Comprehensive Characterization of Cerebrovascular Dysfunction in Blast Traumatic Brain Injury Using Photoacoustic Microscopy

Abstract

Blast traumatic brain injury (bTBI) is a leading contributor to combat-related injuries and death. Although substantial emphasis has been placed on blast-induced neuronal and axonal injuries, co-existing dysfunctions in the cerebral vasculature, particularly the microvasculature, remain poorly understood. Here, we studied blast-induced cerebrovascular dysfunctions in a rat model of bTBI (blast overpressure: 187.8 ± 18.3 kPa). Using photoacoustic microscopy (PAM), we quantified changes in cerebral hemodynamics and metabolism—including blood perfusion, oxygenation, flow, oxygen extraction fraction, and the metabolic rate of oxygen—4 h post-injury. Moreover, we assessed the effect of blast exposure on cerebrovascular reactivity (CVR) to vasodilatory stimulation. With vessel segmentation, we extracted these changes at the single-vessel level, revealing their dependence on vessel type (i.e., artery vs. vein) and diameter. We found that bTBI at this pressure level did not induce pronounced baseline changes in cerebrovascular diameter, blood perfusion, oxygenation, flow, oxygen extraction, and metabolism, except for a slight sO₂ increase in small veins (<45 μm) and blood flow increase in large veins (≥45 μm). In contrast, this blast exposure almost abolished CVR, including arterial dilation, flow upregulation, and venous sO₂ increase. This study is the most comprehensive assessment of cerebrovascular structure and physiology in response to blast exposure to date. The observed impairment in CVR can potentially cause cognitive decline due to the mismatch between cognitive metabolic demands and vessel's ability to dynamically respond to meet the demands. Also, the impaired CVR can lead to increased vulnerability of the brain to metabolic insults, including hypoxia and ischemia.

Introduction

As a major cause of disability and death, traumatic brain injury (TBI) poses a massive health and financial burden in the United States, affecting over 2 million people and costing more than $76 billion in medical expenses and lost productivity every year.^1,2 TBI due to explosive blast exposure is a leading cause of morbidity and mortality in military service members and is a key contributor to combat-related mental health illnesses.³ Although significant progress has been made in recent decades in understanding the TBI associated with conventional injury mechanisms, such as impact and acceleration-deceleration shearing forces, the precise pathophysiology and injury mechanisms associated with primary blast overpressure exposure (blast TBI or bTBI) are less well understood.

Current understanding of bTBI suggests that the brain may be injured either directly by blast waves impacting the cranium or indirectly by transvascular mechanisms.⁴ However, the cellular and molecular underpinnings of each path are still missing. Limited pre-clinical translational studies have demonstrated injuries to neurons, oligodendrocytes, astrocytes, and cytoskeletal proteolysis in the cortex and hippocampus, as well as alterations in the white matter.⁵ Also, very limited studies have been performed to explore macrovascular and microvascular responses to blast exposures, suggesting that the co-existing cerebrovascular dysfunctions may also play a key role in the injury cascade.^6–14 However, current studies of the blast-induced vascular dysfunctions largely focus on structure-related issues, including breakdown of the blood–brain barrier (BBB), hemorrhage, and reductions of tight-junction proteins in the cerebral vasculature.¹⁵ For instance, transient increase in BBB permeability was observed following blast exposure by measuring the extravasation of Evans blue or immunoglobulin-G.^16,17 Also, blast-induced subdural hemorrhage was reported recently, showing structural damages to the cerebral vasculature in experimental bTBI.¹⁸ Although promising, bTBI, especially that induced by low-level blast waves, may not always present directly observable structural changes.¹⁵ Thus, functional and physiological investigation of cerebrovascular responses to blast exposure is imperative. However, it has been limited by the lack of imaging tools.

