- Email: [email protected]
Mechanisms and pathophysiology of the low-level blast brain injury in animal models
NeuroImage 54 (2011) S83–S88
Contents lists available at ScienceDirect
NeuroImage j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y n i m g
Mechanisms and pathophysiology of the low-level blast brain injury in animal models Annette Säljö ⁎, Maria Mayorga, Hayde Bolouri, Berndt Svensson, Anders Hamberger Institute of Biomedicine, Department of Medical Chemistry and Cell Biology, Sahlgren Academy, University of Gothenburg, Gothenburg, Sweden
a r t i c l e
i n f o
Article history: Received 15 October 2009 Revised 17 May 2010 Accepted 18 May 2010 Available online 24 May 2010 Keywords: Blast overpressure Brain injury Brain edema Cognitive function Intracranial pressure Shock tube
a b s t r a c t The symptoms of primary blast-induced mTBI, posttraumatic stress disorder and depression overlap. Evidence of an organic basis for these entities has been scarce and controversial. We present a review of animal studies demonstrating that low-level blast causes pathophysiological and functional changes in the brain. We monitor a time period from minutes to approximately 1 week after blast exposure from multiple modes (air, underwater, localized and whole body). The most salient findings observed were (1) the peak pressures (Pmax) in the brain, elicited from the blast from the firing of military weapons (Pmax 23–45 kPa), have a similar magnitude as that registered in air close to the head. Corresponding measurements during the detonation pulse from explosives under water show a Pmax in the brain, which is only 10% of that in water outside the head. (2) The rise time of the pressure curve is 10 times longer in the brain as compared with the blast in air outside the head during firing of military weapons. (3) The lower frequencies in the blast wave appear to be transmitted more readily to the brain than the higher frequencies. (4) When animals are exposed to low levels of blast, the blast wave appears mostly transmitted directly to the brain during air exposure, not via the thorax or abdomen. (5) Low levels of blast cause brain edema, as indicated by increased bioelectrical impedance, an increase in the intracranial pressure, small brain hemorrhages and impaired cognitive function. Published by Elsevier Inc.
Introduction Today, with the increased incidence of mild TBI (mTBI) in the Iraqi and Afghanistan conflicts, there is renewed interest in understanding the mechanism of this injury. There is now emerging evidence for a pathophysiologic basis for blast-induced mTBI (Chavko et al., 2007; Moochhala et al., 2004; Saljo et al., 2008, 2009; Warden, 2006). Scientists in Sweden have had an interest in blast phenomena, injury mechanisms and the development of protective strategies since the formulation of dynamite by Alfred Nobel in the mid-1800s. Their contributions to blast-induced traumatic brain injury (TBI) are of particular relevance (Clemedson, 1956a) because the question of whether non-lethal, non-penetrating blast exposure was associated with a pathologically based clinical entity was a controversial topic more than 50 years ago. The research on blast started at FOA (Defence Research Establishment), Sweden, in the late 1940s with Clemedsons PhD thesis “An experimental study on blast injuries”, in 1949, followed by the publication of a series of reports on blast injuries (Clemedson, 1951, 1954), propagation of blast waves through the body (Clemedson and Pettersson, 1956) and the use of shock tubes ⁎ Corresponding author. Institute of Biomedicine, Section of Medical Chemistry and Cell Biology, Sahlgren Academy, University of Göteborg, P O Box 440, Göteborg, SE 405 30 Sweden. E-mail address: [email protected] (A. Säljö). 1053-8119/$ – see front matter. Published by Elsevier Inc. doi:10.1016/j.neuroimage.2010.05.050
(Celander et al., 1954). Clemedson showed that the pressure wave, which is transmitted to the brain, does not differ much with respect to amplitude from the recorded blast wave in air outside the head (Clemedson, 1956b). This led him to conclude that perhaps the blast wave was being conducted through the skull. Other studies, determining the criteria for suitability of soldiers for training with powerful weapons, such as the Karl Gustaf bazooka (Aschan, 1954), were also conducted. There was already knowledge that these powerful weapons caused symptoms which only more “tough” soldiers could withstand. Soldiers underwent an extensive neuropsychiatric examination before and after training. Nine out of 11 of the randomly selected soldiers performed worse in the retest, i.e. did not fulfill the criteria for suitability as gunners. In the late 1990s, blast studies were resumed in FOA (“Impulse noise and brain injury”, Säljö, PhD thesis, 2001). At that time, Swedish blast research moved to the University of Gothenburg. Säljö's thesis was focused on how impulse noise affects the central nervous system. Neuronal apoptosis and alterations in a number of neuronal proteins were demonstrated immunohistochemically (Saljo et al., 2000, 2002a, b). An increased reactivity was also demonstrated in the supporting cells in the brain, the astrocytes and the microglia (Saljo Moberg, 2001). Blast-induced leakage of marker proteins from neurons and glial cells was also detected (Säljö et al., 2003). Today, with the increased incidence of mild TBI in military conflicts, there is renewed interest in understanding the mechanism of this injury. This review
S84
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
describes our research which adds to the increasing body of evidence supportive of a pathophysiologic basis for mTBI (Chavko et al., 2007; Moochhala et al., 2004; Saljo et al., 2008; Saljo et al., 2009; Warden, 2006). Low-level blast and brain injury Our research efforts were prompted by a paucity of information regarding the early blast wave–brain interactions leading to the cellular and molecular changes described by Säljö (PhD thesis, 2001). We also describe the pathophysiologic changes leading to injury and dysfunction. The studies were undertaken to explore the risk for brain injury from multiple exposures of blast in the peri-threshold regions of military occupational standards. Porcine and rodent models were used to investigate the interaction between the blast wave and the body. The character of the pressure curve recorded in the brain is compared with that in air and water. We have also attempted to determine the contribution of each parameter of the pressure time function in air and water to causation of brain injury. This allows us to propose a mechanistic hypothesis for blast related TBI. Changes in the macroscopic and microscopic appearance of the brain, the intracranial pressure and cognitive function are also described.
