Review of blast noise and the auditory system

Abstract

The auditory system is particularly vulnerable to blast injury due to the ear’s role as a highly sensitive pressure transducer. Over the past several decades, studies have used a variety of animal models and experimental procedures to recreate blast-induced acoustic trauma. Given the developing nature of this field and our incomplete understanding of molecular mechanisms underlying blast-related auditory disturbances, an updated discussion about these studies is warranted. Here, we comprehensively review well-established blast-related auditory pathology including tympanic membrane perforation and hair cell loss. In addition, we discuss important mechanistic studies that aim to bridge gaps in our current understanding of the molecular and microstructural events underlying blast-induced cochlear, auditory nerve, brainstem, and central auditory system damage. Key findings from the recent literature include the association between endolymphatic hydrops and cochlear synaptic loss, blast-induced neuroinflammatory markers in the peripheral and central auditory system, and therapeutic approaches targeting biochemical markers of blast injury. We conclude that blast is an extreme form of noise exposure. Blast waves produce cochlear damage that appears similar to, but more extreme than, the standard noise exposure protocols used in auditory research. However, experimental variations in studies of blast-induced acoustic trauma make it challenging to compare and interpret data across studies.

1. INTRODUCTION

Blast injuries are a prevalent concern in modern society and occur in many settings, such as improvised explosive device (IED) detonations in military combat, terrorist bombings unleashed on unsuspecting civilians, and accidental catastrophic explosions (Bruins and Cawood, 1991) like the recent warehouse explosion in Beirut. There are four categories of blast injuries: primary injury from the blast wave directly impacting the human body, secondary injury caused by projectiles lifted by the blast wave, tertiary injury from trauma caused by the displacement of the victim by the blast wind, and quaternary injury from all other causes not otherwise categorized, such as burns, toxic exposure, and post-traumatic stress disorder. Primary blast injury chiefly affects gas-filled organs, namely the lungs, the gastrointestinal tract, and the external and middle ear space. The auditory system is especially sensitive to blast injury since it is designed to detect minute changes in air pressure. Blast-related auditory pathologies can range in severity and often include hearing loss, tinnitus, hyperacusis, and otalgia.

Animal studies of blast injury of the auditory system gained prevalence in the mid to late-1900s after the widespread use of explosive weaponry during World War II. Many of these early studies simply described the structural damage sustained by the peripheral auditory system including tympanic membrane perforation, ossicular chain discontinuity, hair cell loss, and tearing of the organ of Corti (Garth, 1994; Henderson and Hamernik, 1986; Hirsch, 1968). Since then, a breadth of literature has emerged seeking to further map out the structural and molecular events occurring in the peripheral and central auditory system that underlie blast-related hearing disturbances. Scientists have also begun to explore therapeutic strategies to regulate the markers and pathways involved in auditory blast injury. Though recent review articles cover some of this work, with a review by Zhang (2019) thoroughly summarizing findings from animal studies of blast-induced tinnitus and another review by Choi (2012) providing a general discussion about known mechanisms of blast-induced hearing loss, there still remains a significant range of studies on blast-related auditory sequelae not otherwise covered by these articles due to the reviews’ specific scopes and the developing nature of this field. A comprehensive discussion about the effects of a blast overpressure wave on the peripheral and central auditory pathways is warranted to discuss key recent findings like the association between endolymphatic hydrops and loss of cochlear synapses after blast (Kim et al., 2018), novel applications of previously studied therapeutic agents (Lu et al., 2021), and notable biomarkers of the auditory pathways affected by blast injury (Lu et al., 2021; Y. Wang et al., 2020). In the present review, we provide an overview of the sequelae of blast injury on middle and inner ear structures, introduce recent findings that provided better mechanistic explanations of blast-related cochlear and central auditory system disorders, discuss new therapeutic strategies, and bring attention to areas of research that require further investigation to bridge gaps in our knowledge about blast-induced auditory dysfunction.

2. EXPERIMENTAL WAYS TO DELIVER A BLAST

2.1. Blast, impulse noise, and impact noise

First, it is important to define what a blast is and how it is used to create a model for hearing loss for research purposes. A blast is characterized by an instantaneous rise in atmospheric pressure, typically caused by the conversion of explosive material into compressed gas. This sudden release of energy creates a large pressure front, or positive overpressure, that propagates at supersonic speeds, also known as a shock wave. The initial shock wave is often followed by wind and combustion products that all impose significant harm to the structures in their path, including structures of the peripheral and central auditory system.

Though distinctions are often made between blast, impulse noise, and impact noise, these realistically exist on a continuum of short-duration noise exposures that can cause acute acoustic trauma (Figure 1). The clearest distinction can be made between blast and impact noise. Blast waves ideally resemble the Friedlander waveform with a sharp onset positive peak, followed by a decay with a negative pressure phase (Figure 2A). Impact noise, on the other hand, is frequently encountered in the industrial setting and refers to the noise created by objects colliding (e.g. hammering, mechanical clatter). Impact noises have the sharp onset of a positive peak pressure and then rapidly decay back to baseline. However, they are often repetitive and so waveforms may overlap one another. Also, depending upon the source of the impact noise, they can be reverberant, which is represented by oscillations in their pressure-time waveform (Crocker and Arenas, 2021) (Figure 2C).

Figure 1: Impulsive Noise Continuum.

Impact noise, impulse noise, and blast exist on a continuum of high-intensity, short-duration noise exposures.

Figure 2: Blast-Impulse-Impact-Noise Waveforms.

(A) The Friedlander waveform represents an idealized blast waveform with a positive phase and a negative phase. Blast stimuli are often recreated in the lab. (B) The impulse waveform is characterized by a rapid rise and decay of the sound pressure level and is generally less intense than a blast. Impulse stimuli are often recreated in the lab. (C) The impact waveform is frequently characterized by oscillations and reverberations and will generally have a peak pressure of less than 140 dB SPL. Impact noise is not often recreated in the lab. (D) The noise waveform is the least specific waveform as noise is often administered continuously. Continuous noise is not further discussed in the text.

Hamernik and Hsueh (1991) cite an approximate peak overpressure of 140 dB SPL as the threshold differentiating blasts from impact noise. This approximate threshold is significant because (1) shock waves from peak sound pressure levels below 140 dB SPL are weak and not sustained for long, and (2) pure tones in the mid-frequency range at or above 140 dB SPL have waveforms that begin to be distorted by nonlinear physical properties such as speed, temperature, and viscosity (Hamernik and Hsueh, 1991; Webster and Blackstock, 1976). Most industrial impact noises fall below 140 dB SPL and lack the intense shock wave associated with a blast.

The distinction between impulse noise and blast noise is more arbitrary. The term ‘impulse’ is defined as the absolute energy or pressure associated with a blast wave (Shirbhate and Goel, 2021). However, ‘impulse noise’ as a term is often used broadly to describe various types of high-intensity, short-duration noise exposures with a rapid rise and decay of the sound pressure level (Figure 2B). While this may refer to the noise generated from blast or from gun fire, it has also referred to industrial impact noises depending on the author’s definition. For example, Hamernik and Hsueh (1991) use impulse noise and blast interchangeably, with the only distinction made being that lower intensity impulse noises are indistinguishable from impact noise. Garth (1994) on the other hand likens impulse noise more to impact noise based on his following criteria: (a) the peak overpressure of impulse noise is typically less than 2 kPa (160) dB while a muzzle blast from gunfire can produce tens of kPa, (b) blast involves significant air and combustion products while impulse noise does not, (c) Impulse noise is often associated with low frequency mechanical clatter.

