In order to bring you the best possible user experience, this site uses Javascript. If you are seeing this message, it is likely that the Javascript option in your browser is disabled. For optimal viewing of this site, please ensure that Javascript is enabled for your browser.
Did you know that your browser is out of date? To get the best experience using our website we recommend that you upgrade to a newer version. Learn more.

We use cookies to optimise the design of this website and make continuous improvement. By continuing your visit, you consent to the use of cookies. Learn more

Assessing stroke in patients with atrial fibrillation

An article from the e-journal of the ESC Council for Cardiology Practice

Be reminded here how observation of symptoms can help determine location of the embolus or emboli, involvement of anterior or posterior circulation, and what diagnostic work-up will establish the cause-effect relationship between left atrial appendage and cardioembolic (ischemic) stroke. A primary target is to initiate a reperfusion therapy no later than 60 minutes from hospital admission - provided patient is eligible. The next important step is to start anticoagulant therapy in the hospital phase because only initiation of anticoagulant therapy in hospital can reduce the relative risk of recurrent ischemic embolic stroke up to 80%.

Atrial Fibrillation


The risk of stroke is inherent to atrial fibrillation and culminates in a patient's elderly years.  While valvular atrial fibrillation risk of stroke is most often related to its rheumatic origin and therefore has a specific time onset, non-valvular AF risk of stroke is related to underlying pathophysiological mechanisms which usually each have separate time onsets. Arterial hypertension and systemic changes in haemostatic functions may differ in their courses yet interaction between these culminate during a patient's elderly years. Virchow's triad are the three categories of factors contributing to thrombogenesis that might be detected in elderly patients with atrial fibrillation. These are 1) vessel wall abnormality (atrial and left atrial appendage dilatation, endocardial denudation and fibroelastic infiltration) 2) Abnormalities of blood constituents (platelet and hemostatic activation) and 3) Flow abnormalities (reduced left atrium appendage velocity, spontaneus echo - contrast on transoesophageal echo). 
The provocative events for which an embolus from the left atrial appendage (LAA) follows the bloodstream to a part of the arterial bed - which can be any - and causing a stroke or systemic embolism still remain unidentified, (see here the case of a disappearing left atrial appendage thrombus captured by echocardiography, in a patient with atrial fibrillation, resulting in a stroke).
Strokes occur up to 10 times more often than systemic embolism due to the fact that small thrombi cannot bypass the carotid orifice - these thrombi in strokes eventually lodge in the cerebral circulation as a result of the hydrodynamic, anatomic and physical factors related to atrial fibrillation (1,2).
Aproximately 80 % of emboli primarily lodge into the anterior circulation of the brain (middle cerebral artery) and 20 % into posterior circulation (3,4,5).
Thrombolytic therapy with intravenous tissue plasminogen activator (tPA) is of net benefit if performed in stroke units (6) in eligible patients with acute ischemic stroke, intra-arterial thrombolysis is preferable in patients with the new anti-coagulation treatments to lessen the effect on the coagulation time.
This review aims to offer tools for fair assessment of cardioembolic stroke.

1 - Observation of symptoms 

Observation of symptoms can help determine 1) location of the embolus or emboli and 2) involvement of anterior or posterior circulation.
Severity of symptoms lessens with division of the middle cerebral artery (MCA) - one of the three major arteries together with the anterior and posterior artery. Click here for a reminder of localisation of the three major arteries here and here for one regarding the branching of the MCA into M1 (Horizontal), M2 (MCA after the division on upper and lower vessels - Sylvian), M3 (Insular artery - Cortical) and M4.
Indeed, emboli in the M1 artery (in trunk of the MCA before the division) induces symptoms such as hemiparesis (weakness on one side of the body), hemihypesthesia (reduction in sensitivity in one side of the body) hemianopsia (decreased vision or blindness in half the visual field of one or both eyes), conjugated eye and total aphasia (disturbance of the comprehension and formulation of language - left lesion, find here and here, all types of aphasia) or multimodal hemineglect (failure to be aware of items to one side of space - right lesion) whereas when either of the M4 arteries (leptomeningeal arteries - prefrontal, precentral, angular temporal, anterior parietal, posterior parietal are involved, only isolated Wernicke's aphasia can be observed (patient has serious comprehension problems).
Anterior circulation location: A
A – refers to anterior cerebral artery (ACA
A1 - first part of anterior cerebral artery (before anterior communicating artery); 
A2, A3 - parts of the artery after communicating artery.
Posterior circulation: Involvement of the posterior circulation can be suspected through findings of symptoms such as headache, somnolence with progression to coma, true vertigo, drunken - type gait, disconjugated gaze, hiccup, slurred speech, bilateral motor symptoms, Wallenbergs syndrome (facial numbness, spinothalamic sensory loss, contralateral hemiparesis (usually mild), ataxia (lack of coordination of muscular movements - cerebellar), nystagmus (involuntary eye movement), diplopia (double vision), Horner's syndrome (drooping of the eyelid (ptosis) and constriction of the pupil (miosis), dysphagia (difficulty swallowing) and dysarthria (difficulty articulating), or locked - in syndrome. 
Locations are PICA (posterior inferior cerebellar artery); AICA (anterior inferior cerebellar artery); SCA (superior cerebellar artery); Distal basilar artery.
Ischemic infarctions which are caused by emboli from LAA fall into specific classifications of ischemic stroke (7,8) which we go on to detail.