higher frequencies in that of the rifle (Saljo et al., 2008). Though the exact mechanisms are not yet clear, the transmission of low frequencies into the brain appears more efficient than that of high frequencies, an observation supported by results from under water experiments (see below). This efficiency can be assessed using the brain/air Pmax ratio, which decreases with increasing high frequency content (Table 1). The brain/air ratios of other parameters of the pressure time function, such as the A-duration (the duration of the first positive peak) and the impulse (integration of the pressure time curve), were between 1.0 and 2.0, while the ratios for the Pmax were between 1.0 and 0.4 (Tables 1 and 2). The parameter, which differed most markedly between brain and air, was the rise time. The brain/air ratio for the rise time was in the range of 10 for the various weapons (Table 2). The contribution of head acceleration to injury induction was determined with high-speed video and an accelerometer, which was aligned with the axis of impact and attached to the animal's head (Saljo et al., 2008). At the levels of blast overpressure used in our experiments, the acceleration of the head was very small and not considered to be causative to the measured pathophysiologic and functional changes. Bioelectrical impedance
Characterization of pressure curves in air and brain Anesthetized pigs were placed in the gunner's position of weapons and received 3 consecutive exposures (Saljo et al., 2008). The following weapons and maximal peak pressures (Pmax expressed in kPa) were used: a howitzer (Haubits F77, 9 or 30 kPa), a bazooka (Karl Gustaf, 45 kPa) and an automatic rifle (AG-90, 23 kPa). The pressure levels in air were measured at the gunner's position. During blast exposure, the pressure as a function of time was recorded in the brain and in air, close to the head of the animal (Table 1, Saljo et al., 2008). A hydrophone (diam. 9.5 mm) or a pen-shaped piezo-electric sensor was used for recordings in air. Hydrophones were routinely used as pressure sensors in the pig brain (Saljo et al., 2008); however, occasionally, miniaturized optic probe sensors (diameter 0.42 mm. Samba, Saljo et al., 2009) were used in parallel. The most striking finding was that the pressure recorded in the brains had a Pmax, which was of similar or slightly reduced magnitude as that in air close to the head. The brain/air Pmax ratio was high for the bazooka and the 30-kPa howitzer but low for the automatic rifle (Table 1). Our conclusion of this was that the scalp, skull bone and cerebrospinal fluid, which separate the brain from the surrounding air, do not constitute an appreciable protection for the brain. While the pressure time curves in brain were very similar to those in the air after exposure to the howitzer, this similarity was reduced with the bazooka and was even less evident with the rifle (Saljo et al., 2008). The frequency distribution of the blast wave differs among the weapons: lower frequencies contained in the howitzer blast and
Bioelectrical impedance is used as a means to detect cerebral edema following hypoxia (Klein and Krop-Van Gastel, 1993; Verheul et al., 1994; Williams et al., 1991; Lingwood et al., 2003) and TBI (Harting et al., 2010). The bioelectrical impedance (Olsson et al., 2006) was measured in pilot experiments on anaesthetized pigs that were exposed in the gunner's position (bazooka, Pmax 45 kPa). In exposed animals, the electrical impedance in the brain increased after a few seconds and reversed within a few minutes in most experiments but occasionally remained elevated for 20–30 min. (Fig. 1). This pattern was similar to the changes in brain impedance during hypoxia (Lingwood et al., 2003). We conclude that the increased bioelectrical impedance support the microscopic findings and the increased ICP and is indicative of the development of brain edema following blast exposure. Neuropathology Pigs were sacrificed 3–7 days after exposure to weapon detonation and the brains were fixed in formalin by perfusion of the vascular system. While the surface of the brains from the control animals was consistently pale, 30% of the brains from the animals exposed to the bazooka or the rifle displayed, on the surface, erythematous regions of varying size and intensity. Microscopical examination revealed a significant increase in the amount of erythrocyte-filled capillaries in the erythematous regions, which we interpreted as a sign of brain edema (Saljo et al., 2008), as previously described in another model of
Table 1 Blast characteristics (Pmax and frequency distribution) in air or water close to the animals head and in the frontal cerebral cortex of a pig in crew positon for the weapons and in water, 96 m from the explosive. Pmax
Weapons in air Karl Gustav Haubits 30 kPa Haubits 9 kPa Automatic rifle Explosives under water (250 g, 96 m) Detonation pulse 1st secondary pulse 2nd secondary pulse
Frequency distribution
kPa
kPa
Ratio
Air 42 30 9 23 Water 482 77 33
Brain 28 20 4 10 Brain 51 25 16
Brain/air 0.7 0.7 0.4 0.4 Brain water 0.1 0.3 0.5
Air Medium Low Low Medium Water High Medium Medium
Brain Low Low Low Low Brain Low Low Low
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
S85
Table 2 Blast characteristics (A-duration, impulse and rise time) in air or water close to the animals head andin the frontal cerebral cortex of a pig in crew positon for the weapons and in water, 96 m from the explosive. Weapons in air
Karl Gustav Haubits 30 kPa Haubits 9 kPa Automatic rifle Explosives under water (250 g, 96 m) Detonation pulse 1st secondary pulse 2nd secondary pulse
A-duration
Impulse
Air (ms)
Brain (ms)
2 5 2 1.5
4 8 3 1.3
Ratio Brain/air 2 1.6 1.5 0.9
Rise time
Air (Pa × s)
Brain (Pa × s)
30 59 9 7 Water 51
47 79 10 6 Brain 21
TBI (Foda and Marmarou, 1994). Foda described parenchymal microvascular changes consisting of brain edema and vascular congestion with “sludging” of red blood cells. Vasospasm may be of special importance at exposure to higher levels of blast (Armonda et al., 2006). Microscopical examination also showed small subarachnoidal and brain tissue hemorrhages in pigs exposed to the howitzer at 30 kPa, the bazooka or the rifle (Fig. 2). Hemorrhages were found particularly in the occipital lobe, in the cerebellum and the medulla oblongata (Saljo et al., 2008).
Ratio
kPa/ms
kPa/ms
1/(brain/air)
970 1097 304
76 87 25
13 13 12
Water 51972 1341 390
Brain 5005 280 83
1/(brain/water) 10 5 5
Brain/air 1.6 1.3 1.1 0.9 Brain/water 0.4
the detonation pulse, while such frequencies were virtually lacking in the secondary pulses (Saljo et al., 2008). Hardly any frequencies above 1000 Hz appeared to be transmitted into the brain. Furthermore, there was little difference in the brain in frequency distribution after exposure to the detonation and the secondary pulses and frequencies around 100 Hz dominated. The brain/water ratio for the rise time was 10 for the detonation pulse while 5 for the secondary pulses. This is probably due to the predominantly low frequencies in the secondary pulses but not in the detonation pulse (Saljo et al., 2008).
Characterization of pressure waves in water and brain Transmission of pressure waves in the body
Intubated pigs, connected to an air tank, were placed 5 m below the surface in 10-m deep water in lake Vättern (Sweden) and exposed to 60, 125 or 250 g explosives, detonated at distances of 66–96 m (Saljo et al., 2008). At these distances, the recordings show a primary detonation pulse, which is followed by secondary pulses of smaller magnitude, the first two after 140 and 250 ms, respectively. The pulses have a rapid peak, followed by oscillations of decreasing magnitude. The maximal response in the brain differed considerably from the detonation pulse in water outside the head, while the responses in the brain were similar to those in water during the secondary pulses. This was reflected in the brain/water Pmax ratios, which were 0.1 for the detonation pulse and 0.3 and 0.5 for the first and second secondary pulse, respectively (Table 1; Saljo et al., 2008). The frequency distribution of the pulses appeared to play a similar role as for the abovementioned pulses from different weapons in air. In water, frequencies in the 5000- to 10,000-Hz range were present in
The experiments were undertaken in order to understand the transfer of pressure from a local blast exposure to the abdomen or to the head. In these experiments, pigs were exposed locally by means of a shock tube with a diameter of 0.2 m (Saljo et al., 2008). Pressures in the abdomen and the brain were compared after local exposure of either the abdomen or the head. When a Pmax of 30 kPa was recorded within the tube, a Pmax of 15 kPa was registered in the air outside the abdomen, 20–30 mm lateral to the tube opening. A similar Pmax, 15 kPa, was recorded inside the abdomen. In the same experiments, the Pmax in air outside the head was 1.8 kPa since the upper body could not be completely shielded. The registered Pmax in the brain was 0.5 kPa. This pressure could be transmitted exclusively from the air outside the head, as judged from brain/air Pmax ratios during free field exposures (Table 1). At the level of blast exposure used, the results do
Fig. 1. Time course of changes in bioelectrical impedance in the brain during 3 successive exposures of pigs to blast overpressure from the firing of a bazooka (Pmax 45 kPa). 1, 2 and 3 indicate the timing of exposures.
Fig. 2. Light microscopy of section from the left occipital cortex of the pig brain exposed to blast overpressure from the firing of a bazooka. Stained for unspecific peroxidase. Red blood cells in brain tissue outside blood capillaries and erythrocyte-filled capillaries.
S86
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
not indicate an appreciable transmission of a pressure from the abdomen to the brain. In other experiments, the vertex of the head was locally exposed in a similar fashion as the abdomen. The Pmax in air outside the head was then 22 kPa and that within the brain 9 kPa, again in agreement with the air/brain ratios during free field exposures (Table 1). The peak pressure in air outside the abdomen was now 2 kPa and that in the abdomen 0.8 kPa. The results do not support a transfer of pressure from the head to the abdomen. Intracranial pressure (ICP) In this and the following section, male Wistar rats (230–330 g) were exposed to blast in the shock tube, mentioned in the previous section. The animals were placed in the tube, 0.25 m from the free opening of the shock tube and had the long axis of their body parallel to and its head facing the direction of the blast front. Pulses with Pmax of 10, 30 and 60 kPa and duration of 4–6 ms as measured in the air in close proximity to the head were generated with pressures of 0.2, 0.6 and 1.2 bar, respectively (Saljo et al., 2009). The ICP was measured with the miniaturized optic probe (Samba) at 0.5, 2, 6 and 10 h and 1, 2, 3, 5 and 7 days after exposure 10 kPa, 30 kPa or 60 kPa. In addition to the pressure curve recorded in the brain during blast exposure, there is a slowly rising, long lasting, although finally reversed, increase in the pressure, labeled ICP. However, an increase in the ICP is not a specific reaction to blast exposure (Saljo et al., 2009). The mean ICP in control rats was close to 6 mm Hg (Saljo et al., 2009), whereas the ICP in rats exposed at 60 kPa increased significantly after 30 min (Fig. 3). A corresponding increase was not seen until after 2 h in rats, exposed at 30 kPa and as late as after 6 h in rats exposed at 10 kPa. The 30- and the 60-kPa groups reached a peak value for ICP 10 h after exposure, at a level of 11 and 16 mm Hg, respectively. There was a slow return towards control levels during the subsequent days. Rodent studies, simulating the concussions sustained by football players (Viano et al., 2009), provided an opportunity to compare ICP measurements from a blast and non-blast model of TBI in rats of similar weight and strain (Fig. 3). The concussive trauma resulted in a peak value and peak time, which is similar to that after a blast exposure at 60 kPa (unpublished). Cognitive function For tests of cognitive function, rats were trained and tested in a circular, water filled pool (Morris, 1984; Saljo et al., 2009). Their task was to find a platform, 10 mm under the water level, which was mounted in one of the pool quadrants. Rats were trained for 6 days to find the platform within 10 s. Exposure to a blast overpressure of 10 kPa
on the 7th day increased the time to find the platform to 17 s 2 days later. The time to find the platform increased even more in rats exposed to 30 kPa, i.e. 21 s (Saljo et al., 2009). The speed of swimming was similar in all rats. This increase in time to find the platform in blast exposed rats is indicative of cognitive dysfunction.
