The subtypes/etiologies of ischemic stroke can be categorized as large artery disease, small artery disease, and cardiac embolism with this classification based on careful history taking, neurological examinations, cardiac evaluations, and the use of imaging techniques such as brain computed tomography/magnetic resonance imaging (CT/MRI) and angiographies. However, despite these thorough workups, the etiologies of approximately 1/3 of ischemic strokes remain unknown and are classified as cryptogenic strokes [1]. Notably, in patients with stroke due to cardiac embolism, the responsible heart diseases include atrial fibrillation (AF), acute myocardial infarction, sick sinus syndrome, valvular disease, and significant heart failure. Furthermore, recent studies have identified new cardiac diseases that can generate embolic stroke, likely accounting for a large portion of cryptogenic strokes.

Paradoxical embolism due to the presence of a patent foramen ovale (PFO) is thought to cause about 50% of cryptogenic strokes [2], with other possible causes including unstable plaques at large arteries, vasculitis, hypercoagulable states, unidentified AF, or other cardiopathies [3]. When cryptogenic strokes are suspected to be caused by embolism based on clinical and imaging findings (e.g., non-lacunar, cortical infarctions), they are referred to as “embolic stroke(s) with unknown source(s) (ESUS).” [4] As stated above, paradoxical embolism associated with a PFO has emerged as the single most important etiology of ESUS, and various therapeutic methods, including percutaneous PFO closure, have been developed to prevent further ischemic stroke or transient ischemic attack (TIA) in these patients. However, there are still controversies in the management of PFOs. Moreover, emerging evidence suggests that even without the presence of AF or PFO, cardiopathies, especially left atrial enlargement (LAE), may lead to thrombus formation and play a role in the development of ESUS. This narrative review describes the current status of these novel embolic sources such as PFOs and cardiopathy.

The foramen ovale is a hole that exists in the wall between the left and right atria of a human fetus. It normally closes during infancy; however, this does not occur in approximately 25% of the general population, leaving a PFO [5]. While the PFO is usually asymptomatic, allowing blood to leak from the right atrium to the left atrium, problems may arise when the blood contains a clot. Studies have shown that PFOs are detected in up to 40% of patients with ESUS [6]. However, the presence of a PFO itself is not definite evidence of a causal association. It has been shown that several factors are associated with an increased risk of stroke occurrence in patients with a PFO, including a large right-to-left shunt (RLS) [7] and the presence of specific anatomical features such as atrial septal aneurysm (ASA) or increased interatrial septal mobility [8].

For the diagnosis of PFOs, echocardiography plays an important role in both detecting and characterizing the anatomical and functional aspects of the defect. Among the available techniques, transthoracic echocardiography is the most commonly used non-invasive diagnostic method for detecting PFOs [4]. However, its sensitivity is low at 46%, though its accuracy in identifying the paradoxical RLS improves when agitated saline contrast is used [9]. Alternatively, transesophageal echocardiography (TEE) has been proven to be accurate in detecting PFOs and provides information on shunt dimension and anatomic features, including ASA. TEE is also helpful in identifying other thromboembolic sources such as significant aortic arch plaques [10]. Another effective method is transcranial Doppler (TCD) with a bubble study, which has shown effectiveness in detecting the presence of RLS and indirectly suggesting the presence of PFOs [11]. If TCD results are negative, further studies may not be necessary. Meanwhile, to assess the severity of shunting, the Spencer scale, which classifies shunting into five grades (0–5) from no shunt to severe shunt, is used to assess the severity of shunting. A grade ≥3 is considered diagnostic [12]. However, despite its high sensitivity (97%), the Spencer scale is limited in distinguishing between cardiac and pulmonary shunting. Overall, although the diagnosis strategies may vary among stroke centers, an initial evaluation with TCD followed by TEE when necessary is a widely accepted approach in clinical practice.

Regarding the risk of future ischemic stroke or TIA, one study showed that the risk ranged from 0% to 4.4% for stroke and from 0% to 14% for stroke or TIA [13]. This wide range suggests that the risk varies based on patient characteristics. A meta-analysis indicated that cryptogenic stroke patients with a PFO did not have a higher risk of recurrent stroke or TIA compared to cryptogenic stroke patients without a PFO [14]. However, this observation does not mean that the risk is low, but rather that the risk of other stroke causes in patients without a PFO is similar to those with a PFO.

