Inhalational and Acute Lung Injury

INTRODUCTION As the research in lung injuries has become more sophisticated, the techniques that are used have become divergent. On one hand, there has been the development of techniques for cellular and molecular biology. On the other, there has been the development of techniques to insure the credibility of investigations in intact animals. Studies of chemical systems, cells, and isolated organs provide invaluable information relating to the pathophysiology found in trauma, however, these findings must be confirmed in vivo. Many investigators consider it necessary to accomplish such studies in chronically prepared animals because acute surgical manipulation and anesthetic agents may alter the physiological status of the animal. Efforts have also been made to insure that the animals to be studied have recovered from the various operative procedures because these traumatic insults affect body function. These procedures are also preferentially performed in large animals because the physiological status of these animals can be monitored and their cardiopulmonary resuscitation can be accomplished to physiological endpoints. The removal of blood from large animals for the measurement of hematological, immunological, and biochemical responses produces minimal physiological change, and the animals can be monitored to ensure that this is the case. In addition, these animals can be monitored and compared with the clinical situation to ensure that the experimental response to injury models the human situation. Burn and smoke inhalation injury Smoke inhalation injury is a serious health threat to victims of house fires, explosions, and other disasters involving fire and smoke (1-3). The clinical symptoms and prognosis of this injury are often exacerbated by additional burn injury or bacterial infection followed by sepsis (1, 4). Adults and children suffer the direct consequences associated with the application of thermal energy to the skin, which is most commonly from an open flame or hot liquid and is less commonly caused by direct contact with hot objects, toxic chemicals, or high-voltage electricity (1, 3). In the human burn patient, inhalation injury is usually seen in flame burns. It can occur with the other situations if the toxic chemicals or the hot liquid or steam are inhaled. In addition, one aspect of lung injury associated with any form of trauma is related to systemic reperfusion (5-7). The latter will occur with inadequate or delayed resuscitation. Inadequate fluid resuscitation is a common occurrence with burn patients and is one of the major factors of mortality (8, 9). The use of animal models to study burn injury is necessary because it allows interventions that might compromise the status of a patient. In animals, we are better able to manipulate the systems to control the injuries and to use invasive techniques to better understand the pathophysiology. In addition, as the pathophysiology of the injury is better understood, therapeutic interventions can be tested in the animal model. If the proper model has been established, the efficacy as well as serious side effects of the treatment can be identified and thus, the safest and most effective therapeutic intervention can be identified for testing in the thermally injured patient. The sheep was used by most investigators evaluating the lethal combination of burn and smoke inhalation (10-15). Once the injury is established, the animals can be studied in the unanesthetized state. A total thickness burn is analgesic, and with smoke inhalation, the animals do not have any discomfort as long as their blood gases can be maintained in a relatively normal range with a ventilator. The animals easily tolerate a ventilator if it is connected through a tracheostomy. Lung Lymphatic Increased transvascular fluid flux is a hallmark of the development of acute lung injury. Since the early work of Ernest Starling (16), lymph flow has been considered a reliable index of interstitial fluid formation. This concept was reinforced by the work of Vreim et al. (17) who reported that during periods of pulmonary edema formation, the protein content of the interstitial fluid is similar to the lung lymph. Sheep are unique in that they have a single lymphatic that drains the lung (18). Multiple small afferent lymphatics drain the lung and then pass across the pulmonary ligament and enter the caudal mediastinal lymph node, which is drained by a single efferent (Fig. 1) (19, 20). Relatively pure lung lymph can be obtained by tying off the caudal end of the node and cauterizing the surface of the diaphragm (21, 22). The preparation of the chronic lung lymphatic allows the measure of pulmonary transvascular fluid flux over several weeks. There is a 10-fold increase in lung lymph flow after burn and smoke inhalation in the sheep (23, 24). There is a graded increase in lung lymph flow as well as extravascular