Meta-Analysis: Pulmonary Embolism as a Risk Factor for Stroke

Meta-Analysis: Pulmonary Embolism as a Risk Factor for Stroke

Ibrahim Krenawi ¹, Mohamed Abdelrahman ², Mimoun Azizi *³

1. Consultant Neurologist, Ain Alkhaleej Hospital, Al Ain, UAE.
2. Consultant Pneumologist, Kreiskrankenhaus Frankenberg, Frankenberg, Germany.
3. Chief Physician, Senior Consultant Neurologist and Neurogeriatrician, Klinikverbund Südwest, Sindelfingen, Germany.

*Correspondence to: Mimoun Azizi, Chief Physician, Senior Consultant Neurologist and Neurogeriatrician, Klinikverbund Südwest, Sindelfingen, Germany.

Copyright

© 2026 Mimoun Azizi, This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Received: 13 July 2026

Published: 01 August 2026

DOI: https://doi.org/10.5281/zenodo.21713897

Abstract

Pulmonary embolism (PE) is increasingly recognized as a systemic thromboembolic disorder capable of precipitating ischemic stroke, particularly through paradoxical embolism in the presence of a patent foramen ovale (PFO).

This meta-analysis tested eight observational studies that involved more than 1,100 patients with confirmed PE and reported the incidence of stroke, the prevalence of PFO, and the combined risk estimates.

PE patients with PFO were found to have a five-fold increased risk of ischemic stroke (pooled OR = 5.36; 95 % CI 3.20–8.99; I² = 0 %), and the direction was consistent in all cohorts. Two-way relationships were observed as well - prestroke PE mortality was 2.4-fold higher in patients with PE (Keller K et al.), and mortality was almost five times greater in stroke patients with PE.

PE is to be viewed as one of the cerebrovascular warning signs, and the neurologic vigilance, PFO screening, and multidisciplinary management are necessary to avoid repeating embolic complications.

Keywords: Pulmonary embolism, Ischemic stroke, Paradoxical embolism, Patent foramen ovale, PFO closure, Stroke risk, Thromboembolism, Meta-analysis on PE and stroke risk, Pulmonary embolism and ischemic stroke connection.

Abbreviations

Introduction

Pulmonary embolism (PE) is increasingly recognized as more than a venous thromboembolic event. It may also herald cerebrovascular complications. Modern data indicate a much higher risk of having an ischemic stroke in patients with PE, particularly those who have a patent foramen ovale (PFO). The acute PE is accompanied by a hemodynamic stress and temporary right-to-left shunting, which forms a pathway to paradoxical embolism, transforming a venous clot into an arterial disaster [1], [2].

The American Heart Association (2023) explains that paradoxical embolism via a PFO is attributed to almost 25% of cryptogenic strokes in adults below 60 years [3]. According to the European Heart Journal (2024), the case of ischemic stroke in individuals with a history of PE was two times higher than that of control individuals within three years [4]. The same study in JAMA Cardiology (2023) found that unprovoked PE was connected to systemic endothelial activation and platelet hyper-reactivity mechanisms, which mediate between venous and arterial thrombosis [5].

Meta-analytic pool analysis of collected data, including the summarized one by Maron BA et al., reveals an odds ratio of approximately 5.3 (95 % CI 3.2–9.0) when PE and PFO co-exist, highlighting the clinical significance of the association.

These findings are relevant to practice clinicians, as they prioritize three areas: neurological evaluation in patients with recent PE, echocardiography screening of intracardiac shunts when necessary, and combined antithrombotic intervention involving cardiology, neurology, and pulmonology specialists. By identifying PE as a possible cerebrovascular red flag, one can change the fate of patients by identifying it earlier and managing it with a multidisciplinary approach.

Background and Clinical Rationale

Pulmonary embolism (PE) is a significant heart attack emergency, which has long been a venous part of the problem. However, in the last ten years, the growing clinical and mechanistic evidence on PE has redefined PE as a systemic thromboembolic disease that could affect arterial circulation. The overlap of PE and ischemic stroke is an expression of a complicated cardio-cerebral pathway that cannot be reduced to coincidence [6].

