This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Cerebral concussion is a mild traumatic brain injury (MTBI) which is notoriously difficult to assess with current imaging techniques. A traumatic incident to the head revolves initially around a CT scan, which typically demonstrates an absence of physical findings. However investigations into the outcome of patients with post concussive syndrome (PCS) indicate the presence of diffuse axonal injury (DAI) suggesting that brain injury persists. In this review, a comprehensive search of the literature for assessment of concussive injuries following MTBI was undertaken with an emphasis placed on CT, MRI, PET, SPECT and DTI. CT was found to be lacking in the detection of neuronal damage; however remains the standard for acute injury. Overall it was found that there has been a paradigm shift towards functional studies in the assessment of concussion and axonal injury. These techniques, such as diffusion tensor imaging (DTI), serve as an adjunct to conventional MRI by providing an insight into the detailed functional consequences of concussive injury. Most importantly, the literature demonstrated that there needs to be further investigation into the clinical relevance of axonal injury post concussion and the neurological integrity of athletes with an otherwise normal CT scan. The need to ascertain which technique holds the most promise in the accurate and specific diagnosis of MTBI was also made paramount.
Keywords: Concussion, Diffuse Axonal Injury (DAI), Neuroimaging, Mild Traumatic Brain Injury, Magnetic Resonance Imaging (MRI)
Cerebral concussion is a form of mild traumatic brain injury commonly observed in the young and the athletic. Typically this injury occurs in sporting and recreational events, when the brain undergoes an acceleration or deceleration differentially within the skull1. A concussion to the head manifests itself as either a loss of consciousness (LOC) of less then half an hour or as a period of amnesia and memory deficit lasting less then a day. However it can include any type of altered mental status at the time of injury. On the Glasgow Coma Scale, or GCS, which is a neurological scale defining the conscious state of a patient, the concussed individual is given a score of 13-15, placing cerebral concussion in the category alongside minor brain injury2. Although the impact of a cerebral concussion on the human brain is not yet fully understood, it is known to induce a complex pathophysiological response due to the traumatic biomechanical forces exerted through the injury3. A common consequence of a mild traumatic brain injury such as a concussion is PCS. PCS or post concussive syndrome refers to the concussion-related symptoms often present for at least three months following the initial concussive injury3. This cluster of symptoms, commonly including dizziness, irritability, headache, cognitive dysfunction and fatigue have been debated since the beginning of the nineteenth century over whether their cause is truly organic and structural in nature. Much research over the past decade has been conducted in an attempt to improve the notably limited concussion and PCS management which as it stands, includes a triad of symptom monitoring, physical examination and neurocognitive testing. Though there are cases where lesions are detected post concussion, the majority of athletes presenting with mild traumatic brain injury have negative CT findings4. On conventional MRI sequences too, the only typical findings are small cortical contusions or subdural hematomas, and small white matter haemorrhages that are generally interpreted to be reflective of diffuse axonal injury5. Throughout the literature, there is increasing evidence to suggest that this structural diffuse axonal injury is present following concussive brain injuries, with the severity and extent of axonal damage correlating to the Glasgow coma scale score6. The forces present in the initial concussive injury act to decelerate the brain where the shearing forces of the motion act on the axons of the neurons to stretch, disrupt and eventually separate nerve fibers. Diffusion tensor imaging (DTI) has the capacity to detect these axonal changes by measuring the movement of water in six or more nonlinear directions. This allows the determination of the main water movement direction within axons in white matter5. Even though the conventional modalities of CT and MRI do not detect subtle neuronal change, other functional techniques complement DTI including fMRI, PET, and SPECT, and are yielding modest results, whereas newer techniques such as MR spectroscopy have yet to display their full diagnostic potential. On the basis of this review, the target in terms of imaging in concussion is the identification of diffuse axonal injury. Many diagnostic modalities presented in the current review (MRI, MRS, PET scan, DTI, SPECT scan etc.) hold much promise for uncovering axonal injury following concussion. Before this can be done, we first must determine which modality correlates best with pathologically defined axonal injury after concussion, and thus is best able to serve as a diagnostic criterion standard.
