Atrial fibrillation (AF) is the most common arrhythmia seen in clinical practice. It has an estimated prevalence of 1–2% in the general population, and this prevalence increases with age.1 It is a significant contributor to cardiovascular morbidity, including palpitations, syncope, heart failure and stroke, as well as to increased mortality.2–4 Patients with non-valvular AF have an incidence of stroke of 3–5% per year, representing a two- to seven-fold increase in risk compared with people without AF.5–8 This risk may be much greater in patients with abnormal substrates such as rheumatic heart disease, where there is a 17.5-fold increased stroke risk.9 In addition, there is evidence to suggest that patients with AF have more frequent episodes of silent cerebral infarction, and that the resultant neurological deficits are greater. The heightened risk of stroke that results from AF has been attributed to the presence of atrial mechanical dysfunction.10–13 This review will focus on current knowledge of atrial mechanical function associated with AF, its potential mechanisms and their implications for the evolution of techniques to restore and maintain sinus rhythm.
Tachycardia-related ventricular cardiomyopathy results from sustained rates in excess of 100 beats per minute. This is a recognised phenomenon that is defined as an impairment of left ventricular function secondary to chronic tachycardia that is partially or completely reversible after normalisation of the heart rate or the restoration of sinus rhythm. Similarly – and perhaps less recognised – atrial tachycardia can result in an atrial cardiomyopathy that manifests as mechanical remodelling and progressive loss of contractile function.
Atrial mechanical dysfunction as a result of AF could largely be attributed to the structural abnormalities that are known to result from atrial arrhythmia. However, while these processes take weeks to develop, atrial mechanical dysfunction is observed with even short durations of atrial arrhythmia, suggesting a role for more functional cellular mechanisms in its development. While the pathophysiological link between increased rates and cardiomyopathy is not well understood, several recent studies argue in favour of an imbalance in cellular calcium handling. Critical to myocardial contraction is the intracellular handling and regulation of calcium. Schotten et al. described downregulation and altered function of the L-type Ca2+ channel and an increased Ca2+ efflux via the Na+/Ca2+ exchanger.14 Thus, over time there is a relative cellular depletion of calcium that results in depressed contractility in AF, which is further aggravated by the reduced rate of sinus rhythm. Schotten et al. showed that β-adrenoceptor density remained unchanged in atrial cardiomyopathy.14
Evaluation of Atrial Mechanical Function
Left atrial (LA) function is classically described in three phases of the cardiac cycle.15 First, the LA receives blood from the pulmonary venous return and acts as a reservoir that stores pulmonary venous return during LV contraction and isovolumic relaxation after the closure and before the opening of the mitral valve. Second, during the early phase of ventricular diastole the LA is a conduit into the left ventricle through which blood flows passively from the pulmonary veins. Finally, the LA contracts to actively empty and serves to augment the left ventricular stroke volume by approximately 20%.16 The relative contribution of each phase to left ventricular filling is dependent on the left ventricular diastolic properties.17,18
The LA appendage has been implicated as the atrial site most likely to develop thrombus. Trans-oesopohgeal echocardiographic assessment with a multiplane probe can visualise the LA appendage in a horizontal short-axis view at the base of the heart and provide 2-chamber longitudinal views of the left chambers. In investigating patients with recent embolic events, LA appendage thrombus has been found in 15% of those with AF lasting for 48 hours and in 27% of those with AF lasting for more than three days.19 LA appendage function has been characterised using pulsed Doppler to determine the LA appendage emptying velocity (LAAEV). The LAAEV and LA spontaneous echocardiographic contrast (LASEC) are the best characterised measures of atrial mechanical function that have been correlated with an increased risk of thrombus and stroke.10 In addition, it is well established that LAAEV <20cm/s or <25cm/s is strongly correlated with the presence of LASEC and thrombus, respectively.20,21
Further to these well-studied markers, others have variably used pulmonary vein blood flow velocity, the peak mitral inflow A wave and its velocity time integral as surrogate markers of atrial function.22–24 Recently, tissue Doppler imaging and LA ejection fraction using computed tomography (CT) or magnetic resonance imaging (MRI) have been reported in studies evaluating LA contractile function; however, the reproducibility of these techniques for evaluating atrial mechanical function remains questionable.25–27
Atrial Mechanical Function and Arrhythmias
Clinically, atrial mechanical dysfunction has been best studied to characterise ‘atrial stunning’ associated with cardioversion of atrial arrhythmia. Atrial stunning is associated with a reduction in LAAEV and an increase in LASEC, and is implicated in the heightened risk of stroke following cardioversion.10–12 At least initially, it was thought to be due to the electrical cardioversion itself. However, there is accumulating evidence suggesting that this process is related to the arrhythmia and not to the mode of cardioversion, with the demonstration of stunning with termination of atrial arrhythmia spontaneously, pharmacologically or after the ablation of chronic atrial flutter.10,11,28 In addition, electrical cardioversion during sinus rhythm or for ventricular fibrillation has not been observed to result in impairment of atrial mechanical function.29,30 Furthermore, immediate reversal of atrial mechanical stunning has been observed by pacing the atria at faster rates or using isoproterenol or intravenous calcium, suggesting that this process is functional in nature.31
Recent studies have expanded our understanding of atrial stunning by demonstrating that the severity of the mechanical dysfunction is related to the duration of arrhythmia. Sanders et al. demonstrated that after cardioversion of AF of short duration (six months) atrial mechanical dysfunction could be reversed, but a severely attenuated response was observed after cardioversion of AF of long duration (three years).32 While these findings may reflect a greater degree of calcium loss and therefore contractile dysfunction, it is possible that with longer durations of arrhythmia the structural abnormalities that are known to occur may have a greater contribution to determining atrial mechanical function. The absence of reversal of atrial mechanical function after cardioversion of long-duration arrhythmia has potentially significant implications for the risk of stroke in patients undergoing ablation of permanent AF.