Combining optics and ultrasound in a unique way, photoacoustic imaging has emerged as an enabling technology in both basic and translational brain research.¹⁹ In this modality, pulsed laser excitation is absorbed by endogenous or exogenous chromophores in the biological tissue, which induces transient heating, thermoelastic expansion, and ultimately ultrasonic emission. Enabling high-resolution high-sensitivity detection of the optical absorption contrast across the optical and acoustic dimensions, photoacoustic imaging is highly complementary to conventional optical and ultrasound imaging modalities. As a major embodiment of photoacoustic imaging, photoacoustic microscopy (PAM) is uniquely capable of imaging optically absorbing biomolecules with cellular-level resolution.²⁰ Capitalizing on the optical absorption of blood hemoglobin, PAM is capable of label-free microvascular imaging in the rodent brain.²¹ Further, our newly developed multi-parametric PAM enables comprehensive and quantitative characterization of cerebral vasculature and hemodynamics at the microscopic level, including the concentration and oxygen saturation of hemoglobin (C_Hb and sO₂, respectively), cerebral blood flow (CBF), oxygen extraction fraction (OEF), the cerebral metabolic rate of oxygen (CMRO₂), and cerebrovascular reactivity (CVR),^22–24 presenting an ideal tool to study bTBI-induced cerebrovascular dysfunctions.

Using our pioneering multi-parametric PAM, we investigated the anatomical, functional, and metabolic dysfunctions of the cerebral vasculature in an established rat model of bTBI. With our self-developed vessel segmentation algorithms,²⁵ we also characterized the differential responses of different vessel types to the blast exposure.

Methods

Animals

In this study, 6- to 8-week-old male Sprague Dawley rats (Charles River Laboratory) were randomly assigned to two different groups (control vs. bTBI). Animals were housed under a cycle of 12-h light and 12-h dark with access to food and water. All animal procedures and experiments were approved by the Institutional Animal Care and Use Committee at the University of Virginia.

Rat model of bTBI

Four h prior to the PAM experiments, animals were anesthetized with pentobarbital and placed into a high-pressure shock tube located in the Aerospace Research Laboratory at the University of Virginia (Supplementary Fig. 1A; see online supplementary material at http://www.liebertpub.com). The detailed layout of the shock tube is shown in Supplementary Figure 1B. The tube has a square section of 20 × 20 cm². The total length of the test section is 4.3 m. At the end of the tube, the shockwave enters a catch chamber 6.7 m long, 2.5 m in diameter and disintegrates. The animal was placed ∼5 m away from the diaphragm, with the head facing the flow. The diaphragm was made of five polyester discs with a thickness of 0.014 in, which were glued together to break at the desired pressure. To achieve a blast-type pressure profile, only the diaphragm section (10 cm long) was rapidly pressurized by helium gas up to the point when the diaphragm broke (∼400 psi). This shock tube is capable of generating blast waves with a static pressure higher than 30 psi (∼208 kPa). In the experiments presented herein, the pressure profile had a sharp peak, which then dropped ∼50% in 2 μsec and ∼90% in 5 μsec (Supplementary Fig. 1C).

The time history of the pressure in the test section is typical of the positive phase of a Friedlander wave, which is associated with an explosion in a free field. However, the measured profile did not capture the negative phase of the Friedlander wave. During the blast overpressure procedure, the animal was closely monitored by a high-speed camera (10,000 fps) to confirm that it was maintained in the right position (Supplementary Video; see online supplementary material at http://www.liebertpub.com). The total time for animal handling and blast exposure was less than 5 min. Following the experimental overpressure exposure, the animal was placed on a homeothermic blanket with rectal probe to maintain the body temperature and was supplied with pure oxygen, until recovering from general anesthesia. Animals within the control group were handled in the same way as those in the blast-exposed group, with the exception of blast overpressure exposure.

