Content uploaded by Jussi Posti
Author content
All content in this area was uploaded by Jussi Posti on Apr 05, 2022
Content may be subject to copyright.
1
Blood-based biomarkers and traumatic brain injury––a clinical perspective
Jussi P. Posti, MD, PhD
Neurocenter, Department of Neurosurgery and Turku Brain Injury Center, Turku University Hospital and University of Turku,
Finland
Olli Tenovuo, MD, PhD
Neurocenter, Turku Brain Injury Center, Turku University Hospital and University of Turku, Finland
Manuscript classification: Review article
Running title: TBI biomarkers, a clinical perspective
Corresponding author: Jussi Posti, Neurocenter, Department of Neurosurgery and Turku Brain Injury Centre, Turku University
Hospital, P.O. Box 52, FI-20521 Turku, Finland; jussi.posti@utu.fi
Word count: 5670; Title character count: 65; Abstract word count: 132
Number of references: 80; Number of figures: 0; Number of tables: 2
Sources of funding statement: Dr. Posti has received funding from Academy of Finland (#17379), Government’s Special Financial
Transfer tied to academic research in Health Sciences, Finland (#11129) and the Maire Taponen Foundation.
Acknowledgements: None
Conflict of interests: None
2
Abstract
Blood-based biomarkers are promising tools to complement clinical variables and imaging findings in the diagnosis,
monitoring, and outcome prediction of traumatic brain injury (TBI). Several promising biomarker candidates have been
found for various clinical questions, but the translation of TBI biomarkers into clinical applications has been negligible.
Measured biomarker levels are influenced by patient-related variables such as age, blood-brain barrier integrity, and renal
and liver function. It is not yet fully understood how biomarkers enter the bloodstream from the interstitial fluid of the
brain. In addition, the diagnostic performance of TBI biomarkers is affected by sampling timing and analytical methods.
In this focused review, the clinical aspects of glial fibrillary acidic protein, neurofilament light, S100 calcium-binding
protein B, tau, and ubiquitin C-terminal hydrolase-L1 are examined. Current findings and clinical caveats are addressed.
Keywords: traumatic brain injury, biomarkers, diagnosis, outcome
3
Introduction
Traumatic brain injury (TBI) is one of the most complex diseases. Since the advent of imaging modalities such as
computed tomography (CT) and magnetic resonance imaging (MRI), there has been a call for complementary diagnostic
tools to assess the severity of the injury and predict outcome more accurately. The progress has been slow. Due to the
lack of laboratory methods to reveal and monitor ongoing pathophysiological processes, the treatment of TBI still remains
largely symptomatic and reactive rather than proactive.
A major obstacle to the development of diagnostic, stratification, and prognostic tools for TBI has been the heterogeneous
nature of the disease.1 It ranges from diffuse injuries, focal contusions, various types of haemorrhagic mass lesions to
penetrating injuries, and from single to repetitive injuries. No two injuries are identical. Demographic and genetic
differences, as well as common concomitant injuries and treatment heterogeneity, add to this complexity.2 In many ways,
TBI lags behind other medical and public health challenges, especially in relation to its incidence and global impact. For
example, cardiovascular diseases can largely be diagnosed in a pathophysiologically meaningful way, and their response
to treatment is usually easily measured. In the case of TBI, the treatment is inherently about treating the consequences
and symptoms of pathology once it has developed––rather than identifying patients in whom a pathological process will
develop and thus preventing it or controlling the ongoing pathophysiology. Diagnostic tools for characterising severity
and progression of TBI remain simplistic, inaccurate, and rely largely on nonspecific symptoms. They can also be
considered as snapshots that are not able to depict the dynamic nature of the disease.
Blood-based biochemical markers are promising tools to complement clinical variables and imaging findings in the
diagnosis, monitoring, and outcome prediction of TBI. In TBI, neuronal, astroglial, and oligodenroglial cells as well brain
vasculature are damaged by axial and extra-axial lesions and shear forces, which variously contribute to the clinical course
and symptomatology, as well as fate of brain tissue and function, which mainly determine patient recovery. There is an
unmet clinical need for biomarkers that could detect the presence of TBI in patients with polytrauma, assess the true initial
severity of TBI, predict disease progression, and improve prediction of outcomes. Treatment monitoring is almost fully
based on gross physiology and clinical symptoms, as if patients with acute myocardial infarction would be treated based
on pulse and blood pressure monitoring and severity of chest pain. In most severe cases of TBI, intracranial pressure and
brain oxygen can be monitored, without possibilities to prevent imminent secondary injuries or to detect cascades leading
to common brain emergencies.
Several promising blood-based biomarker candidates have been found for various clinical questions3,4, but the translation
of TBI biomarkers into clinical applications has been negligible. Inconsistent research methods, such as variable sampling
times, variably representative and poorly characterised study cohorts, and use of heterogeneous analytical kits and
instruments across studies, have hindered this process. Therefore, clinical interpretation of the current body of literature
is challenging. Perhaps the most significant problem with previous studies has been the focus on a single sampling point
rather than longitudinal sampling. The greatest work remains to be done in determining the optimal sampling timing for
various biomarkers and understanding the underlying biological processes reflected in them.5 Measured levels may be
influenced by patient age6,7, integrity of the blood-brain barrier (BBB)8, glymphatic system function9, extracellular
proteolysis10, and renal and hepatic function11,12. This list is not exhaustive, and these different modifiers are not mutually
exclusive. The transport of biomarkers from the brain to the bloodstream is much more complex than that from internal
organs. It is not yet fully understood how biomarkers accumulated in the interstitial fluid of the brain after TBI enter the
4
bloodstream, and how injuries or diseases may modify this process. Many studies have reported half-lives for the most
studied biomarkers. However, TBI as well as frequent concomitant injuries, dynamic nature of TBI itself and given
treatments may greatly alter the mechanisms behind biomarker half-lives, e.g., by affecting protease activities and altering
clearance.12 Thus, an injury-dependent half-life may be very different from the half-life in physiological conditions.
Over the many years of disease-oriented, blood-based biomarker research, the understanding of what various biomarkers
can do has changed dramatically. No longer is it expected that a “brain troponin” can solve most clinical questions, but a
new understanding of the complementary role of biomarkers in TBI medicine has evolved, they are being an important
part of a systems medicine approach for this complex disease. In this focused review, we discuss the current understanding
of blood-based protein biomarkers in TBI. We address the feasibility of biochemical diagnostics in various clinical
settings and consider where the greatest efforts in biomarker research should be directed.
Clinical background
TBI is initiated by an external physical force to the head and causes immediate structural damage and dysfunction of the
brain to varying degrees.13 This primary injury is followed by a complex series of events that include structural, functional,
neurovascular, metabolic, and inflammatory changes. These events are vaguely called secondary brain injury. Many of
these initially aim to limit damage and restore homeostasis and functionality, but they may also contribute further to brain
damage and dysfunction.12,14 These processes are reflected in the levels of brain-enriched proteins, genetic markers (such
as microRNAs), exosomes and metabolites that are released into the bloodstream.
Over 90% of patients who sustain TBIs are classified in the mild end of the spectrum (mTBI), including those injuries
labelled as concussions. The vast heterogeneity of TBIs is reflected in the often obscure and inconsistent definitions.
Concussion is a clinical syndrome in which there are mainly slight and presumably transient symptoms caused by external
impact to the head. There is no universal definition of concussion, and its underlying pathology is still poorly known.
