Stroke associated with patent foramen ovale (PFO): Evaluation

INTRODUCTION — When embolic ischemic stroke occurs in an individual with a patent foramen ovale (PFO), the PFO may or may not be causally related to the stroke. This topic will review the approach to the evaluation of patients who have an ischemic stroke in the setting of a PFO. The management of PFO-associated stroke is reviewed elsewhere. (See "Stroke associated with patent foramen ovale (PFO): Management".)

Cryptogenic stroke is reviewed in detail separately. (See "Cryptogenic stroke and embolic stroke of undetermined source (ESUS)".)

The risk of stroke related to atrial septal abnormalities and indications for treating atrial septal defects in adults are also discussed elsewhere. (See "Atrial septal abnormalities (PFO, ASD, and ASA) and risk of cerebral emboli in adults" and "Management of atrial septal defects in adults".)

COMPREHENSIVE EVALUATION

Goals and approach — Determining whether a PFO is incidental or pathogenic in relation to an ischemic stroke is essential to guide decisions about PFO management and secondary stroke prevention. The approach incorporates an evaluation for other possible causes of ischemic stroke, PFO features, and methods (ie, the Risk of Paradoxical Embolism [RoPE] score and PFO-associated stroke causal likelihood [PASCAL] classification) that estimate the likelihood of paradoxical embolism through a PFO as the mechanism of the stroke.

Clues from the history — The history should include specific questions that define the circumstances immediately preceding the event. Was the patient doing something that might increase right-to-left shunt flow through a PFO, such as straining, coughing vigorously, or lifting or pushing a heavy object? Was there any risk for deep venous thrombosis (DVT), such as prolonged immobility (eg, postoperative status, sitting in a cramped airline seat with legs dependent and knees flexed), dehydration, or venous hypercoagulability?

The presence of the above historical features raise concern for paradoxical embolism. However, absence of these features does not exclude paradoxical embolism.

Is the stroke embolic? — By definition, PFO-associated stroke is embolic. On evaluation, an embolic mechanism is especially likely when neuroimaging reveals infarcts in multiple vascular territories or a single wedge-shaped infarct involving cortex and the underlying subcortical white matter. With embolic stroke, the neurologic deficit is typically maximal from the onset, with a possibility of rapid improvement if there is spontaneous recanalization. (See "Overview of the evaluation of stroke", section on 'Determining a presumptive diagnosis of stroke subtype'.)

Whether or not an embolic stroke is related to PFO or to another mechanism cannot be determined reliably based upon neuroimaging features alone. However, certain findings may be more suggestive of PFO. In an analysis of data from subjects with cryptogenic stroke and PFO (n = 1141) or no PFO (n = 1539) in the RoPE study database, characteristics associated with a significantly higher prevalence of PFO were a large index stroke (odds ratio [OR] 1.36), index stroke seen on imaging (OR 1.53), and superficial (ie, involving the cerebral or cerebellar cortex) stroke location (OR 1.54) [1]. The last of these characteristics is included in the RoPE score.

Exclusion of other sources of ischemic stroke — Patients with an ischemic stroke and a PFO should undergo a comprehensive evaluation by both a stroke neurologist and a cardiologist to ensure that other causes of ischemic stroke are excluded and that PFO-associated stroke (ie, paradoxical embolism through a PFO) is the most likely mechanism [2]. (See "Cryptogenic stroke and embolic stroke of undetermined source (ESUS)", section on 'Evaluation and diagnosis'.)

●PFO-associated stroke – Patients with an embolic-appearing ischemic stroke who have a medium- or high-risk PFO (see 'PFO assessment' below) and who have no other identified stroke etiology should be recognized as having a PFO-associated stroke [3].

●Findings – The evaluation of a PFO-associated stroke is characterized by the following findings [4]:

•No large vessel stenosis (≥50 percent) or occlusion in the territory of the infarct. Intracranial and extracranial neurovascular imaging should be performed to exclude vascular causes of stroke, including large artery atherosclerosis, dissection, and other vasculopathy. (See "Neuroimaging of acute stroke".)

•No radiographic acute lacunar infarction (ie, a small [≤1.5 cm] deep perforator infarct) and no clinical lacunar stroke syndrome (ie, hemiparesis/plegia, hemianesthesia without cortical signs) if imaging shows no infarct. (See "Lacunar infarcts", section on 'Clinical features'.)

•No evidence of occult atrial fibrillation or other high-risk cardioembolic source (table 1) on cardiac monitoring and echocardiography.

