Military Blast Injury and Chronic Neurodegeneration: Research Presentations from the 2015 International State-of-the-Science Meeting

Pathological characteristics of blast-induced TBI

Researchers use neuropathological evidence to identify and characterize CTE. These neuropathological observations are also a window into the mechanisms that may drive chronic neurodegeneration after TBI.

Dr. William Stewart, Queen Elizabeth University Hospital, presented on behalf of Dr. Douglas Smith from the University of Pennsylvania on lessons learned from studies in animal models regarding the neuropathology of blast-induced TBI. Diffuse axonal injury (DAI) is generally believed to be the primary clinical neuropathological sequela in impact (nonblast) mTBI or concussion. Pre-clinical studies in blast, however, have failed repeatedly to demonstrate DAI acutely in primate, ²¹ swine, ²² and rodent ^23,24 models. Evidence in support of axonal injury is also an inconsistent finding in imaging and pathology studies in humans; potential confounding factors where axonal pathology has been described at autopsy include drug overdose, ante-mortem anoxia, and concurrent impact TBI. ²⁵

While DAI is not a consistent observation in pure blast TBI models, reports across species have documented evidence of glial, ²¹ microglial, ²⁶ ) and vascular pathologies, however. ²²

Better neuropathological characterization of blast-induced pathology in human tissue is a high-priority research need that will inform and direct the development of appropriate animal models. In addition, ongoing efforts to define common data elements in pre-clinical research will help drive consensus between groups. Greater collaboration in the field is needed.

Opportunities for clinical neuropathological research are limited. Dr. Daniel Perl, of the Uniformed Services University of the Health Sciences (USUHS), reviewed neuropathology findings from case studies of acute and chronic blast TBI in humans, including observations of unique pathological lesions in the form of glial scarring.

Few neuropathological studies exist describing acute or chronic effects of blast TBI in humans. Observations of unusual glial scarring in an index case of chronic blast TBI, however, led to a recent collaboration between the USUHS Center for Neuroscience and Regenerative Medicine and the Joint Pathology Center comparing neuropathology after chronic blast TBI (five cases), acute/subacute blast TBI (three cases), chronic civilian impact (nonblast) TBI (five cases), and civilian controls with no TBI history (three cases). A notable pattern of glial scarring was observed in all chronic blast TBI cases. A parallel pattern was seen in two service members who survived four days post-blast: reactive astrocytes (which are thought to reflect the initiation of glial scarring) were found in the subpial region and gray-white matter junction.

Ex vivo high-resolution magnetic resonance imaging (MRI) (7 Tesla) of tissue from a single chronic blast TBI case provided a preliminary indication that the detection of gray-white matter abnormalities associated with glial scarring may be possible in living persons. Glial scarring was observed in structural areas associated with known functions thought to be adversely affected by chronic TBI or CTE. The current theory regarding how blast waves interact with human physiology predicts that damage will occur at structures consisting of, or proximate to, tissues of different densities, which is the case of structures with glial scarring in this preliminary study. Tau pathology was not observed consistently across chronic blast TBI cases; however, it is possible that tau aggregation would have developed in these patients over time. Observed patterns of axonal pathology also did not differentiate between acute/subacute and chronic blast cases.

These preliminary findings were observed in spite of small sample size, limited data on blast exposure number and nature of injuries, limited post-concussive symptomology data, and limited exposure histories for nonblast (impact) TBI cases. Larger, controlled studies comparing blast and civilian impact TBI and using available clinical and operational data to inform exposure history are recommended. High-resolution imaging of lesions associated with gliosis may also benefit clinical studies of blast-exposed service members. Based on these preliminary observations, the development of animal models with human tau or that characterize post-blast gliosis would also be beneficial.

