Atrial fibrillation and stroke: anticoagulation in trials and real-world settings

For the podcast associated with this article, please visit https://academic.oup.com/eurheartj/pages/Podcasts. Stroke is a devastating event for patients and their families alike. For cardiologists, embolic strokes are most important1 as they are potentially preventable by appropriate measures such as drugs and interventions.2 Besides hypertension,3 a patent foramen ovale4 and aortic and carotid plaques atrial fibrillation are the most important causes of stroke.5 Indeed, the importance of left atrial thrombus formation6 and neurological events, as well as the protective role of anticoagulants, is well established.7,8 In their clinical review ‘Stroke prevention in atrial fibrillation: redefining ‘real-world data’ within the broader data universe’, Renato Lopes and colleagues from the Duke Medical Center in Durham, North Carolina, USA9 remind us that although real-world data have received increasing attention from regulatory authorities and professional societies, dividing evidence into that derived from real-world vs. ‘non-real-world’ sources provides only one element of a larger framework for evidence evaluation. Evidence should be evaluated on the basis of the source of the data, the method of treatment allocation (i.e. randomized or simply observed as used in practice), and the context in which the evidence was generated (i.e. overall study design). A study incorporates real-world data when it primarily uses data collected for non-research purposes, such as insurance claims or health records, regardless of study design. Separation of study design, data source, and context enables the parallel evaluation of critical elements: (i) whether a study can support claims of causal inference, which can be assured with confidence only where patients are assigned to treatments by protocol, and (ii) whether the study population and clinical context mirror clinical practice, a strength of observational studies. Observational studies employing real-world data are useful for describing how oral anticoagulants are used in clinical practice (see below10), but generally cannot be used to make claims regarding comparative treatment effects. Questions regarding treatment effects generally are best answered through a randomized clinical trial, and additional pragmatic randomized clinical trials are needed to evaluate key clinical questions about the prevention of thrombotic events in atrial fibrillation. Atrial fibrillation ablation is usually performed with continuous anticoagulation,11 although evidence for it is lacking. In their FAST TRACK paper ‘Apixaban in patients at risk of stroke undergoing atrial fibrillation ablation’, Paulus Kirchhof and colleagues from the University of Birmingham in the UK note that the safety of vitamin K and non-vitamin K antagonist oral anticoagulants (NOACs) regarding cognitive function after ablation are not known.12 Thus, they randomized 674 patients with stroke risk factors undergoing atrial fibrillation ablation to continuous apixaban (5 mg b.i.d.) or vitamin K antagonists (International Normalized Ratio 2–3) in a prospective, open, multicentre study with blinded outcome assessment. The primary outcome occurred in 6.9% on apixaban, and in 7.3% on vitamin K antagonists (P=0.0002 for non-inferiority; Figure 1). Cognitive function increased by 1 Montreal Cognitive Assessment unit at the end of follow-up without differences between groups. In analysable magnetic resonance imaging (MRI), acute brain lesions were found in 27.2% randomized to apixaban, and in 24.8% randomized to vitamin K antagonists. Thus, continuous apixaban is a safe and effective alternative to vitamin K antagonists in patients undergoing atrial fibrillation ablation at risk of stroke with respect to bleeding, stroke, cognitive function, and MRI-detected brain lesions. These intriguing findings are put into context in an Editorial by Christopher B. Granger from the Duke University Medical Center in Durham, North Carolina, USA.13 The Kaplan–Meier curve for time to first event for (A) stroke/SE, (B) death, (C) major bleeding, and (D) major/CRNM bleeding from randomization. CRNM, clinically relevant non-major; SE, systemic embolism; VKA, vitamin K antagonist (from Ezekowitz MD, Pollack CV, Halperin JL, England RD, VanPelt Nguyen S, Spahr J, Sudworth M, Cater NB, Breazna A, Oldgren J, Kirchhof P. Apixaban compared to heparin/vitamin K antagonist in patients with atrial fibrillation scheduled for cardioversion: the EMANATE trial. See pages 2959–2971). Cardioversion