Atrial Mechanical Function After Ablation of Atrial Fibrillation
Beukema et al. have shown that after successful radiofrequency ablation of patients with highly symptomatic AF there was a significant reduction in LA size as determined by echocardiography.33 However, while the reports on the reduction in atrial size have been consistent, reports on atrial function after ablation have been varied. Thomas et al. studied the recovery of atrial mechanical function in patients undergoing radiofrequency linear ablation during surgery compared with those undergoing cardioversion.27 LA function was evaluated by determining the LA volume and LA ejection fraction.
LA function was found to be reduced after both ablation and cardioversion compared with controls. However, the LA ejection fraction was lowest in the ablation group (15.8±10%), intermediate in the cardioversion group (26±10%) and highest in the normal controls (33±7%). Although the patients undergoing ablation had longer duration of AF (51±9 versus 7±2 months; p=0.0001), the investigators raised the concern that the worse atrial mechanical function in the ablation group may be an additional effect of ablation, over and above that caused by chronic AF alone. However, one should expect that a number of factors may have contributed to the impaired atrial function after surgery: for example, trauma from the surgery, including incisions and ischaemia, may have independent deleterious effects on the LA.
Lemola et al. examined the effects of LA circumferential ablation on LA function in 36 consecutive patients with paroxysmal AF using gated multiphase dynamic contrast-enhanced CT scans of the chest.34 CT scans were processed using volume analysis software to calculate LA volumes. LA circumferential ablation resulted in an approximately 30% decrease in LA ejection fraction in patients in sinus rhythm undergoing ablation of paroxysmal AF, suggesting a potential deleterious effect of catheter ablation on LA function.34 In contrast, Verma at al. reported on patients undergoing ablation of paroxysmal or persistent AF who were followed using echocardiography as well as CT, and concluded that LA volumes declined and LA ejection fraction improved post-ablation.35 Unlike in the study by Lemola et al., these investigators observed deterioration in LA ejection fraction in only a single patient undergoing catheter ablation of paroxysmal AF (1%). While further studies are awaited, this discrepancy may be explained by the variability in the parameter measured in these studies. LA ejection fraction as determined by echocardiography or CT has several limitations, including the angulations of the image during acquisition and the manual tracing of the LA endocardium. In addition, this parameter has not previously been demonstrated to be associated with an increased stroke risk.
Sanders et al. have recently presented data on atrial mechanical function using the well-established markers of LAAES and LASEC by performing serial trans-oesophageal or intra-cardiac echocardiography.36 In this study, in patients with paroxysmal AF undergoing ablation during sinus rhythm there was no change in atrial function, indicating that ablation itself is not associated with impairment of function. In contrast, patients with persistent or permanent AF undergoing ablation of AF during AF there was significant atrial stunning on termination of AF. At follow-up evaluation six months after the ablation procedure, there had been a significant recovery of atrial mechanical function. However, the LAAEV did not reach the same magnitude as in those patients undergoing ablation in sinus rhythm.
Reduction of Stroke Risk after Ablation of Atrial Fibrillation
These data on the recovery of atrial mechanical function after ablation to restore sinus rhythm provide the impetus to suggest that such therapy may reduce the risk of stroke associated with this condition. In agreement with this, Oral et al. recently presented observational data in their series of 755 consecutive patients undergoing AF ablation.37 Of this group, 411 (56%) had at least one risk factor for stroke, with 34 having had a previous stroke. Anticoagulation was ceased in 79% of 256 patients in sinus rhythm with no stroke risk factors and in 68% of 266 patients in sinus rhythm with at least one stroke risk factor. During follow up, seven of 755 patients (0.9%) had an embolic event within two weeks of ablation and a further two (0.3%) had late embolic events (180 and 300 days post-procedure). In the same cohort, two patients (0.3%) had cerebral haemorrhage while anticoagulated.
Abnormal atrial mechanical function as assessed by trans-oesophageal echocardiography and newer emerging modalities such as tissue Doppler imaging and contrast echocardiography are characterised by a reduction in LAAEV, an increase in LASEC and the formation of thrombus. Atrial mechanical function is determined by the duration of sustained arrhythmia. Ablation does not significantly affect atrial mechanical function, but restoration of sinus rhythm by ablation is associated with an improvement in atrial mechanical function. Further prospective studies are required to determine the effectiveness of catheter ablation of AF in stroke prevention.
Acknowledgement of Support
Dr Dimitri is supported by a Post-graduate Scholarship from the Cardiac Society of Australia and New Zealand.