Risk factors and CTE

Dr. Robert Cantu and Mr. Philip Montenigro, both from Boston University Alzheimer's Disease and CTE Center, reviewed research on the neuropathological and clinical features of CTE. CTE is defined as a progressive neurodegenerative disorder in those with a history of brain trauma (frequently with multiple exposures) that results in cognitive, behavioral, mood, and/or motor deficits. In post-mortem analysis, CTE is distinguished from Alzheimer disease (AD) by its characteristic pattern of tauopathy, including perivascular accumulation of neurofibrillary tangles at sulcal depths and superficial cortical layers. ²⁷ A recent case series of neuropathologically confirmed CTE identified the presence of amyloid-beta (Aβ) deposition in 52% of CTE subjects. ²⁸ The presence of Aβ pathology in these subjects occurred at an earlier age with altered dynamics compared with the normal aging population and those with AD. In addition, Aβ deposition was associated with a greater severity of CTE tauopathy and dementia symptoms. With support from the NINDS and National Institute of Biomedical Imaging and Bioengineering, consensus neuropathological diagnostic criteria for CTE were established recently. ²⁹

Currently, CTE can be diagnosed only by a post-mortem neuropathological examination. In an effort to identify a common clinical presentation of CTE, ³⁰ investigators performed a retrospective analysis of clinical histories, medical records, and next-of-kin reports among autopsy-confirmed CTE cases without comorbidities. In more than 70% of cases, symptoms occurred in three clinical domains: behavior, mood, and cognition. Symptom clusters gave rise to two subtypes: (1) behavioral/mood, which had an earlier age of onset and predominant symptoms of explosivity, rage, depression, with mild cognitive symptoms later in life; and (2) cognitive, which had a later age of onset and predominately memory, executive, and attention issues.

Recently, Cantu and Montenigro and associates proposed clinical criteria for the diagnosis of CTE ³¹ and for the first time applied these criteria to two case reports. ³² To date, studies to validate and improve on the proposed clinical criteria are ongoing; an expert clinical consensus panel has already utilized the proposed criteria to diagnose clinically 99 cases retrospectively while panel members remain blinded to the neuropathological diagnosis. ³³

Brain injury is the primary known risk factor for CTE. For professional football players, the number of years played and years since retirement both correlate with CTE severity. ²⁷ Career length may be associated with CTE severity by serving as a proxy for the number of injuries accumulated; years since retirement may be associated with the amount of time neuropathology was allowed to progress. Next of kin reported concussion frequency and steroid use are not correlated with severity of CTE pathology. It is not yet known which specific aspect(s) of brain injury (source, frequency, location) influence the risk for or natural progression of CTE. Closed-head impacts in collision sports (e.g., football, boxing) cause sudden acceleration or deceleration of the brain inside the skull that may introduce internal brain injuries without visible fracture or contusion.

These closed-head brain injuries may result from forces propagated through two biomechanical pathways. The first is the “common” pathway to injury, which is homogeneous in nature and occurs where the brain is structurally vulnerable to injury. When strains overwhelm the underlying brain tissue through the common pathway, a previously described acute closed-head brain injury pattern occurs termed the “gliding contusion.” ³⁴ The chronic neuropathological characteristics of CTE share certain similarities to, and may be associated with, the acute closed-head brain injury pattern of the gliding contusion. The second aspect of a closed-head brain injury is the “focal” or “local” pathway, which is heterogeneous in nature and specific to the type and location of impact. Thus, closed-head brain injuries result from a combination of both common (homogeneous) and local (heterogeneous) pathways to tissue injury.

The common pathway is thought to play the major role in the accumulation of injury that results in long-term impairment and CTE tauopathy, as game play results in repetitive closed-head impacts. Cantu and Montenigro hypothesized, however, that sport specific differences in impact exposure might alter the pathological and clinical manifestations of CTE through differences in the biomechanical forces introduced through the local pathway of closed-head impacts.