Animal preparation for PAM imaging

Before PAM imaging, a craniotomy was carefully performed following established protocols.^26,27 The animal was anesthetized with isoflurane (3% for induction and 1.5–2% for surgery). Toe pinch was performed to confirm that the animal was fully sedated before surgery. The scalp was shaved using a trimmer and then depilated using hair removal cream (Surgi-cream). After sterilization of the scalp using 70% ethanol and povidone-iodine, a surgical incision was made and the periosteum was removed to expose the targeted craniotomy site. An open-skull window (∼4 mm in diameter) was created using a dental drill. To avoid overheating, drilling was paused every 30 sec and saline flush was administered. Once the skull became thin and deformable, it was removed using a fine forceps to expose the underlying tissue, followed by application of a pre-soaked soft-gel foam to prevent bleeding. Next, ultrasound gel was applied to the cranial window, after which the animal was transferred to the PAM apparatus. After being placed on the imaging stage, the exposed brain was gently coupled with a transparent membrane at the bottom of a water tank for subsequent delivery of ultrasound impulses. During imaging, the animal was maintained under anesthesia with 1.5% isoflurane and the body temperature was kept at 37°C using a heating pad.

Multi-parametric PAM system

As shown in Supplementary Figure 2 (see online supplementary material at http://www.liebertpub.com), two nanosecond-pulsed lasers (BX40-2-G and BX40-2-GR, Edgewave) were used in the multi-parametric PAM system for spectroscopic measurement of sO₂, along with C_Hb and CBF. Two beams with orthogonal polarization states were combined using a polarizing beam splitter (48–545, Edmund Optics). Then, the combined beam was attenuated by a neutral density filter (NDC-50C-2M, Thorlabs), reshaped by an iris diaphragm (SM1D12D, Thorlabs), focused by a condenser lens (LA1608, Thorlabs), and filtered by a pinhole (P50C, Thorlabs) before being coupled into a single-mode optical fiber (P1-460B-FC-2, Thorlabs) via a microscope objective (M-10X, Newport). For accurate quantitative measurement, a beam sampler (BSF10-A, Thorlabs) and a high-speed photodiode (FDS100, Thorlabs) were used to monitor and compensate for the laser fluctuation. The fiber output was collimated and then refocused by a pair of identical achromatic doublets (AC127-025-A, Thorlabs) into the rat brain through the central opening of a customized ring-shaped ultrasonic transducer (central frequency: 35 MHz; 6-dB bandwidth: 70%; f-number: 1.5). An iris diaphragm (SM05D5, Thorlabs) was used to control the beam width. To compensate for the optical aberration at the water–air interface, a correction lens (LA1207-A, Thorlabs) was used. To achieve maximum imaging sensitivity, the optical and acoustic foci were coaxially and confocally aligned.

Principles of PAM measurements

Using two optical wavelengths (532 nm and 558 nm in this study), the oxy- and deoxy-hemoglobin (HbO₂ and HbR, respectively) can be differentiated by PAM, from which the sO₂ can be derived in absolute values as²⁸:

where and are the relative concentrations of HbO₂ and HbR, respectively. With the aid of vessel segmentation,²⁵ the sO₂ values of individual feeding arteries and draining veins (s_aO₂ and s_vO₂, respectively) within the region of interest (ROI) can be extracted. With this, the regional OEF can be calculated as:

where and are the average sO₂ of all feeding arteries and that of all draining veins, respectively.

Meanwhile, the absolute C_Hb can be quantified by analyzing statistical fluctuation in the PAM-acquired A-line signals, which is induced by the Brownian motion of red blood cells (RBCs). The more RBCs, the higher the fluctuation. Given that each RBC contains ∼15 pg of hemoglobin,²⁹ the total amount of hemoglobin within the detection volume of PAM can be calculated. Based on the lateral resolution (2.7 μm) and the penetration of 532-nm light in rodent blood (46 μm),³⁰ the detection volume of our PAM is estimated to be 263 μm³. Therefore, the average RBC count within the detection volume can be derived as^22,31:

where and respectively denote the mean and variance operation, is the amplitude of the PAM signal, and is the electronic thermal noise of our PAM system.

