Multimodality Cardiac Imaging in Ischemic Stroke: Insights into the Heart-Brain Interaction

Awareness of the heart-brain interplay and knowledge of cardiac disorders leading to brain injury are of utmost importance. Clinicians need to recognize a wide range of cardiac disorders that may cause ischemic stroke. Cardioembolic strokes are commonly seen when an intracardiac thrombus is formed in the left atrium (LA), left atrial appendage (LAA), or in patients with atrial arrhythmias. Occasionally, thrombi may also be found in the left ventricle (LV) in patients with cardiomyopathies, Takotsubo syndrome (TTS), myocarditis, or acute myocardial infarction (AMI), or originating from heart valves in patients with valvular disease or cardiac masses. In some cases, a patent foramen ovale (PFO) acts as a passage where paradoxical embolisms bypass directly from the venous system into the systemic arterial circulation.

Pathology of the heart may also impact the brain in a more subtle manner, i.e., predispose to vascular cognitive decline and dementia, mostly by repetitive micro- or macroembolism to the brain, which is the second most common subtype of dementia in Europe and North America [1]. The proposed mechanisms of this association include recurrent silent cerebral infarctions, cerebral microbleeds, cerebral hypoperfusion, inflammation, and genetic factors [2]. Heart failure may induce brain injury through neurohormones, inflammation, and nutritional disturbances, which may subsequently cause gray matter atrophy and cognitive dysfunction [3]. Hypertension is regarded as the most important risk factor for cardiovascular and cerebrovascular events such as myocardial infarction, ischemic and hemorrhagic stroke, as well as subclinical arterial disease and brain damage (atherosclerosis; arterial remodeling; and small vessel disease, including the combination of white matter lesions, lacunar infarcts, microbleeds, and enlarged perivascular spaces) [4, 5]. In particular, systolic blood pressure is directly related to stroke and stroke-related mortality [6]. Almost all randomized controlled trials have shown a consistent benefit of optimal antihypertensive treatment with regard to the prevention of strokes in hypertensive patients (office blood pressure exceeding 140/90 mm Hg) after previous stroke or transient ischemic attack (TIA) [7].

Failure to identify cardioembolic sources of stroke and other cardiac disorders predisposing for neurological damage substantially increases the risk of mortality and morbidity. Clinicians need to thoroughly pursue the underlying cause of stroke. Clinical suspicion is often necessary to detect cardiac thrombi as it may not be evident on routine cardiac imaging or other diagnostic tests.

In the first part of the present review, we will focus on the importance of multimodality cardiac imaging in the post-stroke screening in terms of etiological factors and summarize cardiac disorders related to ischemic stroke, with a particular focus on PFO as a culprit for paradoxical embolism. Other rare but important cardiac conditions that increase stroke risk and traditionally have received little attention, such as cardiac amyloidosis, Fabry disease (FD), and LA abnormalities, will be discussed in detail. Importantly, the heart is not always the source of brain damage but can per se be damaged by a stroke leading to neurogenic stunned myocardium, transient LV dysfunction, and arrhythmias. Hence, in the second part of this review, we will discuss the brain-heart axis.

In patients with ischemic stroke or TIA, identifying and treating the underlying etiology is crucial to prevent recurrence. The initial diagnostic cardiac workup in these patients includes a 12-lead ECG, extended or long-term heart rhythm monitoring, and transthoracic echocardiography (TTE), followed by transesophageal echocardiography (TEE) in selected cases. Depending on the clinical situation, cardiac computed tomography angiography (CCTA) and cardiac MR (CMR) may also provide complementary information when indicated. Although not routinely used, ambulatory blood pressure monitoring and arterial stiffness assessment by pulse wave velocity can also be useful to identify treatable risk factors, especially in young stroke survivors who are more likely to have early vascular aging [8, 9]. Early recognition and modification of underlying vascular risk factors in these patients are essential in order to reduce the rate of recurrent strokes and vascular events [9]. Furthermore, 24-h ambulatory blood pressure monitoring improves diagnosis of hypertension and may reveal masked hypertension and non-dipping blood pressure patterns, not only in hypertensive patients but also in normotensive stroke survivors [10]. Interestingly, non-dipping pattern of blood pressure has also emerged in young patients with cryptogenic stroke, who otherwise appeared to have relatively favorable cardiovascular health state [11].

