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

Key Points Question Are levels of serum brain-derived tau (BD-tau) at admission associated with clinical outcome and long-term change in patients with severe traumatic brain injury (sTBI)? Findings In this cohort study of 39 patients with sTBI, the mean fold difference in serum BD-tau concentrations on day 0 for patients with sTBI with unfavorable clinical outcomes vs those with favorable clinical outcomes 1 year after the injury was higher than the mean fold differences in serum total tau and phosphorylated tau231. Serum BD-tau demonstrated slower clearance from the blood (56.6% of baseline levels remaining by day 7) compared with total tau and phosphorylated tau231, which had only 19.0% and 7.5% of baseline levels, respectively, remaining at day 7. Meaning This study suggests that concentrations and longitudinal trajectories of serum BD-tau differ among patients with sTBI depending on clinical outcome; serum BD-tau could be used as an accessible biomarker to monitor clinical outcome in patients with sTBI at admission and 7 days after the injury.


Introduction
Traumatic brain injury (TBI) is one of the leading causes of morbidity, disability, and mortality across all ages. 1,2Around the world, more than 50 million individuals are affected by TBI every year. 2 Posttraumatic complications of TBI can range from minor neurological and psychosocial problems to long-term disability, 3 making it crucial to follow up with patients after injury to ascertain longitudinal outcomes.
Traumatic brain injury is often classified as mild or severe according to the intensity of the injury. 4Severe TBI (sTBI) can be more life threatening and has lower rates of survival. 2In clinical settings, sTBI is commonly classified using the Glasgow Coma Scale (GCS) at admission to the hospital, while the Glasgow Outcome Scale (GOS) is used to assess long-term clinical outcome. 5reover, structural damage after sTBI may be detected by neuroimaging techniques. 6Despite the proven effectiveness of these approaches, they have limitations in providing biochemical brainrelated changes reflected in the bloodstream within a few hours after trauma.Circulating blood biomarkers provide biochemical information and prognostic insights into clinical severity to guide patient management and monitor long-term outcome. 4rum total tau (T-tau) is one of the most well-characterized biomarkers for sTBI, 4,7,8 showing high increases within hours of the injury. 7However, studies have suggested that current assays for T-tau quantify both central nervous system (CNS) and peripheral tau when measurements are performed on blood (serum or plasma) samples. 9,10Therefore, we hypothesized that a blood-based biomarker that is selective for CNS tau will be more accurate at reflecting the brain-associated tau released into the bloodstream while avoiding potential influences from peripheral tau.To this end, we evaluated the association of the novel brain-derived tau (BD-tau) marker 11 with baseline clinical severity and longitudinal outcome compared with T-tau in serum samples from participants with sTBI followed up clinically over a 1-year period.We also examined changes in serum phosphorylated tau 231 (p-tau 231 ) and neurofilament light chain (NfL) concentrations as other neuronal injury-related markers.

Study Cohort, Design, and Outcome
This study included 42 participants (39 with data on all 4 serum biomarkers) from the prospective Swedish TBI Neurointensivvårdsavdelning cohort of patients with sTBI who were receiving clinical care at the Sahlgrenska University Hospital, Gothenburg, Sweden, and followed up for 1 year. 12,13rticipant recruitment, clinical assessments, and blood sample collection were performed between September 1, 2006, and July 1, 2015.The inclusion criteria were (1) TBI with a GCS score of 8 or less on admission, (2) admission to the neurointensive care unit within 48 hours of head injury, (3) aged 18 years or older, (4) acceptance from next of kin to participate in the study, and (5) residence in Sweden for 12 months of follow-up.The exclusion criteria included no provision of informed consent, known history of neurological and/or autoimmune disease, and pregnancy.The ethics committee at the University of Gothenburg approved the study.Written informed consent was obtained from the Traumatic brain injury outcome was clinically assessed with the GOS at 12 months 5 ; those with a GOS score of 1 to 3 were classified as having an unfavorable outcome, and those with a GOS score of 4 to 5 were classified as having a favorable outcome.The 12-month outcome assessments were collected using a mixed-methods approach, including interviews performed either in person or via telephone.For participants with substantial impairment, their proxies were interviewed.
There were 39 participants on day 0, 39 on day 7, and 15 on day 365.Loss at follow-up was mainly due to death or disability, particularly in the group with unfavorable outcomes.

