Individualized MRI-Based Stroke Prediction Score Using Plaque Vulnerability for Symptomatic Carotid Artery Disease...

Area of Science:

• Neurology
• Cardiovascular Medicine
• Medical Imaging

Background:

• Symptomatic carotid stenosis treatment decisions traditionally rely on stenosis degree.
• Intraplaque hemorrhage on MRI is a significant predictor of stroke risk.
• A need exists for improved risk stratification beyond traditional measures.

Purpose of the Study:

• To develop and validate the Individualized Magnetic Resonance Imaging-Based Stroke Prediction Score Using Plaque Vulnerability for Patients With Symptomatic Carotid Artery Disease (IMPROVE) model.
• To incorporate intraplaque hemorrhage and other risk factors for more accurate stroke risk prediction.
• To identify patients with symptomatic carotid disease who are at high risk for ipsilateral ischemic stroke.

Main Methods:

• The IMPROVE model was developed using data from 760 patients across 5 cohorts with symptomatic carotid disease on optimal medical treatment.
• Cox regression analysis was employed, incorporating predictors like stenosis degree, intraplaque hemorrhage, last event classification, sex, and age.
• The model was internally validated for accuracy and calibration, and its risk stratification was compared to care-as-usual.

Main Results:

• The IMPROVE model demonstrated good accuracy (C statistic, 0.82) and calibration.
• Key predictors included degree of stenosis, presence of intraplaque hemorrhage, and classification of the last ischemic event.
• Patients stratified as high-risk by IMPROVE had a higher incidence of ipsilateral ischemic stroke (24.0%) compared to care-as-usual (20.7%).

Conclusions:

• The IMPROVE model accurately estimates individual stroke risk in patients with symptomatic carotid artery disease.
• Incorporating intraplaque hemorrhage via MRI enhances stroke risk prediction.
• The IMPROVE model may aid in stratifying patients for revascularization decisions pending external validation.