2.2. Blast Waveforms Vary with Environment and Experimental Condition

A blast waveform in realistic conditions will vary based on the environment in which it occurs. Free field blasts in open environments most closely resemble a simple Friedlander wave because surface reflections are minimized (Hamernik and Hsueh, 1991). In contrast, blasts in enclosed spaces produce surface reflections that can prolong the positive phase of the waveform with additional peaks and oscillations, as shown in a study by Richmond et al., (1959) in which steel plates were placed in the blast chambers to study the effects of tertiary blast injury. Similarly, modifications of the blast waveform will occur with varying experimental conditions in the laboratory setting which can contribute to the challenge of inter-study comparisons (Needham et al., 2015). Thus, it is important to consider the experimental factors such as type of blast chamber, source of blast wave, and animal model when recreating an experimental model of blast-induced acoustic trauma.

2.3. Blast Chamber Designs

The most common blast chamber designs used in animal models of blast-induced hearing loss are shock tubes (Choi, 2012; Hickman et al., 2018) and advanced blast simulators (ABS) (Figure 3). Conventional shock tubes consist of a high-pressure driver section constraining the pressurized or explosive charge and a low-pressure expansion section through which the shockwave propagates (Figure 3). They are simple and compact in design, making them more accessible for laboratory use (Newman et al., 2015). However, issues commonly encountered with their use include late-time reflections generated from the blast wave reverberating along the length of the tube which can cause additional injury and blockage caused by the animal subject obstructing the tube, which alters the blast wave profile (Needham et al., 2015). Use of a conical tube versus a constant-diameter tube, utilizing a larger ratio of tube size to animal subject size, and placement of the animal subject outside the tube may circumvent some of the issues mentioned, though at the risk of introducing additional biomechanical alterations to the blast wave such as exaggerated blast winds (Needham et al., 2015).

Figure 3: Conventional Shock tube versus Advanced Blast Simulator.

Shock tubes are generally composed of a high-pressure driver chamber containing the source of the blast and a low-pressure expansion chamber through which the blast propagates. Advanced blast simulators are larger and more sophisticated versions of shock tubes with absorptive material placed at the end of the tube to reduce reflections.

Advanced Blast Simulators (ABS) are larger and more sophisticated versions of shock tubes created to address the shortcomings of conventional shock tubes. The ABS typically has an absorptive material placed at the end of the test section to diffuse the oncoming shockwave, thereby mitigating end-time reflections and exaggerated blast winds to produce a waveform that best matches that of free-field blasts using explosives (Arun et al., 2020; Gan et al., 2020). However, an ABS is much larger and more expensive than a shock tube and may not be a realistic option for many labs given the constraints of resources and space.

Other considerations for blast chamber designs include: (1) the use of diaphragms or membranes separating high and low-pressure areas within the chamber, which allows for a sharper onset of blast peak overpressure with better customization of the peak overpressure intensity (Arun et al., 2020; Hickman et al., 2018; Manohar et al., 2020; Newman et al., 2015) and (2) source of the blast. Blasts may be generated using explosive detonations or pressurized air/gas. Some of the earliest animal studies of blast injury were conducted using high-explosive detonations on dogs, pigs, rabbits, guinea pigs, and mice. These studies provided initial data points about threshold relationships between blast pressure and biological events like tympanic membrane perforation and mortality (Bowen et al., 1968; Richmond et al., 1989, 1959). Studies have increasingly moved away from explosives and now the predominant approach is to use pressurized air/gas due to for simplicity and safety. The advantages and disadvantages of blast chamber designs and considerations are further discussed in Table 1.

Table 1.

Experimental Blast Chamber Types – Advantages and Disadvantages

Blast Chamber Designs	Advantages	Disadvantages	Relevant studies/citations
Shock Tubes	- Simple design, easier to assemble - More affordable - Conical shock tubes may more closely approximate spherical blast wave	-Can generate additional reflected pressures including late-time reflections that injure the animal after the initial insult - Blast wind larger than expected - Constant diameter tubes may create blast wind that are non-spherical	- Cho et al., 2013 - Hickman et al., 2018 - Newman et al., 2015 -Needham et al., 2015
Advanced Blast Simulators	-Produces free field Friedlander waveforms comparable to field blasts using explosives -Eliminates artifacts commonly seen with shock tubes	- Expensive - Design may be more complicated - More difficult to manipulate/customize	- Y. Wang et al., 2020 - Arun et al., 2020
Gas/air-driven vs. explosive-driven blast	Gas - Experiments may be more repeatable - Less expensive - More customizable Explosive - May better replicate field blasts in real-world scenarios	Gas -Less closely approximates a Friedlander blast wave Explosive - Costly - Safety concerns with storage and handling - Smoke and gas emission can cause quaternary blast injuries - Not conducive to large number of tests	- Risling and Davidsson, 2012 - Stewart and Pecora, 2015 - Liu et al., 2015 - Kumar and Nedungadi, 2020
Rupturable diaphragm/membrane separating high and low pressure regions	- Improved control of blast wave intensity and onset depending on type of material used and number of layers (i.e., Mylar, brass foil, aluminum foil)	- Membrane/diaphragm may slip at higher pressures - Challenging to perform multiple consecutive blasts within short/realistic timeframe because need to replace membrane with each blast	- Newman et al., 2015 - Hickman et al., 2018 - Manohar et al., 2020

2.4. Laser Induced Shock Waves

Laser-induced shock waves (LISWs) are photomechanical waves generated by irradiating a solid material with high-powered laser pulses. The laser target, composed of an absorptive material such as black rubber, is typically placed atop tissue at the area of interest and covered by an optically transparent material like polyethylene terephthalate. A short laser pulse pointed at the target is absorbed by the rubber, which creates a plasma. The subsequent expansion of the plasma creates a shock wave that is directed into the tissue by the transparent covering sheet. Using this method, the shock wave energy can be easily controlled by adjusting the laser fluence, defined as the time integrated flux or energy per unit area, and size of the laser target. Historically, LISWs have been used to replicate blast-induced pulmonary injury (Satoh et al., 2010), traumatic brain injury (Sato et al., 2014), and acoustic injury (Kurioka et al., 2014; Niwa et al., 2016). When positioned post-auricularly on research animals, LISWs cause direct disruptions to the stereociliary bundle of the cochlear outer hair cells (OHCs), resulting in pure sensorineural hearing loss (SNHL). Reduced numbers of inner hair cells (IHCs) and spiral ganglion cell (SGC) synaptic ribbons have also been observed after LISW, which may further contribute to inducing SNHL. Because the lasers are directly aimed to deliver shock waves to an individual target, a major advantage of LISW is its anatomical specificity and the minimization of collateral organ damage that may produce confounding effects. Additionally, the ability to induce one-sided injury using this model allows animal subjects to serve as their own experimental controls. Lastly, the relatively simple experimental setup does not require blast generators or sizable blast containment apparatuses, which may appeal to experimenters with physical space limitations. The downside, of course, is that there is no actual blast wave created using this model.

3. ANIMAL MODELS USED IN HEARING LOSS RESEARCH

There are four main animal models widely used in hearing loss research: mouse, rat, guinea pig, and chinchilla. Guinea pigs and chinchillas have more anatomical and frequency range similarities to humans. They also have more docile temperaments and longer lifespans (5–7 and 15–20 years, respectively) than laboratory mice and rats (both <3 years), making long-term behavioral studies possible. Additionally, because of their larger body sizes, guinea pigs and chinchillas have larger tympanic membranes and middle ear spaces than mice and rats, allowing for easier anatomical and physiologic manipulation in surgery and drug delivery. However, their highly specific feeding and housing requirements, breeding limitations, and higher expense preclude their use more commonly. Additionally, guinea pigs and chinchillas are protected species under the Animal Welfare Act (AWA) which strictly regulates their use and handling in research. Mice and rats, on the other hand, are excluded from the AWA and are more commonly used because they are readily available from commercial vendors, easily housed in large numbers, and are genetically manipulable. Interspecies comparisons between animal models are challenging due to differences in cochlear anatomy and spectral tuning. For example, a 1 ms duration Friedlander wave contains predominantly low-frequency energy (Henderson and Hamernik, 1986) which may be more traumatic to the ear of a chinchilla tuned to lower frequencies versus a mouse’s ear which is tuned to higher frequencies. Further explanations of the applications of the animal model along with references to excellent in-depth review articles about each species are provided (Table 2).

