Atrial septal defect and PFO: closure indications

Overview and definitions

ESC 2020 opens its ASD section with a reminder that the defect can stay undiagnosed until adulthood.^[1] ESC 2020 names five types, and the 2025 ACC/AHA guideline classifies isolated ASDs by location with similar shares.^[1][2]

Types of atrial septal defect

^{[1] [2]}

The two guidelines differ only on the unroofed coronary sinus: ESC 2020 gives under 1%, ACC/AHA 2025 gives 1%.^[1][2] ESC 2020 lists associated lesions that include anomalous pulmonary venous connection, persistent left SVC, pulmonary valve stenosis and mitral valve prolapse. ACC/AHA 2025 calls partial anomalous pulmonary venous return the most common one.^[1][2]

^{[1] [2]}

Where ASD sits in the ACHD AP classification

The ACC/AHA anatomic and physiological (AP) classification assigns an anatomic complexity (I–III, simple to complex) and a physiological stage (A–D). Stages run from A (clinically well without residual anatomic or physiological sequelae) to D (for example, advanced heart failure or high-risk PAH).^[2] Its Table 4 places ostium secundum ASD among the simple lesions (class I), and partial or complete AVSD and sinus venosus defect among the lesions of moderate complexity (class II).^[2] Stage A requires all five: no cardiac symptoms, no haemodynamic or anatomic sequelae, no sustained arrhythmias, normal exercise capacity and normal pulmonary pressure.^[2] Stage C criteria include a haemodynamically significant shunt and a BNP or NT-proBNP at least twice the upper limit of normal. Stage D criteria include Eisenmenger syndrome and heart failure hospitalisation in the past 12 months.^[2]

Patent foramen ovale

SCAI 2022 describes the PFO as a vestigial congenital structure present in around 25% of adults. In most cases it is entirely benign and needs no treatment.^[8] The 2025 ACC/AHA ACHD guideline leaves anatomic variants such as the PFO outside its scope, so the PFO rows in this topic come from stroke and interventional guidelines.^[2] SCAI 2022 defines PFO-associated stroke as an ischaemic stroke with cortical, large white matter or retinal infarct in the presence of a PFO and no other identified likely cause.^[8] ESO 2024 describes an atrial septal aneurysm as a bulging of the atrial septum into the left and/or right atria. It can facilitate a large interatrial shunt, which ESO calls the second highest-risk feature.^[7]

Epidemiology and risk factors

^{[1] [2] [8] [7]}

• ESC 2020: severe pulmonary vascular disease is very rare (under 5%); its development presumably needs additional factors, and its course is similar to idiopathic PAH.^[1]
• ESO 2024: despite a 25% prevalence in the general population, only a small minority of people with a PFO will have a PFO-related ischaemic stroke in their lifetime.^[7]
• ESO 2024: on medical treatment, recurrence in patients with no other apparent cause of stroke than a PFO is low, about one event per 100 person-years, a cumulative incidence of 4.6% after 3.8 years of follow-up in the randomised trials.^[7]
• Australian and New Zealand consensus (2026): a PFO is implicated in 25%–50% of cryptogenic strokes in patients aged under 60 years.^[10]

Pathophysiology

ESC 2020 sets out the haemodynamics: shunt volume depends on RV and LV compliance, defect size and LA and RA pressure.^[1] Because the RV is more compliant than the LV, a simple ASD shunts left to right. The shunt is in general relevant once the defect is 10 mm or more, and it causes RV volume overload and pulmonary overcirculation.^[1] Anything that stiffens the LV or raises LA pressure (hypertension, ischaemic heart disease, cardiomyopathy, aortic and mitral valve disease) increases the left-to-right shunt. As a consequence, an ASD may become more important with age.^[1] Conversely, reduced RV compliance (pulmonary stenosis, PAH, other RV disease) or tricuspid valve disease may reduce the shunt or eventually reverse it, causing cyanosis.^[1]

ACC/AHA 2025 adds that left-to-right shunting may enlarge the right heart and cause RV dysfunction. A minority of patients may develop PAH, right-to-left shunting and/or paradoxical embolism.^[2] The left atrium matters for closure. When LA pressure is significantly raised, the ASD provides an egress, and closure without fenestration may worsen symptoms and reduce functional capacity (ACC/AHA 2025).^[2] ESC 2020 makes the same point. With impaired LV function (systolic and diastolic), closure may worsen heart failure, and an increase in filling pressure due to closure may worsen symptoms and outcome.^[1]