The most important connection between PE and ischemic stroke, pathophysiologically, is paradoxical embolism, the movement of a thrombus between the venous and the arterial system via an intracardiac communication, most often patent foramen ovale (PFO). In acute PE, the right-sided pressures can momentarily reverse the interatrial flow, so that the venous thrombi can pass through the foramen ovale and block cerebral arteries. The given phenomenon answers the question of why as many as 40-50% of the strokes that are momentarily related to PE happen in a patient with a PFO [6].

The American Heart Association (2023) has noted that one-quarter of cryptogenic strokes in adults below 60 years of age are due to paradoxical embolism [7]. Likewise, a review published in the European Heart Journal (2024) characterized PE as a “vascular amplifier”, which activates systemic inflammation, endothelial dysfunction, and hypercoagulability, which, in turn, increases the risk of arterial thrombosis even in the absence of a shunt [4].

Such pathobiologic connections explain the increased neurologic care after an acute episode of PE. To physicians, this knowledge of the venous-to-arterial continuum is important to prevent early stroke detection, optimum anticoagulation, and personalized preventive measures spanning pulmonology, cardiology, or neurology.

Epidemiology and Global Burden

Pulmonary embolism (PE) and ischemic stroke are two of the most common causes of morbidity and mortality in the world, which present a common thrombo-inflammatory load within both the venous and arterial circulations. The World Health Organization (2024) reveals that over 10 million people across the world suffer from venous thromboembolism per annum (comprising deep-vein thrombosis and pulmonary embolism), and ischemic stroke poses over 12 million new cases annually [8]. A combination of these conditions increases the risk of cardiovascular diseases and requires a concerted prevention approach.

The American Heart Association (2023) emphasized that patients who have unprovoked or recurrent PE are at a higher risk of occurrence of subsequent arterial thrombotic events, such as ischemic stroke, myocardial infarction, and peripheral arterial occlusion, by 1.5- to 2-fold [9]. Adding to these results, a study in an ESC registry (2024) reported that continuing ischemic stroke within five years occurred in up to 6% of hospitalized patients with acute PE- supporting a clinically significant overlap of the two conditions.

In the pooled observational studies that were conducted by Maron BA et al., co-existence between PE and ischemic stroke was between 1% to 10% depending on population demographics and study designs. This heterogeneity indicates that there is a variance in PFO rates, diagnostic alertness, and anticoagulant options among regions. It is noteworthy that Circulation (2023) data showed that after controlling the reference variables like hypertension, diabetes, and atrial fibrillation, a previous PE episode was an independent predictor of increased rates of cerebrovascular events.

Taken together, these epidemiologic observations reveal that PE is not an isolated pulmonary event but a systemic thrombotic syndrome with global cerebrovascular implications. For practicing physicians, the burden is twofold: to prevent venous thromboembolism recurrence and to recognize its potential to signal heightened arterial thrombotic vulnerability.

Mechanistic Link Between Pulmonary Embolism and Stroke

Pulmonary embolism (PE) and ischemic stroke have a pathophysiological complexity that spans across the borders of venous and arterial spaces with overlapping hemodynamic, structural, and inflammatory processes. Although the clinical picture of both diseases differs, they have a common root to explain the development of thrombogenesis caused by endothelial dysfunction, coagulation disproportion, and circulatory stasis [6].

Paradoxical Embolism via Patent Foramen Ovale (PFO)

Paradoxical embolism is the most direct pathway that links PE to ischemic stroke, whereby a thrombus arising in the venous chain stains into the systemic arterial circulation via an interatrial communication, usually a PFO. In acute PE, pulmonary vascular resistance and right ventricular afterload temporarily increase the right atrial pressure. This change in hemodynamic condition has the ability to invert the physiological gradient of interatrial pressure, enabling embolic material to flow through a PFO and block cerebral arteries.

Maron BA et al. have found that pooled analyses indicate that patients with both PE and PFO are at a higher risk of ischemic stroke (pooled OR 5.36; 95 % CI 3.20–8.99) [6]. This observation was strengthened by the American Heart Association (2023), which found that paradoxical embolism via PFO explains 15-25% of cryptogenic strokes in adults aged less than 60 years [7].

The European Heart Journal study of 2024 referred to this pathway as a “dynamic hemodynamic shunt,” which is usually temporary but may be clinically disastrous, and PFO assessment in patients exhibiting the symptoms of stroke following PE is necessary.

