|Home | About | Journals | Submit | Contact Us | Français|
Traumatic brain injury (TBI) caused by explosive munitions, known as blast TBI, is the signature injury in recent military conflicts in Iraq and Afghanistan. Diagnostic evaluation of TBI, including blast TBI, is based on clinical history, symptoms, and neuropsychological testing, all of which can result in misdiagnosis or underdiagnosis of this condition, particularly in the case of TBI of mild-to-moderate severity. Prognosis is currently determined by TBI severity, recurrence, and type of pathology, and also may be influenced by promptness of clinical intervention when more effective treatments become available. An important task is prevention of repetitive TBI, particularly when the patient is still symptomatic. For these reasons, the establishment of quantitative biological markers can serve to improve diagnosis and preventative or therapeutic management. In this study, we used a shock-tube model of blast TBI to determine whether manganese-enhanced magnetic resonance imaging (MEMRI) can serve as a tool to accurately and quantitatively diagnose mild-to-moderate blast TBI. Mice were subjected to a 30 psig blast and administered a single dose of MnCl2 intraperitoneally. Longitudinal T1-magnetic resonance imaging (MRI) performed at 6, 24, 48, and 72h and at 14 and 28 days revealed a marked signal enhancement in the brain of mice exposed to blast, compared with sham controls, at nearly all time-points. Interestingly, when mice were protected with a polycarbonate body shield during blast exposure, the marked increase in contrast was prevented. We conclude that manganese uptake can serve as a quantitative biomarker for TBI and that MEMRI is a minimally-invasive quantitative approach that can aid in the accurate diagnosis and management of blast TBI. In addition, the prevention of the increased uptake of manganese by body protection strongly suggests that the exposure of an individual to blast risk could benefit from the design of improved body armor.
Traumatic brain injury (TBI) from explosive munitions, also known as blast TBI, is the signature injury in the Iraq and Afghanistan war theaters.1,2 Most cases of blast TBI are mild in severity and present with clinical symptoms indistinguishable from those of blunt concussion. Blast TBI often occurs in a repetitive pattern, and this pattern may be the cause of chronic neuropsychiatric conditions in veterans of these conflicts, just as in the case of athletes with histories of repeat concussions from contact and collision sports. In some cases, such repetitive TBI exposure may lead to chronic traumatic encephalopathy, a degenerative tauopathy.3,4 Prevention of chronic symptomatologies and neurodegenerative disease after concussion is based on the avoidance of repeat exposure to TBI-producing events until patients become asymptomatic. This principle is at the root of several versions of return-to-play or combat guidelines issued by international conferences, clinical professional associations, the Centers for Disease Control and Prevention, and the U.S. military (www.dvbic.org). Although such guidelines are based on established clinical practice and are generally designed to be on the “safe side” of decision-making, it is impossible to predict, for each individual soldier or athlete, the long-term outcome of re-exposure to TBI.
It has been argued that the best way to base decisions regarding redeployment or return-to-play is to follow relevant quantifiable post-TBI changes based on brain imaging or non-imaging biomarkers.5–8 Such markers could help determine at which point re-exposure to TBI risk is acceptable. Further, because of the acute and chronic neuropathological consequences of TBI, early detection also may assist in prompt and thus more effective interventions.
Magnetic resonance imaging (MRI) is a particularly appealing imaging modality because it is non-invasive and causes only minimal distress to the patient. Unfortunately, blast TBI, especially of the more common mild variety, is not easy to diagnose with structural imaging because it is neither associated with marked morphometric or pathological changes, such as overt edema, bleeding, midline shift, or atrophy, nor does it generate sufficient imaging contrast when using blood pool contrast agents currently available in the clinic.
