Welcome to Stroke Education.CO.UK


Cardioembolism refers to strokes due to emboli emanating from the heart. There are several underlying reasons why this can happen. The high flow rate of blood to the brain means that it is inherently susceptible to cardioembolism. Cardioembolic strokes related to AF and PAF are very preventable with anticoagulation and so detecting AF/PAF is incredibly important. There are other less common aetiologies. Cardioembolism accounts for 20% of ischaemic strokes. Cardioembolic strokes are generally more severe and are more prone to recurrence than other stroke types. The risk of long term recurrence and mortality are high after a cardioembolic stroke. By far the most common cause of cardioembolic stroke disease is non valvular Atrial fibrillation.

Cardiac emboli tend to flow to the intracranial vessels. Over two-thirds of blood goes via the anterior circulation and most of the time thrombi occlude the proximal MCA causing often massive, superficial or single large striatocapsular or multiple MCA territory infarcts. Clinical effects will depend on the proximity of occlusion in the MCA, preexisting collateralisation, rapidly or reperfusion of the vessel, thrombus break up and distal embolisation and thrombolyis. Be aware that a large clot or a pre-existing carotid stenosis can cause thrombus to cause acute internal carotid artery subacute or total occlusion. Posterior circulation emboli can occlude vertebrals and basilar arteries as well as reaching distal posterior cerebral artery. Brainstem, cerebellar and occipital lesions can be seen which may be multiple.

Classical teaching is that embolic strokes are clinically maximal at onset. A shower of emboli or emboli breakup and distal carriage can cause simultaneous multi territory strokes. Coma may be seen with rapid improvement. It is also recorded that embolic strokes can show sudden improvement. Recanalisation and haemorrhagic transformation also commoner with embolism. Lacunar strokes especially multiple are less likely to be cardioembolic and tend to be more likely small vessel related.

Brain imaging is useful but until there are bilateral or anterior and posterior circulation large vessel stroke lesions the CT does not help greatly with aetiology. MRI in this respect can show more detail, can identify older lesions not seen on CT and show that subcortical strokes extend to cortex and more likely to be large vessel. CT may show Haemorrhagic transformation seen in up to 71% of cardioembolic strokes [J├Ârgensen L et al. 1969]. Haemorrhagic infracts are predominately cardioembolic [Fisher CM et al. 1951].

Making a diagnosis of Cardioembolism is usually supported by finding of either AF or a structural cardiac lesion. The history of examination findings of AF or rheumatic valve disease or prosthetic valves makes the diagnosis straightforward. Difficulties arise when AF is paroxysmal but in acute stroke it should be detectable acutely and if not present then holter monitoring should be considered really for at least 48 hours. There should be a very low threshold for looking for PAF or sick sinus syndrome both of which are associated with embolism. Transthoracic echocardiogram can detect the typical findings of mitral stenosis or dilated cardiomyopathy and other structural ventricular diseases. Left ventricular thrombus, valve vegetations, or atrial myxomas can also be detected. Left atrial size and left ventricular systolic function can be assessed. In younger patients or when valve or other structural lesions need better defined then Transoesophageal echocardiogram is used. Diagnostic views of aortic arch, ascending aorta, left atrium and left atrial appendages, intra-atrial septum, pulmonary veins, and valve vegetations can be seen. Any PFO or atrial septal aneurysm can be detected and bubble contrast may be given to look for shunting.

Echocardiography in the first hours after the onset of stroke is necessary only in rare cases, such as if infective endocarditis is suspected. In the days thereafter, transthoracic echocardiography or, preferably, transoesophageal echocardiography can be undertaken to support or refute a possibility of cardioembolism.

Causes and Risk of Cardioembolic stroke [Tudor G. et al. 2008]

High RiskLow or uncertain Risk
Atrial FibrillationPatient Foramen Ovale
Sustained Atrial FlutterAtrial septal aneurysm
Sick Sinus SyndromeSpontaneous atrial contrast
Left Atrial ThrombusMitral Valve prolapse
Left Atrial Appendage ThrombusCalcific aortic stenosis
Left Atrial MyxomaFibroelastoma
Mitral StenosisGiant Lambel Excrescence
Mechanical ValveAkinetic Ventricular wall segment
Infective EndocarditisSubaortic hypertrepophic cardiomyopathy
Non Infective EndocarditisCongestive cardiac failure
Left Ventricular Myxoma
Recent Anterior MI
Dilated Cardiomyopathy

Structual Heart Disease as a source of Cardioembolism

These include any condiution that allows stasis and the formation of thrombi. Severely Impaired LV function, LV aneurysm formation. Atrial or Ventricular myxomas. There may be coexisting AF. Others include Cardiomyopathy, Mitral valve prolapse and prosthetic heart valves.

Cardiomyopathy: Risks of cardioembolism may be seen in those with very poor LV function and those with underlying Chronic AF with enlarged LA.

Mitral Valve Prolapse: Studies have shown that mitral valve prolapse (MVP) is associated with fibrinous deposits on the valve, endothelial denudation and annular thrombus at the junction with the atrial wall. The myxomatous redundant valve leaflets appear to increase the predisposition to thromboembolic events. There is no evidence for antithromotic treatment other than antiplatelets in those who have experienced stroke. There should be a search for undetected PAF.

Mitral annulus calcification: Is seen with mitral stenosis, mitral regurgitation and may be increased in those with cardiogenic brain embolism. Detected on echocardiography it suggests a twofold increase in risk for stroke. Embolism of fibrinated cell clot or calcium spicules has been reported.

Prosthetic Heart Valves: Diseased and damaged heart valves are often replaced with either mechanical or bioprosthetic (tissue) valves. Tissue prosthetic valves are believed to be associated with a smaller risk of thromboembolism than mechanical valves. Mitral valve prostheses are associated with a greater risk of thromboembolism than Aortic, possibly because of the higher incidence of atrial fibrillation and other thromboembolic risk factors in these patients.

Paradoxical Embolism

The pulmonary circulation provides a large filter for any thrombi returning from the systemic circulation. A small microthrombus to the lungs is probably insignificant but if the pulmonary circulation can be bypassed and shunted then there is a risk of thrombi entering the systemic arterial circulation. This can happen when there are areas allowing right to left shunting in the heart such as a patent foramen ovale or Atrial septal defect or Ventricular septal defect or other form of congenital heart disease. However clots bypassing the lungs can be seen in those with pulmonary arteriovenous malformations.

To reliably diagnose paradoxical embolism I would suggest you need to try to prove

  • A stroke that is Radiologically cardioembolic - i.e. large artery and perhaps multiple territories
  • Evidence of a DVT at the time - Doppler USS or at least a raised Dimer
  • Evidence of a shunt with Right to left flow
  • It helps if there is an underlying venous procoagulant state e.g. malignancy, pregnancy etc.

Management is difficult. First of all any PFO which is seen in 25% of the population may simply be a bystander and association does not mean causation. Closure of PFOs is not risk free and may even increase long term risk of (P)AF and so limits any reduction, cancels or even increases any risks of stroke. Unfortunately the interventionalist have not proven their case. All those having PFO closure should be done as part of a proper clinical trial.

Top of Page