has traditionally been recommended after prolonged anticoagulation with vitamin K antagonists, unless atrial fibrillation was of very short duration. Today, NOACs are an alternative.5,14 This has been tested in a FAST TRACK paper entitled ‘Apixaban compared to heparin/vitamin K antagonist in patients with atrial fibrillation scheduled for cardioversion. The EMANATE trial’, in which Michael Ezekowitz and colleagues from the Sidney Kimmell Medical College in Bryn Mawr, Pennsylvania, USA compared apixaban 5 mg b.i.d. (or 2.5 mg b.i.d. in those with two of the following: age ≥ 80 years, weight ≤ 60 kg, or serum creatinine ≥ 133 µmol/L) to heparin/vitamin K antagonism in 1500 patients with atrial fibrillation with less than 48 h anticoagulation prior to cardioversion.15 Comparing apixaban to heparin/vitamin K antagonists, there were 0 vs. 0.8% strokes, no systemic embolism, and two vs. one deaths. In the safety population, there were 0.4 vs. 0.8% major and 11 vs. 13 clinically relevant non-major bleeding events. On imaging, 60 out of 61 patients with thrombi continued randomized treatment, all without outcome events. Thus, rates of stroke and bleeding were low with both treatments, supporting the use of apixaban for anticoagulation of patients with atrial fibrillation undergoing cardioversion. The results are further discussed in a thoughtful Editorial by Kenneth W. Mahaffey from the Stanford University Medical Center in Palo Alto, California, USA.16 Oral anticoagulation has been advocated with increasing evidence to prevent stroke in patients with atrial fibrillation for many years.17,18 However, its uptake by physicians and patients alike was initially slow. In their article ‘A 10 year study of hospitalized atrial fibrillation-related stroke in England and its association with uptake of oral anticoagulation’, James Campbell Cowan and colleagues from the Leeds General Infirmary in the UK determined in a real-world setting whether changing patterns of anticoagulant use in atrial fibrillation have impacted on stroke rates in England.10 The number of patients with atrial fibrillation increased linearly over a decade from 692 054 to 983 254 with a prevalence of 1.29 and 1.71%, respectively (Figure 2). Hospitalizations for atrial fibrillation-related stroke/100 000 patients increased from 80/week in 2006 to 98/week in 2011, and declined to 86/week in 2016. Anticoagulant use amongst those with CHA2DS2-VASc > 2 increased from 48 to 79%, and antiplatelet use declined from 43 to 16%. After adjustment, a 1% increase in anticoagulant use was associated with a 0.8% decrease in the weekly rate of atrial fibrillation-related stroke. Thus, within a decade, atrial fibrillation prevalence and anticoagulant use in England increased, leading to a decline in hospitalized atrial fibrillation-related strokes. These robust results are put into clinical context in a balanced Editorial by Ben Freedman from the Heart Research Institute in Sydney, North South Wales, Australia.19 Temporal trend in finished consultant episodes of atrial fibrillation-related stroke and national prevalence of atrial fibrillation (from Cowan JC, Wu J, Hall M, Orlowski A, West RM, Gale CP. A 10 year study of hospitalized atrial fibrillation-related stroke in England and its association with uptake of oral anticoagulation. See pages 2975–2983). Atrial fibrillation is often associated with comorbidities such as hypertension and heart failure.20 In their article ‘Targeted therapy of underlying conditions improves sinus rhythm maintenance in patients with persistent AF: results of the RACE 3 trial’, Isabelle C. Van Gelder and the RACE 3 Investigators investigated whether treating comorbidities such as mild–moderate heart failure would improve sinus rhythm maintenance in early persistent atrial fibrillation.21 They randomized 245 patients to rhythm control or intervention therapy consisting of: (i) mineralocorticoid receptor antagonists, (ii) statins, (iii) renin–angiotensin inhibitors, and (iv) cardiac rehabilitation. At 1 year, sinus rhythm was maintained in 75% in the intervention and 63% in the conventional group (odds ratio 1.8, P=0.042). Thus, RACE 3 shows that targeted therapy of heart failure improves sinus rhythm maintenance in patients with persistent atrial fibrillation, a finding that is further assessed in an Editorial by Paulus Kirchhof from the University of Birmingham in the UK.22 The editors hope that readers of this issue of the European Heart Journal will find it of interest to them. With thanks to Amelia Meier-Batschelet for help with compilation of this article.