Generally, in mTBI, the duration of unconsciousness is a few minutes (up to 30 minutes), and post-traumatic amnesia is
up to 24 hours. These limits are artificial and not based in any scientific evidence, and the underlying pathophysiology of
unconsciousness or amnesia is not properly understood. In many cases of mTBI, patients may not lose consciousness, or
reliable information on this is impossible to get. An mTBI diagnosis is challenging in the emergency department setting
because neurological symptoms are nonspecific, neurological signs are often absent, confounding conditions are common,
and clinical imaging tools are usually unable to reveal the injury. Yet even an mTBI or concussion may result in complex
events at molecular, cellular and tissue levels, leading to persistent or permanent symptoms, threatening the person’s well-
being, working ability, and quality of life. These events can potentially be measured and monitored using blood-based
biomarkers.
In concussion and mTBI, the most important clinical issues for which blood-based biomarkers can potentially be used are
determining the true severity of the injury and identifying patients who require further measures, such as brain imaging,
sick leave, or follow-up. Verification of the presence of TBI is particularly important in patients who have concomitant
injuries that also may cause alterations in consciousness and cognition, such as polytrauma, and painful orthopaedic
injuries. Another diagnostically challenging group is patients who are inebriated or under influence of drugs. Patients
who have such diagnostic confounders and an mTBI, are often at risk for developing secondary injuries that may
exacerbate the injury. Although most patients with mTBI recover well, a substantial percentage still have post-injury
5
symptoms after one year.15 Despite some recovery may take place also after one year16, it has been shown that there is
also a risk for long-term or late decline.17 Predicting outcome after mTBI is challenging, whether outcome is measured
by time to recovery, persistence of symptoms, or development of long-lasting symptoms. Reliable models for predicting
the outcome of mTBI are lacking, although preinjury characteristics have some predictive value18. There is also an unmet
clinical need to reliably assess the late phase of TBI. A minority of patients with TBI may develop progressive brain
volume loss and neurological deterioration19, where underlying pathophysiology is still largely unconfirmed.
The diagnosis of acute moderate and severe TBI does not require biochemical diagnostics. The decision to perform an
acute head CT is also easy to make in these patients. In these more severe forms of TBI outcome prediction and monitoring
of treatment effects are the most important clinical questions that may be answered with biomarker assessment alone or
together with clinical, physiological, or imaging covariates. It is noteworthy that the only blood-based biomarker
application used and officially approved in clinical practice to date is to assess the need for head CT imaging.20,21
The most common metric to describe the diagnostic performance of biomarkers is the area under receiver operating
characteristic curve (AU-ROC), which illustrates the diagnostic ability of a binary classifier system as its discrimination
threshold varies. There are no absolute standards for using the AU-ROC to evaluate the predictive ability of a diagnostic
test. An AU-ROC value of 0.80 or above generally indicates that the test is clinically useful.22 However, in many cases
in clinical medicine, a point on the AU-ROC curve where sensitivity is high (e.g., above 90%) also gives a good indication
of the diagnostic or prognostic ability of the biomarker as sensitivity percentage. Yet. statistical sensitivity cannot directly
be translated into clinical usability, which depends also on the nature of the clinical question. In high-risk questions,
clinicians may expect 100% sensitivity.
Most studied biomarkers
This review focuses on the five most studied TBI biomarkers because of their potential clinical relevance. The main
properties of the biomarkers presented below are summarised in Table 1.
Glial fibrillary acid protein (GFAP) is a monomeric filament protein of the cytoskeleton that is expressed in astrocytes in
the white and grey matter.23 Regulation of GFAP increases during astrogliosis and GFAP is released into the bloodstream
after TBI both as an intact protein (50 kDa) and as a degradation product (18–42 kDa).24,25 GFAP is detectable within one
hour of injury and peaks within 20–24 hours with proposed half-life of 24–48 hours.26,27 GFAP is also expressed in cells
outside the central nervous system, such Schwann cells, chondrocytes, testicular Leydig cells, enteric glia, and liver and
pancreas.28 GFAP levels in blood after TBI are affected by age.6 Persistently high levels of GFAP have been reported
after TBI.19
The neurofilament protein exists in three forms: the 200 kDa heavy chain, the 150 kDa medium chain, and the 68 kDa
light chain (NF-L). NF-L is a component of the axonal scaffold of neurons.29 The half-life of NF-L is not known, but it
is apparent that it is among the longest of the currently known TBI biomarkers. NF-L levels are increased after TBI in
both blood and cerebrospinal fluid.30 Although NF-L is highly brain-enriched, plasma levels of NF-L have been shown
to be also specific for peripheral axonal damage.31 There is also evidence that extracranial injury may affect its levels.32
NF-L levels in blood after TBI are affected by age.7 NF-L levels appear to increase in the first two weeks after injury and
decrease thereafter though persistently high levels have been reported.19,33
6
S100 calcium-binding protein B (S100B) is an 11 kDa low molecular weight calcium-binding protein expressed in glial
and Schwann cells. S100 proteins are involved in the regulation of a variety of processes, such as cell cycle progression
and differentiation.34 Blood levels of S100B increase within one hour after TBI with a peak in less than six hours35 and
half-life of 30 minutes to 2 hours. Blood levels of S100B increase during various athletic activities and are also elevated
in patients who have sustained injuries to body parts other than the head due to expression outside the central nervous
system.36,37 S100B levels in blood after TBI are affected by age.38 Immediately after TBI with extracranial injuries,
extracranial sources of S100B efflux may be significant.28 After TBI, levels in cerebrospinal fluid may be up to 100 times
higher than in blood.36 Salivary testing for S100B has also been proposed as an alternative to S100B blood testing in the
diagnosis of TBI.39
Tau is a microtubule-associated protein expressed in the axons and axon terminal cytoskeleton in the central and
peripheral nervous systems.40,41 In the central nervous system, there are six isoforms with molecular weights ranging from
33 kDa to 46 kDa.40 Most tau biomarker research in neurotrauma has focused on total tau (T-tau).42 TBI results in both
hyperphosphorylation and cleavage of tau proteins. Therefore, research interest has also focused on both cleaved tau (c-
tau) and phosphorylated tau (p-tau).3,43 The current literature is insufficient to determine the half-life of tau after TBI. Tau
levels appear to peak at 12–24 hours, but levels decline relatively slowly and persistently high levels have been reported43–
45 Tau levels in blood after TBI increase with aging.46
Ubiquitin C-terminal hydrolase-L1 (UCH-L1) is a 25-kDa neuronal enzyme involved in the addition and removal of
ubiquitin from proteins destined for internal metabolism within neurons.47 UCH-L1 is detectable within one hour from
TBI, peaks eight hours after TBI and has a fairly short half-life among TBI biomarkers, 7–9 hours.48,49 UCH-L1 is also
expressed in cells outside the central nervous system, such as in testis, ovary, and kidney.50,51
Biomarkers in mild TBI diagnostics
Clinical questions related to traumatic brain injury biomarkers are summarized in Table 2. As discussed above, there are
important patient groups where detecting mTBI in emergency departments is important but difficult. Recognizing mTBI
is relevant in also many other circumstances, for example in various sports and in the military. However, the threshold
between a trivial head impact and an mTBI/concussion is very complex, as shown in many studies done mainly in athletes
(see below). Therefore, and considering the interindividual variations, defining relevant biomarker cut-offs for an mTBI
diagnosis is a challenging task.
Current evidence
Summarising existing research findings, blood-based biomarkers have shown somewhat disappointing accuracy in
distinguishing patients with concussion/CT-negative mTBI from healthy controls or orthopaedic injury controls. The
results are inconsistent because of differences in study settings, populations, and methods. Most AU-ROCs are less than
0.80, indicating only mediocre discrimination.