Considering that a PFO is present in 1/4 of the general population, it is important to determine whether PFOs are indeed the causes of stroke or TIA in specific patients. To address this, the Risk of Paradoxical Embolism (RoPE) score was developed [15]. This score is based on the understanding that a PFO-related stroke is more likely to occur in younger patients without vascular risk factors and with a superficial infarct on neuroimaging. The RoPE score can be easily calculated by summing the following criteria: (A) no history of hypertension, diabetes, stroke/TIA, or smoking (1 point each); (B) presence of a cortical infarct (1 point); and (C) age categories: 18–29 (5 points), 30–39 (4 points), 40–49 (3 points), 50–59 (2 points), and 60–69 (1 point).

The RoPE score provides an estimate of the probability that a PFO discovered in a patient with cryptogenic stroke is the actual cause of the stroke, rather than an incidental finding. A higher RoPE score corresponds to a higher probability. However, it is important to note that the score does not include high-risk anatomical features of the PFO. Additionally, patients with higher RoPE scores are typically young individuals without risk factors, resulting in lower stroke recurrence rates compared to those with lower scores [16].

When managing patients, determining the role of the PFO as the cause of stroke and ruling out other potential causes is crucial [5]. There are three options available: antiplatelet therapy, oral anticoagulant therapy, and percutaneous PFO closure. Trials comparing the efficacy of anticoagulants and antiplatelets have yielded heterogeneous results. One meta-analysis [17] found that anticoagulants were more effective in preventing recurrent stroke/TIA compared to antiplatelets (7.7% vs. 9.8% event rates, respectively, P=0.03). However, this result was accompanied by a greater risk of major bleeding (7.1% vs. 1.3%; odds ratio [OR] 6.49, 95% confidence interval [CI]: 3.25–12.99, P<0.00001). Thus, the authors suggested that the best management approach should be individualized based on both the risk of recurrent stroke and the bleeding risk. It is worth noting that the study aforementioned was published in 2015 and primarily involved the use of warfarin. In reality, patients with PFOs are not commonly prescribed warfarin due to factors such as their young age (which necessitates the use of anticoagulants for a longer duration than older patients), the narrow therapeutic window of warfarin (which requires repeated international normalized ratio measurements), and potential interactions with food and other drugs. In contrast, non-vitamin K antagonist oral anticoagulants (NOACs) are expected to offer similar efficacy with fewer bleeding complications. Unfortunately, the use of NOACs is limited in many parts of the world, including Korea, as reimbursement by the government is only provided for patients with concomitant AF. Therefore, further studies are needed to gather evidence on the efficacy and safety of NOACs in managing patients with PFO-related stroke.

There have been multiple randomized clinical trials (RCTs) conducted to assess the effectiveness of percutaneous PFO closure for secondary prevention in patients with PFO-associated stroke. The CLOSURE I (Evaluation of the STARFlex Septal Closure System in Patients with a Stroke and/or Transient Ischemic Attack due to Presumed Paradoxical Embolism through a Patent Foramen Ovale) trial [18] enrolled 909 patients with cryptogenic stroke or TIA associated with PFOs. The primary outcome (a composite of stroke or TIA within 2 years of follow-up, death from any cause within the first 30 days, or death from neurologic causes between 31 days and 2 years) was 5.5% in the closure group and 6.8% in the medical-therapy only group. The adjusted hazard ratio (HR) was 0.78 with a 95% CI of 0.45–1.35, resulting in a P-value of 0.37. These findings indicate that there was no clear benefit of PFO closure over medical therapy alone for preventing recurrent stroke or TIA.

Studies such as the RESPECT (Randomized Evaluation of Recurrent Stroke Comparing PFO Closure to Established Current Standard of Care Treatment) trial [19] and the PC (Percutaneous Closure of PFO Using the Amplatzer PFO Occluder with Medical Treatment in Patients with Cryptogenic Embolism) trial [20] failed to confirm the benefit of PFO closure over medical therapy alone. However, a meta-analysis incorporating these three RCTs [21] showed that PFO closure was superior to medical therapy alone in reducing recurrent ischemic stroke (adjusted HR 0.58; 95% CI: 0.34–0.99). This analysis also found a greater benefit when the analysis was limited to the two trials (PC and RESPECT) that utilized the Amplatzer PFO occluder device (adjusted HR 0.41; 95% CI: 0.20–0.88).