Table 2.

Animal Models of Blast-Induced Acoustic Trauma

Animal Model	Advantages	Disadvantages	Relevant studies / Review articles
Mouse Common strains: - CBA/Ca - CBA/CaJ - C57BL/6J	- Generally obtained from commercial vendors and are therefore commonly pathogen-free - Wide variety of genetically modified strains readily available - Easy to breed and house - Familiarity of veterinary and husbandry staff to care and handling of species - Auditory function and pharmaceutical effects have been thoroughly established and well-documented in most strains	- Higher frequency range of hearing than humans - Relatively short lifespan (~6 mo to 3 years) does not allow for long-term studies - Age-dependent sensitivity to noise makes experimental timing a crucial component to consider - Inbreeding may cause a multitude of confounding health issues - Rapid metabolism and small size make drug dosing and testing difficult	- Cho et al., 2013 - Lynch et al., 2016 - Ohlemiller, 2019 - Liberman and Liberman, 2015 - Kim et al., 2018 - Mizutari et al., 2013 - Gratton et al., 2011
Rat Common strains: - Sprague-Dawley - Long-Evans - Fischer 344	- Larger body size than mice, making auditory functional testing, surgical procedures, and experimentation easier - Well-established auditory thresholds in several rat strains	- Higher frequency range of hearing than humans - Sensitivity to noise is agedependent	- Niwa et al., 2016 - Escabi et al., 2019 - Holt et al., 2019 - Y. Wang et al., 2020 - Ewert et al., 2012 - Mao et al., 2012 - Luo et al., 2014a - Masri et al., 2018 - Kurioka et al., 2014 - Yu et al., 2020
Guinea Pig Common strains: - Hartley - Dunkin-Hartley	- Audible frequency range comparable to that of humans; better low-frequency hearing than mice and rats - Larger anatomy allows for easier experimentation - Docile nature and relative ease of training allows for behavioral studies - Longer lifespans (5 to 7 years) makes long-term experiments possible - An established animal model for studies involving acquired SNHL, noise injury, drug injury, and otoprotection against injury	- Highly regulated by USDA - Veterinary and husbandry staff may be less familiar with care and handling of species - Endemic CMV in many commercially available colonies may negatively impact use in experiments - Breeding limitations given lower fecundity and longer gestation periods compared to mice and rats - Albino and pigmented strains vary in peripheral and central auditory functions due to differences in melanin production	- Naert et al., 2019 - Atkinson et al., 2014 - Chen et al., 2013 - Hu, 1991 - H. W. Lin et al., 2011 - Robertson, 1983 - Scheibe et al., 1993 - Shi et al., 2013
Chinchilla	- Widely considered the gold-standard animal model for studying middle-ear diseases, NIHL, and functional diagnostic tests - Larger TMs and middle ear spaces than other rodent models make for easier experimental manipulation - Audible frequency range (50 Hz to 33 kHz) and intensity sensitivity are most comparable to humans compared to other animal models - Anatomical similarities to humans (e.g. wide TM area, # of cochlear turns - humans 2 ¾, chinchillas 3) - Relative genetic heterogeneity reduces issues that arise from inbreeding - Docile temperament allows for behavioral experiments - Long lifespan (15–20 years) makes long-term studies possible - Robustness of chinchillas to withstand anesthesia and surgery allows for detailed data to be collected from all levels of the auditory system - May be used to model various types of CHL and SNHL that mimic pathologies observed in humans - Long history of the chinchilla model in hearing research allows pharmacological studies to be done to investigate various drug efficacies and effectiveness of HL prevention methods	- Relative genetic heterogeneity - may preclude use in genetics-related studies - Protected species in the US with strict research regulations - Often acquired from furrier ranches, where animals with lower quality fur are sold for research – possibly associated with poorer general health - Less familiarity to the species for research veterinary and husbandry staffs - Highly specific housing requirements: cool temperature, low humidity, diet rich in fiber and protein, regular dust baths to cleanse oil and moisture buildup in fur - Breeding limitations due to longer gestation periods, extended weaning timelines - Potentially more susceptible to noise damage than other animal models - Metabolism is notably different from other rodent models, which may affect drug metabolism and comparative studies with other models	- Chen et al., 2019 - Trevino et al., 2019 - Radziwon et al., 2019 - Guan et al., 2015 - Lupo et al., 2009 - Margolis et al., 1995 - Martin, 2012 - Thornton et al., 2013 - Heffner and Heffner, 1991 - Miller, 1970 - Trautwein et al., 1996 - Wake et al., 1994 - Wang et al., 1997 - Bohne and Carr, 1979 - Salvi and Boettcher, 2008 - Snyder and Salvi, 1994 - Gerhardt et al., 1980 - McFadden et al., 1997 - Dobie and Humes, 2017 - Hickman et al., 2018 - Davis et al., 2002

4. TYMPANIC MEMBRANE AND MIDDLE EAR CHANGES

As a blast pressure wave enters the ear canal, it encounters the tympanic membrane (TM), which couples the impulse energy with the middle ear ossicles to ultimately transmit the sound and energy to the fluid-filled cochlea. The extent of injury to the TM and middle ear is modified by blast wave characteristics and orientation of the blast relative to the animal subject (Zhang, 2019). Disruption of the TM and middle ear results in conductive hearing loss (CHL) which most frequently occurs in conjunction with SNHL due to downstream effects of blast injury on the cochlea.

4.1. Tympanic Membrane Rupture and Surface Defects

TM perforation is amongst the most common injuries reported after blast exposure, occurring in up to 16% of explosion-wounded patients in the military population (Ritenour et al., 2008) and 90% of hospitalized patients after the Boston marathon bombing (Remenschneider et al., 2014). 184 dB SPL is commonly regarded as the TM rupture threshold, or the minimum level needed to produce any TM rupture, in humans (Hirsch, 1968). It is also estimated that 50% of human TMs will rupture at peak blast pressures of about 194 dB SPL (Hirsch, 1968). Similar data collected from animal models of blast-induced acoustic trauma allow for rough numerical comparisons with other species (Table 3). Peak blast pressures of ~194 dB SPL cause TM rupture in 95–100% of mice and rats (Cho et al., 2013; Lien and Dickman, 2018; Mao et al., 2021; Wang et al., 2020; Sandlin et al., 2018). Guinea pigs also demonstrate a 100% TM rupture rate in response to a 194 dB SPL peak blast wave, according to Yokoi and Yanagita (1984) and a 46% rate of rupture at 187 dB SPL according to Richmond et al. (1959). In chinchillas, the threshold for TM rupture is often cited as 190 dB SPL, though the study from which this value was derived defined the term “threshold” as the mean peak pressure needed to rupture all animal subjects’ TMs. Another study cites 175 dB SPL as the TM rupture threshold in chinchillas, defined as the lowest intensity stimulus that produced a TM rupture (Hickman et al., 2018). This definition is more comparable to the definition used to establish the TM rupture threshold in humans (Hirsch, 1968). From these data, it appears that mice, rats, guinea pigs, and chinchillas may be slightly more susceptible to blast-induced tympanic membrane perforation than humans. However, a more definitive interspecies comparison is challenging to conduct because susceptibility to TM rupture is affected by orientation to blast, pressure intensity, age, and numerous other experimental factors that have not been controlled for across different studies.

Table 3.