In a study with university student athletes, AU-ROC of the combination of GFAP and UCH-L1 in discriminating athletes
with concussion from contact sports controls in the acute postinjury period was 0.71.52 In another corresponding study,
both S100B and UCH-L1 had AU-ROCs of 0.79.53 T-tau was measured in blood samples from professional ice hockey
players within hours after a sports-related concussion, and the levels were increased compared with the preseason baseline
7
with an AU-ROC of 0.80.44 In a recent study, levels of GFAP, NF-L, and UCH-L1 were only modestly able to discriminate
U.S. cadets with acute concussion from those without (AU-ROCs 0.73–0.75), but when the biomarkers were combined,
AU-ROC increased to 0.80. In that study, T-tau did not show discriminatory power (AU-ROC 0.50).
Most studies including all severities of TBI have shown that many biomarkers can distinguish mTBI from moderate to
severe TBI––including the recent CENTER-TBI study, which includes the largest biomarker study cohort to date.4
Analogously with aforementioned concussion studies, it seems that distinguishing CT-negative patients with mTBI
acutely from patients with orthopaedic injuries using with clinically meaningful AU-ROCs may not be possible.54,55
Traumatic axonal damage, including diffuse axonal damage currently thought to be part of the pathophysiology of
concussion, may explain the significantly elevated levels of biomarkers in imaging-negative TBI even days after injury.
Clinical issues and confounding factors
One of the inadequacies of blood-based biomarker studies of mTBI is that most studies report values at a single time point
after injury, usually at admission or within a few hours from the injury. Studies at a single time point can provide important
information for diagnostics and outcome prediction but are just snapshots of the dynamic pathophysiology and thus
difficult to interpret reliably. Therefore, serial sampling consisting of at least two assessments will add the reliability.
Blood biomarker levels change significantly over time also after concussion and mTBI44,53,56,57, underpinning importance
of studies monitoring biomarker levels over time after such an event. This is especially important because patients with
mTBI typically seek for medical care with variable delays. Under what circumstances biomarkers indicate a significant
mTBI, whether biomarkers vary depending on the mechanism of injury (e.g., impact or deceleration), and whether
pathologic levels depend on, e.g., ethnicity or co-existing medical confounders (such as pre-injury health, medications,
and inebriation) remains to be determined. It is already apparent that age and extracranial injuries affect the levels of most
biomarkers discussed in this review (Table 1).
In the acute diagnostics of TBI, longitudinal sampling consisting of at least two samples and a panel of biomarkers may
be needed to assess the trend in biomarker levels and to provide clinically useful information to verify the presence of
TBI in a clinically complex setting. A recent study showed promising results for using repeated GFAP levels in the acute
diagnosis of stroke.58
The modest AU-ROCs in the above clinical questions may be related to the subtle biochemical response to concussion
and some mTBIs, or just reflect the heterogeneity of the patient population and sampling time after the injury. There are
also unexplained discrepancies in the existing studies––for example, AU-ROCs for T-tau have varied considerably
between fairly similar clinical settings when measured with similar assays.44,53 In our own studies, the levels of none of
the biomarkers have been normally distributed, which often stems from outliers who have abnormally high biomarker
levels. This was particularly evident in our previous study, where we observed unexpectedly high GFAP and UCH-L1
levels in some patients with orthopaedic injuries, although they had not any clinical signs of TBI, no head impact, and
showed normal head MRI.54 Indeed, most studies have shown that there is considerable variability in biomarker levels
also in controls, the reasons for which are still largely unknown.
Biomarkers in assessing the need for head imaging after mTBI
8
Determining the need for acute head imaging is a clinical question, which greatly benefits from a reliable biochemical
test with a single measurement in a predetermined time window. This is probably the most studied clinical question in
the field of TBI-related blood-based biomarker research. Most patients with symptoms suggestive of mTBI do not meet
the established clinical criteria for head CT. On the other hand, having sustained a head injury in leisure-time accident,
sporting event, or battlefield one must determine whether the patient can be followed up by nurses or paramedics in the
field or should be transferred to the nearest hospital with the capability of performing an emergency head CT. On the
other hand, most patients who undergo a head CT in the emergency department have a negative scan, thus being exposed
for seemingly unnecessary ionising radiation. Intracranial mass lesions or swelling that develop after an initially normal
CT scan are rare, and risk for neurosurgical measures after an initial normal head CT is negligible. To be clinically useful,
a biomarker for screening the need for acute head CT has to be rapidly and easily available, preferably with a portable
point-of-care device. More studies including serial samplings are needed to determine the optimal sampling window for
each potential biomarker.
Current evidence
The use of S100B in the emergency department has been recommended in the Scandinavian Guidelines for Initial
Management of Minimal, Mild and Moderate Head Injuries in Adults to preclude the need for head CT (cut-off = 0.1
ug/L) in patients with isolated mild head injury who are at low risk for intracranial hemorrhage and are seen within six
hours of injury. Three previous studies have shown that the use of S100B within the Scandinavian guidelines is a safe
and cost-effective method to reduce the number of unnecessary head CT scans.20,21,59
Later, GFAP has also been extensively studied in identifying patients with traumatic intracranial findings on head CT
after TBI. The recent ALERT-TBI study demonstrated that a blood test incorporating GFAP and UCH-L1 in identifying
CT-positive findings has better sensitivity (0.98) and specificity (0.36)60 than the S100B in the validation studies for the
Scandinavian guidelines (sensitivity 0.94 and specificity 0.19).20,21 The U.S. authorities (FDA) approved this test for
identifying patients requiring head CT. The role of UCH-L1 in this combination test has been questioned.61 Unlike the
Scandinavian guidelines, the FDA-approved test does not consider clinical covariates such as extracranial injury or other
clinical factors such as GCS score, injury mechanism, or use of anticoagulants that predispose to intracranial hemorrhage.
GFAP is not yet included in any clinical guidelines. In studies where GFAP has been evaluated in identifying patients
with traumatic intracranial findings on head CT without clinical covariates, it has outperformed (AU-ROCs 0.74–
0.89)4,48,62–65 NF-L (0.81–0.82)4,62, S100B (0.58–0.76)4,62,65, T-tau (0.78–0.82) 4,62, and UCH-L1 (0.62–0.83)4,48,63,64.
Clinical issues and confounding factors
This review article can only cover a small portion of the studies that address this topic. Because of the large number of
studies, AU-ROC results vary widely. This is largely explained by the fact that some of the studies were conducted only
in cohorts with mTBIs and some in TBIs of all severities. The latter studies have almost systematically higher AU-ROC
values, which are driven by severe injuries and higher biomarker levels. Another important cause for variation in AU-
ROC values is heterogeneous sampling times. Because of the rapid kinetics of S100B and UCH-L1, their performance
decreases already within the first 24 hours from injury. Extracranial injuries are also potential confounders because they
affect at least some biomarker levels.36,62 There is limited evidence if serum or plasma samples should be used. In a recent
study, serum and plasma GFAP had similar but not identical ability to discriminate between patients with mTBI with and
without intracranial abnormalities (AU-ROC serum 0.81; plasma 0.78).66 Age is an important confounding factor in this
9
clinical diagnostic question. The TRACK-TBI investigators observed that the ability of GFAP to discriminate between
CT -positive and CT-negative patients with mTBI decreased with age.6
Epidural hematoma is an important topic––although it is a progressive, space-occupying lesion, it may result in minimal
injury to the parenchyma with minor elevations in blood biomarkers at the time of initial evaluation in the emergency
department. Yet, the clinical outcome of an overlooked epidural hematoma may be catastrophic. Because of the rarity of
epidural hematomas, the literature is insufficient to reliably assess the diagnostic performance of any biomarker in
identifying patients with these mass lesions.
