- Email: [email protected]
Monitoring Hemodynamic and Metabolic Alterations during Severe Hemorrhagic Shock in Rat Brains
Monitoring Hemodynamic and Metabolic Alterations during Severe Hemorrhagic Shock in Rat Brains Nannan Sun, BS, Weihua Luo, PhD, Lin Z. Li, PhD, Qingming Luo, PhD Rationale and Objectives: Our long-term goals are to identify imaging biomarkers for hemorrhagic shock and to understand how the preservation of cerebral microcirculation works. We also seek to understand how the damage occurs to the cerebral hemodynamics and the mitochondrial metabolism during severe hemorrhagic shock. Materials and Methods: We used a multimodal cerebral cortex optical imaging system to obtain 4-hour observations of cerebral hemodynamic and metabolic alterations in exposed rat cortexes during severe hemorrhagic shock. We monitored the mean arterial pressure, heart rate, cerebral blood flow (CBF), functional vascular density (FVD), vascular perfusion and diameter, blood oxygenation, and mitochondrial reduced nicotinamide adenine dinucleotide (NADH) signals. Results: During the rapid bleeding and compensatory stage, cerebral parenchymal circulation was protected by inhibiting the perfusion of dural vessels. During the compensatory stage, although the brain parenchymal CBF and FVD decreased rapidly, the NADH signal did not show a significant increase. During the decompensatory stage, FVD and CBF maintained the same low level and the NADH signal remained unchanged. However, the NADH signal showed a significant increase after the rapid blood infusion. FVD and CBF rebounded to the baseline after the resuscitation and then declined again. Conclusions: We present for the first time simultaneous imaging of cerebral hemodynamics and NADH signals in vivo during the process of hemorrhagic shock. This novel multimodal method demonstrated clearly that severe hemorrhagic shock imparts irreversible tissue damage that is not compensated by the autoregulatory mechanism. Hemodynamic and metabolic signatures including CBF, FVD, and NADH may be further developed to provide sensitive biomarkers for stage transitions in hemorrhagic shock. Key Words: Mitochondrial metabolism; cerebral microcirculation; decompensation; NADH; CBF. ªAUR, 2014
H
emorrhagic shock is a condition caused by a rapid and significant loss of intravascular volume, which may lead to hemodynamic instability, decreased tissue perfusion, cellular hypoxia, organ damage, and death (1,2). Usually, the process of hemorrhagic shock is described in three stages (or in four stages if defining an initial stage
Acad Radiol 2014; 21:175–184 From the Britton Chance Center for Biomedical Photonics, Wuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology, Wuhan, Hubei, PR China (N.S., W.L., Q.L.); Department of Biomedical Engineering, Key Laboratory of Biomedical Photonics of Ministry of Education, Huazhong University of Science and Technology, 1037 Luoyu Road, Wuhan, Hubei 430074, PR China (N.S., W.L., Q.L.); Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA (N.S., L.Z.L.); Department of Biochemistry and Biophysics, Britton Chance Laboratory of Redox Imaging, Johnson Research Foundation, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA (N.S., L.Z.L.); and Institute of Translational Medicine and Therapeutics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA (L.Z.L.). Received August 21, 2013; accepted November 26, 2013. This article is dedicated to the 100th birthday of Dr. Britton Chance, whose guidance and support were instrumental for the establishment of our research laboratories. Address correspondence to: Q.L. e-mail: [email protected] ªAUR, 2014 http://dx.doi.org/10.1016/j.acra.2013.11.017
before the compensatory stage) (1,3). In the first stage (i.e., the compensatory stage) the body undergoes microcirculatory ischemia. This is characterized by compensatory hypotension during which the minimum perfusion level of vital organs is maintained by autoregulatory mechanism. In the second stage (i.e., the decompensatory stage) the body exhibits microcirculatory congestion, characterized by decompensatory hypotension that indicates the failure of autoregulatory mechanism. The last stage is the refractory stage, characterized by refractory hypotension and microcirculation failure followed by irreversible cell damage and organ failure. Whether the cerebral circulation is protected by the autoregulatory mechanism during the process of the shock has long been controversial (4,5). Previous studies have highlighted that cerebral blood flow (CBF) in rats is maintained when mean arterial pressure (MAP) is kept between 60 and 140 mmHg (6). Animals would have lower MAP if they underwent a greater blood loss. Wan et al. (7) found that the microcirculation of the cerebral cortex could be maintained after losing 35% of blood volume (MAP could be reduced to 60 mmHg or less). However, they found that the buccal microcirculation supplied by the carotid artery showed 175
SUN ET AL
damages that are consistent with the changes in peripheral circulation. Furthermore, the protection of cerebral circulation resulting from autoregulatory mechanism may be temporary during the process of shock (1). The autoregulatory mechanism will show signs of failure once the blood pressure (BP) exceeds the lower threshold for an extended period (4,8). Therefore, during the process of shock, particularly during the decompensatory stage, it is important to pay close attention to the dynamic changes in the circulatory states of vital organs, such as the heart and brain (9). These factors have important clinical significance for determining an appropriate fluid resuscitation strategy and timely initiation of protective measures for heart and brain (7). Typically, there are eight indicators used for monitoring the shock under clinical settings: BP, heart rate (HR), central venous pressure, cardiac output, urine volume, blood gas analysis, hematocrit, and arterial blood lactate. However, these indicators neither can be monitored in real time nor provide local hemodynamic and metabolic information. Some studies have shown that monitoring the changes in local microcirculation, hemodynamics, and tissue metabolic states of vital organs are conducive to understanding the autoregulatory mechanisms, tissue function, and injury during hemorrhagic shock (4,10). Ultimately, these real-time monitoring may help clinicians to determine a patient’s stage of shock, grasp the safety time window, and administer appropriate treatments. To study the dynamic vascular perfusion (including dural vessels supplied by the external carotid artery and pial vessels supplied by the internal carotid artery), vascular response, and metabolic change in a rat model of severe hemorrhagic shock (11–14), we used a multimode optical imaging technology, that is, an integration of laser speckle contrast imaging (LSCI) (15–17), optical intrinsic signal imaging (OISI) (18,19), and endogenous nicotinamide adenine dinucleotide (NADH) fluorescence imaging (20). The rat model was achieved through rapid bleeding to obtain a low BP level of about 40 mmHg. The results of our study indicate that the transitional process from compensatory stage to decompensatory stage can be simulated by this rat model. During the initial and compensatory stages, cerebral parenchymal circulation was protected by inhibiting the perfusion of dural vessels. The blood supplies of dural vessels are originated from the external carotid artery, and the perfusion changes of dural vessels are consistent with the changes of systemic circulation. In addition, the cerebral hemodynamic and metabolic signal signatures may reflect underlying tissue damages during the decompensatory stage and provide potential biomarkers for the shock stages and the stage transition. MATERIALS AND METHODS
Academic Radiology, Vol 21, No 2, February 2014
Figure 1. A schematic representation of experimental procedures performed during severe hemorrhagic shock of the rat model. Blood was withdrawn from the femoral artery until the mean arterial pressure (MAP) decreased to about 40 mmHg in 15 minutes. MAP was maintained at 40 mmHg in the compensatory and decompensatory stages by continually withdrawing or injecting blood. Fluid resuscitation was performed during the rapid infusion period by returning all blood withdrawn to the rat. Periods: A, baseline; B, rapid blood bleeding; C, compensatory stage; D, decompensatory stage; E, rapid blood infusion; and F, after blood infusion.
(50 mg/kg) and urethane (600 mg/kg). Anesthesia was maintained by supplements of one-sixth of the initial dose at intervals of 30–45 minutes. The body temperatures were maintained at 37.0 0.5 C using a rectal probe and a feedback-controlled heating blanket. PE-50 cannulas were inserted into the femoral arteries on both sides and into the femoral vein on the right side. The heads of the rats were held in position using a stereotaxic apparatus and craniotomies were performed over the right parietal bone. The dura stayed intact and was continually bathed with artificial cerebrospinal fluid. Experimental Procedures
Rats were randomly divided into a shock group (n = 7) and a sham group (n = 5). The experimental procedure has been described previously (11–14) and a brief overview is provided in Figure 1. We measured MAP and HR with a pressure transducer (model RM6240, multichannel physiological parameters monitoring instrument; Chengdu Instrument Factory, Chengdu, China) in the left femoral artery. The rapid bleeding from the right femoral artery was conducted until MAP dropped to 40 mmHg; MAP was maintained at 40 mmHg by continual blood withdrawal (up to 50% of the total blood volume) and blood injection for 2 hours. We assumed that the rat entered the decompensatory stage if the BP had been maintained at 40 mmHg for 90 minutes. The fluid resuscitation was carried out half an hour after the outset of the decompensatory stage with a rapid infusion of the total amount of blood withdrawn back into the rat. The exposed rat brain tissues were continuously monitored by the multimodal optical imaging system during the entire experiment. The experimental procedure of the sham group was the same as the shock group except no blood withdrawal and injection. Imaging Instrument
Animal Preparation
The experimental protocol was approved by the Institutional Animal Care and Use Committee at Huazhong University of Science and Technology. Male Wistar rats (220–280 g, n = 12) were anesthetized by intraperitoneal injection of a-chloralose 176
The multimodal optical imaging system (20,21) integrates NADH autofluorescence imaging, blood flow LSCI, and dual-wavelength OISI (reflectance imaging). It can simultaneously monitor changes of a few physiological parameters in vivo as shown in Figure 2. Two identical 12-bit
Academic Radiology, Vol 21, No 2, February 2014
HEMODYNAMIC AND METABOLIC CHANGES DURING SHOCK
Figure 2. A technical diagram for multimodal optical imaging. DHbO, DHbR, and DHbT are relative concentration changes compared to the baselines of oxyhemoglobin (HbO), deoxyhemoglobin (HbR), and total hemoglobin (HbT), respectively. CBF, cerebral blood flow; FVD, functional vascular density; HR, heart rate; MAP, maen arterial pressure.
