Imaging of Neurotrauma in Acute and Chronic Settings.

Traumatic injuries of the brain and spinal cord are a significant source of mortality and long-term disability. A recent systematic study in a rat model of spinal cord injury (SCI) indicates severe, destructive, and very protracted inflammation as the key mechanism initiated by the massive injury involving the white matter. Although the severe inflammation is localized and counteracted by astrogliosis, it has a damaging effect on the blood vessels in the surrounding spinal cord, leading to persistent vasogenic edema. Evaluation of these injuries with imaging of the brain and spinal cord plays a crucial role in the acute trauma work-up, allowing clinicians to quickly identify abnormalities that require immediate medical or surgical intervention or to exclude them from the workup. Recently, anti-inflammatory agents have been shown to inhibit and accelerate the elimination of post-SCI inflammation in preclinical studies, and an exciting potential has arisen for the use of antiinflammatory drugs in clinical studies to achieve neuroprotection (i.e., inhibition of destruction caused by inflammation) and to inhibit vasogenic edema in SCI, traumatic brain injury, and stroke. In both subacute and chronic settings, imaging can guide therapy and provide important prognostic information. In this review, we discuss the imaging workup and evolving imaging findings of neurotrauma in the acute and chronic setting, including conventional and advanced imaging techniques. As neuroimaging is the primary mode of diagnostic analysis in neurotrauma, it is a critical component in future clinical trials evaluating neuroprotective therapies. Copyright© Bentham Science Publishers; For any queries, please email at epub@benthamscience.net.

INTRODUCTION

Traumatic brain injury (TBI) and spinal cord injury (SCI) represent significant global health problems that affect millions of people annually [1]. Falls and road injuries are the largest contributors to both types of injuries, and TBI has shown a steady increase in global incidence since 1990 [1]. TBI remains the leading cause of morbidity and mortality globally for people younger than 45 years, and the resulting long-term effects of TBI can carry significant economic and social consequences [2-4].

Current clinical treatment of acute trauma to the neuroaxis after hemodynamic optimization is predominantly aimed at addressing structural issues, both contributing to primary damage (e.g., fixation of an unstable spine fracture or evacuation of an extra-axial hematoma) and avoiding those that could contribute to secondary damage (e.g., a decompressive craniotomy for increased intracranial pressures) [5, 6].

In pathologic terms, trauma to the neuroaxis in the first 2 days is considered the acute phase and is characterized by cellular necrosis, edema, and hemorrhage. The subsequent inflammatory phase starts around day 3 and is characterized by destructive inflammation from macrophage infiltration, which can last beyond 16 weeks [7]. Recent elucidation of the pathogenesis of neurotrauma in a rat model of SCI [8] indicates that the severity of the inflammation is associated with two fundamental pathologic processes: (1) destruction of the spinal cord and (2) vasogenic edema in the spinal cord surrounding the localized lesion.

In current clinical practice, pharmacological treatment of this inflammation has been limited. Early corticosteroid therapy has long been controversial in the treatment of SCI, and it is not supported in the treatment of TBI [5, 6]. Rat model studies of SCI have shown that 1-2 weeks of intradural dexamethasone showed the neuroprotective effect of marked inhibition of macrophage infiltration [9, 10]. However, the inflammation-inducing necrotic myelin-rich debris remained in the cavity of injury (COI), which would likely result in recurrent inflammation upon cessation of steroid therapy [9-11]. Longer therapy could not be administered secondary to the deleterious systemic effects of long-term steroid therapy [11]. There are numerous potential pharmacological neuroprotective agents that are being investigated in animal and clinical trials, which include estrogen, riluzole, glibenclamide, cetherin, fumaric acid esters, endaravone, N-Palmythiolethalomine-oxazoline, and mytramycine A. These candidate treatments so far have been used only in the short term, and their efficacy with respect to long-term inflammatory changes is yet to be seen. Recently, Serp-1 and M-T7 (myxoma virus–derived immunomodulatory proteins) have been studied in rat model SCI, including longer-term intradural infusion, with improvement in inflammatory changes pathologically [7, 12].

