Introduction
Traumatic brain injury (TBI) is a slow epidemic in India as well as in the world.
About 25% of global deaths due to TBI occur in India. It is one of the major causes
of morbidity and mortality in India. High-quality research is lacking and is the need
of the hour. The increased incidence of TBI in a productive population is of serious
concern.[1] The Glasgow coma scale (GCS) has been used to assess severity of TBI. This is purely
clinical and does not give any information about the underlying pathology. Computed
tomographic (CT) scan is used to diagnose TBI for immediate evaluation but does not
correlate with long-term outcomes particularly in mild head injury.[2] Biomarkers are used for diagnosis, risk stratification, and predicting outcomes
in TBI patients.[3] The TBI results in neurologic, neuropsychological, or behavioral changes caused
by mechanical trauma.[4] The entire spectrum of sequelae of TBI cannot be predicted by GCS and CT scan alone;
hence, there is need of addition of serum biomarkers for the evaluation of TBI.[5]
[6]
[7]
[8]
[9]
Biomarkers in TBI
Improvement in scientific advances and knowledge of TBI biomarkers has improved our
understanding of complicated pathological processes related to TBI. Various drugs
in animal experiments are neuroprotective but similar results in humans are not satisfactory.[10]
[11]
[12]
[13] Failure in clinical trials is due to a lack of temporal measurement of biomarkers
and heterogeneous pathophysiological processes involving TBI. It is a combination
of an acute and chronic event resulting in a progressive delayed degenerative process
involving reactions at the cellular level. As the axons are lengthy, they are particularly
susceptible to direct trauma to the brain.[14]
[15] Hence, a thorough understanding of these processes in greater detail is important
for further research and knowledge for therapeutic intervention.
TBI Biomarker Requirement
For a serum biomarker to be useful, it should have the following characteristics:
-
It should be easily measured in serum /plasma or cerebrospinal fluid (CSF),
-
Their levels must change in TBI patients after an injury as compared with the normal
population,
-
It should have some baseline level in the control population,
-
It should be derived from the brain as the main source,
-
Their levels can be easily quantified by using sandwich enzyme-linked immunosorbent
assay or similar assays.
-
Their levels should be able to classify the severity of TBI, should correlate with
CT and GCS findings.[16]
There are currently two hypotheses as to how the brain proteins reach the peripheral
circulation: first is damage to blood–brain barrier (BBB), and second is following
the bulk movement of fluids through central nervous system (CNS) that may clear the
proteins from the central to a peripheral system called as glymphatic system.[17] These biomarkers can be classified as below:
-
Inflammatory markers like interleukin-1β (IL-1β), IL-10, and tumor necrosis factor-α
(TNF-α).
-
Markers of astrocyte activation: glial fibrillary acidic protein (GFAP), S100β.
-
Markers of neuronal injury: myelin basic protein (MBP), neuron-specific enolase (NSE),
ubiquitin carboxy-terminal hydrolase-l (UCHL-1).
IL-1 is an important mediator of inflammation in the CNS as well in the peripheral
nervous system. These molecules of the IL1 family are the widely studied cytokines
about TBI in different models of focal and diffuse injury.[18]
[19]
[20]
[21] The most commonly studied isoform in TBI is the IL-1β. The mature human IL-1β is
17.5 kDa.
IL-1β is the main endogenous mediator of multiple hosts. Response to injuries like
fever, alterations of neuroendocrine, immune, and cardiovascular systems affect IL-1
levels. IL-1β induces various responses such as alteration of BBB, migration of inflammatory
cells, the release of membrane arachidonic acid metabolites, free radicle generation,
and complement-mediated damage.[22]
Clinical Evidence
IL-1β is normally absent in the blood–cerebrospinal fluid of healthy individuals.