Echocardiography plays a crucial role in diagnosing cardiac and aortic sources of embolism in patients with ischemic stroke and TIA. In some cases, the intracardiac thrombus may be directly visualized by echocardiography, but usually it reveals conditions that are associated with the risk of cardioembolic stroke. Standard TTE is recommended in all patients after ischemic stroke [12]. The use of TEE is recommended in patients aged under 50–55 years and in cases where there is a suspicion of cardioembolic source and TTE is not sufficient [13, 14]. This is often dependent on the clinical assessment by the attending/treating neurologists, who usually have already performed a pre- and transcranial ultrasound, including a bubble test when the patient´s history is suspicious of paradoxical venous embolism [15]. If endocarditis is suspected, additional cardiac imaging such as TEE and PET-CT imaging may be necessary to confirm the diagnosis, assess complications, and monitor the clinical course. In general, TTE is superior when assessing the ventricles, while TEE is particularly useful for visualization of the LAA, native and prosthetic valves, PFO, atrial septal aneurysm, cardiac tumors, and complex aortic plaques (ascending, arch, descending aorta). In stroke secondary to non-valvular atrial fibrillation (AF), the LAA is the location of thrombus 90% of the time [14], and TEE or CCTA is often necessary for its visualization (Fig. 1).

Echocardiography can also provide insight into other low-risk features associated with stroke, such as mitral annular calcification (MAC), a finding which is often observed in patients with hypertension and aortic stenosis. In a study by De Marco et al. [16], MAC was identified as an independent predictor of incident ischemic stroke in treated hypertensive patients with ECG LV hypertrophy (LVH). Spontaneous echo contrast is also associated with stroke and is found in 2% of unselected TTE series. It is caused by erythrocyte aggregation and erythrocytes under low-flow conditions, and is associated with a dilated LA, mitral valve disease, reduced cardiac function, and hypercoagulable states. Intracardiac tumors are also related to ischemic stroke and may lead to thrombus formation with subsequent embolization or to embolization of neoplastic material (Fig. 2).

When performing echocardiography in stroke patients, it is also important to visualize the ascending aorta and aortic arch to look for atherosclerotic plaques, aneurysms, dissections, and intramural hematomas (Fig. 3). Particularly, aortic arch atherosclerosis with high-risk features such as thick plaques (≥4 mm), ulcerated plaques, or mobile thrombi should be looked for during TEE [17]. Plaque thickness ≥4 mm in the ascending aorta or aortic arch has been shown to correlate with cerebral embolic events.

Dilated LA (LA maximum volume index ≥34 mL/m²) is an independent predictor of heart failure, AF, ischemic stroke, and death [18]. In addition to LA maximum volume index, other echocardiographic measures can be obtained, including LA minimum volume and LA pre-atrial systolic volume. Calculated parameters include LA volume change (minimum to maximum), reservoir volume (LA volume reduction in early ventricular diastole), and LA stroke volume during atrial systole. Furthermore, conduit volume can be calculated as a difference between LV stroke volume and the sum of LA reservoir and stroke volumes [19, 20]. When using transthoracic three-dimensional (3D) echocardiography for data acquisition and analysis, the LA phasic volume cycle can be readily obtained. Data from a recent SECRETO sub-study showed that impaired LA function, LA myopathy, and increased LA stiffness, reflecting early mechanical dysfunction (before this can be detected by conventional measures of LA volume and function), may substantially increase the risk of cryptogenic ischemic stroke, particularly in patients <50 years of age [21]. Furthermore, in patients with newly detected AF, assessment of LA reservoir function by strain imaging and total atrial conduction time derived from tissue Doppler imaging improved risk stratification for stroke independent of the CHA2DS2-VASc score [22].

Which atrial parameters are most predictive of future events or determine prognosis is still under investigation. However, many of these parameters are readily available with echocardiography.

