Traditionally, this model uses gravitational force and a weight of a known mass that falls over the skull of the experimentation animal (Fig. 1A). Thus, after sedating the rodent, it is placed and fixed on a horizontal surface to avoid any head movements after the weight has been dropped. The magnitude of the injury depends on the mass of the weight and the height it was dropped from.15

Figure 1.

Models of cranioencephalic traumas in rodents. (A) Model of weight falling over the skull. (B) Impact-acceleration model due to free falling weight: Marmarou weight drop injury model. (C) Model of controlled cortical impact. (D) Fluid-percussion model.

(0.16MB).

Source: Morales DM, Marklund N, Lebold D, Thompson HJ, Pitkanen A, Maxwell WL, et al. Experimental models of traumatic brain injury: do we really need to build a better mousetrap? Neuroscience. 2005;136:971–89. doi:10.1016/j.neuroscience.2005.08.030.

Although this model is easy to use, inexpensive and fast, it is not widely used because it has a great variability since the skull breaks into several bone fragments of different sizes depending on the energy interchange used in the experiment. Another setback here is that there is possibility of weight rebound after impacting the animal's skull.

In order to optimize this model, several researchers have been developing a series of modifications in an attempt to increase reproducibility and achieve a specific brain damage maybe by modifying the severity of the TBI sustained.16

This model generates a focal-type experimental traumatic brain injury. It has been tested in rodents only. To develop it we need to conduct one craniotomy procedure in such a way that a piston pumps compressed air over the brain.13

This device consists of a pneumatic cylinder usually connected using a 4–5cm in length and 1cm in diameter threaded system to an impactor (Fig. 1C). The intensity of the traumatic brain damage sustained will depend mainly on the speed of this impactor and on the deformation depth generated. In most studies, the speed used is between 0.5m/s and 1.0m/s and the depth of cortical deformation is modulated using a bar that holds the impact system. The duration of the impact is usually adjusted between 25m/s and 250m/s.13

The histological injuries generated by this model after sustaining mechanical impact are focal, damage intracranial pressure, and compromise the cerebral perfusion pressure.17,18 In general, the brain damage achieved by this model is characterized by an area of focal necrosis in the cortex surrounded by an area of brain swelling similar to the bruises we can see in a patient with TBI.19 On the other hand, different studies have shown that, to a lesser extent and after the controlled cortical impact, it is also possible to cause, among others, DAIs at the subcortical, cerebellum, and midbrain levels.20,21

There are several advantages that this model has to offer. In the first place, it allows us to efficiently control the different mechanical variables that unfold here in such a way that we can make a more precise quantification of the damage sustained by the brain.

Secondly, this model allows us to evaluate the effects produced by the post-traumatic brain swelling and see what role the BBB actually plays here. Basically, using this model, Kiening et al. described the role that acuaporin-4 (AQP4) plays in the development of brain swelling.22 Thirdly, this model allows us to analyze the BBB integrity after sustaining the TBI and explore different pharmacological therapies that may run through the BBB. Thus, the activated signaling pathways after sustaining the TBI are guaranteed here as it happens with melatonin or with a mimetic peptide derived from the ApoE.23–25 Fourthly, the lower mortality rate of this model since it does not damage the brainstem allows us to know what mid-term cellular changes occur when sustaining TBIs at both molecular and genetic levels.26–28 The use of microarray techniques allows us to analyze alterations in a great number of genes after applying this controlled cortical impact model. Gene sequences associated with inflammatory processes (IL1α, IL1β and COX2, among others) have been described with functions in the cytoskeleton (synaptotagmin 4, MAP2, GFAP, nestin and vimentin, among others), signal transduction pathways (ERK and MAPK, among others), functions of cell-cycle regulation pathways (such as cyclin),26 and genes regulating calcium-signaling pathways (calcium/calmodulin dependent-protein kinase C beta and protein kinase 1β).28 This model has also been used to study the pathophysiological changes associated with the GABAA receptor in cultures of dented granulosa cells.27 Lastly, from the clinical point of view, the role played by repeated brain concussions usually associated with contact sports (boxing, rugby, etc.) has been studied too, which allows us to analyze, at an experimental level, anxiety and motor issues derived and identify corticosterone as a possible biomarker of these injuries.29

