A RAT MODEL OF SMOKE INHALATION INJURY

Xie E.F., Yang Z.C., Wang D., Fu Q.F., Wei J.J.

Institute of Burn Research, Southwestern Hospital, Third Military Medical University, Chongqing 400038, People's Republic of China

SUMMARY. Using a self-designed apparatus consisting of a smoke generator and a traurnatogenic chamber, we successfully established a rat model of smoke inhalation injury and observed smoke components and the situation of animal pulmonary injuries. This model could be applied in experimental research on the pathogenesis and treatment of inhalation injury.

Introduction

Although great advances in the prevention and clinical management of burn shock and burn wound infection have enhanced the survival of severely burned patients, the prognosis of large cutaneous burns complicated by concurrent inhalation injury is still compromised. Smoke is among the most frequent factors causing complicated inhalation injuries in clinical burned patients. As the pathogenesis of smoke inhalation injury is not totally clear, and in view of the poor specific methods available for effective treatment, inhalation injury remains a major contributor to the morbidity and mortality of burn injuries and is badly in need of further study.
Inhalation injuries are usually produced by gaseous and/or particulate products of incomplete combustion. The generation of these toxic materials depends on conditions of combustion such as oxygen supply, temperature, and heating rate in the fire.' In the light of the specific condition that most inhalation injuries occur in a relatively closed space, we designed and established a rat model of smoke inhalation injury for further study of this injury on the basis of our original dog' and rabbit' models.

Patients and methods

1. Model EU-2 electric respirator (made in Shanghai, China) (Fig. 1). This connected the smoke generator (combustion chamber) with a rubber tube for delivering smoke into the traumatogenic chamber.
2. Auto-equilibrium recorder (Model SC-11, made in Sichuan, China). This recorded the temperature in the smoke generator.
3. Smoke generator. The combustion chamber was refitted from a 32 cm diameter portable pressure sterilizer. At a distance of 2 cm from the sterilizer base along the wall, two 1.5 cm diameter holes were drilled, one for pulling through electric lines, the other being a rubber tube I m in length and 1.5 cm internal diameter connected to the respirator. A screw hole 1.5 cm diameter was bored on the sterilizer cover to be fixed with a 10 em long iron pipe, which linked another rubber tube, 1.5 m long and 1,5 cm internal diameter, to reach the traumatogenic chamber. At the base of the generator, a 1200 W electric stove was installed, with an asbestos net over it.
4. Traumatogenic chamber. The smoke chamber is a specially made air-tight wooden rimless drawer with an inner area measuring 66 x 36 x 30 em. There was a glass window 19 x 16 em set into the case ceiling to observe the rat in the traurnatogenic chamber, as well as a thermometer and a sampling tube. Four 60 W incandescent light bulbs were set up on the four corners of the ceiling in order to maintain the air temperature and lighting in the traumatogenic chamber, which was interconnected with the smoke generator by a rubber tube through a 1.5 cm diameter hole drilled 2 em from the base of the case along the rear chamber wall.
5. Smoking material. Dry pine sawdust (100 g, dried for 72 h at 70 'C) was well mixed with 30 mI kerosene in a 24 em diameter stainless steel basin before use.

Fig.1 Sketch of induce smoke inhalation injury.

Forty adult healthy male Wistar rats weighing 250-300g (supplied by the Experimental Animal Center of the Sichuan Institute of Chinese Traditional Medicine) were divided into five groups. There were one control group (n = 8) and four injured groups (n = 8 in each). The animals were anaesthetized with 2% pentobarbital sodium (50 mg/kg) administered intraperitoneally, then settled in the dorsal position. An intubation was placed in the right femoral artery, while another 2 mm diameter silica gel tracheal intubation was inserted per os and well fixed perorally. All the animals in the injured groups were injured with smoke inhalation 30 min after operation, then observed and killed by bleeding respectively 2, 6, 12 and 24 h after smoke inhalation injury. Control animals underwent similar procedures except that pure air replaced smoke.
A Constantan thermistor-thennometer was used to record temperature. The electric stove in the smoke generator was connected to the power at 220 V steady voltage, and preheated for 10-20 min with the cover. When the temperature at the point 1 cm below the sterilizer edge registered 70 'C, the smoking material was placed into the combustion chamber and covered tightly for 5 min before connecting the generator with the traumatogenic chamber. The respirator was then switched on (600 ml/count and 24 counts/min) to drive air into the smoke generator and at the same time to deliver smoke into the traumatogenic chamber, where the air temperature was maintained at 38-40 'C. Five minutes later, when the chamber was filled with smoke, the drawer of the chamber was opened swiftly, and each rat was confined in the sealed traumatogenic chamber in a constant position, inhaling smoke for 2 min (reckoned by stopwatch). The rat was taken out of the chamber in order to breathe fresh air for 5 min before being re-injured with smoke in the chamber for another 2 min. The procedure was repeated three times, following the traumatogenic pattern 2-5-2-5-2 (min), ensuring that smoke inhalation lasted 6 min. During the whole procedure, the electric stove in the smoke generator and respirator were at work continuously. The arterial and tracheal incubations were removed 2 h after smoke exposure. On recovering consciousness from anaesthesia, the animals were provided with food and water ad libitum.
The smoke samples were collected from the traurnatogenic chamber 1 min before and I min after each traumatization. The concentrations of oxygen and carbon monoxide in smoke samples were measured using gas chromatography (Model SC-3, Sichuan, China) immediately after collection.
The clinical signs of injured animals were observed. Arterial blood was collected from the femoral artery respectively 5 min post-injury and prior to death. Arterial blood gases and pH were determined by a Model AVL990 blood gas analyser made in Sweden. The anterior chest wall was removed as soon as the animal was killed. In addition to observation of gross pathological changes in the lung, samples taken from each low lobe of both lungs were fixed in 10% neutral buffered fortnalin for 48 h and then embedded in paraffin, sectioned at 6 [tm, stained with haematoxylin and eosin, and studied by light microscope for micropathological changes. Total lung water JLW), extravascular lung water (EVLW) and residual pulmonary blood water (RPBW) were measured according to the method of Noble et al.
All the experimental results were presented as mean values ± standard errors. Data were compared with normal controls by analysis of variance and t-test. Results were considered significant if p < 0.05 (*) and highly significant if p < 0.01 (**).