Since anti-inflammatory agents can inhibit [9-11] and, after 8 weeks of administration, eliminate [12] severe inflammation in the spinal cord in a rat model, neuroimaging used in a systematic fashion can monitor the neuroprotective effect of candidate anti-inflammatory agents in clinical trials, involving SCI and TBI patients. In addition, magnetic resonance imaging (MRI) may prove very useful for monitoring edema around the areas of inflammation that apparently persist as long as inflammatory macrophages persist, which is greater than 16 weeks in the rat [8], indicating inflammatory vascular damage and vasogenic edema [13]. Whether effective anti-inflammatory therapies will achieve neuroprotection and eliminate edema in TBI and SCI remains to be seen, but a required protracted course of the treatment opens an opportunity for developing novel neuroimaging protocols designed to systematically monitor changes in the size of the lesion and perilesional edema, thus directly and in real-time determining the effectiveness of an anti-inflammatory agent. Given the lack of other in vivo analytic methods to monitor the long progression of destructive inflammation initiated by SCI or TBI and the inhibitory effect of anti-inflammatory agents, neuroimaging with its established and widely available technology is well-positioned to provide the first accurate analytical monitoring of treatment efficacy in the clinical setting.

TBI is defined as traumatically induced structural injury, physiological disruption, or both to brain function caused by an external force, and the patient must also exhibit a loss or decreased level of consciousness, loss of memory surrounding the time of injury, neurological deficits, or an intracranial lesion on imaging [14]. Typically, TBI is assessed clinically with the Glasgow Coma Scale (GCS) at admission, where a score of less than 8 is considered severe, a score of 9-12 is considered moderate, and a score of 13-15 is considered mild. Severe and moderate TBIs each makeup about 10% of TBIs, while 80% fall into the mild category [15, 16]. Although this scale is a simple classification system, it is still frequently used in clinical practice and research because of its good interobserver reliability and prognostic capabilities [15, 16]. The GCS score is limited by factors such as intervention prior to the patient’s arrival at the healthcare facility, any additional injuries the patient may have sustained, and patients who are difficult to assess because of altered mental status or the patient’s baseline difficulty with communication for reasons unrelated to the trauma [15, 16]. Also, in mild TBI, which accounts for approximately 80% of cases, this score is a poor indicator of outcome [15, 16]. Many other classification systems can be found in the medical literature. Of note, the United States Veterans Affairs and the Department of Defense have recommended against using only the GCS, and they have published a combined classification that takes into account the presence of imaging abnormalities, presence and duration of loss of consciousness, presence and duration of altered consciousness, and presence and duration of post-traumatic amnesia [14].

In contrast to TBI, the definition of SCI does not take into account transient symptoms. Instead, motor and sensory deficits are assessed at different levels after medical stabilization. The Neurological Standard Scale published by the American Spinal Injury Association/International Spinal Cord Society is commonly used to assess the severity of deficits, replacing the older Frankel scale [17].

TBI and SCI both share the principal mechanism of focal damage from contact injuries. TBI has the additional potential mechanism of acceleration/deceleration injuries and rotational/shear injuries causing diffuse damage. A combination of the primary mechanical insult and secondary insults, such as ischemia or intracranial hypertension, contribute to the extent of the injury [2, 18]. This paper primarily focuses on imaging findings of direct injury to the parenchyma of the brain and spinal cord. Other commonly associated injuries, such as extra-axial hemorrhage, osseous fractures, and vascular injuries, are outside the scope of this article.

TRAUMA IMAGING INDICATIONS

Diagnostic imaging, along with the patient’s medical history and physical examination, plays a crucial role in assessing trauma to the head and spine. Imaging is not only critical for triage but also can quickly identify patients with injuries who would benefit from neurosurgical intervention and early medical therapy [19-21]. Over the years, clinicians have employed many different guidelines to determine whether imaging is necessary, such as the New Orleans Criteria [22], the Canadian CT Head Rule [23], and the National Emergency X-Radiography Utilization Study (NEXUS) [24] for head trauma, and the Canadian C-Spine Rule [25] and NEXUS [26] for cervical spine imaging. Each of these guidelines uses a different variation of patient history and findings from the physical examination, such as the mechanism of injury, patient demographics, and mental status. In addition, the American College of Radiology Appropriateness Criteria publishes guidelines that rate the appropriateness of different imaging studies in a wide range of clinical scenarios, including trauma to the head and spine, and that include information about relative radiation doses for radiography and computed tomography (CT) examinations [27, 28].