lung water with increasing durations of smoke exposure (22).FIG. 1: Cannulation of the lung lymphatic. The figure also illustrates the afferent entering the node from the right lung across the pulmonary ligament.In his original study on lymph flow, Starling (16) developed an equation to define transvascular fluid flux. The equation was later modified by Landis and Pappenheimer (25): where Jv is the transvascular flux that under equilibrium is equal to lung lymph flow. In our laboratory, we measure lung lymph flow with a stopwatch and a graduated tube. Kf is the filtration coefficient that is a function of the permeability of water and small molecules and surface area of the pulmonary microvasculature. Ppc and Pi are the hydrostatic pressures in the pulmonary microvasculature and interstitium. We assume the latter to be zero. Sigma (σ), the reflection coefficient, is an index of pulmonary microvascular permeability to protein. When completely permeable to protein, it is 0. When completely impermeable, it is 1. Pπp and Pπi are the plasma and lymph oncotic pressures, respectively (which can be measured from the plasma and lymph from with a colloid osmometer; model 4100; Wescor, Logan, UT). We measure Ppc by the method of Holloway et al. (26). When the pulmonary arterial pressure catheter is wedged, there is a rapid and then slow fall to the wedge pressure. Pulmonary microvascular pressure is taken from the J point of the rapid and slow pressure slopes. We are able to determine this pressure from normal and injured sheep (27). After injury, the pulmonary microvascular pressure rose from 13 ± 1 mm Hg to 17 ± 1 mm Hg. Sigma (σ), the reflection coefficient, is determined by the wash-down technique developed by Taylor et al. (28, 29). In these studies, a urinary retention catheter was placed into the mitral valve to raise pulmonary microvascular pressure and thus raise the transvascular fluid flux to prevent back diffusion. The low protein concentration usually occurs within 3 h of injury. Unfortunately, although the use of a balloon catheter to produce the injury was used by most investigators (7, 30-33), we found that unless the animals were anticoagulated, they would have cerebrovascular accidents caused by small emboli to the brain. We have determined that heparinization reduces the lung injury seen with smoke inhalation (12, 34, 35). We thus developed a new technique for the determination of the reflection coefficient. We used pneumatic occluders placed on the pulmonary veins to raise microvascular pressure (36). With this technique, we determined that the reflection coefficient fell from 0.81 ± 0.02 to 0.64 ± 0.02 after injury (37). With the determination of the various parameters of the Starling-Landis equation, we could then calculate the filtration coefficient. We determined that the filtration coefficient rose from 0.020 ± 0.002 mL/min/mm Hg to 0.042 ± 0.009 mL/min/mm Hg (37). Using a paradigm developed by Taylor et al. (29), we determined that 24 h after injury, 66% of the pulmonary microvascular fluid flux was the result of increases in small and large molecular permeability, whereas 34% was the result of increases pulmonary microvascular pressure (37). In an effort to determine if the actions of smoke on the lung were direct or indirect, we set up a preparation in which the lung lymphatic only drained the left lung (38). We are also able to ventilate each lung separately using a modified double-lumen endotracheal tube and thus we smoked each lung separately. When the left lung was injured, there was a 4-fold increase in left lung lymph flow, confirming a direct injury to the lung. However, when the right lung was injured, there was a 2-fold increase in lung lymph flow to the left lung, confirming that there was also an indirect mediation of the injury (39). These data are in agreement with reports from Hales' group (10). The veins of the bronchial circulation drain into the pulmonary microvasculature (40). Therefore, we hypothesized that activated leukocytes or cytotoxins released into the bronchial circulation could be responsible for the lung parenchymal changes. Sheep have a single bronchial artery. To test the hypothesis that the bronchial circulation was playing a role in the parenchymal lung injury, we tied off the vessel, and in another group, we also sclerosed the vascular bed. In both cases, the lung injury seen with smoke inhalation was markedly reduced (41). The bronchial microvasculature is thus an important component of acute respiratory distress syndrome (ARDS) associated with smoke inhalation with or without burn injury. With inhalation injury, there is a marked increase in bronchial blood flow (42). This increase in bronchial blood flow is also associated with an increase in bronchial microvascular permeability in a canine model of inhalation injury (43). Simultaneous with these two changes, there is a loss of the ciliated columnar epithelium (44) and