MTBI: Functional Vs Organic
Like other minor head injuries, cerebral concussion is often considered simply a temporary fluctuation in consciousness with an absence of any long term conseqences7. However, in many cases it has been noted that there is an enduring degree of neuropsychological deficits. Most particularly, these are related to attention, working memory and the executive functions associated with the frontal lobe. Studies conducted in 2001 identified a pattern of diffuse activation primarily in the pre-frontal cortex (PFC) of concussed individuals. This was one of many observational studies with evidence suggesting a variety of functional anomalies in concussed individuals which correlate with imaging studies. One such study, conducted by Mathias demonstrated that there was significant reduction in information exchange between the hemispheres of the brain8. A comprehensive study by Zhang however concluded that no consistent findings across advanced brain imaging techniques (fMRI and DTI) were observed, possibly due to the time frame of the study, technological issues with the diffusion weighted imaging and the subtle nature of neuronal damage in concussion9. Other functional studies have also focused on MTBI and its effects on working memory and other functional deficits post injury. In the working memory study, individuals within the MTBI group complained of significantly higher incidence of memory impairment than those without any history of concussion. In addition, the fMRI patterns of brain activation in those with concussion varied to those of the controls, particularly in the right dorsolateral prefrontal cortex and the right lateral parietal region where a heightened activation in cognitive tasks was noted10. This difference in the allocation of cognitive processing recourses was hinted at being due to a possible impairment in the patients with MTBI, with different brain regions being 'recruited' to perform the functions of the damaged areas in their abscence. In another prospective observational study twenty patients with PCS following MTBI had deficits on neuropsychological tests and of these same individuals eleven had identifiable lesions on MRI11.
In the majority of functional studies, the fact that brain activation was generally witnessed as being more widespread in those following injury compared to controls, and the fact that the performance levels of the chosen cognitive tasks were unchanged between the groups leads to a number of conclusions. Namely, that a form of induction or recruitment of neurons was undertaken by other areas of the brain when injured regions were no longer able to function approriately12. This would effectively 'compensate' for the cognitive impairment in regions compromised by the concussive injury.
Diffuse Axonal Injury
The white matter tracts of the axonal processes of the neurons are highly susceptible to damage from shear and stress forces experienced during a concussive injury. This is primarily because of their viscoelastic properties and their highly rigid organisation within the white matter. The name diffuse axonal injury is used to signify the scattered and haphazard pattern of injury the axons experience. Unfortunately, it is generally microscopic in nature, rendering it essentially invisible to current medical imaging modalties13. Despite the brain being quite compliant, when a severe degree of force is applied, the violent stretch and tensile elongation of the axons disrupts the cytoskeleton and the normal axonal organisation . Initially following the concussive injury, this may take hours or even days to manifest itself14. In addition, soduim channels are compromised resulting in a large influx of sodium ions, followed by swelling as water follows the ions osmotically. The last stage in the progression of axonal damage, alongside progressive disorganisation of the axonal cells, is termed axotcoomy. This is a disconnection of the axon with a 'terminal clubbing' bulb being formed. Pathologically this is the final stage and results in the neuron being permanently unable to communicate. However above all, the predominant and most characteristic pathology of DAI is the microscopic axonal swellings, which have been shown to be extremely difficult to illuminate with non-invasive methods of medical imaging13
The initial workup for patients with concussion, particularly of athletes, involves a triad of physical examination, symptom monitoring and neurocognitive testing. This may or may not be accompanied by a CT scan of the brain to exclude gross pathology. Neurocognitive testing involves a number of cognitive tasks, for example recalling words and objects, simple tests of reaction time and in some cases can include motor and balance testing. This is advantageous as concussed patients frequently suffer from unsteadiness and imbalance post injury. These tests have an advantage in that they are non-invasive and can detect occult brain injury and subtle functional deficit when a more complex neurological examination fails to reveal any underlying condition. However neurocognitive testing relies on having a cooperative patient, and one with no underlying learning difficulties or neurological conditions. In addition, finding a baseline value for ambiguous cognitive function is near impossible, further putting this investigative method at a disadvantage.