Football players are exposed generally to repeated linear forces because of frontal, head-to-head contact. Boxers are exposed generally to rotational acceleration resulting from the hook-punch to the chin. Rotational acceleration is thought to have a greater effect on the mid-brain and cerebellum, which was observed in a recent case comparison of a single professional football player and single professional boxer. ³² Tissue analysis from both individuals met consensus based neuropathological criteria for stage III CTE. Yet the boxer sustained greater neuronal dropout and tau-positive neurons within the dentate of the cerebellum. These neuropathological differences were mirrored in a secondary analysis of the data from McKee and colleagues ²⁷ : when professional boxers and football players were compared, boxers were more likely than football players to have CTE with motor symptoms and neurofibrillary tangles in the cerebellar dentate, a higher proportion of which were severe. ³²

Blast-induced neurodegenerative mechanisms

Dr. Gregory Elder, of the James J. Peters VA Medical Center, reviewed animal model research of blast-induced behavioral alterations and pathology. Investigators at the Naval Medical Research Center developed an animal model designed to mimic blast-induced mTBI. Exposure to multiple blasts resulted in no major neuropathological abnormalities, but was associated with transiently impaired spatial learning as assessed by the Morris Water Maze test. ³⁵

Distinguishing the overlapping symptoms of PTSD and post-concussion syndrome has proven difficult for clinicians and has created the need for animal models to study and tease apart the behavioral components of these conditions. Recent experiments have demonstrated the induction of PTSD-related traits after exposure to blast. In behavioral assessments, animals exposed to multiple blasts exhibited increased anxiety, acoustic startle, and cued fear responses when compared with controls. In blast-exposed animals, abnormal behavioral responses persisted for three days after exposure to a predator scent. These behavioral abnormalities were also accompanied by changes in stathmin 1, a microtubule-associated protein highly expressed in the amygdala that plays a role in learning and fear responses. Subsequent DNA methylation profiling in these animals has indicated epigenetic changes potentially underlying long-term, chronic effects. ³⁶

An increase of Aβ plaques has been reported after closed-impact TBI in humans ³⁷ and in animal models. ³⁸ Recent work in blast-exposed rodents, however, indicates a decrease of Aβ in rats at 24 h and one week post-blast. ³⁹ In addition, amyloid precursor protein (APP) levels increased in blast-exposed animal models, similar to observations after closed-impact TBI. To date, tau pathology or neurofibrillary tangles have not been observed in this blast-exposed animal model.

Blast-induced vascular pathology may be relevant to the study of CTE given that perivascular tau accumulation is one defining characteristic of the disorder. Exposure to multiple blasts is associated with vascular abnormalities, with dysmorphic vessels exhibiting irregular vessel walls, and debris in the lumen; the neurophil surrounding such vessels was normal, suggesting that blast effects are specific to the vasculature. Acute blast exposure in this animal model was also associated with a degradation of the glycocalix, a supportive matrix lining the inner walls of blood vessels, and poorer spatial learning performance; both abnormalities could be rescued with prophylactic application of hyaluronidase during blast exposure. Chronic effects of blast exposure include remodeling of the extracellular matrix in rats. In addition to chronic vascular effects, neuroinflammatory-related changes after blast exposure are being considered as a potential mechanism of injury. ⁴⁰

Dr. Peethambaran Arun, of the Center for Military Psychiatry and Neuroscience at the Walter Reed Army Research Institute, presented findings from two animal models of brain injury describing effects on phosphorylated tau (p-tau) and APP levels in the brain. Rats given a single exposure to blast injury (blast tube, 19 psi) or impact-acceleration injury (weight drop, 500 g from 250 cm above the skull's surface) showed decreased levels of tissue nonspecific alkaline phosphatase (TNAP) and elevated levels of p-tau at 6 and 24 h in the brainstem, hippocampus, and cortex compared with a sham control group. Plasma levels of alkaline phosphatase enzyme activity were also reduced at 6 and 24 h after exposure to an impact injury, but not blast; this result may be explained by alkaline phosphatase coming into the plasma from internal organs after blast exposure. Levels of alkaline phosphatase in the brain decreased after both blast and impact injuries. At 24 h, APP levels were elevated in the brainstem, hippocampus, and cortex of both injury groups. These data suggest two processes are set in motion by blast or impact injury: (1) a decrease in TNAP, which may lead to the accumulation of p-tau, and eventual development of CTE; and (2) an accumulation of APP, which may trigger an AD-like pathology.