The divalent manganese ion (Mn2+) is a paramagnetic compound that was recognized as the first MRI T1 contrast agent several decades ago. In fact, the use of Mn2+ as a contrast agent parallels the history of the application of nuclear magnetic resonance (NMR) in biological studies.9–14 The use of Mn2+ as a contrast agent in MRI, termed manganese-enhanced MRI (MEMRI), was pioneered in studies of neural activity, tract-tracing, and neuroarchitecture by Koretsky and colleagues in the 1990s.10,15–18 When administered orally or parenterally, Mn2+ accumulates in all body tissues and is excreted via the biliary system,19–21 a condition that explains its use, in a chelated form, to assess liver and pancreatic pathologies in Europe.19,22–24 After systemic injection, Mn2+ enters the brain promptly through the cerebrospinal fluid (CSF), thereby first reaching the ventricles, choroid plexus, and regions that lack a blood–brain barrier (BBB), such as the pituitary and pineal glands and the median eminence. After crossing the BBB, Mn2+ can be incorporated into brain cells via active transport through calcium-channels and other metal transporters25–29 and it can persist inside cells for weeks or months.30–32 The usefulness of MEMRI in visualizing the effects of trauma on brain structures was first demonstrated in a mouse model of diffuse TBI.33
In the present study, we explored changes in the MEMRI signal using a mouse model of mild-to-moderate blast exposure. We found that mice exposed to blast showed a rapid and significant increase in brain Mn2+ uptake. In addition, we found that such an increase was prevented by shielding of the torso, a strategy previously shown by us to protect the brain from the neuropathological consequences of blast.34,35 Our findings indicate that MEMRI may serve as a non-invasive tool for the rapid and accurate diagnosis of blast TBI.
Mice (C57BL6/J, male; Jackson Laboratories, Bar Harbor, ME) were anesthetized with 4% isoflurane in a gas mixture of 30% oxygen and 70% nitrous oxide, fixed on a wire-mesh holder, and positioned at the open end of the shock tube facing the shockwave front. Our blast model was designed to minimize head movement and thus prevent, to the extent possible, acceleration/deceleration injury. Toward this goal, the head was tethered on the mesh through the frontal teeth. Anesthetized animals were exposed to a single mild-moderate blast event (corresponding to 30 psig of membrane rupture pressure) in the supine position in a helium-driven shock tube as previously described.35,36 Animals were divided into three groups that were exposed to 1) whole–body blast (n=12); 2) blast with a polycarbonate jacket for torso protection with chest, abdomen, and limbs covered (n=5) as described previously35,36; and 3) no blast exposure (sham; n=10). Fifteen minutes after exposure, animals were injected intra-peritoneally with a single dose of 40mg/kg of MnCl2 and subcutaneously with 200μL of neutral phosphate-buffered saline.37 The MnCl2 solution was prepared as described previously.37 Briefly, 100mM of highly purified MnCl2·4H20 (product number 529680; Sigma-Aldrich) were dissolved in 100mM bicine solution and adjusted to pH 7.4. Control mice were not exposed to blast nor did they receive MnCl2. All procedures involving animals were approved by the Animal Care and Use Committees of Johns Hopkins and Georgetown Universities.
MRI was performed in a 7-Tesla horizontal Bruker spectrometer run by Paravision 5.1 as previously described.38–44 Animals were imaged after a 6-h recovery post-blast. Anesthetized mice (1.5% isoflurane in a gas mixture of 30% oxygen and 70% nitrous oxide) were placed on our in-house designed and custom-manufactured stereotaxic device (ASI Instruments, Warren, MI) with built-in temperature and cardio-respiratory monitoring engineered to fit a Bruker 4-channel phased-array mouse brain coil and a corresponding transmit volume coil. T2-weighted rapid acquisition with relaxation enhancement (RARE) sequences were used to identify key neuropathologies such as bleeding, hematomas, edema, or gross changes in brain structure. T1-weighted sequences were then used to visualize changes in contrast related to MnCl2 uptake. The MEMRI imaging protocol was a T1-weighted two-dimensional RARE sequence with the following parameters: repetition time (TR), 1650msec; echo time (TE), 10.6msec; inversion time (TI), 650msec; field of view (FOV), 3.0cm; RARE factor, 2; matrix, 256×256; averages, 1; slice thickness, 1mm. A custom-designed 0.01mM MnCl2 phantom was positioned inside the brain coil above the animal head and used as an internal standard.