Discussion The exact mechanism of how a blast wave interacts with the body is unknown. Several modes have been hypothesized to include formation of an electromagnetic field, direct transmission of the blast wave to the brain through the skull, the formation of a secondary shear wave (Moore et al., 2008), transmitted pressure from the vasculature and soft tissues of the thorax (Cernak and Malizevic, 1997) and acceleration of the head. The different modes may operate singly or in concert and one mechanism may predominate in lowlevel blast versus another at higher level blast exposure. Low, occupational levels of blast affect both cognitive performance and ICP (Saljo et al., 2009). Similar effects are observed in our model for non-blastproduced, sport-related concussion (Viano et al., 2009). It has been assumed that only the auditory system is sensitive to the high frequency components in a blast wave, whereas the nonauditory system was more so to lower frequency (Patterson and Hamernik, 1997). Little is known about the brain's response to the frequency distribution of a blast wave. Our studies demonstrate for the first time that the frequency spectrum of a blast wave is important in the transmission of energy after primary blast exposure (Table 1), a result that has been verified in a recent study (Chavko, 2009). Animals exposed to powerful, low frequency weapons (i.e. howitzer) showed the highest brain/air Pmax ratios. In our underwater experiments, the relation between the brain/water Pmax ratio and frequency distribution was evident; however, injury correlates are not possible to make. Furthermore, frequencies above 1000 Hz do not transmit to the brain (Saljo et al., 2008). Our observation, that the lower frequencies contained in a blast wave appear to be transmitted more readily to the brain, is consistent with studies of blast pressure transmission from air into the abdomen of pregnant sheep (Gerhardt et al., 2000). Our studies indicate that blast overpressure is mostly transmitted directly to the brain in air exposure, not via the thorax or abdomen. This is in contrast to recent studies (Cernak and Malizevic, 1997; Cernak et al., 2001). In our experiments with the bazooka and the howitzer, the pressure transmission into brain was high, the Pmax air/ brain ratio being 0.7 (Table 1). Clemedson was the first to report that there are no appreciable Pmax differences in air versus brain during blast exposure (Clemedson and Pettersson, 1956) and recent studies (Chavko et al., 2007; VandeVord, 2009) also corroborate our findings.
Fig. 3. Time course of the intracranial pressure (ICP, mm Hg) in rats during 1 week after blast exposure at 10 (●), 30 (■) and 60 kPa (▲). The results are compared with ICP measurement (unpublished) after translational acceleration of the head produced in a rat concussive model simulating professional football brain injuries (Viano et al., 2009). In comparison with blast exposure at 60 kPa, the peak ICP values and peak time course are similar. Baseline levels are given as time zero. Data are presented as mean ± SEM.
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
A dose–response increase in ICP was shown after blast exposure and an inverse relationship was noted between the level of blast and the time of onset of ICP changes, i.e. in agreement with findings in models of closed head injury (Engelborghs et al., 1998). We hypothesize that the increased ICP after blast exposure is related to brain edema (Fig. 3). However, there are a number of possible causative factors for increased ICP, such as hemorrhage, CSF outflow obstruction and BBB damage (Hoane et al., 2006; Vink et al., 2003). The immediate and late increases in ICP elevation are being attributed to vascular changes and more complex changes, respectively (Engelborghs et al., 1998; Fritz et al., 2005; Jamali et al., 1998; Rooker et al., 2003). ICP measurements seem to be a more sensitive method to detect mTBI than markers for neuropathology (Teranishi et al., 2009), both with respect to the onset of injury and exposure conditions. The measurement of electrical impedance in the brain is a promising means to detect cerebral edema following hypoxia-ischemia (Klein and Krop-Van Gastel, 1993; Verheul et al., 1994; Williams et al., 1991; Lingwood et al., 2003) and TBI (Harting et al., 2010). The usefulness of animal studies is limited by the uncertainty of scaling relationships, validity of laboratory models and species differences. There are no studies comparing species differences in neurological susceptibility, anatomy, tissue composition, stress strain responses, etc. Furthermore, differences in blast generating sources (explosives, shock tube), exposure to multiple effects of blast (secondary, tertiary, quaternary, etc), environmental design (free field, enclosures) and orientation of the subjects make comparison of animal studies difficult. Few studies have addressed the questions of the influence of the frequency spectra of blasts (Chavko, 2009) or the effects of repeated exposures. The correlation of blast levels with the neuropathology of brain injury induced by low levels of blast is also fairly unexplored (Moochhala et al., 2004; Saljo et al., 2009; Teranishi et al., 2009). Our results pose questions with regard to the validity of current occupational standards since brain injury was not a factor considered during their development. More studies are needed to establish routes and mechanisms of pressure transmission from the blast wave to the brain and to identify initial and secondary mechanisms of injury. Also, the role of the blast wave parameters and injury correlates of brain injury development should be explored further. The creation of a comprehensive pathologic scoring system would assist in comparing results from different biological models. Lastly, advancement of numerical models are needed to better deal with the multitude of experimental variables, which ultimately may result in protective strategies. Conclusions 1. The peak pressures (Pmax) in the brain, elicited by the blast from the firing of military weapons (Pmax 23–45 kPa), have a similar magnitude as that registered in air close to the head. Corresponding measurements during the detonation pulse from explosives under water showed a Pmax in the brain, which was only 10% of that in water outside the head. 2. The rise time of the pressure curve was 10 times longer in the brain as compared with the blast in air outside the head during firing of military weapons 3. The lower frequencies in the blast wave appeared to be transmitted more readily into the brain than the higher frequencies. 4. When animals are exposed to low levels of blast in air, the blast wave was mostly transmitted directly to the brain, not via the thorax or abdomen 5. Low levels of blast cause • brain edema, as indicated by increased bioelectrical impedance (bazooka 45 kPa) • an increase in the intracranial pressure, peaking 6–10 h after exposure (10, 30 and 60 kPa) • small brain hemorrhages (bazooka, automatic rifle)
S87
