Atrial fibrillation in patients with cardiac amyloidosis and heart failure: a desperate cause?

This editorial refers to ‘Adverse outcomes of atrial fibrillation ablation in heart failure patients with and without cardiac amyloidosis: a nationwide readmissions database analysis (2015–19)’, by W. Ullah et al., https://doi.org/10.1093/ehjopen/oead026. Recent epidemiological data show an increasing number of patients with cardiac amyloidosis (CA) among heart failure (HF) patients, with a prevalence of 8–17 per 100 000 person-years.1 Infiltrating deposition of insoluble amyloid fibrils in the atrial myocardium impairs atrial conduction, which promotes atrial fibrillation (AF). The presence of CA at the ventricular level also plays a role in the pathophysiology of atrial fibrillation through an increase in atrial and ventricular filling pressures that induces non-specific fibrosis.2 As a result, an estimated 20% of patients with CA also have AF.1 Restoration and maintenance of sinus rhythm remain challenging in patients with HF and AF, particularly because of the higher prevalence of persistent vs. paroxysmal AF in this population.3 Another important challenge is the management of AF in patients with CA. Indeed, as in most patients with impaired ventricular diastolic function, AF-related loss of atrial systole has a significant negative impact on haemodynamics in patients with CA. At the same time, rate-control therapies (particularly beta-blockers) are deemed to worsen HF symptoms in CA patients.4 A rhythm control strategy (spearheaded by catheter ablation of AF) therefore appears to be the only treatment option that can provide clinical benefit to patients with CA and symptomatic AF. To date, however, data regarding AF ablation in patients with CA are scarce and come mainly from single-centre studies involving a small number of patients. In this issue of the European Heart Journal Open, Ullah et al.5 use global health data to obtain a sufficient population to draw initial conclusions about the early outcomes of AF ablation in patients with CA. To conduct this study, data (September 2015 to November 2019) were extracted from the National Readmissions Database (>15 million discharge data points from 22 US states) to identify 148 133 patients with AF and concomitant HF who underwent AF catheter ablation. Based on a balanced distribution of baseline comorbidities, a propensity score matching analysis selected 616 patients (293 with CA and 323 without CA) from this population for further evaluation. Ullah et al. report that AF ablation is associated with significantly higher net adverse clinical events in patients with CA, primarily due to higher in-hospital all-cause mortality (IH-ALM), at both initial admission (adjusted odds ratio for IH-ALM = 9.03, P < 0.0001) and 30-day readmission (adjusted odds ratio for IH-ALM = 4.28, P = 0.04). This study shows that catheter ablation of AF is associated with higher mortality in patients with CA. However, association does not mean causation. For the hypothesis of a causal link to be supported, at least the precise cause and time of death in both groups of patients should be known. Indeed, from international registries, it is known that the occurrence of fatal stroke and fatal tamponade are the two leading causes of death attributable to AF.6 In their study, Ullah et al.5 report no significant difference in the odds of stroke, cardiac tamponade, and major bleeding between the two groups, but the respective rate of fatal events for each specific category of adverse events is unknown. After AF ablation, the vast majority of deaths from periprocedural complications occur (with the exception of atrio-oesophageal fistula) within a few days of the procedure. However, Ullah et al.5 observed significantly higher all-cause mortality at both index admission [odds ratio (OR = 9.03)] and 30-day readmission (OR = 4.28) in patients with CA. Combining this finding with a higher readmission rate (mainly due to cardiovascular causes, including acute exacerbations of HF) in patients with CA, one cannot help but think that CA may ‘simply’ be a much worse prognostic factor in patients with HF, with AF ablation being a confounding factor. This speculation is also supported by the significantly higher proportion of patients with major or extreme loss of function in the CA group. The authors acknowledge important limitations due to the design of this retrospective study based on database analysis. Some information is not accessible in such a register. In particular, the type of AF (paroxysmal or persistent), or the circumstances of ablation of the AF (elective ablation or bailout in unstable patients). If these characteristics were different between the groups, this could influence the differences in populations’ outcomes. Recently, Donnellan et al.3 suggested that catheter ablation of AF, when compared with antiarrhythmic drugs, is more amenable to maintain sinus rhythm in CA patients. In addition, AF catheter ablation was associated with a significant reduction in the rate of hospitalization for HF and AF recurrences.3 Atrial fibrillation catheter ablation remains effective to maintain sinus rhythm in patients with CA, but if it may influence more robust endpoints such as mortality remains unknown. Likewise, the most opportune moment to carry out the ablation remains to be established. Finally, independently of rhythm management, end-stage heart failure is still the leading cause of death in patients with CA. Conflict of interest: None declared.