Handbook of Clinical Neurology, Vol. 127 (3rd series) Traumatic Brain Injury, Part I J. Grafman and A.M. Salazar, Editors © 2015 Elsevier B.V. All rights reserved

Chapter 20

Advances in imaging explosive blast mild traumatic brain injury 1

H. HETHERINGTON1, A. BANDAK3, G. LING2, AND F.A. BANDAK2,3* Department of Neurosurgery, Yale University School of Medicine, New Haven, CT, USA

2

Department of Neurology, F. Edward Hébert School of Medicine, Uniformed Services University of the Health Sciences, Bethesda, MD, USA 3

Integrated Services Group Inc., Potomac, MD, USA

INTRODUCTION Shock wave pressures from explosive blast can have a serious effect on the brain without causing visible external injury to the head (Chs 6 and 11). Explosive blast mild traumatic brain injury (mTBI) may present concussive symptoms similar to those of conventional mTBI from blunt head impact; however, the physics and pathophysiology of the two are likely different (Ch. 6). One in five, or even up to one half of veterans of recent wars with a history of explosive blast exposure are reported to have experienced some form of dysfunction 1 year postinjury (Vasterling et al., 2009; Rosenfeld and Ford, 2010; Bass et al., 2012; Sayer, 2012). Memory impairment and in many cases, post-traumatic stress disorder (PTSD), are two common comorbidities of explosive blast mTBI. This relationship remains at best controversially debated (Hoge et al., 2008). Clinical assessment of explosive blast mTBI is difficult on several levels. First, accurate field data on the intensity of explosive blast exposure is usually not available, and when it is, it is difficult for the clinician to decipher. Second, conventional MRI cannot always detect physical evidence of injury. Therefore, clinical assessment of explosive blast mTBI relies largely on selfreporting of symptoms and cognitive testing which can either be inconclusive or compromised by other psychiatric and motivational factors. Methods to identify injury anatomically and link it with specific cognitive domains could significantly enhance clinical management. Although most anatomic imaging studies using MRI (T1- and T2-weighted imaging) have found minimal consistent changes in explosive blast mTBI, imaging

modalities focusing on structural (diffusion tensor imaging (DTI)), functional (functional magnetic resonance imaging (fMRI)), and metabolic (positron emission tomography (PET) and magnetic resonance spectroscopic imaging (MRSI)) aspects have reported significant differences between veterans with explosive blast mTBI and control subjects.

DIFFUSION TENSOR IMAGING DTI provides a measure of the directions of diffusion of molecules, mainly water, within the brain. An increase in overall diffusion (mean diffusivity (MD)) is typically consistent with increased water content (i.e., edema and inflammation) and thus relatively less resistance, and therefore, higher diffusion rates. Diffusion within the healthy brain is a highly anisotropic process, with substantially greater rates of diffusion parallel to and along white matter tracts (axial diffusivity, i.e., funneling of diffusion along the tracts), and substantially slower diffusion in the perpendicular direction (radial diffusivity, i.e., barriers to diffusion across tracts) (see Bandettini, 2009, for review). Disruption of the white matter structure can lead to changes in radial and axial diffusion (removal of the structures facilitating longitudinal diffusion and restricting axial diffusion), thus resulting in decreased fractional anisotropy (FA). In the acute setting, primary axonal pathology as well as inflammatory processes and edema can contribute to the overall rates of measured diffusion making interpretation challenging. In conventional TBI, DTI studies provide a reasonably sensitive measure of axonal pathology. In mTBI, within 2 weeks, either FA is decreased, MD is

*Correspondence to: Dr. Faris A. Bandak, Professor, Department of Neurology, A1036, F. Edward Hébert School of Medicine, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, Maryland 20814, USA. Tel: +1-301299-7357, E-mail: [email protected]

310

H. HETHERINGTON ET AL.

increased, or both (see Sharp and Ham, 2011, for review). DTI changes are often seen in large white matter tracts such as the corpus callosum (Aoki et al., 2012) but are also commonly seen throughout many brain regions (Kinnunen et al., 2011; Smits et al., 2011). In conventional mTBI, it has been reported that these changes are correlated with a variety of cognitive performance measures including processing speed (Levin et al., 2008), learning, and memory (Salmond et al., 2006; Kinnunen et al., 2011). Thus, there has been an expectation that DTI in explosive blast mTBI will provide similar sensitivity. MacDonald and coworkers reported an increased number of DTI abnormalities in 18 of 63 veterans with explosive blast mTBI (MacDonald et al., 2011) with consequent secondary head impact in comparison to 21 blast-exposed veterans without a diagnosis of mTBI. The subjects were studied twice, once within 90 days of the exposure and again 6–12 months later. Abnormalities were identified based on the criterion of reduction in FA by two standard deviations (2 SD), in comparison to controls in 12 anatomically defined brain regions within major white matter tracts. In the 63 subjects with mTBI, 18 showed significant decreases in FA in two or more brain regions, satisfying the criteria for multifocal brain injury. In the other 45 subjects, one defect was detected in 20 subjects, while 25 subjects demonstrated no significant abnormalities. The most common sites of these abnormalities were the cingulum, uncinate fasciculus, anterior limb of the internal capsule, cerebellar peduncles, and orbitofrontal white matter. The 47 subject subgroup studied 6–12 months postexposure showed decreased FA consistent with diminishing progressive loss of white matter structure and normalized axial and mean diffusivity consistent with resolution of