Blast-Induced TM Rupture Thresholds in Animal Models

Animal Model	Blast Peak Overpressure(s)	Number of Consecutive Stimuli	Rate of TM Rupture	Studies/Reviews
Chinchilla	175 dB SPL	1	20%	Hickman et al. (2018)
	190 (9.1 psi)	2–3	N/A (measured mean threshold)	Gan et al. (2016)
Mice	189 dB SPL	1	100%	Lien and Dickman (2018)
	189–199 dB SPL	1	100%	Cho et al. (2013)
	195 dB SPL (16 psi)	1	100%	Mao et al. (2021)
Rat	~193 dB SPL (14 psi)	2–3	“minimal damage”	Ewert et al. (2012)
	~194 dB SPL (15 psi)	1	95–100%	Sandlin et al. (2018)
	196 dB SPL (16 psi)	2 (separated by 2 minutes)	100%	Y. Wang et al. (2020)
Guinea Pig	~187 dB SPL (6.7 psi)	1–4	46%	Richmond et al. (1959)
	~194 dB SPL	1	100%	Yokoi and Yanagita (1984)

TM perforation is thought to provide some level of protection against further damage to the inner ear because it reduces power transmission through the middle ear. Hickman et al. (2018) exposed chinchillas to multiple blast overpressures and found that cochlear synaptopathy, defined as the degeneration of synaptic connections between the IHCs and the auditory nerve, and hair cell loss were virtually never seen amongst ears that sustained TM perforation compared to those that didn’t. Similarly, Hamernik et al. (1984) found damage to the organ of Corti (OoC) to be less severe in chinchillas after multiple impulse noise exposures that sustained TM rupture compared to those with intact TMs. Findings in human studies are limited due to the lack of histologic data, but there appear to be no significant differences in hearing outcomes between patients exposed to a single blast with or without TM rupture (Shah et al., 2014; Taşlı et al., 2021). This makes sense since the TM perforation is most likely to mitigate damage to subsequent blasts.

Even when the blast pressure does not rupture the TM, it can alter the mechanical properties of the TM. Luo et al. (2016) investigated post-blast properties of human cadaveric ears and reported a decrease in the stiffness of the TM in the circumferential direction, suggesting weakening of the circumferential collagen fibers. Liang et al. (2017) reported similar findings in chinchillas exposed to blast pressures below the rupture threshold. Although the TM did not rupture, there was a decrease in the overall stiffness of the TM and an increase in compliance. Blast-exposed TMs also had lower thresholds for rupture with subsequent blast stimuli compared to TMs that were initially unexposed, suggesting weakening of the TM despite blast pressures not reaching rupture thresholds. These post-blast changes correlate with microanatomical defects on the surface of the TM visible on scanning electron microscopy (Liang et al., 2017).

4.2. Middle Ear Changes

Middle ear damage has been reported to varying degrees after blast exposure. These involve disruption of the ossicular chain and middle ear hemorrhage and swelling (Garth, 1994; Messervy, 1972; Ríos et al., 2021; Y. Wang et al., 2020; Yu et al., 2020). In humans, ossicular chain damage often occurs at the malleolar-incudal joint, though damage to the incudostapedial joint and fracture of the stapes footplate have also been described (Garth, 1994; Messervy, 1972; Yu et al., 2020). Like TM perforation, ossicular chain disarticulation disconnects the middle ear from the cochlea, so that much of the sound energy is not transferred to the cochlea. Additionally, macrophages, intact and degenerate neutrophils, and cellular debris have all been detected in the middle ear after blast exposure, suggesting the onset of an acute inflammatory response with subsequent necrosis (Ríos et al., 2021). Some studies, however, report no evidence for middle ear damage after blast exposure (Cho et al., 2013; Newman et al., 2015; Yokoi et al., 1982). A possible explanation is that TM rupture, which was noted in a couple of these studies (Cho et al., 2013; Yokoi et al., 1982), protected downstream middle and inner ear structures during the blast exposure (Hamernik et al., 1984; Hickman et al., 2018). More likely, however, is that these small mammals have a fused malleus-incus joint (Mason, 2013), so they are not ideal models for studying how a blast would affect the human middle ear. Study variations due to the dependence of middle ear damage on stimulus intensity and orientation of the animal subject in relation to the blast shock wave may provide an alternative explanation (Zhang, 2019).

5. GROSS COCHLEAR MORPHOLOGICAL CHANGES

5.1. Mechanical Trauma to Intracochlear Tissues

It has long been thought that blast exposure causes substantial and traumatic disruption of intracochlear soft tissues. Early studies have reported findings ranging from disrupted tight cell junctions, to breaches in the reticular lamina resulting in mixing of perilymph and endolymph, to separation of the OoC from the basilar membrane (Garth, 1994; Hamernik et al., 1984; Henderson et al., 1994; Patterson and Hamernik, 1997). More recent studies, however, have found these patterns of damage to be atypical. Cho et al. (2013) saw no evidence of disrupted tissue or grossly altered morphology on cochlear histology after exposing mice to a blast wave equal to or greater than that of a typical IED injury. In addition, Kim et al. (2018) perfused gold nanoparticles throughout the perilymph of blast-exposed mice cochleae and found no breaching of the nanoparticles into the endolymphatic space, suggesting that the membranes separating the perilymphatic and endolymphatic spaces were intact. It is quite possible that tissue artifact may confound the histological analyses of earlier studies. Henderson et al. (1994) explicitly examined potential artifacts in histological preparation in both control and experimental cochleae after blast exposure. Longitudinal and radial cracks along the cochlea were observed even in unexposed control cochlear samples. Additionally, most of the earlier studies were conducted in chinchilla using multiple blast exposures versus just a single blast exposure. Another recent study exposed rats to two consecutive 16 psi blast shock waves and noted blood cells in scalae tympani and vestibuli, suggestive of disruption of the blood-labyrinth barrier. However, no mention was made of other features of gross cochlear trauma (Y. Wang et al., 2020). The intensity of the blast wave is sure to impact the degree of histologic findings after blast exposure, but it is difficult to compare the amount of blast pressure reaching the cochlea between different labs as there are often different chamber configurations, different ways of measuring the blast pressure, and different animal models. Future studies may consider systematically testing aforementioned variables while holding other factors constant to identify the experimental variables that critically affect blast-related intracochlear histologic findings.

5.2. Cochlear Blood Flow

Regulation of cochlear blood flow (CoBF) after acoustic trauma is complex, likely involving an interchange of multiple processes including autonomic innervation and metabolic factors (Shi, 2011). Most studies investigating changes in CoBF were conducted using noise exposures of varying durations and intensities, with only a handful specifically utilizing blast overpressures. Prolonged noise exposures are generally associated with an immediate and persistent reduction in CoBF (Arpornchayanon et al., 2011; Hawkins, 1971; Nakashima et al., 2003), although some studies deviate from this finding (Scheibe et al., 1993; Shin et al., 2019). Reduced CoBF is associated with various hearing abnormalities including SNHL, endolymphatic hydrops, and presbycusis (Nakashima et al., 2003). In contrast, studies measuring changes in CoBF after blast exposures or short duration high intensity noise exposures demonstrate an acute period of increased CoBF followed by a return to baseline (Chen et al., 2013; Hu, 1991; Perlman and Kimura, 1962; Prazma et al., 1983; Quirk et al., 1991). Chen et al. (2013) reported that a blast pressure threshold of 45 kPa was needed to observe these effects, with no changes in CoBF for pressures below this threshold. This supports previous findings that CoBF may vary with stimulus intensity (Scheibe et al., 1993). Transient increases in CoBF after acoustic trauma may serve a compensatory and protective role to meet the metabolic demands of an acutely injured cochlea (Attanasio et al., 2001; Chen et al., 2013). Variations in study results are likely a consequence of a wide range of study methodologies, including differing stimulus intensities, durations of exposure, and methods of CoBF measurement. In particular, the limited availability of methods to measure CoBF may play a significant role, with many earlier studies utilizing postmortem histological parameters as markers of CoBF (i.e. vessel lumen diameter and density, number, and spacing of red blood cells in the vessels) instead of directly measuring flow rate (Axelsson and Dengerink, 1987; Nakashima et al., 2003). Future studies taking advantage of new and emerging technologies for CoBF visualization and measurement may help clarify the inconsistent findings reported in the current literature (Dziennis et al., 2012; Kong et al., 2018; Reif et al., 2013, 2012; Subhash et al., 2011).