The initial workup should be supplemented with imaging studies should symptoms progress, and
Of all modalities used in the imaging of the brain, Computed Tomography (CT) is the most commonly used in routine, clinical care. Despite this modalities ability to resolve soft tissue structures being lesser to that of MRI, the advantage of being able to visualise surgically correctable intracranial bleeds remains. Cerebral contusions, subdural and epidural hematomas, all of which are evidence of significant brain injury are associated with post concussive patients and readily identified on CT. In actuality, the CT scan detects blood and occasionally oedema however does not effectively detect subtle neuronal injury such as DAI15. In the comparison of CT with conventional MRI only four previous studies have focused solely on individuals with mild concussive injuries, with only one including CT scanning. Within this cohort of research, the prevalence of abnormalities on MRI varied from 10 to 57% while on those with a nondescript CT scan, 30% of this population were observed to have a lesion.16-19. In brief, CT scanning is the initial method of choice for evaluating and excluding haemorrhage in a patient with MTBI, but its ability to evaluate the microscopic damage on a neuronal level is absent.
Magnetic Resonance Imaging
Magnetic resonance imaging plays a key role in the imaging of traumatic brain injury. A study conducted in 2009 on conventional MRI scanning of individuals with concussive injuries did not rule out subtle traumatic abnormalities below the detection threshold20. The authors recommendation was to perform studies in a greater population with concussion, preferably utilising a scanner with a higher field strength (3 tesla or more), as it was uncertain whether lesions that were too discrete to be shown with a 1.0 Tesla MR machine could explain the clinical symptoms of PCS20. In addition to the standard sequences used for the brain, typically a T1 sagittal and an axial proton density/T2 weighted dual echo, a fluid-attenuated inversion recovery or (FLAIR) sequence for the imaging of mild traumatic white matter injuries has been shown to offer some additional diagnostic value. The FLAIR sequence is an inversion recovery technique where the time to inversion (the TI) is adjusted to match the relaxation time of the tissue type (in this case, fluid) in order to suppress its signal contribution in the image. For example, to suppress fluid, the inversion time is set to the zero crossing point of fluid, resulting in the signal being 'erased'. It is in this way that FLAIR sequences allow lesions which would normally be obscured by the bright fluid signal in a typical T2 weighted image to be readily identified and made visible. In addition to the FLAIR sequences, Topal advocates for the use of GRE and DWI sequences which were found to be superior to conventional spin-echo sequences in the accurate detection of axonal injury21. GRE, otherwise known as a Gradient Echo sequence relies on a pair of bipolar gradients and the use of a small flip angle between zero and ninety degrees with which to flip the magnetization into its transverse component. A180 degree rephrasing pulse is not used as it is in its counterpart, spin echo. Despite these sequences being inherently more susceptible to magnetic field incoherency and inhomogencicity (as the FID is a function of T2*); GRE protocols are more adept at detecting hemorrhagic and bleeds22. The other recommended MRI sequence, diffusion weighted imaging (DWI), offers perhaps the more promising sensitivity to axonal injury. The diffusion of water molecules along a magnetic gradient acts to reduce the MR signal strength, rendering areas with low diffusion a comparatively brighter signal, implicating axonal injury and disruption to the axonal cytoskeleton. These DWI sequences identify shearing injuries which are not obvious on any T2, FLAIR or GRE sequences, thus making them valuable in evaluating closed head injuries such as concussion23. However, overall, for both conventional MRI and CT, early and exact identification of the extent of axonal injury is a major diagnostic challenge. This is due to the injuries seldom being identifiable; with both modalities demonstrating poor prediction of the functional outcome of patients with mild traumatic injury. This is predominately due to axonal injury that is not adequately detected24.