Cardioembolic strokes account for about 20–30% of all ischemic strokes [23], but the actual rate may be higher as there is a high proportion of strokes of unknown cause [24], especially in young patients. Cardioembolic strokes typically cause more severe brain injury and affect multiple brain artery territories than other ischemic stroke subtypes. They also have higher recurrence rates and mortality compared to stroke due to large-artery atherosclerosis or small vessel disease [23, 25]. Patients with embolic stroke of undetermined source have a recurrence rate of 4–5% per year [26]. Common cardiac abnormalities associated with stroke are listed in Table 1. Among these, the most common is AF, followed by AMI, cardiomyopathies, and cardiac masses [27]. Overall, cardiac pathologies range from conditions where the causality may be easily determined, such as cerebral embolism after thrombus formation in post-AMI akinetic myocardium, to conditions where it may be impossible to determine a causal relationship, such as the presence of MAC.

AF is associated with a 4- to 5-fold increased stroke risk depending on demographics and comorbidities. Approximately 20–30% of patients admitted with ischemic stroke are diagnosed with AF after the initial diagnostic workup [28], and these strokes appear to carry a high mortality risk [29]. Multiple trials have demonstrated the efficacy of anticoagulation in reducing the risk of stroke by 65%–80% in patients with clinically detected AF and other risk factors [30‒32]. The recent European Society of Cardiology (ESC) guidelines on AF recommend anticoagulation in patients with CHA2DS2-VA scores ≥2 points and that anticoagulation should be considered in patients with ≥1 point [28]. Since different clinical predictors do not contribute equally to the individual risk of developing AF, patients with a score <2 have a broad range of risk. Hence, the CHA2DS2-VA score may be flawed in certain clinical scenarios, and it is therefore recommended to also account for additional factors that may modify an individual’s stroke risk, such as the presence of cancer, chronic kidney disease, certain ethnicities (Asian, Hispanic, Black), cardiac biomarkers, atrial enlargement, hyperlipidemia, smoking, and obesity [28].

Atrial pathology may be associated with stroke even in the absence of AF [33, 34] and is more prevalent in patients with cryptogenic stroke, possibly independent from the presence of AF [35]. Both echocardiographic and electrocardiographic findings, such as increased size (volume), mitral regurgitation, and high-voltage P-wave terminal force in lead V1, may predict stroke risk [18, 36‒38]. However, the clinical benefit of using markers of atrial cardiopathy to guide anticoagulation has not been shown [39], and the value of including such markers in risk scores guiding anticoagulation thus remains limited. Risk scores must also balance precision and simplicity, and all known risk factors cannot be included. In patients in whom TEE is contraindicated or not available, cardiac CT is an excellent option for visualization of the LAA (Fig. 1) and may detect LAA thrombus with a sensitivity and specificity close to 100% [40]. LAA morphologies can also be easily assessed with CT, and different morphologies are associated with different stroke risk [41]. CT also allows visualization of the entire aorta and excludes classical dissection, as well as intramural hematoma and thrombi (Fig. 3).

AMI may also give rise to thrombus formation, particularly in the presence of LV dysfunction and subendocardial or transmural injury, especially when involving the anterior wall [42]. TTE has a lower sensitivity when it comes to diagnosing LV thrombi, but fresh, protruding, and mobile LV thrombi are usually easily diagnosed by TTE (Fig. 4a, b) [43]. Older mural thrombi may require the use of ultrasound contrast, cardiac MRI, or CCTA to be diagnosed (Fig. 4c, d) [44, 45]. Additionally, the LV apex may not always be adequately visualized on TTE. In general, CMR is more accurate for detecting LV thrombus compared to TTE and TEE [46, 47], with a sensitivity of 82%–88% and specificity of almost 100% [48], but lower availability and higher costs may limit its widespread use. With CMR, when a complete dataset is acquired, a set of LV short-axis cine images from LV apex to the roof of the LA is obtained. With this one can also analyze phasic volumes and Gadolinium wash in and wash out [49, 50]. Also, venous-phase gated cardiac CT can be used as an alternative to detect LV thrombus. In a study of 129 patients with stroke of unknown etiology, cardiac CT detected LV thrombus in 10.1%, and 8.2% were diagnosed with intra-cardiac thrombus that would have been missed on routine examinations [45]. Cardiac CT with delayed enhancement is also utilized to detect myocardial scarring, and it is anticipated that in the future, CCTA could provide a comprehensive imaging approach for assessing coronary anatomy, myocardial function, scarring, and intracardiac thrombus [51, 52]. The reported incidence of LV thrombus increases during the first 2 weeks after MI [53]. Accordingly, the ESC guidelines suggest delayed imaging at 2 weeks for high-risk patients [54].