Six hours after rats are repeatedly exposed to blast (two 19 psi blasts within 1 min), p-tau levels are elevated in many brain areas (e.g., brainstem, cortex, cerebellum, midbrain), and remain significantly elevated in the hippocampus at 24 h and 14 days. In rats exposed to two blasts, p-tau levels are compared with rats exposed to a single blast. The APP levels, while also higher after exposure to two blasts, did not increase at the 6-h time point, but were elevated at 24 h and remained so through the 14-day time point.

The lack of a common pattern of elevation for p-tau and APP supports the hypothesis that the neurobiological mechanism of CTE may be distinct from AD, although the accumulation of APP at later time points suggests that blast can trigger an AD-like pathology. Future neuropathological work will examine the pattern of p-tau deposits to determine whether it is located in the superficial layer of the cortex, as in CTE, or in the global layer, as is seen in AD.

Dr. Alexander Lin, of the Brigham and Women's Hospital, reviewed research using magnetic resonance spectroscopy (MRS) to identify post-concussion neurochemical alterations that may serve as biomarkers for CTE. The MRS measures relative concentrations of brain metabolites in living persons and can characterize acute (minutes to days) and chronic (years) neurometabolic changes after injury that may be associated with CTE.

Early application of MRS demonstrated a potential role for predicting functional outcomes after severe head injury. Studies in comatose patients with severe head injury resulting from car accidents found that those with a less than 20% reduction of N-acetylaspartate (NAA; a chemical found within neural cell bodies that is thought to correlate with the number of functioning neurons) had positive recovery outcomes; patients with a larger than 50% reduction in NAA remained in a persistent vegetative state. ⁴¹ The MRS measurements later were shown to be predictive of Glasgow Coma Scores in adults with severe head injury. ⁴²

MRS can also characterize metabolic changes after mild brain injury. In a study of subjects with a single concussion, compared with controls, MRS showed an acute reduction of NAA that recovered over time. In subjects with two concussions, NAA reduction was similar but recovered more slowly. ⁴³

Research in athletes exposed to head trauma aims to identify metabolic changes that may serve as CTE biomarkers. The MRS of retired football players with CTE-related complaints and a history of multiple concussions indicated elevated levels of choline (a marker of cell damage) and lower levels of glutathione (a marker for neural formation) in the anterior cingulate compared with controls. ⁴⁴ Elevated levels of choline and myo-inositol (a potential marker of glial cells) were observed in soccer players compared with a control group of noncontact sport athletes. ⁴⁵

Development of two-dimensional (2D) MRS has enabled identification and quantification of previously undetectable brain metabolites. Application of 2D MRS in retired football players with CTE-related complaints and a history of multiple concussions indicated elevated levels of glutamate; however, the degree of glutamate did not correlate with the degree of concussion exposure. These and other observations have led researchers to hypothesize about changes in brain metabolite levels during exposure to injury and development of CTE. For example, a better understanding of the timing of glutamate decrease could help identify time points for the potential application of intervention strategies that mitigate its neurotoxic effects. Future research combining MRS with neuroimaging modalities may advance understanding of how neurometabolic changes after brain injury or CTE are localized.

Dr. David Cook, of the VA Puget Sound Health Care System and University of Washington, reviewed research findings describing alterations in p-tau expression, blood–brain barrier dysfunction, and cerebellar damage in the brain after blast injury. To investigate these phenomena, an animal model of blast injury was developed that uses a shock tube to generate a blast wave similar to trinitrotoluene in the open field. Rodents exposed to this blast wave exhibited aberrant p-tau in the hippocampus for up to 30 days after injury and transient perivascular p-tau accumulation was detected before abating within 4 h after exposure.

In vivo two-photon laser scanning microscopy was used to study blood–brain barrier integrity in blast-exposed animals. Transient microvascular alterations and microglial responses were observed in focal areas after blast exposure. Delayed blood vessel dysfunction and microglial/macrophage accumulation was observed, indicating that some instances of these phenomena are not directly related to the immediate biomechanics of blast exposure.