• impaired cognitive function 2 days after the exposure (10 and 30 kPa) Conflict of interest statement The authors declare that there are no conflicts of interest. Acknowledgments The authors want to express their gratitude to Svante Höjer, Samba Sensors AB, Västra Frölunda, Sweden, for sharing with us his knowledge on pressure sensors and pressure recording. We also thank Torsten Olsson, professor emeritus, Department of Signals and Systems, Chalmers University of Technology, Göteborg, Sweden, for introducing us into the field of electrical impedance and for carrying out experiments with us. The study was supported by grants from the Swedish Armed Forces/FMV and by Svenska Militärläkareföreningen. References Armonda, R.A., Bell, R.S., Vo, A.H., Ling, G., DeGraba, T.J., Crandall, B., Ecklund, J., Campbell, W.W., 2006. Wartime traumatic cerebral vasospasm: recent review of combat casualties. Neurosurgery 59, 1215–1225 discussion 1225. Aschan, Sälde, 1954. Medicinska undersökningar av närpvskyttar. Försvarets Forskningsanstalt, Sundbyberg, Sweden. Dnr 10043-8112. Celander, H., Clemedson, C.-J., Ericsson, U., Hultman, H., 1954. The use of compressed air operated shock tube for physiological blast research. Försvarets Forskningsanstalt Sundbyberg, Sweden. Dnr 1690-8112. Cernak, I., Malizevic, Z., 1997. Indirect neurotrauma caused by pulmonary blast injury: development and prognosis. Int. Rev. Armed Forces Med. Serv. 70, 114–120. Cernak, I., Wang, Z., Jiang, J., Bian, X., Savic, J., 2001. Ultrastructural and functional characteristics of blast injury-induced neurotrauma. J. Trauma 50, 695–706. Chavko, M., 2009. Transfer of energy inside the body after exposure to blast. Military Health Research Forum, Kansas City, MO, USA. Chavko, M., Koller, W.A., Prusaczyk, W.K., McCarron, R.M., 2007. Measurement of blast wave by a miniature fiber optic pressure transducer in the rat brain. J. Neurosci. Methods 159, 277–281. Clemedson, C.J., 1951. Detonation injuries. Svenska Lakartidning 48, 1337–1343. Clemedson, C.J., 1954. Correlation between respiratory phase and extent of lung damage in air blast injury. J. Appl. Physiol. 7, 38–42. Clemedson, C.-J., 1956a. Shock wave transmission to the central nervous system. Acta Physiol. Scand. 37, 204–213. Clemedson, C.J., 1956b. Shock wave transmission to the central nervous system. Acta Physiol. Scand. 37, 2–3 Fasc. Clemedson, C.J., Pettersson, H., 1956. Propagation of a high explosive air shock wave through different parts of an animal body. Am. J. Physiol. 184, 112–119. Engelborghs, K., Verlooy, J., Van Reempts, J., Van Deuren, B., Van de Ven, M., Borgers, M., 1998. Temporal changes in intracranial pressure in a modified experimental model of closed head injury. J. Neurosurg. 89, 796–806. Foda, M.A., Marmarou, A., 1994. A new model of diffuse brain injury in rats. Part II: Morphological characterization. J. Neurosurg. 80, 301–313. Fritz, H.G., Walter, B., Holzmayr, M., Brodhun, M., Patt, S., Bauer, R., 2005. A pig model with secondary increase of intracranial pressure after severe traumatic brain injury and temporary blood loss. J. Neurotrauma 22, 807–821. Gerhardt, K.J., Abrams, R.M., Huang, X., Griffiths, S.K., Peters, A.J., 2000. Intra-abdominal sound pressure levels during impulse noise exposure in sheep. Mil. Med. 165, 153–156. Harting, M.T., Smith, C.T., Radhakrishnan, R.S., Aroom, K.R., Dash, P.K., Gill, B., Cox Jr., C.S., 2010. Regional differences in cerebral edema after traumatic brain injury identified by impedance analysis. J. Surg. Res. 159, 557–564. Hoane, M.R., Kaplan, S.A., Ellis, A.L., 2006. The effects of nicotinamide on apoptosis and blood–brain barrier breakdown following traumatic brain injury. Brain Res. 1125, 185–193. Jamali, S., Bodjarian, N., Vigue, B., Mazoit, J.X., Samii, K., Tadie, M., 1998. Increase in the chronically monitored cerebrospinal fluid pressure after experimental brain injury in rats. Brain Inj. 12, 525–536. Klein, H.C., Krop-Van Gastel, W., 1993. Prediction of specific damage or infarction from the measurement of tissue impedance following repetitive brain ischaemia in the rat. Neuropathol. Appl. Neurobiol. 19, 57–65. Lingwood, B.E., Dunster, K.R., Healy, G.N., Ward, L.C., Colditz, P.B., 2003. Cerebral impedance and neurological outcome following a mild or severe hypoxic/ischemic episode in neonatal piglets. Brain Res. 969, 160–167. Moochhala, S.M., Md, S., Lu, J., Teng, C.H., Greengrass, C., 2004. Neuroprotective role of aminoguanidine in behavioral changes after blast injury. J. Trauma 56, 393–403. Moore, D.F., Radovitzky, R.L., Shupenko, L., Klinoff, A., Jaffe, M.S., Rosen, J.M., 2008. Blast physics and central nervous system injury, pp. 243–250. Morris, R., 1984. Developments of a water maze procedure for studying spatial learninng in the rat. J. Neurosci. Methods 11 (1), 47–60. Olsson, T., Broberg, M., Pope, K.J., Wallace, A., Mackenzie, L., Blomstrand, F., Nilsson, M., Willoughby, J.O., 2006. Cell swelling, seizures and spreading depression: an impedance study. Neuroscience 140, 505–515.
S88
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
Patterson Jr., J.H., Hamernik, R.P., 1997. Blast overpressure induced structural and functional changes in the auditory system. Toxicology 121, 29–40. Rooker, S., Jorens, P.G., Van Reempts, J., Borgers, M., Verlooy, J., 2003. Continuous measurement of intracranial pressure in awake rats after experimental closed head injury. J. Neurosci. Methods 131, 75–81. Saljo Moberg, A., 2001. Impulse noise and brain injury: an experimental study in rats (PhD thesis), University of Gothenburg, Sweden. Saljo, A., Bao, F., Haglid, K.G., Hansson, H.A., 2000. Blast exposure causes redistribution of phosphorylated neurofilament subunits in neurons of the adult rat brain. J. Neurotrauma 17, 719–726. Saljo, A., Bao, F., Jingshan, S., Hamberger, A., Hansson, H.A., Haglid, K.G., 2002a. Exposure to short-lasting impulse noise causes neuronal c-Jun expression and induction of apoptosis in the adult rat brain. J. Neurotrauma 19, 985–991. Saljo, A., Bao, F., Shi, J., Hamberger, A., Hansson, H.A., Haglid, K.G., 2002b. Expression of c-Fos and c-Myc and deposition of beta-APP in neurons in the adult rat brain as a result of exposure to short-lasting impulse noise. J. Neurotrauma 19, 379–385. Säljö, A., Huang, Y.-L., Hansson, H.-A., 2003. Impulse noise transiently increased the permeability of nerveand glial cell membranes, an effect accentuated by a recent brain Injury. J. Neurotrauma 20, 787–794. Saljo, A., Arrhen, F., Bolouri, H., Mayorga, M., Hamberger, A., 2008. Neuropathology and pressure in the pig brain resulting from low-impulse noise exposure. J. Neurotrauma 25, 1397–1406.
Saljo, A., Svensson, B., Mayorga, M., Hamberger, A., Bolouri, H., 2009. Low levels of blast raises intracranial pressure and impairs cognitive function in rats. J. Neurotrauma 26, 1345–1352. Teranishi, K., Chavko, Mikulas, Adeeb, Saleena, Carroll, Erica, McCarron, Richard, 2009. Blast-induced neuropathological changes in brain in relation to different orientation to blast. International and National Neurotrauma Societies, Santa Barbara, CA, USA. VandeVord, P., 2009. Understanding of Shock wave transmission to the brain. International State of the Science Meeting on No-Impact Blast-induced Mild Brain Injury. Herndon, VA, USA. Verheul, H.B., Balazs, R., Berkelbach van der Sprenkel, J.W., Tulleken, C.A., Nicolay, K., Tamminga, K.S., van Lookeren Campagne, M., 1994. Comparison of diffusion-weighted MRI with changes in cell volume in a rat model of brain injury. NMR Biomed. 7, 96–100. Viano, D.C., Hamberger, A., Bolouri, H., Saljo, A., 2009. Concussion in professional football: animal model of brain injury—part 15. Neurosurgery 64, 1162–1173 discussion 1173. Vink, R., Young, A., Bennett, C.J., Hu, X., Connor, C.O., Cernak, I., Nimmo, A.J., 2003. Neuropeptide release influences brain edema formation after diffuse traumatic brain injury. Acta Neurochir. Suppl. 86, 257–260. Warden, D., 2006. Military TBI during the Iraq and Afghanistan wars. J. Head Trauma Rehabil. 21, 398–402. Williams, C.E., Gunn, A., Gluckman, P.D., 1991. Time course of intracellular edema and epileptiform activity following prenatal cerebral ischemia in sheep. Stroke 22, 516–521.