edema/inflammation. This led the researchers to view these findings as consistent with a model of progressive injury. Using an alternative method of data acquisition and analysis which highlights the quantification of primary and secondary crossing fibers within an imaging pixel along with tract based coregistration, Morey and colleagues (2013) reported significant differences between controls and veterans with explosive blast mTBI. They found that the volume fraction of primary fibers was decreased in a variety of brain regions including portions of the corpus callosum, corona radiata, forceps major and forceps minor, internal capsule, thalamic radiations, superior fasciculus, and tapetum (Fig. 20.1). The volume fraction of primary fibers in a variety of regions including the inferior frontal fasciculus, uncinate fasciculus, corpus callosum, anterior corona radiata, external capsule, inferior longitudinal fasciculus, cingulum, portions of the internal capsule, and sagittal stratum was significantly correlated with the duration of loss of consciousness. When compared simply to a state of altered consciousness (feeling dazed and confused), even more brain regions were correlated. The authors argue that the extent of injury supports a model of primary blastinduced neurotrauma where biochemical and immune system-based changes mediate long-term downstream changes. These positive findings of altered FA and MD have not been replicated in all studies of mTBI. For example, Levin et al. (2010) reported no difference in FA or apparent diffusion coefficient (ADC) between a group of 37 service members with a history of mild to moderate TBI due to blast in comparison to 15 control service members. Further, no significant correlation was seen between symptom measures and DTI variables.

Fig. 20.1. The red and yellow highlighted regions indicate significant difference ( p < 0.05) in the partial volume of primary fibers for the mild traumatic brain injury (mTBI) and control groups. (Reproduced from Morey et al., 2013.)

ADVANCES IN IMAGING EXPLOSIVE BLAST MILD TRAUMATIC BRAIN INJURY 311 Although the studies of MacDonald et al. (2011) and in PTSD including decreased FA in the anterior cingulate Morey et al. (2013) report significant and widespread cortex (ACC) (Schuff et al., 2011) and left anterior cinchanges in white matter structure providing evidence gulum bundles (Kim et al., 2006). However, these findof structural changes, they did not link the structural ings of significant decreases in FA have not been changes to the extent and severity of clinical symptoms replicated in all DTI studies of PTSD (see Ayling or functional deficits nor were they able to distinguish et al., 2012, for review). Moreover, a study by Morey pure blast effects as opposed to explosive blast with conet al. (2013) failed to show a significant association comitant head impact. Levin’s work specifically identibetween PTSD and DTI-based changes, and unlike prefied a lack of correlation with clinical symptoms. Thus vious reports of DTI changes in PTSD, the extent of it raises the question that while DTI findings are indicabrain involvement was substantially larger in blast tors of a history of mTBI, they may not be predictive of mTBI. Thus some investigators have suggested that blast the extent of functional damage. To assess the impact of mTBI and white matter damage may increase the suscepDTI on brain activity, Huang et al. (2009) investigated tibility of developing PTSD and depression. the relationship between decreased FA and slowing as measured by magnetoencephalography (MEG). FUNCTIONAL IMAGING Although the group studied consisted of a mixture of Although the DTI studies can identify damage to white both conventional (n ¼ 6) and explosive blast mTBI (n ¼ 4) patients, the investigators demonstrated that cormatter tracts, which is not seen on conventional MRI, the tical brain regions with regions giving rise to slowing presence of that damage may not necessarily manifest or were linked by white matter tracts with decreased FA. result in clinical symptoms or cognitive deficits. Sponheim et al. (2011) performed a conceptually similar Although detailed neuropsychological evaluation and study using scalp electrodes and DTI in nine soldiers patient histories can be used in the assessment of injury, with explosive blast mTBI and eight controls. In this their inherent subjective nature and potential biases due to other economic considerations can lead to challenges case, decreased FA in the forceps minor, and left antein clinical management. Thus imaging methods that can rior thalamic radiations were correlated with decreased phase synchrony between electrodes overlying brain provide a more direct and potentially more objective regions connected by these tracts. It is worth noting that measure of dysfunction (such as MEG) can be useful. no such relationships were observed in the control subFunctional MRI has revolutionized the study of human jects. Despite the small sample sizes, these studies report brain function over the past decade by providing maps of a link between white matter injury and a specific change brain activation in response to different stimuli in both in brain function at a detailed anatomic level. Further healthy controls and patients. Most conventional fMRI studies utilize BOLD (blood oxygen level-dependent) work in larger groups will be important to establish this imaging (Ogawa et al., 1993; Norris, 2006). BOLD imagcritical causal link between the structural changes to white mater seen in DTI and ongoing deficits in funcing is based on the physiologic principle that during brain tion. Although Levin’s work failed to identify significant activity, blood flow to the active brain regions and delivgroup differences in DTI between the control group and ery of oxygen outstrips the metabolic demand, causing blast mTBI, the study did demonstrate a statistically sigoxygen levels to rise in the venous system and increases nificant correlation between word recall and verbal in the content of oxygenated hemoglobin. The change in memory and FA and ADC in the internal capsule, left oxygenated hemoglobin alters the MRI characteristics of water, causing a change in signal level. By comparing corticospinal tract, and uncinate fasciculi. Due to the the MRI signal intensity during a task designed to utilize large variability in the severity and conditions including other blunt injuries suffered at the time of blast expoa specific brain region (e.g., recalling a series of digits) sure, it may not be surprising that group comparisons and its control task, the regions of activated brain can be may fail, even in the presence of clear structural identified (Fig. 20.2). Although a variety of related changes. Such approaches, which focus on linking a acquisition methods can give greater specificity as to functional deficit with a specific anatomic abnormality, the region of activation (Duong et al., 2001; Yacoub may be most productive in aiding clinical evaluation and et al., 2001, 2003), the vast majority of studies are performed using BOLD and gradient echo-based echo plamanagement. nar imaging, which primarily reflects contributions from Finally, as a caveat to this work, many patients with blast mTBI also suffer from a variety of psychiatric disdraining veins. orders including PTSD and depression. Just as many of In addition to identifying which brain regions respond the reported symptoms in mTBI due to blast are also to specific tasks, fMRI data have also been interpreted reported in PTSD (Stein and McAllister, 2009; Hicks to identify areas which may be compromised functionet al., 2010), changes in DTI have also been reported ally. For example, in conventional mTBI, increased

