Thromb Haemost 2021; 121(03): 251-254
DOI: 10.1055/a-1336-0576
Invited Editorial Focus

Embolic Stroke of Undetermined Source: The Need for an Integrated and Holistic Approach to Care

1  Department of Cardiology, Medical School of Chinese PLA, Chinese PLA General Hospital, Beijing, China
2  Liverpool Centre for Cardiovascular Sciences, University of Liverpool and Liverpool Heart and Chest Hospital, Liverpool, United Kingdom
Juqian Zhang
2  Liverpool Centre for Cardiovascular Sciences, University of Liverpool and Liverpool Heart and Chest Hospital, Liverpool, United Kingdom
1  Department of Cardiology, Medical School of Chinese PLA, Chinese PLA General Hospital, Beijing, China
2  Liverpool Centre for Cardiovascular Sciences, University of Liverpool and Liverpool Heart and Chest Hospital, Liverpool, United Kingdom
3  Aalborg Thrombosis Research Unit, Department of Clinical Medicine, Aalborg University, Aalborg, Denmark
› Author Affiliations

Atrial Cardiopathy and Likely Pathogenic Patent Foramen Ovale in Embolic Stroke of Undetermined Source

The term embolic stroke of undetermined source (ESUS) was first introduced in 2014 to differentiate it from cryptogenic stroke.[1] ESUS was defined as a nonlacunar brain infarct without proximal arterial stenosis or cardioembolic source, with a clear indication for anticoagulation.[1] The condition consists on average of 17% of ischemic strokes,[2] and is generally associated with relatively young patients, small emboli, and a high stroke recurrence (>4%), partly due to unclear underlying mechanism(s) which led to inappropriate preventive strategies.[2] Much focus has been directed toward improving cardiac rhythm detection and understanding of stroke mechanisms in ESUS.

In a study using machine-learning prediction of stroke mechanisms, 44% ESUS were identified as related to cardiac embolism, and each individual's likelihood of cardiac embolism was associated with eventual detection of atrial fibrillation (AF).[3] Higher probability of cardioembolism for patients with ESUS was associated with older age, vascular disease, heart failure, larger left atrium (LA) size, lower blood pressure, and increased creatinine levels.[3] Although various studies have focused on cardiac rhythm monitoring to identify AF-related cardiac embolism among patients with ESUS,[4] [5] [6] other evidence points toward atrial cardiopathy, patent foramen ovale (PFO), and arterial disease as potential etiologies of ESUS (see [Fig. 1]).

Zoom Image
Fig. 1 Main etiologies of ESUS apart from atrial fibrillation. (A) Arterial disease/atherosclerosis. (B) Atrial cardiopathy. (C) Patent foramen ovale. ESUS, embolic stroke of undetermined source.

Cardiopathy was first introduced as “atrial cardiomyopathy” to define a familial syndrome affecting mainly the atria and atrioventricular-conducting system in 1972.[7] With the accumulation of evidence, the concept of atrial cardiomyopathy has evolved to define any complex structural, architectural, contractile, or electrophysiological changes affecting the atria with the potential to produce clinically relevant manifestations,[8] which might contribute to stroke risk independently of AF. Nonetheless, the mechanism(s) linking atrial cardiopathy and stroke risk is incompletely elucidated.

The clinical, electrocardiographic, and echocardiographic variables and the biochemical markers associated with atrial cardiopathy have been identified among patients with ESUS,[9] [10] [11] [12] [13] most of which reflected left atrial enlargement, although LA fibrosis and dysfunction were also noted as its features. Various studies in patients with continuous monitoring of atrial rhythm also confirmed that atrial high-rate episodes, atrial tachycardia, and subclinical AF were associated with stroke risk.[14] [15] [16] In fact, the anatomical and physiological remodeling of atrial tissue may be the substrate for the thromboembolism associated with stroke.[17] Of note, there are other less pathogenic reasons for atrial cardiomyopathy, e.g., atrial amyloidosis, hereditary muscular dystrophies, gene mutations, genetic repolarization disturbances, drugs, myocarditis, etc. Indeed, assessment of the (atrial) substrate is also recommended as part of the overall characterization of conditions that lead to a high risk of stroke, such AF.[18] While it is assumed that those patients could possibly benefit from anticoagulant treatment, one ongoing clinical trial will address the issue.[19]

Stroke related to PFO was primarily thought to be a consequence of paradoxical embolism originating as venous thrombus in the year 1988.[20] Subsequent studies showed that there was a high prevalence of PFO in patients with ESUS, although at least one-third of PFOs discovered were likely to be incidental.[21] More than one in 10 patients older than 60 years with a cryptogenic event have PFO combined with a large right-to-left shunt.[22] Besides atrial cardiopathy and PFO, there are other heterogeneous pathologies reported for ESUS, including left ventricular disease, covert AF, atherosclerosis, endothelial dysfunction, valvular heart disease, and cancer ([Table 1]).