For the PFO, ESO 2024 lists three potential stroke mechanisms. These are paradoxical embolism of a venous clot through the PFO, clot formed in situ within the PFO, and atrial arrhythmias from disrupted electrical signalling.^[7]

Clinical presentation and natural history

• Patients often stay asymptomatic until adulthood, but most develop symptoms beyond the fourth decade: reduced functional capacity, exertional breathlessness and palpitations (supraventricular tachyarrhythmias), and less often pulmonary infections and right heart failure (ESC 2020).^[1]
• Life expectancy is reduced overall, but survival is much better than was once assumed (ESC 2020).^[1]
• Pulmonary artery pressure can be normal, but on average it rises with age (ESC 2020).^[1]
• With increasing age and pulmonary pressure, tachyarrhythmias (atrial flutter, AF) become more common (ESC 2020).^[1]
• Systemic embolism may come from paradoxical embolism, which is rare, or from AF and atrial flutter (ESC 2020).^[1]
• ACC/AHA 2025: atrial arrhythmia is seen more often when the ASD is closed later in life.^[2]

Differential diagnosis

Isolated anomalous pulmonary venous connection mimics an ASD. It overloads the right heart in the same way, but without the potential for right-to-left shunting, and its left-to-right shunt is not exacerbated by acquired left heart disease (ESC 2020).^[1] ESC 2020 applies the ASD closure principles to its repair. Technical suitability and operative risk are weighed against the potential benefit.^[1]

For a stroke in a patient with a PFO, the differential is another mechanism. AAN 2020 says clinicians should not routinely recommend PFO closure when a higher risk alternative mechanism of stroke is identified (level B).^[9] AHA/ASA 2021 notes that every trial showing benefit from closure excluded lacunar strokes, so "cryptogenic" here means nonlacunar stroke of undetermined source, or ESUS.^[6] SCAI 2022 suggests against routine PFO closure in patients with a history of AF who have had an ischaemic stroke (conditional recommendation, very low certainty of evidence).^[8]

Clinical and bedside assessment

• Examination: fixed splitting of the second heart sound and a systolic pulmonary flow murmur (ESC 2020).^[1]
• ECG: typically incomplete right bundle branch block and right-axis deviation; superior left-axis deviation points to a partial AVSD (ESC 2020).^[1]
• Chest X-ray: increased pulmonary vascularity, frequently overlooked (ESC 2020).^[1]
• ACC/AHA 2025: every patient should have a detailed history and examination, exercise testing, an ECG and transthoracic echocardiography to rule out PAH.^[2]
• Some adults with ASD and PAH may have no signs or symptoms of PAH at rest despite occult disease, so all adults with an unrepaired ASD must be evaluated for PAH, as some may show it only on specific testing such as desaturation with exercise (ACC/AHA 2025).^[2]

Investigations

Imaging the defect

ESC 2020 makes echocardiography the first-line test for diagnosis and quantification.^[1] RV volume overload is the key finding and best characterises the haemodynamic relevance of the defect, better than the shunt ratio. It may be the first, unexpected finding in a patient with an undiagnosed ASD.^[1] Sinus venosus defects generally need TOE for an accurate diagnosis. CMR or cardiac CT is an alternative and is superior for inferior sinus venosus defects.^[1] Before device closure of a secundum defect, TOE should size the defect and explore the residual septum. It should also assess rim size and quality, exclude additional defects and confirm normal pulmonary venous connection (ESC 2020).^[1] CMR is rarely needed, but may help with RV volume overload, inferior sinus venosus defects, Qp:Qs and pulmonary venous connection. For pulmonary venous connection, cardiac CT is an alternative (ESC 2020).^[1]

ACC/AHA 2025 explains its imaging row. Transthoracic echocardiography has limited ability to define all pulmonary venous connections or the whole atrial septum in adults. TOE shows the septum and some anomalous veins well, in particular with sinus venosus defects.^[2] CMR or cardiac CT may be used. CMR avoids radiation and also quantifies the shunt.^[2]

Haemodynamics and PAH

ESC 2020: cardiac catheterisation is required to determine PVR when there are non-invasive signs of raised pulmonary artery pressure. These are a calculated systolic pressure over 40 mmHg, or indirect signs when it cannot be estimated.^[1] Exercise testing should be performed in patients with PAH to exclude desaturation.^[1] ACC/AHA 2025: an adult with an ASD and confirmed precapillary PAH has group 1 PAH, managed according to current PAH guidelines. The definition is a mean PA pressure over 20 mmHg, PVR over 2 Wood units and wedge pressure of 15 mmHg or less.^[2] PAH work-up and treatment are covered in Pulmonary hypertension: five groups and PAH-specific therapy.

