (namely, the effect sizes were closer to 1), but some significant differences remained. More AF patients required intraoperative inotropic support (p = 0.044) and pacing at the end of surgery (p < 0.001). Atrial fibrillation patients were also more likely to require an intra-aortic balloon pump after surgery (p = 0.038). The "relative risk" for hospital discharge was not constant over time and so is reported separately for two epochs, 0 to 6 days and 7 or more days. The split at 6 days was chosen as approximately 50% of patients were discharged within 6 days of surgery. On average, AF patients also stayed in hospital longer; they were less likely to be discharged within 6 days (p = 0.017), but for those who stayed beyond 6 days, there was no difference between the two groups (p = 0.51). No differences were found in the incidence of stroke, renal dialysis, chest infection, perioperative MI, blood loss, transfusion requirement, or the need for readmission to the intensive care unit.
Midterm Outcome
In total, 488 deaths were reported within 5 years of surgery. Of the 428 deaths that occurred after hospital discharge, 388 were in the SR group and 40 in the AF group. The median follow-up for survivors was 5 years in both the AF and SR groups. Percentage survival at 30 days, and 1, 3, and 5 years by group is shown in Table 2. At 5 years, the survival rate for the cohort as a whole was 89% (95% CI: 88% to 90%). However, many fewer patients in the AF group survived to 5 years (p < 0.0001, log rank test; Fig 1). After adjustment for differences between the groups, the overall risk of death was estimated to be 49% higher in the AF group (p = 0.020) when compared with the SR cohort.
View larger version (15K):
[in this window]
[in a new window]
|
Fig 1. Kaplan-Meier estimates of patient survival after coronary surgery. (AF = atrial fibrillation; SR = sinus rhythm.)
|
|
Follow-up data were available for 3,782 patients, which, excluding those who died within a year of surgery or before our follow-up program commenced, represents 81% of the eligible cohort. Cardiac-related event-free survival to 5 years is illustrated in Figure 2; after 5 years, significantly fewer patients in the AF group were free of a cardiac-related event compared with the SR group (57.6%, 95% CI: 48.1% to 65.9%, versus 79.9%, 95% CI: 78.4% to 81.2%, p < 0.0001, log rank test). After adjustment for differences between the groups, the risk of a cardiac-related event for patients with preexisting AF was 43% higher when compared with the SR cohort (p = 0.014).
View larger version (14K):
[in this window]
[in a new window]
|
Fig 2. Kaplan-Meier estimates of cardiac event-free survival after coronary surgery. (AF = atrial fibrillation; SR = sinus rhythm.)
|
|
Post-Hoc Analysis
Sixty-nine patients from the AF group were discharged home in SR after spontaneous surgery-related cardioversion. Of these, 44 (64%) were still in SR at 6 to 8 weeks of follow-up. Longer-term rhythm follow-up data were available for 48 of these 69 patients; and at a median follow-up of 1,483 days (interquartile range, 1,120 to 2,209), 36 were still in SR. Of the remaining 106 who did not spontaneously convert at surgery, 4 died in hospital. For AF patients, conversion to SR did not confer any significant long-term survival benefit (risk-adjusted hazard ratio for death after discharge 0.96, 95% CI: 0.49 to 1.89, p = 0.91).
|
Comment
|
|---|
The worse risk profile of our AF group is in keeping with what reported in the literature. The relationship between age, sex, and AF was recently demonstrated in a cross-sectional study of 1.89 million subjects in a health maintenance organization in northern California [10]. Our study also suggests a relatively low incidence of in-hospital morbidity and mortality in patients with preexisting AF, lower than that reported in a similar study from the Cleveland Clinic [11]. One possible explanation for this difference might be that the Cleveland Clinic study had a larger population with more statistical power (451 AF patients, accrued over a 22-year period). However, it could also be related to advances in anesthetic, perfusion, and surgical techniques, implemented over the last decade, including the use of beating-heart coronary surgery (performed in 29% of our AF group). Patient survival in our study was similar to that reported by the Cleveland Clinic group, with a significantly increased risk of 5-year mortality for patients with preexisting AF [11]. One possible explanation is the reported relationship between AF and congestive heart failure in the sense that each can result in the other [12, 13]. Long-term AF may also cause tachycardia-induced cardiomyopathy [3].