The MRI protocol used for gadolinium experiments was a T1-weighted RARE sequence with the following parameters: TR, 300msec; TE, 10.2msec; FOV, 3.0cm; RARE factor, 1; matrix, 256×256; averages, 4; slice thickness, 0.75mm.
Image processing, normalization, and quantification were performed with Paravision 5.1 Processing Software (Bruker, Billerica, MA). Regions of interest (ROIs) located on different areas of the brain and on the MnCl2 phantom were selected from the MR images, and the mean contrast in the gray scale was measured at 6, 24, 48, and 72h and at 14 and 28 days after blast. (Fig. 1). ROIs were of the same size and shape in all subjects. Image contrast data for each ROI were normalized to contrast data in the MnCl2 phantom.
Data were analyzed by one-way analysis of variance (ANOVA). Mean intensities in the gray scale were measured for each brain ROI at the time-points mentioned above. Table 1 shows the ANOVA data with p values for every region and every time-point.
T2-weighted MRI was first performed to rule out the presence of hematoma, hemorrhage, or edema; these conditions can dramatically increase the uptake of Mn2+ due to the disruption of the BBB and thus produce confounding results. No such pathologies were detected in any of our experimental subjects (data not shown). Next, T1-weighted MEMRI was performed as described in the Methods section. Mice with blast TBI showed a marked increase in positive contrast, compared with sham controls (Fig. 2).
In particular, at 6h post-blast, there was a marked signal increase in the choroid plexus of the interventricular region (tela choroidea) and in associated periventricular structures, such as the fornix and the neostriatum, a distribution pattern that reflects the diffusion of Mn2+ across the blood–CSF barrier, as described previously.31,32 Sham controls that had not been exposed to blast but received the MnCl2 injection did not exhibit this change (Fig. 2, left panel). Twenty-four hours after blast, T1-weighted MEMRI showed that the signal increase had shifted from ventricular to periventricular regions and was homogenously dispersed throughout the brain with a marked accentuation of anatomical structures (Fig. 2, right panel). Once again, sham control subjects showed a significantly lower enhancement at the dose of MnCl2 used. This phenomenon is better visualized after pseudo-color processing of the images (Fig. 3).
The single dose of MnCl2 dose used in our study (40mg/kg) had little effect on mortality and morbidity. Only two mice died during the course of the experiment. These deaths occurred in the second week after the injury and, therefore, 2 weeks post-MnCl2 injection. At this time-point, any potential negative effects of MnCl2 are expected to be minimal. For this reason, we suspect that these deaths were unrelated to the contrast agent.
Post-acquisition image analyses were performed to quantify changes in image contrast. To this effect, several regions of interest (ROIs) were selected on serial T1-weighted MR images and mean intensity values were quantified using Paravision 5.1 (Fig. 4). A MnCl2 phantom was affixed to the brain array coil for normalization as described in the Methods section. Mean intensity values in phantom ROIs demonstrated a greater-than-expected variation over the course of the month-long experiment. This variation was not explained by daily drifts in magnetic frequency, temperature or other factors, and ROIs delineated on the background of the image did not show these variations, a pattern indicating that imaging conditions were consistent over time. We therefore concluded that the variability in the mean intensity values of phantoms was the result of microscopic air bubbles and slight changes in the position of phantoms. For the above reasons, we decided to use instead non-blast exposed mice injected with vehicle (n=4) as imaging quality controls. In the absence of MnCl2, these subjects were not expected to show an increase in contrast and could thus be used to establish baseline T1 values. Therefore, specific ROIs in MnCl2-treated blast and MnCl2-treated sham mice were normalized to ROIs localized in comparable regions in sham-vehicle mice.
