Association of Serum Brain-Derived Tau With Clinical Outcome and Longitudinal Change in Patients With Severe Traumatic Brain Injury

IMPORTANCE: Blood-based measurements of total tau (T-tau) are commonly used to examine neuronal injury in patients with traumatic brain injury (TBI), but current assays do not differentiate between brain-derived tau (BD-tau) and tau produced in peripheral tissues. A novel assay for BD-tau has recently been reported that selectively quantifies nonphosphorylated tau of central nervous system origin in blood samples. OBJECTIVES: To examine the association of serum BD-tau with clinical outcomes in patients with severe TBI (sTBI) and its longitudinal changes over 1 year. DESIGN, SETTING, AND PARTICIPANTS: This prospective cohort study was conducted at the neurointensive unit at the Sahlgrenska University Hospital, Gothenburg, Sweden, between September 1, 2006, and July 1, 2015. The study included 39 patients with sTBI followed up for up to 1 year. Statistical analysis was performed between October and November 2021. EXPOSURES: Serum BD-tau, T-tau, phosphorylated tau(231) (p-tau(231)), and neurofilament light chain (NfL) measured on days 0, 7, and 365 after injury. MAIN OUTCOMES AND MEASURES: Associations of serum biomarkers with clinical outcome and longitudinal change in sTBI. Severity of sTBI was evaluated using the Glasgow Coma Scale at hospital admission, while clinical outcome was assessed with the Glasgow Outcome Scale (GOS) at 1-year follow-up. Participants were classified as having a favorable outcome (GOS score, 4-5) or unfavorable outcome (GOS score, 1-3). RESULTS: Among the 39 patients (median age at admission, 36 years [IQR, 22-54 years]; 26 men [66.7%]) in the study on day 0, the mean (SD) serum BD-tau level was higher among patients with unfavorable outcomes vs those with favorable outcomes (191.4 [190.8] pg/mL vs 75.6 [60.3] pg/mL; mean difference, 115.9 pg/mL [95% CI, 25.7-206.1 pg/mL]), while the other markers had smaller between-group mean differences (serum T-tau, 60.3 pg/mL [95% CI, −22.0 to 142.7 pg/mL]; serum p-tau(231), 8.3 pg/mL [95% CI, −6.4 to 23.0 pg/mL]; serum NfL, −5.4 pg/mL [95% CI, −99.0 to 88.3 pg/mL]). Similar results were recorded on day 7. Longitudinally, baseline serum BD-tau concentrations showed slower decreases in the whole cohort (42.2% on day 7 [from 138.6 to 80.1 pg/mL] and 93.0% on day 365 [from 138.6 to 9.7 pg/mL]) compared with serum T-tau (81.5% on day 7 [from 57.3 to 10.6 pg/mL] and 99.0% on day 365 [from 57.3 to 0.6 pg/mL]) and p-tau(231) (92.5% on day 7 [from 20.1 to 1.5 pg/mL] and 95.0% on day 365 [from 20.1 to 1.0 pg/mL]). These results did not change when considering clinical outcome, where T-tau decreased twice as fast as BD-tau in both groups. Similar results were obtained for p-tau(231). Furthermore, the biomarker levels on day 365 were lower, compared with day 7, for BD-tau but not T-tau or p-tau(231). Serum NfL had a different trajectory to the tau biomarkers, with levels increasing by 255.9% on day 7 compared with day 0 (from 86.8 to 308.9 pg/mL) but decreasing by 97.0% by day 365 vs day 7 (from 308.9 to 9.2 pg/mL). CONCLUSIONS AND RELEVANCE: This study suggests that serum BD-tau, T-tau, and p-tau(231) have differential associations with clinical outcome and 1-year longitudinal change in patients with sTBI. Serum BD-tau demonstrated utility as a biomarker to monitor outcomes in sTBI and can provide valuable information regarding acute neuronal damage.