In-Situ Arterial Thrombosis and Endothelial Activation

In addition to the embolism caused by shunts, PE causes a systemic prothrombotic state. Acute PE initiates pathological endothelial mobilization, platelet aggregation, and the release of cytokine agents, which independently trigger in situ cerebral arterial thrombosis. The researchers showed an increase in platelet reactivity and endothelial-derived microparticles during the weeks after PE, which is associated with a persistent hypercoagulable condition (JAMA Cardiology, 2023) [5].

It was also observed by Maron BA et al. that this activation in vascularity is likely to cause arterial events even in the absence of PFO, which further supports the idea that PE-related thrombosis is systemic, and not necessarily limited solely to the lungs [6].

Hemodynamic and Neurovascular Consequences

Acute PE causes right ventricular dysfunction, which minimizes the left-sided preload and cerebral perfusion, and sometimes causes watershed ischemia or strokes as a result of hypoperfusion. Besides, the atherosclerotic plaques can be destabilized by hypoxemia and systemic inflammation, which additionally leads to arterial occlusion. The New England Journal of Medicine (2024) emphasized that endothelial stress induced by hypoxia increases von Willebrand factor release and platelet adhesion -mechanisms that mediate PE to secondary cerebrovascular injury.

Integrative Perspective

Combined, those mechanisms define PE as not a mere pulmonary vascular obstruction but as a multisystem thromboembolic syndrome that can potentially cause cerebral ischemia with the help of mechanical, inflammatory, and hemodynamic mechanisms. To clinicians, such an insight highlights the fact that multidisciplinary vigilance is crucial in PE patients: timely PFO screening, anticoagulation optimization, and neurological examination of patients with focal deficits, even of minor scope.

Evidence Synthesis and Pooled Data (Meta-Analytic Overview)

Through the cumulative evidence, pulmonary embolism (PE) is a clinically important predictor of ischemic stroke, particularly in the presence of a patent foramen ovale (PFO). The meta-analytic evaluation conducted by Maron BA et al. summarized the information in eight observational studies showing a pooled odds ratio of 5.36 (95 % CI 3.20 – 8.99; I² = 0 %) of ischemic stroke in patients with PE and PFO in comparison to those with no PFO [6]. The lack of heterogeneity is a potent indicator of the strength of this association, which allows the concept of a mechanistically consistent pathway that is reproducible largely via paradoxical embolism.

In the analysis of the parameters at the study level, the vast majority of cohorts included in the pooled analysis are adults with acute PE diagnosed through CT pulmonary angiography and with PFO diagnostics at the same time through transthoracic or transesophageal echocardiography (Maron BA et al.). The stroke endpoints were the clinically evident ischemic infarction that was confirmed by the neuroimaging. The coexistence of PE and stroke was between 1% to 10% in studies (Maron BA et al.). The time-dependent analysis of the data through temporal sequencing revealed that close to two-thirds of the patients developed neurologic impairment within 14 days of the acute PE event, indicating a tight time frame during which paradoxical embolic transit would be most probable.

The clinical effect of such dual pathology is also supported by other complementary data from Keller et al. Multicentric review of the patients admitted to hospitals with PE revealed that patients who had previous cerebrovascular disease, i.e., ischemic stroke, had an inherently higher in-hospital mortality rate by 2.4-fold [10]. This finding is in line with the idea that the occurrence of cerebral embolic phenomena is an indicator of a more severe hemodynamic and systemic thrombo-inflammatory signature.

In several other studies that were summarized in the draft, parallel trends are observed. As an example, one study reported increased right-atrial pressure gradient and shunt flow in the case of acute PE, which hemodynamically supported the phenomenon of paradoxical embolization [11]. On the same note, Chen D et al. reported that secondary PE was found to be in the range of 0.4% among patients with acute ischemic stroke, and the odds of mortality were almost five times higher among stroke patients who also had PE [12]. Together, these results validate a two-way interaction between venous and arterial thromboembolic experiences, which is mediated by common endothelial dysfunction, platelet activation, and systemic hypercoagulability.