This model generates transient intracranial hypertension by reproducing the compression effect generated by a space occupying lesion (SOL) in the brain and swelling after surgical assessment. In this model we perform a small hole through which an elastic probe is placed in the epidural space. Once inserted the probe is inflated like a balloon causing the effect of a SOL (Fig. 2).

Figure 2.

Model of intracranial hypertension due to space occupying lesion in a rat after craniotomy procedure.

(0.04MB).

Source: Lopez-Valdes et al.33

Traumatic SOLs are responsible for intracranial hypertension in between 30% and 50% of all TBIs.30 Such a sustained increase of intracranial pressure is recognized as one of the main mechanisms that casts a shadow on the prognosis of patients with TBIs and, same as it happens with the aforementioned controlled cortical impact model, there is a loss of BBB integrity that ends up leading to brain swelling.

This model is highly reproducible, inexpensive and its methodology is easy, since the variables of volume and filling/evacuation time can be easily controlled here. However, it is known that usually severe traumatic brain injuries are not a single entity given the presence of a high-volume SOL, which makes it hard to extrapolate these results to our routine clinical practice.30–33

On many occasions, the lesions induced by a TBI are due to the fast acceleration of the cranioencephalic region without a direct impact from an object. The oscillation and rotation sustained by the brain inside the skull deforms the brain and leads to a DAI. The movement used here is rotational and at the beginning there is a prolonged acceleration followed by a sudden deceleration in a short period of time. Hence the expression “acceleration–deceleration” at brain level.

This type of indirect energy interchange has been recreated in different animal models of TBI such as pigs, sheep, and primates.7,34 Several studies with small pigs of the Hanford type have shown that the rotational acceleration of the head leads to DAIs with possible damage to the brainstem and even a state of coma following damage to the ascending reticular system.35,36

In general, these models have been able to reproduce effectively the pathophysiological mechanisms that occur in patients with DAI following TBIs. However, their main setback is their high cost since they use big animals, have many technical requirements and require state-of-the art facilities. Lastly, we should mention here that, due to the different brain mass of the different species available, extrapolating the results obtained in each study depends on the animal used. Also, it is difficult to translate these findings into our routine clinical practice.

The model developed by Marmarou et al. is one of the most widely used in rodents due to its low cost and high reproducibility.37 It is an impact-acceleration model of a free-falling weight over the skull. As Fig. 1B shows the animal is placed on a foam rubber surface with its skull exposed. Then, one metallic disk is fixed to the rodent in order to reduce the possibility of cranial fracture and appearance of focal injuries. The degree of brain injury sustained is closely related to the size of the weight used and the height the weight was dropped from.

One of the biggest advantages of this model is that it simulates traumatic DAIs and subsequent brain swelling.13,38 The process of swelling is associated with BBB tears that trigger neuroinflammatory responses inside the intracranial compartment.39 That is why this model is used to study the BBB integrity and functionality and to assess all possible molecules or therapies to improve functional results following a TBI.40 In sum, the Marmarou weight drop injury model appropriately simulates the TBIs sustained in a car crash and has proven useful for the study of repeated TBIs.38,41

The setbacks of this model have to do with how limited the researcher is when it comes to controlling the impact biomechanical forces at stake, which can lead to a weight rebound effect over the skull, thus generating uncontrolled injuries after the first energy interchange.

