Transcatheter closure of a patent foramen ovale (PFO) is an established secondary stroke prevention therapy for patients with an embolic stroke of undetermined source (ESUS). The safety and clinical benefits of this approach have been demonstrated in several randomized controlled trials (RCTs) [1-4]. Patients who experience the greatest benefit are those with a high likelihood of the PFO playing a causal role in the stroke etiology, as shown in a meta-analysis of individual patient data [5]. As outlined by the PFO-Associated Stroke Causal Likelihood (PASCAL) Classification System [5], these include patients with high-risk PFO features, such as a large right-to-left shunt (RLS) and an atrial septal aneurysm (ASA). While mounting evidence supports PFO closure in predominantly Caucasian populations, limited data exist for Asian-Pacific cohorts. Consequently, in 2020, a group of Asian-Pacific clinical experts published an expert opinion and consensus statement for the region [6]. This same group also addressed the need to establish a Heart and Brain Team for managing stroke patients with PFO [7].

Several factors justify the need for updated expert opinions and guidance on PFO closure specifically for the Asia-Pacific region. Although no clear differences in PFO morphology or prevalence between Asians and Caucasians have been identified, it is likely that variations exist in individual physiological characteristics and healthcare delivery systems. For instance, elderly Asian patients may be more prone to hemorrhagic complications and, thus, less likely to benefit from anti-thrombotics compared to other ethnicities [8]. Consequently, PFO closure may offer greater benefits for older Asian patients with PFO-related stroke, who otherwise would require life-long anti-thrombotic therapy. Since the publication of the 2020 Asian-Pacific expert opinion, new evidence has emerged, offering further guidance on several important topics. Recent findings from elderly ESUS cohorts [9-18] support the role of PFO closure in elderly patients, two of which came from an Asian country [17,18]. This calls for a reassessment of PFO closure in patients over 60 years, who were excluded from most RCTs [1-3] that provided the initial evidence for PFO closure.

Additionally, contrast transcranial Doppler ultrasound (cTCD), if appropriately performed, may be a cost-effective tool in Asian countries with limited medical resources. Finally, while PFO closure has shown favorable and reproducible outcomes in ESUS, its role in treating intractable migraine remains debated [19-21]. Given the interest shown by some physicians during the symposium, we believe a discussion on PFO closure in migraine patients is warranted.

This focused update of the expert opinion on PFO closure for the Asia-Pacific region was developed by a group of clinical experts from the region, including neurologists and cardiologists experienced in stroke treatment and catheter-based cardiac interventions. The objective of this update was to review recently published evidence on specific topics related to PFO closure, including its role in elderly and migraine patients, as well as the appropriate use of cTCD as a screening tool for ESUS patients. The evidence was reviewed and discussed during the Asia-Pacific Heart-Brain Summit (October 3, 2024, Bangkok, Thailand). Clinical experts from Japan, China, Hong Kong, Taiwan, Malaysia, Singapore, India, and South Korea (see Acknowledgments) attended the meeting and contributed to the discussions that led to the expert opinions presented in this manuscript. A voting system was used to determine the prevailing opinions among the experts. A writing group (see authors’ list) was formed, consisting of the experts attending the meeting and involved in the discussions. This group was tasked with preparing a manuscript that reflects the leading opinions derived from the meeting.

Evidence for PFO closure in ESUS patients primarily comes from RCTs that enrolled patients younger than 60 years of age. As a result, evidence-based guidelines offer limited guidance on PFO closure in patients aged 60 years or older [22]. While the European Stroke Organization (ESO) [23] guidelines do not provide a specific recommendation for PFO closure in patients over 60, they include a consensus statement advocating for the inclusion of older adults in RCTs, or at least in registries.

The rationale for considering PFO closure in patients older than 60 is supported by several clinical observations. Evidence shows an age-related increase in the risk of recurrent stroke in patients with a PFO who have had a stroke or transient ischemic attack (TIA) of unknown cause [24]. This may be partly due to a higher prevalence of PFO characteristics that increase the risk of paradoxical embolism, such as a large PFO-mediated RLS and the presence of an ASA, in elderly compared to younger patients [9,10]. Additionally, advanced age is associated with a higher risk of venous thromboembolism [25,26] and changes in the structure and function of the right ventricle [27], both of which may elevate the risk of paradoxical embolism [11].