Results

The oxygen and carbon monoxide concentrations of smoke in the smoke chamber during the whole injury process appear in Table 1. The smoke chamber washermetically closed immediately each time after the animal was put into it and/or taken out.

The clinical manifestation in a few injured animals was apnoea during smoke inhalation. Most of the injured animals showed an increased respiratory rate and hyperpnoea immediately after injury. The respiratory rate then gradually decreased and dyspnoea occurred. In the late stages, animals with smoke inhalation injury presented increased oronasal secreta, mouth breathing, and wheezing. Table II presents the survival and death rate of animals in each group.

In addition to PaC02 slightly reducing due to hyperpnoea 5 min after injury, the main findings of arterial blood gas analysis were progressively decreased pH, Pa02 and 02Cont and significantly increased PaC02 and A-aD02. The metabolic acidosis was characterized by the results of arterial blood gas analysis in the early stages after smoke inhalation injury, while metabolic acidosis combined with respiratory acidosis was marked in the later stages (Table III).

Both TLW and EVLW increased dramatically, peaking at 6 h and remaining high 24 h post-injury. No significant change in RPBW was found in the five groups, indicating that RPBW did not influence the changes of lung water content (Table IV).

Gross pathomorphological studies revealed congestion, oedema, necroses and exfoliation of the tracheal mucosa, as well as congestion, swelling, and petechial and/or patchy haemorrhages of lung tissues. Microscopically, haemorrhagic and oedematous cuffs around small blood vessels and bronchioles with aggregation and infiltration of a large number of phagoeytes were observed. The alveolar spaces were flooded with oedematous fluid and some red blood cells (Figs. 2, 3).

Fig. 2 - Six h post~injury, haemorrhage and swelling of lung tissues with aggregation and infiltration of a large number of phagocytes were seen. The alveolar spaces were flooded with oedematous fluid and red blood cells (HE x 400). Fig. 3 - Twelve h post-injury, congestion, oedema, necroses and exfoliation of the tracheal mucosa were observed (HE x 200).

