Study participants

Active duty United States Army personnel (n = 29) within a single site, Fort Leonard Wood, MO engaged in a two-week breacher training course. Heavy wall breaching exercises occurred within one training day during which neurocognitive testing, blood sampling, and symptomology assessments were conducted.

Overpressure measurements

All participants were exposed to two overpressure events (back-to-back heavy wall breaches to defeat concrete walls) in a single training session. Exposure levels were measured as psi (pound per square inch) using the B3-H pressure sensor mounted on the left shoulder of each participant to approximate incident pressure. The B3-H is a small, lightweight, accurate, disposable, and off the shelf device that records and collects data peak pressure, acceleration (rate at which speed changes) and impulse (time exposed to certain levels of overpressure exposure). Peak pressure (psi) and impulse (psi X milliseconds [ms]) are displayed for each incident and as cumulative values for the training session. Participants are in a static position during exposure. Therefore, acceleration does not occur and does not meet the threshold to be automatically recorded by the B3-H sensor.

Assessment of neurocognitive performance

The Defense Automatized Neurocognitive Assessment (DANA) tool was administered prior to (pre: -8h) and after (post: +1h) Exp in accordance with blood-draw time. The DANA consists of three subtasks conducted with a hand-held device and monitor screen. (1) Simple reaction time (SRT) measures pure reaction time. The participant was required to tap on the location of the yellow asterisk symbol as quickly as possible each time it appeared; (2) Procedural reaction time (PRT) is a choice reaction time that measures accuracy, reaction time, and impulsivity. The screen displays one of four numbers for 3 seconds (sec). The participant was required to press a left button (“2” or “3”) or right button (“4” or “5”). This choice reaction time task targets simple executive functioning and working memory; and (3) Go-No-Go (GNG) is a forced choice reaction-time task. A picture of a house is presented on the screen. Either a “friend” (green) or “foe” (white) appeared in a window. The respondent must push a “fire” button only when a “foe” appears. The choice reaction time measures sustained attention and impulsivity. The test quantifies speed and accuracy of target omissions and commissions.

Symptom and TBI/concussion history reporting

Participants completed a 32-item, paper-and-pencil health symptom inventory before (pre-) and after (post-) Exp, in conjunction with each blood draw. The symptoms on the inventory are similar to that of the Rivermead instrument [10, 11], but with additional items and responses relevant to the breaching exercise context rather than exclusively to concussion [12]. Participants were instructed to use a 5-point Likert scale (0 “not experienced at all,” 1 “no more of a problem than before training,” 2 “mild problem–present but don’t really notice and doesn’t concern me,” 3 “moderate problem–I can continue what I am doing but I notice the problem,” 4 “severe problem–constantly present, feels like it could affect my performance”). Participants noted prior history of concussion or mTBI which is reported as a binary metric (No = 0, Yes = 1). Clinical data was not available.

Serum preparation and quantitative biomarker measurements

Venous blood was collected directly into BD Vacutainer SST Serum Separation Tubes (Fisher Scientific, Waltham, MA) and processed within 30 minutes according to the manufacturer’s instructions. Samples were centrifuged at 1,000 x g for 10 minutes, at room temperature. Samples were stored in 1mL aliquots, supplemented with HALT protease/phosphatase inhibitors, and then stored at -80 C until use. GFAP, UCH-L1, Nf-L, tau, Aβ-40, and Aβ-42 were measured using digital immunoassays performed using the Simoa HD-1 according to manufacturer’s instructions (Quanterix Corporation, Lexington, MA). All assays were performed based on manufacturer’s recommendations. Briefly, serum was thawed on ice then centrifuged at 10,200 x g for 10 minutes at 4° C. Thereafter, 120μL of serum supernatant was directly loaded onto a 96 well plate and diluted 1/4 during the assay. Curve fitting analysis was conducted using pre-set programs designed by the manufacturer.

