Atrial fibrillation and heart failure: novel insights into the chicken and egg dilemma

Abstract

Heart failure and atrial fibrillation (AF) are growing cardiovascular disease epidemics worldwide. Heart failure with left-ventricular systolic dysfunction (LVSD) can be found in more than one-third of all patients with AF.1 Conversely, up to half of the subjects with LVSD suffer from AF. Both frequently coexisting comorbidities can cause or exacerbate each other and worsen the prognosis of the patients.1,2 However, in many cases it remains unclear whether LVSD preceded AF or vice versa. Thus, the causal interaction between AF and the left ventricle with potential LVSD remains a—still ambiguous—chicken and egg problem (Figure 1). Atrial fibrillation and heart failure frequently coexist. Both comorbidities can induce each other thereby causing a chicken–egg causality dilemma in these respective patients.1 LV, left ventricular. Recent clinical evidence indicates that AF can induce LVSD, heart failure-related morbidity, and mortality.3 The CASTLE-AF trial demonstrated that catheter ablation of AF can improve mortality, heart failure hospitalization, and LVSD in patients with heart failure with reduced ejection fraction and AF.4 However, it has to be taken into account that included patients also suffered from other causes for LVSD in addition to AF. More specifically, the CAMERA-MRI trial investigated patients with AF and idiopathic cardiomyopathy (LV ejection fraction ≤ 45%), where other identifiable causes of LVSD such as significant coronary artery disease were ruled out. In this trial, rhythm control via catheter ablation improved LV function compared with medical rate control.5 However, previous studies on rhythm control therapy in patients with LVSD also reported contrasting results in particular when using antiarrhythmic drugs for rhythm control.6 While a recent meta-analysis substantiated the favourable effects of rhythm control via catheter ablation on mortality, LV function, and heart failure status in patients with AF and heart failure,6 further studies on rhythm control therapies are needed.3 While AF begets heart failure, also heart failure can cause/precede AF and a significant proportion of patients with different types of heart failure develop AF over time.1 The mechanisms include an increase in left atrial pressure with consecutive atrial enlargement and adverse structural and electrical remodelling. In patients with lone AF also an occult cardiomyopathy may underlie AF induction.7 Thus, identification and classification of patients with AF and heart failure are of utmost importance for appropriate and personalized management and further basic and clinical research is required to shed light into this vicious partnership. Detrimental effects of atrial fibrillation on left-ventricular cardiomyocyte excitation–contraction coupling. ROS, reactive oxygen species; SR, sinus rhythm; CaMKII, Ca2+/calmodulin-dependent protein kinase IIδc. With permission from Pabel et al., Circ Res 2022.9 A critical aspect regarding the beneficial effects of rhythm control therapy is the impact of AF-related loss of sufficient atrial contraction, loss of atrioventricular synchrony along with an atrial cardiomyopathy leading to impaired ventricular filling. Restoring the atrial contribution to ventricular filling improves cardiac haemodynamics, in particular in patients with heart failure. Interesting insights are received from the APAF-CRT mortality trial. Even if AF is accepted to be of permanent nature, ablation of the atrioventricular junction followed by bi-ventricular pacing still significantly reduced mortality compared with rate control therapy in patients with permanent rate-controlled AF, narrow QRS complex in the electrocardiogram, and ≥1 recent hospitalization for heart failure.8 Thus, the detrimental effects of AF may not only be caused by an AF-related atrial cardiomyopathy or by tachycardic heart rates but might also involve direct effects on the LV. This is supported by the above-mentioned CAMERA-MRI trial that illustrated that catheter ablation of AF improved LV function compared with control subjects that had comparable heart rates due to sufficient medical rate control.5 Therefore, even non-tachycardic AF (i.e. arrhythmic ventricular excitation) can negatively influence ventricular function. Scientific investigations on AF mostly focused on the atria and the potential pathophysiological effects of AF on the LV are still unclear. Even less is known about the role of AF-mediated arrhythmic ventricular excitation despite a high number of patients with LVSD express rate-controlled AF. In the light of this, we could recently provide first insights into the effects of AF on the human LV (Figure 2).9 Human ventricular myocardium from patients with rate-controlled AF or sinus rhythm with preserved LV function (LV ejection fraction >50%) undergoing aortic valve