Functional Nuclear Imaging
In contrast to CT, functional neuroimaging techniques that examine the metabolic and physiological state of the brain have shown great potential in demonstrating the effects of mild traumatic brain injury that are undetectable by morphological methods. PET or positron emission tomography is one of these functional imaging techniques, and a member of the nuclear medical imaging modalities family. In this technique, a three dimensional functional image is obtained where a coincidence counter detects gamma rays emitted by a radioactive pharmaceutical fluoro-2-deoxyglucose (FDG). This pharmaceutical is an analogue of glucose, and acts to give an overall map of the metabolic uptake of glucose within the brain. PET has shown particular relevance in the detection of regional cerebral blood flow (rCBF). For instance, in the right prefrontal cortex of patients with PCS during spatial working memory tasks an increase in regional blood flow was reported25. Further confirmation comes from the same study utilising PET and working memory25, with the cohort afflicted by PCS demonstrating a significantly smaller BOLD (blood-oxygenation level dependant) signal change within the right prefrontal cortex. FDG tracers in individuals with concussive injuries who present with a normal CT scan have also implicated hypometabolism in the frontal and temporal brain regions, with which deficits in neuropsychological performance and a continued post-concussive syndrome can be linked26. Shown across a multitude of studies however was that in the resting state, any small differences between healthy and mildly traumatised brain tissue may not be easily shown in a resting cerebral metabolism, and it is only with the presence of external demands such as a spatial working task or a memory test that this mild injury becomes readily identifiable25. As with research concerning conventional modalities, the vast majority of studies have focused on those individuals with severe injuries rather then the relatively mild concussive type. However one study analysed the mild traumatic brain injuries of fifty four patients, fourteen of which had serially perfomed PET27. Cortical cerebral metabolic rate was witnessed to initially increase and then decrease four days post injury, and then was seen to match the notably prolonged decrease in cortical glucose metabolic rate observed in other cohorts with persistent post concussive syndrome.
Single position emission tomography or SPECT is another functional modality which has proven to report a greater range of visible post concussive abnormalities compared with CT and MRI. Akin to PET, frontal and temporal hypometabolism has been demonstrated following working memory tasks in various studies25, and this has been correlated with memory function deficit post injury. Typically, for a functional scan of the brain technetium- 99m is the gamma emitting radioisotope utilised. Upon injection into the bloodstream, the nuclear isomer emits gamma rays which can subsequently be detected by a gamma camera. In practise, 99mHYPERLINK "http://en.wikipedia.org/wiki/Technetium_(99mTc)_exametazime"Tc-HMPAO (hexamethylpropylene amine oxide) is taken up by the brain tissue in a manner proportional to the brain's regular blood flow, which allows both the brain blood flow and cerebral metabolism to undergo assessment. The other ligand often used is known as ethyl cysteinate diethylester, again in combination with the same radioisotope, technetium 99m. Both tracers cross the blood brain barrier and accumulate in regions of the brain in direct proportion relative to the nutrient delivery of the selected volume of brain tissue, giving an estimate of regional cerebral blood flow 15. Generally, SPECT competes with FDG based PET scans and provides similar data concerning cerebral metabolism but in a different method to PET's assessment of the blood glucose uptake. However, because SPECT radioisotopes are less expensive and easier to generate as well as having a greater longevity, SPECT is more commonly seen and more widely available then its functional counterpart. SPECT also has an advantage in that studies have shown that it possesses additional utility in its ability to discriminate lesions which offer a favourable prognosis from those which are unfavourable12. However the strongest case for the use of SPECT scanning is derived from evidence that it is particularly sensitive to lesions associated with mild traumatic injury with early perfusion abnormalities on SPECT appearing larger and more extensive than the same lesions on MRI or CT. This corresponds to the clinical severity of injury and neurological deficit and compromise12. Overall, this functional modality aims to combine the image reconstruction technology of CT with traditional studies of cerebral blood flow using radionuclide's and is more sensitive in the long run compared with CT and MRI in early detection of perfusion defects and following MTBI and concussion.