Data management and statistical analysis

The full dataset containing age (years), sampling time (hours), peak pressure (psi and kPa), impulse (psi X ms, DANA values (ms), biomarker concentrations [pg/mL], and dichotomized symptomology or mTBI/concussion history for each participant is shown (S1 Table). Non-dichotomized, Likert scale symptomology reporting is provided (S2 Table). All data was analyzed using Prism version 7 (GraphPad, La Jolla, CA). Biomarker concentrations [pg/mL] were compared using the Wilcoxon signed rank test, * p ≤ 0.05. Data is shown as the median concentration [pg/mL] +/- IQR. Symptomology was transformed into binary variables (No change or a decrease = 0 or “-”; an increase = 1, or “+”). Biomarker and DANA values were transformed into a delta (d = post-Exp minus pre-Exp) and outliers were removed (ROUT = 1%) prior to correlative analysis using 1-tailed, Spearman rank correlation coefficient, * p ≤ 0.05, or to comparisons against dichotomized (—, no change or decrease vs. +, increased) symptom data. Distribution of delta DANA or delta biomarker values were tested for normality using the D‘Agostino & Pearson test prior to comparison using 1-tailed Mann-Whitney U-Test or Welch’s t-Test as appropriate, * p ≤ 0.05, and are displayed as the median +/- 5–95%-ile range. Delta DANA or biomarker values compared to self-reported symptomology or to mTBI/concussion history are displayed (S3 Table).

A subset of TBI-related proteins are potential biomarkers of acute overpressure exposure based on evaluation of group effects

The utility of blood-based biomarkers are increasingly investigated for concussion or subconcussion. Breacher’s brain symptomology caused by overpressure exposure is similar to that of mTBI or concussion. Therefore, we hypothesized that biomarkers would also have utility for symptomatic overpressure exposure.

GFAP and UCH-L1 are, perhaps, the most thoroughly studied as biomarkers for moderate-severe TBI or closed head hemorrhagic injury. Levels dramatically increase in the serum or plasma within 12-24h [13–16]. GFAP and UCH-L1 are not typically evident in patients with an mTBI unless hemorrhage or intracranial lesions are presented [17, 18]. However, composite assessment of GFAP and UCH-L1 (in addition to spectrin break down product (SBDP-150) revealed that these proteins were elevated in blood collected from study participants who had the most striking decrements in neurocognitive performance, including simple reaction time, as well as symptom reporting [19] in the absence of a clinically defined concussion. UCH-L1 was also increased in serum collected two weeks after training from a subset of participants, but these results were not linked to neurocognitive performance or symptomology [20]. Participants in the present study do not suffer from complicated mTBI within this context. The lack of a robust post-exposure UCH-L1 response among the cohort is not surprising. The moderate drop in median GFAP levels is not known in the context of physiology. However, this observation presents a novel observation that may be deserving of further investigation, specifically in the context of overpressure exposure and its systemic outcomes.

Nf-L and tau are two of the most abundant cytoskeletal proteins in both the peripheral (PNS) and central nervous system (CNS). Both have recently become more prevalent as potential biomarkers of brain trauma, neurological disease, and repeated concussion [21–23]. In the context of low level exposure without a direct impact to the head, this study indicated that median levels of serum Nf-L and tau were elevated (although tau did not meet statistical thresholds). Exposure is reported to impact brain tissues in a way that may mirror a sub-concussive event [24, 25], causing cytoskeletal abnormalities and demyelination in rodent models [26–28]. In animal models, serum tau is elevated within 6h-1d of and the heavy chain isoform of neurofilament is increased within 2h after mild exposure [29]. Serum Nf-L is reported to increase acutely, after sub-concussive head impacts when viewed in the context of TBI status [30]. Both proteins are elevated in blood 1-6h hour after play among athletes who have prolonged return to play status [31–33].

This work indicated that median Aβ-42 was elevated one hour post-exposure. Aβ peptides are toxic monomers shown to be crucial to pathogenesis of chronic neurodegenerative diseases, such as Alzheimer’s disease (AD) [34] or chronic traumatic encephalopathy [35]. Aβ is increased in the brains derived from veterans with a history of between chronic exposures caused by blast, which may offer associative or causative relationship to neurodegenerative diseases. Assessment of Aβ levels in blood is primary viewed in the context of cognitive decline or advanced age wherein Aβ levels in the blood typically decrease in accordance with increased plaque burden in the brain [36]. The effect of Aβ in blood collected from cognitively normal, yet acutely injured and symptomatic, subjects is not fully understood. Recently, increased levels Aβ peptides have been detected in the blood of active duty and veteran populations. Plasma Aβ-40 is elevated among service members who sustained a clinically diagnosed TBI or experienced chronic symptoms associated with PTSD [37, 38]. Interestingly, increased serum Aβ is also associated with hypoxia or hypoxemia [39, 40], which is proposed to occur as a consequence of altered cerebral blood flow after blast overpressure exposure [41, 42].