charge-coupled device (CCD) cameras (2 2 binning, 1024 1396 pixels, Pixelfly QE; PCO Computer Optics, Kelheim, Germany) were attached in parallel to a stereomicroscopy (Olympus SZ6045 TR Zoom, Olympus, Japan) with a field of view 2.62 mm 3.56 mm. One of the CCD cameras was used to image the NADH fluorescence with a band-pass emission filter (475 nm 21 nm, FF01-475/ 42-25; Semrock, Rochester, NY). The other camera was used to image the rest of the three images sequentially. A liquid crystal tunable filter (bandwidth 7 nm, VariSpec; Cambridge Research & Instrumentation, Woburn, MA) was placed in front of the second CCD camera. An ultraviolet light emitting diode (365 nm for NADH excitation, Thorlab, Newton, NJ), a laser diode (660 nm), and a filtered halogen lamp (550– 590 nm, Olympus LG-PS2, Olympus, Tokyo, Japan) were positioned around the rat brain cortex to provide an even illumination or fluorescence excitation. As shown in Figure 2, the NADH fluorescence emission was imaged at 475 nm, OISI at 560 nm and 570 nm, and LSCI at 660 nm (15–17,22). Data Analyses
Using the modified Beer–Lambert relationship, we obtained an equation that relates the signal intensities to the changes of hemoglobin concentrations. log I0l =Itl ¼ mlHbR Dl DHbR þ mlHbO Dl DHbO:
(1)
where l is the light wavelength, I0l and Itl are the measured intensities of reflected light at the light wavelength l for the baseline (t = 0) and time t, respectively; mlHbR and mlHbO are absorption coefficients of oxyhemoglobin (HbO) and deoxyhemoglobin (HbR), respectively; Dl is the differential pathlength factor, that is, the mean optical pathlength of the photons travelling in the tissue; DHbO, DHbR, and DHbT represent the concentration changes of oxyhemoglobin,
deoxyhemoglobin, and total hemoglobin (HbT), respectively. Dl in brain tissue was evaluated and quantified by the Monte Carlo simulation at 350 nm (365 nm), 480 nm (475 nm), 560 nm, and 570 nm using the approach of Kohl et al. and Andrew et al. (23–26). We used the simulated Dl for 560 nm and 570 nm to compute DHbO, DHbR, and DHbT (DHbT = DHbO + DHbR) (24–26) by Equation (1). The intrinsic signals (reflectance) at 350 nm and 480 nm were predicted from DHbO and DHbR using the approach of Yevgeniy et al. (27), and their reflected signals were then used to correct the NADH signals measured from the first CCD (20). CBF was obtained through online processing (customized C program) by the laser speckle temporal contrast analysis of 100 consecutive raw speckles images (15–17,22). FVD is defined as the total length of perfused vessels within a region of interest in a two-dimensional image divided by the area of that region (28). FVD, vascular diameter, and perfusion were computed from laser speckle images, and the details of the quantification methods were shown in previous works (20,29,30). The blood perfusion was obtained by multiplying the CBF and the cross-sectional area of a blood vessel. Vasculatures such as the middle meningeal artery (MMA), middle cerebral artery (MCA), and veins were identified by vessel diameter and morphology (31,32).
RESULTS Alterations of Hemodynamics
Although MAP and HR are physiological representations of the functional state of systemic circulation, MAP is the most important standard for monitoring the shock progression in our model. After rapid bleeding, HR declined rapidly and MAP was maintained at 40 mmHg until fluid resuscitation (Fig 3). After rapid fluid infusion, MAP increased sharply but HR did so slowly. One hour after infusion, the MAP 177
SUN ET AL
Figure 3. Average maen arterial pressure (MAP) and heart rate (HR) recording during the process of severe hemorrhagic shock. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid blood infusion; and F, after blood infusion. **P < .01 comparing between the shock and control group.
and HR of the shock group were not significantly different from those of the control group. Figures 4 and 5 show the changes of CBF, vascular diameter, and vascular perfusion that reflect the microcirculatory and hemorheological conditions. After the rapid bleeding (Fig 4b, Fig 5 ‘‘Period B’’), the blood flow, diameter, and perfusion of MMA decreased. Conversely, the diameter and perfusion of MCA increased, whereas the veins remained relatively stable. At the compensatory stage (Fig 4c), MMA (white arrows in Fig 4a) disappeared along with its diameter and perfusion reduced significantly to zero as shown in Figure 5 ‘‘Period C’’. Diameters of MCA and veins increased and their perfusions first increased and then decreased (Fig 5 ‘‘Period C’’). In general, the CBF image in Figure 4c is darker than that in Figure 4b, which indicates that the blood flow through parenchymal tissue decreased significantly. During the decompensatory stage, a number of vessels dimmed or disappeared in Figure 4d compared to those in Figure 4c. As shown in Figure 5 ‘‘Period D’’, the diameter of MCA initially increased before decreasing, whereas the vein expanded persistently. The perfusion of MCA decreased to the same level as that of the vein and remained steady until 15 minutes prior to rapid infusion. We also observed the interruption of blood flow during the decompensatory stage, which was presumably caused by leukocyte aggregates. The interruptions are shown by the white circles in Figure 6. Consequently, microcirculatory blood flow slowed and arterial perfusion continued to decrease at 15 minutes prior to rapid infusion in Figure 5. The absorptions of excitation and emission light were presumably enhanced by increased blood volume so that the optical images darkened (Fig 4d). The images were lighted up significantly by the blood infusion (Figs 4e and f). During the blood infusion, the blood flow, perfusions, and diameters of MCA and veins increased significantly followed by a dramatic postinfusion decrease that was even below baseline eventually (Fig 5 ‘‘Period F’’). 178
Academic Radiology, Vol 21, No 2, February 2014
The responses of vessels vary significantly dependent on vessel types and diameters (33). Statistical analyses of the mean blood flow changes of MMA, MCA, and vein are shown in Figure 7. During the rapid bleeding, there was no significant difference in the mean blood flow of MCA and veins compared to the baseline (Fig 7 ‘‘Period A’’). However, the mean blood flow of MMA decreased significantly by 20%. The mean blood flow change of MMA is the most visible one after rapid bleeding. It decreased by 70% during the compensatory stage (Fig 7 ‘‘Periods B and C’’) compared to the baseline. In comparison, the mean blood flow changes of MCA and veins decreased by 40%–50% at the compensatory stage compared to the baseline. Blood flow changes of the MMA are significantly different than those of the MCA and veins (P < .01, in Fig 7 ‘‘Period C’’) during the compensatory stage. The differences between MMA and MCA persisted in the decompensatory stage. The blood flows in these three types of vessels largely recovered after fluid resuscitation (Fig 7 ‘‘Period E’’) with MMA flow reaching 65%, MCA 80%, and vein 70% of the baseline level. However, their flows decreased again after rapid infusion. The difference of blood flows between MMA and MCA maintained statistical significance (P < .05) 1 hour after infusion. The changes in hemoglobin concentration were statistically analyzed and shown in Figure 8. In the compensatory stage, the mean HbR increased by 10% and mean HbO decreased by 5% compared to the baseline (P < .05). That indicated tissue hypoxia and/or tissue ischemia with excessive consumption of HbO. In other stages of shock, no significant variations in HbO and HbR were observed. The mean HbT remained stable during the entire shock process despite the blood bleeding. Alteration of Metabolism
The temporal changes of the average values of brain parenchymal CBF, FVD, and NADH fluorescence signals are shown in Figure 9 along with statistical analysis results. All three parameters exhibited no significant changes compared to the sham group and the baseline during the rapid bleeding. During both the compensatory and decompensatory stages, the NADH signal increased relative to the sham group but with no significant differences. The NADH signal was 20% greater than that of the sham group after the fluid resuscitation (P < .05, in Figure 9 ‘‘Period E’’). One hour after infusion, the NADH signal rose further about 40% above the baseline value (P < .05). NADH fluorescence kept increasing after rapid bleeding compared to the initial stage. This indicates that the metabolic state of mitochondria was changed and mitochondria probably became dysfunctional. The brain parenchymal microcirculation (CBF) and FVD decreased significantly, that is, 55% (P < .01) and 20% (P < .05), respectively, during both the compensatory and decompensatory stages, as shown in Figure 9 ‘‘Period C’’. The signals of FVD and CBF were increased by fluid resuscitation and almost totally recovered to the levels of
Academic Radiology, Vol 21, No 2, February 2014
HEMODYNAMIC AND METABOLIC CHANGES DURING SHOCK
Figure 4. Representative images of cerebral blood flow (CBF) at different time points during the hemorrhagic shock. The amount of CBF (arbitrary units) is indicated by the scale bar with the dark color representing the lowest and white color representing the highest flow. The short white lines in a, c, and e indicate the measurement locations of vasculatures shown in Figure 5. Panels a–f correspond to the following Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion. Direction: P, posterior; L, lateral. MMA, middle meningeal artery (white arrows). MCA, middle cerebral artery (white arrows). V, veins.
