Background
Ischemic stroke in people with active cancer is often difficult to classify because more than one stroke mechanism may be present at the same time. Traditional stroke systems such as TOAST were designed for the general stroke population and frequently label these events as cryptogenic, meaning no single clear cause is identified. In cancer patients, however, hidden cancer-related mechanisms such as hypercoagulability, disseminated intravascular coagulation, nonbacterial thrombotic endocarditis, treatment-related thrombosis, and tumor-associated embolism may be the true drivers of stroke. To address this gap, the American Heart Association recently proposed a cancer-related ischemic stroke classification, or CRIS, aimed at identifying strokes more likely linked to cancer itself.
This study evaluated how the CRIS framework changes stroke classification in real-world patients and whether it helps predict prognosis more accurately than the traditional TOAST approach.