^[2]

Balloon testing shows how the two guidelines differ. ESC 2020 recommends it in patients with LV disease (Class I, Level C). The benefit of abolishing the left-to-right shunt is weighed against a possible rise in filling pressure, considering closure, fenestrated closure or no closure.^[1] ACC/AHA 2025 rates it 2b. It notes that retrospective analysis found balloon test occlusion added no risk-stratification information beyond a resting LV end-diastolic pressure cutoff under 15 mmHg.^[2]

Detecting a PFO

• ESO 2024: TOE is conventionally held to be the most accurate test, but its status as the gold standard is not backed by rigorously validated evidence.^[7]
• ESO 2024 expert consensus: TCD has superior sensitivity to TTE for screening for right-to-left shunt (vote 6/9); after a positive screen with no obvious cause of stroke, TOE is suggested to define the presence and anatomy of the PFO (vote 9/9).^[7]
• Australian and New Zealand consensus (2026): a TCD bubble study should be considered as first-line screening (Recommendation 4, grade B1); where TCD is not available, TTE with bubbles is an alternative with lower sensitivity (Recommendation 5, grade B1); in a patient with a positive TCD under consideration for percutaneous closure, the PFO should be confirmed by TOE, which also reveals high-risk features for stroke recurrence, including a large PFO and an associated atrial septal aneurysm (Recommendation 6, grade A1).^[10]

Management: indications for ASD closure

Indications turn on three questions: is the right heart volume-loaded, what is the PVR, and is the left heart healthy enough to lose the shunt.^[1] ESC 2020 gives them in its "Recommendations for intervention in atrial septal defect (native and residual)" table.^[1]

^{[1] [3]}

Two footnotes carry weight. RV volume overload means RV enlargement with increased stroke volume. "Desaturation on exercise" has no precise cut-off, but clinical experience puts it at a fall in arterial saturation below 90%.^[1]

Behind the PVR rows, ESC 2020 says that with PVR under 5 WU closure has been shown to be safe, lowering PAP and improving symptoms. Even in this group, the gain shrinks as PAP rises.^[1] With PVR of 5 WU or more, improvement is unlikely and complete closure is likely to make the outcome even worse.^[1] Vasoreactivity testing is not recommended for the closure decision in that group. It appears safer to treat the PAH and reassess haemodynamics. Fenestrated closure is then considered only if PVR falls below 5 WU with a significant left-to-right shunt; otherwise closure should be avoided.^[1]

^[2]

ACC/AHA 2025 defines significant LV disease as any condition causing a chronically raised left atrial pressure of 15 mmHg or more.^[2] Its supportive text says that a long-standing, unrepaired moderate to large ASD may result in right heart overcirculation, with right-sided chamber enlargement and a Qp:Qs of 1.5 or more. In that setting, regardless of symptoms, closure has been shown to improve functional capacity and reduce morbidity and mortality, particularly at age 40 or older. For safety, PAH and significant LV dysfunction should be ruled out.^[2] It counts PAH as present when PVR is over 2 Wood units. Even so, patients with a PVR of 2 to 5 Wood units have safely undergone and benefited from closure.^[2]

The newer ESC statement: ESC/ERS 2022 pulmonary hypertension guideline

The 2022 ESC/ERS PH guideline is newer than the 2020 ACHD guideline. Its Recommendation Table 18 grades shunt closure by PVR in patients with a pulmonary-to-systemic flow ratio over 1.5:1.^[3] It states that the criteria based on Qp:Qs and PVR (at baseline and/or after targeted PAH treatment) were proposed by the 2020 ESC ACHD guideline.^[3]

^[3]

The table note says closure decisions should not rest on haemodynamic numbers alone but on a multiparametric strategy. ESC/ERS 2022 gives an example: closure is not indicated with desaturation during exercise (6-minute walk test or CPET) or with secondary erythrocytosis suggesting dynamic shunt reversal.^[3] ESC/ERS 2022 also finds no evidence of long-term benefit from treat-and-repair in adult CHD-associated PAH with prevalent systemic-to-pulmonary shunts. It calls for prospective studies.^[3]

Read the 2020 and 2022 ESC rows side by side. The 2022 cut-off for the higher bands is "over 5 WU" where 2020 wrote "5 WU or more". The 2022 IIb row speaks of shunt closure where the 2020 row specified fenestrated closure.^[3][1] For an ASD with a pulmonary-to-systemic flow ratio over 1.5:1, the 2022 rows are the newer ESC statement on the same PVR bands. The 2020 rows for 3–5 WU and for 5 WU or more are superseded for that population and are shown for comparison. The 2022 under-3 WU row (I C) sits beside the 2020 Class I row. The 2020 rows on invasive PVR, device closure, elderly patients, LV disease, paradoxical embolism, Eisenmenger physiology and exercise desaturation have no counterpart row in Table 18.^[3][1]