Recent non-randomized studies further support the consideration of PFO closure in patients over 60 years of age (Supplementary Material 1). These studies report similar rates of procedural complications in elderly and younger patients, including device-related adverse events and paroxysmal episodes of periprocedural atrial fibrillation (AF) [9,10,13,14,16]. Some studies found a higher incidence of new-onset AF at follow-up in elderly compared to younger patients [9,14], which may reflect increased atrial vulnerability in older individuals. However, like younger patients, many new-onset AF episodes occurred within the first month after the procedure [12]. Additionally, older adults have a higher risk of transesophageal echocardiography (TEE)-related complications [28].

From non-age-restricted cohorts undergoing PFO closure, favorable long-term outcomes have been reported in patients over 60 [9,10,12,14,16]. While some studies observed a higher incidence of thromboembolic events in older adults, the rates were lower than those predicted by their Risk of Paradoxical Embolism (RoPE) score [9,10]. A higher mortality in older patients was mainly attributed to age itself [14] with deaths typically non-cardiovascular [10,12]. Studies comparing treatments in elderly populations showed that PFO closure was associated with a significantly lower incidence of recurrent stroke and/or TIA compared to medical therapy [13,15,17]. In a post hoc analysis of the Device Closure Versus Medical Therapy for Cryptogenic Stroke Patients With High-Risk Patent Foramen Ovale (DEFENSE-PFO) randomized controlled trial, the additional benefit of PFO closure over medical therapy was greater in patients over 60 than in younger patients [18].

The RoPE score is a validated tool used to assess the likelihood of paradoxical embolism as a potential stroke cause [29], with a higher score indicating a stronger causal role of the PFO in stroke etiology. However, it may be less suitable for considering PFO closure in older adults. Due to the increased likelihood of other stroke risk factors in elderly individuals, advanced age typically results in a lower RoPE score. Additionally, the RoPE score does not factor in high-risk anatomic PFO features, which are more common in older patients [9,10] and are associated with a greater benefit of PFO closure over medical therapy [6,17]. The PASCAL Classification System incorporates both the RoPE score and high-risk anatomic features [5]. While older adults with a stroke of undetermined source may not be considered for PFO closure based on their RoPE score alone, the PASCAL Classification System may identify them as potential candidates due to high-risk PFO features. A Japanese prospective multicenter study, including 500 patients undergoing PFO closure (30% >60 years of age), further supports this [30]. While 57.5% of patients had a RoPE score <7, 85.5% had at least one high-risk PFO feature, and 92.1% had a “Probable” or “Possible” PASCAL classification.

Guidelines and consensus documents [31-34] recommend excluding subclinical AF before PFO closure. The prevalence of AF in Asia-Pacific countries ranges from 0.5%-5.4%, with an increased prevalence of 4.6%-8.2% in individuals over 70 [35]. These rates are generally lower than those in many Western countries. However, the true incidence of AF may only be detected through continuous monitoring using an insertable cardiac monitor (ICM). Studies of ESUS populations with varying monitoring durations report ICM-detected AF rates of 12% to 41% [36-41]. Given their older age and higher likelihood of AF risk factors, patients over 60 should be considered for intensive monitoring (e.g., long-term ICM monitoring) to exclude AF [33]. Older patients with minimal AF risk may undergo less intensive monitoring. Risk scores predicting AF likelihood in ESUS patients [36,40,42] may be useful for selecting high-risk patients for long-term ICM monitoring.

An expert opinion regarding PFO closure in older adults is provided in Table 1.

cTCD with agitated saline (bubble contrast) is a noninvasive diagnostic procedure used to detect an RLS [43]. Meta-analyses have shown that cTCD has high sensitivity (91%-97%) for detecting an intracardiac shunt when compared with TEE [44,45]. cTCD is also more sensitive than transthoracic echocardiography (TTE) in detecting PFO-mediated RLS [45]. However, cTCD does not reveal the anatomic location of the RLS. It provides a positive result for any shunt from venous to arterial circulation, including intracardiac septal defects and pulmonary arteriovenous malformations. Therefore, cTCD may have slightly lower specificity for detecting a PFO compared to direct imaging of the left atrium [45].