Further, the blood flow speed can be simultaneously quantified by correlation analysis of the same set of A-lines.^32,33 Linearly proportional to the flow speed, the decorrelation rate of the sequentially acquired A-lines can be used to generate flow maps with μm-level spatial resolution. Note that the correlation window consists of 46 A-lines with a constant interval of 0.1 μm, resulting in a window size of 4.6 μm. Because the size of the window is comparable to the average diameter of capillaries, our PAM is able to measure the blood flow in microvessels. By combining the flow speed and the vessel diameter extracted using vessel segmentation, the volumetric blood flow in individual vessels can be calculated as:

where D is the vessel diameter and is the average blood flow speed within the vessel.

By combining the aforementioned hemodynamic parameters, the total CMRO₂ in the ROI can be derived using Fick's law:

where is the oxygen binding capacity of hemoglobin (1.36 mL of oxygen per gram of hemoglobin), is the total volumetric blood flow through the region, and W is the tissue weight estimated by assuming an average cortical thickness of 1.2 mm and a tissue density of 1.05 g/mL.

Cerebrovascular reactivity measurement

To study the CVR, 200 mg/kg acetazolamide (ACZ) was injected through the tail vein of the animal after acquisition of the baseline images. The injection was performed onsite without animal movement. Thus, no image registration was required. Five min post-injection, the same ROI was re-imaged. Side-by-side comparison of the pre- and post-injection images at the single-vessel level, with the aid of vessel segmentation, allowed comprehensive quantification of the CVR to the vasodilatory stimulation, in terms of the vessel diameter, sO₂, blood flow speed, and volumetric blood flow.

Statistical analysis

The unpaired t test was used for statistical comparison of the structural, functional, and associated oxygen-metabolic parameters between the control and bTBI groups in all figures. The paired t test was used to compare these parameters before and after ACZ injection within each group. All figures and data are shown in the format of mean ± standard deviation. Differences with p-values <0.05 were considered statistically significant.

Results

Multi-parametric PAM of blast-induced changes in the cerebral vasculature

Twelve bTBI rats and seven control rats were imaged using PAM. Experimental animals were exposed to 27.2 ± 2.7 psi (187.8 ± 18.3 kPa) static pressure in the high pressure shock tube described above. The C_Hb, sO₂, and blood flow were directly measured using multi-parametric PAM in blast-exposed and control rats before and after ACZ injection (Fig. 1). Marked increases in the venous sO₂ and blood flow speed were observed in the brain of the control rats (blue and red arrows in Fig. 1A, respectively) following ACZ administration. In contrast to the control rats, no significant change was observed in the bTBI cohort (Fig. 1B). Vessel segmentation was used to extract quantitative values related to C_Hb, sO₂, speed of vascular flow, and diameter of individual vessels, from which OEF, volumetric blood flow, and CMRO₂ were derived. To study vessel type-specific responses, the imaged vessels were divided into four groups based on the following values³⁴: large arteries (≥45 μm), small arteries (<45 μm), large veins (≥45 μm), and small veins (<45 μm). As shown in Figure 1C, no significant differences were noted between the two animal groups in baseline diameters of these four types of vessels.

FIG. 1.

Multi-parametric PAM and segmentation-based single-vessel analysis of blast-induced changes in the cerebral vasculature. PAM images of the total hemoglobin concentration, blood oxygenation, and blood flow speed, as well as corresponding vessel segment maps, acquired before and 15 min after injection of ACZ in (A) a control rat and (B) a rat subject to blast exposure (pressure: 196.5 kPa). (C) Statistical comparison of vessel diameters between the control and blast-exposed groups. Sample size: 35 large arteries, 55 small arteries, 62 large veins, and 57 small veins in the control group and 89 large arteries, 94 small arteries, 81 large veins, and 57 small veins in the blast-exposed group. Arrows in (A): representative vessels that show changes in blood oxygenation and flow speed in response to the vasodilatory stimulation. ACZ, acetazolamide; bTBI, blast traumatic brain injury; PAM, photoacoustic microscopy. Color image is available online.