Increased myocardial wall thickness on echocardiography is a common finding in stroke patients and should prompt further diagnostics to determine the etiology. A study enrolling 206 consecutive patients with stroke and TIA found LV hypertrophy in 42% of patients [55]. Most commonly, increased wall thickness or hypertrophy is caused by hypertension, obesity, metabolic syndrome [56], and valvular pathology (such as aortic stenosis). However, other less common etiologies, such as hypertrophic cardiomyopathy (HCM), cardiac amyloidosis, and FD, are also highly important to consider.

Cardiac amyloidosis is not included in the recommendations for post-stroke echocardiographic evaluation, but is now increasingly recognized due to increased awareness and novel diagnostic and therapeutic options. A Mayo Clinic autopsy study demonstrated that subjects with light-chain amyloidosis (AL-amyloidosis) had significantly more intracardiac thrombi (51% versus 16%) and more fatal embolic events (26% versus 8%) than those with other types of amyloidosis, even though they were younger and had less AF [57]. Amyloid deposits in the heart lead to increased myocardial wall thickness, reduced ventricular function, and cardiac arrhythmias, including AF. Atrial and LAA thrombi can even occur in sinus rhythm in patients with amyloidosis [58]. In amyloidosis patients undergoing electrical cardioversion, the rate of intracardiac thrombi was markedly higher compared to controls, despite anticoagulation [59]. Generally, cardiac amyloidosis should be suspected in patients >65 years with LV wall thickness >12 mm, or with at least one “red flag,” including bilateral ventricular hypertrophy, carpal tunnel syndrome, spinal stenosis, or polyneuropathy [60]. Particularly the wild-type transthyretin amyloidosis (ATTR cardiac amyloidosis) can also be found in younger patients, and a high degree of suspicion may be necessary for its detection. Common echocardiographic findings (Fig. 5a, b) include heart failure with preserved ejection fraction (HFpEF), biventricular hypertrophy, sparkling myocardium, biatrial enlargement, pericardial effusion, thickening of the interatrial septum, and apical sparing strain pattern “cherry-on-top” with distinctive strain curves showing a base-to-apex gradient (Fig. 6). Treatment depends on the underlying cause; ATTR cardiac amyloidosis is treated with drugs such as tafamidis, stabilizing transthyretin, preventing its dissociation into monomers, and subsequent formation and deposition of amyloid fibrils in the heart, whereas AL-amyloidosis is generally treated with chemotherapy. ATTR- and AL-amyloidosis are usually easily differentiated by blood tests (including serum electrophoresis and serum free light chains) and technetium-99 m pyrophosphate scintigraphy. However, endomyocardial biopsy may be necessary in selected cases and is still considered the diagnostic gold standard. The prevalence of AF in amyloidosis is as high as 40%, and anticoagulation is recommended regardless of CHA2DS2VA score in the presence of AF [28, 61]. The same holds true for another disorder characterized by increased wall thickness, namely FD, where the CHA2DS2-VA score also has limited predictive capabilities. FD is a rare X-linked disorder where glycosphingolipids accumulate in several organs, including the heart and blood vessels [62]. The prevalence of FD in stroke populations ranges from 0.0 to 4.9% [63‒66], but may be present in 4–5% of men with unexplained LVH [67] and in 3–5% in those with cryptogenic stroke [68]. According to analysis performed on the Fabry registry (fabryregistry.com), including 2,446 patients, stroke occurred in 6.9% of men and 4.3% of women. In these patients, 87% of strokes were ischemic and 13% hemorrhagic [69]. Due to the involvement of both the heart and blood vessels, FD may be associated with both ischemic stroke and intracerebral hemorrhage. FD may present with a spectrum of clinical manifestations and often diffuse symptoms leading to delayed diagnosis. Extracardiac manifestations include renal failure, gastrointestinal symptoms, neuropathic pain, corneal opacities, hypohidrosis, hearing loss, and telangiectasias. Cardiac disease involves increased myocardial wall thickness (Fig. 5c, d), cardiac arrhythmias, valvular thickening and dysfunction, coronary artery disease, and aortic root dilatation [62]. Cerebrovascular manifestations in FD include stroke, vascular abnormalities (vertebrobasilar dolichoectasia and vessel tortuosity), epilepsy, cognitive decline, and depression [62]. Common MRI findings in FD have been reported as white matter hyperintensities, lacunar and cortical/territorial infarcts, cerebral microbleeds, intracerebral hemorrhage, and the Pulvinar sign. In daily clinical practice, it is important to have FD as an actual differential diagnosis when a patient presents with unexplained LVH, stroke at a young age, and renal failure of unknown etiology [62].

