Use of brain injury biomarkers in critical care settings

Biofluid-based biomarkers for detecting brain injury in the ICU

A biomarker is defined as a “biologically based parameter that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention.” During brain injury, a variety of cellular changes can occur, including degeneration, protease activation, oxidative stress, and metabolic disturbances. These changes result in the shedding of specific proteins into the cerebrospinal fluid (CSF) or serum that can be identified and studied for their association with disease presence, outcome, and progression. These biofluid-based biomarkers reflect the earliest changes that occur in the cells before the evidence of injury appears on images. Therefore the use of biomarkers could offer a rapid, noninvasive, and cost-effective tool for the diagnosis of brain injury and determination of the subsequent need for additional diagnostic testing, monitoring, or therapeutic intervention.

Most biofluid-based markers of CNS injury to date are proteins or protein fragments. In the context of basic research, several brain injury biomarkers, including NSE, glial protein S-100B , glial fibrillary acidic protein (GFAP), and myelin basic protein (MBP), have been shown to have great utility in TBI specifically. Alpha-II spectrin protein and its breakdown products (SBDPs) are potential biomarkers of necrosis and apoptosis after TBI. The cleaved form of the axonally located microtubule binding protein tau (c-tau) has been identified as a new biomarker in mouse and rat TBI models. In addition, inflammation markers such as neurofilament-H are promising axonal injury biomarkers of various forms of acute brain damage in experimental TBI models.

Proteomics is the large-scale study of proteins, particularly their structures and functions. This field includes the study of changes in protein expression patterns as related to diseases and environmental conditions. The search for biomarkers of TBI has been approached by integrating biofluid and tissue information. This new approach takes advantage of functional synergy between certain biofluids and tissues with the potential for clinically significant findings. Using differential neuroproteomic methods, a systematic assessment has successfully identified additional protein biomarkers for TBI, such as ubiquitin C-terminal hydrolase-L1 (UCH-L1) and microtubule-associated protein 2 (MAP2), with relevant animal models.

The application of basic scientific discoveries to the clinical setting remains a challenge. Systems biology (the computational and mathematical modeling of complex biologic systems) is an approach to building a holistic, systematic, and unbiased understanding of the structural and behavioral elements of biologic networks. Translating these findings into the clinical, data-driven development cycle and data-mining steps for discovery, qualification, verification, and clinical validation are needed. Data mining techniques used in this field extend beyond the level of data collection to an integrated scheme of animal modeling, instrumentation, and functional data analysis. In the context of TBI proteomics, systems biology tools can be incorporated in several ways to overcome many of the limitations of simple proteomics.

Current biomarker candidates and their properties

Although neurobiomarkers are not yet Food and Drug Administration (FDA) approved for clinical use, considerable research into protein biomarkers for TBI has produced several putative diagnostic and prognostic markers. Table 44.1 describes the most studied potential TBI biomarkers and reflects the current state of the art. Other possible markers include neurofilament proteins MAP2 (2A, 2B), fatty acid binding protein-H (H-FABP), brain-derived neurotropic factor (BDNF), and autoantibodies to brain antigens. But more studies are required for further clinical utility verification and biomarker characterization.

The first listed biomarker is GFAP . It is an astrocyte-specific intermediate filament protein known as a marker of astrocyte activation. Eight different isoforms of this protein are expressed across numerous subsets of astrocytes. Measurements of GFAP and its breakdown products have provided promising data on injury pathways, focal versus diffuse injuries, and prediction of morbidity and mortality. It has also been studied in both CSF and serum of patients with severe TBI. Serum GFAP levels in severe and moderate TBI with Glasgow Coma Score (GCS) <12 are associated with unfavorable outcome at 6 months. New enzyme-linked immunosorbent assay (ELISA) for GFAP and its breakdown products detected both mild-to-moderate TBI and the full spectrum of TBI (TRACK-TBI cohort) in two independent studies. Similarly, Metting and colleagues demonstrated that serum GFAP was increased in patients with an abnormal CT after mild TBI. Another follow-up study using the TRACK-TBI cohort demonstrated that the combination of UCH-L1 with GFAP/BDP further improves its diagnostic utility. A recent paper also found that a caspase-6–generated form of GFAP is elevated at 72 hours after cardiac arrest, but did not predict outcome.

