Abstract 3736: Applying the Carotid Atherosclerosis Score in a Clinical Environment: a Feasibility Study
Background Carotid atherosclerosis is a leading cause of stoke. Recent studies have shown that its severity can be stratified by the Carotid Atherosclerosis Score (CAS) as measured by MRI. However, these studies were performed using customized imaging sequences, coils and comprehensive plaque analysis by specialized human reviewers, which are not generally available in a clinical environment. Whether CAS can be applied for clinical risk assessment is unknown. In this study, we used clinically collected MRI data to compute CAS. We investigated the correlation between CAS and the prevalence of neurological symptoms and disrupted luminal surface (DLS), an indicator of plaque instability, to evaluate the feasibility of translating CAS for clinical use.
Methods Seventy six patients with either symptomatic or asymptomatic carotid disease detected by ultrasound were recruited. All patients underwent a multi-contrast MRI with clinical sequences and a commercially available coil. The lumen and outer wall boundaries and DLS were outlined manually and the plaque components were identified automatically using histologically validated image analysis software. The maximum wall thickness (MWT) and maximum % area (100 x area/[wall area]) of lipid-rich necrotic core (LRNC) were outputted by the software to compute CAS as follows: CAS=1 with MWT<2.0mm, CAS=2 with LRNC<20%, CAS=3 with LRNC between 20% and 40% and CAS=4 with LRNC>40%. Trend analysis was done to evaluate the associations between CAS and prevalence of symptoms and DLS. The area under the receiver operating characteristic (ROC) curve was used to evaluate the classification performance of CAS.
Results Of the 76 patients, there was a significant increasing trend (p=0.01) between CAS scores and neurological symptom status. The presence of DLS also demonstrated a significant increasing trend (p<0.001) versus CAS with rates 0% (0/7), 7% (2/27), 35% (7/20) and 59% (13/22). ROC analysis of CAS also showed it was a good classifier of symptom status (AUC = 0.67, p=0.01) and DLS (AUC = 0.80, p<0.001)
Conclusion In this study, we conclude that the CAS computed from clinically collected MRI data can stratify the severity of carotid atherosclerotic disease as indicated by neurological symptom status and DLS, both of which are important in monitoring patients and evaluating treatment options. Further study is needed to assess how a fully automated CAS analysis affects the quality of the score and how well the score can stratify the risk of future outcomes.
- © 2012 by American Heart Association, Inc.