6. BLAST-INDUCED ENDOLYMPHATIC HYDROPS

6.1. Blast Injury as a Reliable Inducer of Endolymphatic Hydrops

Endolymphatic hydrops (EH), characterized by excess accumulation of endolymph in the scala media, is a disorder of the inner ear associated with symptoms such as fluctuating hearing loss, tinnitus, vertigo, and aural fullness. Despite scientific efforts to understand how EH may lead to auditory symptoms, the pathophysiologic mechanisms of EH remain unclear. Only a few studies have described EH induced by noise trauma (Kumagami, 1992), as studies of EH have historically been limited by indirect electrophysiological measurements or postmortem histology, both of which present challenges to accurately visualizing and quantifying dynamic changes in endolymph volume (Salt, 2004; Schuknecht, 1976). Optical coherence tomography (OCT) is a modern noninvasive imaging tool capable of visualizing the cochlea in vivo (Liu et al., 2017). A recent study by Kim et al. (2018) utilized OCT to image the cochlea of mice exposed to blast overpressures of 196 dB SPL. In all mice, blast exposure led to transient swelling of the scala media that normalized by 1-day post-exposure, consistent with EH. Subsequently, Badash et al. (2021) titrated levels of noise exposures ranging from 80–100 dB SPL and found that a 2 hour exposure to 100 dB SPL noise was the threshold to observe EH in mice. These findings suggest that traumatic noise and blast exposures are reliable inducers of EH in mice and may be used to recreate an animal model of acoustically induced EH.

The mechanism for blast-induced endolymphatic hydrops is thought to be caused by disruption of intracochlear fluid homeostasis. It is hypothesized that blast-induced damage of hair cells and stereocilia prevent removal of endolymphatic K+ by the stereociliary transduction channels, creating an osmotic gradient favoring the influx of water into the scala media (Kim et al., 2018). Findings supporting this hypothesis include: (1) In vivo application of a hypotonic solution to the round window in mice led to an increase in endolymph volume, suggesting a role for osmotic gradients in driving endolymph volume, and (2) Tyr-DT-A mice, which lack K+ secretion due to ablation of stria intermedia cells, did not experience EH after blast exposure (Kim et al., 2018). To date, few studies have directly investigated the role of osmotic gradients in acoustically induced EH but the recent use of OCT to study animal models of blast induced EH is a promising method to better understand how disrupted osmotic gradients regulate endolymph volume.

6.2. Endolymphatic Hydrops as a Surrogate Marker for Cochlear Synaptic Loss after Blast

Recent findings from Kim et al. (2018) and Badash et al. (2021) also demonstrate a correlative relationship between EH and degeneration of cochlear synapses, also termed cochlear synaptopathy, induced by acoustic trauma. Two key findings specifically point to a correlative relationship between the two phenomena. The first is that the threshold for formation of EH after traumatic noise exposure mirrors the threshold for formation of cochlear synaptopathy at around 95–100 dB SPL (Badash et al., 2021). The second is that application of hypertonic saline to the round window reversed endolymphatic hydrops and rescued synaptic loss in both studies (Badash et al., 2021; Kim et al., 2018). These findings together suggest that at the very least, EH can be a surrogate marker for cochlear synaptic loss after traumatic noise and that they may even share a common mechanism or pathway. In addition, osmotic stabilization through hypertonic saline application offers a therapeutic strategy for blast-induced EH and cochlear synaptopathy. It is important to note that a correlative relationship has only been established for acoustically induced endolymphatic hydrops and cochlear synaptopathy. Valenzuela et al. (2020) found that endolymphatic sac ablation in guinea pigs, which has previously been used to create animal models of endolymphatic hydrops, was not associated with cochlear synaptic loss. A shortcoming of this study, however, was that endolymphatic volume was not directly measured or visualized. Nevertheless, the relationship of cochlear synaptopathy to endolymphatic hydrops in the absence of noise exposure is unclear and may benefit from further study.

If blast induced EH and cochlear synaptopathy do share the same pathway, one possible mechanism is that endolymphatic hydrops overstimulates the IHCs, causing glutamate excitotoxicity and leading to cochlear synaptopathy. Excitotoxicity and mechanisms of cochlear synaptic loss are discussed below (sections 8.2 and 8.3). Another possible explanation is the disruption of potassium (K⁺) recycling in the inner ear. Acoustic trauma by damaging the stereocilia, is thought to prevent K⁺ uptake by the stereociliary mechanoelectrical transduction channels leading to K⁺ buildup and osmotic influx of water into the endolymph (Kim et al., 2018). The role of K⁺ in cochlear synaptic degeneration was recently investigated by Zhao et al. (2021) who showed that high levels of extracellular, or perilymphatic, K⁺ led to cochlear synaptopathy. It is unclear how disrupted K⁺ recycling in the endolymph might lead to abnormally high levels of K⁺ in the perilymph but prior studies have reported changes in the permeability of the endolymph-perilymph barrier to K⁺ after noise exposure (Konishi and Salt, 1980). Further research is warranted to clarify the molecular events underlying the relationship between EH and cochlear synaptopathy after blast exposure and traumatic noise.

7. HAIR CELL CHANGES

7.1. Outer Hair Cell Loss and Stereociliary Disruption

Widespread OHC loss is a characteristic finding of noise- and blast-induced acoustic trauma. Morphological evidence of OHC degradation can be seen within hours of a blast exposure (Ewert et al., 2012; Kim et al., 2018). It is well established that mammalian hair cells are post-mitotic and incapable of regeneration after damage (Roberson and Rubel, 1994; Ruben, 1967). As a result, widespread OHC death leads to permanent elevations in hearing thresholds, consistent with permanent hearing loss. Cho et al. (2013) demonstrated elevated DPOAE and ABR thresholds up to 70 days after blast exposure in mice that had also sustained significant loss of OHCs. The pattern of hair cell loss is generally tonotopic, with hair cell damage and death being greatest at the base of the cochlea. In addition, OHCs are more susceptible to damage and cell death than IHCs, which can remain relatively spared after blast exposure. These patterns of loss are not unique to blast exposure and have been observed with age-related changes, noise, and ototoxicity (Jensen-Smith et al., 2012). Some of the earliest findings of this pattern of damage were histopathologic assessments of sensory cells in cats exposed to acoustic trauma and ototoxic drugs (Liberman, 1990; Liberman and Dodds, 1984). Studies specifically investigating blast exposure show similar results (Cho et al., 2013; Kim et al., 2018; Yokoi et al., 1982), with a study in rats reporting IHC loss to be less than 4% at post-blast day 21, compared to OHC loss which ranged from 28–53% along the length of the cochlea (Ewert et al., 2012). The increased vulnerability of high-frequency basal OHCs compared to IHCs and lower-frequency apical OHCs is attributed to differences in energy metabolism and reactive oxygen species production (Hill et al., 2016; Jensen-Smith et al., 2012). Tonotopic differences in OHC calcium homeostasis may also play a role (Fettiplace and Nam, 2019; Furness, 2015; Hill et al., 2016; Yuan et al., 2010).

The inciting cause of OHC loss after blast exposure is likely multifactorial. One mechanism involves direct mechanical stress from the blast waves, which at higher pressures can be severe enough to tear structural components of the OoC including the OHC matrix (Hamernik et al., 1984). This mechanical force acutely disrupts the integrity of the OHC plasma membrane, leading to apoptosis immediately following blast exposure (Hu and Zheng, 2008). More commonly, however, the blast pressure imparts excessive mechanical stress on hair bundles, damaging the stereocilia (Liu et al., 2011). Characteristic morphological changes include fusion, splaying, and bending or breakage of stereocilia at its attachment to the OHC cuticular plate (Hamernik et al., 1984; Kim et al., 2018; Niwa et al., 2016). Kim et al. (2018) compared post-blast cochlear morphology in control CBA/CaJ mice and Tecta^{C1509G/C1509G} mice, which are characterized by a tectorial membrane that does not deflect the stereocilia in response to acoustic stimuli due to the tectorial membrane being detached from the sensory epithelium. The Tecta^{C1509G/C1509G} mice sustained no loss of OHCs compared to blast-exposed control mice, in which stereociliary bundle damage and loss of high-frequency basal OHCs were routinely seen. These findings are consistent with acoustically-induced disruption of the stereocilia being the inciting factor leading to OHC loss.

