Beyond Imaging Features: Dual-Energy CT Reveals Subtle Differences in High-Risk Plaques Between Stroke and Non-stroke...

RATIONALE AND OBJECTIVES: A critical challenge is identifying why some high-risk plaques cause stroke while others remain quiescent. This study aimed to determine if dual-energy CT (DECT) can detect subtle parametric differences in radiologically similar high-risk plaques that may help predict future acute ischemic stroke (AIS) risk.

This retrospective study enrolled patients with non-stenotic carotid plaques (nsCP), categorized into AIS and non-AIS groups. All patients underwent dual-energy CT angiography. Plaque analysis included conventional parameters-maximum wall thickness (MWT), low-attenuation plaque (LAP, <60 HU) thickness, degree of stenosis, and number of high-risk features (HRF)-as well as spectral parameters derived from DECT, such as virtual monoenergetic image (VMI) attenuation values, Z-effective, spectral curve slope_40-140 keV, and iodine concentration. Logistic regression was used to calculate odds ratios (OR), and receiver operating characteristic (ROC) curve analysis was employed to evaluate diagnostic performance.

A total of 87 patients with 216 plaques were analyzed, including 37 AIS patients with 85 plaques. Conventional plaque analysis showed that each additional HRF increased AIS risk by 54%. Similarly, per 1-mm increase in maximum wall thickness and LAP thickness raised AIS risk by 73% and 62%, respectively, while each 1% increase in the degree of stenosis increased risk by 9%. DECT spectral analysis revealed significantly lower attenuation at 140 keV in AIS-group plaques for both napkin-ring sign (22.1 ± 7.8 vs. 29.6 ± 10.4 HU) and LAP (22.2 ± 8.8 vs. 32.3 ± 12.1 HU). ROC analysis demonstrated optimal diagnostic performance at 140 keV for napkin-ring sign (AUC = 0.723) and LAP (AUC = 0.750), with a cutoff of 33.5 HU for low-attenuation plaque, yielding a specificity of 0.952.

Subtle differences exist in radiologically similar plaques between AIS and non-AIS patients. DECT parameters can quantify these differences that are beyond conventional imaging morphology. These DECT parameters serve as stroke-associated biomarkers, providing a direction and a foundation for the future development of effective predictive tools and clinical decision-making.