Recently, new RCTs have been conducted. For example, the RESPECT extended follow-up study [22] provided insights from a 5.9-year follow-up period. Out of 980 patients, recurrent ischemic stroke occurred in 18 patients in the PFO closure group and 28 patients in the medical therapy group (HR 0.55; 95% CI: 0.31–0.999; P=0.046). The GORE REDUCE (GORE HELEX Septal Occluder / GORE CARDIOFORM Septal Occluder and Antiplatelet Medical Management for Reduction of Recurrent Stroke or Imaging-Confirmed TIA in Patients with Patent Foramen Ovale) trial [23] enrolled 664 patients, 81% of whom had a moderate or large amount of shunts. The median follow-up was 3.2 years. The PFO closure group experienced a significantly lower occurrence of ischemic stroke compared to the medication-only group (1.4% vs. 5.4%; HR 0.23; 95% CI 0.09–0.62; P=0.002). Similarly, the CLOSE (Patent Foramen Ovale Closure or Anticoagulants versus Antiplatelet Therapy to Prevent Stroke Recurrence) trial [24] randomized 669 patients who had recently experienced a PFO-associated stroke and had an associated ASA or large interatrial shunt. Patients were followed for a mean of 5.3 years. In the PFO closure group (238 patients), no strokes occurred, while in the medication-only group (235 patients), there were 14 strokes (HR 0.03; 95% CI: 0–0.26; P<0.001). A systematic review and meta-analysis of these RCTs [25], which included a total of 3,627 patients with a mean follow-up of 3.7 years, showed a significant difference in ischemic stroke recurrence between the PFO closure group and the medication-only group (0.53 vs. 1.1 per 100 patient-years, respectively; OR 0.43; 95% CI: 0.21–0.90; relative risk reduction, 50.5%; absolute risk reduction, 2.11%; and number needed to treat to prevent 1 event, 46.5 for 3.7 years).

The trials discussed above have helped researchers identify clinical and anatomical characteristics that can predict the benefit of PFO closure [26]. As a result, the PFO-Associated Stroke Causal Likelihood (PASCAL) risk stratification system has been developed [27]. This system classifies patients into three groups based on two factors: high-risk PFO features (presence of a large shunt, ASA, or both) and RoPE scores (dichotomized into two ranges, 7 to 10 vs. 0 to 6). The three groups are as follows: PROBABLE group (high-risk PFO and RoPE score ≥7), POSSIBLE group (either high-risk PFO and RoPE score <7 or low-risk PFO and RoPE score ≥7), and UNLIKELY group (low-risk PFO and RoPE score <7). Following PFO closure, the relative reduction rates in recurrent ischemic stroke were 90% in the PROBABLE group and 62% in the POSSIBLE group. In the UNLIKELY group, there was no significant difference between PFO closure and medical therapy alone.

Various complications after the procedure, such as aortic erosion, device thrombosis, embolization, endocarditis, and atrio-aortic fistula, have been reported, though they are rare. However, data shows that closing a PFO is associated with a 5-fold higher risk of developing AF compared to medical treatment [28]. Data from randomized controlled trials indicate that up to 6.6% of stroke patients who undergo PFO closure develop AF in the weeks following the procedure, with the majority (>80%) experiencing it within 45 days. Furthermore, the risk of AF after PFO closure increases with age [29].

In a retrospective study conducted at a single medical center, 35 out of 761 patients who underwent PFO closure were monitored with implantable loop recorders for an average of 12 months [30]. Among these patients, 13 (37%) developed AF during the follow-up period. Notably, in 12 patients, AF was first detected within the first 4 weeks of monitoring, resolved within 3 months (spontaneously in 7 patients), and did not recur during the rest of the follow-up. Consequently, ongoing studies are now examining prolonged cardiac monitoring after PFO closure and the implications of AF occurrence on the risk of recurrent stroke. For example, the PFO-AF (Patent Foramen Ovale–Atrial Fibrillation) study [31] is a prospective, multicenter, observational study involving 250 stroke patients who require PFO closure. These patients will have a Reveal Linq device (Medtronic) implanted 2 months prior to percutaneous PFO closure. Follow-up data will include electrocardiograms (ECGs) and analysis of implanted cardiac monitoring data at 2, 12, and 24 months after the closure. The primary aim is to identify the occurrence of AF during the first 2 months of follow-up.