7.2. Blast-related Changes in Hair Cell Gene Expression

Though hair cell loss is one of the most apparent morphological changes sustained by the OoC after blast-induced acoustic trauma, the molecular mechanisms underlying these changes are not well established. Recent work in the past several years has sought to better characterize how exposure to blast overpressures influences cochlear gene expression (Table 4). Notch signaling is a key pathway underlying the regulation of inner ear cell fate and differentiation during embryonic development. Atoh1, an embryonically expressed gene essential for the development and regeneration of hair cells, is regularly repressed by the Notch signaling pathway to inhibit hair cell fate and promote the development of support cells (Ryan et al., 2015). A study by Mizutari et al. (2013) found that exposure to 2 hours of 116 dB SPL noise activated Notch signaling in mice. This was evidenced by the increased mRNA expression of Hes5, a downstream target of the Notch signaling pathway and a known repressor of Atoh1 (Mizutari et al., 2013). Similarly, another study found Atoh1 expression to be significantly reduced in rats exposed to two blast overpressures of 16 psi in rapid succession (Y. Wang et al., 2020).

Table 4.

Changes in Cochlear Gene/Protein Expression after Blast

Hair Cell Gene/Protein Expression Changes After Blast
Name of Gene/Protein	Function	Expression	Animal Model	Citation
Atohl	Hair cell development and regeneration; Inhibited by Notch signaling	Downregulated 1 and 28 days post-blast	Rat	Y. Wang et al., 2020
Hes5	Downstream target of Notch signaling; Repressor of Atohl	Upregulated until day 3 post-blast	Mouse	Mizutari et al., 2013
Pou4f3	Late differentiator of hair cells; Cell survival promoter	Upregulated 1 and 28 days post-blast	Rat	Y. Wang et al., 2020
P21cip1	Anti-apoptotic activity, localized to hair cells after noise exposure; May contribute to resistance to NIHL	Upregulated 6 hours post-noise exposure	Mouse	Gratton et al., 2011
MyoVIIA	Tenses the tip links of the hair cell mechanoelectrical transduction complex (Li et al., 2020)	Upregulated 1 and 28 days post-blast	Rat	Y. Wang et al., 2020
Lhfpl5	Tetraspan membrane protein that regulates transducer channel conductance in hair bundles	Dowregulated 1 day post-blast	Rat	Y. Wang et al., 2020
Cdh23	Cadherins forming fine filaments involved in mechanotransduction	Downregulated 1 and 28 days post-blast	Rat	Y. Wang et al., 2020
Auditory Nerve Gene/Protein Expression After Blast
Name of Gene/Protein	Function	Expression	Animal Model	Citation
Iba1	Protein expressed in activated microglia	Upregulated 1 and 7 days post-blast	Mouse; Humans; Rat	Cho et al., 2013; Liu et al., 2018; Rios et al., 2021
Gpnmb	Gene expressed in microglia	Upregulated 1 and 28 days post-blast	Rat	Y. Wang et al., 2020
MCP-1/CCL2	Gene involved in neuron-immune cell interactions; Involved in microglia activation	Upregulated 1 day post-blast, downregulated 10 days post-blast	Rat; Humans	Y. Wang et al., 2020; Z. Wang et al., 2020

Blast and noise exposure may also be associated with upregulation of genes involved in the promotion of hair cell survival. Pou4f3, a gene recognized as a cell survival promoter and a late differentiator of hair cells (Ryan et al., 2015; Zheng and Zuo, 2017), was found to be significantly upregulated in rats 1 and 28 days after blast exposure (Y. Wang et al., 2020). Similarly, increased expression of P21cip1, a gene associated with anti-apoptotic activity, was localized to hair cells of mice exposed to 1 hour of 105 dB SPL noise (Gratton et al., 2011). Other changes include the significant downregulation of Lhfpl5, Cdh23, and MyoVIIa, all genes involved in various components of hair cell mechanotransduction (Y. Wang et al., 2020). These findings are consistent with the morphological damage sustained by stereocilia after blast exposure. In general, studies regarding hair cell transcriptional changes after blast exposure are sparse and it is difficult to make impactful interpretations of the published data. Further research is needed to understand the genetic basis of hair cell pathophysiology. In addition, it is unclear whether this kind of research would be beneficial in stimulating hair cell regeneration after blast damage if the severe damage also affected supporting cells. Supporting cell damage after blast exposure has barely been studied, but this is relevant because gene therapy for hair cell regeneration is likely going to be based upon targeting supporting cells that have the potential to produce new hair cells.

8. AUDITORY NERVE CHANGES

8.1. Auditory Nerve Degeneration is Characterized by Neuroinflammatory Markers

Blast-induced acoustic trauma leads to significant degeneration of spiral ganglion neurons (SGNs), the cell bodies of the auditory afferent nerve fibers, weeks to months after the initial insult (Cho et al., 2013; Niwa et al., 2016). Similar to hair cells, SGNs in the basal turn of the cochlea appear to be most vulnerable to acoustic trauma (H. W. Lin et al., 2011; Niwa et al., 2016). The degenerative process is associated with an acute neuroinflammatory phase involving microglial cell activation. Numerous studies have demonstrated increased expression of Iba1, a protein indicative of activated microglia, specifically localized to the modiolus and the SGNs after blast (Cho et al., 2013; Liu et al., 2018; Ríos et al., 2021), suggesting that inflammatory processes are likely involved in the auditory nerve degeneration. Furthermore, neuroprotectin D1, an anti-inflammatory therapeutic agent, was recently shown to reduce glial activation in blast-exposed rats (Ríos et al., 2021), which may be a promising approach to mitigating the acute inflammation associated with auditory nerve degeneration.

In addition, a recent study reported blast-induced downregulation of genes related to synaptic transmission, neurocellular homeostasis, and neurogenesis in the rat cochlea (Y. Wang et al., 2020). In contrast, inflammatory genes appeared to be upregulated. Gpnmb, a gene strongly associated with lipid-laden macrophages and thought to induce autophagy during stress conditions, was significantly upregulated in rats on days 1 and 28 after blast exposure, as was RT-1–24, another gene involved in inflammatory processes. On the other hand, expression of chemokine ligand 2 (CCL2) which is also known as monocyte chemotactic protein-1 (MCP-1) was upregulated at post-blast day 1 but downregulated by day 28, suggesting that cytokines in the cochlea may only be upregulated in the acute phase of blast injury (Y. Wang et al., 2020). It is unclear whether these inflammatory markers were specific to the auditory nerve, the hair cells, or the general cochlea. Increased expression of CCL2/MCP-1 has also been documented in military subjects 10 days after exposure to blast, though this finding was obtained as a blood-based biomarker (Z. Wang et al., 2020). These genes are further discussed in Table 4.

8.2. Synaptic Degradation as the Inciting Cause of Auditory Nerve Degeneration

Traditional understandings of hearing loss induced by acoustic trauma cited hair cell loss as the inciting factor, with degeneration of SGNs thought to be a downstream consequence of hair cell damage. This hypothesis was supported by the observation that auditory nerve degeneration is delayed by weeks to months compared to hair cell loss which occurs just hours after noise or blast exposure (Cho et al., 2013; Kujawa and Liberman, 2009; Niwa et al., 2016). However, auditory nerve degeneration may also occur through a separate mechanism known as cochlear synaptopathy, or the degeneration of synaptic connections between the IHC and afferent SGN nerve (Kujawa and Liberman, 2009). It is thought to be one of the earliest signs of noise-induced hearing loss due to its prevalence even in the setting of reversible hearing thresholds and minimal hair cell loss (Fernandez et al., 2020; Kujawa and Liberman, 2009). Cochlear synaptopathy has been described in various animal models including the mouse, rat, chinchilla, guinea pig, and rhesus monkey (Hickman et al., 2018; Liberman and Liberman, 2015; Valero et al., 2017) but is difficult to study in humans due to scarcity of histologic data which has been the cornerstone for detecting synaptic loss in animals (Bramhall et al., 2019). Degenerative changes of the ribbon synapses have been reported as soon as 6 hours after noise or blast exposure and was previously thought to be permanent (Kim et al., 2018; Liberman and Kujawa, 2017), though studies in chinchillas (Hickman et al., 2018; Shi et al., 2013) and recent findings in guinea pigs (Hickman et al., 2020) have now demonstrated widespread regeneration of synapses months after the insult. Clinically, synaptopathy is not readily detected on conventional audiological evaluations, as threshold shifts in animal models of cochlear synaptopathy are often temporary and reversible (Shi et al., 2016). For these reasons, cochlear synaptopathy is also frequently referred to as “hidden hearing loss”.