Contents lists available at ScienceDirect
NeuroImage j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y n i m g
Mechanisms and pathophysiology of the low-level blast brain injury in animal models Annette Säljö ⁎, Maria Mayorga, Hayde Bolouri, Berndt Svensson, Anders Hamberger Institute of Biomedicine, Department of Medical Chemistry and Cell Biology, Sahlgren Academy, University of Gothenburg, Gothenburg, Sweden
a r t i c l e
i n f o
Article history: Received 15 October 2009 Revised 17 May 2010 Accepted 18 May 2010 Available online 24 May 2010 Keywords: Blast overpressure Brain injury Brain edema Cognitive function Intracranial pressure Shock tube
a b s t r a c t The symptoms of primary blast-induced mTBI, posttraumatic stress disorder and depression overlap. Evidence of an organic basis for these entities has been scarce and controversial. We present a review of animal studies demonstrating that low-level blast causes pathophysiological and functional changes in the brain. We monitor a time period from minutes to approximately 1 week after blast exposure from multiple modes (air, underwater, localized and whole body). The most salient findings observed were (1) the peak pressures (Pmax) in the brain, elicited from the blast from the firing of military weapons (Pmax 23–45 kPa), have a similar magnitude as that registered in air close to the head. Corresponding measurements during the detonation pulse from explosives under water show a Pmax in the brain, which is only 10% of that in water outside the head. (2) The rise time of the pressure curve is 10 times longer in the brain as compared with the blast in air outside the head during firing of military weapons. (3) The lower frequencies in the blast wave appear to be transmitted more readily to the brain than the higher frequencies. (4) When animals are exposed to low levels of blast, the blast wave appears mostly transmitted directly to the brain during air exposure, not via the thorax or abdomen. (5) Low levels of blast cause brain edema, as indicated by increased bioelectrical impedance, an increase in the intracranial pressure, small brain hemorrhages and impaired cognitive function. Published by Elsevier Inc.
Introduction Today, with the increased incidence of mild TBI (mTBI) in the Iraqi and Afghanistan conflicts, there is renewed interest in understanding the mechanism of this injury. There is now emerging evidence for a pathophysiologic basis for blast-induced mTBI (Chavko et al., 2007; Moochhala et al., 2004; Saljo et al., 2008, 2009; Warden, 2006). Scientists in Sweden have had an interest in blast phenomena, injury mechanisms and the development of protective strategies since the formulation of dynamite by Alfred Nobel in the mid-1800s. Their contributions to blast-induced traumatic brain injury (TBI) are of particular relevance (Clemedson, 1956a) because the question of whether non-lethal, non-penetrating blast exposure was associated with a pathologically based clinical entity was a controversial topic more than 50 years ago. The research on blast started at FOA (Defence Research Establishment), Sweden, in the late 1940s with Clemedsons PhD thesis “An experimental study on blast injuries”, in 1949, followed by the publication of a series of reports on blast injuries (Clemedson, 1951, 1954), propagation of blast waves through the body (Clemedson and Pettersson, 1956) and the use of shock tubes ⁎ Corresponding author. Institute of Biomedicine, Section of Medical Chemistry and Cell Biology, Sahlgren Academy, University of Göteborg, P O Box 440, Göteborg, SE 405 30 Sweden. E-mail address: [email protected] (A. Säljö). 1053-8119/$ – see front matter. Published by Elsevier Inc. doi:10.1016/j.neuroimage.2010.05.050
(Celander et al., 1954). Clemedson showed that the pressure wave, which is transmitted to the brain, does not differ much with respect to amplitude from the recorded blast wave in air outside the head (Clemedson, 1956b). This led him to conclude that perhaps the blast wave was being conducted through the skull. Other studies, determining the criteria for suitability of soldiers for training with powerful weapons, such as the Karl Gustaf bazooka (Aschan, 1954), were also conducted. There was already knowledge that these powerful weapons caused symptoms which only more “tough” soldiers could withstand. Soldiers underwent an extensive neuropsychiatric examination before and after training. Nine out of 11 of the randomly selected soldiers performed worse in the retest, i.e. did not fulfill the criteria for suitability as gunners. In the late 1990s, blast studies were resumed in FOA (“Impulse noise and brain injury”, Säljö, PhD thesis, 2001). At that time, Swedish blast research moved to the University of Gothenburg. Säljö's thesis was focused on how impulse noise affects the central nervous system. Neuronal apoptosis and alterations in a number of neuronal proteins were demonstrated immunohistochemically (Saljo et al., 2000, 2002a, b). An increased reactivity was also demonstrated in the supporting cells in the brain, the astrocytes and the microglia (Saljo Moberg, 2001). Blast-induced leakage of marker proteins from neurons and glial cells was also detected (Säljö et al., 2003). Today, with the increased incidence of mild TBI in military conflicts, there is renewed interest in understanding the mechanism of this injury. This review
S84
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
describes our research which adds to the increasing body of evidence supportive of a pathophysiologic basis for mTBI (Chavko et al., 2007; Moochhala et al., 2004; Saljo et al., 2008; Saljo et al., 2009; Warden, 2006). Low-level blast and brain injury Our research efforts were prompted by a paucity of information regarding the early blast wave–brain interactions leading to the cellular and molecular changes described by Säljö (PhD thesis, 2001). We also describe the pathophysiologic changes leading to injury and dysfunction. The studies were undertaken to explore the risk for brain injury from multiple exposures of blast in the peri-threshold regions of military occupational standards. Porcine and rodent models were used to investigate the interaction between the blast wave and the body. The character of the pressure curve recorded in the brain is compared with that in air and water. We have also attempted to determine the contribution of each parameter of the pressure time function in air and water to causation of brain injury. This allows us to propose a mechanistic hypothesis for blast related TBI. Changes in the macroscopic and microscopic appearance of the brain, the intracranial pressure and cognitive function are also described.