Table 1

The potential etiologies of patients with embolic stroke of undetermined source



Main finding

Atrial cardiomyopathy

Kamel et al 2015[36]

• 14,542 participants 45 to 64 years

• 904 participants (6.2%) with a definite or probable ischemic stroke

• Left atrial abnormality was associated with nonlacunar stroke (HR: 1.49; 95% CI: 1.07–2.07)

Kamel et al 2015[37]

• Patients in sinus rhythm who subsequently had ischemic stroke (n = 241) and a randomly selected subcohort without stroke (n = 798)

• ECG-defined left atrial abnormality was associated with incident cryptogenic or cardioembolic stroke independently of the presence of AF

Meisel et al 2019[10]

• 18 ESUS patients, average age was 58 years old and 44% were females

• ESUS subjects have LA dysfunction and remodeling at rest and exercise

Patent foramen ovale

Lechat et al 1988[20]

• 60 adults with ischemic stroke under 55 years old, a control group of 100 patients

• The prevalence of patent foramen ovale was significantly higher in the patients with stroke than in the control group

Kim et al 2013[38]

• 157 patients with presumed cryptogenic embolism

Right-to-left shunt was observed in 96 (61.1%) patients

• Right-to-left shunt was more frequently observed in patients with small (<1 cm) infarcts than in those with a large infarct (66·7 vs. 45·9%)

• The mechanisms of stroke other than paradoxical mechanism may play an important role in patients with large cryptogenic embolic stroke

Other potential embolic sources

Ntaios et al 2015[39]

• 2,735 patients with ESUS

• Covert paroxysmal atrial fibrillation: 43.2%; left ventricle: 20.2%; arteriogenic emboli: 13.9%; nonatrial fibrillation atrial dysrhythmias and stasis: 5.9%; paradoxical embolism: 5.1%; mitral valve: 4.7%; aortic valve: 5.5%; atrial structural abnormalities: 3.6; cancer-associated: 1.2%

Perera et al 2018[40]

• 2,144 patients with ischemic stroke

• 24% (n = 78) of young versus 15% (n = 273) of older patients with embolic stroke of undetermined source criteria.

• Fewer young embolic stroke of undetermined source patients (63%) had potential minor risk embolic sources identified versus older embolic stroke of undetermined source patients (77%)

Umemura et al 2019[41]

• 292 consecutive patients (135 men, 157 women; mean age: 74.3 ± 11.6 years) diagnosed with cerebral infarction

• 27 of whom were diagnosed with embolic stroke of undetermined source (9.2%; 14 men, 13 women; mean age: 70.7 ± 11.5 years).

• Valve calcification (11.1%), left ventricle diastolic dysfunction (18.5%), cancer-associated stroke (25.9%), covert atrial fibrillation (7.4%), aortic arch atherosclerotic plaques (11.1%), paradoxical embolism (3.7%), and sick sinus syndrome (3.7%).

Ntaios et al 2019[42]

• 1,382 patients with ESUS

• 397 (29%) had aortic arch atherosclerosis and 112 (8%) had complex aortic arch atherosclerosis

Ikonomidis et al 2019[43]

• 90 patients with ESUS vs. 90 controls with similar risk factors

• Compared with control, ESUS patients had significantly higher arterial stiffness, endothelial glycocalyx damage, and oxidative stress, as well as higher LA volume, and reduced reservoir LA strain (p < 0.05). In ESUS, glycocalyx damage was related with increased pulse wave velocity which was linked with LA reservoir strain.

Abbreviations: AF, atrial fibrillation; CI, confidential interval; ESUS, embolic stroke of undetermined source; HR, hazard ratio; LA, left atrium.

In this issue of Thrombosis and Haemostasis, Leventis et al[23] reported that the presence of atrial cardiopathy was inversely related to the presence of likely PFO in patients with ESUS. The prevalence of atrial cardiopathy was lower in patients with likely pathogenic PFO compared with patients with likely incidental PFO (31.2%) or without PFO, indicating that atrial cardiopathy and likely pathogenic PFO were two competing etiologies for ESUS.

Their findings were in keeping with two recent studies, both used hierarchy clustering analysis, an experimental machine-learning technique. One study identified three different clusters of clinical phenotypes among 127 patients with ESUS: young age, PFO, posterior circulation infarct, and male sex; hypertension, left atrial cardiopathy, severe stroke, involvement of multiple vascular territories, and diabetes; and dyslipidemia, ipsilateral vulnerable substenotic carotid atherosclerosis, smoking, anterior territory infarct.[24] The other study noted four distinct clusters of potential embolic sources for ESUS among 800 patients with ESUS: the largest cluster was associated with arterial disease, followed by atrial cardiopathy, PFO, and left ventricular disease in decremental sizes.[25] These studies support the heterogeneity of ESUS where various embolic sources and risk factors play distinct roles ([Supplementary Table S1], available in the online version).[26]

Thus far, large clinical trials (NAVIGATE-ESUS and RE-SPECT-ESUS) have found that anticoagulation using rivaroxaban or dabigatran as nonsuperior to aspirin in preventing stroke recurrence following the onset of ESUS, and in the case of rivaroxaban, it was associated with more bleeding events.[27] [28] There are two ongoing trials comparing apixaban with aspirin in secondary prevention of stroke, one in cryptogenic stroke and atrial cardiopathy[19] and the other in ESUS with insertable cardiac monitor for AF.[29] Their findings will shed more light on a more individualized stroke prevention strategy for subgroups of ESUS.

The optimal approach for the prevention of recurrent stroke in ESUS relates to age and addressing combined risk factors among those patients with ESUS.[30] An integrated approach to stroke care facilitates a more holistic approach, especially since ESUS management is likely to involve multidisciplinary teams. For example, improved monitoring will aid detection of associated AF, especially if we “look harder, look longer, and look in more sophisticated ways.”[31] Indeed, an integrated approach to AF management is now recommended in guidelines, being associated with good long-term outcomes.[32] [33] [34] Older patients with ESUS might possibly benefit from an anticoagulation strategy, given increased comorbidities for thromboembolic risk over aging.[35] Time will provide more evidence to help guide the optimal approach to investigation and management of AF.

* The review process for this paper was fully handled by Christian Weber, Editor-in-Chief.

Supplementary Material

Publication History

Publication Date:
11 December 2020 (online)

© 2020. Thieme. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany