LAUSR

Imaging tests always misdiagnose anatomical variants of cerebral sinuses as cerebral venous sinus thrombosis (CVST). Anatomical variants of cerebral sinuses were called CVST mimics. This study aimed to identify the role of inflammatory markers in differentiating CVST from mimics. A total of 146 patients diagnosed as CVST and 93 patients with mimics were recruited in this study. ROC analysis was performed to demonstrate the sensitivity and specificity of inflammatory markers for diagnosing CVST. Rank logistic regression analysis was performed to identify the association of markers to CVST severity and prognosis. CVST presented higher inflammatory reactions compared with mimics, demonstrated by the neutrophil count(5.11 [3.97, 6.80] vs. 3.06 [2.34, 3.86]), IL-6 (7.42 [3.85, 14.22] vs. 2.47 [1.50, 4.00]) and neutrophil-to-lymphocyte ratio (NLR, 3.19 [2.18, 4.62] vs. 1.66 [1.16, 2.22]). ROC analysis showed markers with area under the curve (AUC)>0.8 including IL-6 (optimal cutoff, 3.790; Kappa value, 0.499), neutrophil count (3.975; 0.522) and NLR (2.070; 0.476). After propensity score matching, only IL-6 had an AUC>0.8, with an optimal cutoff of 3.060 and a Kappa value of 0.636. Ranked logistic regression showed that IL-6 (OR, 95% CI, 1.063, 1.026-1.101; 1.029, 1.009-1.050), CSF IgA (0.279, 0.110-0.706; 0.398, 0.162-0.974), CSF IgM (22.399, 3.004-167.001; 9.545, 1.382-65.928), and CSF IgG (1.287, 1.124-1.473; 1.232, 1.091-1.392) were independently correlated with the baseline and follow-up mRS. In conclusion, inflammatory markers in CVST were different from those in mimics. These markers, especially IL-6, could not only differentiate CVST from its mimics, but also evaluate CVST severity and prognosis.