Blood Biomarkers in Brain Injury Medicine

Abstract

Purpose of Review:

This review seeks to explore blood-based biomarkers with the potential for clinical implementation.

Recent Findings:

Emerging non-proteomic biomarkers hold promise for more accurate diagnostic and prognostic capabilities, especially in the subacute to chronic phase of TBI recovery. Further, there is a growing understanding of the overlap between TBI-related and Dementia-related blood biomarkers.

Summary:

Given the significant heterogeneity inherent in the clinical diagnosis of Traumatic Brain Injury (TBI), there has been an exponential increase in TBI-related biomarker research over the past two decades. While TBI-related biomarker assessments include both cerebrospinal fluid analysis and advanced neuroimaging modalities, blood-based biomarkers hold the most promise to be non-invasive biomarkers widely available to Brain Injury Medicine clinicians in diverse practice settings. In this article, we review the most relevant blood biomarkers for the field of Brain Injury Medicine, including both proteomic and non-proteomic blood biomarkers, biomarkers of cerebral microvascular injury, and biomarkers that overlap between TBI and Dementia.

Introduction

Traumatic Brain Injury (TBI) is a significant cause of morbidity and mortality in the United States, but existing clinical classification systems of TBI have shown limited value regarding diagnosis, prognosis, and research utility, particularly in cases of mild TBI (mTBI) [1–5]. Different TBI subtypes, including diffuse axonal injury (DAI) [6] and diffuse cerebral microvascular injury[7], trigger different biochemical pathways that have the potential to produce accessible and measurable specific blood biomarkers of resulting individual or overlapping TBI endophenotypes [1, 2, 7–9]. Thus, there has been significant interest in identifying blood-based biomarkers that can accurately classify TBI into severity level, identify specific TBI-induced neuropathological endophenotypes, predict adverse outcomes, and accurately identify patients who could benefit from targeted treatment(s) or participation in clinical trials [1, 7–10].

There has been an exponential increase in the number of TBI-related blood biomarker studies performed over the past two decades [2], yet there remains a paucity of clinically meaningful blood biomarkers available to Brain Injury Medicine clinicians. Some of the past stalemate has been due to the significant heterogeneity implicit with TBI exposure, including injury severity, anatomical location(s) injured, resulting pathophysiological endophenotype(s), and neurological and non-neurological comorbidities prior to TBI exposure. Given this complexity, members of our team recently performed a quantitative assessment of the diagnostic and prognostic accuracy of available TBI-related blood biomarkers, parsed by their discriminative ability to: 1) confirm the presence of a mTBI/concussion; 2) predict intracranial damage requiring neuroimaging assessment acutely after mTBI/concussion exposure; 3) predict delayed recovery after mTBI/concussion; and 4) predict long-term outcome following severe TBI exposure [1]. In this present article, we first review the most relevant blood biomarkers within the field of Brain Injury Medicine, including both proteomic and promising non-proteomic blood biomarkers, and then present research into the specific TBI endophenotype of cerebral microvascular injury. Finally, we introduce evolving research demonstrating overlap between TBI-related and Dementia-related blood biomarkers.

Proteomic Blood Biomarkers of Traumatic Brain Injury

Despite their brief half-lives, protein biomarkers have been the primary focus of TBI blood biomarker research over the past two decades [1, 2, 8]. S100 calcium binding protein B (S100B), an astrocytic protein upregulated and released in response to stress, has been one of the most widely studied TBI-related biomarkers to date and is proposed to be a clinical predictor for the need to perform neuroimaging acutely following TBI exposure [8]. Further, elevated levels of S100B have been associated with increased glial activation, secondary injury progression, and a worsened prognosis following TBI exposure [2]. However, little evidence exists to show that it outperforms current emergency assessments of TBI by trained clinicians [4, 11]. In addition, significant limitations make S100B a blood biomarker with poor clinical utility [4, 8, 12–18]. These limitations include its lack of neuronal specificity, as it is also upregulated in response to physical exertion and polytrauma, as well as the short time frame of three to six hours post-TBI exposure within which it must be collected and measured [4, 8, 15–18]. In addition, the lack of congruity amongst reference values for different age groups further limits its utility [13, 14]. As a result, new TBI-related protein blood biomarkers have been under investigation in recent years [2, 8], including: neurofilament light (NfL), ubiquitin C-terminal hydrolase (UCH-L1), glial fibrillary acidic protein (GFAP), and tau.