Most examples of cochlear synaptopathy after blast exposure occur alongside marked OHC loss and permanent threshold shift due to the severity of blast compared to noise (Cho et al., 2013; Kim et al., 2018; Y. Wang et al., 2020). These additional findings may more closely mimic what military subjects or civilians exposed to blast injury would experience but make it challenging to study synaptopathy in a controlled manner. Blast injury using LISW overcomes this issue, as it can generate a less intense stimulus and create cochlear synaptopathy without inducing hair cell loss (Niwa et al., 2016). A recent study by Hickman et al. (2018) also addressed this issue by titrating blast pressure waves, delivered via a shock tube, to levels that only led to temporary threshold shifts with minimal hair cell loss. Despite intact hair cells and reversible threshold shifts, there was significant degeneration of IHC ribbon synapses, suggesting that the synaptic connections between the IHC and afferent cochlear nerve fibers are more vulnerable than hair cells to blast exposure (Hickman et al., 2018). This is consistent with what has been found with continuous noise exposures, although the synaptopathy after blast exposure was more focal (Hickman et al., 2018). These subtle but lasting changes to the IHC ribbon synapses may contribute to persistent perceptual abnormalities like tinnitus, hyperacusis, and speech-in-noise difficulties reported by some patients with recovered auditory thresholds after a blast injury (Remenschneider et al., 2014).

8.3. Cochlear Synaptopathy Mechanisms and Treatment

Cochlear synaptopathy is thought to be a consequence of glutamate excitotoxicity (Kim et al., 2019; Liberman and Kujawa, 2017). Glutamate is the primary neurotransmitter released at the IHC ribbon synapse and has been shown to induce swelling and retraction of the postsynaptic auditory nerve terminal (Puel et al., 1994; Robertson, 1983). An acoustic overexposure is thought to produce an overabundance of glutamate which may be toxic to the IHC ribbon synapse. Studies investigating the mechanisms of cochlear synaptopathy after blast exposure are limited, as most have been conducted using prolonged high-level noise exposures. While a continuous two-hour noise exposure may feasibly cause an over-release of glutamate due to prolonged overstimulation, it remains unclear as to how a blast exposure, which just lasts a few microseconds, also induces toxic levels of glutamate release (Liberman and Kujawa, 2017). Hickman et al. (2018) postulated that brief overstimulation of the IHC could lead to an extended depolarization causing an over-release of glutamate. Kim et al. (2018) postulated that post-traumatic EH and disruption of cochlear fluid homeostasis may also play a role. This was supported by their finding that osmotic stabilization both prevented cochlear synaptopathy and reversed EH (Kim et al., 2018).

Recent animal studies offer promising therapeutic strategies for cochlear synaptopathy. Kim et al. (2018) found that application of hypertonic saline to the round window rescued synaptic loss in blast-exposed mice, suggesting that restoring cochlear osmotic gradients disrupted by the blast may be a viable therapeutic approach. Another therapeutic target is neurotrophin-3 (NT3), which is a neurotrophic factor that promotes SGN survival (Kempfle et al., 2021). Round window delivery of NT3 or smaller analogues has been shown to promote the regeneration of pre- and post-synaptic components of the ribbon synapses in mice (Kempfle et al., 2021; Suzuki et al., 2016). Finally, selective blockade of calcium-permeable AMPA receptors, a subtype of glutamate receptors, via intracochlear perfusion with IEM-1460 prevented noise-induced cochlear synaptopathy presumably by preventing excitotoxicity while still mediating hearing and neurotransmission through other glutamate receptor subtypes (Hu et al., 2020). These exciting discoveries of therapeutic agents for blast- and noise-related cochlear synaptopathy relied upon a good understanding of its mechanisms, highlighting the need for further studies to clarify the molecular events that lead to synaptic degeneration after acoustic trauma.

9. CHANGES IN THE BRAINSTEM AND CORTEX

Until recently, blast-induced auditory dysfunction was primarily attributed to pathologies of the peripheral auditory system (PAS). However, there is a growing body of literature examining the structural and functional changes that occur in the central auditory system (CAS) after noise and blast exposure. Concurrent to damage of the PAS, blasts can cause direct injury to various other solid organs including the brain and central nervous system. In fact, hearing loss is frequently found in patients with blast-induced traumatic brain injury (TBI), with up to 43% of veterans with TBI affected by sub-acute hearing loss and 9% afflicted with chronic hearing loss (Hoffer and Balaban, 2011). The degree of contribution of PAS damage versus CAS damage to post-blast auditory disturbances is largely unclear. Also poorly delineated is the degree of blast-related CAS pathology that is caused through acoustic trauma occurring first at the level of the PAS versus through direct biomechanical damage to the brain parenchyma.

9.1. Changes in the Auditory Brainstem

The first relay station from the PAS to the CAS occurs at the level of the cochlear nucleus (CN) in the brainstem. Findings in the CN of rats exposed to both single and double 19 psi blast overpressures show localized degeneration of neuronal cytoskeletal elements on silver staining that is otherwise not observed in the medial geniculate nucleus (MGN) or the auditory cortex (AC), though silver staining of these larger auditory structures is a suboptimal form of analysis (Arun et al., 2021). Nevertheless, such localized degenerative findings in the CN support the notion that blast-related changes in the CN start at the PAS and are relayed to the brainstem, perhaps through axonal degeneration.

Though axonal injury was not directly investigated in the previous study, other studies have used magnetic resonance diffusion tensor imaging (MR-DTI) to analyze axonal integrity in the CN with conflicting findings. Kaliyappan et al. (2021) showed reduced axial diffusivity on MR-DTI suggestive of axonal injury in the CN of blast-exposed chinchilla while Mao et al. (2012) did not find significant changes in the CN of blast-exposed rats. The reason for this discrepancy is unclear, and additional MR-DTI studies of the CN after blast exposure may help clarify these conflicting findings. Neuroplastic hyperactivity in the CN has also been linked to the development of tinnitus after blast (Luo et al., 2014a), which will be further discussed below (Section 9.3). Understanding the blast-related changes occurring at the first relay station between the PAS and CAS can facilitate a clearer understanding of how PAS damage might mediate CAS damage, highlighting the importance of these studies.

9.2. Changes in the Auditory Midbrain and Auditory Cortex

Physiological studies of blast-related changes in the auditory midbrain and auditory cortex (AC) reveal the deleterious effects of blast on central auditory processing. Race et al. (2013) obtained audiometric recordings of blast and noise-exposed rats and reported findings suggestive of deficits in thalamocortical transmission and cortical activation that were specific to the blast exposure. Blast-exposed rats also experienced significant temporal processing impairments in the envelope following responses (EFRs), suggestive of functional damage to the inferior colliculus (IC) (Race et al., 2017).

Blast overpressures also lead to reorganization of auditory frequency maps in the primary auditory cortex. Masri et al. (2018) showed that compared to control auditory frequency maps that were organized in a tonotopically with representation of the entire frequency spectrum (2–32 kHz), the auditory frequency maps of rats exposed to a single 22-psi blast were random and disorganized with over-representation of narrower frequency ranges. These observed changes may explain some central auditory processing pathologies including speech-in-noise (Saunders et al., 2015) and pitch discrimination difficulties (Papesh et al., 2014) commonly experienced by those afflicted by blast injury. A limitation of these studies is their lack of biochemical, histologic, or in vivo imaging data to develop a detailed understanding of the biophysical mechanisms underlying these findings.