Blood Sample Handling and Biomarker Measurements
Serum samples were obtained at the indicated time points according to standard procedures and stored at −80 °C until use.Serum BD-tau and p-tau 231 were measured on the Simoa HD-X platform (Quanterix) using validated in-house assays, 11,14 and T-tau and NfL with Quanterix assays (Nos. 101552 and 103670, respectively).

Statistical Analysis
Biomarker measurements and statistical analyses were performed between October and November 2021 at the University of Gothenburg with Prism, version 9.3.1 (GraphPad).The distributions of data sets were examined for normality using the Kolmogorov-Smirnov test.Because the data were nonnormally distributed, nonparametric tests were used, and continuous data are presented as median (IQR) values.To compare serum biomarker levels between 2 groups (ie, unfavorable and favorable outcome groups at each time point), the mean fold differences (95% CI) were calculated and statistical comparisons examined using the Mann-Whitney test.For examining biomarker levels at all 3 time points (days 0, 7, and 365) within the whole cohort or the specific outcome groups, the Kruskal-Wallis test with the Dunn multiple comparison was used.P values (including those adjusted for multiple comparisons) were considered significant at the 2-sided P < .05level.

Cohort Characteristics
The study included 42 participants with at least 1 biomarker measured at baseline.

Serum BD-Tau Levels in sTBI Clinical Outcome Groups on Admission and 7 Days Later
Initial levels of BD-tau (on days 0 and 7) were associated with GOS outcome at 1 year.Thus, on day 0, mean (SD) serum BD-tau levels were higher in the unfavorable outcome group (191.4 [190.8][from 44.0 to 13.0 pg/mL]) outcome group (Table ).However, despite the decrease in concentrations, the mean differences between outcome groups were similar at days 0 and 7.By day 365, serum BD-tau levels in both groups had further decreased to concentrations that were much lower than the corresponding day 0 and day 7 levels (Figure , A).
Serum T-tau and p-tau 231 levels were also decreased between days 0 and 7 (  , B and C).Another distinction from BD-tau levels was that the mean differences between groups on days 0 and 7 tended to vary (Table ).Because of the decrease in concentration for T-tau and p-tau 231 from day 0 to day 7, the between-group mean differences on day 7 vs day 365 were similar.

Discussion
In the present study, we found that serum BD-tau level could have utility for evaluating clinical outcome in sTBI, both on the day of the event and 7 days later.These results, which were not observed for serum T-tau level, suggest that the selective measurement of tau of CNS origin in the bloodstream has the capacity to improve the accuracy of the clinical outcome and management of sTBI.In agreement with recent findings indicating that current blood-based T-tau assays quantify tau of both CNS and peripheral origin and that the latter makes up approximately 80% of T-tau signal in the bloodstream, 15 our results suggest that CNS tau differences in groups of patients with sTBI and in different clinical outcomes can be masked if a nonselective blood-based tau assay (ie, T-tau) is used.
In addition, the inability of p-tau 231 and NfL to differentiate between the clinical outcome groups suggests their limited value for the clinical evaluation of sTBI, despite their well-validated functions for Alzheimer disease pathophysiology and general neurodegeneration, respectively. 14,16e results indicate that all 3 tau biomarkers (BD-tau, T-tau, and p-tau 231 ) are released from the brain into the bloodstream within minutes to hours of sTBI, possibly due to the opening of the blood-brain barrier.This initial increased release of both total (unphosphorylated) and phosphorylated forms of tau agrees with previous reports showing that brain trauma leads to the rapid release of tau of various molecular forms into extracellular fluids. 7,8The consistent longitudinal reduction in these biomarker levels was due to a lack of replenishment of the initial (day 0) signals during physiologically regulated tau turnover. 7Serum T-tau was cleared much more rapidly (81% removed by day 7) than BD-tau, which could be explained by the ratio of CNS tau to peripheral tau in the bloodstream, which was in favor of CNS tau on day 0 (due to increased release of brain tau) returning to pre-sTBI levels over time.However, BD-tau, which exclusively quantifies brainoriginating tau, 11 showed that CNS tau is not cleared so quickly and that substantial amounts do remain for up to 1 year.This slow clearance of BD-tau proved useful for the clinical monitoring of outcome and recovery after sTBI.For example, while considering T-tau alone might suggest recovery by day 7 (due to significantly decreased levels of BD-tau compared with day 0), BD-tau indicates otherwise because the levels were still statistically indifferent from day 0 regardless of clinical outcome.Continuous evaluation of BD-tau levels between days 7 and 365 would be informative to ascertain the point at which the decrease was significantly lower compared with days 0 and 7 and whether patients with favorable outcomes reached this point earlier than those with unfavorable outcomes.We also anticipate that individuals with mild TBI will show faster decreases in BD-tau compared with those with sTBI.Finally, NfL had a different trajectory, similar to previous reports, 12 suggesting slower release into the bloodstream with the tau markers.However, the peaking of the signal at 7 and its difference from day 365 suggests its value for outcome monitoring after sTBI.