-Screening for atrial fibrillation should include a 12-lead electrocardiogram (ECG) and 24-hour cardiac monitoring (ambulatory or by telemetry). Adults over age 40 years with cryptogenic ischemic stroke or cryptogenic transient ischemic attack (TIA) with no atrial fibrillation detected on 12-lead ECG and 24-hour monitoring should undergo extended ambulatory monitoring, as discussed elsewhere. (See "Cryptogenic stroke and embolic stroke of undetermined source (ESUS)", section on 'Advanced evaluation'.)

-Transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE) not only enable identification and characterization of PFOs but also enable identification of relevant structural findings including atrial septal aneurysms and other potential embolic sources such as high-risk valve lesions (eg, mitral stenosis), cardiac tumors (eg, myxoma), intracardiac thrombus, fibroelastomas, and other valvular vegetations. Test selection is discussed below and reviewed separately. (See 'PFO assessment' below and "Echocardiography in detection of cardiac and aortic sources of systemic embolism".)

•No hypercoagulable condition with a high risk for arterial thrombotic events, such as the antiphospholipid syndrome. (See "Clinical manifestations of antiphospholipid syndrome" and "Diagnosis of antiphospholipid syndrome" and "Management of antiphospholipid syndrome".)

PFO assessment — Methods to detect a right-to-left shunt associated with a PFO include TTE, TEE, and transcranial Doppler (TCD), in conjunction with agitated saline contrast (a "bubble study")Among these methods, TCD cannot confirm intracardiac shunting or relevant structural findings (e.g., atrial septal aneurysm), and generally only TEE enables visualization of the type (eg, PFO, atrial septal defect [ASD]), site, and size of the shunt. (See "Cryptogenic stroke and embolic stroke of undetermined source (ESUS)", section on 'Cardiac and aortic evaluation'.)

While cardiac CT and cardiovascular magnetic resonance (CMR) imaging have been used to identify PFO, they are less sensitive than TEE [5-7].

Identification of PFO — The diagnostic evaluation of PFO is reviewed here briefly and discussed in detail separately. (See "Patent foramen ovale", section on 'Diagnosis and evaluation'.)

●TTE or TEE – These tests are performed with intravenous injection of agitated saline contrast at rest, with Valsalva, and with cough; the study is considered positive if microbubbles (typically three or more) appear in the left heart within three cardiac cycles of bubbles filling the right atrium [8]. Multiple agitated saline contrast injections with provocative maneuvers may be required to enhance sensitivity for identification of a shunt via the PFO [9]. In some cases, intermittent flow through the PFO is also visualized by color Doppler. (See "Patent foramen ovale", section on 'Diagnosis and evaluation'.)

TTE is often used as the initial study because it is better tolerated than TEE and is more widely available than TEE or TCD. Also, TTE provides a better assessment (with a microbubble ultrasound-enhancing agent) for left ventricular thrombus (particularly apical), which is an important component of the evaluation in patients who may have ischemic heart disease or cardiomyopathy. In addition, some patients are better able to perform the Valsalva maneuver and cough during TTE than during TEE (due to the effects of sedation and/or discomfort during TEE). Thus, in some cases, TTE may be more sensitive than TEE for PFO shunt detection [10].

However, TEE is the reference standard for the detection of PFO [9], is generally more sensitive than TTE [11], and better demonstrates its anatomic features. Therefore, patients who are candidates for PFO closure (ie, those ≤60 years of age with an embolic-appearing stroke or a TIA who have no evident source of stroke or TIA other than a PFO, despite a comprehensive evaluation) should undergo TEE as either the initial study or as a follow-up study to TTE or TCD, even when a preceding TTE is negative or nondiagnostic for PFO. (See "Stroke associated with patent foramen ovale (PFO): Management", section on 'Patient selection for PFO closure'.)

For patients who do not meet selection criteria for PFO closure, a follow-up TEE may still be indicated to identify a stroke etiology but is not required to assess for PFO. Note that all patients who are candidates for PFO closure should have preprocedural imaging with TEE, including those with features suggesting PFO detected by TTE or TCD. (See "Stroke associated with patent foramen ovale (PFO): Management", section on 'Preprocedural imaging'.)

●TCD – Insonating the middle cerebral artery through the temporal bone window, TCD is performed with intravenous injection of agitated saline contrast at rest, with Valsalva, and with cough; the study is considered positive for a right-to-left shunt if bubbles are detected in the middle cerebral artery. This approach is at least as sensitive as TEE to identify right-to-left shunt [12-15]; TCD also has advantages compared with TEE in being noninvasive and relatively easy to perform at the bedside. However, TCD cannot confirm the location of the shunt (which may be intracardiac or extracardiac), identify an atrial septal aneurysm, or rule out other cardioembolic sources. Thus, when TCD is positive, imaging (generally with TEE) is performed to identify shunt location [11].