In our study, surgery seems to trigger spontaneous cardioversion to SR in a significant number of patients. This benefit, however, is transient, and does not seem to confer any midterm survival benefit. On the contrary, the Cox-maze procedure has produced durable long-term outcomes for the surgical treatment of AF [14, 15]. Furthermore, in recent years, intraoperative ablation using radiofrequency energy has been developed to simplify the surgical treatment of AF [16–20], and subsequent modifications have been applied to reduce the number of lesion sets required [15, 21]. The efficacy of these techniques is based on the achievement of a scar/healing process at the ablation site, which has been demonstrated to be necessary to obtain a lasting effect [16–20]. However, most studies of combined AF ablation and cardiac surgery reported in the literature involve valvular patients [22]. Theoretically, similar results might be obtained in CABG patients; however, no recommendation can be made from our study as the majority of patients were treated before the introduction of modern techniques of AF ablation. Furthermore, the current evidence in the literature supporting the use of these techniques in CABG patients is rather limited and inconclusive.
This study is not without its limitations. First, there were important preoperative imbalances between groups. Regression modelling can never entirely account for all imbalances, and the adjusted estimates may still be subject to some residual confounding. Nevertheless, such confounding is unlikely to change the direction of the differences seen. Furthermore, without blinding, the assessment of some outcomes could have been biased by the knowledge of the patients' clinical history. For example, as a precaution, patients with a history of AF may have been kept in hospital longer. However, in our institution strict local guidelines are used to make decisions about perioperative patient care, and these guidelines were applied throughout the study period.
A further limitation is the lack of data on the efficacy of warfarin-related long-term anticoagulation needed for patients suffering long-lasting AF, and on the occurrence of late hemorrhagic or thromboembolic events. These events might have contributed to the reduced midterm survival outcome in the AF group, rather than the presence of AF itself. However, management with warfarin, and its potential detrimental effect, is a specific AF-related issue.
The similarity between our results for patient survival and cardiac event-free survival are not surprising given that death was the first reported event for the majority of patients. As clinical follow-up was not sought before 2002, nonfatal cardiac events are necessarily underreported, and hence event-free survival is overestimated. However, we do not expect this bias to strongly influence the comparison between the SR and AF groups as the introduction of the follow-up program was the same for both groups.
Finally, our definition of preexisting AF does not match exactly the American College of Cardiology/American Hospital Association/European Society of Cardiology guidelines [23] since the majority of patients in our study were treated before their publication in 2001.
In conclusion, our study suggests that CABG patients with a history of preexisting AF have a much worse risk profile than those in SR. Preexisting AF does not seem to be associated with increased in-hospital mortality and major morbidity. Surgery-related spontaneous cardioversion to SR appear to be transient in the majority of cases, and does not seem to confer any survival benefit.
|
Acknowledgments
|
|---|
This study was supported by the Garfield Weston Trust and the British Heart Foundation.