profound mucous secretion (43, 45). The combination exudation/transudation of fluid into the airway can have profound effects on airway resistance (46). This material clots to form obstructive airway materials (47). These so-called cast materials are filled with leukocytic infiltrates that present at a time when there is an increase in the chemokine interleukin (IL)-8 (11, 45). We have been able to develop a technique for quantitating the degree of airway obstruction (45). Moreover, we have noted that the airway contains inducible nitric oxide synthase (iNOS) and its mRNA. Thus, we have found increased levels of NO2/NO3 in lung homogenates and in the lung lymph and plasma (11, 48). In addition, we have noted increased levels of nitrotyrosine in lung tissues. This can be formed from a reaction of the hydroxyl group of tyrosine with reactive nitrogen species such as peroxynitrite (49). We were able to reverse many of the aspects of smoke and smoke and burn injury by the administration of an iNOS inhibitor (48, 50). Of interest, these iNOS inhibitors likewise prevented much of the dilation seen in untreated sheep subjected to combined burn and smoke inhalation injury. Treatment with these compounds reduced the number of obstructed airways and lowered ventilatory pressures. In our studies, we have also noted the presence of poly (ADP-ribose) in the lung tissues of the injured animals (51). Poly (ADP-ribose) is formed by the enzyme poly (ADP-ribose) polymerase (PARP). This enzyme is activated when there is a single strand break of DNA. The breakage occurs as a result of exposure to an oxidant such as peroxynitrite (52). This enzyme can produce cell injury by depletion of cellular energy stores (53). PARP can also function in a similar manner as a nuclear transcription factor because it is reported to activate NF-κB (54, 55). The enzyme is also reported to up regulate iNOS and IL-8, both of which are increased in burn and smoke inhalation (56, 57). We found that a PARP inhibitor reduced the response to burn and smoke. In a similar manner as iNOS inhibition, inhibition of PARP reduced the increase in bronchial blood flow. Figure 2 illustrates the responses to burn and inhalation injury and how this response was affected by the administration of the iNOS and PARP inhibitor, suggesting that PARP was responsible for the formation of iNOS.FIG. 2: Comparison of the effects of iNOS and PARP inhibition on the responses to burn/smoke injury. These data were obtained from two different studies of the iNOS inhibitor BBS-2 (see Ref. 50) and the PARP inhibitor INO-1001(see Ref. 51). The similarity of the actions on the variables after burn/smoke injury is interesting. The inhibitors were administered 1 h after injury. Values are mean ± SEM.We are in the process of performing studies in which we anticoagulants and fibrinolytic agents. We have thus far tested three compounds. Models of sepsis There have been several models of sepsis that have worked within the sheep: single bolus of endotoxin or bacteria (58-60), continuous infusion of endotoxin or bacteria (61, 62), and cecal ligation and puncture (63). In practically every case, there are several major points to make. The sheep is quite sensitive to endotoxin, on the same order as humans. The administration of endotoxin or bacteria is associated with an increased lung lymph flow and lung water, wet/dry, etc. In the situation in which there is a continuous infusion of endotoxin or bacteria and a steady state, the arterial pressure is reduced, the cardiac output is elevated but at a plateau, and we determined that the pulmonary microvascular reflection coefficient was normal. The increased lung lymph flow is sustained by in elevated filtration coefficient and increased microvascular pressure (62). Unfortunately, these models do not show acute lung injury by its definition of a PaO2/FiO2 ratio of less than 300 (64). There is a increased shunt blood flow in the animal, but the mixed venous oxygen tension is very high, so that the PaO2 is maintained in the normal range (65). More recently, we have developed a pneumonia model of sepsis in which the sheep reach the criteria of ARDS within 12 h (66). In this model, the animal is given an inhalation injury and then Pseudomonas aeruginosa is placed in the lungs. This model has been used to test some of the drugs that were useful in our ARDS model. We found that a PARP inhibitor was effective in attenuating a substantial amount of the lung injury seen with pneumonia (67). On the other hand, although nonspecific NOS inhibitors were effective in reducing the lung injury seen in this model, specific iNOS had little effect on the response (68). In this case, a neuronal NOS (nNOS) inhibitor seemed to be more effective than the iNOS inhibition (68). Actually and surprisingly, treatment with recombinant antithrombin III and heparin seemed to not only be effective in reducing airway cast formation, but