The most prominent observation in this study was the marked increase in the width of the sulci of soccer players and the determination that the increased width was most evident in the athletes who experienced concussions. The reduction in brain volume of some soccer players was at an age when meta-analyses indicate that normally brain growth occurs rather than reduction in brain volume [33
]. A reduction in brain volume of a similar magnitude following mild traumatic brain injury was reported by Jarrett and colleagues [34
]. Brain volume changes can vary as shown by Poldrack et al. [32
], but our changes show a persistent and increasing atrophy. The significance of the finding of reduced volume of brain of the women athletes over a four year period remains to be determined because of the longitudinal study of brain volume in a single adult who was in his fifth decade of life [32
]. Our study reveals no clear association between overt concussive events that occurred during collegiate competition and reductions in brain volume. Changes in brain volume similar to those reported here were observed over the one year daily study of that individual; however, the consistent decrease in volume of the brain of the women athletes over the four year time span during an age when the brain is generally increasing in volume suggests that this finding may be significant.
The widened sulci suggest that the brain is deformed during repetitive impact of heads as a result of rapid acceleration/deceleration of an elastic brain interacting with non-compressible spinal fluid and a non-compressible cranial vault during heading of the ball and collisions between players. The widening of the sulci, and the presence of low-intensity punctate regions in the white matter interfaces with gray seen on SW and T2 images of our soccer study, is indicative of a “water hammer” causation during high-velocity head impacts [35
]. During the impact, the deformable brain parenchyma is driven against a non-compressible cranial vault and the non-compressible CSF is then driven into the sulci, as illustrated in Figure 5
. The region of brain exposed to the highest force is at the base of the sulcus where the “water hammer” force must dissipate if brain integrity is retained. The regions at the base of the sulcus are primary sites of vascular injury and hemorrhage into the brain parenchyma. Post mortem evidence obtained from brains of athletes who had TBI is consistent with the base of the sulci being most involved pathologically [11
]. Neuronal damage consisting of axonal bulbs and swellings is most commonly located in the deep gyri at the interface between the gray and white matter [11
]. The SWI reported here shows small focal hemorrhage where small vessels and U fibers are impacted.
The vulnerability of the interface between gray and white matter of the brain to injury from water hammer effects results from both the differing mechanical properties of the gray and white matter as well as the orientation of major dendritic and axonal processes at the base of the gyri. The white matter is, on average, 39% stiffer (average modulus 1.895 kPa) than gray matter (1.389 kPa) [36
]. The white matter is more viscous than gray matter and responds less rapidly to mechanical loading that is imposed by the water hammer effect. In addition, the orientation of the major dendritic and axonal processes of gray matter are aligned with the vector of force delivered by the spinal fluid driven against the base of the sulcus, while the axonal processes of the U fibers in the white matter are oriented perpendicular to the vector of force. These two dynamics result in the shearing force and rupture of vessels near the interface of the U fibers with gray matter. Figure 5
illustrates this sequence.
Proposed Mechanism for Development of CTE Following Repetitive Traumatic Events
Our hypothesis is that high-force impact on the torso and head of soldiers exposed to IEDs or of athletes experiencing body impacts causes a release of macrophages from the spleen, an activation of the macrophages to M1 and M2 macrophage cells [20
], and a transient increased permeability of the blood–brain barrier [41
]. In addition, as a result of high-force impact of the head, neuronal and glial populations at the base of the sulci are injured and release proteins into the vascular compartment and CSF. These proteins generate antibody responses to the released proteins originally sequestered in the brain. With subsequent impact forces on the body, the activated M1 and M2 cells enter the brain parenchymal compartments across the permeablized BBB, as do the antineuronal and antiglial antibodies. The activated M1 cells produce interferon (IFN) gamma that then induces expression of MHC/HLA histocompatibility markers [43
] by the neurons leading to neuronal silencing and subsequent death. Figure 6
illustrates this sequence of events. Support for this sequence of events is described below.
An excellent review of molecular, cellular, and system responses to traumatic head and body events is presented by Nizamutdinov and Shapiro [20
]. There is a release of macrophages from the spleen into the blood which subsequently penetrates the CNS compartment at regions of disrupted BBB [41
]. There are increases in levels of inflammatory response modifiers [46
]. Finally, the entry of the activated macrophages through the permeabilized BBB will lead to the expression of elevated interferon gamma levels that then result in increased expression of MHC/HLA markers on the neurons [43
], leading to silencing of the neuronal discharge and subsequent death.
The increased permeability of the BBB immediately following head impact and concussion will release neuronal and glial proteins (including neurofilament light, medium, and heavy [14
]) from the CNS compartment. The released proteins may then initiate production of antibodies to the neuronal proteins, as shown by Kornguth in patients with small cell carcinoma of the lung [26
]. During subsequent repetitive head injuries, the antibodies reactive/cross reactive to the neural proteins enter the CNS compartment and may cause the clinical signs of TBI. Each repetitive event is likely to exacerbate the injury, particularly when the interval between concussions is short.