Discussion

The components of smoke are complex and vary considerably, in relation to conditions of incomplete combustion, the combustion materials, temperature, circumstances and timing. The inhalation of aldehydes, especially acrolein - the major lesive substance contained in wood smoke and/or coated upon carbon particles - may cause degeneration of airway mucosal proteins, loss of cilia and the respiratory epithelium, and activation of alveolar macrophages, leading to inflammatory reactions that result in damage to the alveolar capillary membrane, an increase in its permeability, and pulmonary oedema.
We did not assay the contents of aldehydes in smoke but we determined the concentrations of oxygen and carbon monoxide for the purpose of bringing the traurnatogenic conditions and injury severity under control. It has been reported that carbon monoxide poisoning and asphyxiation from hypoxia may be the main causes of fatalities in persons suffering from smoke inhalation injury at the scene of an accident.' In the initial experiments when dry pine sawdust only was used as combustion material, the concentration of oxygen detected in smoke was only 15%, compared with that of carbon monoxide which was as high as 5%, a figure directly associated with the high mortality of animals in the traumatogenic chamber (30%) was a result of carbon monoxide poisoning and hypoxia. Kerosene, in contrast, is considered to assist combustion and reduce the concentration of carbon monoxide and aldehydes in smoke. The ratios of the concentrations of carbon monoxide, total aldehyde and acrolein in kerosene smoke, in relation to those in wood smoke, are respectively 1:12.5, 1:20 and 1:50.' We conducted our experiment with a combination of wood sawdust and kerosene as combustion material, which decreased mortality in the traumatogenic chamber below 10%. The continuous measurement conditions showed well-controlled traumatogenic conditions, as the concentrations of oxygen and carbon monoxide in smoke remained consistent, and differences remained low throughout the traumatogenic procedure. Also, the controlled temperature was maintained at 38-40 OC, preventing the thermal injury from affecting the extent of the injury. If asphyxiation developed soon after injury, a 5 ml syringe supported by artificial ventilation was helpful for resuscitation.
In the past, there have been two major two ways of establishing animal models of smoke inhalation injury, i.e., perfusion by smoke via endotracheal intubation and smoking in a traumatogenic chamber. The first is applicable in larger animals' and the second in sirnaller ones.' In a preliminary test, we initially put the rats in a specially designed cage. Conscious animals inhaled smoke freely in the traumatogenic chamber. Some animals were found to struggle and present hyperpnoea, while others showed exhaustion and breath-holding. These manifestations and others such as reflex closure of the glottis because of smoke irritation may contribute to the varying amount of smoke inhaled, the uncontrollable degree of injury, the remarkable variation in the extent of pulmonary injury, and the high mortality (30%) in the traumatogenic chamber. We therefore proposed an integration of two methods, in which endotracheal intubation was performed before the effecting of smoke inhalation injury under narcosis in the traumatogenic chamber. This integrated method kept the airways open, allowed relatively smooth respiration, and diminished the factors affecting the amount of smoke inhaled, thus ensuring stability and consistent injury.
The experimental results - clinical manifestations, arterial blood gas analysis, determination of lung water content, and the pathological examination of lung tissues - showed a number of changes in injured animals subjected to smoke inhalation according to this model. The animals quickly developed acute pulmonary oedema and serious respiratory failure; PaO- decreased progressively; PaC02 was clearly elevated; and lung water content was markedly increased. Severe inhalation injury was also attested by pathological examination. Pulmonary oedema peaked at 6 h and failed to return to control level 24 h post-injury. These changes were consistent with the onset and development of severe smoke inhalation lung injury and caused the predicted injuries. Of the 31 injured rats, considered after 12 h and 24 h respectively, 16 animals survived and 15 died 12-24 h post-injury, with a natural mortality of 48.4%, indicating that with this model mortality at 24 h accounted for a near median lethal number (LD50). The model is therefore applicable not only to the study of the pathogenesis of inhalation injury in the early stages post-injury but also to research on early experimental treatment.
In conclusion, the rat model of smoke inhalation injury established in this study satisfied three requirements for an ideal animal model, i.e., the acquisition of expected injuries, stability and easy replication, and low mortality during traumatogenic procedures. Owing to our limited facilities, we were unfortunately unable to determine and control the content of acrolein, one of the main traumatogenic agents in wood smoke - otherwise the model would be perfect.

RESUME. Les Auteurs ont projeté un appareil composé d'un générateur de fumée et d'une chambre traumatogénique pour créer un modèle murin de l'inhalation de fumée. Ils ont ainsi observé les composantes de la fumée et la situation des lésions pulmonaires dans les rats. Ce modèle pourrait être utilisé dans les recherches expérimentales sur la pathogenèse et le traitement des lésions dues à l'inhalation.

1. Clark W.R.: Smoke inhalation: diagnosis and treatment. World J. Surg., 16: 24-9, 1992.
2. Sobel J.B., Goldfarb I.W., Slater H. et al.: Inhalation injury: a decade without progress. J. Burn Care Rehabil., 13: 573-5, 1992.
3. Terrill J.B., Montgomery R.R., Reinhardt C.F.: Toxic gases from fires. Science, 200: 1343-7, 1978.
4. Zhu P.F., Jiang K.Y., Chen F.M. et a].: Establishment of a smoke inhalation model in dogs. J. Trauma (China), 3: 7-10, 1987.
5. An J., Yang A.C., You Z.Y. et al.: A model for smoke inhalation injury in rabbits. Acad. J. Third Military Med. Univ. PLA, 9: 71-3, 1987.
6. Noble W.H., Obdrzalek J., Kay J.C. et al.: A new technique for measuring pulmonary edema. J. Appl. Physiol., 34: 508-12, 1973.
7. Prien T., Traber D.L.: Toxic smoke compounds and inhalation injury: a review. Burns, 14: 451-60, 1988.
8. Levine M.S., Radford E.P.: Fire victims: medical outcomes and demographic characteristics. Am. J. Public Health, 67: 1077-80, 1977.
9. Zirkia B.A., Ferrer J.M., Floch H.F.: The chemical factors contributing to pulmonary damage in "smoke poisoning". Surgery, 71: 704-9, 1972.

Acknowledgements. The authors would like to thank Ms Sumei Lin, Assistant of the Foreign Liaison Department, MEBO International Group (USA) for participating in the preparation of the manuscript. This study was supported by grants from the General Logistics Department of the Chinese People's Liberation Army.