The methodology of cTCD for detecting an RLS is well-documented in the literature [46]. In brief, a low-frequency pulsed Doppler probe (2.0 MHz) is secured to the scalp, typically using a commercially available headframe. The ultrasound probe is aimed at obtaining Doppler spectra from the middle cerebral artery through an appropriate temporal acoustic window. An agitated saline solution (typically 9 mL saline mixed with 1 mL air and a few drops of the patient’s blood [46]) is injected via a large-bore cannula, preferably placed in one of the antecubital veins. The solution is vigorously shaken through a tri-way connector. The addition of blood significantly increases the number of microbubbles, enhancing the contrast mixture’s effectiveness [47]. Approximately 4-6 seconds after injection of the saline-air mixture, the patient is instructed to perform a strong Valsalva maneuver, which reduces observed flow velocities. Continuous TCD monitoring is then conducted for 22 seconds, and the number of microbubbles is counted. In the case of an RLS, microbubbles will travel from the venous to the arterial circulation. Detection of a microbubble in the intracranial artery produces a distinct micro-embolic signal (MES), characterized by an audible high-intensity transient signal (HITS) and visualized on the TCD screen (Figure 1). A grading system based on the number of MES is used to determine the functional grade of RLS (Supplementary Material 2). Since RLS is position-dependent, it is recommended to perform cTCD evaluation twice: once in the supine position and once sitting upright [48].

In a study of 572 patients using cTCD to screen all patients with ischemic stroke or TIA, cTCD was feasible in 91% of the patients [49]. Reasons for failure included unsuitability for contrast injection due to pregnancy, patient intolerance, lack of a suitable temporal bone window, and cannulation issues. An RLS was found in 37% of patients with a cryptogenic event and 23% of patients with a known cause.

cTCD offers several advantages over TTE or TEE. It is relatively inexpensive, easy to perform as a bedside procedure, non-invasive, well-tolerated by patients, and has a low-risk profile. It is also easier for patients to perform a Valsalva maneuver during cTCD than during TEE. Therefore, cTCD is an ideal first-line screening tool in the setting of ESUS. If an RLS is revealed in a patient diagnosed with ESUS, TEE is recommended to confirm the presence of a PFO and its role in the RLS. Given its high sensitivity, cTCD should also be considered for detecting residual shunts after PFO closure [34].

An expert opinion regarding the use of cTCD in patients diagnosed with ESUS is provided in Table 2.

Although not fully explained, the PFO is suggested to play a role in the pathophysiology of migraine. Proposed mechanisms include the passage of micro-emboli, metabolites, and vasoactive substances through the PFO, as well as transient hypoxemia causing microinfarcts in the brain [50]. Nevertheless, the association between migraine and PFO appears to be variable and has been questioned by previous studies [21]. Given the mixed outcomes from studies on this topic, the debate regarding migraine as an exclusive indication for PFO closure remains unresolved.

Recent meta-analyses [50,51] on the effect of PFO closure for migraine symptoms provide a comprehensive overview of the current evidence. RCTs showed that PFO closure was associated with a significant reduction in the number of migraine attacks. However, complete resolution of migraine was only observed in observational studies, not RCTs. These findings provide insufficient support for PFO closure as a standalone treatment for migraine [21], though reducing migraine symptoms may be an additional benefit of PFO closure. This view is supported by recent reviews [19,20], emphasizing the need for further RCTs. Instead, monoclonal antibodies targeting the calcitonin gene-related peptide pathway, as recommended by guidelines, offer another effective pharmacological alternative [52,53].

An expert opinion on PFO closure for migraine is provided in Table 3.

Since the publication of the Asian-Pacific expert opinion on PFO closure in 2020, new clinical evidence has provided further guidance on several topics addressed in the expert opinion. These include growing support for considering PFO closure in selected patients over 60 years of age. Additionally, the need for further support in using cTCD as a screening tool to identify potential candidates for PFO closure has been emphasized. This method may be particularly beneficial in Asian-Pacific healthcare settings, aligning with current practices.

While the safety and efficacy of PFO closure in ESUS have been demonstrated in patients under 60, particularly those with high-risk PFO features [5], there are opportunities for broader indications and further optimization of the therapy. This expert opinion encourages physicians to consider PFO closure in ESUS patients older than 60 with high-risk PFO features, guided by the PASCAL Classification System [5] and clinical judgment in an individualized approach. Evidence supports the idea that, in addition to younger populations, cryptogenic embolic events in older adults are also associated with RLS [49] and favorable outcomes from PFO closure are expected [9,10,12-17].

