Device closure for patent foramen ovale in patients with cryptogenic stroke: a paradigm in evidence

Recent randomised trials will alter management of patent foramen ovale in patients with cryptogenic stroke

Cryptogenic stroke is a devastating event in young patients. Patent foramen ovale (PFO) may be related to this type of stroke in certain cases and can be closed with a percutaneous device. However, due to perceived lack of benefit in early trials, this procedure has not been routinely performed in the past. Guidelines from the American Heart Association and the American Stroke Association published in 2014 state that “for patients with a cryptogenic [ischaemic] stroke or [transient ischaemic attack] and a PFO without evidence for [deep vein thrombosis], available data do not support a benefit for PFO closure”.1 But the results of recent trials and newly revised guidelines recommendations will see a change in landscape.

The foramen ovale is a hole that exists in the wall between the left and right atria of every human fetus. It enables oxygenated blood to flow from the right to the left atrium for fetal development and normally closes during infancy, but it does not close in about 25% of the general population.2 Cryptogenic strokes are now thought to comprise about one‐third of all ischaemic strokes,3 in which the presence of a PFO with paradoxical embolism is one cause, particularly in young patients. Treatment includes medical management (with antiplatelet therapy or anticoagulant therapy) or percutaneous device closure.

Previous randomised controlled trials failed to demonstrate superiority of device closure versus medical therapy (antiplatelet or anticoagulant).4,5,6 A potential reason for lack of benefit was the inclusion of patients with any PFO without specifying high risk anatomical features. Hypermobility of atrial septum (atrial septal aneurysm) and PFO size are now known predictors of recurrent stroke.7 Subgroup analyses of one trial confirmed that PFO closure significantly reduced recurrent stroke when such features were present.8 Another reason was that the use of a closure device in one trial was associated with more frequent complications and lower procedural success compared with other closure devices, and is no longer available.4 Furthermore, consent for randomisation was challenging across trials. Physicians and patients may be biased in which treatment they consider superior, and patients may opt for device closure through off‐label indication rather than trial enrolment. This hampers recruitment and the likelihood of finding significant between‐group differences when event rates are low. These limitations paved the need to re‐address this clinical question, and more recent trials have shown definitive benefit of device closure in select patients with PFO‐related stroke.8,9,10

Two randomised trials (CLOSE and Gore REDUCE) and one extended follow‐up trial (RESPECT), published in 2017, demonstrated a reduction in recurrent stroke following device closure versus antiplatelet therapy.8,9,10 The overall hazard ratio of recurrent stroke in a meta‐analysis was 0.32 (95% CI, 0.13–0.82; P = 0.02; annualised weighted event rate 0.61% and 1.17%, respectively).11 The strengths of these trials include cohort selection and extended follow‐up duration. In the CLOSE trial, only patients with high risk features on echocardiography, such as large right‐to‐left shunt or atrial septal aneurysm, were included.9 In the REDUCE trial, other causes of stroke such as large artery atherosclerotic disease and small vessel lacunar infarctions identified on cerebrovascular imaging were strictly excluded.10 Across the studies, the median follow‐up ranged between 4 and 6 years, with extensive enrolment periods (8–9 years) needed to sufficiently power these large trials (663–980 patients) for what is a relatively uncommon pathology, particularly given that recurrent events are infrequent. Overall, these recent trials with stringent inclusion criteria addressed limitations of previous trials, and the findings support investigating for PFO in the setting of cryptogenic left circulation thromboembolism.

Comprehensive evaluation for potential causes of cryptogenic stroke individualises management. Multidisciplinary involvement is key, including a stroke unit and a structural heart team to identify the most appropriate patients.2 Exclusion of other stroke causes include haematological testing for hypercoagulable states, cardiac monitoring for atrial fibrillation and vascular studies, such as carotid ultrasound.2 A contrast transthoracic echocardiography with an agitated saline bubble study is the initial step in suspected PFO‐related stroke.2 This test should be followed by transoesophageal echocardiography to confirm the PFO, assess shunt size and evaluate interatrial septum anatomy. A potential clinical pathway for investigating cryptogenic stroke is depicted in the Box.

For patients proceeding with PFO device closure in Australia, only select devices are approved by the Therapeutic Goods Administration. The current approved indication is for patients with stroke or transient ischaemic attack and PFO diagnosed by echocardiography, with right‐to‐left shunting during Valsalva manoeuvre. Trends of PFO device closure declined between 2008 and 2014, as major guidelines recommended against the procedure due to perceived lack of efficacy.1,12 However, recent international guidelines and expert consensus, revised following the results from these latest trials,8,9,10 suggest device closure after multidisciplinary consensus in adults aged less than 60 years with cryptogenic stroke and high risk anatomical features of PFO.2,13 It is now anticipated that this trend will reverse with the release of further supportive recommendations from society guidelines and the growing number of referrals from neurologists. In patients who proceed with device closure, technical success is high (~ 99%) and is followed by dual antiplatelet therapy for 3–6 months.9,10 Complications of device closure include atrial tachyarrhythmias (5–7%), venous vascular access complications (2%), device thrombosis (< 1%), and device embolisation (< 1%).8,9,10 Procedural long term safety (> 10 years) is promising, with a recurrent ischaemic stroke event rate of 1% and a major bleeding event rate of 2%.14

Nevertheless, percutaneous device closure has limitations. Despite reducing recurrent stroke, there is no evidence that device closure has an impact on overall survival, as no trial has been powered to detect this endpoint alone.4,5,6,8,9,10 It may be argued, however, that stroke with significant neurological deficit may be a worse outcome. Moreover, the number needed to treat to prevent one recurrent ischaemic stroke per year is 96.11 This appears to be a large number, but it does not reflect potential disease‐free life years achieved from a single procedure, particularly in young patients. As trials only included patients with stroke (> 24 hours of symptoms or neuroimaging diagnosis), it is uncertain whether patients with transient ischaemic attack would benefit. Moreover, among the latest trials, only one study directly compared device closure with anticoagulant therapy, which could be considered standard of care over antiplatelet therapy, and found no significant difference.8 However, in view of an increased risk of major bleeding and ongoing, periodic testing while receiving anticoagulation, experts still recommend device closure over anticoagulation therapy, although both options should be discussed with each patient.13 Current approved indications for device closure in PFO are stringent; thus, case load, operator and institution experience may affect procedural outcomes. Device closure may not be indicated when PFO does not demonstrate high risk anatomical features, and the case should be discussed in a multidisciplinary context balanced with patient preference.

Ultimately, these are challenging trials to conduct, but the clinical impact is apparent. Stroke with residual neurological deficit is catastrophic, particularly for young, high functioning and otherwise healthy individuals. PFO device closure has existed for several decades, and the recent randomised data finally establish the procedure as a reliable and safe secondary preventive measure in select candidates with cryptogenic left circulation thromboembolism. We look forward to further data that clarify the relative efficacy of available closure devices and how we can better risk‐stratify patients with PFO.

Box – Potential clinical pathway for investigating cryptogenic stroke with patent foramen ovale

---

CT = computed tomography; ECG = electrocardiogram; MRI = magnetic resonance imaging; PFO = patent foramen ovale; TOE = transoesophageal echocardiography; TTE = transthoracic echocardiography. * Identifiable mechanisms: large artery atherosclerotic disease, lacunar stroke syndrome, hypercoagulable disorder, arterial dissection. † Risk factors: congestive heart failure, diabetes mellitus, structural heart abnormalities (left atrial enlargement, left ventricular hypertrophy), hypertension.