Figure 5. Representative changes of vascular diameters and perfusions associated with hemorrhagic shock. (a) The vascular CBF profiles versus time for middle meningeal artery (MMA), middle cerebral artery (MCA), and vein. The cerebral blood flow (CBF) is scaled in gray color as shown in Figure 4. The width of CBF (bright color) profile indicates the vascular diameter. The darkening indicates the decrease of blood flow. The measurement locations have been marked in Figure 4. (b) The perfusion in MMA, MCA, and vein versus time. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion. (Color version of the figure is available online.)
179
SUN ET AL
Academic Radiology, Vol 21, No 2, February 2014
Figure 6. Blood flow interruption observed for some vessels during the decompensatory stage. The images in the first row are intrinsic (Ref [reflectance]) signals of 560 nm and those in the second row are CBF images (CBF [cerebral blood flow]). T1–T4 indicates different time points during the decompensatory stage.
Figure 7. Comparison of the mean blood flow changes in middle meningeal artery (MMA), middle cerebral artery (MCA), and vein during different stages of hemorrhagic shock. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; F, after infusion. **P < .01, *P < .05.
sham group and baselines. One hour after infusion, the CBF signal appeared significantly lower (30%) (P < .01) similar to FVD (P < .05) compared to baseline (Fig 9 ‘‘Period F’’). These results suggest that after fluid resuscitation the tissue metabolic state and vascular function cannot return to normal and brain injury may have taken place. The representative results from one rat are shown in Figure 10. CBF decreased rapidly during the compensatory stage, whereas HbR rose and Hb Olympus declined. Especially at the time point around 52 minutes, NADH began to rise corresponding with the decrease of FVD and CBF and 180
the increase of HbR and HbT. This phenomenon of simultaneous changes in hemodynamic and metabolic indices indicates that the autoregulatory system entered another distinct functional state probably with functional failure. Also, the onset of both hemodynamic and metabolic changes might indicate the beginning of the transition from the compensatory to the decompensatory stage. Note that this time point cannot be predicted from the MAP and HR curves which had plateaued. NADH, HbR, HbT, FVD, CBF, MAP, and HR plateaued during the decompensatory stage. NADH, FVD, HbO, and HbR started to change simultaneously again 15 minutes prior to rapid fluid infusion. Significant changes in these parameters resulted from fluid resuscitation. HbR decreased first and started to increase around 10 minutes after infusion and reach above the baseline at 2 hours. HbO increased first and started to decrease around 10 minutes after infusion and remained significantly above the baseline at 2 hours. NADH, HbT, and HR mostly increased with time whereas FVD, CBF, and MAP decreased with time. Two hours after infusion, NADH and HbT remained dramatically above the baseline whereas FVD, CBF, MAP, and HR remained below the baseline.
DISCUSSION In this study, we used a multimodal optical imaging system to study a controlled isobaric model for severe hemorrhagic shock with a cranial window. LSCI technique was used to capture the hemodynamic information of cerebral cortex during hemorrhagic shock. The changes of CBF, vascular diameter, and vascular density can be observed at high
Academic Radiology, Vol 21, No 2, February 2014
HEMODYNAMIC AND METABOLIC CHANGES DURING SHOCK
Figure 8. Relative changes of the average oxyhemoglobin (HbO), deoxyhemoglobin (HbR), and total hemoglobin (HbT) concentrations associated with the process of severe hemorrhagic shock. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion. *P < .05.
Figure 9. Mean relative changes in parenchymal microcirculation (cerebral blood flow, functional vascular density, and nicotinamide adenine dinucleotide [NADH]) recorded during the process of severe hemorrhagic shock. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion. **P < .01, *P < .05.
temporal and spatial resolutions in real-time and in vivo. The large field of view and the deeper imaging depth (compared to endogenous fluorescence imaging) allowed us to simultaneously monitor the response of different types of blood vessels (MMA, MCA, and vein) in the process of shock (34). OISI was used to provide relative changes in HbO, HbR, and HbT. Our results showed that this is an effective approach to study the microcirculation in the process of shock.
We found that the blood supply of dural vessels such as MMA is the first to be affected under emergency, such as acute bleeding. The cerebral autoregulatory mechanism was activated during rapid bleeding period and blood circulation was redistributed by, for example, decreasing the blood flow to MMA while maintaining or even increasing the blood flow to MCA. The MMA is supplied by the external carotid artery whereas the MCA is supplied by the internal carotid 181
SUN ET AL
Figure 10. Representative changes of hemoglobin concentration, functional vascular density (FVD), nicotinamide adenine dinucleotide (NADH), cerebral blood flow (CBF), maen arterial pressure (MAP), and heart rate (HR) from a rat during hemorrhagic shock. The dashed line around 52 minutes in period C indicates the time point when NADH, FVD, HbR, HbO, and HbT started to change simultaneously. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion.
artery. The different responses of these two vessels may indicate that the blood supply regulator in the brain is located in capillaries or postcapillary brain tissues rather than in the carotid artery. During the compensatory stage, we identified regions of blood vessels with congested or stagnated blood flow. We found that venules and arterioles shrank and microcirculatory perfusion declined sharply. Microcirculatory alteration is important for the hemodynamic compensation. The shrinkage of arterioles may not only increase peripheral resistance to maintain BP but also redirect blood volume from peripheral tissues and organs (such as the skin) to vital organs (such as the brain and heart). Our data showed that the autoregulatory mechanism protected the brain circulation more than the peripheral circulation. We also found that the mean HbR increased and HbO decreased significantly during the shock. The average concentration of HbT remained stable during the entire shock process, likely because of the regulation of the compensatory mechanism. Note that a couple of other factors may confound the observation. First, the cerebral blood flow slowed down along with the local blood retention, which caused the tissue hemoglobin concentration to appear mildly elevated. Second, the physiological changes modified the scattering and absorption coefficients of tissue, which introduced experimental errors in calculating the changes of hemoglobin concentration (13). More detailed studies are needed to investigate these possibilities in the future. We have monitored NADH fluorescence signals in vivo, in addition to hemodynamic changes. The metabolic state indicates directly the oxygen and blood supply of local brain tissue during hemorrhagic shock in real time. Combining the metabolic and hemodynamic information may more accurately represent the microcirculation status during the 182
Academic Radiology, Vol 21, No 2, February 2014
process of shock. NADH, an electron donor in the mitochondrial respiratory chain, can reflect the mitochondrial metabolic state in real time (35–37). Currently, people have not found an effective method that can provide quantitative analysis of NADH concentration in vivo. Commonly, the blue fluorescence with the ultraviolet excitation is regarded as mainly contributed by NADH and represented metabolic and pathologic changes (36,38,39). The NADH fluorescence signal can be influenced by hemodynamics (light absorption and scattering) (37,40), which were partially corrected with reflectance images in our study. Some studies use redox ratio to indicate the metabolic state (41,42). The redox ratio is calculated by the fluorescence signals of flavin adenine dinucleotide (FAD), another electron carrier in the mitochondrial respiratory chain, divided by NADH signals. However, the apparent FAD fluorescence in tissue might be weak and have a low specificity to mitochondrial metabolic state, because the FAD channel might be contributed by other fluorophores, such as lipofuscin. So we did not monitor the FAD signals in this study. Our measurement indicated that parenchymal NADH remained stable at the beginning of the compensatory stage. NADH fluorescence increased to reach a plateau at the decompensatory stage and remained high after blood reinfusion. This abnormal NADH signal may indicate that metabolic processes in the cerebral circulation were damaged. However, Clavijo et al. (10) reported that urethral blood flow decreased and NADH increased in swine during a shorter period (20 minutes) of hemorrhagic shock and these parameters recovered during resuscitation. Therefore, our study might indicate that cerebral function was more vulnerable to shock than urologic function. Also the longer shock period in our study might cause irreversible damages to tissue metabolism. Further research is needed to confirm the tissue damage by histology. The most interesting phenomenon observed is the simultaneous changes of hemodynamic and metabolic signals at certain time points during the course of shock in some rats. At one time point (52 minutes in Fig 10 ‘‘Period C’’) in the compensatory stage, both the FVD and CBF started to decrease and the Hb signals started to increase when NADH started to increase. At another time point (15 minutes prior to rapid blood infusion), NADH, HbR, and HbT started to decrease, and FVD and HbO started to increase slightly. No significant changes were observed for MAP and HR at both time points. The similar phenomena were observed in two more rats. The underlying mechanisms for the simultaneous changes observed at these two time points are not known and more studies are needed to illustrate them. However, the first time point might provide a potential marker for the stage transition from the compensatory to the decompensatory stage, if we consider the following: (1) significantly decreased FVD and CBF indicated the onset of a severely limited blood/nutrition supply to cerebral tissue, (2) increased HbR and NADH indicated tissue hypoxia and
Academic Radiology, Vol 21, No 2, February 2014
HEMODYNAMIC AND METABOLIC CHANGES DURING SHOCK
inhibition of mitochondrial metabolism which might eventually cause tissue damage. We used a cranial window to observe the response of a large area of the parietal cortex. This approach inevitably changed the intracranial pressure. It is difficult to evaluate whether this change has differentially affected the hemodynamics and metabolism between normal and pathologic conditions. Moreover, it is known that severe hemorrhagic shock causes global change to the brain, and the distribution of cerebral blood flow is heterogeneous in the whole brain (4,43). The tissue responses to pathologic condition, such as hemorrhagic shock or ischemia, are heterogeneous in different regions (4). Some encephalic regions may be more susceptible to ischemic injury. It is unknown whether the damage susceptibility of deeper tissue is consistent with the cortex during hemorrhagic shock. Therefore, further studies on subcortical tissue of brains may provide more complete and important information about the brain damages and regulatory mechanism of brain circulation during hemorrhagic shock. The severe hemodynamic and metabolic damages may be associated with the onset of regulatory function failure. The most crucial aspect of effective shock treatment is to quickly and accurately diagnose the patient’s stage of shock. The biomarkers for the stage transition and the severities of metabolic damages will be of great significance for understanding not only shock physiology but also some other clinical conditions. For example, the optical techniques employed by this study may be applicable to real-time monitoring of organs during surgeries. The hemodynamic/metabolic biomarkers, if successfully developed in future, may help physicians to decide the safety time window and when to conduct fluid resuscitation and the amount of fluid needed for ischemic conditions.