2 - Ischemic stroke classification

The Oxfordshire Community Stroke Project and TOAST (Trial of Org 10172 in Acute Stroke Treatment) developed two classes of ischemic strokes. The former concentrates on initial symptoms, and the latter on initial symptoms + diagnostic tests, however both do not have the ideal qualities for diagnostic purposes. This validated classification system has facilitated the study of differences in stroke subtypes with respect to etiology, risk factors, treatment, and outcomes (9). Etiologic (10) investigations and effective secondary stroke prevention are thus influenced by the correct categorization of stroke syndromes.
In OCSP classification TACI (total anterior circulation infarction), PACI (partial anterior circulation infarction) and POCI (posterior circulation infarction) can be cardioembolic strokes in nature. It is believed that cardioembolic strokes can be diagnosed in 15-25 % of all ischemic strokes (with recent know-how in diagnostic approaches this number might reach as high as 35%) (12, 14). In this classification, there is also the LACI (lacunar infarction). Click here for a look at the common sites of cerebral aneurysms. 
TOAST classification is the fourth and main classification for substantiating diagnosis. There is:

  • Atherosclerosis with stenosis: 50% narrowing or occlusion of the relevant extracranial or intracranial large artery in the absence of another aetiology.
  • Atherosclerosis without stenosis: 50% diameter narrowing or more in the absence of another aetiology in patients with two or more of the following risk factors: current smoking, 50 years of age or more, hypertension, diabetes, high cholesterol. 
  • Cardio-embolism: Atrial fibrillation, intracardiac thrombus or tumor, rheumatic mitral stenosis, prosthetic valves, endocarditis, sick sinus syndrome, left ventricular aneurysm or akinesia after AMI, acute (3 months) AMI, cardiomyopathy – in the absence of another aetiology
  • Small artery: lacunar syndrome and normal CT/MRI or relevant lesion up to 1,5 cm and absence of cardiac source of embolism and 50% stenosis. 
  • Stroke of other determined aetiology: arterial dissection, fibromuscular dysplasia, AVM, vasculitis, venous sinus thrombosis, hypercoagulable states, migrainous ischaemia.
  • Stroke of undetermined aetiology: none of the above causes could be determined. Two or more potential causes of infarction (7).

A large number of cases of ischemic stroke may actually fall into a fifth type of classification where stroke identification is impossible. Lately, new efforts have been to for improvement in the area of diagnostic classification for an ischemic stroke. For example the Multicentre Retinal Stroke Study is being initiated (where retinal photographs are being used for the purpose of ischemic stroke classification).

3 - Diagnostic work-up 

Diagnostic work-up is based on establishment of the cause-effect relationship between left atrial appendage (cardiac embolus source in 90 % cases) and cardioembolic (ischemic) stroke development. In elderly patients, diagnosis is not as simple as it may seem because two or more etiological factors for stroke can almost certainly be present (for example - presence of significant carotid stenosis and permanent atrial fibrillation at the same time). This situation requires a multiprofessional team within a stroke unit with all of their acumen and experience. It is also very important in case of cardioembolic stroke to make a correct diagnosis taking into account the recurring rate of this type of ischemic stroke. (9-11)

I – Stroke assessment

A. Neurological 
First assessment of 11 neurologic functions (NIHSS score) is important in that the NIHSS score is strongly related to outcome. For example - ischemic stroke with a NIHSS 20 and more portends a very serious prognosis: a 3 month mortality in stroke like this is around 40% (these patients should be considered potential candidates for endovascular treatment).
Severity classification of ischemic stroke according the neurological deficit (NIHSS classification):