312

H. HETHERINGTON ET AL.

Fig. 20.2. Increased bilateral amygdala activation to fear in MDD (major depressive disorder) versus non-MDD. A region of interest (ROI) functional MRI (fMRI) analysis revealed significantly greater fear-related activation in bilateral amygdalae of blast mild TBI patients with MDD versus non-MDD subjects (left panel). Mean activation for angry, happy, fear and shape matching trials extracted from the amygdalae ROIs is displayed in the right panel. (Reproduced from Matthews et al., 2011.)

levels of activation and increased areas of response associated with working memory have been reported in response to memory tasks (Smits et al., 2009). However, it is important to note that if blood flow becomes limited and cannot sufficiently increase to meet demand, or if blood volume increases substantially, the BOLD signal will decrease, disappear, or reverse in sign despite elevated brain activity (Kamba et al., 2007; Bianciardi et al., 2011). Finally, decreasing or missing responses will mimic the response seen in normal regions with minimal or no involvement in the task. Thus fMRI data must be carefully interpreted, considering the magnitude of the task and the cerebrovascular status of the subject. In a study of 22 subjects with explosive blast mTBI, one half with major depression and the other without, Matthews et al. (2011) reported significantly greater activation in the amygdala, a site known to be involved in processing of negative emotions. He also reported significantly less activation in the dorsolateral prefrontal cortex (DLPFC), middle frontal gyrus, and subcortical brain regions. Additionally, FA decreases in the superior longitudinal fasciculus, which is known to connect the DLPFC (an emotional control center) to many brain regions, were correlated with greater depressive symptom severity. In this case, although not all mTBI subjects developed major depressive disorder (MDD), those who did demonstrate a pattern of white matter injury (DTI data) consistent with brain regions showing altered function (fMRI data) which are involved in the pathology being expressed (MDD). The authors speculate that the specific damage and its anatomic localization resulting from mTBI may predispose this group to develop MDD. In 15 explosive blast mTBI subjects and 15 control subjects, using a stimulus response task in which the

subject is required to discriminate between red and blue arrows and their orientation, Scheibel et al. (2012) reported greater activation in the anterior cingulate gyrus, medial frontal cortex, and posterior cerebral areas involved in visual and visual-spatial functions in TBI subjects. After correcting for the effects of decreased reaction time, PTSD and depression; the extent of activation increased. Further, there was a correlation between PTSD symptoms and activation within posterior brain regions. Although in relatively small groups, these studies demonstrate the potential for bridging the gap between identifying a structural deficit seen in DTI with an objective change in brain function, which is anatomically linked with a specific symptom or behavior.

