Abstract WP214: Serum Claudin-5, a Potential Biomarker Reflecting Blood-Brain Barrier Dysfunction in Acute/Subacute Cerebral Venous Thrombosis With Venous Infarct

Abstract

Objective: To evaluate whether claudin-5, occludin and MMP-9 proteins in serum and cerebrospinal fluid(CSF) reflect blood-brain barrier(BBB) dysfunction in cerebral venous thrombosis(CVT) with venous infarct. Methods: It is a prospective observational study. Blood and CSF samples were obtained from CVT patients in acute, subacute and chronic stages respectively. 7 healthy subjects without central nervous system disease were included as control group. Serum and CSF Claudin-5, occludin and MMP-9 were measured using ELISA kits. Values of continuous variables are presented as mean ± SD and compared with ANOVA. Results: Between Jul 2015 to Dec 2016, 52 CVT patients(age: 35.9±14.1 years; female:53.9%) were enrolled. 43 patients were at acute/ subacute stage(30 had venous infarct, while 13 did not have). 9 patients were at chronic stage. CSF occludin and MMP-9 were not detectable in most of the subjects. Serum occludin and MMP-9 were similarly expressed in all four groups. The value of CSF Claudin-5 was 257.1±44.1pg/ml for the control group, 458±223.6pg/ml for chronic CVT patients, 593.1±262.7pg/ml for acute/subacute CVT patients with venous infarct, and 529.1±334.6pg/ml for acute/subacute CVT patients without venous infarct. Compared with the control subjects, CSF Claudin-5 significantly increased in acute/subacute CVT patients with and without venous infarct(p=0.003, p=0.031, respectively). The values of CSF claudin-5 between acute/subacute CVT patients with and without venous infarct were similar (p=0.464). The value of serum Claudin-5 was 1760.6±374.2pg /ml for control subjects, 3150.7±1197.6pg/ml for acute/subacute CVT patients with venous infarct, and 2411.6±565.2pg/ml for acute/subacute CVT patients without venous infarct. Compared with control subjects, serum Claudin-5 significantly increased only in acute/subacute CVT patients with venous infarct(p=0.001), which had also significantly higher value than without venous infarct(p=0.025). Conclusions: Serum claudin-5 might reflect disruption of BBB in acute/subacute CVT patients with venous infarct. In future, larger cohorts are warranted to precisely evaluate cutoff value of serum claudin-5 in reflecting BBB dysfunction in acute/subacte CVT patients with venous infarct.