Cerebral hemodynamic and oxygen-metabolic responses to blast exposure

To determine whether blast overpressure exposure caused functional changes in cerebral vessels in this experimental bTBI model, all hemodynamic and oxygen-metabolic parameters accessible to PAM were quantified and compared with the baseline without administration of vasoactive agents. As expected, the arterial and venous C_Hb showed no difference between the two groups (Fig. 2A). However, the average venous sO₂ in the blast-exposed rats (79.0 ± 2.5%) was higher (p-value: 0.039) than that in the control rats (76.0 ± 3.1%), whereas the average arterial sO₂ values of the two groups were similar (90.8 ± 1.2% vs. 91.8 ± 1.1% for the control and blast-exposed groups, respectively; p-value: 0.14; Fig. 2B). No significant differences were observed between groups with respect to the average venous blood flow speed, OEF, CBF, or CMRO₂ (Fig. 2C–F).

FIG. 2.

Cerebral hemodynamic and oxygen-metabolic responses to blast exposure. Statistical comparison of (A) total hemoglobin concentration, (B) arterial and venous blood oxygenation, (C) arterial and venous blood flow speed, (D) regional oxygen extraction fraction, (E) regional cerebral blood flow, and (F) regional cerebral metabolic rate of oxygen between the control and blast-exposed groups. Sample size: seven control rats and 12 bTBI rats. *p < 0.05. bTBI, blast traumatic brain injury.

To examine whether the blast-induced functional changes differed between vessel types, we repeated this analysis for each of the four different types of vessels—including 35 large arteries, 55 small arteries, 62 large veins, and 57 small veins in the control group and 89 large arteries, 94 small arteries, 81 large veins, and 57 small veins in the blast-exposed group. Single-vessel analyses demonstrated that blast-induced changes in the venous sO₂ were seen primarily in small veins (Fig. 3A). Also, significant increases were observed in the blood flow speed in large veins of the blast-exposed animals (Fig. 3B), but corresponding volumetric flow values were comparable between the two groups (Fig. 3C).

FIG. 3.

Vessel type-specific cerebrovascular responses to blast exposure. Statistical comparison of (A) blood oxygenation, (B) blood flow speed, and (C) volumetric blood flow between the control and blast-exposed groups. Sample size: 35 large arteries, 55 small arteries, 62 large veins, and 57 small veins in the control group (seven control animals) and 89 large arteries, 94 small arteries, 81 large veins, and 57 small veins in the blast-exposed group (12 blast-exposed animals). *p < 0.05. bTBI, blast traumatic brain injury.

Influence of blast exposure on cerebrovascular reactivity to vasodilatory stimulation

In addition to studies exploring baseline differences in the vascular structure and hemodynamics between the blast-exposed and control animals, the current effort also incorporated a vasodilatory challenge, using intravenously administered ACZ, to examine whether blast overpressure exposure caused changes in CVR. Grossly observable differences were seen between the blast-exposed and control groups following ACZ administration, as shown in Figure 4A and 4B. In quantifying these observed differences, ACZ induced a 7.8% increase in the venous sO₂ in the control animals, which was significantly different from the baseline control values (p-value: 0.008). No significant elevation in the venous sO₂ was seen in blast-exposed rats (Fig. 4B). Additionally, the arterial flow was significantly elevated compared with the baseline in both groups. However, the degree of arterial flow elevation was significantly higher in the control animals than in the blast-exposed ones. The venous blood flow was significantly elevated in the control animals but not in the blast-exposed ones (Fig. 4C). Reduced OEF was observed in both the control (74.5% of the baseline) and the blast-exposed (87.2% of the baseline) animals. However, these changes were not significantly different (p-value: 0.163; Fig. 4D). Accompanying this reduced OEF was increased regional CBF (Fig. 4E). The control animals showed significantly higher (p-value: 0.018) increases in regional CBF (126.2% of the baseline) than the blast-exposed ones (106.2% of the baseline). The coupling of OEF and CBF resulted in compensated regional CMRO₂ following vasodilatory stimulation in both groups (Fig. 4F), consistent with previous observations.^35–37

FIG. 4.