  1. 1-7 points  - mild 
  2. 8-14 points - moderate to severe 
  3. 15-42 points - severe (9-11)

B. Neuroimaging and laboratory 
Neurovisualisation and laboratory assessment are usually carried out simultaneously. A primary target is to initiate a reperfusion therapy (as in the case of acute myocardial infarction) no later than 60 minutes from patient hospital admission. 
The cornerstone of diagnosis is rapid confirmation of the ischemic lesion and hemorrhagic stroke exclusion. Today, all CT scan examinations for ischemic stroke are performed according to the special ASPECTS scale by a very experienced neuroradiologist because ischemic changes are very subtle - specificity of CT in cases of ischemic stroke in the first 3 hours is no more than 31%! Broad collaboration between the neurologist and the neuroradiologist is vital because of this fact. 
The 10 regions of interest in different areas of the brain need to be assessed so that that ischemia may be excluded or verified. Each region of ischemia is to be subtracted from the number 10 (10-1, if there is 1 ischemic region, minus 2 if there are two etc). If total ischemic score equals 7 or less than total lesion is more than 1/3 of CMA - prognosis is then very serious. Ischemic score of 9 or 10 - generally portends a good prognosis. The aim is to make definitive diagnosis within 25 minutes of hospital admittance without any delay, especially among patients who have been admitted in the first 3 hours since symptom development (the category of patients who must be considered for treatment with fibrinolytic therapy). For patients with posterior location of ischemic infarction PS - ASPECTS is being investigated where 8 regions have been identified as requiring exploration (9-11).

  • CT - angio or MRI - angiography

Can help in exact verification of occlusion and assessment of therapy efficacy (after fibrinolysis with fibrinolytics for vessel patency assessment and /or as preparation for endovascular treatment. (9-11)
See here slides to explore concept of penumbra, acute carotid syndromes, and advantages and disadvantages of diagnostic techniques, non contrast CT, perfusion CT and diffusion weighted CRT in stroke. 

  • Blood analysis

Blood work involves study of glucose level, full blood count, platelets level, coagulation INR parameters - international normalization ratio, aPTT-activated partial thromboplastin time, PT- prothrombine time and fibrinogen (10-11).
If after neurovisualisation, cardioembolic ischemic stroke is suspected (usually TACI, PACI , POCI stroke types fall into this category) - then diagnostic modality for evaluation of embolic potential and atrial fibrillation identification should be applied and oral anticoagulants must be prescribed before discharge to suitable candidates for long-term therapy. Initiation of the anticoagulant therapy can reduce the relative risk of the recurrent ischemic embolic stroke up to 80 % which makes anticoagulant therapy one of the most effective therapy for secondary prevention to date.
II - AF detection 

  • ECG: serial ECG in hospital for the fifth consecutive days after the event is preferable to single ECG)
  • Holter monitoring for 24 - 48 hours (12-13)
  • Continuous ECG monitoring with automated analysis (a CEM) for a 7 day duration has been recently successfully studied (significant additional cases with AF were identified in cases with negative Holter monitoring data). Improved methods for detecting AF are currently being evaluated in pilot trials, such as a 30-day cardiac event monitoring belt in the EMBRACE study. The Reveal XT device can be potentially used for long- term monitoring of rhythm (1-3 year). The device can be implanted (studies that include stroke patients are ongoing) (9-13).

The embolic potential can be evaluated with the following tools:

  • Echocardiography - transthoracic (TTE), transoesophageal (TEE) and 3D - echocardiography. TEE - is a golden standard for an evaluation of the left atrial appendage (LAA). Transoesophageal echocardiography has been shown to identify features correlating with high thromboembolic risk. (3,4,5,10,11) They are:
  • Left atrial thrombus, increased left atrial appendage size, low left atrial appendage peak velocities, spontaneous echocardiographic contrast (often called “smoke”), left ventricular dysfunction, left ventricular hypertrophy and complex aortic plaque. The stroke rates were 1.3% per year if neither risk factor was present, 7.8% per year if a left atrial abnormality was present, 12.0% per year if complex aortic plaque was present, and 20.5% per year if both were present.
  • Cardiac MRI can be also used for left atrium assessment (11)
  • Transcranial Doppler Ultrasonography can show high intensity transient signs that indicate microembolism. This mode of investigation can provide supportive evidence for embolism as a cause of stroke in a patient who presents with symptoms of stroke.
  • D-dimer analysis (in one recent study d - dimer, when measured within 6 h from stroke onset, excluded cardioembolic stroke with a sensitivity of 100% and a specificity of 52 % with D-dimers level below 300 μg/l )(14). 