Despite the promising results of PFO closure methods, there is a lack of evidence regarding its effectiveness in preventing recurrent ischemic stroke in elderly patients (≥60 years) with PFOs and ESUS. This is because current guidelines recommend device closure only for patients under 60 years old [32], as most randomized controlled trials have excluded patients over this age limit [18-20,22-24]. However, it is noteworthy that although these guidelines impose age limitations, there have been no reports of inefficacy or increased adverse effects of PFO closure in older patients. Notably, previous studies excluded older patients, in part because they are more likely to have other undetected causes such as covert AF, hidden malignancy, or aortic atheroma, compared to younger patients. The presence of the RoPE score has also led to some confusion, with some arguing that older patients should be excluded from PFO closure due to low RoPE scores. However, this approach does not make sense, as the RoPE score was primarily developed to estimate the causal relationship between PFO and ESUS in young patients, not to predict the risk of recurrent strokes. In fact, studies have shown that older patients with PFOs are more likely to develop future strokes compared to younger patients [33]. Additionally, older patients are more likely to develop venous thrombosis and pulmonary hypertension, which may increase the risk of PFO-related stroke.

Only one RCT, the DEFENSE-PFO (Device Closure Versus Medical Therapy for Cryptogenic Stroke Patients With High-Risk Patent Foramen Ovale) trial [34], did not impose an age limit of <60 years. In this study conducted in South Korea, the authors enrolled patients aged 18–80 years, with a mean age of 52. A recent sub-analysis from the DEFENSE-PFO trial compared the efficacy of PFO closure versus medical therapy alone in patients ≥60 years old [35]. Of the 21 ESUS patients receiving medical therapy alone, four experienced a recurrent ischemic stroke or TIA (primary endpoint), whereas none of the 13 patients who underwent PFO closure experienced recurrent events (Cox proportional HR for medication arm, 7.36 [95% CI: 0.28–195.8]; P=0.23). Furthermore, all recurrent strokes in the medical arm exhibited an embolic pattern on brain MRI, further suggesting a potentially beneficial role of PFO closure in the older population.

Notably, the Korean healthcare policy does not employ age limitations in the reimbursement for device closure of PFOs. This allowed us to perform another analysis. We enrolled patients aged ≤60 years who had a cryptogenic stroke and PFO from 10 hospitals in South Korea [36]. We assessed the effect of PFO closure over medical therapy alone using a propensity-score matching method in the overall cohort and in those with a high-risk PFO (large shunt and/or ASA). Out of the 437 patients (mean age, 68.1 years), 303 (69%) had a high-risk PFO, and 161 (37%) underwent PFO closure. In addition, over a median follow-up of 3.9 years, 64 (14.6%) patients developed recurrent ischemic stroke or TIA. Furthermore, in the propensity score-matched cohort of overall patients (130 pairs), PFO closure was associated with a significantly lower risk of a composite of ischemic stroke or TIA (HR 0.45; 95% CI: 0.24–0.84; P=0.012). Moreover, in a subgroup analysis of patients with high-risk PFO features (116 pairs), PFO closure was significantly associated with lower risks of both the composite of ischemic stroke or TIA (HR 0.40; 95% CI: 0.21–0.77; P=0.006) and ischemic stroke (HR 0.47; 95% CI: 0.23–0.95; P=0.035). These results suggest that PFO closure might be equally effective in elderly patients as it is in younger subjects with cryptogenic stroke and PFOs, especially for those with high-risk PFOs.

As discussed above, older patients are more likely to present with hidden AF and other risk factors. Moreover, old age is a risk factor for developing supraventricular arrhythmias after PFO closure. Therefore, therapeutic decisions should be made carefully. Based on the lack of robust data from clinical trials, current clinical guidelines still do not recommend PFO closure for older patients and instead recommend intense cardiac monitoring for at least 6 months to rule out subclinical AF [28]. Overall, further RCTs are needed to assess the efficacy and safety of PFO closure versus medical therapy in older (age ≥60 years) patients with ESUS, and such trials are ongoing [37].