Cerebral hemodynamic and oxygen-metabolic responses to vasodilatory stimulation. Statistical comparison of ACZ-induced changes in (A) total hemoglobin concentration, (B) arterial and venous blood oxygenation, (C) arterial and venous blood flow speed, (D) regional oxygen extraction fraction, (E) regional cerebral blood flow, and (F) regional cerebral metabolic rate of oxygen against their corresponding baseline values (significance, if any, is indicated on the top of the columns) and between the control and blast-exposed groups (significance, if any, is indicated between the two compared columns). Sample size: seven control rats and 12 bTBI rats. *p < 0.05. ACZ, acetazolamide; bTBI, blast traumatic brain injury.

To examine whether the bTBI-induced CVR changes varied based on vessel types, we repeated the above described analyses for each of four different types of vessels. As shown in Figure 5A, ACZ induced significant sO₂ increases in both the large and small veins in the control animals (107.9% and 111.5% of the baseline, respectively), but failed to elicit similar increases in the blast-exposed animals (100.7% and 101.1% of the baseline for the large and small veins, respectively). The ACZ-induced arterial dilation was abolished in the blast-exposed animals, but not in the controls (106.6% and 114.0% of the baseline for the large and small arteries, respectively; Fig. 5B). Along with the larger increases in venous sO₂ and arterial diameter, more significant elevations in the blood flow were observed following ACZ administration in all types of vessels in the control animals (117.2%, 131.8%, 120.4%, and 133.7% of the baseline for the large arteries, small arteries, large veins, and small veins, respectively). In contrast, much impaired responses in the blood flow were seen in the blast-exposed animals (109.4%, 102.5%, 103.8%, and 107.1% of the baseline for the large arteries, small arteries, large veins, and small veins, respectively; Fig. 5C). Not surprisingly, the impaired vasodilation and flow upregulation in the blast-exposed animals led to diminished responses in the volumetric blood flow (110.0%, 109.1%, 107.1%, and 114.5% of the baseline for the large arteries, small arteries, large veins, and small veins, respectively; Fig. 5D), compared with those in the control animals (134.0%, 180.5%, 126.2%, and 146.2% of the baseline for the large arteries, small arteries, large veins, and small veins, respectively).

FIG. 5.

Vessel type-specific cerebrovascular reactivity to vasodilatory stimulation. Statistical comparison of ACZ-induced changes in (A) blood oxygenation, (B) vessel diameter, (C) blood flow speed, and (D) volumetric blood flow against their corresponding baseline values (significance, if any, is indicated on the top of the columns) and between the control and blast-exposed groups (significance, if any, is indicated between the two compared columns). Sample size: 35 large arteries, 55 small arteries, 62 large veins, and 57 small veins in the control group (seven control animals) and 89 large arteries, 94 small arteries, 81 large veins, and 57 small veins in the blast-exposed group (12 blast-exposed animals). *p < 0.05. ACZ, acetazolamide; bTBI, blast traumatic brain injury.

Discussion

In recent years, our understanding of the influence of bTBI on the brain has undergone substantial expansion. Evidence emerging from pre-clinical and clinical studies suggests that cerebrovascular dysfunctions may play an important role in the pathophysiology of bTBI and warrants in-depth investigation. Advances in clinical neuroimaging technologies, including computed tomography, magnetic resonance imaging (MRI), and positron emission tomography, enable direct visualization of the cerebral vasculature in TBI patients. However, in mild or moderate bTBI, subtle changes in the cerebral microvasculature³⁸ often escape detection of conventional neuroimaging technologies. Two-photon fluorescence microscopy, the technology of choice for intravital imaging of the rodent brain, has enabled cerebrovascular imaging at the microscopic level. However, most applications of this technology require administration of exogenous angiographic agents,³⁹ which might alter vascular physiology, thereby introducing a confound into these experimental studies. Even with the aid of exogenous agents, comprehensive evaluation of cerebrovascular function and associated oxygen metabolism remains largely inaccessible in in vivo models. The lack of enabling tools for label-free, quantitative, and comprehensive imaging of the cerebral vasculature at high resolution has fundamentally limited our understanding of the vascular component of bTBI and other brain disorders.^38,40–43