METABOLIC IMAGING Positron emission tomography Positron emission tomography (PET) studies use radioactive tracers to image the uptake and retention of the labeled substrates or their metabolites. The most common tracer for assessing brain metabolism is the radioisotope [18]F-labeled fluorodeoxyglucose (FDG). This compound mimics the structure of glucose, which is the primary fuel source for brain oxidative metabolism. After the compound is taken up in the brain, its structure prevents its further metabolism (lack of one hydroxyl group in comparison to glucose) and it accumulates within the cell. Thus the presence of the radioactive signal within the brain is used as a measure of glucose uptake. Since glucose is the primary substrate for brain metabolism and glucose uptake is driven by metabolic/ functional demand, reduced FDG uptake in the absence

ADVANCES IN IMAGING EXPLOSIVE BLAST MILD TRAUMATIC BRAIN INJURY of volume loss is typically interpreted as a decrease in brain activity and thus impaired function. In 12 veterans with mTBI, Peskind et al. (2010) reported that, in comparison to 12 controls, the mTBI group displayed reductions in FDG uptake in the cerebellum, vermis, pons, and medial temporal lobe. These findings of cerebellar injury/dysfunction were consistent with cognitive deficits seen in this group, including decreases in verbal fluency, cognitive processing speed, attention, and working memory.

Magnetic resonance spectroscopy imaging Unlike most MRI methods, which are concerned with the content and properties of the tissue water, magnetic resonance spectroscopy imaging (MRSI) studies focus on the measurement of millimolar concentrations of small molecular weight molecules. These molecules typically include, N-acetylaspartate (NAA), a compound found only in neurons (Urenjak et al., 1993), choline (Ch), a trimethylamine which is associated with membrane damage and repair (Brooks et al., 2001), and creatine (Cr), reflecting both phosphorylated and unphosphorylated forms of phosphocreatine, the primary buffer for adenosine triphosphate (ATP)-requiring processes. Although there are many other compounds that can be measured by MRSI, including glutamate, glutamine, g-aminobutyric acid (GABA), glutathione, glucose, lactate, and taurine, these compounds are more difficult to measure accurately and have yet to establish their broad utility in studies of TBI. Due to the localization of its synthesis to neuronal mitochondria, measurements of NAA have proven highly useful in assessing a wide variety of neurologic and psychiatric disorders (see Moffett et al., 2007, for review). Reductions in NAA are typically interpreted as reflecting neuronal injury. In epilepsy, NAA levels return to normal when the frequency of seizures is reduced by treatment (Hugg et al., 1996; Cendes et al., 1997; Serles et al., 2001; Vermathen et al., 2002). In contrast, choline increases are routinely seen in diseases that result in axonal damage, such as the demyelinating phases of multiple sclerosis (De Stefano and Filippi, 2007). In mTBI, MRS measurements of decreased NAA levels have been reported in MRI normal brain regions (Govindaraju et al., 2004) and have been used to predict long-term cognitive outcome in children (Babikian et al., 2006) and adults (Holshouser et al., 2006). In athletes who have experienced a sport-related concussive head injury (with or without loss of consciousness), reductions in NAA have been reported in frontal white matter which persisted beyond the resolution of clinical symptoms, and whose recovery was significantly slowed by exposure to a second concussive injury (Vagnozzi et al.,

313

2008, 2010). Finally, recent work from Henry et al. (2011) has demonstrated chronic changes in football players. Thus, despite minimal changes on conventional MRI, mTBI, even in the absence of loss of consciousness, can result in MRS-detectable changes, both acutely and chronically. Based on some of the findings mentioned above, studies to evaluate NAA/Ch and NAA/Cr ratios in warfighters with reported symptomatology associated with exposure to blast seems prudent. Given the high frequency of complaints of memory impairments, the localization of memory to the hippocampus and temporal lobe, and previous studies correlating NAA changes with memory impairment (Pan et al., 2009), Hetherington et al. (2013) have developed techniques to carry out MRSI measurements of the human hippocampus of warfighters. MRSI studies of the human hippocampus are challenging due to the strong distortions in B0 field homogeneity induced by the ear canals and frontal sinuses. The distortions in B0 field homogeneity lead to broader lines in the MR spectra and increased spectral overlap, making accurate quantification difficult. Although the signal-to-noise ratio increases linearly with field strength, the performance of conventional volume coil detectors decreases dramatically at 7 T, resulting in poor homogeneity (especially from more inferior brain structures such as the temporal lobe) and increased power deposition, limiting the use of conventional MRSI sequences. To overcome these two limitations we used specialized hardware including: (1) a specialized gradient set with 3rd degree B0 shimming (Pan et al., 2012) and (2) an eight element inductively decoupled transceiver array with RF shimming (Hetherington et al., 2010). B0 shimming is the process by which distortions in the homogeneity of the static magnetic field (B0) are minimized by generating fields to oppose the distortions. Shim coils are used to generate these fields, with their spatial dependence modeled by spherical harmonics. Virtually all clinical MRI systems are equipped with 1st and 2nd degree shims, which allow only corrections for distortions that display a spatial dependence of up to the 2nd power of their distance from the center of the magnet. Unfortunately, in the human brain, dependencies of up to the 7th power of the distance can be seen (Pan et al., 2012) resulting in poor quality spectra acquired from a subject at 3 T using 1st and 2nd degree shimming. Such images are displayed in Figure 20.3A. Although the spectral quality is sufficient for analysis from the posterior portions of the hippocampus, it degrades rapidly anteriorly with the anterior rows of the hippocampus being uninterpretable. Addition of 3rd degree shims, however, results in a 38% improvement in homogeneity (Pan et al., 2012), enabling the acquisition of good quality spectra even