After blast exposure, the cerebellum appears to be susceptible particularly to blood–brain barrier disruption, especially in ventral areas, and cerebellar Purkinje cells exhibit p-tau accumulation and plasma membrane integrity alterations. In addition, the loss of Purkinje cells was observed after exposure to a single blast; cell loss was greater after exposure to multiple blasts. Purkinje cell losses are observed in patch-like microdomains. These observations are similar to reports of Purkinje cell loss in post-mortem neuropathological examination of former boxers. ⁴⁶ In animal models, blast exposure causes alterations in a motor task (rotarod) and elevated astrogliosis 30 days after injury. Similarly, veterans exposed to blast exhibit hypometabolism in the cerebellum that correlates with the number of exposures and sensorimotor integration symptoms.

Dr. Charles Rosen, Department of Neurosurgery at West Virginia University, discussed research investigating the role of endoplasmic reticulum (ER) stress responses in blast-induced neurodegeneration. The ER stress responses are thought to be associated with neurodegeneration through the induction of p-tau pathology. ^{47 –49} The ER stress response is mediated in part by two intracellular cascades: (1) the PRKR-like ER kinase (PERK) pathway, and (2) the inositol-requiring enzyme 1 pathway. Stressors to the ER disrupt normal equilibrium between these two pathways and induce relative overactivation of the PERK pathway, which elevates the transcription factor C/enhancer binder protein homologous protein (CHOP), leading to cell apoptosis through activation of caspase proteins.

To investigate intracellular signaling pathways associated with ER stress response and blast exposure, a rodent model was developed using a blast tube apparatus. Validation experiments demonstrated induction of a physiologically relevant blast wave, and animals exposed to the blast wave exhibited astrogliosis, p-tau pathology, and activation of the PERK pathway with elevation of CHOP. Exposure to single or repeated blast induced maladaptive behavior (reduced anxiety), as assessed by the Elevated Plus Maze, despite retention of normal motor activity. In addition, these animals exhibited delayed learning and memory deficits 21 days after blast, as assessed by the Morris Water Maze. Administration of salburinal 30 min before blast exposure inhibited activation of the PERK cascade and restored normal behavior on the Elevated Plus Maze. These data indicate that blast exposure induces ER stress responses that mediate TBI-relevant behavioral outcomes. Modification of the ER stress response could prove to be a beneficial strategy for preventing blast-related neurodegeneration and abnormal behavior. Complete details on the study by Dr. Rosen and colleagues can be found in the article “Endoplasmic Reticulum Stress Modulation as a Target for Ameliorating Effects of Blast Induced Traumatic Brain Injury” in this Supplement on page S-62.

Neuroimaging and biomarkers

The successful development of objective biomarkers could enable the identification of CTE pathology in living persons, which would greatly enhance research capabilities and would inform potential diagnostic and treatment strategies. Investigators are pursuing neuroimaging modalities and biospecimen analysis as potential biomarkers of CTE.

Dr. Denes Agoston, from USUHS, presented findings from pre-clinical research characterizing the long-term neuropathological and behavioral effects after single and repeated mild, blast-induced mTBI (rmbTBI), considered a primary risk factor for development of neurodegenerative conditions such as CTE. The pathophysiological processes linking single and multiple blast exposures to neurodegeneration, however, are not fully elucidated. These findings can be found on page S-44 of this Supplement.

Using a shock tube model in rats, exposure to a single blast resulted in (1) memory impairment that persisted for the entire 71-day experimental period and (2) transient elevation of anxiety behavior that recovered to normal levels by 71 days. Proteomics analysis of brain regions mediating the above mentioned neurobehavioral functions (e.g., dorsal hippocampus, ventral hippocampus, amygdala, prefrontal cortex) from blast-exposed rats revealed region-specific changes in the levels of the select biomarkers (e.g., interferon gamma, interleukin-6), indicating that different brain regions have different vulnerability to blast exposure. Measurement of serum biomarkers suggested neuroinflammation, vascular damage, and axonal damage in these animals. These results indicate long-lasting functional and molecular changes resulting from a single blast exposure.