ACKNOWLEDGMENTS This work was supported by the National Major Scientific Research Program of China (Grant No. 2011CB910401), the Science Fund for Creative Research Group of China (Grant No.61121004), the Director Fund of Wuhan National Laboratory for Optoelectronics, and the Specific International Scientific Cooperation (Grant No. 2010DFR30820). We appreciate the help from Xiaoli Sun, Cui Yin, Anle Wang, and April Peng. REFERENCES 1. Gutierrez G, Reines HD, Wulf-Gutierrez ME. Clinical review: hemorrhagic shock. Crit Care 2004; 8:373–381. 2. Gann DS, Drucker WR. Hemorrhagic shock. J Trauma Acute Care Surg 2013; 75:888–895. 3. Spaniol JR, Knight AR, Zebley JL, et al. Fluid resuscitation therapy for hemorrhagic shock. J Trauma Nurs 2007; 14:152–160. 4. Kovach AG, Sandor P. Cerebral blood flow and brain function during hypotension and shock. Annu Rev Physiol 1976; 38:571–596. 5. Taccone FS, De Backer D. Is cerebral microcirculation really preserved in shock states? Crit Care Med 2010; 38:1008–1009. 6. Werner C, Lu H, Engelhard K, et al. Sevoflurane impairs cerebral blood flow autoregulation in rats: reversal by nonselective nitric oxide synthase inhibition. Anesth Analg 2005; 101:509–516.
7. Wan Z, Sun SJ, Ristagno G, et al. The cerebral microcirculation is protected during experimental hemorrhagic shock. Crit Care Med 2010; 38:928–932. 8. Golden PF, Jane JA. Experimental study of irreversible shock and the brain. J Neurosurg 1973; 39:434–441. 9. Szopinski J, Kusza K, Semionow M. Microcirculatory responses to hypovolemic shock. J Trauma 2011; 71:1779–1788. 10. Clavijo JA, van Bastelaar J, Pinsky MR, et al. Minimally invasive real time monitoring of mitochondrial NADH and tissue blood flow in the urethral wall during hemorrhage and resuscitation. Med Sci Monit 2008; 14: BR175–BR182. 11. Zhao KS, Junker D, Delano FA, et al. Microvascular adjustments during irreversible hemorrhagic shock in rat skeletal muscle. Microvasc Res 1985; 30:143–153. 12. Thiemermann C, Szabo C, Mitchell JA, et al. Vascular hyporeactivity to vasoconstrictor agents and hemodynamic decompensation in hemorrhagic shock is mediated by nitric oxide. Proc Natl Acad Sci USA 1993; 90:267–271. 13. Song R, Bian H, Wang X, et al. Mitochondrial injury underlies hyporeactivity of arterial smooth muscle in severe shock. Am J Hypertens 2011; 24: 45–51. 14. Mongan PD, Capacchione J, Fontana JL, et al. Pyruvate improves cerebral metabolism during hemorrhagic shock. Am J Physiol Heart Circ Physiol 2001; 281:H854–H864. 15. Li P, Ni S, Zhang L, et al. Imaging cerebral blood flow through the intact rat skull with temporal laser speckle imaging. Opt Lett 2006; 31: 1824–1826. 16. Liu S, Li P, Luo Q. Fast blood flow visualization of high-resolution laser speckle imaging data using graphics processing unit. Opt Express 2008; 16:14321–14329. 17. Jiang C, Zhang H, Wang J, et al. Dedicated hardware processor and corresponding system-on-chip design for real-time laser speckle imaging. J Biomed Opt 2011; 16:116008. 18. Pouratian N, Sheth SA, Martin NA, et al. Shedding light on brain mapping: advances in human optical imaging. Trends Neurosci 2003; 26:277–282. 19. Hillman EM. Optical brain imaging in vivo: techniques and applications from animal to man. J Biomed Opt 2007; 12:051402. 20. Sun XL, Wang YR, Chen SB, et al. Simultaneous monitoring of intracellular pH changes and hemodynamic response during cortical spreading depression by fluorescence-corrected multimodal optical imaging. Neuroimage 2011; 57:873–884. 21. Dunn AK, Devor A, Bolay H, et al. Simultaneous imaging of total cerebral hemoglobin concentration, oxygenation, and blood flow during functional activation. Opt Lett 2003; 28:28–30. 22. Qiu J, Li P, Luo W, et al. Spatiotemporal laser speckle contrast analysis for blood flow imaging with maximized speckle contrast. J Biomed Opt 2010; 15:016003. 23. L’Heureux B, Gurden H, Pain F. Autofluorescence imaging of NADH and flavoproteins in the rat brain: insights from Monte Carlo simulations. Opt Express 2009; 17:9477–9490. 24. Kohl M, Lindauer U, Royl G, et al. Physical model for the spectroscopic analysis of cortical intrinsic optical signals. Phys Med Biol 2000; 45: 3749–3764. 25. Dunn AK, Devor A, Dale AM, et al. Spatial extent of oxygen metabolism and hemodynamic changes during functional activation of the rat somatosensory cortex. Neuroimage 2005; 27:279–290. 26. Prahl S. Optical Absorption of Hemoglobin. Available at: http://omlc.ogi. edu/spectra/hemoglobin/index.html.1998;1. Jan. 11, 2010. 27. Sirotin YB, Hillman EMC, Bordier C, et al. Spatiotemporal precision and hemodynamic mechanism of optical point spreads in alert primates. Proc Natl Acad Sci USA 2009; 106:18390–18395. 28. Schmid-Schoenbein GW, Zweifach BW, Kovalcheck S. The application of stereological principles to morphometry of the microcirculation in different tissues. Microvasc Res 1977; 14:303–317. 29. White SM, George SC, Choi B. Automated computation of functional vascular density using laser speckle imaging in a rodent window chamber model. Microvasc Res 2011; 82:92–95. 30. Chen S, Li P, Zeng S, et al. Using threshold segmentation methods to measure dynamic vasodilatation in a series of optical images. Proc SPIE 2005;151–158. 31. Nielsen AN, Fabricius M, Lauritzen M. Scanning laser-Doppler flowmetry of rat cerebral circulation during cortical spreading depression. J Vasc Res 2000; 37:513–522.
183
SUN ET AL
32. O’Farrell AM, Rex DE, Muthialu A, et al. Characterization of optical intrinsic signals and blood volume during cortical spreading depression. Neuroreport 2000; 11:2121–2125. 33. Brookes ZLS, Brown NJ, Reilly CS. Response of the rat cremaster microcirculation to hemorrhage in vivo: differential effects of intravenous anesthetic agents. Shock 2002; 18:542–548. 34. Bolay H, Reuter U, Dunn AK, et al. Intrinsic brain activity triggers trigeminal meningeal afferents in a migraine model. Nat Med 2002; 8:136–142. 35. Chance B, Baltscheffsky H. Respiratory enzymes in oxidative phosphorylation. J Biol Chem 1958; 233:736–739. 36. Chance B, Cohen P, Jobsis F, et al. Intracellular oxidation-reduction states in vivo. Science 1962; 137:499–508. 37. Chance B, Schoener B, Oshino R, et al. Oxidation-reduction ratio studies of mitochondria in freeze-trapped samples. NADH and flavoprotein fluorescence signals. J Biol Chem 1979; 254:4764–4771.