^{[1] [3] [2] [1] [3] [2]}

Treat-and-repair and fenestration

• ACC/AHA 2025: in proven PAH, PAH-directed therapy is indicated, and when PVR is 5 Wood units or more, regression below 5 may make closure safe to consider.^[2]
• Closure with raised PVR does not remove the need for continued PAH therapy and specialised PAH care, which ACC/AHA recommends in this group; average follow-up in the studies was short to medium term, about 3 years.^[2]
• Fenestrated closure has commonly been used to keep an outlet in case PA pressures keep rising, but no fenestrated transcatheter device is currently available in the United States (ACC/AHA 2025).^[2]
• ACC/AHA 2025 says there is likely to be a narrow range within which fenestrated closure may be reasonable for an adult with ASD and PAH whose PVR is 5 Wood units or more, needing careful reanalysis and repeat invasive haemodynamics after PAH therapy, as offering closure could be dangerous if PAH remains progressive.^[2]

Paradoxical embolism through an ASD

ESC 2020 says closure should be considered after suspected paradoxical embolism, other causes excluded, regardless of defect size, provided there is no PAH and no LV disease (Class IIa, Level C). ACC/AHA 2025 recommends it with evidence or strong suspicion of paradoxical embolism (COR 1, LOE C-EO).^[1][2] ACC/AHA adds that paradoxical embolism often reflects a systemic disease that predisposes to venous thromboembolism (clotting disorders, malignancy, kidney disease, connective tissue disease). It says it is reasonable to use, for example, the Risk of Paradoxical Embolism score to identify patients with a thrombophilia who may benefit more from closure than medical therapy alone.^[2] With known or suspected paradoxical embolism, particularly with a thrombophilia, ACC/AHA calls closure indicated to prevent recurrence. This holds regardless of defect size, chamber size or Qp:Qs, in the absence of significant PAH or LV disease.^[2]

Contraindications and cautions

ESC/ERS 2022 adds that defect closure may also harm patients with small or coincidental defects that behave like idiopathic PAH.^[3] ESC/ERS 2022 says shunt closure, surgical or interventional, may only be considered in patients with prevalent systemic-to-pulmonary shunting without significantly increased PVR.^[3]

Device versus surgical closure

ESC 2020: device closure has become the first choice for secundum defects when the morphology allows, which is the case in about 80% of patients. The morphology criteria include a stretched diameter of 38 mm or less and a sufficient rim of 5 mm except towards the aorta.^[1] Mortality cannot be assumed to be zero, but several recent studies reported none, and serious complications occurred in 1% or fewer.^[1] Early atrial tachyarrhythmias are mostly transient. Erosion of the atrial wall, anterior mitral leaflet or aorta, and thromboembolic events, appear to be very rare.^[1] Antiplatelet therapy is required for at least 6 months (aspirin 75 mg once daily minimum). The incidence of late arrhythmias or adverse events still needs study.^[1]

Surgical repair has low mortality (under 1% without significant comorbidity) and good long-term outcome when done early (childhood, adolescence) without pulmonary hypertension (ESC 2020).^[1] ESC 2020 recommends carefully weighing surgical risk against benefit in elderly patients not suitable for device closure (Class I, Level C).^[1]

^{[1] [2]}

ACC/AHA 2025 says transcatheter closure has been used mostly for secundum defects with adequate rims and, traditionally, without partial anomalous pulmonary venous return. Morbidity and mortality are low with both approaches.^[2] For superior sinus venosus defects, ACC/AHA 2025 says nonsurgical options are attractive. In eligible transcatheter cases, procedural success is typically over 95%, with no deaths and few complications. Its row makes transcatheter closure, when technically feasible, a possibly reasonable alternative to surgery to reduce operative morbidity (COR 2b, LOE B-NR).^[2] In patients with atrial arrhythmias, an electrophysiology study is often done before closure (ACC/AHA 2025).^[2]

Age shapes the result. ESC 2020 says outcome is best with repair before 25 years, and closure after 40 appears not to change the frequency of later arrhythmias. Morbidity (exercise capacity, breathlessness, right heart failure) benefits at any age, particularly with catheter closure.^[1]

PFO-associated stroke: selecting patients for closure

The selection rows come from four bodies; each keeps its own population and grading.^[6][7][8][9]

^[6]

