Results of the current study provide promising insights into the use of DTI as an imaging method sensitive enough to capture mild WM injury in individuals with sports-related concussion who experience persistent symptoms with no LOC. The TBSS image analysis technique utilized here reduces the alignment issues of traditional voxel-wise analyses, and reduces the limitations of a priori
defined region-of-interest analyses that may dilute or eliminate detection of small structural lesions post-concussion. Furthermore, this is the first report of increased MD in individuals with sports-related concussion who experience persistent symptoms from an injury that did not warrant assessment by the GCS. Previous DTI studies of concussion or mild TBI have focused on subjects with GCS scores ranging from 13–15, and the assessment of FA (Niogi et al., 2008a
; Rutgers et al., 2008
), with the exception of Arfanakis and colleagues (2002
), Inglese and co-workers (2005
), Lipton and associates (2008
), Wilde and colleagues (2008
), Mayer and associates (2010
), and Messé and co-workers (2010
), all of whom assessed both FA and MD. Studies assessing WM fiber tract integrity in the early phase of recovery (days post-injury) in individuals with mTBI (GCS scores 13–15) describe conflicting results, with reports of decreased FA and/or increased MD (Arfanakis et al., 2002
; Inglese et al., 2005
; Miles et al., 2008
), and reports of increased FA and/or decreased MD (Bazarian et al., 2007
; Mayer et al., 2010
; Wilde et al., 2008
). Additionally, during later phases of recovery (months post-injury), studies assessing WM fiber tract integrity in individuals with persistent cognitive impairment reported decreased FA (Niogi et al., 2008a
), and increased MD (Lipton et al., 2008
). Collectively, these studies have reported abnormalities in a variety of brain regions, including the corona radiata, uncinate fasciculus, corpus callosum, inferior longitudinal fasciculus, superior longitudinal fasciculus, inferior fronto-occipital fasciculus, and capsula interna, many of which were also identified in the current study.
Results of the current analysis suggest a lack of structural integrity located in the left temporal lobe, where association and projection fibers running anterior-posterior and superior-inferior merge or cross. The most significant cluster spans the retrolenticular part of the internal capsule and the posterior thalamic radiation, which contain reciprocal thalamo-cortical pathways, including the optic radiation that connects the lateral geniculate nucleus to the occipital lobe. In this region, anterior-posterior-oriented long association fibers from the inferior fronto-occipital fasciculus (connecting the frontal and occipital lobes), and the inferior longitudinal fasciculus (connecting the temporal and occipital lobes), merge with the thalamic fibers. Part of the identified cluster also follows the C-shaped trajectory of the SLF as it bends superior-inferiorly from the frontal lobe towards the temporal lobe. In the superior temporal lobe, several voxels of the cluster with increased MD lie along the acoustic radiation, which contains projection fibers that connect the medial geniculate body of the thalamus to the primary auditory cortex of Heschl's gyrus. This centrally-localized junction of several WM tracts might be particularly vulnerable in concussion, and hence may be more sensitive than other areas of the brain to injury from translational (linear) acceleration/deceleration forces, or rotational forces that disrupt the structural integrity of the tracts (Broglio et al., 2009
; Viano et al., 2007
). These findings are in close agreement with the results of a recent TBSS study by Messé and colleagues, that revealed increased MD in long association tracts of the inferior fronto-occipital fasciculus and inferior longitudinal fasciculus bilaterally, as well as in the corpus callosum in a group of mTBI subjects with persistent neurobehavioral impairment during the subacute phase.
The results of the current study are also surprisingly similar to those of a recent TBSS pilot study of veterans who sustained combat-related closed head TBI following active duty in Operation Iraqi Freedom and Operation Enduring Freedom (Kim and Jorge, 2010
). Although the effects of blast-related TBI in veterans remain controversial (Levin et al., 2010
), Kim and Jorge (2010
) reported significantly higher MD and lower FA in veterans with TBI than controls in the following WM fiber tracts: left inferior longitudinal fasciculus, left superior longitudinal fasciculus, left retrolenticular internal capsule, and left sagittal stratum. These are the same regions identified in the current study, suggesting a similarity in the manifestation of military blast injuries and repetitive sports-related concussions with persistent symptoms, or identifying a region of the brain that is particularly susceptible to mild head trauma. Although the mechanisms of injury may differ, these results, when taken in combination, may open the door to future translational studies aimed at better understanding and treating the effects of mild brain injury in both civilian (specifically sports-related concussion) and military populations.