T1 contrast enhancement was significantly increased in mice exposed to blast, compared with sham mice, in all brain regions examined (Fig. 4). Differences in mean intensity values were significant in all ROI during the first week after blast injury (Table 1). Differences were still present in select regions at 2 weeks but disappeared one month after injury.
Mn2+ is known to accumulate in tissues progressively over the course of several days. In our study, the MEMRI T1 signal continued to increase up to 72h post-blast, after which it progressively diminished until it reached baseline by 1 month. It should be mentioned that mice were not imaged before administration of MnCl2 and therefore the first value depicted in graphs (at 6h) does not correspond to a true baseline contrast.
To determine whether the rapid and marked uptake of Mn2+ was primarily caused by a breakdown of the BBB, we performed contrast-enhanced MRI with the contrast agent gadopentetate dimeglumine (Gd). MRI with Gd-based contrast agents is the accepted standard for the assessment of the integrity of the blood–brain barrier.45–50 When the brain vasculature becomes abnormal, Gd leakage causes signal enhancement in the surrounding tissues. A dose of Gd of 0.3mmol/kg was injected subcutaneously at 6, 24, and 72h and at 7, 14, and 28 days after blast. MRI was performed 20min after injection; this time frame was chosen based on previous DCE-MRI experiments we performed on C57/Bl6 mice with brain tumors and penetrating TBI in which we determined that Gd contrast in mice with BBB disruption reached a maximum 15min after bolus injection and remained constant for approximately 1.5h. Blast-injured mice did not show contrast enhancement in any of the brain ROI selected but did exhibit hyperintensity in midline vascular structures and a lateral ventricular entity that is likely the choroid plexus (Fig. 5). On the other hand, an ROI placed on neck muscle showed a 35% increase in positive contrast. This result indicates that the excessive influx of Mn2+ into the brain after blast exposure is not simply due to vascular pathology and extravasation.
Based on our previous findings showing that torso and upper abdomen protection prevents neuropathological and behavioral changes after blast exposure,35 we explored whether body protection with a stiff polycarbonate jacket can prevent or minimize Mn2+ contrast increase. As shown in Figure 4, MEMRI contrast values in mice exposed to blast in the presence of torso protection were indistinguishable from those of sham subjects. These results show that shielding of the torso prevents not only neuropathological changes caused by blast,35 but also the changes in Mn2+ uptake by the mouse brain.
Our findings establish that the exposure of mice to a 30 psig overpressure shockwave, which corresponds to a mild-to-moderate blast, results in the rapid and enduring increase in uptake of Mn2+ by brain structures, compared with sham controls, based on an increase in positive contrast in T1-weighted MRI. This augmentation of positive contrast progresses from ventricular to periventricular brain sites within the first few hours post-blast and, by 48–72h, it becomes diffuse; it then slowly returns to baseline within a month after exposure to blast. Importantly, torso protection with a polycarbonate body shield covering chest, abdomen, and limbs prevents the increase of MEMRI signal. Further, a single dose of MnCl2 at 40mg/kg is sufficient to track the evolution of contrast agent permeability and retention over a month without evidence of Mn2+-associated toxicity in experimental subjects. These findings suggest that MEMRI generates anatomically and temporally predictable patterns of progressive increase of image contrast in subjects exposed to mild-moderate blast TBI and that this signal increase might be prevented by strategies, such as torso shielding, that also ameliorate neuroinflammation1 and traumatic axonal injury.35 Thus, MEMRI contrast increase may serve as a relatively simple and reliable marker of acute mild-moderate TBI relevant to blast and blunt injury.33
MEMRI signal is significantly different in all ROIs analyzed between injured and sham (control) mice up to 2 weeks after injury. The difference is lost at 1 month post-injury. It is known that Mn2+ has a very slow efflux rate from the brain and can remain in mouse tissue between 68 and 168 days after injection.51 However, the usefulness of MEMRI uptake increase as a biomarker is only apparent in the first couple of weeks after blast injury, at least under the conditions of the present study. Future studies should establish the threshold of detectability of MEMRI signal as a biomarker for blast injury in mice using lower, and thus safer, doses of MnCl2, with different manganese compounds and varying degrees of blast pressure.