For acute head trauma, CT of the head without contrast is the most appropriate initial examination for mild, moderate, or severe trauma, according to the American College of Radiology Appropriateness Criteria, assuming imaging is not contraindicated by the New Orleans Criteria, the Canadian CT Head Rule, or the NEXUS criteria [27]. Plain radiography is not considered appropriate in any patient with acute head trauma. Intravenous contrast is not indicated unless vascular injury is a possibility. If a vascular injury is suspected, CT angiography of the head and neck is recommended; MRI of the head and magnetic resonance angiography (MRA) of the head and neck without contrast are also considered appropriate. Otherwise, in no other situation is MRI considered the appropriate initial imaging examination for head trauma [27].

After initial imaging, if there is a change in the patient’s neurological status, repeating CT without contrast is again the modality of choice, with MRI as the next step if the repeated CT does not explain the new symptoms. If there is no neurological change, short-term follow-up CT is considered appropriate only if the patient has an increased risk for developing hemorrhage or for worsening of existing hemorrhage, such as patients with frontal or temporal contusions, age greater than 65 years, or intracranial hemorrhage volume greater than 10 mL, or those who are undergoing anticoagulation therapy. In the subacute or chronic setting, if the patient has persistent symptoms, MRI of the head without contrast is the most appropriate study, unless the patient experiences a rapid onset of new symptoms, in that case, CT would again be recommended [27].

The guidelines for acute trauma in the spine are similar to head CT such that the gold standard for initial imaging is CT. Again, if imaging is contraindicated by the Canadian C-Spine Rule and NEXUS, then no imaging (including radiographs) is considered appropriate. MRI is considered appropriate if there is any clinical concern for SCI or acute fracture is seen prior to CT. As is the case for head trauma, CT angiography or MRA can be obtained if there are clinical or imaging concerns for trauma-related vascular injury [28].

The advantages of CT over MRI include wide availability, speed, low cost, and lack of need for metal screening. CT demonstrates excellent sensitivity for spine and skull fractures and a sensitivity close to that of MRI for acute intracranial hemorrhage. The advantages of MRI over CT include superior imaging of the brain and spine parenchyma and lack of ionizing radiation [19, 21, 29].

It should be noted that in the above standard imaging guidelines, serial imaging is usually not indicated and is typically only used in evaluating postoperative changes. In the aforementioned article [8], multiple serial MRIs of rat spinal cords were taken over the course of 16 weeks after induced SCI. A similar imaging protocol could be employed in clinical trials of neuroprotective therapies.

CONTUSION

Primary traumatic injury to the neural tissues results in vasogenic edema from mechanical injury to the blood-brain barrier and cytotoxic edema from direct cellular injury. Subsequent secondary damage is caused by alterations in blood flow, loss of cerebrovascular autoregulation, vasospasm, metabolic dysfunction, inflammation, and excitotoxicity [19]. The primary traumatic injury, when seen on imaging, is commonly referred to as a contusion, both in brain and spine imaging. Contusions can be nonhemorrhagic or contain a mixture of hemorrhagic and nonhemorrhagic components, whereas a purely hemorrhagic lesion is considered a parenchymal hematoma [30]. Contusions in the brain are most commonly seen in the inferior frontal lobe and anterior inferior temporal lobe. They can significantly increase in size in the first 48 hours, a phenomenon often referred to as blooming [19]. One large clinical study showed that approximately 50% of the contusions that were managed conservatively developed new or more hemorrhage within 48 hours [31]. Initial contusion size and the presence of coexisting subdural hematoma were demonstrated as positive predictors of progression. Also of note, the patients who required neurosurgical intervention all were presented with abnormal GCS scores. The spinal cord sustains injury in 10% to 14% of patients with spinal fractures, and the cervical spine is the most common site of injury with neurological deficits (40%) [32]. If a spinal cord contusion is hemorrhagic, it carries a worse clinical prognosis [32].

