Differentiating Thrombotic from Cardioembolic Stroke: The TOAST Classification System
The TOAST (Trial of Org 10172 in Acute Stroke Treatment) classification system is the standard scoring system used to differentiate large-artery atherosclerotic (thrombotic) stroke from cardioembolic stroke in adults with acute ischemic stroke. 1, 2
The TOAST Classification Framework
The TOAST system categorizes ischemic strokes into five subtypes based on clinical features and diagnostic test results 1, 2:
1. Large-Artery Atherosclerosis (Thrombotic Stroke)
Diagnostic criteria include:
- Cortical infarct in the distribution of a large cerebral artery with demonstrated luminal occlusion or ≥50% stenosis of atherosclerotic origin 1
- Often preceded by a TIA in the same arterial distribution 1
- Mechanism presumed to be artery-to-artery embolism or hemodynamic insufficiency 1
- Must exclude potential cardiac sources of embolism 1
2. Cardioembolism
Diagnostic criteria include:
- Cortical or large subcortical infarctions with a recognized high-risk cardiac source 1
- Arterial occlusion caused by an embolus originating from the heart 1
- Moderate-risk cardiac source alone qualifies for "possible cardioembolism" 1
- Must exclude atherosclerotic narrowing in the parent large artery 1
3. Small-Artery (Lacunar) Stroke
- Symptoms consistent with lacunar syndrome (e.g., pure motor hemiparesis) with small (≤1.5 cm) lesion on neuroimaging 1
- Generally associated with diabetes or hypertension 1
4. Stroke of Other Determined Etiology
- Uncommon causes identified by diagnostic testing (nonatherosclerotic vasculopathies, hypercoagulable states, hematologic disorders) 1
5. Stroke of Undetermined Etiology
- Accounts for more than one-third of all strokes 1
- Includes patients with two potential mechanisms or incomplete evaluation 1
Clinical Application and Validation
The TOAST system demonstrates excellent interobserver reliability: two independent neurologists disagreed in only 1 of 20 patients (95% agreement), and both reached specific etiologic diagnoses in 55% of cases 2. The system allows for "possible" and "probable" diagnoses based on diagnostic certainty 2.
Imaging-Based Differentiation Strategies
Vascular Territory Analysis
Middle cerebral artery territory infarctions are most commonly caused by large-artery atherosclerosis (37.3% overall prevalence), while cardioembolism accounts for 23.2% of carotid territory strokes 3. Multiple-vascular-territory infarctions strongly suggest cardioembolism in carotid territory (44.2%) but large-artery atherosclerosis in vertebrobasilar territory (52.1%) 3.
Infarct Volume and Number
Each 5 mL increase in infarct volume is associated with both cardioembolic (OR 1.07,95% CI 1.01-1.14) and large-artery occlusions (OR 1.10,95% CI 1.02-1.18), but dramatically lower odds of lacunar stroke (OR 0.18,95% CI 0.06-0.55) 4.
Critical distinction: A single large-volume infarct favors large-artery atherosclerosis, whereas multiple infarcts of varying ages favor cardioembolism 4.
Essential Diagnostic Workup
Immediate Imaging Requirements
- Non-contrast CT within 25 minutes of arrival to exclude hemorrhage and assess early ischemic changes 1
- Vascular imaging (CTA or MRA) to identify large-vessel occlusion or ≥50% stenosis 1
- MRI with diffusion-weighted imaging provides superior sensitivity (88-100%) and specificity (95-100%) for detecting acute ischemia, including small cortical, subcortical, and posterior fossa lesions 1
Cardiac Evaluation
- 12-lead ECG immediately to detect atrial fibrillation or acute myocardial infarction 1, 5
- Continuous cardiac monitoring for 24 hours to detect paroxysmal atrial fibrillation, which occurs in up to 25% of cryptogenic strokes 5
- Echocardiography (TTE or TEE) when cardiac source is suspected 1
Laboratory Studies
Obtain complete blood count, coagulation profile (PT/INR, aPTT), glucose, hemoglobin A1c, creatinine, and lipid profile 1, but do not delay imaging or treatment while awaiting results unless clinical suspicion exists for bleeding diathesis or anticoagulant use 6.
Prognostic Implications
Cardioembolic stroke carries the highest mortality: 22.6% death rate within 90 days compared to 3.3% for small-vessel disease 1. Large-artery atherosclerotic stroke has intermediate mortality and the highest risk of early recurrent stroke, particularly with severe stenosis 1. This prognostic information guides the intensity of secondary prevention strategies 1.
Emerging Biomarker Approaches
Gene expression profiling in peripheral blood can differentiate cardioembolic from large-vessel atherosclerotic stroke with 95.2% sensitivity and specificity using a minimum of 23 genes 7. Genes regulated in large-artery atherosclerotic stroke are expressed in platelets and monocytes (modulating hemostasis), while genes in cardioembolic stroke are expressed in neutrophils (modulating immune responses) 7. This approach remains investigational but may aid diagnosis of cryptogenic strokes 7.
Common Pitfalls in Classification
Substantial inconsistency exists between bedside and centralized expert subtyping: agreement is highest for large-artery atherosclerosis (77.4%) but only 38.7% for cardioembolism and 12.2% for other determined etiologies 8. This misclassification has critical therapeutic implications: only 15.3% of newly diagnosed cardioembolic patients received appropriate oral anticoagulation at discharge when subtyping was inaccurate, compared to 51.3% when classification was correct 8.
Key errors to avoid:
- Assuming stroke type based on severity alone—substantial overlap exists between subtypes 9
- Failing to exclude cardiac sources before diagnosing large-artery atherosclerosis 1
- Missing paroxysmal atrial fibrillation by relying on single ECG rather than prolonged monitoring 1, 5
- Inadequate vascular imaging to detect high-grade stenosis 1