Another important entity to be considered is HCM, present in about 1:500 individuals. In HCM, the presence of AF also warrants anticoagulation regardless of the conventional CHA2DS2-VASC score [70]. When evaluating patients with increased myocardial wall thickness, it is important to appreciate that different LV geometry patterns are associated with different stroke risks, and concentric LVH and concentric remodeling seem to be associated with the highest stroke risk [71]. This may be in part due to its relationship with hypertension as a confounder.

Non-ischemic cardiomyopathies with LV dilation and dysfunction also predispose for LV thrombus formation and cerebral embolization [72], although less common compared to ischemic heart disease [48]. Most commonly encountered cardiomyopathies include dilated cardiomyopathy, arrhythmogenic cardiomyopathy, and HCM. Other cardiac conditions, such as LV non-compaction [73], are also associated with thrombus formation due to the trabeculated myocardium and intertrabecular recesses predisposing for blood stasis and subsequent thrombus formation. Thrombus formation is especially profound if non-compaction is associated with LV scar or decreased systolic function [74]. The trabeculations are typically seen in the apex and middle segments of the inferior and lateral walls [73]. Although conventional TTE or contrast echocardiography is the first-hand modality, CMR is often needed to confirm the final diagnosis. Especially, CMR is useful in differential diagnosis between HCM, non-compaction, and arrhythmogenic right ventricular cardiomyopathy and can also be used for tissue characterization of myocardium (low native T1 in Fabry and high native T1 in cardiac amyloidosis) and cardiac tumors, as well as thrombus following AMI (Fig. 4d). Notably, the exact role of non-compaction as an embolic source is not well established, albeit studies have found even minor non-compaction to be associated with prolonged gadolinium washout in the LV and increased stroke risk [49, 75].

PFO is categorized as a minor risk factor for stroke in the current recommendations [44]. The prevalence of PFO in the general adult population is estimated to be 20–25% based on autopsy studies [76] and around 30% in studies with living individuals [77]. Case-control studies have shown that PFO is more common in young and middle-aged patients with cryptogenic stroke compared with matched controls [78] and is implicated in approximately 5% of all ischemic strokes and 10% of strokes in young and middle-aged adults [79]. Assessment of intracardiac right-to-left shunt (RLS) is normally done by TEE bubble study using agitated saline with and without blood. However, the addition of blood to agitated saline may substantially improve the sensitivity of transcranial Doppler for the detection compared with other conventional contrast agents [80].