Its detection in patients with TBI has been difficult.[23]
[24]
[25] The studies that correlated the levels of IL-1β with the outcome have found that
serum levels of IL-1β taken within 6 hours of TBI correlated well with TBI severity
in a cohort of 48 patients.[26] In recent studies of severe TBI, patient's increase in CSF levels of IL-1β has been
linked with worse outcomes.[27]
[28]
TNF-α is another prime chemokine involved in initiating and upregulating the inflammatory
response and other cytokine production. It is a 17-kDa active cytokine derived from
a 26 codon precursor molecule after being cleaved by the converting enzyme.[29] Normally this cytokine is not expressed in a healthy brain. Because of this fact,
its role in physiological conditions is not understood. But in a state of inflammation
or disease, it is produced in abundance along with other inflammatory cytokines by
activated microglia.[30]
Clinical Evidence
It has been proven that the serum and CSF levels of patients with TBI at 24h intervals
are significantly elevated as compared with controls. It was found in patients that
died within 17 minutes of trauma, they had increased TNF-α mRNA and proteins.[33] Hayakata et al have investigated CSF of 23 patients with GCS <8, and they have noted
a peak of 20 to 30pg/mL within 24 hours. No significant correlation is seen with TNF-α
levels and intracranial pressure (ICP) or Glasgow outcome scale (GOS) outcomes after
6 months.[27] In a recent study by Stein et.al, the blood and CSF samples of 24 patients with
severe grades of TBI at 12 hours intervals for 7 days have been investigated. They
have also monitored the ICP and cerebral perfusion pressure (CPP) in the same patients.
Their study has shown a correlation with serum levels of TNF-α and subsequent change
in ICP or CPP, but not associated with any prediction for the outcome.[34]
It is a chemokine with a weighing around 40 kDa. IL-10 is synthesized in the brain
by the microglia and astrocytes. In the periphery, it is mainly generated by the lymphopoietic
cells.[35]
[36]
[37] It is a main anti-inflammatory cytokine and considered to be neuroprotective.
Clinical Evidence
Csuka et al have measured IL-10 in serum and CSF of 28 patients with GCS <8. They
have found increased levels in CSF and serum. In CSF, it ranged from 1.3 to 41.7pg/mL
and in serum, the levels ranged from 5.4 to 23pg/mL. The temporal variation has been
such that there was an early rise followed by a slow decline. This is noted both in
CSF and serum. The BBB function was also assessed using the CSF/serum albumin ratio
and then correlated with the IL-10 levels. But they did not find a significant correlation.[39] When considering outcome after 6 months, it is reported that the IL-10 levels have
been higher in patients with poor outcome (GOS < 4). The IL-10 levels alterations
in polytrauma patients could be the reason for the poor association between IL-10
and grades of TBI.[28]
[40]
Though there are contrary reports on IL-10 in TBI, it is established that it reduces
neuroinflammation centrally and causes peripheral immune-suppression. This plays a
role especially in secondary brain injury by increasing the chances of infection.[24]
S-100β is the most commonly studied biomarker in TBI. S-100 family has a low molecular
weight of 10.5 kDa; it regulates intracellular calcium levels. It is synthesized mainly
by astrocytes microglia, oligodendrocytes, and neurons.[41]
[42] Their effect seems to be dependent on their levels in serum, regardless of its spatial
distribution. It is toxic at higher concentrations and protective at minimal concentrations
in healthy subjects. S-100β is nonspecific to the brain, as it is secreted in extra-CNS
areas such as adipose tissue, cartilage cells, cardiac cells, and pulmonary alveolar
cells.[43] Variations in levels of S-100β have been seen in conditions such as heart failure,
musculoskeletal damage, fractures, and obesity.[44] Low serum levels of 0.05 ng/mL are noted in healthy humans; also levels are not
affected by gender or age of patients.[41] Increased S-100β is expressed by astrocytes and considered as a marker for impaired
BBB, and an increased level of S-100β corresponds to BBB damage.[45]
In a recent study, it was shown that increased CSF/ serum albumin suggests disruption
of BBB; also increased concentrations of S-100β predicted the severity of TBI subjects.[46] Another author studied 14 TBI patients posttrauma with magnetic resonance imaging
and single photon emission computed tomography for BBB disruption; data showed well