This model was initially designed to mimic focal brain injuries. But even though the theoretical mechanics of this model is oriented toward the origin of focal damage, in some cases, the energy released in the impact can produce DAIs, which is why it is considered a mixed traumatic brain damage model.42–44

The use of this model requires a prior craniotomy in such a way that energy interchange occurs following the exertion of pressure through a fluid directly over the dura mater of the experimentation animal. In order to exert this pressure, a pendulum linked to a piston is used to move a certain mass of fluid originating the pressure pre-established in the experiment (Fig. 1D). This impact causes displacement and damage due to the deformation of neuronal tissue. The magnitude of the damage depends on the amount of fluid used.13,45

In the first models used, the impact occurred at brain midline level.46 Then, other working groups started to cause the injury in one of the two hemispheres of the brain, the so-called “fluid-induced lateral perfusion model”. It has been confirmed that the lateralization of damage allows us to study and compare the extent of the injury sustained in both hemispheres separately. Also, small variations in the position of craniotomy can cause significant differences in the severity of brain damage.47 This fluid-induced lateral perfusion model produces focal cortical concussions and diffuse subcortical damage. Using this mode, Hicks et al. showed the persistence of a series of degenerative cascades following a TBI in vulnerable areas of the brain such as the thalamus, the ipsilateral hippocampus, the septum pellucidum, the striated septum, and the amygdala.48

The histological manifestations obtained with the fluid-percussion model are multiple and it can reproduce the appearance of intracerebral hemorrhages, swelling, and gradual damage to gray matter, all common injuries of traumatic brain damage in humans.49 Also, in this model we have been able to see an increase of blood pressure and also an increase of intracranial pressure. Although, as we said before, this model tries to simulate focal lesions, it also causes mixed lesions since it triggers processes such as DAIs, neuroinflammation and swelling in the cerebral parenchyma.50,51 This model has been used in studies of memory, cognitive, learning, and behavioral impairment, at experimental stages.50–54

The fluid-percussion model has a series of advantages such as being inexpensive, being capable of recreating cerebral concussions similar to the ones sustained by humans and being highly reproducible. But even though the use of this model is relatively common, its main setback is how difficult it is to maintain an adequate control of calibration and the percussion system since air bubbles may originate inside the fluid. This can lead to a certain variability in the degree of lesion caused. On the other hand, this model does not exclude multiple damage and it can affect other areas such as the brainstem.45,46

This model simulates TBIs induced by the detonation of an explosive. The brain damage induced by this type of traumas is multiple and varied. Explosions can associate multiple inert elements that are launched or make a projectile-like impact in the skull. Also, brain damage can be caused by the shock waves that follow the detonation. That is why when it comes to designing blast-like experimental models, we need to make sure that they are reproducible and take all these aspects into consideration both in isolation and simultaneously. In these blast-like experimental models, TBIs are both focal and diffuse and damage is mixed.55

Within the different blast-like experimental models available today, one of the most widely used ones is ballistic penetration-induced brain damage. This model has been designed to simulate the damage caused by a projectile or any other type of object to the brain55–57 (Fig. 3). Also, there are other blast-like experimental models that specifically combine DAI damage and development of cerebral concussions55,57 (Fig. 4).

Figure 3.

Projectile impact simulation model.

(0.14MB).

Source: Risling and Davidsson.55

Figure 4.

Blast injury model.

(0.25MB).

Source: Risling and Davidsson.55

In another version of this blast model, damage is caused by the shock waves released following the detonation of an explosive. To this end, one metallic cylinder is used with one of its edges closed and in contact with the detonation. On the other side of the tube, the sedated rodent is introduced in a compartment to prevent any movement from happening after the detonation.58,59 Different types of histological phenomena have been recorded following this injury such as DAI, damage to white matter, neurodegeneration and glia activation.60

Blast-like experimental models are usually highly reproducible and they allow us to study multiple responses after sustaining the TBI such as secondary inflammatory responses, concussions or axonal injuries.55,57,58 Depending on the model used, the greatest setback here is that it may be very expensive and depending on the case we are dealing with, very complex due to the type of materials required to implement it.