Inducing rmbTBI in the shock tube (five exposures, 24-h interval) resulted in more pronounced neuropathological and behavioral effects when compared with single blast exposure. Rats exposed to rmbTBI had more pronounced memory deficits (including spatial memory) and increased cell death 21 days post-injury when compared with rats exposed to a single blast. Preliminary experiments indicate that the most damaging between-exposure interval is 1 h in the rat model; this time frame is thought to be equivalent to several days or weeks in humans. Further defining the complex temporal relationship between human and rodent pathologies involved in the neurodegenerative processes is critical to translating pre-clinical findings to clinically relevant information useful for diagnosis of brain injury, and treatment and rehabilitation for those with such injury.

Dr. Meghan Robinson, of the VA Boston Healthcare System, presented findings from neuroimaging studies at the Translational Research Center for TBI and Stress Disorders (TRACTS) that examined the chronic effects of blast injury. Recent research suggests that acute concussion symptoms are not predictive of long-term changes to the brain. In addition, service members exposed to blast often underreport concussion, indicating that there is a population with the effects of blast injury without assessable concussion symptoms. Given these observations, investigators at TRACTS are seeking to explore variables other than concussion symptoms that could potentially characterize the chronic effects of blast injury. Two recent studies have drawn from a cohort of veterans and service members recruited by TRACTS who have completed neuropsychological testing, clinical assessment, physical assessment, and MRI scans.

One study examined alterations of functional connectivity after blast exposure. ⁵⁰ Subjects exposed to blast at close range (<10 m) exhibited decreases in functional connectivity as assessed by seed-based resting state connectively analysis of MRI scans; decreases were not observed for those exposed to longer range (>10 m) blasts and were independent of blast concussion. Clinical factors such as sleep, pain, or PTSD were controlled for in these comparisons. Differences in functional connectivity were not observed in subjects with concussion history.

A second study sought to determine whether blast exposure influenced age-associated degradation of white matter tissue integrity. ⁵¹ Diffusion tensor imaging indicated more pronounced cross-sectional age associations with white matter integrity among subjects exposed to blast (<100 m away). Independent clinical factors, such as sleep, pain, or PTSD, could not account for this effect on white matter integrity. These observations suggest the existence of an underlying neurodegenerative process.

Taken together, these studies indicate that blast exposure may be an important consideration, independent of acute concussion symptoms, for potential chronic health effects among service members and veterans. They identify a potential clinical guideline for determining which blasts (i.e., <10 m) pose a higher relative risk to health. Complete details of Dr. Robinson's study can be found in the article “Characterization of Differences in Functional Connectivity Associated with Close-Range Blast Exposure” on page S-53 of this Supplement.

Dr. Jessica Gill, of the National Institutes of Nursing Research, reviewed the results in service members of a study of TBI biomarkers that could potentially be applied to CTE. Enhanced detection of tau and Aβ protein in peripheral blood has been made possible by development of a sensitive analytical technique that amplifies the signal yielded by the traditional enzyme-linked immunosorbent assay by a factor of 300 to 1000. Analysis of peripheral blood samples was performed at Madigan Army Medical Center in service members who were deployed within the last 16 months. ⁵² Among 70 service members who self-reported exposure to TBI during deployment, 58 (82.8%) reported exposure to blast.

The presence of tau protein in blood was associated with TBI exposure and chronic injury symptoms. The concentration of total tau protein in peripheral blood plasma was significantly higher in subjects with self- reported TBI exposure compared with nonexposed controls. Among TBI subjects, plasma concentration of tau was significantly higher in individuals reporting three or more exposures compared with those with fewer than three exposures. Total plasma levels of tau were also correlated with chronic TBI symptoms, as assessed by the Neurobehavioral Symptom Inventory.

The study's findings also indicate a potential relationship between plasma Aβ and blast injury. Concentration levels of Aβ protein were higher in subjects self-reporting that they were exposed to at least one TBI compared with controls; however, this difference was not statistically significant. Among TBI subjects exposed to blast, Aβ levels were significantly higher than those of nonexposed controls.

