https://orcid.org/0000-0003-4893-0824
The interaction and mediation analysis to identify mediators of treatment success for early rhythm control in atrial fibrillation (AF) patients with respect to the primary composite cardiovascular (CV) outcome.
This editorial refers to ‘Attaining sinus rhythm mediates improved outcome with early rhythm control therapy of atrial fibrillation: the EAST-AFNET 4 trial’, by L. Eckardt et al., https://doi.org/10.1093/eurheartj/ehac471.
Clinicians have long been focused on achieving sinus rhythm for patients with atrial fibrillation (AF). It seems only natural that pushing a disease’s distinguishing feature into quiescence will permit patients to live longer and better. History, however, has proven that this is a difficult task and does not always translate into benefit in patient-important outcomes.
The landmark randomized EAST-AFNET 4 trial found that an early rhythm control (ERC) strategy in patients with new-onset (<1 year from time of diagnosis) AF reduces a composite of cardiovascular death, stroke, or hospitalization for heart failure or acute coronary syndrome.1 This trial was pragmatic in its design in that a large number of approaches to rhythm control were used. It offered promise for patients and clinicians, but left many asking ‘How did the rhythm control approach work?’ That being in sinus rhythm is beneficial for our patients confirms our clinical feeling, and it has been speculated that a rhythm control strategy early in the disease course may lead to improved cardiovascular outcomes. This could work by halting pathological and deleterious atrial cardiomyopathy characterized by structural remodelling, impaired function, and electrophysiological instabilities.2 However, data supporting evidence for the role of these mechanisms and their contribution to the beneficial outcome in the EAST-AFNET 4 trial have been lacking.
In this issue of the European Heart Journal, Eckardt et al.3 performed a pre-specified subanalysis of the EAST-AFNET 4 trial aiming to identify which clinical variables explained the benefit of the ERC strategy. They performed a causal interaction and mediation analysis to identify mediators of treatment success with respect to the primary composite outcome and two of its key components, cardiovascular death and stroke. Their applied statistical model decomposed the overall effect of the ERC approach on the outcome in EAST-AFNET 4 into a direct effect and a pathway via a potential mediator to allow the analysis of exposure–mediator interactions and a causal interpretation (Graphical Abstract). ERC reduced the first primary outcome in the period beginning at 12 months and continuing up to the study end (median follow-up 4.1 years) by 27%. The presence of sinus rhythm as documented by a resting 12-lead ECG at 12 months explained 81% of this decrease. In contrast, AF recurrence in the first 12 months of follow-up explained only 31% of the treatment effect and systolic blood pressure at the 12-month visit explained only 10%.
Is restoring or maintaining sinus rhythm really the key to unlocking this protective effect? Or perhaps patients in whom we succeed in maintaining sinus rhythm represent a lower risk group? Importantly, EAST-AFNET 4 did not study the effect of a specific intervention but investigated ERC as a pragmatic clinical approach. The benefit of ERC was achieved using a variety of combinations of antiarrhythmic drugs and AF ablation. The AF better care (‘ABC’) approach4 and oral anticoagulation were used in all study participants irrespective of rhythm outcome or randomization allocation.5 Although the interaction and mediation analyses presented by Eckardt et al. suggest a key role of being in sinus rhythm for the reduction in cardiovascular outcomes,3 it does not prove a causal relationship. Sinus rhythm on a single resting ECG at 1-year follow-up may reflect the degree of atrial cardiomyopathy, which may mediate the response to ERC and has also been shown to be associated with the risk of cardiovascular complications including heart failure, stroke, and death.2 The progression of atrial cardiomyopathy may be delayed or reversed by direct and pleiotropic effects of the multimodal approach, including systematic risk factor control, anticoagulation, and rhythm control interventions by antiarrhythmic drugs or ablation.2,6,7
From this analysis, the question arises about how to treat patients who are in AF at the 12-month visit. Is this a signal that the window of opportunity for rhythm control has closed or rather a call to double down on our efforts to achieve sinus rhythm at all costs? Future trials exploring the best escalation strategy involving hybrid approaches with AF ablation and antiarrhythmic drugs or re-do ablation procedures with or without experimental treatment combinations are required. A rate control strategy may always be an acceptable option in appropriately selected patients.