NfL is a cytoskeletal protein mainly found in the myelinated axons of subcortical neurons, making it more specific to the brain [1, 8, 19]. Following TBI exposure, NfL is released into the extracellular space due to DAI (specifically damage to the axonal membrane and cytoskeleton), resulting in elevated serum levels of NfL [8]. Considerable research has shown that serum NfL levels can distinguish between mild, moderate, and severe TBI not only acutely, but also months to years post-injury, indicating that TBI induced DAI is a chronic neuropathology [5, 8, 9, 20–22]. Indeed, traumatic DAI has been proposed as an underlying mechanism for both persistent post-concussive syndrome (PCS) and late-life neurodegeneration following TBI exposure [5, 9]. Thus, it is notable that following TBI exposure, increased serum NfL has shown associations with longitudinal progression of DAI, delayed recovery, and overall worse patient outcomes [1, 5, 8, 9, 19]. In addition, there is some evidence that elevated serum NfL is associated with long-term reductions in brain volume and white matter integrity following TBI exposure [5, 8, 9]. Importantly, serum NfL levels remain unchanged in response to exercise in healthy controls [5]. Collectively, there is mounting evidence that NfL is a higher quality protein blood biomarker of TBI compared to S100B.

UCH-L1, another neuronal protein, helps to tag cytosolic proteins for ubiquitination or degradation [1, 8, 23]. Due to this function, UCH-L1 has been proposed as a blood biomarker of TBI exposure, as it is upregulated in response to neuronal damage [8, 24]. To date, more evidence supports a role for UCH-L1 to aid in TBI diagnosis than for TBI prognostication [11]. For example, in cases of severe TBI, serum UCH-L1 levels have been found to be higher in patients with neuroimaging findings of DAI compared to patients with focal injuries [21, 25]. However, some studies have claimed that UCH-L1 does have some prognostic value following severe TBI exposure, including elevated serum UCH-L1 levels being associated with worse memory performance and increased mortality [21, 26, 27]. Unfortunately, other studies have shown that the clinical utility of UCH-L1 alone (as opposed to clinical utility when combined with GFAP, as detailed below) only applies to severe TBI, as UCH-L1 levels only show weak associations with injury severity and neuroimaging results in mild-to-moderate TBI [9]. Overall, there needs to be more rigorous investigations into the individual clinical utility of UCH-L1 as an individual blood-based biomarker in TBI diagnosis and prognosis [24].

GFAP is a cytoskeletal intermediate filament expressed in astroglia [1, 8, 23, 28, 29]. It has been suggested as a blood biomarker of TBI due to its upregulation during astrogliosis, a protective process that occurs in response to central nervous system (CNS) injury [8]. For example, elevated serum GFAP levels have been observed in response to a cerebral vascular accident (CVA), with evidence to support its association with 3-month CVA outcome [28, 30]. In moderate-to-severe TBI, increased GFAP levels have shown associations with unfavorable outcomes and reduced perceptual reasoning [26, 28]. Additionally, elevated GFAP levels have shown significant accuracy in predicting 6-month mortality following severe TBI [21, 31]. In mTBI, some studies have suggested that serum GFAP can guide treatment by predicting the need for neuroimaging and neurosurgical intervention post-injury [1, 21, 32]. However, other studies claim that GFAP is not clinically useful in mTBI because the protein is also expressed at low levels outside of the brain [28, 33, 34]. Interestingly, the clinical utility of both GFAP and UCH-L1 increases considerably when the two proteins are measured together [1, 8, 35]. A current FDA-approved dual marker assay for serum GFAP and UCH-L1 has shown promise as a tool for documenting mTBI/concussion, as well as predicting the need for neuroimaging following mTBI exposure [1, 8, 35]. Due to their longer half-lives, these two proteins have the potential to outperform S100B in diagnostic utility, though more research is needed [8, 22, 23].