UCH-L1 is a deubiquitinating enzyme highly expressed in neuronal cells. It is one of the few markers identified using proteomic methods. In addition, its high brain specificity and abundance in brain tissue make it an attractive candidate marker. CSF and serum UCH-L1 levels were found to be elevated in patients with severe TBI, correlating with the severity and outcome of injury. ^, Increased levels of UCH-L1 post-TBI are proposed to be secondary to blood-brain barrier (BBB) dysfunction. Several other studies also report the detectability of UCH-L1 in blood after mild TBI and with UCH-L1 levels correlating with traditional clinical assessments. ^, However, the utility of UCH-L1 in mild TBI with respect to its sensitivity and specificity requires further clinical assessment.

In the ALERT-TBI trial (1959 participants), UCH-L1 and GFAP assays in combination were found to discriminate those with CT abnormalities defined clinically with mild to moderate TBI when measured within 4 hours postinjury. These findings in part led to US FDA clearance for testing of this combination among the mild TBI cohort.

Tau is an intracellular MAP with a molecular mass of 48–67 kDa that is highly enriched in axons. TBI was found to cause the cleavage of tau protein, with elevated levels of c-tau in CSF and serum. c-tau possesses many desirable characteristics of a biochemical marker and is associated with both disruption of the BBB and postinjury cleavage of tau protein. Other studies from two groups have demonstrated the significance of tau/c-tau in predicting outcome in severe TBI patients. ^, Similarly, several studies report the utility of tau or c-tau in the prediction of outcome in mild TBI. ^, However, other studies have reported the poor ability of tau protein to predict outcome and postconcussion syndrome in mild TBI. Recently, two ultrasensitive assay platforms were developed to enable the robust detection of tau (and a phosphorylated form of tau) in serum from acute-phase TBI patients with various levels of TBI severity.

S100B is a glia-specific calcium-binding protein. Elevated S100B levels accurately reflect the presence of neuropathologic conditions, including TBI or neurodegenerative diseases linked to astroglial injury. More importantly, S100B levels are reported to rise before any detectable changes in intracranial pressure, neuroimaging, or neurologic examination findings. S100B is considered a prognostic biomarker of BBB permeability and CNS injury. Several studies have reported that S100B protein might detect brain death after severe TBI. ^, Another study showed that serum and urine levels of S100B after TBIs have prognostic significance for survival and disability. A similar study on serum S100B measured 24 hours after injury reported that it predicts unfavorable outcome (i.e., Glasgow Outcome Scale [GOS] score <4 or death at 3 months after injury in severe TBI patients). S100B elevation has also been found after mild TBI. Although S100B remains promising as an adjunctive marker, the main limitation toward its use is the lack of specificity for brain trauma; this is likely, given that S100B can be released by cells other than astrocytes.

αII-spectrin is a cytoskeletal protein enriched in neuronal axons and presynaptic terminals. SBDPs are produced by the breakdown of αII-spectrin by calpain and caspase, which are activated in the brain after TBI. SBDPs thus reflect axonal damage. SBDP150 and SBDP145 are indicative of calpain activation, often associated with acute necrotic neuronal cell death, and SBDP120 is generated by the action of caspase-3 and is associated with delayed apoptotic neuronal death. αII-spectrin has been studied primarily in the context of severe TBI. Elevation of SBDP150 and/or SBDP145 levels in CSF was reported as a possible outcome predictor in patients with severe TBI versus initial CT diagnosis with Marshall grade. SBDPs, especially CSF SBDP150, may be useful as a differential diagnostic biomarker for its ability to distinguish between focal and diffuse injury in the acute phase of TBI. ^, Whereas αII-spectrin is present in various nucleated cells and most tissues, its high abundance and enrichment in brain and the fact that SBDPs are injury generated make SBDPs potentially useful TBI biomarkers, especially in combination with other more brain-specific markers.

