Atrial fibrillation: from biomarkers to CABANA

For the podcast associated with this article, please visit https://academic.oup.com/eurheartj/pages/Podcasts. Atrial fibrillation is the most common arrhythmia. It is associated with symptoms and hence reduced quality of life, as well as an increased rate of stroke and death.1 While traditionally antiarrhythmic drugs have been used to treat the condition, catheter-based ablation has become a preferred treatment option thanks to the encouraging results of the procedure overall.2,3 and in patients with heart failure.4 It remains unclear, however, to what extent ablation improves outcomes in patients without heart failure. Indeed, the CABANA trial did not meet its primary endpoint based on intention-to-treat analysis, while there was a positive signal on a per-protocol analysis.5 Since trials, particularly those recruiting slowly such as CABANA, may not reflect the real-world situation, the current FAST TRACK ‘Atrial fibrillation ablation in practice: assessing CABANA generalizability’ by Peter A. Noseworthy and colleagues from the Mayo Clinic Minnesota in Rochester, Minnesota, USA6 is a timely contribution. In a propensity analysis of 12 032 patients undergoing ablation and 171 728 treated medically, they found that ablation was associated with a reduction in the primary composite endpoint of all-cause mortality, stroke, major bleeding, and cardiac arrest, particularly in patients who would have been eligible for the trial, which were about three-quarters of the real-world patients (Figure 1). These findings give some weight to the disputed per-protocol analysis of CABANA and support the generalizability of the trial, as discussed in a balanced Editorial by John A. Camm.7 Hazard ratio for primary outcome stratified by potential trial-eligibilty in the overall population (N = 183 760) (from Noseworthy PA, Gersh BJ, Kent DM, Piccini JP, Packer DL, Shah ND, Yao X. Atrial fibrillation ablation in practice: assessing CABANA generalizability. See pages 1257–1264). Undetected atrial fibrillation is a major health concern, particularly because of its association with stroke.8–10 Thus, diagnostic and prognostic biomarkers have been proposed11,12 as they could simplify screening and allow for stratified prevention and treatment of atrial fibrillation. In their article entitled ‘Data-driven discovery and validation of circulating blood-based biomarkers associated with prevalent atrial fibrillation’, Paulus Kirchhof and colleagues from the University of Birmingham in the UK investigated whether blood biomarkers associated with atrial fibrillation would be clinically useful.13 Forty cardiovascular biomarkers were quantified in 638 consecutive patients with known atrial fibrillation or ≥2 CHA2DS2-VASc risk factors, but not paroxysmal or silent atrial fibrillation. Atrial fibrillation was significantly associated with age, male sex, body mass index, elevated brain natriuretic peptide, elevated fibroblast growth factor-23,14 and reduced tumour necrosis factor-related apoptosis-induced ligand-receptor 2, but not other biomarkers (Figure 2). These biomarkers improved the prediction of atrial fibrillation compared with clinical risk factors alone, with a net reclassification improvement of 0.178. Both logistic regression and machine learning predicted atrial fibrillation well during validation, with a receiver operating characteristic (ROC) curve of 0.684. Thus, three simple clinical risk factors such as age, sex, and BMI, and BNP and FGF-23 may help to identify patients with atrial fibrillation. Data-driven discovery identifies BNP and FGF-23 as biomarkers for AF. Brain natriuretic peptide and fibroblast growth factor 23 identified by regression and machine learning to be robustly associated with atrial fibrillation in a cohort of 638 patients presenting to hospital. AF, atrial fibrillation; BNP, brain natriuretic peptide; FGF-23: fibroblast growth factor 23 (from Chua W, Purmah Y, Cardoso VR, Gkoutos GV, Tull SP, Neculau G, Thomas MR, Kotecha D, Lip GYH, Kirchhof P, Fabritz L. Data-driven discovery and validation of circulating blood-based biomarkers associated with prevalent atrial fibrillation. See pages 1268–1276). Increases in fat-free mass and fat mass have been associated with higher risk of atrial fibrillation in observational studies,15,16 but others described an obesity paradox in atrial fibrillation.17 It is not known whether these associations reflect independent causal processes. In their article ‘Body composition and atrial fibrillation: a Mendelian randomization