The increase in positive contrast observed by MRI after systemic injection of MnCl2 is directly attributable to an increase in the concentration of Mn2+ in brain tissue.52 Doses of MnCl2 administered to rodents have been correlated with concentrations achieved in brain tissue using a variety of techniques such as inductively coupled plasma mass spectrometry,53,54 graphite furnace atomic absorption spectroscopy,55 inductively coupled plasma atomic emission spectroscopy,56 and high resolution autoradiography and gamma counting with 54MnCl2.31,57 Doses and routes of administration of MnCl2 that result in different degrees of brain MRI contrast enhancement also have been established.37,54,58
MEMRI is accepted as a valid and useful preclinical imaging approach to study neuroarchitecture, neural tracts (connectivity), and cerebral function in vivo.10,15,17,37,59 The paramagnetic nature of divalent manganese ions (Mn2+), which is administered to animals in the form of the inorganic salt MnCl2, allows it to function as a T1 MRI contrast agent. Mn2+ uptake by the central nervous system (CNS) appears to be both a passive process (secondary to the breakdown of the BBB), and an active transfer utilizing specific transporters. Mn2+ may compete with Ca2+ for passage through several types of voltage-gated Ca2+ channels25,26 and with iron for transport through the divalent metal transporter (DMT-1, DCT-1, ZIP-8, etc.) in neurons and astrocytes.26–29 In this fashion, Mn2+ uptake imitates the physiological uptake of Ca2+ and iron and thus becomes a marker of neuronal activity.15,37,60 Its rapid uptake, slow efflux from the brain, and general tissue retention over time allows for MEMRI to be used to image acute events, such as exposure to drugs, environmental aggressors or, in our case, blast injury. Several studies focusing on MEMRI and other MRI techniques in the diagnosis of TBI have been published, but most have focused on chronic or acute penetrating TBI61,62 or diffuse blunt trauma.33 As far as we know, ours is the first study using MEMRI as a biological marker for acute TBI related to blast.
The temporal and anatomical pattern of Mn2+ distribution in the brain following blast exposure suggests a compromise of BBB integrity. However, the exploration of BBB permeability with Gd was negative in our study. It is known that in order for the BBB to become permeable to Gd, a significant vascular network breakdown and/or necrosis must be present.63,64 In fact, small brain tumors and lesions with intact vessels are frequently not detectable with Gd-enhanced MRI. Therefore, in blast TBI, even though a BBB breakdown might be a factor in Mn2+ uptake, Gd is unable to detect it and therefore, MEMRI appears to be a more sensitive indicator of BBB integrity than Gd. Active uptake by Ca2+-gated channels is another possible mechanism involved in the incorporation of Mn2+ after blast exposure. Several studies have shown disturbance of calcium homeostasis in TBI and promising beneficial effects of calcium channel blockers.65–68 Future studies should explore the potential effect of calcium channel blockers in reducing MEMRI signal enhancement after TBI.