Though exact mechanisms are not well-established and would benefit from further research, recent animal studies investigating microstructural and biochemical changes suggest that axonal degeneration in the auditory cortex (Kallakuri et al., 2018), similar to what has been seen in the CN, and alterations in the balance of excitatory and inhibitory neurotransmitters (acute reductions in GABA and NMDA receptors) (Shao et al., 2021) may contribute to impaired central auditory processing after blast-exposure. There is also evidence of bilateral degeneration in the auditory cortex following unilateral blast-exposure (Kallakuri et al., 2018) and bilateral findings of altered neurotransmitter levels despite the control ear being protected by an earplug with structures of the PAS left intact (Shao et al., 2021). These findings suggest that some component of blast-related damage to the auditory cortex may arise from mechanical forces on the brain exerted by the shockwave. However, it is difficult to ascertain the relative contribution of TBI to these bilateral degenerative changes given that there are many binaural connections in the central auditory pathway.

9.3. Role of CAS in Blast-Induced Tinnitus

Tinnitus is a common perceptual disturbance experienced after blast injury. The hypothesized mechanism of blast-induced tinnitus is similar to that of noise-induced tinnitus; most acoustic insults causing a loss of PAS sensitivity will cause tinnitus, likely through adaptive changes that occur in the CAS (Bauer, 2018). Mao et al. (2012) exposed rats to a 14-psi blast which caused early onset tinnitus at broad frequency ranges that shifted to a higher frequency range over time. These findings were associated with signs of demyelination, ischemia, and compensatory neuroplastic changes in the auditory centers detected on MR-DTI. To better understand the mechanisms underlying blast-related tinnitus, Luo et al., 2014a examined CAS structures of rats with post-blast tinnitus and showed significant increases in spontaneous activity in the dorsal cochlear nucleus (DCN), the first relay station of the CAS. This hyperactivity occurred immediately after the initial insult but was significantly reduced 3 months after exposure (Luo et al., 2014a). Subsequent studies by the same group reported hyperactivity in the inferior colliculus (IC) and auditory cortex (AC) that was persistent 3 months later (Luo et al., 2017, 2014b). Similar results have been reported in chinchillas (Bauer et al., 2008). These results suggest that acute tinnitus immediately after blast exposure may initially arise from hyperexcitability of neurons in lower-level CAS structures. Over time, compensatory neuroplastic changes cause the spontaneous firing to become less dependent on afferent input from the cochlea and more intrinsic to the neurons in higher-level CAS structures such as the IC and AC leading to chronic tinnitus (Luo et al., 2017). These studies also demonstrated hyperactivity in all frequency regions of the IC that shifted towards a higher frequency region with time, possibly because the increased higher frequency regions in the PAS are more vulnerable to deafferentation.

The therapeutic effects of antioxidant treatment, particularly a combination of N-acetylcysteine (NAC) and disodium 2,4-disulfophenyl-N-tert-butylnitrone (HPN-07), on neuronal injury in the central and peripheral auditory system have been demonstrated in multiple animal studies conducted within the past 15 years (Du et al., 2013; Ewert et al., 2012; Kopke et al., 2007). A recent study of blast-exposed rats treated with these antioxidants found normalization of ABR wave V/I amplitude ratios and reduced behavioral evidence of tinnitus, both of which correlated with histological evidence of improvement in tinnitus-related neurodegeneration and neuroplasticity. Specifically, blast-exposed rats treated with antioxidants experienced (1) significant recovery of neuronal Arc expression, a neuroplasticity protein widely used as a tinnitus biomarker when downregulated and (2) changes in levels of GABA receptors indicative of increased inhibitory signaling in CAS structures shown to be hyperexcitable in prior animal models of tinnitus (Lu et al., 2021). Taken together, it appears as though tinnitus following blast injury involves initial short-term changes in the CAS, followed by long-term, progressive reorganization. Drug therapies designed to reduce the initial damage are likely to be the most effective. However, a lot of work remains to elucidate ways to promote normalization of CAS function long after the blast has occurred in patients with chronic tinnitus. Future studies may benefit from investigating the effects of antioxidants on spontaneous hyperactivity in the CAS associated with tinnitus.

8.3. Global Brain Changes Affecting Auditory Function

The auditory system is interconnected with a vast array of neural networks within the brain that extend beyond just auditory function, which may explain why hearing loss is a known risk for cognitive decline and dementia (Deal et al., 2017; F. R. Lin et al., 2011a, 2011b; Liu and Lee, 2019). On manganese-enhanced MRI (MEMRI), blast-exposed rats with electrophysiological and behavioral evidence of tinnitus were found to have limbic hyperactivity localized to the amygdala and nucleus accumbens (Ouyang et al., 2017). The same was not true of blast-exposed rats with negative measures of tinnitus, suggesting a potential role of tinnitus in limbic hyperactivity after blast and highlighting blast-related interactions between the auditory and limbic system (Ouyang et al., 2017). These interactions may underlie known associations between tinnitus and anxiety (Hesser and Andersson, 2009). Additional studies utilizing MEMRI or functional MRI, an imaging tool that similarly measures neural activity, may help further delineate global brain changes that contribute to blast-induced auditory dysfunction.

Blast-induced hearing loss has also been associated with reduced hippocampal neurogenesis in rats weeks after the insult and deficits in long-term spatial memory (Manohar et al., 2020). This is likely secondary to neurodegenerative and inflammatory changes commonly seen in the brain parenchyma after blast. Shock waves from a single blast can impair the membranous integrity of neural and glial cells, causing intracellular contents to leak into the cerebrospinal fluid. Additionally, activated microglia cells may release pro-inflammatory cytokines, such as TNF-α, which result in significant neuroinflammation (Masri et al., 2018). These pathophysiological changes likely contribute to the widespread degenerative changes observed in many regions of the brain including CAS structures, characterized by axonal swelling, vacuoles, and GFAP reactive astrocytosis (Arun et al., 2021; Du et al., 2013; Kallakuri et al., 2018). As previously discussed, treatment with antioxidants has been shown to reduce neuronal and axonal injury markers in the brain and CAS (Du et al., 2013), highlighting their potential as therapeutic agents that can simultaneously treat blast-injury of the brain and auditory system.

10. Conclusion

In conclusion, blast exposure is an extreme form of noise exposure, causing pathophysiologic changes within the auditory system including hair cell loss, cochlear synaptopathy, auditory nerve degeneration, inflammatory changes, and compensatory changes in the CAS secondary to decreased peripheral input (Le et al., 2017). The only clear differences between blast and noise exposure are: (1) TM perforation and middle ear injury can occur with blast but not noise and (2) blast also can cause traumatic brain injury that may further exacerbate auditory symptoms although the mechanisms by which this occurs are not well delineated. The benefit of studying hearing loss using a blast is that the exposure is instant and so changes within the cochlea can be investigated immediately, as compared to a noise exposure which is often administered continuously (e.g., 2-hour exposure). However, the variety of ways in which blast overpressures can be administered to the auditory system make it difficult to interpret differences in findings across a wide range of studies. Nevertheless, this field of research would benefit from further studies investigating mechanisms of cochlear synaptopathy, the specific contributions of traumatic brain injury and central auditory system pathologies to blast-related hearing disturbances, and therapeutic strategies at all levels of the auditory system.

Highlights.

• Experimental models of blast vs noise demonstrate similar patterns of cochlear damage

• Variations in experimental methodology in blast studies make data synthesis challenging

• Endolymphatic hydrops is an indicator of subsequent cochlear synaptic loss

• Central auditory system injury is mediated by both cochlear damage and brain trauma

• More mechanistic studies are needed to develop effective treatments