Table 1: The Common features of cardioembolic ischemic stroke 

Local radiological signs such as cortical infarction: for example, embolus location in M4 of the MCA in connection with Vernicke's aphasia in elderly patient might be considered a typical localisation in cardioembolia in AF patient, or cerebellar infarctions.
Sudden onset of maximal deficit with decreased level of consciousness.
Rapid regression of initially massive symptoms: "spectacular shrinking deficit".
Multiples infarct sites: Finding on neuroimaging of acute infarcts involving multiple vascular territories in the brain, or multiple levels of posterior circulation (can be noticed in up to 40 % of all patients).
Recurrency and asymptomatic character: with maximum recurrent strokes in the first 6 month after the first event and 15-26 % of patients already have symptomless lesion of unknown onset on neuroimaging. 

4 - Outcome

Ischemic stroke is the second cause of death worldwide. It is also true that 60 % of stroke related death occurs in women. What is more, approximately 77 % of strokes are first events and about 10-30 % of survivals will have a second stroke within 5 years with acute mortality in some cases after second stroke up to 70 %. Finally more than 25 % of ischemic stroke survivors older than 65 years of age are institutionalized 6 months later (9-11). The main purpose of the second European declaration on ischemic stroke (Helsinborg, 2006) is to 1) reduce 1- month mortality from acute ischemic stroke to 15 % and less, 2) increase chances of being independent 90 days after the acute event in 70 % and more among those who have been afflicted by ischemic stroke in the European Union to 2015.
Regarding disability, patients with cardioembolic stroke can also be classified according to the modified Rankin scale which is internationally well accepted. This scale has long been used as surrogate marker of outcomes in ischemic strokes. 

Table 2: The modified Rankin scale*

Level 1. No symptoms: no symptoms at all
Level 2. No significant disability despite symptoms: able to carry out all usual duties and activities
Level 3. Slight disability: unable to carry out all previous activities but able but able to look after own affairs without assistance
Level 4. Moderate disability: requiring some help, but able to walk without assistance
Level 5. Moderate severe disability: unable to walk without assistance and unable to attend to own bodily needs without assistance
Level 6. Severe disability: bedridden, incontinent, and requiring constant nursing care and attention

*Original Rankin scale - from 1957, consisting of a single item, with five grades was modified in 1988 as part of a study of aspirin in stroke prevention from the UK-TIA study group. 

  • Not enough patients are on needed anticoagulant therapy: Only 10 % to 59 % of patients, who have suffered cardioembolic ischemic stroke, are on anticoagulant therapy after the cardioembolic ischemic stroke which goes against current recommendations (17,18).
  • More disability without therapy: According to the data from pooled randomised controlled trials (RCTs) of alteplase in placebo arm: proportion of patients at Level 3 of modified Rankin scale is 15%, level 4 is 20% and level 5 is 8%, which means that without treatment, 43% of patients are severely disabled after ischemic stroke. 
  • Three month independence from disability: SITS–ISTR (2003–2010) results (at the time almost 80,000 patients with ischemic stroke were around 25 % to have atrial fibrillation) and RCT (active treatment arm) results were respectively 50.4 % (53.5–56.0) and 50.1% (44,5–54,7).

In cases of hospital discharge without oral anticoagulant after cardioembolic stroke, prognosis for recurrency is high.
In a study of recurrent cerebral embolism in nonvalvular atrial fibrillation, Hart el al reported a 20% recurrence rate of cerebral embolism within 11 days of the initial embolus in patients who had not undergone anticoagulant therapy. Wolf et al. reported, in a population- based study, that the risk of recurrent stroke at 30 days was only slightly more frequent in stroke patients with atrial fibrillation (25%) than in those without atrial fibrillation (20%). However, the 6-month recurrence rate was 47% in the atrial fibrillation group compared with 20% in the group without atrial fibrillation (3-4).


According to the population-based studies patients with cardiembolic strokes are older, their ischemic strokes tend to be more severe and they are considered high-risk patients in terms of further possible bleeding complications.
Those facts are translated into the situation where large numbers of potentially suitable candidates are left without oral anticoagulation treatment for secondary prevention of stroke.
Diagnosis, treatment, follow-up and prevention cannot be effective enough without team-work between cardiologist and neurologist to reach a consensus regarding the treatment of senior people with multiple comorbidities and a high bleeding risk per se.


