Association between low levels of serum miR-638 and atherosclerotic plaque vulnerability in patients with high-grade carotid stenosis

Abstract

OBJECTIVE Carotid artery atherosclerosis is a major cause of ischemic stroke. However, reliable criteria to identify patients with high-risk carotid plaques beyond the severity of stenosis are still lacking. Circulating microRNAs (miRNAs) are being postulated as biomarkers for a variety of vascular immune-inflammatory diseases. The authors investigated whether cell-free circulating miR-638, highly expressed in vascular smooth muscle cells and implicated in proliferative vascular diseases, is associated with vulnerable atherosclerotic plaques in high-risk patients with advanced carotid artery stenosis undergoing carotid endarterectomy (CEA). METHODS The authors conducted a prospective study in 22 consecutive symptomatic patients with high-grade carotid stenosis undergoing CEA and 36 age- and sex-matched patients without ischemic stroke history or carotid atherosclerosis (control group). In addition, they reviewed data from a historical group of 9 CEA patients who underwent long-term follow-up after revascularization. Total RNA was isolated from all serum samples, and relative miR-638 expression levels were detected by reverse transcription-quantitative polymerase chain reaction (RT-qPCR) and compared among groups. A correlation analysis of serum miR-638 levels with vascular risk factors and treatments, and with plaque features, was performed. The ability of serum miR-638 to discriminate between the non-CEA control group and the different CEA groups was assessed by receiver operating characteristic evaluation. A logistic regression model was employed to examine the association between stratified CEA patients and serum miR-638 levels. RESULTS Serum levels of miR-638 were significantly lower in symptomatic CEA patients (p = 0.009) and particularly in the subgroup of CEA patients who had experienced stroke (p = 0.0006) than in non-CEA controls. Discrimination of high-risk plaques was accurate (area under the curve [AUC] 0.66 for symptomatic CEA patients in general and 0.76 for those who had experienced stroke). When only patients with high cardiovascular risk were considered, the diagnostic value of serum miR-638 from symptomatic CEA patients and CEA patients who had experienced stroke improved (AUC 0.79 and 0.85). Moreover, serum miR-638 was negatively correlated with the occurrence of stroke, smoker status, presence of bilateral pathology, coronary artery disease, and cholesterol treatment; and with the high-risk fibroatheroma plaques extracted from CEA patients. Multivariate logistic regression analysis demonstrated that serum miR-638 was an independent predictor of plaque instability. Furthermore, serum miR-638 appeared to attain good discrimination for atherosclerotic stenosis in CEA patients based on analysis of blood samples obtained in the historical group before and 5 years after intervention (p = 0.04) (AUC = 0.79). CONCLUSIONS According to this preliminary proof-of-concept study, serum miR-638 might constitute a promising noninvasive biomarker associated with plaque vulnerability and ischemic stroke, particularly in individuals with elevated cardiovascular risk.