AHA/ASA notes that each study defines high-risk anatomic features differently.^[6] AHA/ASA 2021 reports a meta-analysis of all the trials. The number needed to treat to prevent one recurrent stroke was 131 over 1 person-year of follow-up, or 13 over 10 person-years.^[6] It also reports administrative claims data. Serious periprocedural complications, including AF, occurred in 4.9% of patients aged 60 or under and were significantly more frequent (10.9%) above 60, where randomised data are extremely limited.^[6]

ESO 2024 evidence-based treatment recommendations (selected from Table 12, synoptic table of all recommendations)

^[7]

ESO 2024 found insufficient evidence for an evidence-based recommendation on closure above 60 or below 18 years. It did not count TIA as an index event, because only one randomised trial included TIA patients.^[7] It also suggests against long-term anticoagulation in PFO-associated stroke unless anticoagulation is indicated for another reason.^[7]

^[8]

SCAI marks strength by wording: "recommends" is strong and "suggests" is conditional.^[8] AAN 2020 practice advisory: clinicians should ensure a thorough evaluation has ruled out other stroke mechanisms (level B). Under 60 years with a PFO, an embolic-appearing infarct and no other mechanism identified, they may recommend closure after discussing benefits and risks (level C).^[9] AAN also advises counselling that a PFO is common, found in about 1 in 4 adults. Patients should hear that causation is hard to establish with certainty, and that closure probably reduces recurrent stroke risk in selected patients (level B).^[9]

Estimating causality: RoPE and PASCAL

AHA/ASA 2021 describes the Risk of Paradoxical Embolism (RoPE) score as an index of how likely a PFO is to be stroke-related, which may help selection.^[6] Higher scores go with no hypertension, diabetes, prior stroke or TIA, a cortical infarct on imaging and younger age. The estimated probability that the PFO is stroke-related runs from 0% (95% CI 0–4) at scores of 0 to 3 to 88% (95% CI 83–91) at 9 to 10.^[6] ESO 2024 Table 6 gives one point each for no history of hypertension, diabetes, or stroke or TIA, for being a non-smoker and for a cortical infarct on imaging. Age adds points (18–29 years 5; 30–39, 4; 40–49, 3; 50–59, 2; 60–69, 1; over 70, 0), to a maximum of 10.^[7]

PASCAL classification (ESO 2024, Table 7)

^[7]

In ESO Table 7, a large shunt means more than 20 bubbles in the left atrium on TOE. An atrial septal aneurysm means more than 10 mm of excursion from the midline.^[7] ESO 2024 lowered the quality of evidence for PASCAL-based selection by one level. PASCAL was neither an inclusion criterion nor a stratification variable in any of the six randomised trials, and it has not been prospectively validated.^[7]

ESO 2024 reports the SCOPE individual patient data meta-analysis (IPDMA) of randomised trials of closure versus medical therapy in patients with cryptogenic stroke and a PFO. Hazard ratios for recurrent ischaemic stroke were 1.14 (95% CI 0.53–2.46), 0.38 (0.22–0.65) and 0.10 (0.03–0.35) in the unlikely, possible and probable PASCAL groups (Table 8).^[7] The 2-year absolute risk reductions were −0.7%, 2.1% and 2.1% in the same three groups.^[7] In a further IPDMA analysis, patients with a large shunt plus an atrial septal aneurysm benefited most. The absolute risk reduction was 5.5% at 2 years and 7.1% at 5 years, which ESO 2024 translates into a number needed to treat of 14.^[7]

^[7]

Older patients, anticoagulation and thrombophilia

• Older patients: AHA/ASA 2021 says the procedure should rarely be done in patients over 60, and only in very unusual clinical circumstances; ESO 2024 gives no evidence-based recommendation above 60 and, by expert consensus (vote 8/9), encourages trials or at least a registry and otherwise suggests using PASCAL and clinical judgement to guide therapy; for patients aged 60 or older with a prior PFO-associated stroke, SCAI 2022 suggests PFO closure rather than long-term antiplatelet therapy alone (conditional, very low certainty).^[6][7][8]
• Closure versus anticoagulation: AHA/ASA 2021 says that in patients aged 18 to 60 with a nonlacunar ischaemic stroke of undetermined cause despite a thorough evaluation and a PFO, the comparative benefit of closure versus warfarin is unknown (2b, C-LD), and notes closure in appropriately selected patients is superior to aspirin but not known to be superior to warfarin.^[6]
• In its supporting text, not a graded recommendation, ESO 2024 notes that a history of non-cerebral embolism, deep vein thrombosis, pulmonary embolism or PAH, Valsalva at stroke onset and the presence of venous thrombophilia, among other features, may be taken into consideration when managing patients after PFO-associated stroke.^[7]
• ESO 2024 expert consensus suggests PFO closure in selected patients aged 13 to 17 with PFO-related stroke, according to PFO anatomy.^[7]