When study participants are viewed collectively, assessment of median biomarker values indicate that serum GFAP and NfL decrease while Aβ-42 increases after overpressure exposure. These biomarker changes occur within the same time frame as SRT decrement, although there is no overall correlation to the DANA metrics within the cohort of participants. Dichotomizing biomarker changes according to breacher’s brain symptomology may offer additional insight regarding individual and group post-exposure responses.

Acute shifts in biomarker levels and neurocognitive decrements are associated with post-exposure symptomology, not mTBI or concussion history

Breacher’s brain symptomology, such as post-exposure headaches, dizziness and taking longer to think, are well established [43], yet typically studied in relation to mTBI/concussion diagnosis and chronic blast exposure. The current study indicated that low levels of overpressure exposure consistently have this effect, even at early time frames. Neurocognitive decrements have been proven to be useful during acute timeframes [44]. Therefore, association of SRT, PRT, and GNG decrements with the top three reported symptoms among participants is fitting for blast exposure.

Interestingly, decreased GFAP or NfL, elevated UCH-L1, and, to a greater extent, increased Aβ peptide levels were detected in serum of participants who reported increased dizziness after exposure. Acute GFAP and NfL suppression among symptomatic participants was surprising, and may appear to be contrary to the effects shown for concussed athletes, wherein biomarkers generally increase among cohorts with a clinically defined concussion [32]. However, it is notable the pre-game levels were not determined and that the temporal dynamics indicate a decrease 1–12 hours after play. Suppressed GFAP in relation to symptomology remains a novel observation.

The change in serum UCH-L1 was positive (e.g. increased) among the subset of participants who reported post-exposure dizziness. Previously, UCH-L1 levels were shown to be unchanged in serum derived from symptomatic concussion patients who were negative for CT abnormalities [45]. However, UCH-L1 was elevated in serum one hour after game-play among subconcussive football players, although the relationship to specific symptoms were not determined [46].

Direct relevance of acute Aβ-peptide elevation is not well known in the context of symptomology caused by overpressure exposure or subconcussive paradigms. Rather, fluctuation in blood Aβ are largely understood in the context of subacute-chronic symptomatic mTBI, after activities such as boxing, without stratification of specific symptoms [47]. However, mass spectrometry-based proteomics of serum revealed peptides (proteins) that were specifically associated with the decree of PTSD or post-concussive syndrome symptomology among veterans who suffered a mTBI [48]. It is possible that acute Aβ elevation among symptomatic participants in this study are aligned with these observations. Overall, this study is the first to show dysregulation of blood biomarkers, specifically Aβ, are aligned with symptomology that is common among participants exposed to blast overpressure.

This study is not without a few potential caveats. First, symptomology and mTBI/concussion history may be under-reported [49–51]. Changes in UCH-L1 levels were higher in serum of participants who reported prior mTBI/concussion based on self-report. However, definitive medical evidence is not available. The change in UCH-L1 after low levels of overpressure exposure is considered a small effect that will remain under consideration for future studies. Second, amyloid precursor protein expression and release of Aβ peptides may occur outside of the CNS, including the epidermis and muscle and leak into the blood stream [52, 53]. However, there were no reports of tissue injury among participants. Lastly, circadian variation of Aβ levels among healthy controls (< 5%) [54, 55] has been reported, but the changes induced by exposure within this study eclipsed those found associated with circadian patterns.

To our knowledge, this work is the first to explicitly report changes in biomarker levels in the context of acute overpressure exposure compared to pre-exposure values. The strict definition of a clinical mTBI/concussion was not met within this paradigm. However, acute measurement of proteins in serum, particularly Aβ peptides, coupled with symptomology and neurocognitive assessment may provide a novel biomarker relevant to subconcussive effects of blast overpressure exposure. Acute serum biomarker dynamics among overpressure-exposed persons, who have subconcussive breacher’s brain symptoms, may be worth further collective evaluation particularly in the absence of a clinically defined mTBI/concussion.