Currently, two Asian studies are ongoing that include elderly ESUS patients with high-risk PFO features. The multicenter, prospective, observational Prospective Registry of Elderly ESUS With PFO (COACH_ESUS) registry study [54] compares outcomes in patients treated with PFO closure after 3-6 months of monitoring for paroxysmal AF versus patients receiving medical therapy. The Evaluation of Prevalence and Clinical Impact of Atrial Fibrillation in Elderly Patients With Cryptogenic Stroke and High-Risk Patent Foramen Ovale (DefenseElderly) study is a prospective cohort study comparing the incidence of paroxysmal AF in elderly ESUS patients [55]. Patients with a PFO and no AF during 6 months will be recommended for PFO closure and will be compared with a control group of ESUS patients without a PFO. Both studies may provide additional insights into the safety and efficacy of PFO closure in the elderly Asian population. Nevertheless, RCTs comparing PFO closure and medical therapy in elderly populations, particularly those with high-risk PFO features, are warranted.

While this expert opinion encourages broadening the indications for PFO closure in older adults, there are reservations about PFO closure as an exclusive indication for migraine patients. Despite the benefits experienced by migraine patients after PFO closure [56], there is currently no clear and reproducible evidence justifying PFO closure as a first-line treatment for these patients.

This expert opinion encourages the use of cTCD as a routine, first-line screening tool to detect RLS in all ESUS patients. This may expedite the diagnostic process and facilitate early identification of patients with a PFO-associated stroke. TCD has all the characteristics of a routine screening tool: it is easy to learn and apply, well tolerated by the patient, and most importantly, highly sensitive for detecting the pathology of interest.

Currently, insufficient evidence from RCTs is available to recommend an optimal antithrombotic therapy after PFO closure. Guidelines and consensus statements suggest dual antiplatelet therapy (DAPT) for several months after PFO closure, followed by long-term single antiplatelet therapy (SAPT) [23,34]. These recommendations are based on the antithrombotic regimen used during RCTs on PFO closure in ESUS patients. Nevertheless, a lower-intensity antithrombotic regimen may be considered specifically in Asian patients, who are known to have a higher bleeding risk and lower risk of thrombotic events compared to Caucasian patients [57,58]. Such an approach seems to be supported by results from a French/Canadian non-randomized multicenter study including 1,532 relatively young (<50 years) patients who underwent PFO closure. This study reported no difference between short-term DAPT followed by SAPT versus SAPT only in the 5-year rate of all-cause death, stroke, TIA, peripheral embolism, myocardial infarction, or BARC type ≥2 bleeding [59]. Outcomes from other studies suggest that discontinuation of antithrombotic therapy within 6-12 months after PFO closure may be a reasonable strategy in relatively young patients [60,61]. However, given the lack of strong evidence from RCTs, the decision to stop antithrombotic therapy should be the result of an individualized risk-based approach and left to the discretion of the treating neurologist. Whether a short-term antithrombotic strategy after PFO closure is safe for older Asian patients remains an unanswered question and requires further research.

Regarding the prevention of thrombus formation on the occlusion device, a study reported favorable results from a light antithrombotic regimen of 100 mg of aspirin for 6 months after closure of a PFO or an atrial septal defect [62]. After this period, endothelialization of the device is typically complete, and device-related thrombus is rarely observed. Further research on this topic is desired to develop an evidence-based recommendation. Major issues to be clarified include the need for DAPT versus SAPT during the first months after closure and the optimal duration of antithrombotic therapy in the long-term after PFO closure.

Like the initial expert opinion published in 2020, this update was the result of discussions among neurologists and cardiologists. It is important to emphasize that both clinical specialties should be involved in decision-making related to PFO closure in general, and more specifically in the topics addressed by this expert opinion, i.e., considerations regarding PFO closure in older adults or in patients with intractable migraine.

In summary, while mounting clinical evidence supports the safety and efficacy of PFO closure in ESUS patients <60 years of age, further research is warranted to confirm the benefits of PFO closure in older adults and to guide optimal antithrombotic therapy after PFO closure.