Strengths and Limitations
This study has some strengths, including the longitudinal design and that multiple biomarkers were compared head to head.It also has some limitations, including the lack of sampling time points between days 7 and 365, as well as the restriction of the study to patients with sTBI without including those with mild TBI.Moreover, many participants were lost at follow-up due to death and disability.
Despite the important and novel information provided in this study, the results should be

JAMA Network Open | Neurology
Brain-Derived Tau and Longitudinal Change Among Patients With Severe Traumatic Brain Injury In this cohort study of 39 patients with sTBI, the mean fold difference in serum BD-tau concentrations on day 0 for patients with sTBI with unfavorable clinical outcomes vs those with favorable clinical outcomes 1 year after the injury was higher than the mean fold differences in serum total tau and phosphorylated tau 231 .Serum BD-tau demonstrated slower clearance from the blood (56.6% of baseline levels remaining by day 7) compared with total tau and phosphorylated tau 231 , which had only 19.0% and 7.5% of baseline levels, respectively, remaining at day 7. Meaning This study suggests that concentrations and longitudinal trajectories of serum BD-tau differ among patients with sTBI depending on clinical outcome; serum BD-tau could be used as an accessible biomarker to monitor clinical outcome in patients with sTBI at admission and 7 days after the injury.
Longitudinal trajectories of serum biomarker levels following severe TBI, in the whole cohort and according to clinical outcome.The plots show the median serum concentrations of brain-derived tau (BD-tau) (A), total tau (T-tau) (B), phosphorylated tau 231 (p-tau 231 ) (C), and neurofilament light chain (NfL) (D).In each plot, the serum biomarker values at different time points (on days 0, 7, and 365) are shown for the whole cohort (left) as well as in the 2 clinical outcome groups.

JAMA Network Open | Neurology Brain
-Derived Tau and Longitudinal Change Among Patients With Severe Traumatic Brain Injury

by a University College London User on 07/23/2023
The longitudinal trajectory of serum NfL was different from the longitudinal trajectories of the tau-based biomarkers.Instead of decreasing from day 0 to day 7, serum NfL increased by 255.5% (from 86.8 to 308.9 pg/mL).There were increases of 156.0%(from 91.1 to 233.2 pg/mL) and 343.5% (from 85.8 to 380.5 pg/mL) from day 0 in the favorable and unfavorable outcome groups, respectively (Tableand Figure).The highest levels were recorded on day 7 in the whole cohort and in both clinical outcome groups, with increased mean differences between days 0 and 7 (Table).The levels decreased by 97.0% (from 308.9 to 9.2 pg/mL) from day 7 to day 365 (TableandFigure, D).