The only blood-based biomarker that has been used in clinical practice is S100B, which is used to identify selected patients
with mTBI in need for a head CT in some Scandinavian and European countries.20,21 S100B has excellent negative
predictive value for pathologic intracranial CT findings in this selected group of patients. However, S100B must be used
together with clinical covariates. Its adoption in hectic emergency medicine has been slow because of the lack of a rapid
point-of-care test. Although the sensitivities of both the GFAP–UCH-L1 test and the Scandinavian Guidelines with S100B
are adequate for clinical use, the specificities unfortunately remain low. Two-thirds (GFAP–UCH-L1)60 or 4/5
(S100B)20,21 tested positive for head imaging eventually show a normal head CT, why clinicians easily discontinue using
the test despite savings in cost and radiation exposure.
Biomarkers in outcome prediction and late assessment of traumatic brain injury
The outcome of TBI is a complex issue with various definitions, from brain tissue fate to subjective symptoms and quality
of life. Acute severity indices have shown only modest association with outcome, especially in patients with mTBI.
However, also in patients with moderate–severe TBI, clinical predictors and imaging combined explain only 35% of the
outcome variance.67 In patients with mTBI, the most important predictions concern the risk of long-term or permanent
sequelae or time to full recovery. In more severe cases, the risk of death, need for neurosurgical intervention, or permanent
dependence on others are the most important. Even in clearly severe TBI, some patients can make a full recovery, and
these cases are almost impossible currently identify in the acute setting. There is also a risk for withdrawing from medical
care in cases who are deemed unsalvageable, although this estimate is not based on measurable firm evidence.
Most studies use the overall or functional outcome defined by the Glasgow Outcome Scale (GOS) or its extended version
(GOSE). In case of mTBI, functional outcome is usually divided into complete recovery (GOSE 8) and incomplete
recovery (GOSE 1–7), whereas in moderate to severe traumatic brain injury, a distinction is usually made between
favourable (GOSE 5–8) and unfavourable outcome including death (GOSE 1–4).
Different types of outcome prediction questions provide an excellent place for biomarkers because they can inform about
events that are beyond the predictive power of clinical covariates. The best-known clinical prediction models for moderate
and severe TBIs are the International Mission for Prognosis and Analysis of Clinical Trials in TBI (IMPACT) prediction
model68 and the Corticosteroid Randomization After Significant Head injury (CRASH)69 model. These clinical models
are based on clinical variables at admission and do not have the same potential dynamic prognostic power as serial
biomarker samples. To be clinically meaningful, any outcome prediction should have a very high reliability. This is
mandatory when making decisions on potentially lifesaving but risky treatment decisions. These kinds of questions
always also need quick decisions, why biomarker use in these setting requires rapid results.
10
There is also an important but unmet clinical need to evaluate the late state of an initially mTBI in the chronic phase in
cases where head imaging results are negative, initial clinical records are inaccurate/insufficient to assess the duration of
posttraumatic amnesia and/or level of consciousness or these assessments are impossible because of confounders. Several
blood-based biomarkers are elevated in chronic neuroinflammatory and neurodegenerative diseases, and some are
currently under investigation as markers for diagnosis and disease progression for these conditions.70 There is also an
association between TBI and later neurodegenerative disease, because TBI may act as a trigger for progressive neurologic
deterioration in some individuals, with chronic neuroinflammation as a probable mechanism.71
Current evidence
GFAP correlates with outcome in both mild and moderate-severe TBI. GFAP is reported to modestly discriminate
between patients with favourable and unfavourable outcome (assessed at 6–12 months after injury) in mTBI, with an AU-
ROC of 0.76. In this study, NF-L outperformed GFAP with an AU-ROC of 0.83. Furthermore, in the same study, NF-L
predicted complete recovery, whereas GFAP did not.72 The patients with mTBI were clearly from the severe end of mTBI
definition, and many showed CT abnormalities. In a cohort that included all severities of TBI from the same study, both
UCH-L1 (AU-ROC 0.73) and GFAP (AU-ROC 0.72) at admission modestly discriminated patients with unfavourable
outcome from those with favourable outcome. In patients with complete and incomplete recovery, the discriminatory
power of UCH-L1 and GFAP was not clinically useful, 0.54 and 0.63, respectively. In predicting mortality, the AU-ROCs
of UCH-L1 and GFAP were not much better, 0.66 and 0.72, respectively.73 In another study, predictive performance of
UCH-L1 and GFAP within 24 hours from injury for complete recovery at three months in patients with TBIs of all
severities was not adequate, 0.59 and 0.65 respectively. In predicting favourable outcome, the AU-ROCs were somewhat
better, 0.80 and 0.75, respectively. The performance improved when the biomarkers were combined.64
NF-L has shown promise in predicting outcomes. Blood and cerebrospinal fluid levels of NF-L distinguished players with
post concussive symptoms lasting over a year from patients with post concussive symptoms resolving within one year
with AU-ROCs of 0.81 and 0.80, respectively. However, in that study, no associations were observed between NF-L and
functional outcome measured by GOSE at 30 days after TBI. In a study examining NF-L levels in patients with severe
TBI over 12 days of injury, NF-L levels were predictive of patient survival, both at baseline and when comparing
cumulative 12-day levels with patient outcomes a year later.74
S100B has been thoroughly examined as an outcome prediction biomarker. In a study that included patients with moderate
and severe TBI, S100B levels were found to be significant predictors for poor outcome and death, with AU-ROCs of 0.82
and 0.86, respectively.75 Another study found that S100B levels at admission were not correlated with outcome at one
year after injury. However, when 12-day cumulative S100B levels were assessed, they differentiated survivors from
nonsurvivors.74 A third study examined S100B levels in the first three days after TBI. S100B levels measured within 12
hours of injury or 36 hours after injury did not predict outcomes, whereas S100B levels 12–36 hours after injury correlated
with outcomes at 6–12 months after injury. In a follow-up study by the same group, S100B levels at approximately 28
hours after injury were found to correlate with outcomes at three or six months, in contrast to assessments at earlier or
later time points. Concurrent extracranial injury was found to increase S100B levels in the first 12 hours after injury,
decreasing thereafter. This suggests that S100B levels should be determined at a time point between 12–36 hours after
injury in patients with polytrauma to predict outcome.76
11
Tau levels have also been shown to predict outcome following TBI. The P-tau levels and P-tau/T-tau ratio outperformed
the T-tau level in discriminating patients with favourable and unfavourable outcome in patients with moderate to severe
TBI.43 In patients with mTBI, it has been observed that though T-tau levels are correlated with outcome, they are not able
to predict full recovery.42
Biomarker panels have also shown promise in outcome prediction as shown earlier in case of combining GFAP and UCH-
L1.64 In a study including multiple biomarkers, a combination of GFAP, heart fatty acid binding protein (H-FABP), and
NF-L yielded a specificity of 40% when sensitivity was set at 95–100% in predicting complete recovery in patients with
mild TBI. The inclusion of various clinical covariates was found to significantly improve the ability of the biomarkers to
predict outcome. When injury severity score was combined with H-FABP and NF-L, specificity increased to 56% when
sensitivity was maintained at 95–100%. With respect to TBIs of all severities, a panel of GFAP, H-FABP, and interleukin-
10 (IL-10) yielded a specificity of 63%. Inclusion of clinical covariates dramatically improved the results, reaching a
specificity of 80% with a sensitivity at 95–100% when IL-10, age, and TBI severity (admission GCS and duration of
posttraumatic amnesia) were combined. Another study including multiple biomarkers and clinical covariates showed that
combination of GFAP and NF-L provided the best improvement in the predictive performance of the IMPACT model.77
To date, three studies have been published on blood biomarker levels and microstructural injury derived from diffusion
tensor magnetic resonance imaging (DTI). In a study of patients with subacute and chronic TBI, both GFAP and NF-L
were found to be persistently elevated with different kinetic profiles. NF-L decreased constantly during the study period,
whereas GFAP showed a biphasic profile with an initial decrease followed by a secondary increase. It was also observed
that biomarker levels at 30 days after injury were associated with changes in functional outcome at 90 days, and that NF-
L at 30 days was associated with later outcome and grey and white matter loss at 90 days. This correlation was not found
with GFAP.33
Another study examined biomarker trajectories during the first year after injury. NF-L was observed to peak 10 days to
six weeks after injury and to remain elevated at one year, with peak NF-L correlating with the extent of white matter
injury and predicting white matter atrophy between six and 12 months after injury. Both NF-L and GFAP predicted grey
matter atrophy in the first six months after injury.78
In the most recent study on this topic, many patients were shown to have persistent and temporally variable elevation of
GFAP and NF-L up to 13 years after TBI. It was found that GFAP levels were initially normal at about eight months but
tended to increase by over five years later, whereas NF-L levels showed a reversed pattern. The persistent elevation of
GFAP and NFL at six months was significantly related to metrics of microstructural injury on DTI. 19 The results support
that patients with TBI show accelerated brain ageing, as also reported before.79
Clinical issues and confounding factors
Prediction of outcome with blood-based biomarkers is mostly influenced by the same factors as other clinical questions
discussed earlier in this review. Above all, sampling time and extracranial injuries may affect the results. Unlike in
identifying patients with concussion/mTBI or the need for acute imaging, clinical variables play an important
accompanying role in predicting outcome after TBI with biomarkers. As discussed, the outcome may be seen from various
viewpoints, and the contribution of biomarkers in predicting, e.g., permanent symptoms. cognition, or quality of life is
largely unknown.