Multi-parametric PAM has filled this gap and shed new light on blast-induced cerebrovascular dysfunction. Capitalizing on the optical absorption of blood hemoglobin, our PAM simultaneously imaged cerebrovascular anatomy, C_Hb, sO₂, and blood flow speed in both blast-exposed and control animals in a label-free manner (Fig. 1), with an imaging depth of ∼500 μm from the brain surface.⁴⁴ With the aid of vessel segmentation, these structural and hemodynamic parameters were extracted at the single-vessel level, from which volumetric blood flow, regional CBF, OEF, and CMRO₂ were also quantified. The ability of multi-parametric PAM to comprehensively characterize the cerebral vasculature has enabled us to extend CVR evaluation from the blood flow to other functional and metabolic parameters. Moreover, the high-resolution structural and physiological measurements of PAM allow for insight into vascular response based on cerebral vessel subtypes.

Using this enabling technology, we showed that the blast exposure did not induce significant changes in the cerebral microvascular structure, hemodynamics or associated oxygen delivery and metabolism, except for slight increases in venous sO₂ and blood flow (Figs. 2 and 3). Interestingly, the reduced OEF (due to the increased venous sO₂) and increased CBF remained tightly coupled in the blast-exposed rats, resulting in maintenance of CMRO₂. Of note, these observations were made in animals at rest. It remains to be determined whether CMRO₂ would have been maintained under stress conditions, where cerebral metabolic demands might be higher and may challenge the impaired vascular responses in the blast-exposed brain. Despite the lack of overt changes in the cerebrovascular structure, hemodynamics and oxygen metabolism, the blast exposure significantly impaired CVR, including eliminated arterial dilation and compromised autoregulation of CBF and sO₂ (Figs. 4 and 5).

Our observations reinforce and expand current knowledge of cerebrovascular dysfunctions in bTBI. Previous MRI studies have demonstrated no obvious changes following blast exposure to the rat brain,⁴⁵ echoing the lack of baseline differences between the control and blast-exposed rats observed by PAM. Although inducing no detectable changes in the vascular structure, low-level blast exposures have been shown to impair vascular autoregulation in experimental bTBI models.⁴⁶ Indeed, a recent study by Rodriguez and colleagues has shown, in an isolated segment of the middle cerebral artery (MCA), that blast exposure results in impaired CVR.¹⁴ Here, we extend the understanding of blast-induced changes in CVR from a single macroscale MCA segment ex vivo to a network of microvessels of different types (i.e., artery vs. vein) and diameters in vivo, and from diameter only to the microvascular structure, perfusion, oxygenation, and tissue metabolism. Our results not only quantitatively demonstrate the influence of blast overpressure exposure on the multi-faceted CVR (including blood perfusion, oxygenation, flow, tissue oxygen extraction, and oxygen metabolism) to vasodilatory stimulation (Fig. 4), but they also show the dependence of the CVR impairment on the vessel type and size (Fig. 5), thereby providing exciting new insights into blast-induced anatomical and functional changes in the cerebral microvasculature and suggesting a state of cerebrovascular hyporeactivity within the first few hours following blast exposure.