MBP is one of the most abundant proteins in white matter, composing 30% of the myelin protein. It is important in the myelination of nerves. As a constituent of the sheath, MBP is essential for normal myelination and axonal signal conduction. Several studies on severe TBI patients have reported that MBP levels could track the occurrence of post-TBI hypoxia, predict the outcome, and prompt adequate treatment. ^, Serum MBP is elevated in the majority of children with acute TBI, including well-appearing children with TBI from child abuse in whom the diagnosis might otherwise have been missed. Most recently, a review by Kochanek and colleagues suggests that MBP is a potential biomarker for pediatric TBI. Because MBP lacks clinical sensitivity, the interest in MBP as a biomarker for TBI is lower than that of S100B , NSE, and GFAP. One possible explanation for this lack of sensitivity is that MBP undergoes extensive fragmentation/degradation after TBI, thus complicating its robust detection by traditional sandwich ELISA.

NSE is a glycolytic enzyme that is present in central and peripheral neurons and neuroendocrine cells, with serum levels rising after cell injury. NSE is passively released into the extracellular space only under pathologic conditions during cell destruction. Acute post-TBI levels of NSE and MBP were correlated with outcome in children, particularly those under 4 years of age. ^, In the setting of diffuse axonal injury in severe TBI, levels of NSE at 72 hours after injury have shown an association with unfavorable outcome. However, in very early studies, serum or CSF NSE was considered of limited utility as a marker of neuronal damage. ^, A limitation of NSE is the occurrence of false-positive results that occur because NSE is also present at high levels in red blood cells.

Neurofilament (NF) proteins are the key intermediate filaments in neurons and a major component of the axonal cytoskeleton. The major neuronal filaments in the CNS are those assembled from NF triplet proteins: neurofilament light (NF-L; 61 kDa), medium (NF-M; 90 kDa), and heavy (NF-H; 115 kDa). After TBI, calcium influx into the cell contributes to a cascade of events that activates calcineurin, a calcium-dependent phosphatase that dephosphorylates neurofilament side-arms, presumably contributing to axonal injury. Phosphorylated NF-H (pNF-H) was found to be elevated in the CSF of adult patients with severe TBI compared with controls. Similarly, hyperphosphorylated NF-H has also been shown to significantly correlate with neurologic deficit in severe TBI children. More recently, phosphorylated NF-H was shown to stratify lower grades of injury, with significant rises in pNF-H seen up to 3 days after mild TBI. pNF-H levels in CSF are also elevated in amateur boxers. Although pNF-H is showing promise as both a sensitive and specific marker of axonal injury after TBI, consideration of other NF isoforms is needed to stratify injury severity in TBI patients. For example, amateur boxers also have elevated CSF levels of NF-L.

TBI is a complex injury: primary injury occurs at the moment of trauma, when tissues and blood vessels are stretched, compressed, and torn; secondary injury then follows. Secondary injury events include damage to the BBB, release of factors that cause inflammation, free radical overload, excessive release of the neurotransmitter glutamate (excitotoxicity), influx of calcium and sodium ions into neurons, and dysfunction of mitochondria. We thus anticipate a growing list of putative TBI biomarkers with different cell or subcellular origins and different diagnostic and prognostic properties. Such findings are also very likely for other relevant CNS insults in neurointensive care. Preclinical and clinical studies suggest the potential alteration of the following proteins in some cases of TBI: neurite degeneration markers, MAP2, amyloid β peptide (Aβ1-40, Aβ1-42), neuroinflammatory markers (microglial ionized calcium-binding adapter molecule 1, inflammasome proteins-caspase-1, Nacht leucine-rich-repeat protein-1 (NALP-1), apoptosis-associated speck-like protein containing a caspase recruitment domain (ACS), ^, biofluid levels of neurotropic markers (BDNF, nerve and growth factor, and heart-type fatty acid—binding protein). Last, autoimmune markers (autoantibodies to brain antigens such as GFAP) have also been reported as potential biomarkers for subacute or chronic phases of TBI. However, future work is required for their clinical utility verification and biomarker characterization.