Although percutaneous PFO closure has been established as an effective therapy for patients with PFOs and ESUS, there are still many questions that need to be answered before a therapeutic decision can be made. In order to address these questions, it is necessary to establish a “heart-brain team” consisting of various clinical specialties, including neurologists, cardiologists, and hematologists if necessary. It is also important to involve the patient closely in the decision-making process. This team can provide a collaborative platform to facilitate a multidisciplinary approach to decision making. It will also create opportunities for education and future research on PFO-associated strokes. Some common clinical debates that arise during these meetings include: (1) treatment for older ESUS patients (60 years and older) with high-risk PFOs; (2) treatment for patients with high-risk PFOs but with another identified cause; (3) treatment for patients with high-risk PFOs and an uncertain diagnosis of TIA; (4) treatment for ESUS patients with PFOs but without high-risk features; and (5) decisions regarding the type and duration of antithrombotic therapy following PFO closure, as well as treatment for patients with severe migraines and high-risk PFOs [38].

It has been reported that atrial cardiopathy may be the possible cause of embolism in patients without AF or PFOs. Various methods have been used to assess atrial cardiopathy [39], including electrocardiographic markers such as P-wave terminal force velocity in V1 (PTFV1) [40], imaging markers such as left atrial volume index (LAVI) (mL/m²) [41], atrial fibrosis [40], LAE [42], and serum biomarkers such as N-terminal probrain natriuretic peptide (NT-proBNP) [43].

One of the most important markers for cardiopathy is LAE. The Framingham Heart Study [44] and Olmsted County cohort study [45] showed that left atrial size was a significant predictor of stroke after adjustment for AF. In those without documented AF at baseline, left atrial volume was independently associated with a composite outcome of major cardiovascular events, including stroke. However, this association was not detected in subjects with AF [45]. More recently, Edwards et al. [46] conducted a retrospective longitudinal cohort study involving consecutive community-dwelling adults. Subjects with a history of AF or anticoagulation use were excluded. They were followed for up to 5 years. LAE was measured as the antero-posterior linear left atrial diameter (mm) on the baseline 2D echocardiogram [47]. The results showed that after adjustment for age, sex, and various vascular risk factors, each 10 mm increase in left atrial diameter increased the incidence of ischemic stroke by about 2-fold (2-year HR 1.72, 95% CI: 1.16–2.55, P=0.007; 5-year HR 1.87, 95% CI: 1.41–2.49, P<0.0001). A >2-fold increase in the hazard of incident AF was also observed. However, excessive atrial ectopy, measured as the frequency of atrial premature beats per hour on baseline Holter monitoring, did not show significant associations with stroke risk in this study. Meanwhile, a systematic review [48] of nine cohorts analyzing 67,875 participants and 3,093 stroke outcomes also showed that LAE was significantly associated with increased stroke risk in patients with sinus rhythm.

The potential mechanisms that cause stroke in patients with LAE can be explained as follows. First, in an enlarged atrium, there is an increased risk of thrombogenesis due to blood stasis. While appropriate studies have not clearly documented this, there was a report showing that the volume of the left atrium was correlated with the severity of the atrial appendage thrombus in patients with dilated cardiomyopathy [49]. However, LAE is closely associated with other health conditions such as hypertension, diabetes, obesity, and sleep apnea [50]. Overall, although these risk factors were taken into account in previous studies, it is possible that the increased risk of stroke in patients with LAE can partly be attributed to these or other unidentified vascular risk factors. Second, LAE is closely associated with left ventricular (LV) dysfunction, which may also be a source of embolic stroke. Therefore, LAE may simply serve as an indicator of the severity of LV dysfunction [51,52]. Finally, considering that left atrial remodeling and AF are interconnected, the elevated stroke risk in patients with LAE might be partially attributed to the subsequent development of AF. This development may not be easily detected by physicians [53,54]. Supporting this notion, a recent study of consecutive ESUS patients [55] found that 42% had LAE, defined as a LAVI >34 mL/m². This study revealed that LAE was significantly associated with old age, hypertension, supracardiac atherosclerosis, and later detection of AF. However, there was no significant association between LAE and stroke recurrence. This result supports the idea that the recurrence of stroke in patients with LAE could be related to the development of AF.