The improved understanding is made possible by the unique capability of multi-parametric PAM for quantitative, high-resolution, comprehensive imaging of the microvasculature, in contrast to the semi-qualitative, macroscopic-resolution, and single-contrast (e.g., perfusion only) nature of laser Doppler flowmetry and the ex vivo only arteriography technique used in the previous study.¹⁴ The impairment of cerebrovascular autoregulation following blast exposure may contribute to secondary injuries by compromising the ability of cerebral vasculature to deliver oxygen to the brain, resulting in increased vulnerability of the brain to metabolic insults and injury cascades associated with this condition.^47,48

Besides impaired cerebrovascular autoregulation, other forms of cerebrovascular dysfunction have been associated with bTBI—including vasospasm and BBB breakdown¹⁵—which warrants future investigation using PAM. BBB disruption, which may occur even in mild TBI without overt bleeding, contributes to vasogenic edema—an important factor underlying the clinical outcome of TBI.⁴⁹ The impairment of the BBB and consequent vasogenic edema are essential vascular factors in the secondary injury process of bTBI.⁵⁰ Exploiting the strong optical absorption of Evans blue, a commonly used dye for vascular permeability testing,⁵¹ our PAM technology holds great potential to extend the classic Evans blue assay to in vivo. Moreover, label-free imaging of brain edema has been demonstrated by photoacoustic computed tomography.⁵²

Future development of the multi-parametric PAM for high-resolution imaging of BBB permeability and vasogenic edema will add to our current understanding of bTBI-induced cerebrovascular dysfunctions. Cerebral vasospasm is another prominent secondary insult in bTBI and is often associated with severe brain damage.⁵³ According to clinical observations, blast exposure-induced cerebral vasospasm typically lasts 14–30 days.¹⁵ Thus, visualization of this chronic dysfunction requires longitudinal monitoring. To this end, a long-lifetime cranial window technique needs to be developed for PAM. To date, the two main types of chronic window for intravital brain imaging are open-skull window and reinforced thinned-skull window.^27,54 Although providing excellent visibility, the former is invasive and may activate microglia and astrocytes, thereby affecting brain hemodynamics and activity. In contrast, the thinned-skull window provides chronic access to the brain with reduced inflammatory response and acceptable visibility, but requires labor-intensive maintenance. Adoption and refinement of these window techniques will not only grant PAM the access to chronic changes in cerebrovascular parameters, including vasospasm, but also enable it to longitudinally monitor how blast-induced cerebrovascular dysfunctions progress over time.

It is worth pointing out that a new study by Esenaliev and associates demonstrates the therapeutic effect of high-energy (5 mJ; optical fluence, 300 J/cm²), near-infrared (808 nm), nanosecond laser pulses for blast-induced neurotrauma.⁵⁵ The underlying mechanisms may be related to the photothermal effect and/or low-intensity ultrasonic neuromodulation.⁵⁵ Although nanosecond-pulsed lasers were also used in the present study, our lasers were of much lower pulse energy (∼100 nJ; fluence, <0.02 J/cm²). At this fluence level, the estimated temperature rise and generated ultrasonic pressure²⁰ are insufficient for photothermal/ultrasonic therapy. Moreover, our lasers output visible light (532 and 558 nm), different from the near-infrared light commonly used for laser therapy. Thus, it is unlikely that the nanosecond lasers used in the present study have therapeutic effects that may confound the results.

A caveat of the present study is that animals of a single sex (i.e., male) were studied. It has been shown that differences in sex contribute to the heterogeneity in the outcomes of mild TBI.⁵⁶ Thus, future studies incorporating both sexes are warranted to provide a more comprehensive understanding of the injury mechanisms and to better inform clinical practice.

Conclusion

Using the state-of-the-art multi-parametric PAM, this study represents the most comprehensive assessment of the cerebrovascular structure and physiology following experimental blast exposure in rats. Our results show that the blast exposure can significantly impair vasodilation, blood flow autoregulation, as well as blood oxygenation and extraction. These findings suggest the potential for a primary role of the alteration in CVR in the pathophysiology of bTBI and may facilitate the identification of new and promising therapeutic targets for this disease.

Author Disclosure Statement

No competing financial interests exist.