Tau is an axonal protein abundant in the CNS that promotes microtubule viscoelasticity via phosphorylation [1, 8, 36]. In response to TBI exposure, tau is disrupted from microtubules, and abnormally phosphorylated tau (P-tau) levels increase [1, 8, 37, 38]. This type of tau hyperphosphorylation has also been observed in neurodegenerative conditions in which toxic deposits containing tau form in the brain [1, 8, 39]. As a blood biomarker of TBI exposure, proteomic investigations of total tau (T-tau) and P-tau have been mixed [8, 9, 40, 41]. In a 2014 prospective cohort study of DAI and astroglial injury following mTBI/concussion in professional ice hockey players, researchers measured serial plasma samples from players and found that athletes’ T-tau levels post-mTBI/concussion were significantly higher than their preseason levels, both immediately after injury and several hours later [40]. Athletes with more severe symptoms (defined as symptoms that lasted more than ten days or mTBI/concussion exposure that resulted in a loss of consciousness) also showed higher levels of T-tau, prompting researchers to ultimately suggest that T-tau levels may be an indicator of mTBI/concussion severity [40]. In contrast, as an acute blood biomarker of TBI exposure, studies have shown that serum T-tau only shows a notable association with neuroimaging findings in cases of severe TBI exposure, does not correlate with clinical outcomes in the chronic stage of TBI recovery, and increases in healthy individuals with physical exertion, like S100B [8, 9]. While there is also some evidence to suggest that the P-tau/T-tau ratio can discriminate between TBI severities acutely, as well as predict the need for acute neuroimaging following mild-to-moderate TBI exposure, more research is needed before proteomic analysis of T-tau or P-tau is utilized in routine TBI clinical care [8, 21, 38, 42].

Several studies have identified new potential protein biomarkers of TBI beyond those discussed above, pointing to specific TBI endophenotypes with potential therapeutic specificity. One study suggests that serum high sensitivity C-reactive protein (hsCRP) may help identify TBI patients who could benefit from anti-inflammatory interventions [10], while another supports serum cardiac troponin as a means for identifying TBI patients who could benefit from beta-blockers [43]. Other current approaches to identifying these types of biomarkers include quantitative plasma proteomic profiling, examining the contents of extracellular vesicles released following TBI, and investigating biomarkers associated with specific symptoms post-injury [21, 44–48]. However, while there is evidence to suggest that protein panels may be more useful as TBI biomarkers than any one serum protein alone, the current literature points to the critical need to examine non-protein biomarkers following TBI exposure, given the general limitations of proteins as blood biomarkers following any type of injury, including their relatively short half-life [21].

Emerging Non-Proteomic Blood Biomarkers of Traumatic Brain Injury

While the majority of TBI-related blood biomarker research to date has been focused on proteomic biomarkers, multiple drawbacks exist regarding their clinical utility. These drawbacks include: vulnerability to degradation in the bloodstream, poorly validated degree of enrichment within the brain parenchyma, expressional variability across brain regions, and often significant extra-neuronal expression [8, 49]. For example, many proposed proteomic TBI-related blood biomarkers – including S100B and UCH-L1 as detailed above, along with alpha-II spectrin (spectrin-αII), amyloid beta (Aβ), neurofilament heavy chain (NfH), and neuron-specific enolase (NSE) – have demonstrated expression that is poorly restricted to the CNS. Thus, their diagnostic or prognostic specificity following TBI exposure is limited when directly measured through traditional proteomic analysis of blood products. This is especially true in the setting of polytrauma, when other non-CNS injuries have occurred [49]. In their 2021 review of emerging blood biomarkers for TBI, Huibregtse, et al. describe how it is possible that non-proteomic blood-based biomarkers, including extracellular vesicles (EVs) and microRNAs, will yield more diagnostic and prognostic accuracy for the clinical practice of Brain Injury Medicine [8]. While non-proteomic biomarker research is in its infancy relative to the study of proteomic biomarkers, recent research into these non-proteomic biomarkers provides reason for optimism regarding their potential to enhance diagnostic and prognostic accuracy following TBI exposure.

