Neurofilament Light Protein: A Marker for Injury Severity, Clinical Course, and Outcome Following Moderate to Severe Traumatic Brain Injury

BACKGROUND: Recent studies demonstrate increased neurofilament light (NfL) levels across all types of traumatic brain injury (TBI). We evaluated inter-relationships between acute temporal NfL profiles with dynamic Intensive Care Unit (ICU) neurophysiological burden and tested the hypothesis that ICU burden is a significant, modifiable factor mediating NfL relationships to outcome. METHODS: Stored serum collected days 0-10 from 97 patients with severe TBI were analyzed. Temporal NfL patterns were examined using group-based trajectory (TRAJ) analysis. Mean NfL was calculated over days 0, 1-5, 6-10, and 0-10; ICU injury burden scores were generated using ridge regression associations with Glasgow Outcome Scale Extended (GOSE) over the same time-bins. We used univariable and multivariable regression to assess NfL TRAJ and mean values with respect to ICU injury burden, demographics, clinical factors, and outcome. NfL TRAJ and days 0-10 ICU injury burden were added to a baseline model to evaluate outcome discrimination. Multivariate mediation analysis was applied to assess the causal effects of ICU injury burden on NfL associations with GOSE. RESULTS: Of the patients, 63 had low NfL temporal profile, and 34 had higher profile (high-TRAJ). Individuals in the high-TRAJ group had lower Glasgow Coma Scale scores (GCSs), more frequently required barbiturate coma or decompressive surgery, and had worse 6-month outcomes. NfL means at all time epochs were significantly associated with unfavorable GOSE in univariate regression. In stepwise multivariate logistic regression, the area under the curve (AUC) increased progressively from ~ 0.74 to ~ 0.87 with the addition of computed tomography (CT) burden, NfL TRAJ, and ICU injury burden. NfL means and ICU injury burden were both negatively correlated with the initial GCS and positively correlated with CT burden, which was the most consistently significant baseline predictor. Covariate-adjusted mediation analysis suggests that ICU injury burden was a causal factor contributing to both CT burden and NfL-TRAJ associations with unfavorable GOSE. CONCLUSIONS: Temporal NfL profiles provide novel information about intracranial pathophysiology as a causal (and potentially modifiable) mediator of axonal injury following severe TBI. In conjunction with standard-of-care clinically derived metrics, NfL improved explanation of variance in outcomes.