In patients with acute injuries, cerebral contusions on CT demonstrate a focal area of hypoattenuation from the edema (Fig. ). When contusions contain areas of hemorrhage, the blood products are hyperattenuating. The regions of the brain in which contusions are most commonly seen, the anterior and inferior portions of both the anterior and middle cranial fossae, present imaging challenges because the contusions are frequently in an area where beam hardening artifact from the adjacent skull base is present [19]. Cerebral contusions are most effectively identified by T2-weighted MRI (particularly T2-weighted fluid-attenuated inversion recovery [FLAIR] MRI sequences), which are significantly more sensitive for detecting contusions than CT. On FLAIR, contusions show increased T2 signal intensity. The presence of hemorrhage in a contusion can have a variety of appearances in different MRI sequences, but susceptibility artifact on gradient-recalled echo (GRE) or susceptibility-weighted imaging (SWI) is most sensitive (Fig. ). For acute injuries, contrast is not indicated for use with MRI; however, subacute contusions can enhance with contrast similar to subacute infarcts [20, 30]. Diffusion-weighted imaging (DWI) has also demonstrated sensitivity for detecting contusions and for identifying areas of ischemia from secondary injuries, such as herniation or vascular injury [20, 33].

CT is not sensitive for evaluating SCI, although it is highly sensitive for evaluating unstable fractures of the spine that are often associated with SCI. In one study, only 1 out of 367 patients with SCI diagnosed on MRI had a CT negative for post-traumatic findings, giving a false negative rate of 0.3% [34]. MRI is highly sensitive for detecting spinal cord contusion, ligamentous injury, disc injury, and injury to paraspinous soft tissues. A spinal cord contusion is apparent as increased T2 signal intensity with swelling in the acute phase (Fig. ). As in the brain, susceptibility artifact is the most sensitive sign of associated hemorrhage (Fig. ). Any patient with suspected SCI should be evaluated using MRI because MRI can provide not only the location and relative severity of the SCI but also information regarding the cause of injury, which could include hematoma, osseous fragment, or disc protrusion. The differentiation between various causes of SCI has significant implications for treatment planning. In addition, when primary injury to the spinal cord is severe enough, complete transection can occur (Fig. ) [21].

There is a relative paucity of clinical trials showing serial MRI in the acute and subacute phase after SCI or TBI, likely secondary to the current clinical imaging indications discussed previously as well as patient condition. Recently, a small prospective study with serial MRI in the first 3 weeks following cervical SCI demonstrated spinal cord edema that expanded in the longitudinal dimension in the first 48 hours, followed by a gradual decrease on the subsequent scans [35].

Recent systematic histologic analysis of the SCI in the rat model has revealed a novel mechanism of removal of scattered extravasated red blood cells in the spinal cord around the main lesion. Extravasated red blood cells probably originate from microvascular injury and are apparently phagocytized by astrocytes within the first 3 days after SCI and are absent 7 days after SCI [8]. This mechanism is distinct from the macrophage phagocytosis of red blood cells and myelin-rich necrotic debris active in the main lesion. It may be significant in concussions where neuroimaging reveals no changes and indicates efficient, noninflammatory removal of microhemorrhages by the CNS tissue reaction, specifically reactive astrogliosis.

DIFFUSE AXONAL INJURY AND TRAUMATIC AXONAL INJURY

The other form of primary injury is known as diffuse axonal injury (DAI). This term presents difficulty because it has been defined differently in radiological, clinical, and pathological terms [30]. In pathological terms, the entire brain experiences tissue deformation during an acceleration/deceleration event. Because of the organized and anisotropic structure of white matter axon tracts, these tracts are uniquely susceptible to rapidly applied mechanical forces, and white matter tracts can therefore be selectively damaged in these events [36]. The radiological definition of DAI has evolved as the development of new and improved imaging techniques have demonstrated better sensitivity for the diffused changes that occur in the brain parenchyma after acceleration/deceleration type injuries. In radiological terms, DAI refers to the widespread distribution of white matter lesions, which are defined as signal abnormalities on CT, FLAIR, GRE/SWI, contrast-enhanced T1-weighted MRI, DWI, or diffusion tensor imaging (DTI). These white matter lesions can be hemorrhagic or nonhemorrhagic. Traumatic axonal injury (TAI) refers to the same white matter lesions but in a more confined distribution [30, 33].