In EAST-AFNET 4, sinus rhythm was controlled by predominantly class 1 antiarrhythmic drugs in the majority of patients (84% in the ERC group) and ablation was performed in a minority (24% in the ERC group). Interestingly, in the ATTEST trial, progression from paroxysmal to persistent AF at 3-year follow-up was delayed by rhythm control, and AF ablation was superior to antiarrhythmic drugs.8 Additionally, an exploratory analysis of CABANA suggested that AF ablation may improve cardiovascular outcomes in younger patients with fewer comorbidities.9 In order to investigate the best approach to how ERC should be achieved, Eckardt et al. also studied the timing of AF ablation in an exploratory analysis of 98 patients. They showed that early ablation within 8 weeks after randomization was associated with a reduction in cardiovascular events, as compared with AF ablation performed later. Additionally, the exploratory analysis did not find that AF ablation was associated with better outcomes than antiarrhythmic drug therapy. Although these data support the concept that AF ablation should be performed earlier rather than later, the number of patients undergoing ablation was very small and the analysis was consequently underpowered. Additionally, the stratification of ablation vs. other means of ERC was not incorporated in the randomization process. Although the EARLY-AF and STOP-AF trials10,11 did not have the power to deduce if early ablation would lead to decreased cardiovascular mortality as compared with antiarrhythmic drugs, they showed that cryoballoon AF ablation as an initial treatment strategy in patients with treatment-naïve AF significantly improved arrhythmia outcomes (freedom from any, or symptomatic, atrial tachyarrhythmia). Whether this improvement in arrhythmia outcomes results in reduced cardiovascular outcomes warrants further study with AF ablation as the primary mean of ERC. This also raises further questions about the implementation of ERC in clinical practice. Waiting lists for AF ablation are generally long and may disqualify AF ablation as the initial treatment option in most countries.
In EAST-AFNET 4, sinus rhythm in a resting 12-lead ECG at 12 months follow-up was identified as the main mediator for cardiovascular outcomes in the 4.1 years follow-up period. A previous analysis showed that ERC reduced the first primary composite outcome in all AF patterns, but a minority of patients had persistent AF.12 Therefore, the single ECG at 1 year has probably missed several AF episodes during follow-up, but may, however, distinguish patients with low or no AF burden from patients with high AF burden. To accurately prove freedom from AF during ERC, a continuous rhythm-monitoring approach would be required to determine AF burden. According to the study protocol, patients randomized to ERC were required to transmit a single-lead ECG twice per week and while symptomatic, and in total 97 978 abnormal ECGs were collected.4 These data may add interesting additional information on arrhythmia outcomes and efficacy of rhythm control in EAST-AFNET 4, but may also provide an additional interesting factor mediating the cardiovascular outcome risk. We also need to address the questions of whether the rhythm on a single 12-lead ECG performed 1 year after beginning ERC and cardiovascular endpoints at 4.1 years follow-up are also the most relevant issues to address for our AF patients. Effective symptom control and improvement in health-related quality of life remain the important outcome for most of our patients. The CIRCA DOSE trial showed13 that in primarily low-burden paroxysmal AF patients with continuous rhythm monitoring by implantable cardiac monitors, the reduction in AF burden following ablation is associated with a clinically meaningful improvement in quality of life, making AF burden the optimal electrical complementary endpoint14 to predict, monitor, and guide the improvement of symptoms and quality of life in our patients.
In general, the implementation of ERC will require a multidisciplinary approach with a shared decision-making process in order to discuss the benefits of rhythm control therapy. To give all patients with recently diagnosed AF and concomitant cardiovascular conditions access to an ERC approach, remote digital care pathways15 may be required to allow immediate initiation of ERC strategies and ensure continuous patient education, risk factor and lifestyle modification, and monitoring of the success of ERC focusing on a reduction in AF burden and improvement of AF-related symptoms and quality of life. It follows naturally that with such an approach, achieving and maintaining sinus rhythm is a marker of success. However, when and how to measure remain open questions. A more pressing question, though, is what to do for the patient who remains in AF despite our best efforts. Do we fight at all costs to get them out of AF, or do we have to accept that for some patients sinus rhythm was not meant to be?
Data availability
There are no new data associated with this article.
References
Author notes
The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.
Conflicts of interest: None declared.