12
Most prognostic biomarker studies involve a one-time determination of biomarker levels at admission to an emergency
department. Most blood-based biomarker levels––although not all––are initially elevated after and gradually decrease
over time, depending on the kinetics of the particular biomarker. This expected course reflects only a single injury event
and is unable to measure the impact of secondary injuries and neuroworsening caused by them. Therefore, it is important
to determine the temporal pattern of changes in biomarker levels in TBIs of different severities and types to understand
the results of biomarker assessments at different time points. Last, it is worth mentioning that complex models that include
panels of biomarkers and various clinical variables are poorly suited for situations that require rapid clinical decisions. In
addition, many clinical variables are still subjective, which can lead to sources of error in assessment. It is obvious that
emerging dynamic outcome prediction models including objectively and rapidly measurable variables will probably be
the most adequate choice for improving the diagnostic and care in patients with TBI.80
Conclusion
This article presents an overview of the current potential for clinical use of five selected TBI biomarkers. Due to active
research in this field, articles on TBI biomarkers have been published at a rapid pace in recent years, of which
unfortunately only a small portion can be covered in this targeted review article. Considering many far greater clinical
needs, it is noticeable that the identification of the need for an acute head CT imaging with blood-based biomarkers has
attracted the most interest. Developing clinically useful biomarkers for other, more important needs should have also lots
of commercial interest.
There is a need to explore a new generation of TBI biomarkers. For example, mircoRNAs81,82, extracellular vesicle
proteins83, and autoantibodies84 have attracted research interest. However, there are technical limitations and challenges
in using exosomes as blood-based biomarkers, mostly due to problems in isolating pure exosomes and quantification.85
Blood-based metabolomics has also shown promise in the acute diagnostics and outcome prediction of TBI.86,87
The examples given in the above sections illustrate the complex challenges of biomarker research in TBI. Theoretically,
a clinician should know what biomarkers are needed in each case, what their sensitivity and specificity are, what time
window is reasonable, and how time affects the panel needed for sensitivity/specificity, as well as how, for example, age
and other injuries affect the results. Validated biomarker assessment methods and appropriate training will likely reduce
this knowledge gap in the future, but systems medicine approaches and machine learning algorithms will most likely be
needed to aid the clinician. Given the dynamic nature of TBI, particularly in more severe cases, it is highly unlikely that
a biomarker/biomarker panel at a single time point will be sufficient. Despite the many difficulties discussed above, there
is a lot of potential in using biomarkers in future TBI medicine.
13
References
1. Agoston D V., Elsayed M. Serum-based protein biomarkers in blast-induced traumatic brain injury spectrum
disorder. Front Neurol. 2012;JUL. doi:10.3389/fneur.2012.00107
2. Carney N, Totten AM, O’Reilly C, et al. Guidelines for the Management of Severe Traumatic Brain Injury, Fourth
Edition. Neurosurgery. 2017;80(1):6-15. doi:10.1227/NEU.0000000000001432
3. Wang KK, Yang Z, Zhu T, et al. An update on diagnostic and prognostic biomarkers for traumatic brain injury.
Expert Rev Mol Diagn. 2018. doi:10.1080/14737159.2018.1428089
4. Czeiter E, Amrein K, Gravesteijn BY, et al. Blood biomarkers on admission in acute traumatic brain injury:
Relations to severity, CT findings and care path in the CENTER-TBI study. EBioMedicine. 2020;56:1-11.
doi:10.1016/j.ebiom.2020.102785
5. Blennow K, Hardy J, Zetterberg H. The Neuropathology and Neurobiology of Traumatic Brain Injury. Neuron.
2012;76(5):886-899. doi:10.1016/j.neuron.2012.11.021
6. Gardner RC, Rubenstein R, Wang KKW, et al. Age-Related Differences in Diagnostic Accuracy of Plasma Glial
Fibrillary Acidic Protein and Tau for Identifying Acute Intracranial Trauma on Computed Tomography: A
TRACK-TBI Study. J Neurotrauma. 2018;35(20). doi:10.1089/neu.2018.5694
7. Iverson GL, Reddi PJ, Posti JP, et al. Serum neurofilament light is elevated differentially in older adults with
uncomplicated mild traumatic brain injuries. J Neurotrauma. 2019;36(16). doi:10.1089/neu.2018.6341
8. Villaseñor R, Lampe J, Schwaninger M, Collin L. Intracellular transport and regulation of transcytosis across the
blood–brain barrier. Cell Mol Life Sci. 2019;76(6). doi:10.1007/s00018-018-2982-x
9. Plog B, Dashnaw M, Hitomi E, et al. Biomarkers of traumatic injury are transported from brain to blood via the
glymphatic system. J Neurosci. 2015;35(2):518-526. doi:10.1523/JNEUROSCI.3742-14.2015
10. George N, Geller HM. Extracellular matrix and traumatic brain injury. J Neurosci Res. 2018;96(4).
doi:10.1002/jnr.24151
11. Akamine S, Marutani N, Kanayama D, et al. Renal function is associated with blood neurofilament light chain
level in older adults. Sci Rep. 2020;10(1). doi:10.1038/s41598-020-76990-7
12. Mcdonald SJ, Shultz SR, Agoston D V. The Known Unknowns: An Overview of the State of Blood-Based Protein
Biomarkers of Mild Traumatic Brain Injury. J Neurotrauma. 2021;38(19). doi:10.1089/neu.2021.0011
13. Menon DK, Schwab K, Wright DW, Maas AI, Health D and CAWG of the I and II toward CDE for R on TBI
and P. No Title. Vol 91. University of Cambridge, United Kingdom. dkm13@wbic.cam.ac.uk; 2010:1637-1640.