CT is relatively insensitive for detecting DAI; findings that can be detected are microhemorrhages (seen as small areas of hyperattenuation) (Fig. ) or diffuse post-traumatic edema [37]. MRI is considerably more sensitive for detecting both hemorrhagic and nonhemorrhagic lesions in DAI because of the susceptibility artifact caused by hemosiderin (Fig. ). GRE MRI sequences are more sensitive than traditional MRI sequences, and moving from 1.5T to 3T magnet field strength demonstrates a doubling of sensitivity for hemorrhagic lesions. Although GRE sequences are more sensitive than traditional MRI, they are not as sensitive as SWI sequences for these lesions; SWI sequences have been shown to be 3 to 6 times more sensitive than GRE sequences for microhemorrhages [20]. White matter lesions that are nonhemorrhagic can be seen on T2-weighted MRI and T2-weighted FLAIR MRI sequences as foci of hyperintensity [19, 20, 38]. DAI lesions can be seen on DWI as foci of restricted diffusion. This method is less sensitive than GRE imaging for detecting DAI/TAI hemorrhagic lesions; however, it has been shown to identify shearing injuries that are not detected on GRE or T2-weighted FLAIR images (Fig. ) [20, 33].

CHRONIC CHANGES

In the subacute and chronic phase after cerebral contusion, edema decreases and encephalomalacia develops. Encephalomalacia can be seen as atrophy of the injured parenchyma with decreased attenuation on CT images and increased signal on T2-weighted MRI sequences. Encephalomalacia can also progress to a more cystic appearance (Fig. ). A localized severe inflammatory response of very prolonged duration may follow neurotrauma, and the surrounding edema was recently proposed to result from inflammatory damage to the blood-brain barrier (BBB) or blood–spinal cord barrier (BSCB) and vasogenic edema [13]. The interstitial water is moved out of the central nervous system (CNS)

via astrocytes with expressing aquaporin-4 (AQP-4) via the pia limitans externa into the subarachnoid space, via the pia limitans interna into the ventricular system and the central canal, and via the BBB/BSCB into the blood vessels [39-43]. In SCI and TBI and severe supervening inflammation [8], vasogenic edema in the surrounding CNS is likely moved by hypertrophied astrocytic networks with hyperexpressed AQP-4 to all 3 compartments and probably also to a COI [13] forming in deep sites of trauma, indicating the role of CNS tissue response and, specifically, astrogliosis in removing excess edema water to balance its pathologic supply via vasogenic edema. These putative mechanisms need to be considered in the imaging of chronic neurotrauma patients for proper diagnostic interpretation. Recent progress with anti-inflammatory treatment of SCI [11, 12] and its potentially beneficial effect in inhibiting vasogenic edema [13] may prove to enhance the importance of MRI in clinical trials involving SCI and TBI patients. Kwiecien et al. described differentiating COI and fibrous scarring of arachnoiditis in rats, and it may be possible to identify in vivo in humans on T2-weighted imaging as the fluid-filled COI should have greater T2 hyperintensity than adjacent scarring [8]. However, a number of additional limitations on MRI in scanning live patients include limitation in scanning time, motion artifact, and artifact from surgical hardware. This possible use of imaging requires additional investigation.

Associated hemorrhagic products can resolve or persist chronically as hemosiderin. The spinal cord has a similar pattern of findings of atrophy and increased T2 signal intensity, referred to as myelomalacia.

Wallerian degeneration is another chronic finding that can also be seen in the chronic phase in the brain and spinal cord. This condition is presented as T2 hyperintensity and atrophy in the anterograde or retrograde directions along with injured white matter tracts [44, 45].

As discussed previously, white matter lesions (both hemorrhagic and nonhemorrhagic) from DAI often persist in the chronic phase as signal abnormalities on MRI. A recent study has demonstrated the long-term persistence of microhemorrhagic DAI lesions on MRI [46]. This phenomenon can be a helpful tool as a long-term imaging marker of prior trauma because many of the initial findings in TBI, such as extra-axial hemorrhage or pachymeningeal enhancement, may resolve before a patient undergoes MRI after trauma. However, the lesions are nonspecific, since a variety of common entities can closely mimic the appearance of both the hemorrhagic and nonhemorrhagic white matter lesions of DAI [44, 46].

Whole-brain atrophy is a known long-term effect of TBI. In a study on severe TBI, volumetric analysis of a 1-year follow-up scan showed a mean volume loss of 8.4% and a significant correlation between the amount of volume loss and long-term functional status [47]. Mild-to-moderate TBI also showed whole brain atrophy that correlates with functional testing. These findings suggest that volumetric analysis may prove useful in evaluating the neuroprotective effects of new treatment agents. Loss of consciousness was also found to be associated with increased loss of brain volume [48].