In PFO, the shunt is normally directed from the high-pressure (LA) to the low-pressure chamber (RA). However, in situations where the shunt is diverted from the right-to-left direction (acute deep venous thrombosis or pulmonary embolism, recent prolonged travel and immobilization, or Valsalva maneuver preceding the ischemic stroke event), the PFO may act as a mediator for a paradoxical embolism, and the thrombus is not developed in situ in the PFO structure or elsewhere in the heart [13]. Hence, in these cases a PFO-associated stroke should be dealt with as its own major category among the cardioembolic sources, separate from the traditional classification of high, medium, low-, and uncertain-risk sources depending on the presumed likelihood of being the cause of the stroke. The difficulty thus lies in determining the clinical relevance and causality of the PFO or ASD in the patient with cryptogenic stroke. This may be obvious in young patients with cryptogenic stroke and large RLS, high-risk echocardiographic features (hypermobile atrial septum, RLS at rest, tunnel-like PFO (≥10 mm), prominent Eustachian valve, or Chiari’s network in the right atrium, as well as a sharp angle between the inferior vena cava and PFO). However, this is rarely the case. Certain PFO features, such as atrial septal aneurysm, were found to be the only ancillary echocardiographic feature predicting recurrent stroke in a pooled analysis of two observational and two randomized trials [81]. A transcranial Doppler ultrasound with bubble test is a useful screening tool and ancillary examination to quantify the magnitude of RLS in PFO, with a sensitivity and specificity of 97% and 93%, respectively, for detecting PFO compared with TTE [82]. It may provide an alternative method for diagnosing a probable PFO in patients with swallowing difficulties, contraindications for or poor visibility in TEE, or problems with performing the Valsalva maneuver. However, transcranial Doppler ultrasound with bubbles does not directly visualize the PFO, and the PFO alone or combined with other septal abnormalities can thus only be diagnosed by cardiac imaging. Also, high-resolution optical coherence tomography can detect in situ thrombi and abnormal myocardium within a PFO. In a study of 117 patients with PFO, in situ thrombi were found in 84% of stroke patients, compared with 57% in the migraine group and 0% in asymptomatic individuals [83]. Detecting in-situ thrombus could influence antithrombotic therapy or PFO-closure decisions, but with the risk of an invasive procedure, i.e., procedure-related embolization. Further research is needed to determine how, and if, this technology can be integrated into clinical practice.

Risk scores such as the Risk of Paradoxical Embolism (RoPE) score and PASCAL classification are designed to aid in determining the likelihood of a PFO being clinically relevant [79]. When a PFO with high-risk features is detected and other causes are excluded, PFO closure is recommended after a joint decision by a neuro-cardiology team [84]. In the absence of high-risk features, the benefit is less clear, and in patients >60 years, PFO closure is generally not recommended. Also, the presence of PFO in young patients with clinically silent infarcts that appear on MRI scans, performed for headaches or other complaints, can pose a major management dilemma. Complications after PFO closure occur in about 4.5% and include atrial arrhythmias (most commonly AF), infective endocarditis, bleeding thrombus formation on the device, and device embolization/dislodgment. Follow-up for arrhythmias has been limited in all PFO closure trials, and the AF rate is about 5% in PFO closure versus 1.6% in patients receiving medical/non-invasive treatment [85]. Informed decision by a team comprising cardiologists and neurologists is thus of great importance, especially when the cause is less clear or other causes in addition to PFO are present [84].

To conclude, post-stroke cardiac imaging is no longer limited to echocardiography, but also includes novel imaging techniques such as CMR and cardiac CT. To prevent future ischemic strokes, especially in younger individuals, it is crucial to detect and treat direct and indirect cardioembolic sources. Clinicians must be familiar with an extensive number of cardiac disorders related to stroke, some more important than others, and also to evaluate the possibilities that novel cardiac imaging may offer. Furthermore, a PFO has traditionally been concidered as a minor or uncertain risk factor for ischemic stroke. However, PFO-associated strokes are a distinct category among the cardioembolic sources as their clinical significance is dependent on different associated features and other stroke risk factors.

The bidirectional interplay between the heart and the brain in its physiological state involves the autonomic nervous system as well as endocrine pathways. In central nervous system (CNS) injury and stress, such as stroke, trauma, tumors, and epilepsy, these pathways and inflammatory mechanisms may cause neurogenic cardiac injury.

In post-stroke patients, neurogenic cardiac injury can lead to a number of pathological changes, including ECG changes (most commonly ST-segment elevation), arrhythmias, acute coronary syndromes due to severe coronary vasoconstriction, LV dysfunction, including stress cardiomyopathy (TTS), and in severe cases cardiac arrest [86, 87]. This is often referred to as the stroke-heart syndrome. The increased sympathetic activity/drive with catecholamine release from sympathetic nerve fibers and adrenal glands subsequently causes myocardial calcium influx and produces an excitatory effect on the myocardium that in some cases leads to catecholamine-induced cardiac injury [88], also called neurogenic stunned myocardium. Increased sympathetic tone and elevated concentrations of noradrenaline can also make coronary plaque more susceptible to rupture. However, according to a meta-analysis of 131,299 patients, the incidence of acute coronary syndromes after stroke was only 1.7% per year [89]. Additionally, LV dysfunction can occur following a stroke and includes clinical heart failure [90] and TTS. The LV dysfunction in TTS is reversible, typically presenting as an apical-midventricular ballooning pattern, but various subtypes, including midventricular and focal types, may occur. By contrast, TTS with subsequent akinetic regions within the myocardium can lead to thrombus formation and stroke. For this reason, patients with TTS should always be assessed for the presence of a thrombus. TTE is the first-line imaging modality for suspected LV dysfunction, but cardiac MRI can provide additional information, in particular to confirm myocardial edema using STIR (Short Tau Inversion Recovery) imaging in TTS and myocarditis and for detecting myocardial scarring or fibrosis associated with myocardial infarction with LGE (Late Gadolinium Enhancement) [91].