Given the potential mechanisms of stroke in LAE, including blood stasis, LV dysfunction, or hidden AF, anticoagulation appears to be more effective than aspirin for stroke prevention in these patients. Based on this assumption, several RCTs have been conducted. One such RCT, the NAVIGATE ESUS (New Approach Rivaroxaban Inhibition of Factor Xa in a Global Trial Versus ASA to Prevent Embolism in Embolic Stroke of Undetermined Source) [56], compared the efficacy of rivaroxaban and aspirin in patients with ESUS and various heart conditions associated with a higher risk of AF. The primary outcome was the occurrence of ischemic stroke. A total of 7,112 patients with a mean (standard deviation, SD) age of 67 (9.8) years were enrolled and stratified based on different predictors of AF. The HAVOC score (hypertension, age, valvular heart disease, peripheral vascular disease, obesity, congestive heart failure, and coronary artery disease), which is determined by clinical patient characteristics and validated weights, was calculated. This score takes into account various factors: 4 points for a history of congestive heart failure, 2 points each for a history of coronary artery disease, age 75 or older, hypertension, and cardiac valvular disease, and 1 point each for obesity and peripheral vascular disease. The mean HAVOC score was found to be 2.6 (SD, 1.8). Additionally, the mean left atrial diameter was measured to be 3.8 (SD, 1.4) cm, and the median daily frequency of premature atrial contractions was recorded as 48. The detection of AF during the study’s follow-up increased with each HAVOC score tertile (2.3% for scores 0–2, 3.0% for scores of 3, and 5.8% for scores >3). However, neither HAVOC score tertiles nor the frequency of premature atrial contractions had an impact on the association between rivaroxaban and recurrent ischemic stroke (P for interaction=0.67 and 0.96, respectively). The annual incidence of AF also rose with larger left atrial diameters (2.0%, 3.6%, and 5.2%) and tertiles of premature atrial contractions (1.3%, 2.9%, and 7.0%).

In the aforementioned study, among the predefined subgroup of patients with a left atrial diameter greater than 4.6 cm (9% of the overall population), the risk of ischemic stroke was lower in the rivaroxaban group (1.7% per year) than in the aspirin group (6.5% per year) (HR 0.26; 95% CI: 0.07–0.94; P for interaction=0.02). These results suggest that the HAVOC score, left atrial diameter, and premature atrial contraction frequency can predict subsequent clinical AF. Additionally, rivaroxaban has been found to be more effective than aspirin in reducing the risk of recurrent stroke specifically in patients with moderate or severe LAE. Therefore, while NOACs may not be superior to antiplatelets in preventing future strokes in ESUS patients overall, they may be more effective in certain patients with severe atrial cardiopathy.

The ATTICUS (Apixaban for Treatment of Embolic Stroke of Undetermined Source) trial [57] was a study that focused on ESUS patients with risk profiles for cardiac thromboembolism. These risk factors included left atrium (LA) sizes greater than 45 mm, the presence of spontaneous echo contrast in the left atrial appendage, left atrial appendage flow velocities of 0.2 cm/s or less, atrial high-rate episodes, a CHA2DS2-VASc score of 4 or higher, or presence of a PFO. Participants in the study were randomly assigned to receive either aspirin (100 mg once daily) or apixaban (5 mg twice daily). Cardiac monitoring was compulsory, and if AF was detected, participants on aspirin were switched to apixaban. The primary outcome of interest was the occurrence of new ischemic lesions on brain MRI during the 12-month follow-up period. Secondary outcomes included major and clinically relevant non-major bleeding. A total of 353 patients were assigned to either the apixaban group (178 patients) or the aspirin group (175 patients) at a median of 8 days after experiencing an ESUS event. MRI follow-up data at 12 months were available for 325 of the participants (92.3%). Among the apixaban group, 23 of 169 participants (13.6%) experienced new ischemic lesions, while 25 of 156 participants (16.0%) in the aspirin group had these lesions (adjusted OR 0.79; 95% CI: 0.42–1.48; P=0.57). Major bleeding occurred in 5 participants, and 7 participants experienced clinically relevant non-major bleeding (HR 0.68; 95% CI: 0.22–2.16). The study was terminated prematurely following a pre-specified interim analysis due to a lack of promising results.