ADVANCED IMAGING

Perfusion imaging can be done with CT using intravenous contrast, it can be done with MRI in a variety of ways with or without contrast, and it can be done with single-photon emission CT using a variety of radiotracers. After the initial raw data is acquired from imaging, post-processing software can map out the perfusion characteristics of the cerebral blood flow, cerebral blood volume, and mean transit time. These studies are traditionally used for evaluating ischemia and brain tumors [30]. In a study by Wintermark et al., the performance of CT perfusion imaging on patients with severe head trauma at admission was shown to provide increased sensitivity for contusion over standard non-contrasted CT [49]. After TBI, patients have impaired cerebrovascular autoregulation, which can result in focal or global hyperperfusion or hypoperfusion, which manifest as differences in cerebral blood flow on perfusion images. Hyperperfusion at baseline is predictive of better clinical outcomes, and hypoperfusion is associated with poor outcomes [2, 49]. Additionally, the symptomatology of extra-axial hematomas has been found to correlate with focal perfusion abnormalities in the underlying brain parenchyma; this correlation could potentially be used to guide intervention [30, 50].

DTI is an MRI technique that measures diffusion anisotropy in white matter tracts and can be used to calculate the metric of fractional anisotropy. Fractional anisotropy is a marker of microstructural integrity and is decreased by disorders that reduce this integrity, including infarct, demyelination, and tumors, in addition to trauma [19, 30, 51]. A study of patients with mild TBI with persistent cognitive defects showed that amount of damage in white matter structures in DTI significantly correlated with mean reaction time on a cognitive test, whereas the number of microhemorrhages did not correlate [51]. DTI can also be performed to evaluate SCI and has been shown to detect changes in the spinal cord that appear normal on conventional MRI. There is also a correlation between motor function scores and DTI metrics in nonhemorrhagic SCI [21]. The biggest drawback to using DTI in trauma imaging is the lack of specificity for decreases in the fractional anisotropy due to the variability in different populations and because many processes can also decrease it. In research, premorbid imaging can be done to show change related to TBI, which is obviously not feasible in the clinical setting [19].

Another advanced imaging application that can be performed with MRI is magnetic resonance spectroscopy. This imaging study is conducted by first performing standard MRI so that a targeted area can be selected. Then a spectrum is obtained based on different properties of hydrogen atoms in different molecules [52]. A number of different molecular spectra are used as markers for different biochemical processes. N-acetylaspartate (NAA) is a marker for neurons, Choline is a marker for membrane synthesis and repair, lactate indicates anaerobic metabolism, and glutamate is an excitatory neurotransmitter. All of these markers have been noted to be altered in TBI [30]. A study by Cohen et al. of MR spectroscopy in patients with mild TBI showed that there was an average reduction of 12% of NAA in white matter [53]. This reduction was seen diffusely throughout the white matter, and the majority of patients in the study had no findings suggestive of TBI on conventional MRI. This reduction of NAA was even seen in patients who had no measurable cortical atrophy, implying an increased sensitivity of white matter to damage from TBI relative to gray matter. The study also noted the increased white matter NAA reductions in older patients, who have been shown to have worse clinical outcomes than younger patients after isolated TBI [53].

Functional MRI is another imaging modality that has been used in research studies without being a part of normal clinical work-up. This modality uses MRI signal changes from variations of blood oxygen level that correspond to changes in blood flow tied to neuronal activity. This imaging study can be done while the patient is in a resting state or while the patient performs repetitive tasks. Several studies have demonstrated abnormalities in task-related functional MRI studies in patients with a history of TBI whose findings on conventional MRI show no abnormalities [30].

REPORTING AND CLASSIFICATION

Over the years, multiple classification systems and methods to standardize the reporting of radiological findings in neurotrauma have been proposed and used. The common data elements were published in an attempt to provide standardization of terminology in reporting radiological findings of TBI; these definitions guided the description of pathological findings discussed previously [30].