314

H. HETHERINGTON ET AL.

Fig. 20.3. (A) Scout images for a normal subject, showing region of spectra (white box) and spectra. The data was acquired at 3 T using 1st and 2nd degree shims. (B) Scout image and spectra acquired at 7 T in the same subject using 1st to 3rd degree shims.

Fig. 20.4. (A) Maps of the radiofrequency (RF) excitation field generated when each coil in the array is driven individually with the same amplitude. (B) Map of the excitation field when all coils are driven simultaneously with the same amplitude. (C) Map of the excitation field when the amplitudes are adjusted to maximize the homogeneity of the RF excitation field. (D) The “ring” distribution which excites only the periphery of the head. From Hetherington et al. (2013).

from the most anterior portions of the hippocampus (Fig. 20.3B). The data displayed in Figure 20.3B were acquired from the same subjects as Figure 20.3A, but at 7 T where the effects of the field distortions are 2.3 times greater. At most clinical field strengths 1.5–3 T, conventional volume coils typically provide a homogeneous radio frequency (RF) excitation field, which varies by 15% or less. However, at 7 T, the interactions between the RF field and the head result in dramatic decreases in homogeneity reaching up to 50% (Vaughan et al., 2001) and decreases in efficiency (strength of the excitation field per watt of applied power) thereby dramatically increasing the deposited power by up to a factor of five. To overcome these limitations a transmission array using eight independent coils for both transmission and reception was developed (Avdievich et al., 2009).

The maps of the strength of the RF excitation field when using equivalent drive amplitudes to each of the RF coils are displayed in Figure 20.4A. For the temporal lobe, although the coils are driven with the same power, the excitation field is much smaller for the anterior coils. When these equal amplitudes are applied simultaneously (Figure 20.4B), the excitation field is very inhomogeneous. This mimics the response of conventional volume coils. By adjusting the amplitude and phase of the RF pulses applied to each of the eight coils (“RF shimming”), the homogeneity was improved achieving a standard deviation of 11% (Fig. 20.4C). To limit power deposition, we used a different set of RF amplitudes and phases to generate a second RF distribution (“ring distribution”), which excites only the periphery of the head (Fig. 20.4D). This distribution is used to eliminate the signal from the periphery while not perturbing the

ADVANCES IN IMAGING EXPLOSIVE BLAST MILD TRAUMATIC BRAIN INJURY signal from interior brain regions and suppressing unwanted lipid resonances from the skin and muscle. It uses a factor of four less power than the homogeneous distribution thereby decreasing the power deposition and avoiding unwanted tissue heating. Adopting these improvements for MRSI imaging of the hippocampus, de Lanerolle et al. (2014) conducted a study of 25 warfighters (36  10 years) with explosive blast mTBI and self-reported persistent memory impairments and 25 control subjects (31 + 10 years). Of the 25 warfighters, three subjects had only one exposure, eight had only blast exposures, and 14 had blunt trauma and blast exposure. Our study utilized a single-slice acquisition sequence (Hetherington et al., 2010) using a 10 mm thick slice angulated along the temporal pole and images were acquired with 24  24 phase encoding steps over a FOV of 192  192 mm providing 1 cc resolution after postacquisition spatial filtering. A repetition time of 1.5 seconds was used yielding an acquisition time of 14.4 min. An echo time of 40 ms was used to minimize spectral overlap of NAA, Cr, and Ch with amino acids and macromolecule resonances (Behar and Ogino, 1993). To improve the accuracy of their measurement by correcting for naturally occurring variations in metabolite content along the hippocampal formation, image-guided single voxel reconstructions (Hetherington et al., 2007) were used to provide

