We determined the incremental value of carotid plaque components – beyond traditional cardiovascular risk factors – for predicting first-ever stroke events.
Between 2007 and 2012, participants from the population-based Rotterdam Study with asymptomatic carotid wall thickening >2.5 mm on ultrasonography, were invited for carotid MRI. Among 1349 participants (mean age: 72 years, 49.5% women) without cardiovascular disease, we assessed plaque thickness, carotid stenosis(>30%), presence of intraplaque hemorrhage, lipid-rich necrotic core, and calcification. Follow-up for stroke was complete until January 1, 2016. Using Cox proportional hazards regression we fitted prediction models including traditional cardiovascular risk factors (base model), and calculated 5-year predicted risks of stroke. We extended the base models by single and simultaneous additions of plaque characteristics, and calculated the c-statistics.
During a median follow-up time of 5.6 years, 56 strokes occurred. The median predicted 5-year risk of stroke based on traditional cardiovascular risk factors was 3.3%. Intraplaque hemorrhage and carotid stenosis were associated with incident stroke (adjusted HR intraplaque hemorrhage: 2.04, 95%CI=1.17-3.53, stenosis: 1.80, 95%-CI=1.00-3.24). After simultaneous addition of all plaque components, plaque thickness, and carotid stenosis to the base model, only intraplaque hemorrhage remained statistically significantly associated with incident stroke. The c-statistic (95%CI) improved from 0.67 (0.60-0.73) to 0.69 (0.62-0.75) after single addition of intraplaque hemorrhage to the base model. Other plaque characteristics did not improve the model discriminative ability for predicting first-ever stroke beyond traditional risk factors.
Presence of intraplaque hemorrhage is a predictor of first-ever stroke, independent of traditional cardiovascular risk factors, and other plaque characteristics.
Not applicable