doi:10.1016/j.apmr.2010.05.017
14. Posti JP, Dickens AM, Orešic M, Hyötyläinen T, Tenovuo O. Metabolomics profiling as a diagnostic tool in
14
severe traumatic brain injury. Front Neurol. 2017;8(AUG). doi:10.3389/fneur.2017.00398
15. Carroll LJ, Cassidy JD, Cancelliere C, et al. Systematic Review of the Prognosis After Mild Traumatic Brain
Injury in Adults: Cognitive, Psychiatric, and Mortality Outcomes: Results of the International Collaboration on
Mild Traumatic Brain Injury Prognosis. Arch Phys Med Rehabil. 2014. doi:10.1016/j.apmr.2013.08.300
16. Carroll EL, Outtrim JG, Forsyth F, et al. Mild traumatic brain injury recovery: a growth curve modelling analysis
over 2 years. J Neurol. 2020;267(11). doi:10.1007/s00415-020-09979-x
17. Mac Donald CL, Barber J, Jordan M, et al. Early clinical predictors of 5-year outcome after concussive blast
traumatic brain injury. JAMA Neurol. 2017;74(7). doi:10.1001/jamaneurol.2017.0143
18. Cnossen MC, Van Der Naalt J, Spikman JM, et al. Prediction of Persistent Post-Concussion Symptoms after Mild
Traumatic Brain Injury. J Neurotrauma. 2018;35(22):2691-2698. doi:10.1089/neu.2017.5486
19. Newcombe VFJ, Ashton NJ, Posti JP, et al. Post-acute blood biomarkers and disease progression in traumatic
brain injury. Brain. 2022;In press.
20. Calcagnile O, Anell A, Undén J. The addition of S100B to guidelines for management of mild head injury is
potentially cost saving. BMC Neurol. 2016. doi:10.1186/s12883-016-0723-z
21. Minkkinen M, Iverson GL, Kotilainen A-K, et al. Prospective Validation of the Scandinavian Guidelines for
Initial Management of Minimal, Mild, and Moderate Head Injuries in Adults. J Neurotrauma. 2019.
doi:10.1089/neu.2018.6351
22. Mandrekar JN. Receiver operating characteristic curve in diagnostic test assessment. J Thorac Oncol. 2010;5(9).
doi:10.1097/JTO.0b013e3181ec173d
23. Yoon H, Walters G, Paulsen AR, Scarisbrick IA. Astrocyte heterogeneity across the brain and spinal cord occurs
developmentally, in adulthood and in response to demyelination. Bradl M, ed. PLoS One. 2017;12(7):e0180697.
doi:10.1371/journal.pone.0180697
24. Halford J, Shen S, Itamura K, et al. New astroglial injury-defined biomarkers for neurotrauma assessment. J
Cereb Blood Flow Metab. 2017;37(10):3278-3299. doi:10.1177/0271678X17724681
25. Papa L, Lewis LM, Falk JL, et al. Elevated levels of serum glial fibrillary acidic protein breakdown products in
mild and moderate traumatic brain injury are associated with intracranial lesions and neurosurgical intervention.
Ann Emerg Med. 2012;59(6):471-483.
26. Welch R, Ellis M, Lewis L, et al. Modeling the Kinetics of Serum Glial Fibrillary Acidic Protein, Ubiquitin
Carboxyl-Terminal Hydrolase-L1, and S100B Concentrations in Patients with Traumatic Brain Injury. J
Neurotrauma. 2017. doi:10.1089/neu.2016.4772
27. Thelin EP, Zeiler FA, Ercole A, et al. Serial sampling of serum protein biomarkers for monitoring human
15
traumatic brain injury dynamics: A systematic review. Front Neurol. 2017. doi:10.3389/fneur.2017.00300
28. Janigro D, Mondello S, Posti JP, Unden J. GFAP and S100B: what you always wanted to know and never dared
to ask. Front Neurol. 2022;In press.
29. Khalil M, Teunissen CE, Otto M, et al. Neurofilaments as biomarkers in neurological disorders. Nat Rev Neurol.
2018;14(10). doi:10.1038/s41582-018-0058-z
30. Gao W, Zhang Z, Lv X, et al. Neurofilament light chain level in traumatic brain injury: A system review and
meta-analysis. Medicine (Baltimore). 2020;99(38). doi:10.1097/MD.0000000000022363
31. Sandelius Å, Zetterberg H, Blennow K, et al. Plasma neurofilament light chain concentration in the inherited
peripheral neuropathies. Neurology. 2018;90(6). doi:10.1212/WNL.0000000000004932
32. Posti JPJP, Takala RSKRSK, Lagerstedt L, et al. Correlation of blood biomarkers and biomarker panels with
traumatic findings on computed tomography after traumatic brain injury. J Neurotrauma. 2019;36(14).
doi:10.1089/neu.2018.6254
33. Shahim P, Politis A, van der Merwe A, et al. Time course and diagnostic utility of NfL, tau, GFAp, and UCH-L1
in subacute and chronic TBI. Neurology. 2020;95(6):e623-e636. doi:10.1212/WNL.0000000000009985 LK -
http://WT3CF4ET2L.search.serialssolutions.com?sid=EMBASE&issn=1526632X&id=doi:10.1212%2FWNL.0
000000000009985&atitle=Time+course+and+diagnostic+utility+of+NfL%2C+tau%2C+GFAp%2C+and+UCH
-
L1+in+subacute+and+chronic+TBI&stitle=Neurology&title=Neurology&volume=&issue=&spage=&epage=&
aulast=Shahim&aufirst=Pashtun&auinit=P.&aufull=Shahim+P.&coden=&isbn=&pages=-
&date=2020&auinit1=P&auinitm=
34. Marenholz I, Heizmann CW, Fritz G. S100 proteins in mouse and man: From evolution to function and pathology
(including an update of the nomenclature). Biochem Biophys Res Commun. 2004;322(4).
doi:10.1016/j.bbrc.2004.07.096
35. Rodríguez-Rodríguez A, Egea-Guerrero JJ, León-Justel A, et al. Role of S100B protein in urine and serum as an
early predictor of mortality after severe traumatic brain injury in adults. Clin Chim Acta. 2012;414.
doi:10.1016/j.cca.2012.09.025
36. Thelin EP, Nelson DW, Bellander BM. A review of the clinical utility of serum S100B protein levels in the
assessment of traumatic brain injury. Acta Neurochir (Wien). 2017. doi:10.1007/s00701-016-3046-3
37. Mehta T, Fayyaz M, Giler GE, et al. Current Trends in Biomarkers for Traumatic Brain Injury. Open access J
Neurol Neurosurg. 2020;12(4).
38. Calcagnile O, Holmén A, Chew M, Undén J. S100B levels are affected by older age but not by alcohol
intoxication following mild traumatic brain injury. Scand J Trauma Resusc Emerg Med. 2013;21(1).
doi:10.1186/1757-7241-21-52
16
39. Janigro D, Kawata K, Silverman E, Marchi N, Diaz-Arrastia R. Is Salivary S100B a Biomarker of Traumatic
Brain Injury? A Pilot Study. Front Neurol. 2020;11. doi:10.3389/fneur.2020.00528
40. Hanes J, Zilka N, Bartkova M, Caletkova M, Dobrota D, Novak M. Rat tau proteome consists of six tau isoforms:
Implication for animal models of human tauopathies. J Neurochem. 2009;108(5). doi:10.1111/j.1471-
4159.2009.05869.x
41. Khan SS, Bloom GS. Tau: The Center of a Signaling Nexus in Alzheimer’s Disease. Front Neurosci.
2016;10(FEB). doi:10.3389/fnins.2016.00031
42. Hossain I, Mohammadian M, Takala RSK, et al. Admission Levels of Total Tau and β-Amyloid Isoforms 1–40
and 1–42 in Predicting the Outcome of Mild Traumatic Brain Injury. Front Neurol. 2020.
doi:10.3389/fneur.2020.00325
43. Rubenstein R, Chang B, Yue JK, et al. Comparing plasma phospho tau, total tau, and phospho tau–total tau ratio
as acute and chronic traumatic brain injury biomarkers. Investigators and the T-T, ed. JAMA Neurol.
