Bilateral symmetry of human carotid artery atherosclerosis: a multi-contrast weighted MR study

The systemic nature of atherosclerotic disease may entail an association in disease severity between left and right carotid arteries. However, the etiology of plaque features in high-risk lesions is presumably attributed to local risk factors. We explored the symmetry of plaque morphology and composition across a broad range of atherosclerotic disease severities. All participants underwent carotid MR imaging on a 3.0 T scanner with a bilateral four-element phased-array surface coil. Vessel boundary and plaque components [calcification, lipid-rich necrotic core (LRNC) and intraplaque hemorrhage (IPH)] of bilateral carotid arteries were outlined. Normalized wall index (NWI) was calculated as follows: NWI = wall volume/total vessel volume. Carotid atherosclerosis score (CAS) was computed for plaque risk stratification. Associations of volume measurements between sides were evaluated using Pearson's correlation. Cohen's kappa was used to assess agreement between dichotomous variables. In the 177 participants with images of sufficient quality, there were very strong correlations between left and right lumen volumes (r = 0.85), total vessel volumes (r = 0.88), and strong correlations between wall volumes (r = 0.79), mean wall thickness (r = 0.66) and NWI (r = 0.71), and a moderate correlation between max wall thickness (r = 0.56) (all P < 0.001). There were moderate between-side agreements for the presence of calcification (κ = 0.54) and LRNC (κ = 0.49), but only fair agreement for IPH (κ = 0.31). The correlation of volume between left and right carotid arteries was strong for calcification (r = 0.62, P < 0.001) and weak for LRNC (r = 0.39, P < 0.001), but there was no significant correlation for IPH (r = 0.01, P = 0.99). Fair agreement (κ = 0.34) for CAS between paired carotid arteries was observed. Only 16 of 47 participants with CAS = 4 on at least one side had the same CAS on the contralateral side. Plaque morphology, calcification, and LRNC may develop symmetrically, but there is a relatively poor correlation for lipid content between sides. The weak symmetry of IPH and CAS indicates that the development of atherosclerosis into high-risk lesions may be regulated by local rather than systemic factors.