Characteristics of non-stenotic carotid plaque in embolic stroke of undetermined source compared with cardiogenic embolism: a retrospective cross-sectional observational study

Background Non-stenotic carotid plaque is considered an important etiology of embolic stroke of undetermined source (ESUS). However, only a few previous studies included a negative control group, and the characteristics of non-stenotic carotid plaque in ESUS have yet to be investigated. The objective of this study is to explore the clinical characteristics of ESUS and the correlation between non-stenotic carotid plaque and ESUS. Methods This is a single-center, retrospective cross-sectional observational study conducted to compare differences in clinical information among ESUS, CE, and large-artery atherosclerosis (LAA), as well as the prevalence of non-stenotic carotid plaque and non-stenotic carotid plaque with low echo between patients with ESUS and CE in Changzhou No.2 People's Hospital from January 2020 to January 2022. Ultrasound was used to evaluate the characteristics of non-stenotic carotid plaque and vulnerable carotid plaque was defined as plaque with low echo. The binary logistic regression model was used to analyze the relationship between the characteristics of non-stenotic carotid plaque and ESUS. The receiver-operating characteristic curve was used to evaluate the diagnostic efficiency of the characteristics of non-stenotic carotid plaque for ESUS. Results We had a final studying population of 280 patients including 81 with ESUS, 37 with CE, and 162 with LAA. There were no differences in clinical features between ESUS and LAA, but in the comparison of CE and ESUS, there were differences in age, smoking, hypertension, levels of triglyceride, total cholesterol, and low density lipoprotein cholesterol. In ESUS, the prevalence of non-stenotic carotid plaque was more common on the ipsilateral side of stroke than in CE [55 (67.90%) vs. 18 (48.65%), p = 0.046], so was the prevalence of non-stenotic carotid plaque with low echo [38 (46.91%) vs. 5 (13.51%), p < 0.001]. Logistic regression analysis showed that the prevalence of non-stenotic carotid plaque (OR: 4.19; 95% CI: 1.45-12.11; p = 0.008) and the prevalence of non-stenotic carotid plaque with low echo (OR: 5.12; 95% CI: 1.55-16.93; p = 0.007) were, respectively, the independent predictors of ESUS. The results receiver-operating characteristic (ROC) curve showed that the combination of age, hypertension, and ipsilateral non-stenotic carotid plaque with low echo had the best diagnostic efficiency for ESUS (0.811; 95%CI: 0.727-0.896; p < 0.001). Conclusion Our results suggest that ipsilateral vu