The relation between plasma miR-126 levels and cerebral collateral circulation in patients with intracranial arterial stenosis

This study aimed to investigate the correlation between the circulating miR-126 regulation pathway and the cerebral collateral circulation (CCC), and to test whether miR-126 could serve as a potential biomarker for CCC formation in patients with intracranial arterial stenosis or occlusion. This single-centre cross-sectional study enrolled patients who underwent cerebral angiography with severe stenosis (≥70%) or occlusion in at least one major intracranial artery. Collateral degree was graded according to the ASITN/SIR classification. The patients were divided into a good CCC group (grade 3-4) or a poor CCC group (grade 0-2). We investigated the plasma levels of miR-126, VEGF, Spred-1 and PIK3R2 by using qRT-PCR, ELISA and Western blot methods, respectively. In addition, we assessed the correlations of plasma miR-126 with VEGF, Spred-1, PIK3R2 and ASITN/SIR grade using the Spearman correlation test and investigated its predictive power for CCC status by using the receiver operating characteristic curve. A total of 68 patients were enrolled (44 with good CCC and 24 with poor CCC). Data showed that plasma miR-126 and VEGF were significantly higher in the good CCC group than in the poor CCC group. Plasma Spred-1 and PIK3R2 level were lower in the good CCC group than in the poor CCC group. In addition, miR-126 and VEGF were positively correlated with ASITN/SIR (miR-126: R = 0.595, P < 0.01; VEGF: R = 0.595, P < 0.01), whereas Spred-1 and PIK3R2 were negatively correlated with ASITN/SIR (Spred-1: R = -0.817, P < 0.01; PIK3R2: R = -0.513, P=0.01). However, the area under the curve of miR-126 level for CCC status was only 0.328 (95% CI: 0.158-0.498; p = 0.067). Plasma miR-126 level may be related to better CCC formation, one of the mechanisms that may be explained by upregulation of VEGF and reduction of Spred-1 and PIK3R2 protein expression. However, miR-126 might not be an independent predictor for CCC, given its low predictive value.