higher frequencies in that of the rifle (Saljo et al., 2008). Though the exact mechanisms are not yet clear, the transmission of low frequencies into the brain appears more efficient than that of high frequencies, an observation supported by results from under water experiments (see below). This efficiency can be assessed using the brain/air Pmax ratio, which decreases with increasing high frequency content (Table 1). The brain/air ratios of other parameters of the pressure time function, such as the A-duration (the duration of the first positive peak) and the impulse (integration of the pressure time curve), were between 1.0 and 2.0, while the ratios for the Pmax were between 1.0 and 0.4 (Tables 1 and 2). The parameter, which differed most markedly between brain and air, was the rise time. The brain/air ratio for the rise time was in the range of 10 for the various weapons (Table 2). The contribution of head acceleration to injury induction was determined with high-speed video and an accelerometer, which was aligned with the axis of impact and attached to the animal's head (Saljo et al., 2008). At the levels of blast overpressure used in our experiments, the acceleration of the head was very small and not considered to be causative to the measured pathophysiologic and functional changes. Bioelectrical impedance
Characterization of pressure curves in air and brain Anesthetized pigs were placed in the gunner's position of weapons and received 3 consecutive exposures (Saljo et al., 2008). The following weapons and maximal peak pressures (Pmax expressed in kPa) were used: a howitzer (Haubits F77, 9 or 30 kPa), a bazooka (Karl Gustaf, 45 kPa) and an automatic rifle (AG-90, 23 kPa). The pressure levels in air were measured at the gunner's position. During blast exposure, the pressure as a function of time was recorded in the brain and in air, close to the head of the animal (Table 1, Saljo et al., 2008). A hydrophone (diam. 9.5 mm) or a pen-shaped piezo-electric sensor was used for recordings in air. Hydrophones were routinely used as pressure sensors in the pig brain (Saljo et al., 2008); however, occasionally, miniaturized optic probe sensors (diameter 0.42 mm. Samba, Saljo et al., 2009) were used in parallel. The most striking finding was that the pressure recorded in the brains had a Pmax, which was of similar or slightly reduced magnitude as that in air close to the head. The brain/air Pmax ratio was high for the bazooka and the 30-kPa howitzer but low for the automatic rifle (Table 1). Our conclusion of this was that the scalp, skull bone and cerebrospinal fluid, which separate the brain from the surrounding air, do not constitute an appreciable protection for the brain. While the pressure time curves in brain were very similar to those in the air after exposure to the howitzer, this similarity was reduced with the bazooka and was even less evident with the rifle (Saljo et al., 2008). The frequency distribution of the blast wave differs among the weapons: lower frequencies contained in the howitzer blast and
Bioelectrical impedance is used as a means to detect cerebral edema following hypoxia (Klein and Krop-Van Gastel, 1993; Verheul et al., 1994; Williams et al., 1991; Lingwood et al., 2003) and TBI (Harting et al., 2010). The bioelectrical impedance (Olsson et al., 2006) was measured in pilot experiments on anaesthetized pigs that were exposed in the gunner's position (bazooka, Pmax 45 kPa). In exposed animals, the electrical impedance in the brain increased after a few seconds and reversed within a few minutes in most experiments but occasionally remained elevated for 20–30 min. (Fig. 1). This pattern was similar to the changes in brain impedance during hypoxia (Lingwood et al., 2003). We conclude that the increased bioelectrical impedance support the microscopic findings and the increased ICP and is indicative of the development of brain edema following blast exposure. Neuropathology Pigs were sacrificed 3–7 days after exposure to weapon detonation and the brains were fixed in formalin by perfusion of the vascular system. While the surface of the brains from the control animals was consistently pale, 30% of the brains from the animals exposed to the bazooka or the rifle displayed, on the surface, erythematous regions of varying size and intensity. Microscopical examination revealed a significant increase in the amount of erythrocyte-filled capillaries in the erythematous regions, which we interpreted as a sign of brain edema (Saljo et al., 2008), as previously described in another model of
Table 1 Blast characteristics (Pmax and frequency distribution) in air or water close to the animals head and in the frontal cerebral cortex of a pig in crew positon for the weapons and in water, 96 m from the explosive. Pmax
Weapons in air Karl Gustav Haubits 30 kPa Haubits 9 kPa Automatic rifle Explosives under water (250 g, 96 m) Detonation pulse 1st secondary pulse 2nd secondary pulse
Frequency distribution
kPa
kPa
Ratio
Air 42 30 9 23 Water 482 77 33
Brain 28 20 4 10 Brain 51 25 16
Brain/air 0.7 0.7 0.4 0.4 Brain water 0.1 0.3 0.5
Air Medium Low Low Medium Water High Medium Medium
Brain Low Low Low Low Brain Low Low Low
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
S85
Table 2 Blast characteristics (A-duration, impulse and rise time) in air or water close to the animals head andin the frontal cerebral cortex of a pig in crew positon for the weapons and in water, 96 m from the explosive. Weapons in air
Karl Gustav Haubits 30 kPa Haubits 9 kPa Automatic rifle Explosives under water (250 g, 96 m) Detonation pulse 1st secondary pulse 2nd secondary pulse
A-duration
Impulse
Air (ms)
Brain (ms)
2 5 2 1.5
4 8 3 1.3
Ratio Brain/air 2 1.6 1.5 0.9
Rise time
Air (Pa × s)
Brain (Pa × s)
30 59 9 7 Water 51
47 79 10 6 Brain 21
TBI (Foda and Marmarou, 1994). Foda described parenchymal microvascular changes consisting of brain edema and vascular congestion with “sludging” of red blood cells. Vasospasm may be of special importance at exposure to higher levels of blast (Armonda et al., 2006). Microscopical examination also showed small subarachnoidal and brain tissue hemorrhages in pigs exposed to the howitzer at 30 kPa, the bazooka or the rifle (Fig. 2). Hemorrhages were found particularly in the occipital lobe, in the cerebellum and the medulla oblongata (Saljo et al., 2008).
Ratio
kPa/ms
kPa/ms
1/(brain/air)
970 1097 304
76 87 25
13 13 12
Water 51972 1341 390
Brain 5005 280 83
1/(brain/water) 10 5 5
Brain/air 1.6 1.3 1.1 0.9 Brain/water 0.4
the detonation pulse, while such frequencies were virtually lacking in the secondary pulses (Saljo et al., 2008). Hardly any frequencies above 1000 Hz appeared to be transmitted into the brain. Furthermore, there was little difference in the brain in frequency distribution after exposure to the detonation and the secondary pulses and frequencies around 100 Hz dominated. The brain/water ratio for the rise time was 10 for the detonation pulse while 5 for the secondary pulses. This is probably due to the predominantly low frequencies in the secondary pulses but not in the detonation pulse (Saljo et al., 2008).
Characterization of pressure waves in water and brain Transmission of pressure waves in the body
Intubated pigs, connected to an air tank, were placed 5 m below the surface in 10-m deep water in lake Vättern (Sweden) and exposed to 60, 125 or 250 g explosives, detonated at distances of 66–96 m (Saljo et al., 2008). At these distances, the recordings show a primary detonation pulse, which is followed by secondary pulses of smaller magnitude, the first two after 140 and 250 ms, respectively. The pulses have a rapid peak, followed by oscillations of decreasing magnitude. The maximal response in the brain differed considerably from the detonation pulse in water outside the head, while the responses in the brain were similar to those in water during the secondary pulses. This was reflected in the brain/water Pmax ratios, which were 0.1 for the detonation pulse and 0.3 and 0.5 for the first and second secondary pulse, respectively (Table 1; Saljo et al., 2008). The frequency distribution of the pulses appeared to play a similar role as for the abovementioned pulses from different weapons in air. In water, frequencies in the 5000- to 10,000-Hz range were present in
The experiments were undertaken in order to understand the transfer of pressure from a local blast exposure to the abdomen or to the head. In these experiments, pigs were exposed locally by means of a shock tube with a diameter of 0.2 m (Saljo et al., 2008). Pressures in the abdomen and the brain were compared after local exposure of either the abdomen or the head. When a Pmax of 30 kPa was recorded within the tube, a Pmax of 15 kPa was registered in the air outside the abdomen, 20–30 mm lateral to the tube opening. A similar Pmax, 15 kPa, was recorded inside the abdomen. In the same experiments, the Pmax in air outside the head was 1.8 kPa since the upper body could not be completely shielded. The registered Pmax in the brain was 0.5 kPa. This pressure could be transmitted exclusively from the air outside the head, as judged from brain/air Pmax ratios during free field exposures (Table 1). At the level of blast exposure used, the results do
Fig. 1. Time course of changes in bioelectrical impedance in the brain during 3 successive exposures of pigs to blast overpressure from the firing of a bazooka (Pmax 45 kPa). 1, 2 and 3 indicate the timing of exposures.
Fig. 2. Light microscopy of section from the left occipital cortex of the pig brain exposed to blast overpressure from the firing of a bazooka. Stained for unspecific peroxidase. Red blood cells in brain tissue outside blood capillaries and erythrocyte-filled capillaries.
S86
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
not indicate an appreciable transmission of a pressure from the abdomen to the brain. In other experiments, the vertex of the head was locally exposed in a similar fashion as the abdomen. The Pmax in air outside the head was then 22 kPa and that within the brain 9 kPa, again in agreement with the air/brain ratios during free field exposures (Table 1). The peak pressure in air outside the abdomen was now 2 kPa and that in the abdomen 0.8 kPa. The results do not support a transfer of pressure from the head to the abdomen. Intracranial pressure (ICP) In this and the following section, male Wistar rats (230–330 g) were exposed to blast in the shock tube, mentioned in the previous section. The animals were placed in the tube, 0.25 m from the free opening of the shock tube and had the long axis of their body parallel to and its head facing the direction of the blast front. Pulses with Pmax of 10, 30 and 60 kPa and duration of 4–6 ms as measured in the air in close proximity to the head were generated with pressures of 0.2, 0.6 and 1.2 bar, respectively (Saljo et al., 2009). The ICP was measured with the miniaturized optic probe (Samba) at 0.5, 2, 6 and 10 h and 1, 2, 3, 5 and 7 days after exposure 10 kPa, 30 kPa or 60 kPa. In addition to the pressure curve recorded in the brain during blast exposure, there is a slowly rising, long lasting, although finally reversed, increase in the pressure, labeled ICP. However, an increase in the ICP is not a specific reaction to blast exposure (Saljo et al., 2009). The mean ICP in control rats was close to 6 mm Hg (Saljo et al., 2009), whereas the ICP in rats exposed at 60 kPa increased significantly after 30 min (Fig. 3). A corresponding increase was not seen until after 2 h in rats, exposed at 30 kPa and as late as after 6 h in rats exposed at 10 kPa. The 30- and the 60-kPa groups reached a peak value for ICP 10 h after exposure, at a level of 11 and 16 mm Hg, respectively. There was a slow return towards control levels during the subsequent days. Rodent studies, simulating the concussions sustained by football players (Viano et al., 2009), provided an opportunity to compare ICP measurements from a blast and non-blast model of TBI in rats of similar weight and strain (Fig. 3). The concussive trauma resulted in a peak value and peak time, which is similar to that after a blast exposure at 60 kPa (unpublished). Cognitive function For tests of cognitive function, rats were trained and tested in a circular, water filled pool (Morris, 1984; Saljo et al., 2009). Their task was to find a platform, 10 mm under the water level, which was mounted in one of the pool quadrants. Rats were trained for 6 days to find the platform within 10 s. Exposure to a blast overpressure of 10 kPa
on the 7th day increased the time to find the platform to 17 s 2 days later. The time to find the platform increased even more in rats exposed to 30 kPa, i.e. 21 s (Saljo et al., 2009). The speed of swimming was similar in all rats. This increase in time to find the platform in blast exposed rats is indicative of cognitive dysfunction.