PFO shunt size — PFO size (small versus large) is a component of the PASCAL approach to risk classification. PFO size is generally inferred from the degree of right-to-left shunting apparent on echocardiography with agitated saline contrast, based on the number of microbubbles appearing in a single frame in the left atrium either spontaneously or after a provocative (Valsalva) maneuver within three cardiac cycles after opacification of the right atrium [16,17].

PFO size is categorized as small or large in the PASCAL classification described below. We use the following categories (based upon thresholds in four of six randomized trials in the meta-analysis discussed below [18]):

●Large: >20 microbubbles (which was the threshold used in four of the six included randomized trials)

●Small: ≤20 microbubbles

Some trials have classified shunt size into small, moderate, and large categories [19]. In one small study, PFO measured ante mortem by assessment of jet size on color Doppler TEE and number of microbubbles on contrast TEE correlated well with autopsy finding; patients with a PFO >10 mm at autopsy had large shunts with >25 microbubbles on contrast TEE [20].

Among patients with an embolic stroke topography, a PFO, and no alternative cause, features associated with medium- to high-risk PFO include factors that increase right-to-left shunt flow (eg, large PFO size, chronic right atrial hypertension, or a Valsalva maneuver at onset of stroke) and concomitant pulmonary embolism or deep venous thrombosis preceding the ischemic stroke [3]. Individual patient characteristics (age and vascular risk factors) are also important in the assessment of PFO risk.

Atrial septal aneurysm — An atrial septal aneurysm (ASA) is defined as redundant and mobile interatrial septal tissue in the region of the fossa ovalis with phasic oscillation (intrusion) of at least 10 to 15 mm into the left or right atrium during the cardiorespiratory cycle and is most commonly identified on TEE (with lower sensitivity of detection by TTE). Although not a consistent finding across all studies, the risk of stroke is likely increased when a PFO is associated with an ASA. (See "Atrial septal abnormalities (PFO, ASD, and ASA) and risk of cerebral emboli in adults", section on 'ASA' and "Patent foramen ovale", section on 'Atrial septal aneurysm'.)

Straddling thrombus — Thrombus trapped in a PFO is rarely encountered [21-24], but this finding confirms paradoxical embolus in transit via a very high-risk PFO, which may in some cases warrant surgical treatment to remove the clot and repair the PFO [3].

Search for venous thromboembolism — We suggest performing a standard evaluation for DVT with D-dimer level and/or ultrasonography of the lower extremities for patients with a PFO-associated embolic infarct and no other evident source of stroke; additional investigations such as CT or MRI venography are generally not warranted. Details of the evaluation are discussed separately. (See "Clinical presentation and diagnosis of the nonpregnant adult with suspected deep vein thrombosis of the lower extremity".)

Identification of DVT or other evidence of venous thromboembolism can help strengthen the clinical inference of paradoxical embolism and also has important implications for therapy (including identifying an indication for anticoagulation, which impacts the timing and need for PFO device closure). Ideally, the evaluation should be done within two to three days of stroke onset, before venous thrombosis develops secondary to stroke-related immobility [3].

Studies using radiograph venography or magnetic resonance venography to examine patients with ischemic cerebral events or other suspected embolic events have found variable rates of proximal leg or pelvic deep venous thrombosis (10 to 22 percent) [25-27]. Among those with documented DVT, most had no symptoms or signs of venous thrombosis. There are several potential explanations for the low rates of proximal DVT in patients with PFO-associated embolic infarct and no other evident source of stroke:

●Evaluation for DVT was incomplete in these studies since venography evaluated either the lower extremities or pelvic veins, but not both.

●The emboli consist of platelet fibrin particles that normally circulate in the systemic venous bed and are too small to be seen with conventional testing. These particles are removed by the efficient lytic system of the lungs; however, when shunted across a PFO or ASD, they may travel unlysed into the cerebral circulation [28].

●The clot forms in the heart at the edges or in the tunnel of the PFO, or in an ASA.

●The embolic source is in the systemic arterial circulation. (See "Atrial septal abnormalities (PFO, ASD, and ASA) and risk of cerebral emboli in adults", section on 'Left-sided sources'.)

●DVTs have resolved (lysed, completely embolized, or recanalized) during the delay between onset of the embolic event and venography.

●Proximal DVTs may be lysed in some patients treated with intravenous thrombolytic therapy [27].

Hypercoagulable evaluation — Hematologic testing to exclude hypercoagulable states (eg, antiphospholipid syndrome and hyperhomocysteinemia) is indicated for patients with stroke being considered for PFO closure. (See "Diagnosis of antiphospholipid syndrome".)