Dr. David Okonkwo, from the University of Pittsburgh Medical Center, reviewed neuroimaging research characterizing neuropathology in relation to neuropsychological outcomes associated with chronic TBI. A recent pilot study in 17 subjects with chronic TBI and four control subjects used High Definition Fiber Tracking (HDFT) to image white matter tracks, position emission tomography (PET) using Pittsburgh Compound B (PiB) to detect Aβ deposition, and fluorodeoxyglucose PET (FDG PET) to detect metabolic activity.

HDFT generates data specific to individual subjects by using novel quantitative metrics to circumvent mathematical limitations of voxel-based analysis; these metrics include generalized Q-space imaging algorithms followed by deterministic tractography. These quantitative metrics do not pursue group mean statistics and are specific to the individual patient. Quantitative tractography metrics can be applied in two ways: (1) examine anatomical asymmetries in white matter to identify and localize specific tracts that have been subjected to loss or abnormalities within the contralateral side of the subject's injury and also within the control subjects and (2) identify and localize abnormalities in white matter arborization, which has been observed in other studies examining subjects exposed to blast injuries.

Ongoing pilot study observations indicate a significant, age-adjusted increase of Aβ deposition in TBI subjects when compared with controls; this signal of increased Aβ was observed in brain regions vulnerable to AD. In gray matter regions of interest in AD, increased Aβ deposition was strongly associated with reduced metabolic activity, which is an expected pattern for those with dementia. In a small cohort of trauma-related dementia patients, increased Aβ deposition was observed in association with decreased metabolic activity, which appears to be similar to AD-related dementia. In addition, white matter anisotropy in this cohort was associated with reduced metabolic activity and increased Aβ deposition.

These observations can enable testing of the hypothesis that white matter injury induces pathophysiological processes that can culminate in long-term change in humans and animal models exposed to TBI. Future research will evaluate whether white matter abnormalities are associated with functional outcomes assessed by neuropsychological tests (e.g., verbal recall, serial reaction time, processing speed index, trail making) in subjects with chronic TBI. In addition, FDG PET imaging to characterize tau protein accumulation is planned. Neuroimaging outcomes will also be correlated with cerebrospinal fluid and serum TBI biomarkers.

Dr. Dara Dickstein, from the Icahn School of Medicine at Mount Sinai, presented on behalf of Dr. Sam Gandy and described molecular imaging research on TBI and CTE, including a review of case studies. Several PET ligands exist that bind with tau or amyloid protein aggregates, which are associated with TBI and CTE neuropathology. Questions remain about the binding specificity of ligands, particularly those that target tau, and whether PET binding signals are accurate reflections of neuropathology or disease states in these conditions.

The PET imaging has identified amyloid deposition after TBI. Application of the PiB radioligand indicated the presence of amyloid in acute TBI, which is retained 48 h, 54 days, and 399 days after injury. Preliminary studies in acute TBI suggest that PiB retention is observed in moderate to severe TBI with visible contusions on MRI. An ongoing PET study in boxers one month post-knockout, however, did not observe amyloid retention with PiB. To date, the presence of amyloid has been associated with severe TBI, with PiB retention in some subjects with TBI, and neuropsychological impairment. ⁵³

Investigators have sought to characterize tau after TBI with PET imaging. ^{54 –56} These studies indicate the presence of p-tau in varying brain distribution patterns, bringing to light questions about the binding specificity of tau ligands and the potential contribution of age-related tauopathy. Studies have indicated that age-related tauopathy is predominately localized in the hippocampus and entorhinal cortex as opposed to a case with severe AD with p-tau staining extending into the temporal neocortex. ⁵⁷ To establish that these ligands can be used to study CTE, it is important to characterize the distribution patterns of p-tau in age-related tauopathy and AD.

Recent case reports in two subjects, aged 68 and 70 years, with TBI history and memory-related neurocognitive problems yielded mixed results: One was tau positive and amyloid negative, while the other was tau negative and amyloid positive. These observations have led investigators to ask whether CTE can be distinguished from AD using PET imaging and cognitive assessment. An ongoing study to address this question is seeking to image amyloid (using the [ ¹⁸ F]-florbetapir ligand) and tau (using the [ ¹⁸ F]-T807 ligand) in conjunction with neuropsychological assessment in subjects with a history of concussion; controls have been the most difficult study group to recruit.