EVs are membrane-bound organelles released from cells, whose collective contents can differ based on their cellular origin and the physiological or pathological condition of the cell [8, 50]. Following neuronal injury, CNS-generated EVs enter the peripheral bloodstream either by crossing the blood brain barrier (BBB) or through the glymphatic pathway and can serve as peripheral indicators of neuropathological processes [8]. EVs are known to cross the BBB under a variety of different circumstances, however the transport routes and the mechanisms for which the EVs are able to cross the BBB remain unclear [51]. The glymphatic pathway describes a waste clearance system within the CNS that functions in a similar manner as a lymphatic system. As the CNS does not have lymphatic vessels to assist in processing and removing waste products, the glymphatic pathway instead depends on a glial perivascular network [52]. Potential advantages of utilizing EVs as TBI-related blood biomarkers include their ability to protect internal protein cargo from degradation, given their membrane encapsulation, and the relative ease associated with their identification in the peripheral bloodstream [53]. For example, the detection and quantification of Aβ-42, GFAP, interleukin-10 (IL-10), NfL, tau, and other cargo proteins within EVs is currently being explored in both the acute and chronic phase of TBI recovery [45, 54–59]. While more research is needed, studies have indicated that EVs have the potential to become meaningful blood-based biomarkers that could positively impact TBI diagnosis and prognosis, across the spectrum of TBI severity and in a variety of treatment settings [8].

MicroRNAs are a class of small single-stranded RNAs that do not directly encode proteins, but instead are involved in the epigenetic regulation of messenger RNAs being translated into proteins [8]. With regard to CNS physiology, microRNAs have been shown to be essential for a range of CNS processes, including neuronal differentiation and maturation, as well as synaptic formation and pruning [8, 60, 61]. Thus, abnormal expression of microRNAs within the CNS have been associated with both TBI exposure and other neurodegenerative diseases [8, 59–66]. Characteristics of microRNAs that could prove advantageous as biomarkers for TBI include their high expression in the brain and their stability in peripheral biofluids, both as freely circulating molecules and as a cargo component of EVs [2, 61]. To date, multiple studies indicate that microRNAs may have particular diagnostic and prognostic utility following severe TBI exposure [8, 60, 62–64]. In contrast, while microRNA studies in acute mTBI have demonstrated mixed results due to heterogenous microRNA response profiles, emerging research indicates microRNA expression profiles may become more apparent and clinically meaningful in the chronic phase of mTBI recovery [8, 59, 65, 66]. Combined with continued advancements in our understanding of non-modifiable genetic factors, further research into microRNAs and other epigenetic biomarkers following TBI exposure could significantly impact the clinical practice of Brain Injury Medicine in the next decade [2].

Blood Biomarkers of Traumatic Cerebral Microvascular Injury

Cerebral microvascular injury is an increasingly recognized and studied TBI endophenotype, in both pre-clinical and clinical TBI-related research [7]. The brain, more than any other organ, requires a consistent blood supply to maintain normal neurophysiological function. Further, the rate of cerebral blood flow to the various sub-components of the brain is tightly regulated by the cerebral microvasculature and closely coupled to the metabolic demands of the given brain region, a process in continual dynamic equilibrium. A unique feature of the cerebral microvasculature is its dependence on both neural and vascular elements, collectively termed the Neurovascular Unit (NVU) in 2001 [67], to regulate this homeostasis [7]. The NVU is composed of perivascular astrocytes, pericytes, and a layer of capillary endothelial cells that connect via intercellular tight junctions or adherens junctions to form the BBB [68]. Because of its central role in the anatomic and functional makeup of the cerebral microvasculature, the NVU is frequently compromised in TBI [7]. Following TBI exposure, the metabolic demands of the brain shift and damaged neurovascular communication networks of the NVU are unable to appropriately respond. Further, the compromised NVU/BBB affords unregulated extravasation of blood products into the brain parenchyma, disrupting the extracellular environment within the brain parenchyma [69, 70]. Evidence of this sequence has been noted across the spectrum of TBI severity [7, 69], necessitating the development of non-invasive blood biomarkers to identify this important TBI endophenotype [7].