1. The ESC Textbook of Cardiovascular Medicine
A. John Camm,Thomas F. Lusher, Patrick W.Serruys
Second edition Oxford University Press, 2009
2. Sex differeneces in stroke: epidemiology, clinical presentation, medical care, and outcomes
Mathew J Reeves, et al.
Lancet Neurology 2008;7(10):915–926
3. Cardioembolic Stroke: An Update
Roger E. Kelley, MD, and Alireza Minagar
Southern Medical Journal Volume 96, Number 4, 2003
4. An update on cardioembolic stroke
N. U. Weir
Postgrad Med J 2008;84:133–142
5. Cardioembolic stroke: an update
Jose M Ferro
Lancet Neurology 2003;2:177–88 
6. Reliability of stroke diagnosis in emergency departments
Martínez Fernández, E.; Sanz Fernández, G.; Blanco Ollero, A.
Neurologia. 2012;27:284-9. - vol.27 núm 05
7. Classification and natural history of clinically identifiable subtypes of cerebral infarction.
Bamford J, Sandercock P, Dennis M, et al.
Lancet 1991 337:1521–1526
8. Classification of subtype of acute ischemic stroke. Definitions for use in a Multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment
HP Adams, Jr et al. 
Lancet Neurol. 2003 Mar;2(3):177-88.
9. Oxfordshire Community Stroke Project Classification Poorly Differentiates Small Cortical and Subcortical Infarcts
Negar Asdaghi, et al. 
10. Acute Ischemic Stroke. Imaging and Intervention. Second Edition 
R.G. González, J.A. Hirsch,M.H. LevP.W. Schaefer L.H. Schwamm
ISBN: 978-3-642-12750-2 Springer-Verlag Berlin Heidelberg 2011
11. Textbook of Stroke Medicine 
Michael Brainin,Wolf-Dieter Heiss,Susanne Heiss
ISBN-13 978-0-521-51826-0 Cambridge University Press 2010
12. The update of the European Stroke Initiative (EUSI) Recommendations for Stroke Management. 
Ringleb, M. et al., European Handbook of Neurological Management: Volume 1,
2nd edition. Edited by N. E. Gilhus, M. P. Barnes and M. Brainin. Blackwell Publishing 2011
13. Implementation and outcome of thrombolysis with alteplase 3–4•5 h after an acute stroke: an updated analysis from SITS-ISTR
Niaz Ahmed, et al, for the SITS Investigators 
Lancet Neurol. 2010 Sep;9(9):866-74. doi: 10.1016/S1474-4422(10)70165-4. Epub 2010 Jul 26.
14. Continuous stroke unit ECG monitoring versus 24 h-Holter ECG for detection of paroxysmal atrial fibrillation after stroke.
Niaz Ahmed, et al, for the SITS investigators 
Rizos T, Guntner J, Jenetzky E, Marquardt L, Reichardt C, Becker R, et al. Stroke. 2012;43:2689–2694.
15. Contribution of Atrial Fibrillation to Incidence and Outcome of Ischemic Stroke: ResultsFrom a Population-Based Study 
Carmine Marini, Federica De Santis, Simona Sacco, Tommasina Russo, Luigi Olivieri, Rocco Totaro, Antonio Carolei
Stroke. 2005;36:1115-1119
16. D-Dimers Predict Stroke Subtype when Assessed Early
Jörg Iseneggera, et al
Cerebrovasc Dis 2010;29:82-86 
17. Stroke Associated with Atrial Fibrillation –Incidence and Early Outcomes in the North
Dublin Population Stroke Study

Niamh Hannon, et al. 
Cerebrovasc Dis 2010;29:43–49
18. Underuse of antithrombotic therapy caused high incidence of ischemic stroke in patients with atrial fibrillation 
Hsiu-Chin Yu, Yun-Fang Tsai, Min-Chi Chen and Chi-Hsiao Yeh
International Journal of Stroke Volume 7, Issue 2, pages 112–117, February 2012


Vol11 N21

Notes to editor

Oleg Lykov MD,
cardiologist, member of ESC since 2003, member of the group Cardiologist of tomorrow
Authors' disclosures: None declared. 

The content of this article reflects the personal opinion of the author/s and is not necessarily the official position of the European Society of Cardiology.