315

reproducible sampling across subjects for the hippocampus. The hippocampal formations are manually outlined (red and green regions of interest (ROIs)), and a midline (yellow) is automatically calculated. The aqueduct is then manually identified (intersection of white lines) and six voxels (yellow circles), three anterior and two posterior to the aqueduct, are reconstructed at 9 mm intervals along the midline (Fig. 20.5). Our findings indicate significant reductions in the ratio of NAA/Ch (single-tailed T-test) in most of the loci of the right hippocampus ( p < 0.05 in loci 2–6 respectively from Fig. 20.5), while the three most anterior loci of the same hippocampus also showed significant reductions in NAA/Cr ( p ¼ 0.022, 0.016 and 0.001 loci 4–6 from Fig. 20.5). The only significant decreases seen from the left hippocampus was NAA/Ch from loci #5 and #6 ( p ¼ 0.014, p ¼ 0.037). We also demonstrated additional findings in two veterans with decrements in NAA/Ch amongst the largest (>2 SD from control), who did not have PTSD and did fail their effort testing. Although the study is small in size and needs to be expanded, these data represent one of the first studies to identify consistent significant metabolic changes reflective of injury in explosive blast mTBI. Further, the finding of significant metabolic alterations in the hippocampus is consistent with the self-reported deficits in memory, which are seen throughout the larger population. Finally, the ability to

Fig. 20.5. (A, B) Data from a healthy control. The spectra are reconstructed from the spatial locations identified by the yellow circles. (C–E) Data from a veteran with blast mTBI. The spectrum in (E) is from the most anterior location from the right hippocampus (image left).

316

H. HETHERINGTON ET AL.

detect injury and validate ongoing impairment in subjects may enable better clinical management in this group. While damage to the hippocampus may play a role in the relationship between explosive blast mTBI and PTSD, it has been proposed that symptoms with associated explosive blast mTBI may be largely attributable to PTSD (Hoge et al., 2008). However, our study did not find any significant differences between mTBI subjects with and without a diagnosis of PTSD (de Lanerolle et al., 2014). Additionally, decreased NAA/Ch was found in all of the subjects who failed effort testing (Hetherington et al., 2013). Given the multitude of psychological processes that can lead to poor performance on these measures, though, it is uncertain whether or not this result is directly attributable to the processes giving rise to the reduced NAA/Ch. Regardless, our study showed that the presence of a large reduction in NAA/Ch is consistent with neuronal injury in this brain region.

SUMMARY Although most conventional imaging studies of explosive blast mTBI have failed to identify significant differences in comparison to controls and link to the anatomic origin of the ongoing persistent deficits, more advanced imaging methods can now detect these injuries. Amongst the most promising imaging methods are DTI, fMRI, and MRSI. Each of these targets different aspects of the pathology involved in mTBI. DTI provides a highly sensitive measure to detect primary changes in the microstructure of white matter tracts. fMRI enables the measurement of changes in brain activity in response to different stimuli or tasks. Although fMRI measurements are indirect and require that the normal coupling between blood flow, metabolism, and electrical activity in the brain remain largely intact, for more subtle injuries, this paradigm is likely to remain valid. Our MRSI imaging studies, although acquired at much lower spatial resolution (1 mL volumes as opposed to < 10 mL for DTI and fMRI), their selectivity to different metabolic and physiologic processes have demonstrated some of the most profound differences on an individual-by-individual basis, suggesting the potential for utility in the management of individual patients. These studies, however, were conducted at 7 T, using advanced hardware and methods, not routinely available on most clinical systems at 3 T. It is becoming quite feasible to adapt these methods, with some limitations, to 3 T systems. Remarkably, all three of these paradigms have found significant success in conventional mTBI where conventional clinical imaging frequently fails to provide definitive differences. It is highly attractive to consider combining the MRI methods presented here into a single

comprehensive imaging sequence to study explosive blast mTBI as well as conventional mTBI. Finally although PET is less frequently used for conventional mTBI, and there are few reports in explosive blast mTBI, its potential to characterize a variety of neurotransmitter systems using target agents will undoubtedly play a larger role, once the basic mechanisms of injury are better understood and techniques to identify the injury are more common.

DISCLAIMER The views and opinions expressed herein are solely those of the authors. They do not represent, should not be interpreted as, and/or do not imply endorsement by the Uniformed Services University of the Health Sciences, Department of Defense, or the US government.

