Measurement of prothrombin fragment 1+2 in cerebrospinal fluid to identify thrombin generation in inflammatory central nervous system diseases

Abstract

BackgroundThe interaction of central nervous system inflammation and coagulation system activation in multiple sclerosis (MS) receives increasing attention for its diagnostic and therapeutic potential. During blood-brain barrier (BBB) disruption, fibrinogen migrates into the CNS and contributes to inflammation. In the coagulation cascade, fibrinogen is converted into fibrin by thrombin, which itself is cleaved from prothrombin by activated factor XII. We hypothesized that the conversion of prothrombin to thrombin can be quantified by prothrombin fragment 1+2 (PF1.2) in cerebrospinal fluid (CSF). Primary endpoint was the correlation between PF1.2, D-dimer and fibrinogen in CSF of patients with neuroinflammatory diseases. Secondary endpoints were PF1.2 levels depending on presence of contrast enhancement (CE) on MRI, and correlation between PF1.2 with serum-CSF albumin quotient (Qalb). Additionally, an exploratory analysis of CSF PF1.2 levels to distinguish between MS-patients and controls without neurological disease was performed. MethodsPatients admitted for a suspected inflammatory CNS disease were prospectively recruited from October 2017 to December 2020. Citrated CSF samples were obtained and analyzed for PF1.2, fibrinogen and D-dimer using a highly sensitive luminescent oxygen channeling immunoassay. Patient clinical data and final diagnoses were retrospectively collected and analyzed. Results187 patients were included, of whom 116 received diagnoses of relapsing-remitting (RRMS), primary-progressive MS, clinically or radiologically isolated syndrome, or anti-aquaporin-4-/anti-myelin-oligodendrocyte-glycoprotein-antibody-related diseases. CSF analysis of those 116 patients revealed a correlation between PF1.2 and CSF fibrinogen (ρ=.315; p<.001) as well as between PF1.2 and CSF D-dimer (ρ=.531; p<.001). Among all 187 patients, CSF PF1.2 was increased in patients with CE on MRI (n=71; 147.38 pmol/l; IQR 83.68-215.36) compared to patients without CE (n=86; 100.03 pmol/l; IQR 33.87-162.80; p=.008). CSF PF1.2 correlated significantly with Qalb (ρ=.445; p<.001). No differences of CSF PF1.2 levels were observed between RRMS (131.48 pmol/l, IQR 42.75–204.10) and disease controls (102.28 pmol/l; IQR 55.60-159.94; p=.606). ConclusionIn patients with autoimmune inflammatory CNS diseases PF1.2 correlated strongly with fibrinogen and D-dimer in CSF, indicating coagulation system activation. The findings suggest that thrombin generation might require acute BBB dysfunction to exert autoimmune effects in the CNS.