correlation with the severity of BBB damage; however, it did not show a positive correlation
of serum S-100β levels.[47]
S-100β levels of nearly 5ng/mL are seen after TBI. Higher levels are useful in predicting
outcomes after 3 to 6 months of injury. Values between 2.0 and 2.5ng/mL were considered
predictive for worse outcomes.[48]
In a recent study, it was observed that S-100β did not increase after patients were
subjected to exercise, which may suggest that serum levels may be more specific to
mild TBI (mTBI)/concussion. In another study, serum levels of S-100β with effects
of soccer ball heading at a particular speed were assessed. It was seen that levels
of S-100β showed no differences in irrespective of the speed of the soccer ball.[49] One of the studies measured S-100β levels in three groups, only head injury, head
injury with other trauma, and those with trauma other than a head injury, it was noted
that blood levels were more in subjects with polytrauma with a head injury, which
suggests extra CNS source of S-100β.[50] Some studies did not find an association between blood levels of S-100 β and the
amount of adipose tissue. Also, S-100β levels do not change in various diseases such
as epilepsy, bladder, lung liver, and renal cancers, which further support that the
levels may correlate with brain injury. The half-life of S-100β is 97minutes; hence,
temporal measurement may be used to identify various grades of TBI and predict outcomes.[51]
[52]
Astrocytes are the major source of GFAP that is one of the important structural proteins
of cytoskeletal-intermediate filaments. There are 10 GFAP isoforms identified till
now. GFAP-α is produced by astrocytes, while GFAP-β is produced by Schwann cells in
peripheral nerves; GFAP-γ is seen in the reticuloendothelial system such as spleen
and bone marrow.[53] GFAP maintains the integrity of the astrocytes cytoskeleton in response to injury.[54] After trauma and astrocyte activation, there is increased migration of GFAP to extracellular
space and levels correlate well with TBI severity.[55]
[56] In normal healthy subjects, levels of GFAP are below the detection limit of 0.012ng/mL,
while certain studies have not shown any detectable levels.[57] Interestingly, in one of the study, it was analyzed that GFAP levels are increased
more in diffuse axonal injury than local injury.[58] These findings were similar to another study which showed that patient with focal
injury had higher levels when compared with diffuse injury, but when diffuse axial
injury (DAI) was further divided in to mild, moderate, and severe, GFAP levels were
higher in moderate and severe types of injury, thus limiting its usefulness in distinguishing
the type of injuries.[59]
Recent studies have shown that GFAP levels are increased in mTBI patients with abnormal
radiological findings compared with those with normal scans.[60]
Enolase is an important enzyme in the glycolytic pathway for ATP production.[61] This enzyme is a protein in the cytoplasm and is expressed depending upon the energy
requirement in a neuron. The levels may be increased in injured axons for homeostasis.
In postmortem examination, NSE is specific for DAI patients with injured axons in
the corpus callosum and is not seen in control groups.[62]
NSE has mainly produced by neurons in normal patients and baseline levels are present
in red blood cells (10ng/mL).[63] Raised levels are seen in stroke, cerebral hemorrhage, and TBI.[64] Many studies have shown that the level of NSE correlates well with moderate and
severe TBI and its outcome after 6 months.[65]
[66] In one of the pediatric TBI studies, it was shown that serum NSE levels were able
to predict poor outcome. Also, the levels in moderate and severe TBI correlate well
with a neurological examination.[67]
In a recent study involving ice hockey players with GCS of 14–15, there was no significant
difference in blood levels of NSE in patients as compared with preinjury status; thus,
it can be concluded that NSE assays may be not specific in detecting altered levels
of NSE after mild injury. Raised NSE levels have been reported in abdominal injuries,
migraine, and femur fracture patients, making it less specific for TBI.[58] In one of the studies with boxers, serum levels of NSE were in a higher range following
2 months of rest as compared with the nonboxers group that suggests impaired neuronal
recovery. Persistent levels of raised NSE even after its half-life of 24 to 48 hours
could suggest repeat TBI.[68]
MBP is one of the most abundant proteins in the CNS and is produced by oligodendrocytes.