REFERENCES Aoki Y, Inokuchi R, Gunshin M et al. (2012). Diffusion tensor imaging studies of mild traumatic brain injury: a metaanalysis. J Neurol Neurosurg Psychiatry 83: 870–876. Avdievich NI, Pan JW, Baehring JM et al. (2009). Short echo spectroscopic imaging of the human brain at 7 T using transceiver arrays. Magn Reson Med 62: 17–25. Ayling E, Aghajani M, Fouche JP et al. (2012). Diffusion tensor imaging in anxiety disorders. Curr Psychiatry Rep 14: 197–202. Babikian T, Freier MC, Ashwal S et al. (2006). MR spectroscopy: predicting long-term neuropsychological outcome following pediatric TBI. J Magn Reson Imaging 24: 801–811. Bandettini PA (2009). What’s new in neuroimaging methods? Ann N Y Acad Sci 1156: 260–293. Bass CR, Panzer MB, Rafaels KA et al. (2012). Brain injuries from blast. Ann Biomed Eng 40: 185–202. Behar KL, Ogino T (1993). Characterization of macromolecule resonances in the 1H NMR spectrum of rat brain. Magn Reson Med 30: 38–44. Bianciardi M, Fukunaga M, van Gelderen P et al. (2011). Negative BOLD-fMRI signals in large cerebral veins. J Cereb Blood Flow Metab 31: 401–412. Brooks WM, Friedman SD, Gasparovic C (2001). Magnetic resonance spectroscopy in traumatic brain injury. J Head Trauma Rehabil 16: 149–164. Cendes F, Andermann F, Dubeau F et al. (1997). Normalization of neuronal metabolic dysfunction after surgery for temporal lobe epilepsy. Evidence from proton MR spectroscopic imaging. Neurology 49: 1525–1533. de Lanerolle NC, Hamid H, Kulas J et al. (2014). Concussive brain injury from explosive blast. Ann Clin Transl Neurol 1 (9): 692–702. De Stefano N, Filippi M (2007). MR spectroscopy in multiple sclerosis. J Neuroimaging 17 (Suppl 1): S31–S35. Duong TQ, Kim DS, Ugurbil K et al. (2001). Localized cerebral blood flow response at submillimeter columnar resolution. Proc Natl Acad Sci U S A 98: 10904–10909.

ADVANCES IN IMAGING EXPLOSIVE BLAST MILD TRAUMATIC BRAIN INJURY Govindaraju V, Gauger GE, Manley GT et al. (2004). Volumetric proton spectroscopic imaging of mild traumatic brain injury. AJNR Am J Neuroradiol 25: 730–737. Henry LC, Tremblay S, Leclerc S et al. (2011). Metabolic changes in concussed American football players during acute and chronic post-injury phases. BMC Neurol 11: 105. Hetherington HP, Kuzniecky RI, Vives K (2007). A subcortical network of dysfunction in TLE measured by magnetic resonance spectroscopy. Neurology 69: 2256–2265. Hetherington HP, Avdievich NI, Kuznetsov AM et al. (2010). RF Shimming for spectroscopic localization in the human brain at 7 T. Magn Reson Med 63: 9–19. Hetherington H, Hamid H, Kulas J et al. (2013). MRSI of the medial temporal lobe at 7 T in explosive blast mild traumatic brain injury. Magn Reson Med 71: 1358–1367. Hicks RR, Fertig SJ, Desrocher RE et al. (2010). Neurological effects of blast injury. J Trauma 68: 1257–1263. Hoge CW, McGurk D, L TJ et al. (2008). Mild traumatic brain injury in U.S. soldiers returning from Iraq. N Engl J Med 358: 453–463. Holshouser BA, Tong KA, Ashwal S (2006). Prospective longitudinal proton magnetic resonance spectroscopic imaging in adult traumatic brain injury. J Magn Reson Imaging 24: 33–40. Huang MX, Theilmann RJ, Robb A et al. (2009). Integrated imaging approach with MEG and DTI to detect mild traumatic brain injury in military and civilian patients. J Neurotrauma 26: 1213–1226. Hugg JW, Kuzniecky RI, Gilliam FG et al. (1996). Normalization of contralateral metabolic function following temporal lobectomy demonstrated by 1H magnetic resonance spectroscopic imaging. Ann Neurol 40: 236–239. Kamba M, Sung YW, Ogawa S (2007). Alteration of blood oxygenation level-dependent signaling by local circulatory condition. J Magn Reson Imaging 26: 1506–1513. Kim SJ, Jeong DU, Sim ME et al. (2006). Asymmetrically altered integrity of cingulum bundle in post-traumatic stress disorder. Neuropsychobiology 54: 120–125. Kinnunen KM, Greenwood R, Powell JH et al. (2011). White matter damage and cognitive impairment after traumatic brain injury. Brain 134: 449–463. Levin HS, Wilde E, Chu Z et al. (2008). Diffusion tensor imaging in relation to cognitive and functional outcome of traumatic brain injury in children. J Head Trauma Rehabil 23: 197–208. Levin HS, Wilde E, Troyanskaya M (2010). Diffusion tensor imaging of mild to moderate blast-related traumatic brain injury and its sequelae. J Neurotrauma 27: 683–694. MacDonald CL, Johnson AM, Cooper D (2011). Detection of blast-related traumatic brain injury in U.S. military personnel. N Engl J Med 364: 2091–2100. Matthews SC, Strigo IA, Simmons AN et al. (2011). A multimodal imaging study in U.S. veterans of Operation Iraqi and Enduring Freedom with and without major depression after blast-related concussion. Neuroimage 54 (Suppl 1): S69–S75.