After PFO closure

• After percutaneous closure in patients with PFO-related stroke, ESO 2024 suggests dual antiplatelet therapy followed by single antiplatelet therapy to reduce the risk of recurrent stroke, based on the protocol of available randomised trials (PICO 5 recommendation; quality of evidence low).^[7]
• The durations are ESO 2024 expert consensus (vote 9/9), following the protocol design of most positive randomised trials: dual antiplatelet therapy for 1–6 months, then single antiplatelet therapy long term, for at least 5 years after closure. ESO 2024 states that no evidence-based recommendation can be formulated on the duration of single antiplatelet therapy.^[7]
• ESO 2024 expert consensus (vote 8/9): in patients aged 18–60 with PFO-associated stroke, basic cardiac monitoring for 24 hours by telemetry or Holter should be done before closure.^[7]

PFO closure outside stroke (selected SCAI 2022 rows)

• Migraine without a prior PFO-associated stroke: suggests against routine PFO closure (conditional, moderate certainty).^[8]
• Divers with prior decompression illness and no prior PFO-associated stroke: suggests against routine closure to prevent decompression illness (conditional, very low certainty).^[8]
• Platypnoea-orthodeoxia syndrome with other causes of hypoxia excluded and no prior PFO-associated stroke: suggests PFO closure rather than no closure (conditional, very low certainty).^[8]

Complications and pitfalls

• Device erosion of the atrial wall, anterior mitral leaflet or aorta appears very rare (ESC 2020); ACC/AHA 2025 notes that suspected erosion is one reason for follow-up advanced imaging.^[1][2]
• After closure above age 40, atrial arrhythmias reach a prevalence of up to 40–60%; without repair or with repair after 40, AF becomes more common (ESC 2020).^[1]
• Late arrhythmias after surgery before age 40 are most often intra-atrial re-entrant tachycardia or atrial flutter, treatable by radiofrequency or cryoablation (ESC 2020).^[1]
• After sinus venosus repair, SVC stenosis and stenosis of redirected pulmonary veins may occur (ESC 2020).^[1]
• PFO closure and AF: in the SCOPE IPDMA reported by ESO 2024, the absolute increase in AF beyond day 45 was 4.41% in the PASCAL unlikely group, 1.53% in possible and 0.65% in probable.^[7]
• In the same analysis, major bleeding included access-site or retroperitoneal haemorrhage in 1%, pericardial tamponade in 0.17% and cardiac perforation in 0.06% (ESO 2024).^[7]
• AHA/ASA 2021 notes that residual shunts after PFO device closure are associated with an increased risk of stroke recurrence, and lists among its knowledge gaps studies to determine the optimal prevention strategy in that setting, including consideration of a second device closure or lifelong anticoagulant therapy.^[6]

Pitfall: some adults with ASD and PAH may have no signs or symptoms of PAH at rest. PAH may show only on specific testing such as exercise desaturation (ACC/AHA 2025).^[2] Pitfall: with impaired LV function, ASD closure may worsen heart failure (ESC 2020). With significantly raised left atrial pressure the ASD provides an egress, and closure without fenestration may worsen symptoms and reduce functional capacity (ACC/AHA 2025).^[1][2]

Prognosis and follow-up

ESC 2020: follow-up should assess residual shunt, RV size and function, TR and PAP by echocardiography. Arrhythmias are assessed by history, ECG and, only if indicated, Holter monitoring.^[1]

• Repaired before 25 years without sequelae (no residual shunt, normal PAP, normal RV, no arrhythmias): no regular follow-up, but patients and referrers should know that tachyarrhythmias can occur late (ESC 2020).^[1]
• Residual shunt, raised PAP, arrhythmias (before or after repair), or repair in adulthood (particularly over 40): regular follow-up, including in specialised ACHD centres, at intervals set by the residual problems (ESC 2020).^[1]
• After device closure: regular follow-up for the first 2 years, then, depending on results, every 3–5 years is reasonable (ESC 2020).^[1]

ACC/AHA 2025 Table 12: ASD routine follow-up and testing intervals by physiological stage

^[2]

ACC/AHA 2025 gives no evidence-based interval for repeat advanced imaging. Some patients may occasionally need it, for example when device erosion is suspected.^[2]

^[2]

Endocarditis prophylaxis

ESC 2020 recommends endocarditis prophylaxis for 6 months after device closure.^[1] The newer 2023 ESC endocarditis guideline places patients with septal defect closure devices in its high-risk group, in whom prophylaxis is recommended or should be considered. This applies during the first 6 months after implantation.^[4] It counts secundum ASD itself among the simple lesions with a low risk of endocarditis.^[4] Prophylaxis regimens are covered in Infective endocarditis.