Evidence from histological, volumetric, and DTI studies supports enhanced structural integrity of the WM in these left hemispheric regions in the normal population. Larger left volumes than right volumes of the lateral geniculate body and the optic radiation, including the part of the optic radiation with increased MD in the current study, were observed in a histological voxel-based morphometry study by Bürgel and associates (1999
), although no asymmetries in the acoustic radiation were identified by Rademacher and colleagues (2002
), who used the same histological approach. A DTI tractography study by Barrick and co-workers (2007
) showed a leftward asymmetry of the WM fiber tracts in the posterior segment of the arcuate fasciculus in the left hemisphere, similar to the region identified in the current study. Furthermore, a voxel-based DTI analysis by Büchel and associates (2004
) revealed an asymmetry in the C-shaped structure of the posterior part of the SLF belonging to the arcuate fasciculus, with higher FA in the left hemisphere regardless of subject handedness. The observed increases in FA extended far into the temporal lobe in a similar manner as the current study's most significant cluster. In addition, Anderson and colleagues (1999
) demonstrated in a post-mortem study that WM increases in the left posterior superior temporal lobe (near the planum temporale region) were due to more thickly myelinated axons. Collectively, these studies of normal controls lend support to a leftward lateralization of increased WM structure, possibly due to thicker myelin, in the centrally localized junction of merging and/or crossing fiber tracts identified by the most significant cluster in the current study. As a result, this brain region may be more susceptible to injury from the translational and rotational forces acting in sports-related concussion. Further research should address this issue, as a larger sample size may reveal other areas of DAI.
Mechanisms of injury were reported to the attending physician for each participant in the current study, although it was frequently difficult for the athlete to remember precisely how the injury occurred, and reported injuries often conflicted with film reviews when available. To date, the relationship between the mechanism of injury and the site of injury remains unclear, and no theoretical thresholds for injury have been established (Mihalik et al., 2007
). As reported by Broglio and associates (2009
), the type of forces (linear and rotational) acting on the brain depends on both the site of the impact and the role of the player. Additionally, a player's physical condition, such as stronger neck muscles, may reduce the likelihood of a concussion (Viano et al., 2007
). Guskiewicz and colleagues (2007b
) and Mihalik and associates (2007
) suggest that it may be difficult to establish a threshold for concussive injury for football players (let alone across other sports disciplines), given the varying magnitudes of impact and frequency of previous concussion (reported and unreported). Complete information regarding the laterality, location, and the mechanics of impact may provide the necessary link between injury mechanism and the location of DAI in future studies of sports-related concussion.
The current study's results also suggest that MD is a more sensitive measure than FA at detecting small structural WM abnormalities, which is in agreement with a recent TBSS study by Messé and associates (2010
), who reported increased MD and no FA changes in individuals with mTBI (GCS score 13–15) and persistent neurobehavioral impairment in the subacute phase. Furthermore, a TBSS study by Acosta-Cabronero and colleagues (2010
), investigating early-stage Alzheimer's disease, revealed that diffusivity measures (axial, radial, and mean) were highly significant and far more sensitive than FA. Collectively, these studies suggest that MD may be a better measure for detecting subtle structural damage such as that caused by concussion, regardless of whether it is due to fiber reorganization, increased membrane permeability, destruction of intracellular compartments, or glial alterations (Acosta-Cabronero et al., 2010
; Beaulieu, 2002
). Enhanced sensitivity of MD to abnormalities in WM structure in the current study is further substantiated through qualitative assessment of average MD and FA values in the most significant cluster with increasing injury severity. For the concussed group, MD was more indicative of structural damage than FA. However, as injury severity increased, FA emerged as the diffusion parameter that is more indicative of structural damage, although age was confounded in the severe TBI group. This finding is substantiated by a 2009 TBSS study by Perlbarg and associates, that examined WM integrity in individuals with severe TBI during the subacute phase (Perlbarg et al., 2009
). These investigators also reported decreased FA and no diffusivity differences between severe TBI subjects with unfavorable 1-year outcomes and severe TBI subjects with favorable 1-year outcomes. Additionally, subjects with favorable outcomes exhibited FA values similar to those of normal healthy control subjects. In another TBI study by Huisman and associates (2004
), of individuals with a mean GCS score4 of 8.7 (SD 3.7), they found decreased FA and no diffusivity differences in the posterior limb of the internal capsule, an area in close proximity to the region identified in the current study. Furthermore, FA correlated with GCS and Rankin scores in the posterior limb of the internal capsule, whereas diffusivity values showed no correlation (Huisman et al., 2004
). The results of the moderate and severe TBI subjects in the current study are only qualitative in nature, due to the very small sample size, and consequently there are limitations to the conclusions that can be inferred. Nonetheless, it is of interest to present these findings to demonstrate that the current method is capable of detecting FA differences in long fiber tracts outside of brain lesions in severe TBI. FA has typically been the DTI measure of choice in previous TBI studies, although the current findings on the continuum of the levels of injury severity suggest that future research in mild TBI should utilize MD as an additional measure. A study by Heiervang and associates (2006
) revealed that FA is more variable among normal subjects than MD. Therefore, if the changes in FA and MD are small, as would be expected in the current study's subject population of concussed individuals, only MD group differences will be detectable with a small sample size. A larger sample size may be needed to overcome the higher between-subject FA variability, and to detect subtle group differences in FA. In summary, minor injuries in a small sample size in whom the lesions are not sizeable enough to provoke large directional (orientation) changes in diffusion, but are substantial enough to provoke small changes in overall diffusion, will be detected with greater sensitivity by MD than by FA.