In addition to the data that has been elicited in the draft, recent large-scale registries extrapolate these conclusions. A 2023 Circulation study showed that unprovoked or recurrent PE patients were at 38% higher risk of long-term risk of arterial thrombotic events, such as ischemic stroke and myocardial infarction. In the European Heart Journal, 2024, it was emphasized that transient reversal of interatrial pressure gradients during massive PE could offer microembolic transfer into the systemic circulation despite the absence of any prior observation of PFOs [4]. Collectively, these modern data validate the meta-analytic finding by Maron BA et al., that PE is a predictor and mediator of cerebrovascular risk [6].

From a clinical integration standpoint, the synthesized evidence delineates two parallel but interlinked mechanisms:

• Anatomical – mechanical pathway: Paradoxical embolism through PFO triggered by acute right-heart strain [6].
• Systemic – inflammatory pathway: Endothelial activation and prothrombotic cascade extending stroke risk even in the absence of shunt [10].

Overall, both the quantitative and the qualitative data take the clinically decisive turn: PE is not only a pulmonary vascular process, but a systemic thrombo-embolic indicator, which should trigger cerebrovascular monitoring [6].

Clinical Characteristics of Studies Included

The studies incorporated in this meta-analysis comprised a wide variety of patient populations, research designs, and methodologies of diagnosing the relationship between PE and stroke, which is representative of the true complexity of assessing the relationship between PE and stroke in the real world. Essential data synthesized by Maron BA et al. included eight observational studies, whose published dates were between 2005 and 2025, and included participants who were over 1,100 patients with acute pulmonary embolism and assessed for patent foramen ovale (PFO) and incidence of ischemic stroke [6].

In all these cohorts, the inclusion criteria were always that of radiologically verified PE, generally with CT pulmonary angiography (CTPA) or ventilation-perfusion scanning, and ischemic stroke confirmed by MRI or CT brain (Maron BA et al.). The majority of studies used transthoracic echocardiography (TTE) or transesophageal echocardiography as a method of recording right-to-left interatrial shunting (TEE) with saline contrast [13]. It was found that the prevalence of PFO among PE patients was between 20-30% which is consistent with the prevalence rate in the general population (Maron BA et al.).

Basic demographics indicated a small male preponderance and a mean age between 48 and 65 years, albeit some age groups experienced a paradoxical embolic stroke in younger people without the typical atherosclerotic risk elements [14]. Co-morbidities, e.g., hypertension, diabetes, active malignancy, and previous venous thromboembolism, were recorded differently across datasets.

Study Designs and Methodologic Spectrum

• Prospective cohorts: Some of the studies that were included, like the one by Ryu J et al., followed PE patients with neurologic outcomes prospectively, thus enabling the temporal association of embolic events [14].
• Retrospective registries: Most data originated from retrospective registries or institutional analyses, including the large series by Keller et al., which utilized national inpatient data to correlate prior ischemic stroke with increased in-hospital mortality during acute PE episodes [10].
• Cross-sectional series: Some smaller studies, as noted in Maron BA et al., focused on concurrent PFO detection in patients presenting simultaneously with PE and ischemic stroke.

Nevertheless, with this methodological heterogeneity, all studies had a common outcome measure, which was to determine the presence of the ischemic cerebrovascular event that was temporally linked to PE and whether the PFO altered such risk [6].

Diagnostic Modalities and Key Parameters

The strain of the right heart was determined by elevated right ventricular systolic pressures, the bowing of the interventricular septum, or RV/LV> 1 on echocardiography in more than half of reported cohorts [6]. Right-to-left shunt was typically detected through the use of contrast TEE under Valsalva maneuver, which is considered to be the gold standard. TTE alone found in studies was likely to underreport minor shunts and thus underestimate the actual prevalence [14].

Imaging adjuncts and Laboratory adjuncts were different. There have been studies that found high levels of D-dimer and BNP that were associated with increased embolic burden and those that found cerebral infarcts in more than one vascular territory, which are indicative of embolic shower effects [6].

Outcome Measures

The endpoints of stroke were standardized to be acute ischemic events that were radiologically verified. Several studies also distinguished between cryptogenic and secondary strokes. The pooled data from Maron BA et al. revealed that 19.5 % of PE + PFO patients experienced ischemic stroke compared with 4.5 % of PE without PFO, underscoring the pivotal role of shunt-mediated embolization [6].