LIKELIHOOD OF A PFO-ASSOCIATED STROKE

RoPE score — The Risk of Paradoxical Embolism (RoPE) score, as shown in the table (table 2) and calculator (calculator 1), estimates the probability that a PFO is incidental or pathogenic in a patient with a seemingly cryptogenic stroke [29]. The PFO-attributable fraction of stroke derived from the RoPE score (table 3) varies widely and decreases with age and the presence of vascular risk factors. High RoPE scores, as found in younger patients who lack vascular risk factors and have a cortical infarct on neuroimaging, suggest pathogenic, higher risk PFOs. By contrast, low RoPE scores, as found in older patients with vascular risk factors, suggest incidental, lower-risk PFOs. The RoPE score is a major component of the PFO-associated stroke causal likelihood (PASCAL) classification system, which provides additional discrimination, as described below.

PASCAL classification — The PASCAL classification system (table 4) estimates the probability that stroke is associated with a PFO in patients with embolic infarct topography and without other major sources of ischemic stroke [3]. The classification is based upon the RoPE score combined with anatomic and clinical factors (see 'PFO assessment' above) including shunt size, presence or absence of atrial septal aneurysm, and/or venous thromboembolism; PASCAL categorizes the likelihood that the stroke is caused by a PFO as unlikely, possible, probable, highly probable, or definite, as shown in the table (table 4).

Given the high prevalence of PFO in the general population and the low risk of stroke related to PFO, there is always some degree of uncertainty about the causal relationship between PFO and an embolic-appearing ischemic stroke with no other evident stroke mechanism despite a comprehensive evaluation [30]. The possibility that the PFO is an "innocent bystander" and that another mechanism is responsible for the stroke is particularly applicable to older patients and to those with known risk factors for stroke (eg, hypertension, hypercholesterolemia, smoking) [31-33]. Causality can best be inferred in younger patients with no other apparent etiology for stroke [34], particularly if DVT is present (as a potential source for paradoxical emboli).

MANAGEMENT — The management of PFO-associated stroke, including selection of appropriate candidates for PFO closure (algorithm 1), is reviewed separately. (See "Stroke associated with patent foramen ovale (PFO): Management".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Stroke in adults".)

INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Patent foramen ovale (The Basics)")

SUMMARY AND RECOMMENDATIONS

●Is the PFO incidental or pathogenic? – Determining whether a patent foramen ovale (PFO) is incidental or pathogenic in relation to an ischemic stroke is essential to guide decisions about PFO management and secondary stroke prevention. This determination incorporates an evaluation for other possible causes of ischemic stroke, PFO features, and methods (ie, the Risk of Paradoxical Embolism [RoPE] score and PFO-associated stroke causal likelihood [PASCAL] classification) that estimate the likelihood of paradoxical embolism through a PFO as the mechanism of the stroke. (See 'Goals and approach' above.)

●Is the stroke embolic? – By definition, PFO-associated stroke is embolic. Neuroimaging features that suggest embolism include infarcts in multiple vascular territories or a single wedge-shaped infarct involving cortex and the underlying subcortical white matter. (See 'Is the stroke embolic?' above.)

●Exclusion of other sources of ischemic stroke – A PFO-associated embolic infarct requires that other causes of ischemic stroke are excluded by the following (see 'Exclusion of other sources of ischemic stroke' above):

•No large vessel stenosis in the territory of the infarct by neurovascular imaging

•No evidence of atrial fibrillation on telemetry and 30-day ambulatory cardiac monitoring

•A PFO but no other high-risk source of cardiogenic embolism identified on transthoracic echocardiography (TTE) or transesophageal echocardiography (TEE)

•No hypercoagulable condition with a high risk for arterial thrombotic events, such as the antiphospholipid syndrome

●PFO assessment – The diagnostic evaluation of PFO should assess for the presence and size of a right-to-left shunting and other features suggestive of high risk for paradoxical embolism (eg, an atrial septal aneurysm or [rarely] a straddling thrombus). (See 'PFO assessment' above.)

●Evaluation for venous thromboembolism and hypercoagulable state – We evaluate for these in all patients with a PFO-associated embolic infarct and no other evident source of stroke. (See 'Search for venous thromboembolism' above and 'Hypercoagulable evaluation' above.)

●Likelihood of a PFO-associated stroke – The RoPE score (table 2) and the PASCAL classification system (table 4) are useful to estimate the probability that stroke is associated with a PFO in patients with embolic infarct topography and without other major sources of ischemic stroke (algorithm 1). (See 'Likelihood of a PFO-associated stroke' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Naser M Ammash, MD, and Robert S Schwartz, MD, and Joseph K Perloff, MD who contributed to earlier versions of this topic review.