The primary focus of the current study was to assess whether DTI offers measures sensitive enough to capture DAI in sports-related concussion not warranting assessment with the GCS, and to additionally identify the locations of DAI in the brain. NP testing and concussion history, done as part of Princeton University's Concussion Program, was administered according to their protocol for use in medical evaluation. As such, it inherently provided some between-subject variability in the testing time intervals post-injury, and in the testing time intervals from date of DTI scan. This may explain the insignificant correlations of MD with neurocognitive (NP testing) measures. Similarly, there was no relationship between MD and concussion history, although self-report of concussions is inherently troublesome, as athletes often minimize symptoms and do not report their injuries (McCrea et al., 2004
). Additionally, the current study was not designed to target a particular brain area with specialized NP tests. Although ImPACT targets many brain areas, it may not be specific enough to cognitively target the brain region identified with MD differences in the current study. Future research should address this issue by utilizing an NP testing protocol with rigorous time intervals and specific tests targeted to the brain region under investigation.
The NP data acquired closest to the DTI scan (at least 1 month post-injury) revealed that this study's concussion subject population was a mixture of subjects testing neuropsychologically normal (n
3), and neuropsychologically abnormal (n
6). Interestingly, all subjects continued to experience persistent symptoms and showed a lack of WM fiber tract integrity despite their NP testing results. Additionally, all subjects with the exception of one were able to continue their normal college coursework following concussion. This raises important questions: Can the brain compensate for potential cognitive deficits resulting from structural abnormalities in sports-related concussion subjects who test NP normal by drawing upon other neurocognitive resources that are individual-specific, such as intelligence quotient (IQ) and level of general cognitive ability, or are the results of NP testing in subjects who performed within the normal range confounded by a learning or motivational effect? If the brain can compensate, this opens the door to future research aimed at identifying the brain regions that are compensating, and in a sense are “working overtime,” to make up for any cognitive deficiencies based on the structural deficit. It also opens the door to future strategies aimed at improving evaluation, management, and treatment of concussion. On the other hand, athletes repeat the same battery of NP tests numerous times after injury to evaluate cognition. Athletes are also aware that the NP test results will be used to determine whether they are allowed to return to play. Therefore athletes may be more motivated to perform well on the tests post-injury than when initially tested at baseline. If the results of NP testing are confounded by learning or motivational effects, this supports the need for a more objective diagnostic tool for the assessment of sports-related concussion, which may involve the use of DTI to measure MD.
Collectively, these results suggest that in addition to current NP tests, DTI may be another useful tool for clinicians when making return-to-play decisions to help reduce vulnerability to recurrent injuries. Whether the structural changes demonstrated in this pilot study of athletes having suffered more than one concussion and demonstrating symptoms that persist for at least 1 month post-injury also occur in athletes with uncomplicated concussions (concussion with symptoms, cognitive function, and balance testing returning to normal within 7–14 days) remains to be evaluated. Ultimately, identifying and tracking the effects of concussion demands a longitudinal assessment of the relationship between structural integrity as assessed by DTI scans, functional ability as assessed by fMRI scans, and performance/cognitive ability on NP tests in the acute, subacute, and long-term phases of recovery after concussion. For these reasons, this pilot study is being followed with a longitudinal combined DTI/fMRI study of a larger sample size to refine, clarify, and provide a more detailed understanding of mild TBI, specifically concussion.
Currently, there are no routine imaging protocols used to evaluate sports-related concussion besides traditional clinical MR imaging (T1-weighted), or CT scans, neither of which can detect subtle structural changes, and thus are inadequate for use in the assessment and management of concussion. This pilot study provides evidence of structural changes in the WM of the brain, creating a link between concussive injury with persistent symptoms and changes on DTI. These findings emphasize the importance of athletes to report any possible concussion, and thus to reduce the chance of re-injury. Furthermore, the current study's results suggest DTI may serve as a biomarker for concussion, and thus may provide an objective diagnostic tool to help determine the severity of injury, and to manage concussions, particularly with regard to return-to-play decisions.