The ARCADIA (Atrial Cardiopathy and Antithrombotic Drugs in Prevention After Cryptogenic Stroke) trial [58] was a multicenter RCT comparing the efficacy of apixaban to aspirin in patients with recent ESUS and arterial cardiopathy. The trial enrolled patients from 185 sites in the Canadian Stroke Consortium and StrokeNet in the United States. Patients who had experienced ESUS within the past 6 months and had evidence of atrial cardiopathy were included in the study. The patients were randomly assigned to receive either apixaban 5 mg twice daily or aspirin 81 mg daily. If AF was detected after randomization, patients were started on anticoagulant therapy at the discretion of their treating physicians. In this study, atrial cardiopathy was defined as having at least one of the following markers: P-wave terminal force >5,000 μV × ms in ECG lead V1, serum NT-proBNP >250 pg/mL, and left atrial diameter index ≥3 cm/m² on echocardiography [58]. The primary efficacy outcome, analyzed using a time-to-event analysis, was recurrent stroke [59]. All participants, including those who were diagnosed with AF after randomization, were analyzed according to their assigned treatment groups. Symptomatic intracranial hemorrhage and other major hemorrhages were the primary safety outcomes of interest. A total of 1,100 participants were enrolled, and the mean follow-up period was 1.8 years. The trial was stopped prematurely due to futility, based on planned interim analysis. Recurrent strokes occurred in 40 patients in both the apixaban group (annualized rate, 4.4%) and the aspirin group (annualized rate, 4.4%) (HR 1.00; 95% CI: 0.64–1.55). Symptomatic intracranial hemorrhage occurred in 0 patients receiving apixaban and 7 patients receiving aspirin (annualized rate, 1.1%). In conclusion, the study found that in patients with cryptogenic stroke and evidence of atrial cardiopathy, apixaban did not show a significant reduction in recurrent stroke risk compared to aspirin.

Recent RCTs have generally failed to demonstrate the benefit of NOACs over antiplatelet therapy. Why is this the case? Firstly, in the ARCADIA trial, the presence of atrial cardiopathy was determined using three readily available biomarkers, based on previous studies that showed an association between these markers (P-wave terminal force in ECG lead V1 [60], serum NT-proBNP [61], and left atrial diameter [44]) and the risk of ischemic stroke. However, additional biomarkers of atrial cardiopathy exist, such as midregional proatrial natriuretic peptides, premature atrial contractions, left atrial fibrosis, left atrial volume, or functional left atrial strain [62]. Therefore, it appears that the markers chosen by the ARCADIA team may not be the most ideal ones to establish a clear link between atrial cardiopathy and stroke. In other words, the association may have been confounded by other markers. Second, another confounding factor is the presence of atherosclerosis, which may play a role as a source of embolism. According to the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria, a diagnosis of stroke secondary to large artery disease can be made when there is a significant (>50%) arterial stenosis. These criteria were established when magnetic resonance angiography was unavailable. However, it has now been recognized that, in addition to the degree of stenosis, the characteristics of arterial pathology are important in predicting the generation of embolism. These characteristics include ulcerative, irregularly shaped walls and intraplaque hemorrhage. As a result, arteries with mild stenosis or even a normal appearance may still result in embolism.

Among patients with ESUS, most recurrent strokes occur in the same cerebral arterial territory as the initial stroke [63]. This finding suggests the possibility of an arterial rather than cardiac embolic source. A meta-analysis of 18,304 patients reported that 51% had non-stenotic carotid plaques in patients with acute ischemic stroke/TIA, and the prevalence was even higher (55%) among patients with ESUS [64]. In the INTERRSeCT (Identifying New Approaches to Optimize Thrombus Characterization for Predicting Early Recanalization and Reperfusion With IV Alteplase and Other Treatments Using Serial CT Angiography) study, non-stenotic carotid plaques were present in 39.1% (54/138) of patients with ESUS. These plaques were significantly more common on the side ipsilateral to the ischemic stroke compared to the contralateral side (60.6% vs. 29.2%, P=0.004) [65]. Another meta-analysis of 323 patients reported that patients with ESUS had a higher prevalence of non-stenotic carotid atherosclerosis with high-risk plaque features in the ipsilateral carotid artery compared to the contralateral artery (32.5% vs. 4.6%, 95% CI: 0.1–13.1) [66]. A recent review [67] also found that vessel wall MRI findings of vulnerable plaque, such as intraplaque hemorrhage, lipid-rich necrotic cores, fibrous cap disruption, and surface ulceration, were more common in the ipsilateral carotid artery than in the contralateral artery. Furthermore, another study that used high-resolution MRI in 243 ESUS patients concluded that among patients without ipsilateral intracranial atherosclerotic plaques, there was an association between the presence of atrial cardiopathy and ESUS (adjusted OR: 4.76, 95% CI: 2.48–9.14). However, this association was not evident among patients with intracranial atherosclerosis [68]. These pieces of evidence support the hypothesis that embolization from large artery atherosclerosis may occur more frequently than previously realized in patients with ESUS. Considering that antiplatelets are regarded as a more suitable treatment option than anticoagulants in patients with large artery disease, the observed lack of benefit from anticoagulation compared to aspirin in the aforementioned trials may be partly due to this preference.