MRI spectroscopy has the goal of detecting individual tissue chemicals and molecules found within the human brain, rather then detecting and differentiating water, lipids and other tissue types as in conventional MR. NAA or N- acetylaspartate is the most common molecule found within the human brain other then the amino acid glutamate and is a clear indicator of the integrity of the neurons15. MRS studies have been undertaken which demonstrate that neuronal damage, as indicated by a decreased level of NAA, can be observed in the pericontusional edematous brain region within as little as seven days post concussion28. Another molecule typically assessed is choline, which acts as an indication of membrane turnover and damage to neurons, and will generally be elevated in cases where neoplasm, inflammation and necrotic regions are present. Once again the large majority of the literature focuses on the use of MR spectroscopy in severe head injury but numerous studies have utilised a smaller cohort of individuals with concussive type injuries29-31. Decreased levels of NAA alongside an increase in choline levels were observed across all three studies, with all patients possessing normal appearing white matter when interrogated by CT. In one case, a six month follow up study, the decreased NAA and increase in choline levels persisted30. These spectroscopy studies reconfirm that subtle neuronal injury following concussive injury may not be identifiable on standard MR or CT imaging, however
Diffusion Tensor Imaging (DTI)
A relatively modern MRI diffusion imaging technique, diffusion tensor imaging has been utilised in numerous studies in recent times. DTI has proven to be useful in identifying early stages of axonal injury in concussion through detecting axonal changes by measuring the movement of water. With diffuse axonal injury, the destruction of neurofilaments and microtubules running down the length of the axon leads to axonal swelling, followed by axonal disconnection and retraction into a ball. In both cases the linear arrangement of the axonal cytoskeleton is lost causing an interruption to the regular flow of water molecules32. Subsequently, the way in which the axon bundles are oriented determines how the water will flow, for example with an ideal case of parallel bundles of axons and their associated myelin sheaths, the diffusion of water would predominate along the main axis of the axon. This property can be imaged and measured with diffusion tensor imaging, and its utility lies in the detection of small changes in water movement which equate to a change in axonal structure in white matter. The two most established scalar quantities used in DTI are fractional anisotropy (FA) and apparent diffusion coefficient or ADC. Fractional anisotropy is a scalar quantity varying between zero and one which describes the degree of anisotropy (how directionally dependant it is). A FA of zero implies the diffusion of water is isotropic and is either unrestricted in all directions, or all equally restricted. Comparatively, a FA of one implies that diffusion is only occurring along a single axis and is restricted in all other directions. The counterpart to FA, the ADC describes the magnitude of the diffusion of water within given brain tissue. A low ADC indicates that the cortical white matter tracts are organized while a high value for the ADC indicates that these tracts are disorganized as would be seen in severe axonal damage. Xu et al made observations in their study on the DTI imaging of patients with mild brain injury that revealed a generalized pattern of white matter changes in major intra- and interhemispheric white matter tracts, consistent with DAI, alongside a generalised increase in ADC values24. In a similar study, but focusing on cases of MTBI, Bazarian et al detected significantly lower trace values and elevated fractional anisotropy with DTI imaging, primarily in the white matter tracts such as the posterior corpus callosum and the internal capsule. This rise in FA and drop in trace compared with controls was believed to be the result of axonal swelling following injury3. The study by Wilde agrees with the findings of Bazarian, and in utilising DTI on individuals with MTBI, an increase in FA and a decrease in ADC was observed33. Although many investigations noted an increase in ADC and a drop in FA, these reports were generally examined or included more severe cases of traumatic brain injury after a much longer post-injury time period then what was used in the case of the Wilde and Bazarian studies, hence the discrepancy in results. An earlier study on the benefits of DTI in 2002 distinguished between lesions with an increased diffusion (vasogenic edema) and restricted diffusion (cytotoxic edema) and the ability of DTI to serve as a valuable adjunct to conventional scans to monitor axonal injury and serve as an indicator of final outcome and injury reversibility34. Thus, from these reports, diffusion tensor imaging MRI can be used successfully to visualize pathology at a microscopic level that is not evident using conventional MRI or other non-invasive methods24.