Discussion The exact mechanism of how a blast wave interacts with the body is unknown. Several modes have been hypothesized to include formation of an electromagnetic field, direct transmission of the blast wave to the brain through the skull, the formation of a secondary shear wave (Moore et al., 2008), transmitted pressure from the vasculature and soft tissues of the thorax (Cernak and Malizevic, 1997) and acceleration of the head. The different modes may operate singly or in concert and one mechanism may predominate in lowlevel blast versus another at higher level blast exposure. Low, occupational levels of blast affect both cognitive performance and ICP (Saljo et al., 2009). Similar effects are observed in our model for non-blastproduced, sport-related concussion (Viano et al., 2009). It has been assumed that only the auditory system is sensitive to the high frequency components in a blast wave, whereas the nonauditory system was more so to lower frequency (Patterson and Hamernik, 1997). Little is known about the brain's response to the frequency distribution of a blast wave. Our studies demonstrate for the first time that the frequency spectrum of a blast wave is important in the transmission of energy after primary blast exposure (Table 1), a result that has been verified in a recent study (Chavko, 2009). Animals exposed to powerful, low frequency weapons (i.e. howitzer) showed the highest brain/air Pmax ratios. In our underwater experiments, the relation between the brain/water Pmax ratio and frequency distribution was evident; however, injury correlates are not possible to make. Furthermore, frequencies above 1000 Hz do not transmit to the brain (Saljo et al., 2008). Our observation, that the lower frequencies contained in a blast wave appear to be transmitted more readily to the brain, is consistent with studies of blast pressure transmission from air into the abdomen of pregnant sheep (Gerhardt et al., 2000). Our studies indicate that blast overpressure is mostly transmitted directly to the brain in air exposure, not via the thorax or abdomen. This is in contrast to recent studies (Cernak and Malizevic, 1997; Cernak et al., 2001). In our experiments with the bazooka and the howitzer, the pressure transmission into brain was high, the Pmax air/ brain ratio being 0.7 (Table 1). Clemedson was the first to report that there are no appreciable Pmax differences in air versus brain during blast exposure (Clemedson and Pettersson, 1956) and recent studies (Chavko et al., 2007; VandeVord, 2009) also corroborate our findings.
Fig. 3. Time course of the intracranial pressure (ICP, mm Hg) in rats during 1 week after blast exposure at 10 (●), 30 (■) and 60 kPa (▲). The results are compared with ICP measurement (unpublished) after translational acceleration of the head produced in a rat concussive model simulating professional football brain injuries (Viano et al., 2009). In comparison with blast exposure at 60 kPa, the peak ICP values and peak time course are similar. Baseline levels are given as time zero. Data are presented as mean ± SEM.
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
A dose–response increase in ICP was shown after blast exposure and an inverse relationship was noted between the level of blast and the time of onset of ICP changes, i.e. in agreement with findings in models of closed head injury (Engelborghs et al., 1998). We hypothesize that the increased ICP after blast exposure is related to brain edema (Fig. 3). However, there are a number of possible causative factors for increased ICP, such as hemorrhage, CSF outflow obstruction and BBB damage (Hoane et al., 2006; Vink et al., 2003). The immediate and late increases in ICP elevation are being attributed to vascular changes and more complex changes, respectively (Engelborghs et al., 1998; Fritz et al., 2005; Jamali et al., 1998; Rooker et al., 2003). ICP measurements seem to be a more sensitive method to detect mTBI than markers for neuropathology (Teranishi et al., 2009), both with respect to the onset of injury and exposure conditions. The measurement of electrical impedance in the brain is a promising means to detect cerebral edema following hypoxia-ischemia (Klein and Krop-Van Gastel, 1993; Verheul et al., 1994; Williams et al., 1991; Lingwood et al., 2003) and TBI (Harting et al., 2010). The usefulness of animal studies is limited by the uncertainty of scaling relationships, validity of laboratory models and species differences. There are no studies comparing species differences in neurological susceptibility, anatomy, tissue composition, stress strain responses, etc. Furthermore, differences in blast generating sources (explosives, shock tube), exposure to multiple effects of blast (secondary, tertiary, quaternary, etc), environmental design (free field, enclosures) and orientation of the subjects make comparison of animal studies difficult. Few studies have addressed the questions of the influence of the frequency spectra of blasts (Chavko, 2009) or the effects of repeated exposures. The correlation of blast levels with the neuropathology of brain injury induced by low levels of blast is also fairly unexplored (Moochhala et al., 2004; Saljo et al., 2009; Teranishi et al., 2009). Our results pose questions with regard to the validity of current occupational standards since brain injury was not a factor considered during their development. More studies are needed to establish routes and mechanisms of pressure transmission from the blast wave to the brain and to identify initial and secondary mechanisms of injury. Also, the role of the blast wave parameters and injury correlates of brain injury development should be explored further. The creation of a comprehensive pathologic scoring system would assist in comparing results from different biological models. Lastly, advancement of numerical models are needed to better deal with the multitude of experimental variables, which ultimately may result in protective strategies. Conclusions 1. The peak pressures (Pmax) in the brain, elicited by the blast from the firing of military weapons (Pmax 23–45 kPa), have a similar magnitude as that registered in air close to the head. Corresponding measurements during the detonation pulse from explosives under water showed a Pmax in the brain, which was only 10% of that in water outside the head. 2. The rise time of the pressure curve was 10 times longer in the brain as compared with the blast in air outside the head during firing of military weapons 3. The lower frequencies in the blast wave appeared to be transmitted more readily into the brain than the higher frequencies. 4. When animals are exposed to low levels of blast in air, the blast wave was mostly transmitted directly to the brain, not via the thorax or abdomen 5. Low levels of blast cause • brain edema, as indicated by increased bioelectrical impedance (bazooka 45 kPa) • an increase in the intracranial pressure, peaking 6–10 h after exposure (10, 30 and 60 kPa) • small brain hemorrhages (bazooka, automatic rifle)
S87