The increased MEMRI contrast after blast may simply be an exaggeration of the normal process of Mn2+ uptake, especially since the temporal–anatomical pattern of uptake is not qualitatively different between blast and sham condition.31,32 MEMRI contrast enhancement is first apparent at the interventricular foramen and then progresses from the ventricles outwards into the brain parenchyma, thus suggesting that impairments in blood–CNS barriers may play a role; such impairments may be more evident in the blood–cerebrospinal fluid barrier of tela choroidea69 and in the BBB itself. Other mechanisms may be related to disturbances in calcium homeostasis associated with TBI as mentioned above; a massive influx of Mn2+ detected by MEMRI may “image” the increased influx of Ca2+ occurring in the first few hours after blast exposure.68,70–72
Mn2+ is a trace element essential for the normal development and function of the brain, but it also has been shown to be neurotoxic particularly in primates and humans.73,74 Specifically, chronic exposure to high levels of the ion can result in teratogenicity75–77 and extrapyramidal motor dysfunction, a neurodegenerative disorder that resembles Parkinson's disease.78–82 Toxicity is partly the result of Mn2+ acting as an analog of Ca2+ in processes such as synaptic neurotransmission and partly, the result of its role as an oxidant.83,84 The single dose of MnCl2 administered to mice in the present study does not produce neurological deficits.85 In fact, mice and rats tolerate well doses of MnCl2 up to 175mg/kg.86,87 In agreement with previous reports, mice did not exhibit signs of toxicity such as changes in weight, attitude and activity, appetite, hydration, coat appearance, posture and gait, movement, and complications at the injection site in our study.58,85 Still, because of evidence that 40mg/kg of MnCl2 may be mildly neurotoxic,31,37,88 the use of MnCl2 as a potential diagnostic for TBI must be presently restricted to preclinical studies. However, the enormous potential for Mn2+ as a contrast agent in the clinic warrants future research efforts to focus on 1) the development of less toxic manganese-based chelates, clusters, and nanoparticles such as those presently being developed 2) the use of higher magnetic fields and improved imaging sequences that require much lower dosage of Mn2+ for detectability by MRI; and 3) the potential co-administration of drugs that protect against the harmful effects of Mn2+ such as the anti-oxidant lycopene.84
All mice were included in data analysis, irrespective of whether they died or not. We do not attribute the deaths of mice in the course of the study to Mn2+ toxicity. Blast injury is known to cause lesions in organs other than the brain.35 There is a possibility, therefore, that these subjects died for reasons directly related to the blast and not to MnCl2 administration.
In the clinic, CT is the routine first-line imaging modality to assess type and extent of TBI. However, MRI is more sensitive in identifying small structural changes, especially in white matter. MRI techniques employed in the assessment of TBI include conventional MRI for gross anatomical changes—such as contusions and atrophy—or special sequences—such as diffusion weighted imaging to diagnose cerebral edema and ischemia and susceptibility weighted imaging to identify micro-hemorrhages accompanying diffuse axonal injury. However, such approaches have limited value in diagnosing or evaluating mild TBI (concussion). In addition, multi-modal MRI approaches have not yet been implemented in most clinical protocols because standardization is difficult to achieve due to variation in quantitative imaging thresholds, to measurement methodologies, and to fractional anisotropy values in various brain structures at different stages of injury. Therefore, the development of imaging modalities with less variability in important measures remains a high priority, especially for subtle changes associated with mild TBI.
Our findings in this paper indicate that MEMRI signal increase may represent a biomarker for mild-moderate blast TBI. Future studies using blast exposures of lower intensity are pivotal in order to establish the limits/threshold of detectability of MEMRI changes in mild TBI. In addition, comparative studies investigating MEMRI contrast increase between blast and blunt TBI are warranted as the early and accurate detection of mild TBI is not only important for diagnostic and dispositional decisions (e.g., return to play), but also to initiate early treatment and thus improve clinical outcomes.
We would like to acknowledge Dr. Ibolja Cernak, University of Alberta, Canada, for the pioneering ideas behind the projects' inception and funding. Blast experiments were performed in the Applied Physics Laboratory at Johns Hopkins University. All imaging experiments were performed in the Georgetown-Lombardi Comprehensive Cancer Center's Preclinical Imaging Research Laboratory, partially supported by the Cancer Center Support Grant P30 CA51008-21. Animal importation into the Preclinical Imaging Research Laboratory was assisted by the Division of Comparative Medicine at Georgetown University Medical Center. Funding provided by: DOD DM102465 (Cernak, PI, Albanese, subcontract PI), NIH P30 CA51008-18 (Weiner) and ABC2 (Albanese); Maryland Stem Cell Research Fund 2011-MSCRF-22-0067-00 (Koliatsos).
No competing financial interests exist.