It has been reported that patients with ischemic stroke who exhibit LV diastolic dysfunction experience poor functional outcomes and an increased incidence of vascular events within a year [69]. This impact is believed to be associated with impaired left atrial contractility and intracardiac stasis, ultimately leading to thromboembolism [69,70]. Furthermore, it is well-established that a severely depressed LV ejection fraction (≤35%) is linked to a higher risk of stroke [71]. However, few studies have investigated whether a modest reduction in LV ejection fraction increases stroke risk. Limited data suggests that even a modest reduction in ejection fraction may result in intracardiac thrombus formation and an increased risk of embolic stroke [72]. One study indicated that LV ejection fraction does not differ between patients with ESUS and those with strokes caused by small or large artery diseases. However, a positive association between lower ejection fraction and ESUS was noted when patients with any evidence of ipsilateral carotid atherosclerosis were excluded [73]. Therefore, ESUS likely represents a diverse group of patients, including those with cardiac embolism and those with non-stenosing atherosclerosis that do not meet the TOAST definition of large artery disease. Additionally, it remains uncertain whether LV wall motion abnormalities, in the absence of recent myocardial infarction, are associated with an increased risk of stroke. One previous study found that LV wall motion abnormality was independently associated with an elevated risk of recurrent stroke [74], whereas another study could not find such associations in patients with ESUS [73].

It is possible that anticoagulation may be more effective than antiplatelet therapy in treating patients with stroke associated with LV dysfunction. A sub-study of the NAVIGATE ESUS [75] study investigated this issue using criteria for LV dysfunction: (1) moderate to severely impaired global LV contractility (ejection fraction ≤40%); and/or (2) any degree or type of wall motion abnormality. The authors found that only 7.1% of ESUS patients met the criteria for LV dysfunction, which may be explained by the exclusion of patients with a history of congestive heart failure in this study. The results showed that patients with LV dysfunction were more likely to be male and have coronary artery disease, as well as a history of stroke/TIA, compared to the overall NAVIGATE ESUS cohort. Additionally, it was discovered that individuals with LV dysfunction had significantly lower rates of stroke recurrence with rivaroxaban (2.4% per year) compared to those treated with aspirin (6.5% per year) (HR 0.36; 95% CI: 0.14–0.93). These findings are consistent with previous studies showing the benefits of anticoagulation over aspirin in patients with reduced ejection fraction. However, further research is needed to determine the optimal medical management strategy for ESUS patients with LV dysfunction.

This review demonstrates that paradoxical embolism associated with PFO is one of the primary causes in ESUS patients. Therefore, it is crucial not to overlook diagnostic investigations such as TCD and echocardiography. The optimal medication for treating these patients remains uncertain, and stroke patients with PFO can be managed with either antiplatelets or anticoagulants, preferably NOACs. Recent studies have indicated that percutaneous closure of PFO is more effective in reducing the recurrence of strokes compared to medical therapy alone in patients with PFO-related stroke. This is particularly true for patients who have high-risk features such as a large shunt or the presence of ASA. However, it is still necessary to study the effectiveness of PFO closure in older patients (≥60 years), and the decision should be made through discussions involving neurologists, cardiologists, and patients.

In ESUS patients without PFO, the cardiopathy itself may be the source of embolism, with LAE being the most significant cause. Although it was previously assumed that NOACs would be an effective therapy, recent RCTs have failed to demonstrate the superiority of NOACs over antiplatelets in preventing further strokes in these patients. This lack of superiority could be attributed to the unclear definition of emboligenic cardiopathy and the presence of other potential embolic sources, such as mild yet emboligenic arterial diseases. Overall, further research is needed to determine an effective management strategy for ESUS patients.