2017;74(9):1063-1072. doi:10.1001/jamaneurol.2017.0655
44. Shahim P, Tegner Y, Wilson DH, et al. Blood biomarkers for brain injury in concussed professional ice hockey
players. JAMA Neurol. 2014;71(6):684. doi:10.1001/jamaneurol.2014.367
45. Randall J, Mörtberg E, Provuncher GK, et al. Tau proteins in serum predict neurological outcome after hypoxic
brain injury from cardiac arrest: Results of a pilot study. Resuscitation. 2013;84(3).
doi:10.1016/j.resuscitation.2012.07.027
46. Chiu MJ, Fan LY, Chen TF, Chen YF, Chieh JJ, Horng HE. Plasma tau levels in cognitively normal middle-aged
and older adults. Front Aging Neurosci. 2017;9(MAR). doi:10.3389/fnagi.2017.00051
47. Tongaonkar P, Chen L, Lambertson D, Ko B, Madura K. Evidence for an interaction between ubiquitin-
conjugating enzymes and the 26S proteasome. Mol Cell Biol. 2000;20(13):4691-4698.
48. Papa L, Brophy GM, Welch RD, et al. Time course and diagnostic accuracy of glial and neuronal blood
biomarkers GFAP and UCH-L1 in a large cohort of trauma patients with and without mild traumatic brain injury.
JAMA Neurol. 2016;73(5):551-560. doi:10.1001/jamaneurol.2016.0039
49. Brophy GM, Mondello S, Papa L, et al. Biokinetic analysis of ubiquitin C-terminal hydrolase-L1 (UCH-L1) in
severe traumatic brain injury patient biofluids. J Neurotrauma. 2011;28(6):861-870. doi:10.1089/neu.2010.1564
50. Wilkinson KD, Lee KM, Deshpande S, et al. No Title. Science (80- ). 1989;246(4930):670-673.
doi:10.1126/science.2530630
51. Meyer Schwesinger C, Meyer TN, Münster S, et al. No Title. 2009;217(3):452-464. doi:10.1002/path.2446
52. McCrea M, Broglio SP, McAllister TW, et al. Association of Blood Biomarkers With Acute Sport-Related
17
Concussion in Collegiate Athletes: Findings From the NCAA and Department of Defense CARE Consortium.
JAMA Netw open. 2020;3(1):e1919771. doi:10.1001/jamanetworkopen.2019.19771
53. Meier TB, Huber DL, Bohorquez-Montoya L, et al. A Prospective Study of Acute Blood-Based Biomarkers for
Sport-Related Concussion. Ann Neurol. 2020;87(6). doi:10.1002/ana.25725
54. Posti JP, Hossain I, Takala RSK, et al. Glial Fibrillary Acidic Protein and Ubiquitin C-Terminal Hydrolase-L1
Are Not Specific Biomarkers for Mild CT-Negative Traumatic Brain Injury. J Neurotrauma. 2017.
doi:10.1089/neu.2016.4442
55. Koivikko P, Posti JP, Mohammadian M, et al. Potential of heart fatty-acid binding protein, neurofilament light,
interleukin-10 and S100 calcium-binding protein B in the acute diagnostics and severity assessment of traumatic
brain injury. Emerg Med J. 2021:emermed-2020-209471. doi:10.1136/emermed-2020-209471
56. Giza CC, McCrea M, Huber D, et al. Assessment of Blood Biomarker Profile After Acute Concussion During
Combative Training Among US Military Cadets: A Prospective Study From the NCAA and US Department of
Defense CARE Consortium. JAMA Netw open. 2021;4(2). doi:10.1001/jamanetworkopen.2020.37731
57. Pattinson CL, Meier TB, Guedes VA, et al. Plasma Biomarker Concentrations Associated with Return to Sport
following Sport-Related Concussion in Collegiate Athletes-A Concussion Assessment, Research, and Education
(CARE) Consortium Study. JAMA Netw Open. 2020;3(8). doi:10.1001/jamanetworkopen.2020.13191
58. Mattila OS, Ashton NJ, Blennow K, et al. Ultra-Early Differential Diagnosis of Acute Cerebral Ischemia and
Hemorrhagic Stroke by Measuring the Prehospital Release Rate of GFAP. Clin Chem. 2021;67(10).
doi:10.1093/clinchem/hvab128
59. Unden L, Calcagnile O, Unden J, et al. Validation of the Scandinavian guidelines for initial management of
minimal, mild and moderate traumatic brain injury in adults TL - 13 (1) (no pagination). Valid Scand Guidel
Initial Manag minimal, mild moderate Trauma brain Inj adults. 2015;13 (1) (no(292). doi:10.1186/s12916-015-
0533-y
60. Bazarian JJ, Biberthaler P, Welch RD, et al. Serum GFAP and UCH-L1 for prediction of absence of intracranial
injuries on head CT (ALERT-TBI): a multicentre observational study. Lancet Neurol. 2018;17(9):782-789.
doi:10.1016/S1474-4422(18)30231-X
61. Maas AIR, Lingsma HF. ALERT-TBI study on biomarkers for TBI: has science suffered? Lancet Neurol.
2018;17(9). doi:10.1016/S1474-4422(18)30275-8
62. Posti JP, Takala RSK, Lagerstedt L, et al. Correlation of blood biomarkers and biomarker panels with traumatic
findings on computed tomography after traumatic brain injury. J Neurotrauma. 2019. doi:10.1089/neu.2018.6254
63. Posti JP, Takala RSK, Runtti H, et al. The Levels of Glial Fibrillary Acidic Protein and Ubiquitin C-Terminal
Hydrolase-L1 during the First Week after a Traumatic Brain Injury: Correlations with Clinical and Imaging
Findings. Neurosurgery. 2016;79(3). doi:10.1227/NEU.0000000000001226
18
64. Diaz-Arrastia R, Wang KKW, Papa L, et al. Acute biomarkers of traumatic brain injury: Relationship between
plasma levels of ubiquitin C-terminal hydrolase-l1 and glial fibrillary acidic protein. J Neurotrauma. 2014;31(1).
doi:10.1089/neu.2013.3040
65. Okonkwo DO, Puffer RC, Puccio AM, et al. Point-of-Care Platform Blood Biomarker Testing of Glial Fibrillary
Acidic Protein versus S100 Calcium-Binding Protein B for Prediction of Traumatic Brain Injuries: A
Transforming Research and Clinical Knowledge in Traumatic Brain Injury Study. J Neurotrauma. 2020;37(23).
doi:10.1089/neu.2020.7140
66. Huebschmann NA, Luoto TM, Karr JE, et al. Comparing Glial Fibrillary Acidic Protein (GFAP) in Serum and
Plasma Following Mild Traumatic Brain Injury in Older Adults. Front Neurol. 2020;11.