184
Academic Radiology, Vol 21, No 2, February 2014
38. Mayevsky A, Rogatsky GG. Mitochondrial function in vivo evaluated by NADH fluorescence: from animal models to human studies. Am J Physiol Cell Physiol 2007; 292:C615–C640. 39. Mayevsky A. Mitochondrial function and tissue viability in vivo: from animal experiments to clinical applications. Forty years of fruitful collaboration with Britton Chance. J Innov Opt Health Sci 2011; 04:337–359. 40. Bradley RS, Thorniley MS. A review of attenuation correction techniques for tissue fluorescence. J R Soc Interface 2006; 3:1–13. 41. Quistorff B, Haselgrove JC, Chance B. High spatial resolution readout of 3-D metabolic organ structure: an automated, low-temperature redox ratio-scanning instrument. Anal Biochem 1985; 148:389–400. 42. Li LZ, Xu HN, Ranji M, et al. Mitochondrial redox imaging for cancer diagnostic and therapeutic studies. J Innov Opt Health Sci 2009; 2:325–341. 43. Strandgaard S, Paulson OB. Cerebral autoregulation. Stroke 1984; 15: 413–416.
H
emorrhagic shock is a condition caused by a rapid and significant loss of intravascular volume, which may lead to hemodynamic instability, decreased tissue perfusion, cellular hypoxia, organ damage, and death (1,2). Usually, the process of hemorrhagic shock is described in three stages (or in four stages if defining an initial stage
Acad Radiol 2014; 21:175–184 From the Britton Chance Center for Biomedical Photonics, Wuhan National Laboratory for Optoelectronics, Huazhong University of Science and Technology, Wuhan, Hubei, PR China (N.S., W.L., Q.L.); Department of Biomedical Engineering, Key Laboratory of Biomedical Photonics of Ministry of Education, Huazhong University of Science and Technology, 1037 Luoyu Road, Wuhan, Hubei 430074, PR China (N.S., W.L., Q.L.); Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA (N.S., L.Z.L.); Department of Biochemistry and Biophysics, Britton Chance Laboratory of Redox Imaging, Johnson Research Foundation, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA (N.S., L.Z.L.); and Institute of Translational Medicine and Therapeutics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA (L.Z.L.). Received August 21, 2013; accepted November 26, 2013. This article is dedicated to the 100th birthday of Dr. Britton Chance, whose guidance and support were instrumental for the establishment of our research laboratories. Address correspondence to: Q.L. e-mail: [email protected] ªAUR, 2014 http://dx.doi.org/10.1016/j.acra.2013.11.017
before the compensatory stage) (1,3). In the first stage (i.e., the compensatory stage) the body undergoes microcirculatory ischemia. This is characterized by compensatory hypotension during which the minimum perfusion level of vital organs is maintained by autoregulatory mechanism. In the second stage (i.e., the decompensatory stage) the body exhibits microcirculatory congestion, characterized by decompensatory hypotension that indicates the failure of autoregulatory mechanism. The last stage is the refractory stage, characterized by refractory hypotension and microcirculation failure followed by irreversible cell damage and organ failure. Whether the cerebral circulation is protected by the autoregulatory mechanism during the process of the shock has long been controversial (4,5). Previous studies have highlighted that cerebral blood flow (CBF) in rats is maintained when mean arterial pressure (MAP) is kept between 60 and 140 mmHg (6). Animals would have lower MAP if they underwent a greater blood loss. Wan et al. (7) found that the microcirculation of the cerebral cortex could be maintained after losing 35% of blood volume (MAP could be reduced to 60 mmHg or less). However, they found that the buccal microcirculation supplied by the carotid artery showed 175
SUN ET AL
damages that are consistent with the changes in peripheral circulation. Furthermore, the protection of cerebral circulation resulting from autoregulatory mechanism may be temporary during the process of shock (1). The autoregulatory mechanism will show signs of failure once the blood pressure (BP) exceeds the lower threshold for an extended period (4,8). Therefore, during the process of shock, particularly during the decompensatory stage, it is important to pay close attention to the dynamic changes in the circulatory states of vital organs, such as the heart and brain (9). These factors have important clinical significance for determining an appropriate fluid resuscitation strategy and timely initiation of protective measures for heart and brain (7). Typically, there are eight indicators used for monitoring the shock under clinical settings: BP, heart rate (HR), central venous pressure, cardiac output, urine volume, blood gas analysis, hematocrit, and arterial blood lactate. However, these indicators neither can be monitored in real time nor provide local hemodynamic and metabolic information. Some studies have shown that monitoring the changes in local microcirculation, hemodynamics, and tissue metabolic states of vital organs are conducive to understanding the autoregulatory mechanisms, tissue function, and injury during hemorrhagic shock (4,10). Ultimately, these real-time monitoring may help clinicians to determine a patient’s stage of shock, grasp the safety time window, and administer appropriate treatments. To study the dynamic vascular perfusion (including dural vessels supplied by the external carotid artery and pial vessels supplied by the internal carotid artery), vascular response, and metabolic change in a rat model of severe hemorrhagic shock (11–14), we used a multimode optical imaging technology, that is, an integration of laser speckle contrast imaging (LSCI) (15–17), optical intrinsic signal imaging (OISI) (18,19), and endogenous nicotinamide adenine dinucleotide (NADH) fluorescence imaging (20). The rat model was achieved through rapid bleeding to obtain a low BP level of about 40 mmHg. The results of our study indicate that the transitional process from compensatory stage to decompensatory stage can be simulated by this rat model. During the initial and compensatory stages, cerebral parenchymal circulation was protected by inhibiting the perfusion of dural vessels. The blood supplies of dural vessels are originated from the external carotid artery, and the perfusion changes of dural vessels are consistent with the changes of systemic circulation. In addition, the cerebral hemodynamic and metabolic signal signatures may reflect underlying tissue damages during the decompensatory stage and provide potential biomarkers for the shock stages and the stage transition. MATERIALS AND METHODS
Academic Radiology, Vol 21, No 2, February 2014
Figure 1. A schematic representation of experimental procedures performed during severe hemorrhagic shock of the rat model. Blood was withdrawn from the femoral artery until the mean arterial pressure (MAP) decreased to about 40 mmHg in 15 minutes. MAP was maintained at 40 mmHg in the compensatory and decompensatory stages by continually withdrawing or injecting blood. Fluid resuscitation was performed during the rapid infusion period by returning all blood withdrawn to the rat. Periods: A, baseline; B, rapid blood bleeding; C, compensatory stage; D, decompensatory stage; E, rapid blood infusion; and F, after blood infusion.
(50 mg/kg) and urethane (600 mg/kg). Anesthesia was maintained by supplements of one-sixth of the initial dose at intervals of 30–45 minutes. The body temperatures were maintained at 37.0 0.5 C using a rectal probe and a feedback-controlled heating blanket. PE-50 cannulas were inserted into the femoral arteries on both sides and into the femoral vein on the right side. The heads of the rats were held in position using a stereotaxic apparatus and craniotomies were performed over the right parietal bone. The dura stayed intact and was continually bathed with artificial cerebrospinal fluid. Experimental Procedures
Rats were randomly divided into a shock group (n = 7) and a sham group (n = 5). The experimental procedure has been described previously (11–14) and a brief overview is provided in Figure 1. We measured MAP and HR with a pressure transducer (model RM6240, multichannel physiological parameters monitoring instrument; Chengdu Instrument Factory, Chengdu, China) in the left femoral artery. The rapid bleeding from the right femoral artery was conducted until MAP dropped to 40 mmHg; MAP was maintained at 40 mmHg by continual blood withdrawal (up to 50% of the total blood volume) and blood injection for 2 hours. We assumed that the rat entered the decompensatory stage if the BP had been maintained at 40 mmHg for 90 minutes. The fluid resuscitation was carried out half an hour after the outset of the decompensatory stage with a rapid infusion of the total amount of blood withdrawn back into the rat. The exposed rat brain tissues were continuously monitored by the multimodal optical imaging system during the entire experiment. The experimental procedure of the sham group was the same as the shock group except no blood withdrawal and injection. Imaging Instrument
Animal Preparation
The experimental protocol was approved by the Institutional Animal Care and Use Committee at Huazhong University of Science and Technology. Male Wistar rats (220–280 g, n = 12) were anesthetized by intraperitoneal injection of a-chloralose 176
The multimodal optical imaging system (20,21) integrates NADH autofluorescence imaging, blood flow LSCI, and dual-wavelength OISI (reflectance imaging). It can simultaneously monitor changes of a few physiological parameters in vivo as shown in Figure 2. Two identical 12-bit
Academic Radiology, Vol 21, No 2, February 2014
HEMODYNAMIC AND METABOLIC CHANGES DURING SHOCK
Figure 2. A technical diagram for multimodal optical imaging. DHbO, DHbR, and DHbT are relative concentration changes compared to the baselines of oxyhemoglobin (HbO), deoxyhemoglobin (HbR), and total hemoglobin (HbT), respectively. CBF, cerebral blood flow; FVD, functional vascular density; HR, heart rate; MAP, maen arterial pressure.
