Current Vascular Pharmacology, 2010, 8, 553-572
553
Postoperative Atrial Fibrillation - What Do We Really Know? Maciej Banach1,*, Antonios Kourliouros2, Kurt M. Reinhart3, Stefano Benussi4, Dimitri P. Mikhailidis5, Marjan Jahangiri2, William L. Baker3, Andrea Galanti4, Jacek Rysz1, John A. Camm2, C. Michael White3 and Ottavio Alfieri4 1
Department of Hypertension, Medical University of Lodz, Poland; 2Departments of Cardiac Surgery and Cardiac & Vascular Sciences St George’s, University of London, UK; 3School of Pharmacy, University of Connecticut, Storrs CT and Divisions of Cardiology and Drug Information, Hartford Hospital, Hartford, CT, USA; 4Division of Cardiac Surgery, S. Raffaele University Hospital, Milan, Italy; 5Department of Clinical Biochemistry, Royal Free Campus, University College London Medical School, University College London (UCL), London, UK Abstract: Postoperative atrial fibrillation (POAF) is a common complication following cardiac surgery, occurring in 20% to 60% of patients. Advanced age, history of atrial fibrillation (AF), heart failure, peripheral arterial disease and chronic obstructive pulmonary disease are predictors of POAF. The pathogenesis of AF seems to be multifactorial, and includes electrical and structural remodeling as well as inflammation (a systemic response caused by cardiopulmonary bypass and cardiotomy). Numerous pharmacologic agents can decrease the incidence of POAF. It is also necessary to evaluate an agent’s ability to decrease stroke, mortality, length of stay and hospital costs. Currently, the use of beta-blockers with adjunctive amiodarone has been shown to reduce POAF and several of its complications. Two therapeutic choices exist in patients with POAF: rate control and rhythm control. The decision which is more important to target should be based on the symptoms of the individual patient. Unlike in patients with chronic AF, POAF is generally transient, and the risks of anticoagulation may outweigh the benefits. Surgical ablation techniques and ablation devices have progressed considerably. This made the procedures quicker and simpler, and therefore feasible in virtually all clinical contexts. In turn, this has raised the issue of post-ablation arrhythmias. Although relapsing AF is generally addressed conservatively, it may require ablation, frequently transseptal. Further research is needed to identify the predictors of POAF and the most effective pharmacological and invasive methods for the prevention and treatment of POAF.
Keywords: Postoperative atrial fibrillation, predictors, pathogenesis, pharmacotherapy, ablation. SEARCH STRATEGY We searched using electronic databases [MEDLINE (1966 - February 2009), EMBASE and SCOPUS (1965 – February 2009), DARE]. Additionally, abstracts from national and international cardiovascular meetings were studied. Where necessary, the relevant authors of these studies were contacted to obtain further data. The main data search terms were: postoperative atrial fibrillation, predictors, pathogenesis, treatment and ablation. EPIDEMIOLOGY AND RISK STRATIFICATION OF POAF Atrial fibrillation (AF) has been described as an ‘epidemic’ due to its increasing prevalence in our ageing population. The prevalence of active AF in the UK has risen from 0.78% in 1994 to 1.31% in 2003 [1]. In another epidemiological study, the incidence of age- and gender-adjusted AF *Address correspondence to this author at the Department of Hypertension, Medical University of Lodz, Zeromskiego 113, 90-549 Lodz, Poland; Tel: +48 42 639 37 71; Fax: +48 42 639 37 50; E-mail:
[email protected]
1570-1611/10 $55.00+.00
rose from 3.04 to 3.68 per 1000 person-years between 1980 and 2000 [2]. A prevalence model structured by the same research group estimated a 3-fold increase in the number of patients with AF in the next 50 years. The increasing prevalence of diabetes, obesity and metabolic syndrome may also play a role in the anticipated AF epidemic [3-7]. Postoperative atrial fibrillation (POAF) is a common postoperative complication that occurs in 20–60% of patients undergoing cardiac surgery [8]. POAF is generally paroxysmal, usually appears 2 and 4 days after surgery, and often spontaneously converts within 30 days [8, 9]. Suitable treatment and prevention of POAF should result in rapid rehabilitation, decreased postoperative complications and reduced cost of hospitalization [10]. Identifying the perioperative predictors of POAF is important because this may provide effective prevention and treatment. The cost implications are substantial. For example, investigators from the Department of Cardiac Surgery in Lodz (Poland) showed that POAF prolonged intensive care unit (ICU) stay by about 1 day and hospitalization by almost 2 days increasing the annual costs of hospitalization by about $45,000 [11-13]. © 2010 Bentham Science Publishers Ltd.