Functional MRI or fMRI is a variant on conventional magnetic resonance imaging in that it relies on the magnetic properties of haemoglobin in order to create images of the blood flow to the brain. The functional component to the study involves patients undertaking certain cognitive tasks and then observing the increases in regional cerebral blood flow. It is this increase in the proportion of oxyhemoglobin relative to deoxyhemoglobin which results in a region of bright MRI signal strength in areas where blood flow is increased. Arising from this is the blood oxygenation level dependent (BOLD) signal which is detected. In many cases this will be superimposed over a surface shaded display of the area of interest. Many studies have used this modality in combination with neuropsychological testing and have witnessed an increased level of activation and an increase in cerebral blood flow in both the right parietal and the right dorsolateral frontal lobe regions following concussion. In one particular study by Chen7, striking similarities were drawn between the overall topographical activation pattern for the control group and concussed athletes. A lesser degree of task related activation was seen in the mid-dorsolateral prefrontal cortex, an area with strong associations with the monitoring of working memory. Other research has also contributed irregular patterns of activation to individuals with concussion after a form of working memory task. For example, in an fMRI study conducted by McAllister et al in 2001, a cohort with mild brain injury demonstrated, in comparison to controls, a large increase in activation compared while undertaking a moderate working memory processing load10. This irregularity in activation was localised to the same area of the right hemisphere, in particular the right dorsolateral prefrontal cortex, which was flagged in previous studies, both of fMRI and PET25.
The return to play
The return to play (RTP) of a concussed athlete presents a difficult challenge, given that in part there is a lack of management guidelines to assist the process35. The athletes must not return to play before the previous concussive injury has resolved, which generally occurs within three weeks in the absence of PCS. A potentially fatal and disabling consequence of a repeat concussion is known as second impact syndrome where the brain swells and the arterioles lose their ability to regulate their own diameter, leading to catastrophic cerebral oedema. Other potential afflictions derived from repeated concussion include dementia puguilistica (also known as Boxers encephalalitis), in which mental function gradually declines alongside associated Parkinsonism and neural dysfunction. Overall, the general emphasis of neuropsychological testing being the cornerstone of concussion management remains36. A focus on a rapid return to play is generally based on a complete resolution of all symptoms, with no new injury and the absence of post concussive syndrome. Often concussion affects the young and the adolescent, if this is the case these patients should be treated as if they were in the paediatric category, and are prescribed with a more lengthy time of recovery35. Despite the fact that there is a large array of return to play guidelines, none have been completely proven as valid nor do all of them have clinical backup. The indication in recent studies has been that recovery and the return to play will be proportionally longer if the concussion occurs in a younger athlete37, a female athlete38 or when the severity of initial injury on onset of the concussion was greater.
The subtle nature of post-concussive syndrome and mild traumatic brain injury belies its long term impact and potential to cause neurological deficits. While computed tomography remains the acute standard for neuroimaging of a mild traumatic brain injury it is only sensitive to gross injuries and is typically performed first as a measure to rule out more serious and life threatening injuries39. The evidence presented in the literature clearly suggests that a normal CT scan does not rule out the presence of brain injury and subtle neuronal and axonal damage. Due to the relationship between the ongoing symptoms of PCS and any underlying brain injury being comparatively weak, clearly, there is a need for a heightened accuracy in the diagnostic method35. The primary functional nuclear medicine studies, PET and SPECT both possess the potential to provide a detailed evaluation of multiple functional consequences after mild or severe traumatic injury and are thus capable of supplementing the clinical evaluation of MTBI and PCS as well as tailoring any therapeutic strategies needed to the individual40. But it is the imaging of diffuse axonal damage which offers the most promise for uncovering the sequaelae and prognosis of individuals with concussion and mild traumatic head injury, with associated PCS. MRI diffusion tensor imaging and MRI spectroscopy in particular were shown to be sensitive. However, in the field of sports injuries and mild concussion the literature has made it clear that many unanswered questions still remain. Further research into the incorporation of metabolic and functional brain imaging, particularly DTI and MR spectroscopy would be beneficial not only to clarify the time course of axonal recovery and long term prognosis, but to better enhance knowledge of the consequences of cerebral concussions with brain imaging. This would further allow a comparison of the efficacy of the various documented RTP protocols. Finally, a heightened understanding of the role of conventional modalities such as CT and its limited appreciation of neuronal damage should be made paramount to the practising physician in the care of a concussed patient.