charge-coupled device (CCD) cameras (2 2 binning, 1024 1396 pixels, Pixelfly QE; PCO Computer Optics, Kelheim, Germany) were attached in parallel to a stereomicroscopy (Olympus SZ6045 TR Zoom, Olympus, Japan) with a field of view 2.62 mm 3.56 mm. One of the CCD cameras was used to image the NADH fluorescence with a band-pass emission filter (475 nm 21 nm, FF01-475/ 42-25; Semrock, Rochester, NY). The other camera was used to image the rest of the three images sequentially. A liquid crystal tunable filter (bandwidth 7 nm, VariSpec; Cambridge Research & Instrumentation, Woburn, MA) was placed in front of the second CCD camera. An ultraviolet light emitting diode (365 nm for NADH excitation, Thorlab, Newton, NJ), a laser diode (660 nm), and a filtered halogen lamp (550– 590 nm, Olympus LG-PS2, Olympus, Tokyo, Japan) were positioned around the rat brain cortex to provide an even illumination or fluorescence excitation. As shown in Figure 2, the NADH fluorescence emission was imaged at 475 nm, OISI at 560 nm and 570 nm, and LSCI at 660 nm (15–17,22). Data Analyses
Using the modified Beer–Lambert relationship, we obtained an equation that relates the signal intensities to the changes of hemoglobin concentrations. log I0l =Itl ¼ mlHbR Dl DHbR þ mlHbO Dl DHbO:
(1)
where l is the light wavelength, I0l and Itl are the measured intensities of reflected light at the light wavelength l for the baseline (t = 0) and time t, respectively; mlHbR and mlHbO are absorption coefficients of oxyhemoglobin (HbO) and deoxyhemoglobin (HbR), respectively; Dl is the differential pathlength factor, that is, the mean optical pathlength of the photons travelling in the tissue; DHbO, DHbR, and DHbT represent the concentration changes of oxyhemoglobin,
deoxyhemoglobin, and total hemoglobin (HbT), respectively. Dl in brain tissue was evaluated and quantified by the Monte Carlo simulation at 350 nm (365 nm), 480 nm (475 nm), 560 nm, and 570 nm using the approach of Kohl et al. and Andrew et al. (23–26). We used the simulated Dl for 560 nm and 570 nm to compute DHbO, DHbR, and DHbT (DHbT = DHbO + DHbR) (24–26) by Equation (1). The intrinsic signals (reflectance) at 350 nm and 480 nm were predicted from DHbO and DHbR using the approach of Yevgeniy et al. (27), and their reflected signals were then used to correct the NADH signals measured from the first CCD (20). CBF was obtained through online processing (customized C program) by the laser speckle temporal contrast analysis of 100 consecutive raw speckles images (15–17,22). FVD is defined as the total length of perfused vessels within a region of interest in a two-dimensional image divided by the area of that region (28). FVD, vascular diameter, and perfusion were computed from laser speckle images, and the details of the quantification methods were shown in previous works (20,29,30). The blood perfusion was obtained by multiplying the CBF and the cross-sectional area of a blood vessel. Vasculatures such as the middle meningeal artery (MMA), middle cerebral artery (MCA), and veins were identified by vessel diameter and morphology (31,32).
RESULTS Alterations of Hemodynamics
Although MAP and HR are physiological representations of the functional state of systemic circulation, MAP is the most important standard for monitoring the shock progression in our model. After rapid bleeding, HR declined rapidly and MAP was maintained at 40 mmHg until fluid resuscitation (Fig 3). After rapid fluid infusion, MAP increased sharply but HR did so slowly. One hour after infusion, the MAP 177
SUN ET AL
Figure 3. Average maen arterial pressure (MAP) and heart rate (HR) recording during the process of severe hemorrhagic shock. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid blood infusion; and F, after blood infusion. **P < .01 comparing between the shock and control group.
and HR of the shock group were not significantly different from those of the control group. Figures 4 and 5 show the changes of CBF, vascular diameter, and vascular perfusion that reflect the microcirculatory and hemorheological conditions. After the rapid bleeding (Fig 4b, Fig 5 ‘‘Period B’’), the blood flow, diameter, and perfusion of MMA decreased. Conversely, the diameter and perfusion of MCA increased, whereas the veins remained relatively stable. At the compensatory stage (Fig 4c), MMA (white arrows in Fig 4a) disappeared along with its diameter and perfusion reduced significantly to zero as shown in Figure 5 ‘‘Period C’’. Diameters of MCA and veins increased and their perfusions first increased and then decreased (Fig 5 ‘‘Period C’’). In general, the CBF image in Figure 4c is darker than that in Figure 4b, which indicates that the blood flow through parenchymal tissue decreased significantly. During the decompensatory stage, a number of vessels dimmed or disappeared in Figure 4d compared to those in Figure 4c. As shown in Figure 5 ‘‘Period D’’, the diameter of MCA initially increased before decreasing, whereas the vein expanded persistently. The perfusion of MCA decreased to the same level as that of the vein and remained steady until 15 minutes prior to rapid infusion. We also observed the interruption of blood flow during the decompensatory stage, which was presumably caused by leukocyte aggregates. The interruptions are shown by the white circles in Figure 6. Consequently, microcirculatory blood flow slowed and arterial perfusion continued to decrease at 15 minutes prior to rapid infusion in Figure 5. The absorptions of excitation and emission light were presumably enhanced by increased blood volume so that the optical images darkened (Fig 4d). The images were lighted up significantly by the blood infusion (Figs 4e and f). During the blood infusion, the blood flow, perfusions, and diameters of MCA and veins increased significantly followed by a dramatic postinfusion decrease that was even below baseline eventually (Fig 5 ‘‘Period F’’). 178
Academic Radiology, Vol 21, No 2, February 2014
The responses of vessels vary significantly dependent on vessel types and diameters (33). Statistical analyses of the mean blood flow changes of MMA, MCA, and vein are shown in Figure 7. During the rapid bleeding, there was no significant difference in the mean blood flow of MCA and veins compared to the baseline (Fig 7 ‘‘Period A’’). However, the mean blood flow of MMA decreased significantly by 20%. The mean blood flow change of MMA is the most visible one after rapid bleeding. It decreased by 70% during the compensatory stage (Fig 7 ‘‘Periods B and C’’) compared to the baseline. In comparison, the mean blood flow changes of MCA and veins decreased by 40%–50% at the compensatory stage compared to the baseline. Blood flow changes of the MMA are significantly different than those of the MCA and veins (P < .01, in Fig 7 ‘‘Period C’’) during the compensatory stage. The differences between MMA and MCA persisted in the decompensatory stage. The blood flows in these three types of vessels largely recovered after fluid resuscitation (Fig 7 ‘‘Period E’’) with MMA flow reaching 65%, MCA 80%, and vein 70% of the baseline level. However, their flows decreased again after rapid infusion. The difference of blood flows between MMA and MCA maintained statistical significance (P < .05) 1 hour after infusion. The changes in hemoglobin concentration were statistically analyzed and shown in Figure 8. In the compensatory stage, the mean HbR increased by 10% and mean HbO decreased by 5% compared to the baseline (P < .05). That indicated tissue hypoxia and/or tissue ischemia with excessive consumption of HbO. In other stages of shock, no significant variations in HbO and HbR were observed. The mean HbT remained stable during the entire shock process despite the blood bleeding. Alteration of Metabolism
The temporal changes of the average values of brain parenchymal CBF, FVD, and NADH fluorescence signals are shown in Figure 9 along with statistical analysis results. All three parameters exhibited no significant changes compared to the sham group and the baseline during the rapid bleeding. During both the compensatory and decompensatory stages, the NADH signal increased relative to the sham group but with no significant differences. The NADH signal was 20% greater than that of the sham group after the fluid resuscitation (P < .05, in Figure 9 ‘‘Period E’’). One hour after infusion, the NADH signal rose further about 40% above the baseline value (P < .05). NADH fluorescence kept increasing after rapid bleeding compared to the initial stage. This indicates that the metabolic state of mitochondria was changed and mitochondria probably became dysfunctional. The brain parenchymal microcirculation (CBF) and FVD decreased significantly, that is, 55% (P < .01) and 20% (P < .05), respectively, during both the compensatory and decompensatory stages, as shown in Figure 9 ‘‘Period C’’. The signals of FVD and CBF were increased by fluid resuscitation and almost totally recovered to the levels of
Academic Radiology, Vol 21, No 2, February 2014
HEMODYNAMIC AND METABOLIC CHANGES DURING SHOCK
Figure 4. Representative images of cerebral blood flow (CBF) at different time points during the hemorrhagic shock. The amount of CBF (arbitrary units) is indicated by the scale bar with the dark color representing the lowest and white color representing the highest flow. The short white lines in a, c, and e indicate the measurement locations of vasculatures shown in Figure 5. Panels a–f correspond to the following Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion. Direction: P, posterior; L, lateral. MMA, middle meningeal artery (white arrows). MCA, middle cerebral artery (white arrows). V, veins.
Figure 5. Representative changes of vascular diameters and perfusions associated with hemorrhagic shock. (a) The vascular CBF profiles versus time for middle meningeal artery (MMA), middle cerebral artery (MCA), and vein. The cerebral blood flow (CBF) is scaled in gray color as shown in Figure 4. The width of CBF (bright color) profile indicates the vascular diameter. The darkening indicates the decrease of blood flow. The measurement locations have been marked in Figure 4. (b) The perfusion in MMA, MCA, and vein versus time. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion. (Color version of the figure is available online.)
