Classification and Management of Ischemic Stroke in Patients With Active Cancer: A Scientific Statement From the American Heart Association.
Review
Overview
abstract
About 10% to 15% of patients with ischemic stroke have a history of cancer, half of whom have active malignancy at the time of stroke. With improved cancer treatments extending patient survival, the coprevalence of these diseases has increased steadily since 2000. This has sparked considerable growth in research and knowledge on this topic. Approximately half of ischemic strokes in patients with active cancer are due to conventional mechanisms, although cancer-related factors may contribute. The remaining half of ischemic strokes in this population are typically classified as cryptogenic or attributed to cancer-specific mechanisms. These cryptogenic strokes often have characteristic risk markers and clinical features and are extremely high risk for recurrent stroke and other adverse events, distinguishing them from other stroke subgroups. Recent epidemiological, translational, and histopathological data indicate that many of these events are likely caused by the cancer itself through multifactorial prothrombotic processes. In this scientific statement incorporating multidisciplinary expertise, we critically appraise and synthesize recent evidence and provide clinical suggestions on the epidemiology, presentation, evaluation, pathophysiology, and treatments for ischemic stroke in patients with active cancer. In addition, we propose a novel classification for ischemic stroke attributed to cancer itself, which we define as cancer-related stroke to enable consistent nomenclature and to harmonize stroke classification across clinical practice and research. This system is based on routinely available clinical data and includes different categories for certainty of causality, relating to the patient's distinctive clinical features and estimated risk for recurrent thromboembolism. We hope this framework spurs dedicated controlled trials to address areas of clinical uncertainty.