Cardiac arrhythmias can be seen in 21.9% of patients with ischemic stroke and in 14.8% with intracerebral hemorrhage [92]. In particular, insular strokes may be associated with an increased risk of arrhythmias due to cardiac autonomic dysregulation [93]. Similarly, it is well known that sudden death may occur in epilepsy despite being well controlled, and “neurogenically provoked” cardiac arrhythmias have been implicated as a potential mechanism [94]. Parasympathetic nerve fibers innervating the heart may also cause bradycardia, atrioventricular blocks, and possibly asystole, as demonstrated in rat studies where insular stimulation led to asystole [95].

Neurogenic pulmonary edema can occur as a result of cerebral injury [96] and stroke. The mechanism of neurogenic pulmonary edema is not fully understood, but it is believed that increased sympathetic activation leads to systemic vasoconstriction and increased blood accumulation in the pulmonary vasculature. Increased pulmonary capillary permeability then causes fluid to accumulate in the pulmonary interstitium and alveoli, resulting in acute respiratory distress [97]. Neurogenic pulmonary edema may also occur following thrombolysis and brain reperfusion. Treatment of neurogenic pulmonary edema involves addressing the underlying neurological cause, as well as supportive measures, and may eventually require mechanical ventilation or treatment with extracorporeal membrane oxygenation (ECMO) [97]. The recognition of cardiac disorders occurring in neurologic patients because of neurogenic cardiac injury is therefore of utmost importance, as failure to do so may have deleterious consequences.

In serious neurologic conditions caused by large ischemic strokes, brain injury, intracranial bleeding, tumors, or CNS infections, increased intracranial pressure may lead to impending brain herniation and a decrease in cerebral perfusion pressure. That, in turn, activates the Cushing reflex that leads first to sympathetic nervous system activation (hypertension, tachycardia) and second to compensatory parasympathetic activation (decrease of heart rate). When the intracranial pressure continues to rise, the condition leads to brain-stem dysfunction, causing irregular breathing and breathing cessation [98, 99].

Cardiac biomarkers like troponins and natriuretic peptides, including N-terminal pro-B-type natriuretic peptide (NT-proBNP), are often elevated after stroke and may reflect cardiac disorders or the physiological response to stroke. To detect myocardial injury in stroke patients, the American Stroke Association therefore strongly recommends routine measurement of cardiac troponins [100].

Advances in cardiovascular imaging have significantly enhanced the detection of cardiac disorders associated with stroke. Clinicians involved in post-stroke workup must be familiar with a range of cardiac disorders associated with stroke and be familiar with the capabilities offered by different imaging modalities for higher individual diagnostic accuracy in order to effectively treat and prevent stroke recurrence.

The authors thank Dr. Terje Larsen for providing cardiac MR images and Dr. Abukar Mohamed Ali for CT images. The graphical abstract was created with BioRender.com. Myrmel, G. (2025) https://BioRender.com/e56g141.

This work was based on the published literature and images from our own imaging laboratories, obtained as part of routine clinical care. Ethical approval for the use of these samples for research purposes was not required in accordance with local guidelines. Written informed consent was obtained from all patients for the publication of the accompanying images.

The authors have no conflicts of interest to declare.

The work was not funded.

G.M.S.M. wrote the first draft, which was critically revised by J.P. and S.S. S.S., U.W.A., and J.P. contributed to data acquisition. U.W.A., J.P., J.S., and V.J. revised the manuscript. All authors approved the final submission.

Jukka Putaala and Sahrai Saeed contributed equally to this work.

All data generated or analyzed during this review are included in the article. Further inquiries can be directed to the corresponding author.