554 Current Vascular Pharmacology, 2010, Vol. 8, No. 4
POAF significantly influences the prognosis of patients who undergo cardiac surgery. Therefore, it is important to implement pre-, intra- and postoperative measures to prevent POAF [14]. Among the many perioperative risk factors of POAF are advanced age, low left ventricle ejection fraction (LVEF) before the surgery, repeated cardiac surgery, concomitant diseases such as chronic obstructive pulmonary disease, peripheral arterial disease, renal failure, and the need for inotropic drugs in the preoperative period [14-16]. Some factors can reduce the incidence of POAF, including offpump coronary artery bypass (OPCAB) and the administration of beta-blockers before surgery [13]. While the prophylactic benefit of beta-blockers (especially cardioselective and those with vasodilator properties) is strongly supported, there is still heated discussion regarding the influence of OPCAB on POAF [15, 17].
The predictors of POAF were also analyzed by the present authors in patients undergoing different types of cardiac surgery. In one study, we included 300 patients with aortic stenosis or regurgitation. The influence of POAF on the occurrence of postoperative complications and its effect on the prognosis of patients who had had aortic valve replacement (AVR) were investigated [18]. We showed that the following predictors of POAF were significant and independent (aortic stenosis group): a history of heart failure (OR, odds ratio = 5.5), low LVEF pre- (5.1) and postoperatively (4.4), advanced age (4.5), body mass index (BMI) 21 kg/m2 (3.9), maximal transvalvular gradient 85 mmHg (3.7) and endsystolic intraventricular septum thickness 1.8 cm (2.9) before the surgery. In patients with aortic regurgitation, we selected the following risk factors of POAF: left atrium diameter 4.25 cm (4.1), low LVEF (3.7) and coexistent insignificant mitral valve defect (2.5) preoperatively, heart failure (4.5), arterial hypertension (3.3) and diabetes (2.6) in the history and left atrium diameter 4.15 cm in the postoperative period (2.9). Analyzing all patients who had had AVR we observed that the type of aortic defect - aortic regurgitation was also an independent predictor of POAF (1.7), and POAF was an independent risk factor for hospital death, increasing mortality almost 4-fold (OR 3.9) [18]. Few studies describe predictors of POAF in patients who have undergone AVR. In one study [19], the authors identified age, NYHA functional class, history of preoperative AF, left ventricular mass index, >300 supraventricular beats on 24 h ECG before surgery, presence of supraventricular tachycardia (SVT), SVT of >5 beats, or a rate >120 beats/min as significant risk factors for POAF. Using multivariate analysis, they identified the following variables as independent predictors: advanced age, history of paroxysmal AF, presence of >300 supraventricular beats/24 h and presence of SVT [19]. In the study by Ducceschi et al. [20], 302 patients were divided into 2 groups according to the absence or evidence of post-aortic valve replacement AF. POAF occurred in 19% of the patients. AF after AVR was associated with advanced age, left atrial enlargement, preoperative episodes of paroxysmal AF, the use of a warm blood cardioplegic solution and normothermia, administration of inotropic agents, prolonged assisted ventilation, electrolyte imbalance, and atrioventricular and intraventricular conduction disorders. Using multivariate logistic regression analysis, age, left atrial enlargement, a prior history of paroxysmal AF, and
Banach et al.
postoperative electrolyte imbalance were identified as independent correlates of POAF, whereas the use of hypothermia appeared to be a protective factor [20]. We described some other predictors that were not mentioned in previous studies, including low BMI and high transvalvular gradient for patients with aortic stenosis, and a history of insignificant mitral regurgitation, essential atrial hypertension and diabetes mellitus for patients with aortic regurgitation [18]. Low BMI as an independent predictor of postoperative AF in patients with aortic stenosis seems interesting, since this confirms previous reports suggesting that low BMI may worsen the prognosis and increase the risk of cardiovascular events and mortality in patients with heart disease. This was especially the case with left ventricular dysfunction [21, 22], and was in accordance with the observation of the so-called “lipid paradox” and “endotoxinlipoprotein hypothesis”, which explained the seemingly protective role of hypercholesterolemia in cardiac cachexia among patients with heart failure [23, 24]. Besides the increase in postoperative mortality, POAF in patients who have had AVR may also be associated with a higher risk of other life-threatening complications such as stroke, especially in patients with prosthetic valves [18, 25]. Such a relation was observed in our study [18]. Ruel et al. (2004) [25] evaluated the determinants of stroke in patients after AVR. POAF, a history of coronary disease, and tiltingdisc mechanical prostheses were independent predictors of embolic stroke after AVR [25]. Next we evaluated the predictors of POAF in 1200 patients who had had isolated surgical revascularization [coronary artery bypass grafting (CABG) including OPCAB], and the influence of POAF on prognosis in these patients [14]. POAF was observed in 278 patients (23.2%). We selected 5 independent predictors of POAF: history of AF (OR = 6.1), preoperative LVEF