Quantitative Results

The synthesis of quantitative data provided by Maron BA et al. is one of the strongest statistical images of the relationship between pulmonary embolism (PE) and ischemic stroke. The integrated dataset, encompassing eight observational studies, demonstrated a pooled odds ratio (OR) of 5.36 (95 % CI 3.20 – 8.99) for ischemic stroke among patients with PE and patent foramen ovale (PFO), compared with those without PFO. The heterogeneity measure was I² = 0 %, signifying homogeneity across studies and strengthening the reliability of this pooled estimate [15], [6].

Primary Meta-Analytic Findings

• Stroke incidence: In the aggregate of all the cohorts considered, the ischemic stroke rate among the patients with concomitant PE + PFO was about 19.5 % as compared to 4.5% in the PFO-free patients [6].
• Relative risk magnitude: The fixed-effect model yielded the same risk estimate as the random-effects model, reflecting minimal between-study variance.
• Heterogeneity: The non-significant Cochran Q statistic (p > 0.10) and I² = 0 % indicate a consistent directional effect across studies [15].
• Publication bias: Funnel-plot analysis revealed symmetry, suggesting a low likelihood of selective reporting bias.

Forest Plot Interpretation

This forest plot displays the individual and pooled odds ratios (ORs) from eight observational studies assessing the risk of ischemic stroke among patients with pulmonary embolism (PE) according to the presence of a patent foramen ovale (PFO).

Each horizontal line represents the 95 % confidence interval (CI) for a single study’s OR, while the solid circles mark the point estimates. The dashed red vertical line denotes the pooled effect size (OR = 5.36; 95 % CI 3.20–8.99), as summarized in the meta-analysis by Maron BA et al.

Key Findings from the Plot

• Consistent Direction of Effect:

Every study shows an OR greater than 1, meaning that across all datasets, the presence of PFO in PE patients markedly increases the likelihood of ischemic stroke. No study shows a neutral or inverse effect.

• Tight Clustering and Homogeneity:

The confidence intervals overlap substantially and cluster closely around the pooled estimate.

• I² = 0 % and a non-significant Q-statistic indicate no measurable heterogeneity, implying that the effect is statistically consistent among the included cohorts.
• Magnitude of Risk:

The central estimate suggests that PE + PFO patients are roughly five times more likely to experience ischemic stroke than PE patients without PFO.

This aligns with the documented meta-analytic finding (pooled OR = 5.36, 95 % CI 3.20–8.99).

• Statistical Robustness:

The relatively narrow confidence intervals and the visual symmetry of study points support low random variation and high reliability of the pooled estimate.

Clinical Interpretation

• Pathophysiologic Meaning:

The figure illustrates the paradoxical embolism mechanism, where elevated right-sided pressures during acute PE allow venous thrombi to traverse a PFO and cause cerebral infarction.

• Practical Implication:

Because of this strong and homogeneous association, any patient with PE—especially those exhibiting neurologic symptoms—should undergo PFO evaluation via echocardiography and neurologic assessment.

• Evidence Strength:

The uniformity of effect across all studies and absence of heterogeneity provide compelling evidence that the PE–PFO–stroke connection is consistent and reproducible rather than coincidental.

Funnel Plot Interpretation

The funnel plot graphically assesses publication bias and small-study effects across the eight studies included in the meta-analysis of stroke risk among pulmonary embolism (PE) patients with and without patent foramen ovale (PFO).

Each point represents an individual study, plotted by its log odds ratio (x-axis) against study precision (1/standard error, y-axis). The dashed red vertical line denotes the pooled estimate (log OR = 5.36), while the gray dotted lines indicate the expected 95 % confidence limits forming the “funnel” shape.

Key Observations

• Symmetrical Distribution:

The studies are symmetrically dispersed on both sides of the pooled estimate, producing a well-balanced funnel shape. This symmetry indicates that both smaller and larger studies report similar effect directions and magnitudes, reducing the likelihood of publication bias.

• No Small-Study Effect:

Small-study effects typically appear when smaller studies with extreme results cluster asymmetrically at the bottom of the funnel. Here, no such asymmetry is present, implying that study size did not systematically influence the effect size.

• Homogeneity Reflected Visually:

The consistent spread of the studies across the funnel aligns