|
References
|
|---|
- Hart RG, Sherman DG, Easton JD, et al. Prevention of stroke in patients with nonvalvular atrial fibrillation Neurology 1998;51:674-681.[Abstract/Free Full Text]
- Benjamin E, Wolf P, D'Agostino R, et al. Impact of atrial fibrillation on the risk of death. The Framingham Heart Study Circulation 1998;98:946-952.[Abstract/Free Full Text]
- Chen MS, Marrouche NF, Khaykin Y, et al. Pulmonary vein isolation for the treatment of atrial fibrillation in patients with impaired systolic function J Am Coll Cardiol 2004;43:1004-1009.[Abstract/Free Full Text]
- Wolf P, Benjamin E, Belanger A, et al. Secular trends in the prevalence of atrial fibrillation. The Framingham Heart Study Am Heart J 1996;131:790-795.[Medline]
- Ascione R, Caputo M, Calori G, et al. Predictors of atrial fibrillation after conventional and beating heart coronary surgerya prospective randomised study. Circulation 2000;102:1530-1535.[Abstract/Free Full Text]
- Watters M, Ascione R, Ryder I, et al. Hemodynamic changes during beating heart coronary surgery with the "Bristol Technique" Eur J Cardiothorac Surg 2001;19:34-40.[Abstract/Free Full Text]
- Angelini G, Taylor F, Reeves B, Ascione R. Early and midterm outcome after off-pump and on-pump surgery in Beating Heart Against Cardioplegic Arrest Studies (BHACAS 1 and 2)a pooled analysis of two randomised controlled trials. Lancet 2002;359:1194-1199.[Medline]
- Ascione R, Williams S, Lloyd C, et al. Reduced postoperative blood loss and transfusion requirement after beating heart surgerya prospective randomised study. J Thorac Cardiovasc Surg 2001;121:689-696.[Abstract/Free Full Text]
- Lloyd C, Ascione R, Underwood M, et al. Serum S-100 protein release and neuropsychological outcome during coronary revascularisation on the beating hearta prospective randomised study. J Thorac Cardiovasc Surg 2000;119:148-154.[Abstract/Free Full Text]
- Go A, Hylek EM, Philips KA, et al. Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention. The AnTicoagulation and Risk Factors In Atrial fibrillation study (ATRIA) JAMA 2001;285:2370-2375.[Abstract/Free Full Text]
- Quader M, McCarthy P, Gillinov M, et al. Does preoperative atrial fibrillation reduce survival after coronary artery bypass grafting? Ann Thorac Surg 2004;77:1514-1524.[Abstract/Free Full Text]
- Van den Berg MP, Tuinenburg AE, Crijns HJ, et al. Heart failure and atrial fibrillationcurrent concepts and controversies. Heart 1997;77:309-313.[Abstract/Free Full Text]
- Stevenson WG, Stevenson LW. Atrial fibrillation in heart failure N Engl J Med 1999;341:910-911.[Free Full Text]
- Cox JL, Schuessler RB, D'Agostino Jr HJ. The surgical treatment of atrial fibrillation III. Development of a definitive surgical procedure J Thorac Cardiovasc Surg 1991;101:569-583.[Abstract]
- Prasad SM, Maniar HS, Camillo CJ, et al. The Cox maze III procedure for atrial fibrillation. Long-term efficacy in patients undergoing lone versus concomitant procedures J Thorac Cardiovasc Surg 2003;126:1822-1828.[Abstract/Free Full Text]
- Chiappini B, Martin-Suarez S, LoForte A, Di Bartolomeo R, Marinelli G. Surgery for atrial fibrillation using radiofrequency catheter ablation J Thorac Cardiovasc Surg 2003;126:1788-1791.[Abstract/Free Full Text]
- Sie HT, Beukema WP, Elvan A, Ramdat Misier AR. Long-term results of irrigated radiofrequency modified maze procedure in 200 patients with concomitant cardiac surgerysix years experience. Ann Thorac Surg 2004;77:512-517.[Abstract/Free Full Text]
- Khargi K, Deneke T, Haardt H, et al. Saline-irrigated, cooled-tip radiofrequency ablation is an effective technique to perform the maze procedure Ann Thorac Surg 2001;72(Suppl):1090-1095.
- Guden M, Akpinar B, Sanisoglu I, Sagbas E, Bayindir O. Intraoperative saline-irrigated radiofrequency modified maze procedure for atrial fibrillation Ann Thorac Surg 2002;74(Suppl):1301-1306.
- Chiappini B, Martin-Suarez S, LoForte A, et al. Cox/Maze III operation versus radiofrequency ablation for the surgical treatment of atrial fibrillationa comparative study. Ann Thorac Surg 2004;77:87-92.[Abstract/Free Full Text]
- Cox JL. Atrial fibrillation II rationale for surgical treatment J Thorac Cardiovasc Surg 2003;126:1693-1699.[Free Full Text]
- Halkos ME, Craver JM, Thourani VH, et al. Intraoperative radiofrequency ablation for the treatment of atrial fibrillation during concomitant cardiac surgery Ann Thorac Surg 2005;80:210-216.[Abstract/Free Full Text]
- Task Force Report. ACC/AHA/ESC guidelines for the management of patients with atrial fibrillation Eur Heart J 2001;22:1852-1923.[Free Full Text]
Top
Abstract
Introduction