study’ Erik Ingelsson and colleagues from Stanford, California in the USA addressed this issue using 487 404 individuals of the UK Biobank with 10 365 events.18 They confirmed that both fat-free mass and fat mass had observational associations with incident atrial fibrillation. The causal effects using the inverse-variance weighted method were 1.55 (for fat-free mass) and 1.30 (for fat mass). The multivariate Mendelian randomization analysis suggested that the causal effects of fat-free mass and fat mass were independent of each other, with causal risk ratios of 1.37 and 1.28, respectively. Thus, genetic increases in fat-free mass and fat mass independently increase the risk of atrial fibrillation. These novel and important findings are further discussed in an informative Editorial by Heribert Schunkert from the Deutsches Herzzentrum in Munich, Germany.19 Heart failure, both that with reduced (HFrEF) and that with preserved ejection fraction (HFpEF), is associated with atrial fibrillation.20–22 However, while atrial fibrillation has been extensively investigated in HFrEF, including in a randomized trial on ablation,4 less is known about atrial fibrillation in HFpEF. HFpEF can occur in several cardiac disorders, but amyloidosis has recently been particularly well characterized23–25 and staged,26 and, importantly, novel treatment options have been developed.27 In their review ‘Heart failure with preserved ejection fraction, atrial fibrillation, and the role of senile amyloidosis’28 Maarten van den Berg and colleagues from the University Medical Center Groningen in The Netherlands remind us that there are two types of senile amyloidosis, i.e. wild-type transthyretin (TTR)-derived amyloidosis (ATTRwt) and isolated atrial amyloidosis (IAA) which are both associated with atrial fibrillation. New diagnostic and therapeutic possibilities have emerged that facilitate clinical management of amyloidosis, which in turn may have implications for the management of HFpEF and atrial fibrillation. The most common complication of atrial fibrillation is embolic stroke. As the risk of stroke has important clinical implications, among them anticoagulation with its risk of bleeding, precise risk stratification29 as a basis of therapeutic decisions is essential.30 This is further discussed in the review ‘Risk stratification for stroke in atrial fibrillation: a critique’ by Ammar M. Killu and colleagues from the Mayo Clinic in Rochester, Minnesota, USA.31 They note that given its burden and associated complications, atrial fibrillation portends a considerable public health problem. Besides stroke and death, there is an association of atrial fibrillation with impaired cognitive functioning.32,33 To mitigate this, anticoagulation has played a major role in atrial fibrillation. While vitamin K antagonists were the mainstay of therapy for decades, non-vitamin K oral anticoagulants or NOACs are increasingly used34 as agents that are modestly more effective at preventing stroke, and substantially safer regarding intracranial haemorrhage, than warfarin. The authors discuss the pathophysiology of atrial fibrillation and the related risk of thrombo-embolic stroke as a basis of recommendations with regards to strengths and limitations of current risk stratification strategies, and discuss gaps in our evidence base. This issue is complemented by the Current Opinion ‘The dilemma of professional and industry relations for medical education’ by Anthony N. DeMaria from the University of California La Jolla in the USA.35 Demaria suggests that there is a relationship between the cardiovascular professional community and the cardiovascular industry, which has always existed. Industry has depended upon physicians to help them define important clinical needs, to assess the efficacy of their innovations in clinical trials, and to disseminate the knowledge and application of their products to the practising community. Significantly, the use of drugs and devices has required prescription by physicians, who have been the primary customers for industry. For their part, cardiovascular specialists have depended upon industry to provide solutions for their clinical needs, to support basic and clinical research, and, especially, to help subsidize medical education. This interaction has fulfilled important needs for both parties and has generally been mutually beneficial. The editors hope that this issue of the European Heart Journal will be of interest to its readers. With thanks to Amelia Meier-Batschelet for help with compilation of this article.