In one study involving the pediatric population, the levels of MBP did not differ
significantly when compared with mTBI patients with, but the peak MBP levels differed
significantly in both groups. MBP in the blood is absent in the initial 48 to 72 hours
of trauma making it unreliable as a screening tool. Once elevated the levels of MBP
persisted beyond 14 days and helped to predict future cerebral bleed after injury.
Increased MBP levels following mild injury are promising and can be used to screen
pediatric populations for mTBI, who are unable to tell symptoms of TBI-related events.[69] Many studies have shown that MBP is released in CSF spaces and then into peripheral
circulation following acute neuronal damage in stroke and multiple sclerosis. In one
of the studies, it was seen that serum MBP levels correlated well with severe TBI
patients.[70] In one of the studies it was shown that in postmortem examination of the brain of
blunt head injury patients, MBP was detected in 17 out of 22 patients.[71]
UCHL-1 is a 24 kDa enzyme with protease activity and constitutes 10% of neuronal proteins
and is used as a histological marker for neurons. UCHL-1 could be detected in the
blood and CSF with a t1/2 of 7 to 9 hours after severe TBI.[72] It has an special role in the ATP-dependent proteasome pathway for the elimination
and the ubiquitination of proteins destined for this pathway and removes the oxidized
and misfolded proteins. The UCHL-1 can be used for the detection of neuronal injury.
But it is not CNS specific as it is produced in extra CNS sites such as endocrine
cells, endothelial cells, aortic endothelium, muscle, and tumors cells. Despite its
presence in other tissues, UCHL-1 is highly expressed in CSF and serum. Due to the
abundance of this biomarker in neuronal tissue and CSF, it was used as a histological
marker to discriminate patients with TBI from patients without traumatic injury to
the head. It is reported that patients, who suffer from head injuries with a consequent
intracranial lesion, had higher levels within the first 4 hours, and the levels are
higher in patients who required surgical management. It demonstrated that UCHL-1 has
an association with injury severity and in-hospital prognosis of mortality and clinical
outcome. It is a good biomarker for diagnosing TBI and intracranial lesions, and it
can differentiate injured TBI from noninjured TBI patients when GCS is altered by
any substance due to unclear cause. TBI patients had significantly elevated serum
and CSF levels of UCHL-1 after injury compared with control patients after injury.
As expected, CSF values of UCHL-1 were substantially higher and more sustained than
levels of UCHL-1 in serum. The mechanism by which UCHL-1 is transported from the brain
compartment into the circulation is unknown. Mondella et al evaluated the exposure
and biokinetic parameters of UCHL-1 in CSF and serum. They found a statistically significant
increase in the median amount and peak concentration of UCHL-1 in serum, and a shorter
time to peak concentration in survivors compared with nonsurvivors.[73]
Recently, Puvenna et al researched subconcussive head injury in 15 American football
players. Serum samples were collected before and after every two different games.
No significant differences were observed between the levels of UCHL-1 between controls
and positive individuals for mild injury within 6 hours regardless of CT brain findings.[45]
Also, there was no correlation between the levels of UCHL-1 and the number of impacts
received. After each game, the levels of S 100β and UCHL-1, markers of BBB disruption,
and neuronal injury, respectively, both were elevated. Only S-100β, unlike UCHL-1,
was correlated with the number of hits received and the UCHL-1 elevation did not correlate
with the increase in S-100β levels. Hence, it was suggested that elevated levels of
UCHL-1 may be due to the release of this protein from the neuromuscular junction .
Panel of Biomarkers
Various novel brain proteins have been identified that potentially identify complicated
mTBI. UCHL-1 and GFAP have emerged as promising biomarkers for use in clinical practice.[74] Initial evidence suggests that both proteins are predictors of CT-scan positive
patients, but were limited by retrospective study, the small size of the cohort, and
variability in the timing of serum sampling. All of these probably biased the estimate
of diagnostic accuracy.