For head trauma, the widely used Marshall and subsequent Rotterdam scoring systems were introduced in 1992 and 2006, respectively. These systems used a combination of CT findings, such as midline shift, cistern effacement, and extra-axial hemorrhage to evaluate trauma on head CT. These scores were focused on severe trauma and demonstrated good predictive values with respect to mortality [30, 54]. More recently, the Neuroimaging Radiological Interpretation System (NIRIS) was proposed. This system models itself after the gold standard of mammography, the Breast Imaging and Reporting Data System, which uses standardized reporting that dictates risk stratification and clinical management. The NIRIS scoring is based on standardized CT imaging findings (as defined in the Common Data Elements Neuroimaging Working Group paper from the National Institutes of Health in 2010 [30]). This scoring system was designed to be outcome-based and to show statistical significance between findings and management actions, in addition to clinical outcomes [54].

For assessing DAI, the Adams score was published in 1989 as a histopathological grading system based on the location of lesions [55]. Grade I lesions are seen in the cerebral hemispheres, predominantly at gray-white interfaces, Grade II lesions include the Grade I location with the additional involvement of the corpus callosum, and Grade III includes the Grade I and II locations with the addition of the dorsolateral or rostral brainstem [55]. This scoring system has been widely used and applied to MRI findings, but a more recent study by Abu Hamdeh et al. showed that the Adams score lacked correlation with outcome [37]. This study found that only the presence of lesions in the substantia nigra and mesencephalic tegmentum had a statistically significant correlation with poor prognosis. In addition, age older than 30 years was also found to be an independent risk factor for poor outcomes [37].

For spinal injuries, the majority of widely used radiological classifications concentrate on osseous and ligamentous injury instead of actual SCI, and these classifications are outside the scope of this article [21]. The extent of T2 signal abnormality and presence of hemorrhage within the spinal cord after trauma have both been shown to be negative prognostic indicators in the past. Recently, the Brain and Spinal Injury Center (BASIC) score was proposed, which classifies spinal cord injuries in the acute phase on axial T2 imaging according to the extent of T2 signal abnormality and the presence of macrohemorrhage. The scoring categories in ascending order of severity are no signal abnormality, central gray matter T2 hyperintensity, T2 hyperintensity of the central gray matter with partial white matter involvement, whole cord T2 hyperintensity, and whole cord hyperintensity with T2 hypointensities indicating macrohemorrhage. This scoring system was shown to have a good correlation with clinical prognosis [56].

Standardized reporting would be necessitated in future clinical trials of neuroprotective agents as a way to quantify imaging findings for research purposes. The NIRIS is based on CT findings and not used in serial imaging; however, it is outcomes-based, which could make it a good model for developing standardized imaging findings in serial MRI in clinical trials. The BASIC score could also represent a good starting point for SCIs, due to its correlation with clinical outcomes. Given that edema and inflammation are the proposed targets of the neurotherapeutics in the studies presented earlier, serial T2- and T2 FLAIR–weighted imaging could be used to assess this. However, it should be noted that edema and inflammation will have overlap in location as well as similar T2 hyperintense signal, limiting differentiation. Additionally, MRI would allow for the possibility of measurements and quantitative analysis.

Current reporting systems generally do not differentiate between white matter and gray matter trauma. Recent pathology studies demonstrate a profound increase in the severity and length of inflammatory changes in white matter injury relative to gray matter injury. These differences should be strongly considered in future attempts at standardized trauma imaging reporting, particularly in the investigation of neuroprotective agents [57, 58].

CONCLUSION

Conventional imaging of the head and spine with CT and MRI are essential modalities for evaluating neurotrauma and guiding management in patients with acute, subacute, and chronic injuries. New advanced imaging techniques have demonstrated improved detection and the subsequent improved understanding of neurotrauma in research, although many are not ready for common clinical use. Serial imaging with MRI is not currently indicated in clinical practice, and there is a relative paucity of research literature utilizing serial imaging in the acute and subacute phase of injury. Recent improvements in the understanding of neurotrauma pathophysiology have demonstrated a protracted destructive inflammatory response after insult. This has led to increased investigations into potentially neuroprotective anti-inflammatory treatments. Animal studies have suggested that short-term serial imaging using conventional MRI sequences, such as T2-weighted imaging, may be able to provide value in assessing the effectiveness of new pharmacological agents by assessing the evolution of a traumatic lesion as well as its associated edema and inflammation. Future clinical investigations for neuroprotective therapies will benefit from the widely available in vivo monitoring that neuroimaging offers.