replacement surgery due to aortic valve stenosis was studied. Left-ventricular myocardium from AF patients showed no structural remodelling with respect to LV fibrosis. However, isolated human ventricular cardiomyocytes from AF patients were characterized by a reduced systolic Ca2+ transient amplitude, which strongly determines systolic contractility. While these results firstly describe the ventricular AF-phenotype in the human, clinical features may also influence the findings. Thus, these results were also confirmed in LV myocardium from non-failing donors with sinus rhythm or AF. Additionally, prospective in vitro simulation of normofrequent AF also resulted in an impaired systolic Ca2+ handling in human LV cardiomyocytes from non-failing donors as well as in human induced pluripotent stem cell cardiomyocytes (iPSC-cardiomyocytes). Thus, the adverse AF-induced ventricular remodelling was reproduced in different human-based approaches encompassing the clinical phenotype in patients and the controlled in vitro models excluding clinical confounders. Further mechanistic experiments revealed an impaired sarcoplasmic reticulum Ca2+ load that is likely caused by an increased diastolic sarcoplasmic reticulum Ca2+ leak in iPSC-cardiomyocytes following AF-simulation. This observation may underlie the depression of systolic Ca2+ release and hence LV contractility. Associated with the disturbances in Ca2+ homeostasis, cytosolic Na+ concentration was elevated, likely via increased late Na+ current, which finally resulted in prolonged ventricular action potentials following AF-simulation. Thus, a profound and adverse electrophysiological remodelling determined by typical heart failure hallmarks is present in the AF ventricle. Investigations of the potential trigger of these mechanisms revealed that levels of reactive oxygen species were elevated in LV myocardium from AF patients leading to higher oxidation of the Ca2+/calmodulin-dependent protein kinase IIδc (CaMKII), which plays a central role in maladaptive remodelling in cardiac disease. This resulted in increased CaMKII activity and consequently increased phosphorylation of the ryanodine receptor at the CaMKII-dependent phospho-site in the AF ventricle. This posttranslational modification is known to shift the open probability of the channel in a more leaky state thereby explaining the increased diastolic Ca2+ leak in the AF ventricle. As a proof of concept, treatment with a CaMKII-inhibitor or an oxidative stress scavenger ameliorated the cellular phenotype caused by AF-simulation. While our translational study provides the first characterization of the detrimental mechanisms of AF on the human LV, further efforts are required to elucidate additional mechanisms including metabolic alterations, ion channel regulation, the role of inflammation and oxidative stress, and a better clinical characterization. Moreover, clinical management of LVSD associated with AF is demanding and the causal understanding remains controversial. Accordingly, some aspects complicate the understanding of the disease: As rhythm disorders are also complications of LVSD, AF-induced LVSD may be a hidden (and thus overlooked) cause of heart failure, particularly when idiopathic cardiomyopathy with concomitant AF is diagnosed. AF-induced LVSD is potentially curable and can be successfully treated by rhythm restoration. Thus, the correct diagnosis is essential but difficult as it can only be established retrospectively after rhythm control-induced recovery of LV function. The current discussion on AF-induced LVSD and interpretation of related clinical trials mixes too many heart failure aetiologies such as patients without other explanations of LVSD and patients with pre-existing structural ventricular abnormalities (i.e. coronary artery disease) and concomitant AF. Therefore, clinical and experimental scientific efforts are required to improve our understanding of this challenging condition. In summary, clinical trials and our experimental investigations have shown that normofrequent AF per se has distinct detrimental effects on the human LV via adverse remodelling of cardiomyocyte excitation–contraction coupling causing LVSD. These findings could contribute to our understanding of the favourable effects of rhythm control via catheter ablation in patients with AF and concomitant heart failure. Our novel data may represent, however, only a first glimpse into this AF-induced condition, which exceeds the atria and may affect the whole heart as well as other organs. The heart image in Figure 2 is licensed by Shutterstock.com/Usama Nasir MD. S.S. is funded by the Deutsche Forschungsgemeinschaft (DFG) through the research grant SO 1223/4-1 and the F. Thyssen Foundation (Az 10.19.2.026MN). S.P. is funded by the Else-Kröner-Fresenius Stiftung (2019_A84) and by the German Society of Internal Medicine (Clinician Scientist Program).

Topics & Concepts