One proposed biomarker for traumatic cerebral microvascular injury is the calcium binding protein S100B due to its production by perivascular astrocytes, a key component of the NVU [1, 2]. However, as detailed in the Proteomic Blood Biomarkers of Traumatic Brain Injury section above, a number of significant limitations reduce the ability of S100B to serve as a reliable blood-biomarker for TBI in general, and traumatic cerebral microvascular injury in particular [4, 8, 11–14]. Additional proposed blood-biomarkers for traumatic cerebral microvascular injury include endothelin-1, von Willebrand Factor (vWF), and fibronectin [71–75]. The capillary endothelial cells of the NVU, like other vascular endothelial cells, produce endothelin-1, a potent vasoconstrictor and inflammatory mediator [72, 74]. Plasma concentrations of endothelin-1 have been shown to increase in the acute period following TBI exposure [73, 75]. Similarly, vWF, a glycoprotein released by endothelial cells in response to injury, and cellular fibronectin, an extracellular matrix glycoprotein involved in angiogenesis and wound healing, have shown promise as blood-biomarkers of traumatic cerebral microvascular injury [71]. While these preliminary studies are promising, additional research is needed to identify and characterize reliable blood-based biomarkers of cerebral microvascular injury, in both the acute and chronic time-points following TBI exposure [7].

Blood Biomarkers of Neurodegeneration: Overlap between TBI and Dementia

Recently, the independent fields studying blood biomarkers for TBI [2] and dementia [39] have provided overlapping evidence to support the link between TBI exposure and progressive neurodegeneration, including accelerated age-related cognitive decline and dementia [76–79]. These findings are expected, given the fact that neuropathological evaluation in the chronic phase of TBI recovery for both repetitive mTBI and single severe TBI have clearly demonstrated neuropathological findings common to neurodegenerative disorders, including Alzheimer’s disease (AD) and other dementias [80]. Indeed, several key proteins associated with neurodegeneration also serve as blood-based biomarkers of TBI and are now being assessed in aging TBI survivors with and without clinical cognitive impairment.

Two of the most studied blood biomarkers for AD and other dementias are P-tau and NfL [39]. Interestingly, a recent 2020 study of veterans with remote TBI exposure found an association between higher levels of the CNS-enriched EV cargo proteins P-tau and NfL, and objective cognitive impairment [76]. Moreover, the differences in P-tau and NfL levels were detected in veterans who, on average, sustained their last TBI exposure over three decades prior to the study [76]. This finding supports the hypothesis that EV-based blood biomarkers can be indicative of past TBI exposure decades after the injury occurs [79]; it also provides direct evidence for the overlap in TBI and dementia blood biomarker investigations [2, 39]. While these results are exciting and promising, future studies investigating diagnostic and prognostic assessment tools for both TBI and neurodegenerative diseases will likely need to incorporate a combination of blood biomarkers (likely as a combined panel of targets) and advanced neuroimaging modalities [79, 81].

Conclusions

The rapidly evolving field of TBI-related blood biomarker research holds significant promise for providing Brain Injury Medicine clinicians with objective markers of CNS injury, with the goal of improving diagnostic and prognostic accuracy. Specifically, it is anticipated that emerging blood biomarkers will be able to confirm the presence of a mTBI/concussion, predict intracranial damage requiring acute neuroimaging assessment after mTBI/concussion exposure, predict delayed recovery after mTBI/concussion, reveal specific TBI endophenotypes, and predict long-term outcome following either repetitive mTBI or single severe TBI exposure. Further, the anticipated ability to identify specific TBI-related endophenotypes will likely improve patient outcomes by affording the research community the opportunity to parse the heterogeneity inherent with a “TBI” diagnosis and investigate novel therapeutic interventions targeting specific pathological processes. Finally, it is anticipated that the overlapping fields of TBI-related and Dementia-related blood biomarker research will continue to highlight common underlying neuropathological processes between TBI exposure and age-related neurocognitive decline and dementia.