Special populations

Pregnancy

The 2025 ESC pregnancy guideline (Table 12) rates pregnancy as low risk with a repaired or unrepaired ASD if there is no PAH. Its ASD row also lists the entries below, among others.^[5]

• For an unrepaired ASD in pregnancy: increased risk of arrhythmia (4%); paradoxical embolism (2%–5%); small-for-gestational-age birth (21%); foetal or perinatal mortality (2%–3%); and pre-eclampsia (7%).^[5]
• Other entries: consider TTE at 28–32 weeks for an unrepaired, uncomplicated ASD; close a large and/or haemodynamically significant ASD before pregnancy; consider aspirin or prophylactic LMWH for prevention of paradoxical embolism with an unrepaired ASD; and consider device closure in pregnancy only for recurrent stroke on medical therapy.^[5]

ESC 2025 Recommendation Table 10 (pulmonary arterial hypertension and pregnancy) includes two Class I, Level C rows on counselling and contraception. It recommends that women of childbearing potential with PAH who wish to become pregnant are counselled by a multidisciplinary team about the very high risk of pregnancy-related adverse events, encouraging shared decision-making about whether to become pregnant. It also recommends clear contraceptive advice for women of childbearing potential with PAH.^[5] Its text on counselling and contraception in PAH says these women should be counselled at diagnosis about the very high risk and uncertainties of becoming pregnant. Clear advice against becoming pregnant, including referral for psychological support if needed, and clear contraceptive advice are required, taking into account the woman's individual needs (no class or level given).^[5] The 2020 ESC ACHD guideline says that patients with pre-capillary pulmonary hypertension should be counselled against pregnancy.^[1] Risk scoring and delivery planning are covered in Cardiovascular disease in pregnancy.

Older adults

ESC 2020 says surgical repair can be performed with very low risk, even in the elderly, but comorbidities that may affect operative risk need to be weighed against the benefit. In elderly patients not suitable for a device, that weighing is a Class I recommendation (Level C).^[1] ACC/AHA 2025 says that in a long-standing, unrepaired moderate to large ASD with right-sided chamber enlargement and a Qp:Qs of 1.5 or more, closure, regardless of symptoms, has been shown to improve functional capacity and reduce morbidity and mortality, particularly at age 40 or older. It says PAH and significant LV dysfunction should be ruled out for safety.^[2]

Evidence, guidelines and regional differences

For ASD, the European rows come from the 2020 ESC ACHD guideline. The newer 2022 ESC/ERS PH guideline gives updated PVR-based shunt-closure rows (Recommendation Table 18) for patients with a pulmonary-to-systemic flow ratio over 1.5:1. The North American rows come from the 2025 ACC/AHA ACHD guideline, a full revision and replacement of the 2018 AHA/ACC guideline.^[1][3][2] The 2025 "What Is New" table records the replaced 2018 row as revised; it is history only. That row was COR 2b: closure may be considered with Qp:Qs of at least 1.5:1, PA systolic pressure at least 50% of systemic and/or PVR above one-third of systemic.^[2]

^{[1] [2]}

ANZ practice

No NHFA/CSANZ guideline on ASD or adult congenital heart disease was found in the census of PubMed and the guideline register for this topic. For PFO-associated stroke, a 2026 Australian and New Zealand consensus statement (Med J Aust) gives practice recommendations. It reports that a PFO is implicated in 25%–50% of cryptogenic strokes in patients aged under 60 years.^[10] Each recommendation met consensus, defined as more than 75% agreement among the participating experts.^[10] Table 1 of the statement grades each recommendation by the quality of the supporting evidence (A high, B moderate, C low) and its strength (1 strong, 2 conditional), or marks it GPP, a good practice point. The recommendations used here include the following, each with the grade Table 1 prints.^[10]