also with some of the lung inflammation that was seen in the pneumonia model (68-70). Creation of the burn/smoke inhalation model The experimental methodology for producing burn and smoke inhalation injuries varies (71). Most investigators induce burn injury with a flame. The body surface of the animal can be determined from its size and weight (body surface in two-thirds function of weight) times a factor dependent on the species (72). In the case of sheep, the body surface area = 0.084 × body weight2/3. The area to be burned is then measured and marked. In sheep and other large animals, this injury is usually performed with a Bunsen burner (73). As the flame is applied to the skin, it contracts. At this point, a third degree (full thickness) burn is obtained. The fire is then moved to a new area of skin. After completion of the burning procedure, the skin is observed and those areas that are pink in color are considered to be second degree. These are given additional injury. In this way, a standardized injury can be induced. In our institute, animals are deeply anesthetized and placed on a ventilator, with positive end-expiratory pressure set to 5 cmH2O, tidal volume at 15 mL/kg, and frequency and FiO2 adjusted to keep PaO2 between 80 and 120 mm Hg and PaCO2 <40 mm Hg. After shaving the wool, a 20% total body surface area third degree flame burn is made on one side and back of the sheep. The burn is produced with a Bunsen burner flame until the skin is thoroughly contracted. We have previously determined this degree of injury to be a full-thickness burn including epidermis and dermis in which the nerve endings are heat-destroyed (74). Thereafter, inhalation injury is induced with a modified bee smoker. The bee smoker is filled with 40 g of burning cotton toweling and is attached to the tracheostomy tube via a modified endotracheal tube containing a thermistor from Swan Ganz catheter. Four sets of 12 breaths of smoke (a total of 48) are delivered and the carboxy hemoglobin level is determined immediately after smoke inhalation to proof grade of carbon dioxide intoxication. The temperature of the smoke is not allowed to exceed 40°C during the smoking procedure. A more detailed description of the method for the smoke insufflation procedure, an analysis of the smoke has previously been published (22). After smoke insufflation, another 20% total body surface area third degree can be made on the remainder of the animal's flank. After the injury, the animals are resuscitated with Ringer's lactate solution following the Parkland formula (4 mL/kg/% burned body surface area/24 h) (75). One-half of the fluid is given during the first 8 h after injury and the remainder during the next 16 h. ANIMAL BEARING AND PERIOPERATIVE MANAGEMENT Most of the animals used in research are yearling males that have been castrated or older females that have been culled from the flock. Reproducing ewes are the main variety used in animal husbandry; the placenta may carry Q fever, a highly contagious microbiologic vector. Young males (about 20 kg) are somewhat smaller than females and need less of space for housing, but are also more rambunctious and noisier than older females (about 40 kg). The latter are docile and in and are very when not in the presence of another sheep. the sheep are placed into (71). The animals, the older are content in these as long as there are other sheep in the and they have and The animals are easily with and is important because it If sheep do not they and will form in their This can occur and is to as If it the animals have to be but not water be may with the but the of the sheep is such that if the animals do not water will flow into the of water can to a in and The sheep is to have a similar to as do humans. in these animals is a anesthetic for and for do not to have with this as do humans. agents used in are to be oxide into the and can the are long in the sheep and be because agents such as may the animal for If the studies are to be performed in the anesthetized state, the animals must be placed on a ventilator. control of is very sensitive to as are humans. the animals be placed in the to the weight of the from the and the If the sheep are to be studied in the unanesthetized state, with and with an such as or be This anesthetic technique will result in an animal within 1 or 2 h of the operative procedure, on the of the When large operative procedures are it is to the with a anesthetic at the end of to and is if the animals their to be up and or to and of the is a very effective also in sheep. As is the case in most animals, the body temperature must be It be noted that the body temperature of most animals is between and It is also important to be in fluid This is during the time periods when body are The sheep models of systemic inflammation are in important into the pathophysiology of acute lung injury and and to the study of treatment in may as the for clinical