317

Moffett JR, Ross B, Arun P et al. (2007). N-Acetylaspartate in the CNS: from neurodiagnostics to neurobiology. Prog Neurobiol 81: 89–131. Morey RA, Haswell CC, Selgrade ES (2013). Effects of chronic mild traumatic brain injury on white matter integrity in Iraq and Afghanistan war veterans. Hum Brain Mapp 34: 2986–2999. Norris NG (2006). Principles of magnetic resonance assessment of brain function. J Magn Reson Imaging 23: 794–807. Ogawa S, Menon RS, Tank DW et al. (1993). Functional brain mapping by blood oxygenation level-dependent contrast magnetic resonance imaging. A comparison of signal characteristics with a biophysical model. Biophys J 64: 803–812. Pan JW, Cavus I, Zaveri HP et al. (2009). Physiologic correlates of metabolic imaging in human epilepsy. In: PA Schwartzkroin (Ed.), Encyclopedia of Basic Epilepsy Research, Academic Press, Oxford, pp. 1568–1573. Pan JW, Lo KM, Hetherington HP (2012). Role of very high order and degree B0 shimming for spectroscopic imaging of the human brain at 7 tesla. Magn Reson Med 68: 1007–1017. Peskind ER, Petrie EC, Cross DJ (2010). Cerebrocerebellar hypometabolism associated with repetitive blast exposure mild traumatic brain injury in 12 Iraq war veterans with persistent post-concussive symptoms. Neuroimage 27: 683–694. Rosenfeld JV, Ford NL (2010). Bomb blast, mild traumatic brain injury and psychiatric morbidity: a review. Injury 41: 437–443. Salmond CH, Menon DK, Chatfield DA et al. (2006). Diffusion tensor imaging in chronic head injury survivors: correlations with learning and memory indices. Neuroimage 29: 117–124. Sayer NA (2012). Traumatic brain injury and its neuropsychiatric sequelae in war veterans. Annu Rev Med 63: 405–419. Scheibel RS, Newsome MR, Troyanskaya M et al. (2012). Altered brain activation in military personnel with one or more traumatic brain injuries following blast. J Int Neuropsychol Soc 18: 89–100. Schuff N, Zhang Y, Zhan W et al. (2011). Patterns of altered cortical perfusion and diminished subcortical integrity in post-traumatic stress disorder: an MRI study. Neuroimage 54 (Suppl 1): S62–S68. Serles W, Li LM, Antel SB et al. (2001). Time course of postoperative recovery of N-acetyl-aspartate in temporal lobe epilepsy. Epilepsia 42: 190–197. Sharp DJ, Ham TE (2011). Investigating white matter injury after mild traumatic brain injury. Curr Opin Neurol 24: 558–563. Smits M, Dippel DW, Houston GC et al. (2009). Postconcussion syndrome after minor head injury: brain activation of working memory and attention. Hum Brain Mapp 30: 2789–2803. Smits M, Houston GC, Dippel DW et al. (2011). Microstructural brain injury in post-concussion syndrome after minor head injury. Neuroradiology 53: 553–563. Sponheim SR, McGuire KA, Kang SS (2011). Evidence of disrupted functional connectivity in the brain after combatrelated blast injury. Neuroimage 54 (Suppl 1): S21–S29.

318

H. HETHERINGTON ET AL.

Stein MB, McAllister TW (2009). Exploring the convergence of posttraumatic stress disorder and mild traumatic brain injury. Am J Psychiatry 166: 768–776. Urenjak J, Williams SR, Gadin DG et al. (1993). Proton nuclear magnetic resonance spectroscopy unambiguously identifies different neural cell types. J Neurosci 13: 981–989. Vagnozzi R, Signoretti S, Tavazzi B et al. (2008). Temporal window of metabolic brain vulnerability to concussion: a pilot 1H-magnetic resonance spectroscopic study in concussed athletes – part III. Neurosurgery 62: 1286–1295. Vagnozzi R, Signoretti S, Cristofori L et al. (2010). Assessment of metabolic brain damage and recovery following mild traumatic brain injury: a multicenter, proton magnetic resonance spectroscopic study in concussed patients. Brain 133: 3232–3242.

Vasterling JJ, Verfaellie M, Sullivan KD (2009). Mild traumatic brain injury and posttraumatic stress disorder in returning veterans: perspectives from cognitive neuroscience. Clin Psychol Rev 29: 674–684. Vaughan JT, Garwood M, Collins CM et al. (2001). 7 T vs. 4 T: RF power, homogeneity, and signal-to-noise comparison in head images. Magn Reson Med 63: 24–30. Vermathen P, Ende G, Laxer KD et al. (2002). Temporal lobectomy for epilepsy: recovery of the contralateral hippocampus measures by (1)H MRS. Neurology 59: 633–636. Yacoub E, Shmuel A, Pfeuffer J et al. (2001). Investigation of the initial dip in fMRI at 7 tesla. NMR Biomed 14: 408–412. Yacoub E, Duong TQ, Ven De Moortele PF et al. (2003). Spinecho fMRI in humans using high spatial resolutions and high magnetic fields. Magn Reson Med 49: 655–664.