179
SUN ET AL
Academic Radiology, Vol 21, No 2, February 2014
Figure 6. Blood flow interruption observed for some vessels during the decompensatory stage. The images in the first row are intrinsic (Ref [reflectance]) signals of 560 nm and those in the second row are CBF images (CBF [cerebral blood flow]). T1–T4 indicates different time points during the decompensatory stage.
Figure 7. Comparison of the mean blood flow changes in middle meningeal artery (MMA), middle cerebral artery (MCA), and vein during different stages of hemorrhagic shock. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; F, after infusion. **P < .01, *P < .05.
sham group and baselines. One hour after infusion, the CBF signal appeared significantly lower (30%) (P < .01) similar to FVD (P < .05) compared to baseline (Fig 9 ‘‘Period F’’). These results suggest that after fluid resuscitation the tissue metabolic state and vascular function cannot return to normal and brain injury may have taken place. The representative results from one rat are shown in Figure 10. CBF decreased rapidly during the compensatory stage, whereas HbR rose and Hb Olympus declined. Especially at the time point around 52 minutes, NADH began to rise corresponding with the decrease of FVD and CBF and 180
the increase of HbR and HbT. This phenomenon of simultaneous changes in hemodynamic and metabolic indices indicates that the autoregulatory system entered another distinct functional state probably with functional failure. Also, the onset of both hemodynamic and metabolic changes might indicate the beginning of the transition from the compensatory to the decompensatory stage. Note that this time point cannot be predicted from the MAP and HR curves which had plateaued. NADH, HbR, HbT, FVD, CBF, MAP, and HR plateaued during the decompensatory stage. NADH, FVD, HbO, and HbR started to change simultaneously again 15 minutes prior to rapid fluid infusion. Significant changes in these parameters resulted from fluid resuscitation. HbR decreased first and started to increase around 10 minutes after infusion and reach above the baseline at 2 hours. HbO increased first and started to decrease around 10 minutes after infusion and remained significantly above the baseline at 2 hours. NADH, HbT, and HR mostly increased with time whereas FVD, CBF, and MAP decreased with time. Two hours after infusion, NADH and HbT remained dramatically above the baseline whereas FVD, CBF, MAP, and HR remained below the baseline.
DISCUSSION In this study, we used a multimodal optical imaging system to study a controlled isobaric model for severe hemorrhagic shock with a cranial window. LSCI technique was used to capture the hemodynamic information of cerebral cortex during hemorrhagic shock. The changes of CBF, vascular diameter, and vascular density can be observed at high
Academic Radiology, Vol 21, No 2, February 2014
HEMODYNAMIC AND METABOLIC CHANGES DURING SHOCK
Figure 8. Relative changes of the average oxyhemoglobin (HbO), deoxyhemoglobin (HbR), and total hemoglobin (HbT) concentrations associated with the process of severe hemorrhagic shock. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion. *P < .05.
Figure 9. Mean relative changes in parenchymal microcirculation (cerebral blood flow, functional vascular density, and nicotinamide adenine dinucleotide [NADH]) recorded during the process of severe hemorrhagic shock. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion. **P < .01, *P < .05.
temporal and spatial resolutions in real-time and in vivo. The large field of view and the deeper imaging depth (compared to endogenous fluorescence imaging) allowed us to simultaneously monitor the response of different types of blood vessels (MMA, MCA, and vein) in the process of shock (34). OISI was used to provide relative changes in HbO, HbR, and HbT. Our results showed that this is an effective approach to study the microcirculation in the process of shock.
We found that the blood supply of dural vessels such as MMA is the first to be affected under emergency, such as acute bleeding. The cerebral autoregulatory mechanism was activated during rapid bleeding period and blood circulation was redistributed by, for example, decreasing the blood flow to MMA while maintaining or even increasing the blood flow to MCA. The MMA is supplied by the external carotid artery whereas the MCA is supplied by the internal carotid 181
SUN ET AL
Figure 10. Representative changes of hemoglobin concentration, functional vascular density (FVD), nicotinamide adenine dinucleotide (NADH), cerebral blood flow (CBF), maen arterial pressure (MAP), and heart rate (HR) from a rat during hemorrhagic shock. The dashed line around 52 minutes in period C indicates the time point when NADH, FVD, HbR, HbO, and HbT started to change simultaneously. Periods: A, baseline; B, rapid bleeding; C, compensatory stage; D, decompensatory stage; E, rapid infusion; and F, after infusion.
artery. The different responses of these two vessels may indicate that the blood supply regulator in the brain is located in capillaries or postcapillary brain tissues rather than in the carotid artery. During the compensatory stage, we identified regions of blood vessels with congested or stagnated blood flow. We found that venules and arterioles shrank and microcirculatory perfusion declined sharply. Microcirculatory alteration is important for the hemodynamic compensation. The shrinkage of arterioles may not only increase peripheral resistance to maintain BP but also redirect blood volume from peripheral tissues and organs (such as the skin) to vital organs (such as the brain and heart). Our data showed that the autoregulatory mechanism protected the brain circulation more than the peripheral circulation. We also found that the mean HbR increased and HbO decreased significantly during the shock. The average concentration of HbT remained stable during the entire shock process, likely because of the regulation of the compensatory mechanism. Note that a couple of other factors may confound the observation. First, the cerebral blood flow slowed down along with the local blood retention, which caused the tissue hemoglobin concentration to appear mildly elevated. Second, the physiological changes modified the scattering and absorption coefficients of tissue, which introduced experimental errors in calculating the changes of hemoglobin concentration (13). More detailed studies are needed to investigate these possibilities in the future. We have monitored NADH fluorescence signals in vivo, in addition to hemodynamic changes. The metabolic state indicates directly the oxygen and blood supply of local brain tissue during hemorrhagic shock in real time. Combining the metabolic and hemodynamic information may more accurately represent the microcirculation status during the 182
Academic Radiology, Vol 21, No 2, February 2014
process of shock. NADH, an electron donor in the mitochondrial respiratory chain, can reflect the mitochondrial metabolic state in real time (35–37). Currently, people have not found an effective method that can provide quantitative analysis of NADH concentration in vivo. Commonly, the blue fluorescence with the ultraviolet excitation is regarded as mainly contributed by NADH and represented metabolic and pathologic changes (36,38,39). The NADH fluorescence signal can be influenced by hemodynamics (light absorption and scattering) (37,40), which were partially corrected with reflectance images in our study. Some studies use redox ratio to indicate the metabolic state (41,42). The redox ratio is calculated by the fluorescence signals of flavin adenine dinucleotide (FAD), another electron carrier in the mitochondrial respiratory chain, divided by NADH signals. However, the apparent FAD fluorescence in tissue might be weak and have a low specificity to mitochondrial metabolic state, because the FAD channel might be contributed by other fluorophores, such as lipofuscin. So we did not monitor the FAD signals in this study. Our measurement indicated that parenchymal NADH remained stable at the beginning of the compensatory stage. NADH fluorescence increased to reach a plateau at the decompensatory stage and remained high after blood reinfusion. This abnormal NADH signal may indicate that metabolic processes in the cerebral circulation were damaged. However, Clavijo et al. (10) reported that urethral blood flow decreased and NADH increased in swine during a shorter period (20 minutes) of hemorrhagic shock and these parameters recovered during resuscitation. Therefore, our study might indicate that cerebral function was more vulnerable to shock than urologic function. Also the longer shock period in our study might cause irreversible damages to tissue metabolism. Further research is needed to confirm the tissue damage by histology. The most interesting phenomenon observed is the simultaneous changes of hemodynamic and metabolic signals at certain time points during the course of shock in some rats. At one time point (52 minutes in Fig 10 ‘‘Period C’’) in the compensatory stage, both the FVD and CBF started to decrease and the Hb signals started to increase when NADH started to increase. At another time point (15 minutes prior to rapid blood infusion), NADH, HbR, and HbT started to decrease, and FVD and HbO started to increase slightly. No significant changes were observed for MAP and HR at both time points. The similar phenomena were observed in two more rats. The underlying mechanisms for the simultaneous changes observed at these two time points are not known and more studies are needed to illustrate them. However, the first time point might provide a potential marker for the stage transition from the compensatory to the decompensatory stage, if we consider the following: (1) significantly decreased FVD and CBF indicated the onset of a severely limited blood/nutrition supply to cerebral tissue, (2) increased HbR and NADH indicated tissue hypoxia and
Academic Radiology, Vol 21, No 2, February 2014
HEMODYNAMIC AND METABOLIC CHANGES DURING SHOCK
inhibition of mitochondrial metabolism which might eventually cause tissue damage. We used a cranial window to observe the response of a large area of the parietal cortex. This approach inevitably changed the intracranial pressure. It is difficult to evaluate whether this change has differentially affected the hemodynamics and metabolism between normal and pathologic conditions. Moreover, it is known that severe hemorrhagic shock causes global change to the brain, and the distribution of cerebral blood flow is heterogeneous in the whole brain (4,43). The tissue responses to pathologic condition, such as hemorrhagic shock or ischemia, are heterogeneous in different regions (4). Some encephalic regions may be more susceptible to ischemic injury. It is unknown whether the damage susceptibility of deeper tissue is consistent with the cortex during hemorrhagic shock. Therefore, further studies on subcortical tissue of brains may provide more complete and important information about the brain damages and regulatory mechanism of brain circulation during hemorrhagic shock. The severe hemodynamic and metabolic damages may be associated with the onset of regulatory function failure. The most crucial aspect of effective shock treatment is to quickly and accurately diagnose the patient’s stage of shock. The biomarkers for the stage transition and the severities of metabolic damages will be of great significance for understanding not only shock physiology but also some other clinical conditions. For example, the optical techniques employed by this study may be applicable to real-time monitoring of organs during surgeries. The hemodynamic/metabolic biomarkers, if successfully developed in future, may help physicians to decide the safety time window and when to conduct fluid resuscitation and the amount of fluid needed for ischemic conditions.