doi:10.3389/fneur.2020.01054
67. Maas AIR, Lingsma HF, Roozenbeek B. Predicting outcome after traumatic brain injury. In: Handbook of
Clinical Neurology. Vol 128. Elsevier B.V.; 2015:455-474. doi:10.1016/B978-0-444-63521-1.00029-7
68. Steyerberg EW, Mushkudiani N, Perel P, et al. Predicting outcome after traumatic brain injury: Development and
international validation of prognostic scores based on admission characteristics. PLoS Med. 2008;5(8).
doi:10.1371/journal.pmed.0050165
69. Collaborators MRCCT, Perel PA, Arango M, et al. Predicting outcome after traumatic brain injury: Practical
prognostic models based on large cohort of international patients. Bmj. 2008;336(7641):425-429.
doi:10.1136/bmj.39461.643438.25
70. Gaetani L, Paolini Paoletti F, Bellomo G, et al. CSF and Blood Biomarkers in Neuroinflammatory and
Neurodegenerative Diseases: Implications for Treatment. Trends Pharmacol Sci. 2020;41(12).
doi:10.1016/j.tips.2020.09.011
71. Shi K, Zhang J, Dong J fei, Shi FD. Dissemination of brain inflammation in traumatic brain injury. Cell Mol
Immunol. 2019;16(6). doi:10.1038/s41423-019-0213-5
72. Hossain I, Mohammadian M, Takala RSKRSK, et al. Early Levels of Glial Fibrillary Acidic Protein and
Neurofilament Light Protein in Predicting the Outcome of Mild Traumatic Brain Injury. J Neurotrauma.
2019;36(10). doi:10.1089/neu.2018.5952
73. Takala RSK, Posti JP, Runtti H, et al. Glial Fibrillary Acidic Protein and Ubiquitin C-Terminal Hydrolase-L1 as
Outcome Predictors in Traumatic Brain Injury. World Neurosurg. 2016. doi:10.1016/j.wneu.2015.10.066
74. Shahim P, Gren M, Liman V, et al. Serum neurofilament light protein predicts clinical outcome in traumatic brain
injury. Sci Rep. 2016;6. doi:10.1038/srep36791
75. Di Battista AP, Buonora JE, Rhind SG, et al. Blood biomarkers in moderate-to-severe traumatic brain injury:
Potential utility of a multi-marker approach in characterizing outcome. Front Neurol. 2015;6(MAY).
doi:10.3389/fneur.2015.00110
19
76. Thelin EP, Jeppsson E, Frostell A, et al. Utility of neuron-specific enolase in traumatic brain injury; relations to
S100B levels, outcome, and extracranial injury severity. Crit Care. 2016;20:285. doi:10.1186/s13054-016-1450-
y
77. Thelin E, Al Nimer F, Frostell A, et al. A Serum protein biomarker panel improves outcome prediction in human
traumatic brain injury. J Neurotrauma. 2019. doi:10.1089/neu.2019.6375
78. Graham NSN, Zimmerman KA, Moro F, et al. Axonal marker neurofilament light predicts long-term outcomes
and progressive neurodegeneration after traumatic brain injury. Sci Transl Med. 2021;13(613).
doi:10.1126/scitranslmed.abg9922
79. Cole JH, Leech R, Sharp DJ. Prediction of brain age suggests accelerated atrophy after traumatic brain injury.
Ann Neurol. 2015;77(4). doi:10.1002/ana.24367
80. Raj R, Luostarinen T, Pursiainen E, et al. Machine learning-based dynamic mortality prediction after traumatic
brain injury. Sci Rep. 2019;9(1). doi:10.1038/s41598-019-53889-6
81. Di Pietro V, Ragusa M, Davies D, et al. MicroRNAs as Novel Biomarkers for the Diagnosis and Prognosis of
Mild and Severe Traumatic Brain Injury. J Neurotrauma. 2017;34(11). doi:10.1089/neu.2016.4857
82. Papa L, Slobounov SM, Breiter HC, et al. Elevations in MicroRNA Biomarkers in Serum Are Associated with
Measures of Concussion, Neurocognitive Function, and Subconcussive Trauma over a Single National Collegiate
Athletic Association Division i Season in Collegiate Football Players. J Neurotrauma. 2019;36(8).
doi:10.1089/neu.2018.6072
83. Karnati HK, Garcia JH, Tweedie D, Becker RE, Kapogiannis D, Greig NH. Neuronal enriched extracellular
vesicle proteins as biomarkers for traumatic brain injury. J Neurotrauma. 2019;36(7). doi:10.1089/neu.2018.5898
84. Wang KKW, Yang Z, Yue JK, et al. Plasma Anti-Glial Fibrillary Acidic Protein Autoantibody Levels during the
Acute and Chronic Phases of Traumatic Brain Injury: A Transforming Research and Clinical Knowledge in
Traumatic Brain Injury Pilot Study. J Neurotrauma. 2016;33(13). doi:10.1089/neu.2015.3881
85. Guedes VA, Devoto C, Leete J, et al. Extracellular Vesicle Proteins and MicroRNAs as Biomarkers for Traumatic
Brain Injury. Front Neurol. 2020;11. doi:10.3389/fneur.2020.00663
86. Orešič M, Posti JP, Kamstrup-Nielsen MH, et al. Human Serum Metabolites Associate With Severity and Patient
Outcomes in Traumatic Brain Injury. EBioMedicine. 2016;12. doi:10.1016/j.ebiom.2016.07.015
87. Dickens AM, Posti JP, Takala RSK, et al. Serum Metabolites Associated with Computed Tomography Findings
after Traumatic Brain Injury. J Neurotrauma. 2018. doi:10.1089/neu.2017.5272
20
Tables
Table 1. Key properties of the most studied traumatic brain injury biomarkers
Table 2. Clinical questions related to traumatic brain injury biomarkers
Question
Assessment
setting
Most promising biomarkers
included in this review
Confounders
Acute identification of mild
TBI
paramedics, ED
NF-L, S100B, UCH-L1
Assessment time, extracranial injuries,
age
Assessment of acute severity
paramedics, ED
GFAP, S100B, UCH-L1
Assessment time, extracranial injuries,
age, dynamic nature of TBI
Need for head acute CT
paramedics, ED
GFAP, S100B, UCH-L1
Assessment time, age, extracranial
injuries, epidural hematoma, S100B
need to be used as a part of a guideline
Outcome prediction
ED, ward, ICU
GFAP, NF-L, S100B
Assessment time, age, extracranial
injuries, outliers with abnormally
low/high biomarker levels
Late identification of TBI
outpatient clinic
GFAP, NF-L
Neurodegenerative diseases, outliers
with abnormally low/high biomarker
levels
Biomarker
Mass (kDa)
Source
Location
Other sources
Half-life (h)
Peak (h)
Confounders
GFAP
50
astrocytes
cytoplasm
Schwann cells,
chondrocytes,
enteric glial cells
liver, pancreas
24–48
20–24
Levels
increase with
age,
extracerebral
expression
NF-L
68
neurons
myelinated
axons
Peripheral axons
unknown
unknown
Levels
increase with
age,
peripheral
nerve injury
S100B
11
astrocytes
cytoplasm
adipocytes,
muscle,
chondrocytes,
enteric glial cells
0.5–2
<6
Levels
increase with
age, short
half-life,
extracerebral
expression
Tau
proteins
33–46
neurons
axon
terminals,
unmyelinated
axons
astrocytes and
oligodendrocytes,
peripheral nervous
system, kidneys
unknown
12–24
Levels
increase with
age,
extracerebral
expression
UCH-L1
25
neurons
cytoplasm
testis, ovary,
kidney
8
7–9
Short half-
life,
extracerebral
expression