The increases in the l NF in serum are associated with initiation of anti-NF antibodies and indicative of an immune process as a major factor in the development of the clinical signs classified as TBI. The increased inflammatory response modifiers described in Nizamutdinov [20
] are consistent with this hypothesis. In research from our laboratory, patients with small cell cancer of the lung (SCCL) that exhibited visual paraneoplastic syndrome were observed to have elevated levels of antibodies to the light chain neurofilament, and several of these patients had antibodies to the medium and heavy neurofilament proteins [26
]. These antibodies to the NF proteins reacted with the small cell cancer as well as the large retinal ganglion cells affected by the visual paraneoplasia. The patients with these antibodies survived longer than those patients with SCCL alone. The anti-neurofilament antibodies resulted in the selective immunoablation of large retinal ganglion cells following injection of these antibodies into cat vitreous [28
]. These observations on the SCCL population are supportive of the observation that the l NF proteins, seen in the mild to moderate TBI, pass from the neuronal compartment into the CSF over a prolonged time period. The continued release may serve as an antigenic stimulus to generate anti-NF antibodies in blood serum. This sequence of events whereby increased levels of l NF appears in the serum of patients at an early stage of progressive neurological disease is consistent with the finding of Byrne [49
] that l NF appears to be an early biomarker correlated with the rate of clinical progression of the autosomal dominant disease Huntington Disease (HD).
As described above, major research efforts funded by the National Institutes of Health, National Football League, the National Collegiate Athletic Association, and Defense Department have made efforts to detect biomarkers that are indicative of emerging CTE in persons exposed to high-impact forces as well as to define the pathological changes in the brain associated with CTE resulting from TBI [7
]. These important investigations have identified neuronal and glial proteins that are released into CSF and blood compartments following impacts, and they also have identified increased amounts of tau and phosphorylated tau as well as neurofibrillary tangles in postmortem brains of athletes who had clinical CTE. The question that then arises is whether there is a specific protein that is responsible for both the initiation of the autoimmune process and for the clinical neurological sequalae that follow that are long term consequences of the process. The authors propose that following multiple traumatic impacts to the head and torso, different proteins are released and the immune response to these proteins is dependent upon the host organism. The factors controlling the variation include (1) activation of splenic macrophages to M1 and M2 subtypes; (2) levels of antibodies generated to the neuronal proteins present in serum and CSF; (3) changes in permeability of the BBB following impacts, fevers, toxicants that will facilitate entry of the macrophages and antibodies into the CNS compartments; (4) production of interferon gamma in the CNS; and, finally, the (5) rate of expression of MHC/HLA markers on the neuronal surface that leads to neuronal silencing and degeneration. This process, rather than singular events, is proposed by us to be the basis of variability among athletes and soldiers to head injury in the susceptibility to and development of the clinical manifestations called CTE. The observations of Bernick and colleagues [16
]—that light NF increases in serum shortly after impact but then declines even while tau protein remains elevated—may suggest that the antibodies in serum that were produced following initiation of the autoimmune process bind the released light NF and thereby lead to a potential erroneous conclusion that there is no prolonged immune response to the NF antigen. The authors suggest that the subsequent impacts continue to release NF proteins that then further stimulate antibody production. The iterative process of antigen release and antibody boost together with the permeabilization of the BBB are proposed by us to be driving factors in the development of CTE.
The hypothesis presented above indicates there is a likely association between traumatic head and torso impact injuries, the water hammer effect on the base of the sulci, resultant release of neurofilament protein, and generation of high titer antibodies to neurofilament protein that then leads to inflammatory responses, interferon gamma production by the monocytes in the CNS, and increased production of MHC markers on neurons. This sequence leads to neuronal silencing and the eventual development of CTE. To demonstrate that this autoimmune process is etiologically involved in developing CTE, the authors propose that, as the clinical symptoms of CTE begin to appear in the second and later decades following TBI, there may be marked increases in the titer of antineuronal/antiglial antibodies in spinal fluid and serum from the athletes or soldiers. If such correlations are made, strong evidence would be provided regarding an etiological role of the autoimmune process.
As further evidence accumulates supportive of the hypothesis that autoimmune mechanisms are primary etiological events leading to CTE after traumatic brain injury, the strategies for mitigation include several approaches: treatment with statins [51
], treatment with minocycline [53
], antibodies to IFN gamma [56
]. Statin treatment has been demonstrated to reduce the extent of brain damage following injury from stroke events [57
]. The statins reduce the release of macrophages from the spleen, stabilize the blood–brain barrier, and decrease production of IFN gamma by the activated macrophages. The minocycline has use as an inhibitor of matrix metallo-proteinases where the proteinases disrupt the BBB [58
]. Administration of minocycline in fact limits infarct size in humans by stabilizing the BBB [55
]. Kantarci and colleagues [59
] have discussed the possible use of anti-interferon gamma antibodies in the treatment of patients with autoimmune diseases. This treatment is also suggested by the study of Zhao [60
], where it is demonstrated that interferon gamma is expressed at higher levels in pathogenic Th 17 cells compared with cholera-toxin-induced Th 17 cells. However, recent studies [61
] suggest that interferon gamma may have both positive and negative effects on the progression of autoimmune diseases depending upon the length of illness, sex of patient, and other factors. The statins and minocycline may offer more promising avenues of exploration than the anti-interferon gamma treatment.
Additionally, the observations of Dretsch and colleagues [62
] that individuals with brain-derived neurotropic factor with Val 66 Met modification are significantly more likely to develop CTE following impact injuries provides a capability for screening of individuals prior to exposure. Pretesting of soldiers and athletes for this brain-derived neurotrophic factor (BDNF) form could markedly reduce exposure and hence long-term risk of the individual for CTE.