An additional case report yielded neuroimaging findings that appear to be associated with behavioral symptoms. The subject was a 39-year-old retired professional football player with multiple concussions and severe functional impairment, including sleep disturbances, headaches, short-term memory loss, as well as impulsive and aggressive behaviors. The subject had kept meticulous neuroimaging, neuropsychological assessment, and concussion history records, enabling longitudinal evaluation.

Neurocognitive assessments indicated a five-year decline in several functional areas, including executive function, verbal reasoning, motor skills, narrative memory, and word retrieval. The MRI records showed diffuse cortical thinning in several brain areas; this was most pronounced in the left frontal and temporal areas, which correlate with areas of the brain governing the neuropsychological functions that declined in this subject. Subcortical volume decrease was greatest in the globus pallidus. The [ ¹⁸ F]-florbetapir PET imaging for amyloid deposition was negative in this subject, but binding of [ ¹⁸ F]-T807 ligand indicated the presence of tau aggregation. The disparate observations of tau aggregation in these case studies and other findings indicate a need for scientific progress and consensus among experts on the interpretation of tau signal in PET imaging.

Treatment strategies

Although the neuropathological and clinical features of CTE have not been definitively established, investigators have made advances in exploring potential treatment strategies for chronic neurodegeneration, including CTE.

Dr. Kun Ping Lu, of the Beth Israel Deaconess Medical Center at Harvard Medical School, reviewed research about the role of tauopathy in TBI, CTE, and AD and the development of novel antibodies that specifically recognize the early driver of neurodegeneration, cis p-tau. Although TBI is a potential risk factor for both CTE and AD, and tauopathy is a primary neuropathological characteristic of CTE and AD, the pathogenic processes linking TBI exposure to the development of tauopathy remain largely unknown.

In AD studies, the cis isomer of p-tau is predominantly localized to axons and has been implicated in the pathogenesis of AD in rodents and human brain tissue. ⁵⁸ Unlike trans p-tau, the cis isomer cannot promote microtubule assembly, is more resistant to dephosphorylation and degradation, and is more prone to aggregation. When cis p-tau is converted to the trans conformation, through Pin1, a peptidyl-prolyl cis-trans isomerase, the progression of tauopathy in AD is inhibited, suggesting that the cis isomer of p-tau may be a promising biomarker and therapeutic target of AD.

Recent neuropathological studies indicate the presence of cis p-tau in human brains with CTE after TBI. The cis isoform appears at early time points in mouse brain tissue after exposure to blast- and impact-induced TBI.⁶¹ Forty-eight hours after blast, cis p-tau is localized in neurons of the cortex in the damaged side of the brain and spreads to hippocampal neurons in the ipsilateral hemisphere by two weeks and then to the cortex in the contralateral hemisphere by three months. Follow-on experiments in cell culture found that repeated TBI brain lysates potently induce cell death in 22% of neurons by apoptosis; when cis p-tau is depleted from the TBI brain lysates by cis p-tau antibody, neuron apoptosis is inhibited. In addition, cis p-tau antibody blocks the ability of neuronal stress, such as hypoxia, to induce cis p-tau and apoptosis in cultured neurons.

The cis p-tau-induced pathogenic process termed “cistauosis” is thought to mediate disease progression of tauopathy through disruption of axonal microtubule networks and mitochondrial transport, ultimately leading to apoptosis; these events are blocked by cis p-tau antibody in neuron cultures and rodents. ⁵⁹ Blockade of cistauosis in mice exposed to TBI through administration of cis p-tau antibody prevents tauopathy; prevents neuron loss and brain atrophy; and restores TBI-related functional outcomes, including risk-taking behavior, urinary dysfunction, and sensorimotor imbalance. Cistauosis is thought to represent a pathogenic mechanism linking TBI to AD and CTE; cis p-tau antibody may be used to target this mechanism for the development of early diagnosis, prevention, and therapeutic strategies for these conditions.