• impaired cognitive function 2 days after the exposure (10 and 30 kPa) Conflict of interest statement The authors declare that there are no conflicts of interest. Acknowledgments The authors want to express their gratitude to Svante Höjer, Samba Sensors AB, Västra Frölunda, Sweden, for sharing with us his knowledge on pressure sensors and pressure recording. We also thank Torsten Olsson, professor emeritus, Department of Signals and Systems, Chalmers University of Technology, Göteborg, Sweden, for introducing us into the field of electrical impedance and for carrying out experiments with us. The study was supported by grants from the Swedish Armed Forces/FMV and by Svenska Militärläkareföreningen. References Armonda, R.A., Bell, R.S., Vo, A.H., Ling, G., DeGraba, T.J., Crandall, B., Ecklund, J., Campbell, W.W., 2006. Wartime traumatic cerebral vasospasm: recent review of combat casualties. Neurosurgery 59, 1215–1225 discussion 1225. Aschan, Sälde, 1954. Medicinska undersökningar av närpvskyttar. Försvarets Forskningsanstalt, Sundbyberg, Sweden. Dnr 10043-8112. Celander, H., Clemedson, C.-J., Ericsson, U., Hultman, H., 1954. The use of compressed air operated shock tube for physiological blast research. Försvarets Forskningsanstalt Sundbyberg, Sweden. Dnr 1690-8112. Cernak, I., Malizevic, Z., 1997. Indirect neurotrauma caused by pulmonary blast injury: development and prognosis. Int. Rev. Armed Forces Med. Serv. 70, 114–120. Cernak, I., Wang, Z., Jiang, J., Bian, X., Savic, J., 2001. Ultrastructural and functional characteristics of blast injury-induced neurotrauma. J. Trauma 50, 695–706. Chavko, M., 2009. Transfer of energy inside the body after exposure to blast. Military Health Research Forum, Kansas City, MO, USA. Chavko, M., Koller, W.A., Prusaczyk, W.K., McCarron, R.M., 2007. Measurement of blast wave by a miniature fiber optic pressure transducer in the rat brain. J. Neurosci. Methods 159, 277–281. Clemedson, C.J., 1951. Detonation injuries. Svenska Lakartidning 48, 1337–1343. Clemedson, C.J., 1954. Correlation between respiratory phase and extent of lung damage in air blast injury. J. Appl. Physiol. 7, 38–42. Clemedson, C.-J., 1956a. Shock wave transmission to the central nervous system. Acta Physiol. Scand. 37, 204–213. Clemedson, C.J., 1956b. Shock wave transmission to the central nervous system. Acta Physiol. Scand. 37, 2–3 Fasc. Clemedson, C.J., Pettersson, H., 1956. Propagation of a high explosive air shock wave through different parts of an animal body. Am. J. Physiol. 184, 112–119. Engelborghs, K., Verlooy, J., Van Reempts, J., Van Deuren, B., Van de Ven, M., Borgers, M., 1998. Temporal changes in intracranial pressure in a modified experimental model of closed head injury. J. Neurosurg. 89, 796–806. Foda, M.A., Marmarou, A., 1994. A new model of diffuse brain injury in rats. Part II: Morphological characterization. J. Neurosurg. 80, 301–313. Fritz, H.G., Walter, B., Holzmayr, M., Brodhun, M., Patt, S., Bauer, R., 2005. A pig model with secondary increase of intracranial pressure after severe traumatic brain injury and temporary blood loss. J. Neurotrauma 22, 807–821. Gerhardt, K.J., Abrams, R.M., Huang, X., Griffiths, S.K., Peters, A.J., 2000. Intra-abdominal sound pressure levels during impulse noise exposure in sheep. Mil. Med. 165, 153–156. Harting, M.T., Smith, C.T., Radhakrishnan, R.S., Aroom, K.R., Dash, P.K., Gill, B., Cox Jr., C.S., 2010. Regional differences in cerebral edema after traumatic brain injury identified by impedance analysis. J. Surg. Res. 159, 557–564. Hoane, M.R., Kaplan, S.A., Ellis, A.L., 2006. The effects of nicotinamide on apoptosis and blood–brain barrier breakdown following traumatic brain injury. Brain Res. 1125, 185–193. Jamali, S., Bodjarian, N., Vigue, B., Mazoit, J.X., Samii, K., Tadie, M., 1998. Increase in the chronically monitored cerebrospinal fluid pressure after experimental brain injury in rats. Brain Inj. 12, 525–536. Klein, H.C., Krop-Van Gastel, W., 1993. Prediction of specific damage or infarction from the measurement of tissue impedance following repetitive brain ischaemia in the rat. Neuropathol. Appl. Neurobiol. 19, 57–65. Lingwood, B.E., Dunster, K.R., Healy, G.N., Ward, L.C., Colditz, P.B., 2003. Cerebral impedance and neurological outcome following a mild or severe hypoxic/ischemic episode in neonatal piglets. Brain Res. 969, 160–167. Moochhala, S.M., Md, S., Lu, J., Teng, C.H., Greengrass, C., 2004. Neuroprotective role of aminoguanidine in behavioral changes after blast injury. J. Trauma 56, 393–403. Moore, D.F., Radovitzky, R.L., Shupenko, L., Klinoff, A., Jaffe, M.S., Rosen, J.M., 2008. Blast physics and central nervous system injury, pp. 243–250. Morris, R., 1984. Developments of a water maze procedure for studying spatial learninng in the rat. J. Neurosci. Methods 11 (1), 47–60. Olsson, T., Broberg, M., Pope, K.J., Wallace, A., Mackenzie, L., Blomstrand, F., Nilsson, M., Willoughby, J.O., 2006. Cell swelling, seizures and spreading depression: an impedance study. Neuroscience 140, 505–515.
S88
A. Säljö et al. / NeuroImage 54 (2011) S83–S88
Patterson Jr., J.H., Hamernik, R.P., 1997. Blast overpressure induced structural and functional changes in the auditory system. Toxicology 121, 29–40. Rooker, S., Jorens, P.G., Van Reempts, J., Borgers, M., Verlooy, J., 2003. Continuous measurement of intracranial pressure in awake rats after experimental closed head injury. J. Neurosci. Methods 131, 75–81. Saljo Moberg, A., 2001. Impulse noise and brain injury: an experimental study in rats (PhD thesis), University of Gothenburg, Sweden. Saljo, A., Bao, F., Haglid, K.G., Hansson, H.A., 2000. Blast exposure causes redistribution of phosphorylated neurofilament subunits in neurons of the adult rat brain. J. Neurotrauma 17, 719–726. Saljo, A., Bao, F., Jingshan, S., Hamberger, A., Hansson, H.A., Haglid, K.G., 2002a. Exposure to short-lasting impulse noise causes neuronal c-Jun expression and induction of apoptosis in the adult rat brain. J. Neurotrauma 19, 985–991. Saljo, A., Bao, F., Shi, J., Hamberger, A., Hansson, H.A., Haglid, K.G., 2002b. Expression of c-Fos and c-Myc and deposition of beta-APP in neurons in the adult rat brain as a result of exposure to short-lasting impulse noise. J. Neurotrauma 19, 379–385. Säljö, A., Huang, Y.-L., Hansson, H.-A., 2003. Impulse noise transiently increased the permeability of nerveand glial cell membranes, an effect accentuated by a recent brain Injury. J. Neurotrauma 20, 787–794. Saljo, A., Arrhen, F., Bolouri, H., Mayorga, M., Hamberger, A., 2008. Neuropathology and pressure in the pig brain resulting from low-impulse noise exposure. J. Neurotrauma 25, 1397–1406.
Saljo, A., Svensson, B., Mayorga, M., Hamberger, A., Bolouri, H., 2009. Low levels of blast raises intracranial pressure and impairs cognitive function in rats. J. Neurotrauma 26, 1345–1352. Teranishi, K., Chavko, Mikulas, Adeeb, Saleena, Carroll, Erica, McCarron, Richard, 2009. Blast-induced neuropathological changes in brain in relation to different orientation to blast. International and National Neurotrauma Societies, Santa Barbara, CA, USA. VandeVord, P., 2009. Understanding of Shock wave transmission to the brain. International State of the Science Meeting on No-Impact Blast-induced Mild Brain Injury. Herndon, VA, USA. Verheul, H.B., Balazs, R., Berkelbach van der Sprenkel, J.W., Tulleken, C.A., Nicolay, K., Tamminga, K.S., van Lookeren Campagne, M., 1994. Comparison of diffusion-weighted MRI with changes in cell volume in a rat model of brain injury. NMR Biomed. 7, 96–100. Viano, D.C., Hamberger, A., Bolouri, H., Saljo, A., 2009. Concussion in professional football: animal model of brain injury—part 15. Neurosurgery 64, 1162–1173 discussion 1173. Vink, R., Young, A., Bennett, C.J., Hu, X., Connor, C.O., Cernak, I., Nimmo, A.J., 2003. Neuropeptide release influences brain edema formation after diffuse traumatic brain injury. Acta Neurochir. Suppl. 86, 257–260. Warden, D., 2006. Military TBI during the Iraq and Afghanistan wars. J. Head Trauma Rehabil. 21, 398–402. Williams, C.E., Gunn, A., Gluckman, P.D., 1991. Time course of intracellular edema and epileptiform activity following prenatal cerebral ischemia in sheep. Stroke 22, 516–521.