Neurofilament light, glial fibrillary acidic protein, and tau in a regional epilepsy cohort: High plasma levels are rare but related to seizures

Objective: Higher levels of biochemical blood markers of brain injury have been described immediately after tonic-clonic seizures and in drug-resistant epilepsy, but the levels of such markers in epilepsy in general have not been well characterized. We analyzed neurofilament light (NfL), glial fibrillary acidic protein (GFAP), and tau in a regional hospital-based epilepsy cohort and investigated what proportion of patients have levels suggesting brain injury, and whether certain epilepsy features are associated with high levels.Methods: Biomarker levels were measured in 204 patients with an epilepsy diagnosis participating in a prospective regional biobank study, with age and sex distribution correlating closely to that of all patients seen for epilepsy in the health care region. Absolute biomarker levels were assessed between two patient groups: patients reporting seizures within the 2 months preceding inclusion and patients who did not have seizures for more than 1 year. We also assessed the proportion of patients with above-normal levels of NfL.Results: NfL and GFAP, but not tau, increased with age. Twenty-seven patients had abnormally high levels of NfL. Factors associated with such levels were recent seizures (p = .010) and epileptogenic lesion on radiology (p = .001). Levels of NfL (p = .006) and GFAP (p = .032) were significantly higher in young patients (<65 years) with seizures & LE;2 months before inclusion compared to those who reported no seizures for >1 year. NfL and GFAP correlated weakly with the number of days since last seizure (NfL: r(s) = -.228, p = .007; GFAP: r(s) = -.167, p = .048) in young patients. NfL also correlated weakly with seizure frequency in the last 2 months (r(s) = .162, p = .047).Significance: Most patients with epilepsy do not have biochemical evidence of brain injury. The association with seizures merits further study; future studies should aim for longitudinal sampling and examine whether individual variations in NfL or GFAP levels could reflect seizure activity.