• Patients aged 60 or under with cryptogenic stroke should undergo screening for PFO (Recommendation 1, grade A1).^[10]
• Above 60, screening may be considered with embolic stroke of undetermined source, no vascular risk factors and no AF on prolonged monitoring (Recommendation 2, grade B1).^[10]
• A high RoPE score alone does not predict benefit from closure and should not be the principal basis of the decision (Recommendation 7, grade B1); PASCAL, which combines RoPE with high-risk PFO features, should be incorporated into the decision (Recommendation 8, grade B1).^[10]
• Patients with PFO-associated stroke and high-risk features for recurrence should be considered for closure (Recommendation 10, grade B1); the decision should be made by a multidisciplinary heart–brain team that includes a stroke physician and a structural heart cardiologist (Recommendation 11, GPP).^[10]
• In PFO-associated stroke, concurrent deep vein thrombosis and pulmonary embolism, and demonstrated thrombophilia, should also be considered in the decision (Recommendation 12, GPP).^[10]

Guidelines checked

A row called the current one is so among the guidelines checked for this topic:

• Sources of the rows and statements used: ESC ACHD (2020); ESC/ERS pulmonary hypertension (2022); ESC endocarditis (2023); ESC cardiovascular disease and pregnancy (2025); ACC/AHA/HRS/ISACHD/SCAI ACHD (2025), which replaces AHA/ACC 2018; AHA/ASA secondary stroke prevention (2021); ESO PFO (2024); SCAI PFO (2022); AAN practice advisory (2020); Australian and New Zealand PFO consensus (2026).^{[1][3][4][5][2][6][7][8][9][10]}
• The 2019 European position paper on PFO management was found in the census but is not held as text for this topic, so it is not used.

Exam pearls

• Secundum ASD is 80% of ASDs; superior sinus venosus defects (5%) are associated with partial or complete connection of the right pulmonary veins to the SVC or RA (ESC 2020).^[1]
• Fixed splitting of S2, a pulmonary flow murmur, incomplete right bundle branch block and right-axis deviation; superior left-axis deviation suggests partial AVSD (ESC 2020).^[1]
• RV volume overload, not the shunt ratio, best characterises haemodynamic relevance (ESC 2020).^[1]
• Device suitability (ESC 2020) depends on morphology, which includes a stretched diameter of 38 mm or less and a sufficient 5 mm rim except towards the aorta; this is the case in about 80% of patients.^[1]
• After device closure: antiplatelet therapy for at least 6 months, aspirin 75 mg once daily minimum (ESC 2020).^[1]
• Vasoreactivity testing is not recommended when deciding on closure with PVR of 5 WU or more (ESC 2020).^[1]
• PASCAL probable means RoPE 7 or more plus a large shunt or atrial septal aneurysm (ESO 2024, Table 7).^[7]
• ESO 2024 after PFO closure: dual antiplatelet therapy followed by single antiplatelet therapy to reduce the risk of recurrent stroke, based on the protocol of available randomised trials (PICO 5 suggestion, low-quality evidence); the durations, 1–6 months then at least 5 years, are expert consensus.^[7]

References10ShowHide

1. [1]Baumgartner H, et al. 2020 ESC Guidelines for the management of adult congenital heart disease. Eur Heart J, 2021.PMID 32860028
2. [2]Gurvitz M, et al. 2025 ACC/AHA/HRS/ISACHD/SCAI Guideline for the Management of Adults With Congenital Heart Disease: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation, 2026.PMID 41411375
3. [3]Humbert M, et al. 2022 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension. Eur Respir J, 2023.PMID 36028254
4. [4]Delgado V, et al. 2023 ESC Guidelines for the management of endocarditis. Eur Heart J, 2023.PMID 37622656
5. [5]De Backer J, et al. 2025 ESC Guidelines for the management of cardiovascular disease and pregnancy. Eur Heart J, 2025.PMID 40878294
6. [6]Kleindorfer DO, et al. 2021 Guideline for the Prevention of Stroke in Patients With Stroke and Transient Ischemic Attack: A Guideline From the American Heart Association/American Stroke Association. Stroke, 2021.PMID 34024117
7. [7]Caso V, et al. European Stroke Organisation (ESO) Guidelines on the diagnosis and management of patent foramen ovale (PFO) after stroke. Eur Stroke J, 2024.PMID 38752755
8. [8]Kavinsky CJ, et al. SCAI Guidelines for the Management of Patent Foramen Ovale. J Soc Cardiovasc Angiogr Interv, 2022.PMID 39131947
9. [9]Messé SR, et al. Practice advisory update summary: Patent foramen ovale and secondary stroke prevention: Report of the Guideline Subcommittee of the American Academy of Neurology. Neurology, 2020.PMID 32350058
10. [10]Chambers BR, et al. Diagnosis and Management of Patent Foramen Ovale for Stroke Prevention: An Australian and New Zealand Consensus Statement Developed by a Modified Nominal Group Approach. Med J Aust, 2026.PMID 42129593