In a recent ALERT-TBI trial, validation of a biomarker test combining UCHL-1 and GFAP
to predict CT-positive patients within 12 hours of TBI was studied. Results showed
the high sensitivity and negative predictive value of the UCHL-1 and GFAP test. This
supports its potential clinical role for ruling out the need for a CT scan among patients
with TBI presenting at emergency departments in whom a head CT is felt to be clinically
indicated.[75] In the United States, the U.S. Food and Drug Administration (FDA) has approved the
Banyan BTI (Brain Trauma Indicator) to predict CT scan abnormalities after mTBI. Blood
is sampled within 12 hours of head injury. Test sensitivity is 97.5% and specificity
is 36.5% on the FDA application.[76] The FDA has recently approved a handheld testing platform for GFAP and UCHL-1 levels
with results available within 15 minutes.[77]
[Table 1] lists sensitivity, specificity, and timing of common biomarkers.[78]
[79] This table shows the sensitivity and specificity of biomarkers to detect an abnormal
CT scan after mTBI.
Table 1
Sensitivity, specificity, and timing of common biomarkers[78]
[79]
|
Biomarker
|
Sensitivity
|
Specificity
|
AUROC
|
Method for estimation
|
Optimal time of testing after TBI
|
|
S100B
|
100%
|
35%
|
0.55–0.78
|
ELISA
|
1–3 hours
|
|
UCHL-1[a]
|
97.5%
|
36.5%
|
0.52–0.77
|
ELISA
|
2–8 hours
|
|
Tau
|
92%
|
100%
|
0.5–0.74
|
Single molecule array (Simoa) assay
|
2–8 hours
|
|
GFAP[a]
|
97.5%
|
36.5%
|
0.65–0.94
|
ELISA
|
6–18 hours
|
Abbreviations: AUROC, area under receiver operating curve; ELISA, enzyme-linked immunosorbent
assay; GFAP, glial fibrillary acidic protein; TBI, traumatic brain injury; UCHL-1,
ubiquitin carboxy hydrolase-1.
a Combined sensitivity and specificity for UCHL-1 and GFAP.
Clinical Utility of Biomarkers TBI
A point-of-care test could be used on the field to help detect a concussion. It could
also be used to determine the severity of concussion and be used to screen patients
for neuroimaging (computed tomography and/or magnetic resonance imaging) and further
neuropsychological testing. Biomarkers could have a role in monitoring recovery and
in managing patients with prolonged postconcussion syndrome, a potential of being
incorporated into guidelines for return to work. Currently available biomarkers reflect
injury severity, and serum GFAP, measured within 24 hours after injury, outperforms
clinical characteristics in predicting CT abnormalities.
The current clinical utility of the biomarkers lies in detection of intracranial injury
defined as abnormal CT scan following mTBI. Most patients with a suspected head injury
are examined using GCS, followed by a CT scan of the head to detect traumatic intracranial
lesions, that may require treatment; however, a majority of patients evaluated for
mTBI/concussion do not have detectable intracranial lesions after having a CT scan.
Availability of a blood test for concussion will help health care professionals determine
the need for a CT scan in patients suspected of having mTBI and help prevent unnecessary
neuroimaging or prompt an urgent neuroimaging if the blood test report is abnormal.[75] Among all the biomarkers UCHL-1 and GFAP have been tested and have been approved.
In the ALERT_TBI study, UCHL-1 and GFAP were measured in serum and analyzed using
prespecified cutoff values of 327 and 22 pg/mL, respectively. UCHL-1 and GFAP assay
results were combined into a single test result that was compared with head CT results.
For 1,920 patients with GCS 14–15, the sensitivity, specificity, positive predictive
value, negative predictive value, positive likelihood ratio, and negative likelihood
ratio were 0.973 (0.924–0.994), 0.367 (0.345–0.390), 0.088 (0.073–0.105), 0.995 (0.987–0.999),
1.5 (1.457–1.618), and 0.07 (0·00–0·159.[76]
The biomarkers may improve prediction of neurological outcomes and mortality in patients
with moderate-to-severe TBI over clinical characteristics alone. GFAP appears to be
the most promising for this.
Conclusion
A panel of several different biomarkers, all associated with injury severity, with
the different cellular origin and temporal trajectories, can help in the prediction
of CT abnormalities after mTBI, and severity and outcome of moderate and severe TBI.