ACKNOWLEDGMENTS This work was supported by the National Major Scientific Research Program of China (Grant No. 2011CB910401), the Science Fund for Creative Research Group of China (Grant No.61121004), the Director Fund of Wuhan National Laboratory for Optoelectronics, and the Specific International Scientific Cooperation (Grant No. 2010DFR30820). We appreciate the help from Xiaoli Sun, Cui Yin, Anle Wang, and April Peng. REFERENCES 1. Gutierrez G, Reines HD, Wulf-Gutierrez ME. Clinical review: hemorrhagic shock. Crit Care 2004; 8:373–381. 2. Gann DS, Drucker WR. Hemorrhagic shock. J Trauma Acute Care Surg 2013; 75:888–895. 3. Spaniol JR, Knight AR, Zebley JL, et al. Fluid resuscitation therapy for hemorrhagic shock. J Trauma Nurs 2007; 14:152–160. 4. Kovach AG, Sandor P. Cerebral blood flow and brain function during hypotension and shock. Annu Rev Physiol 1976; 38:571–596. 5. Taccone FS, De Backer D. Is cerebral microcirculation really preserved in shock states? Crit Care Med 2010; 38:1008–1009. 6. Werner C, Lu H, Engelhard K, et al. Sevoflurane impairs cerebral blood flow autoregulation in rats: reversal by nonselective nitric oxide synthase inhibition. Anesth Analg 2005; 101:509–516.
7. Wan Z, Sun SJ, Ristagno G, et al. The cerebral microcirculation is protected during experimental hemorrhagic shock. Crit Care Med 2010; 38:928–932. 8. Golden PF, Jane JA. Experimental study of irreversible shock and the brain. J Neurosurg 1973; 39:434–441. 9. Szopinski J, Kusza K, Semionow M. Microcirculatory responses to hypovolemic shock. J Trauma 2011; 71:1779–1788. 10. Clavijo JA, van Bastelaar J, Pinsky MR, et al. Minimally invasive real time monitoring of mitochondrial NADH and tissue blood flow in the urethral wall during hemorrhage and resuscitation. Med Sci Monit 2008; 14: BR175–BR182. 11. Zhao KS, Junker D, Delano FA, et al. Microvascular adjustments during irreversible hemorrhagic shock in rat skeletal muscle. Microvasc Res 1985; 30:143–153. 12. Thiemermann C, Szabo C, Mitchell JA, et al. Vascular hyporeactivity to vasoconstrictor agents and hemodynamic decompensation in hemorrhagic shock is mediated by nitric oxide. Proc Natl Acad Sci USA 1993; 90:267–271. 13. Song R, Bian H, Wang X, et al. Mitochondrial injury underlies hyporeactivity of arterial smooth muscle in severe shock. Am J Hypertens 2011; 24: 45–51. 14. Mongan PD, Capacchione J, Fontana JL, et al. Pyruvate improves cerebral metabolism during hemorrhagic shock. Am J Physiol Heart Circ Physiol 2001; 281:H854–H864. 15. Li P, Ni S, Zhang L, et al. Imaging cerebral blood flow through the intact rat skull with temporal laser speckle imaging. Opt Lett 2006; 31: 1824–1826. 16. Liu S, Li P, Luo Q. Fast blood flow visualization of high-resolution laser speckle imaging data using graphics processing unit. Opt Express 2008; 16:14321–14329. 17. Jiang C, Zhang H, Wang J, et al. Dedicated hardware processor and corresponding system-on-chip design for real-time laser speckle imaging. J Biomed Opt 2011; 16:116008. 18. Pouratian N, Sheth SA, Martin NA, et al. Shedding light on brain mapping: advances in human optical imaging. Trends Neurosci 2003; 26:277–282. 19. Hillman EM. Optical brain imaging in vivo: techniques and applications from animal to man. J Biomed Opt 2007; 12:051402. 20. Sun XL, Wang YR, Chen SB, et al. Simultaneous monitoring of intracellular pH changes and hemodynamic response during cortical spreading depression by fluorescence-corrected multimodal optical imaging. Neuroimage 2011; 57:873–884. 21. Dunn AK, Devor A, Bolay H, et al. Simultaneous imaging of total cerebral hemoglobin concentration, oxygenation, and blood flow during functional activation. Opt Lett 2003; 28:28–30. 22. Qiu J, Li P, Luo W, et al. Spatiotemporal laser speckle contrast analysis for blood flow imaging with maximized speckle contrast. J Biomed Opt 2010; 15:016003. 23. L’Heureux B, Gurden H, Pain F. Autofluorescence imaging of NADH and flavoproteins in the rat brain: insights from Monte Carlo simulations. Opt Express 2009; 17:9477–9490. 24. Kohl M, Lindauer U, Royl G, et al. Physical model for the spectroscopic analysis of cortical intrinsic optical signals. Phys Med Biol 2000; 45: 3749–3764. 25. Dunn AK, Devor A, Dale AM, et al. Spatial extent of oxygen metabolism and hemodynamic changes during functional activation of the rat somatosensory cortex. Neuroimage 2005; 27:279–290. 26. Prahl S. Optical Absorption of Hemoglobin. Available at: http://omlc.ogi. edu/spectra/hemoglobin/index.html.1998;1. Jan. 11, 2010. 27. Sirotin YB, Hillman EMC, Bordier C, et al. Spatiotemporal precision and hemodynamic mechanism of optical point spreads in alert primates. Proc Natl Acad Sci USA 2009; 106:18390–18395. 28. Schmid-Schoenbein GW, Zweifach BW, Kovalcheck S. The application of stereological principles to morphometry of the microcirculation in different tissues. Microvasc Res 1977; 14:303–317. 29. White SM, George SC, Choi B. Automated computation of functional vascular density using laser speckle imaging in a rodent window chamber model. Microvasc Res 2011; 82:92–95. 30. Chen S, Li P, Zeng S, et al. Using threshold segmentation methods to measure dynamic vasodilatation in a series of optical images. Proc SPIE 2005;151–158. 31. Nielsen AN, Fabricius M, Lauritzen M. Scanning laser-Doppler flowmetry of rat cerebral circulation during cortical spreading depression. J Vasc Res 2000; 37:513–522.
183
SUN ET AL
32. O’Farrell AM, Rex DE, Muthialu A, et al. Characterization of optical intrinsic signals and blood volume during cortical spreading depression. Neuroreport 2000; 11:2121–2125. 33. Brookes ZLS, Brown NJ, Reilly CS. Response of the rat cremaster microcirculation to hemorrhage in vivo: differential effects of intravenous anesthetic agents. Shock 2002; 18:542–548. 34. Bolay H, Reuter U, Dunn AK, et al. Intrinsic brain activity triggers trigeminal meningeal afferents in a migraine model. Nat Med 2002; 8:136–142. 35. Chance B, Baltscheffsky H. Respiratory enzymes in oxidative phosphorylation. J Biol Chem 1958; 233:736–739. 36. Chance B, Cohen P, Jobsis F, et al. Intracellular oxidation-reduction states in vivo. Science 1962; 137:499–508. 37. Chance B, Schoener B, Oshino R, et al. Oxidation-reduction ratio studies of mitochondria in freeze-trapped samples. NADH and flavoprotein fluorescence signals. J Biol Chem 1979; 254:4764–4771.
184
Academic Radiology, Vol 21, No 2, February 2014
38. Mayevsky A, Rogatsky GG. Mitochondrial function in vivo evaluated by NADH fluorescence: from animal models to human studies. Am J Physiol Cell Physiol 2007; 292:C615–C640. 39. Mayevsky A. Mitochondrial function and tissue viability in vivo: from animal experiments to clinical applications. Forty years of fruitful collaboration with Britton Chance. J Innov Opt Health Sci 2011; 04:337–359. 40. Bradley RS, Thorniley MS. A review of attenuation correction techniques for tissue fluorescence. J R Soc Interface 2006; 3:1–13. 41. Quistorff B, Haselgrove JC, Chance B. High spatial resolution readout of 3-D metabolic organ structure: an automated, low-temperature redox ratio-scanning instrument. Anal Biochem 1985; 148:389–400. 42. Li LZ, Xu HN, Ranji M, et al. Mitochondrial redox imaging for cancer diagnostic and therapeutic studies. J Innov Opt Health Sci 2009; 2:325–341. 43. Strandgaard S, Paulson OB. Cerebral autoregulation. Stroke 1984; 15: 413–416.