Dmitrienko O.D.(1), Golimbievskaya TA.(2), Trofimova T.N., Kossovoy A.L.

St Petersburg Medical Academy of Postgraduate Studies, St Petersburg, Russia
(I) Department of Radiology;
(2) Department of Thermal Lesions

SUMMARY. This paper considers minimal burns with a potentially favourable prognosis, the effects of which were much aggravated by the presence of inhalation trauma. In the group of patients considered, 17 fatalities were expected, against an actual number of 27. Radiological investigations provided a considerable amount of interesting information, as the pulmonary complications could be predicted by evaluation of the dynamics of the radiological changes. This facilitated early intensive therapy.

Introduction

Our research shows that even in cases of minimal burn area the presence of inhalation trauma played a highly negative role in the final outcome (71% lethal). In the cases described in this paper, a lethal outcome was expected on the basis of the prognosis in 17 cases, while in reality 27 deaths were recorded. Radiological investigation of the patients' early stage proved to be highly informative. Pathological pulmonary changes appeared as early as 24 h post-injury. The earliest indications of the development of a pathological process in the lung were an increase and a deformation in the pulmonary pattern in roof segments (central localization) as a result of vascular hyperaemia, caused by hypertransfusion. This is typical in haemodynamic pulmonary oedema. The localization of changes in pulmonary pattern in the periphery of the pulmonary fields was typical of respira~ tory distress syndromes, as a result of haemodynamic violation of the pulmonary capillaries.
The most characteristic form of lesion in fires and burn accidents is a skin burn of various type combined with inhalation trauma (IT). Thermochemical affections of the respiratory system and poisoning due to combustion products remain the primary causes of early death among fire victims. Early diagnosis of damage to the respiratory tract is difficult - it is not possible to determine the degree of damage and thus establish a proper prognosis. Current investigative methods of IT diagnostics, e.g. bronchoscopy, pulmonary scanning using isotope 133x,, and blood gas analysis, are not usually implemented in burn centres in country areas owing to the lack of necessary equipment. Also, implementation of these methods in the event of the mass admission of burn victims would be extremely complicated. The role of radiological methods of the diagnostics of pulmonary complications in fire victims is thus very important.
As long ago as 1947, Schatrkir noted a wide variety of radiological changes in the victims of a fire disaster in a Boston night club. The role and the optimal timing of the radiological analysis of IT-affected lungs remain controversial.Dynamic radiological analysis of pulmonary complications in the pre-clinical period provides a possible line of conduct for early intensive therapy in the presence of such complications .3,4 Most problems arise in diagnos~ tics and in the differential diagnostics of pulmonary corn~ plications such as the respiratory distress syndrome (RDS) in adults. The object of this paper is to study the radiological pattern of critical diffusive lesions of the vascular lung channel in fire victims and the development of criteria for differential diagnostics of RDS and haemodynamic pulmonary oedema (HPO).

Materials and methods

A complex clinical, laboratory, radiological and forensic medical analysis was made of 31 fire victims with both average and high-degree inhalation damage (Table I).
Inhalation trauma was diagnosed in all 31 victims on the basis of anamnesis, facial burns, and changes in the mucous membrane of the nose, mouth and pharynx, and also on the basis of clinical symptoms (unconsciousness, cough with saturated phlegm secretion, cyanosis) and laboratory tests (carboxyhaemoglobin and blood gas level).
More than 10% of the patients suffering from deep burns received anti-shock therapy (colloids) and protein transfusion (3 to 8 litres) in the first three days. Medical therapy (heart preparations, analgesics, heparin, Lasix, corticosteroids and vitamins) was also performed.

Medication was introduced through catheters in the upper and lower cava veins (five patients also received rnedication through the aorta). Hyperbaric oxygenation (through a microtracheostoma) and artificial lung ventilation (usually 1-2 days before death) were administered during treatment.
The radiological symptoms of severe diffuse lesions of the pulmonary vascular channel were compared with data obtained from the study of 113 patients with other primary pathologies leading to pulmonary complications, such as:

Experimental data were obtained through the study of 33 dogs. The experiments were conducted under conditions of hexenal i.v. anaesthesia (20-30 mg/kg mass) and artificial ventilation of the lungs. The model for toxic pulmonary oedema was developed by i.v. injection of silver nitrate at a dose of 1.2 mg/kg-1. Mixed pulmonary oedema was simulated by pressing the dogs' kidney vessels, the flow through which was re-established after three hours' exposure. Haemodynamic oedema was created by raising arterial and venous perfusion pressure in lungs ventilated and perfused through the pulmonary artery. Sodium chloride was added to reduce colloid osmotic pressure. The pulmonary conditions were monitored by radiological, histological and ultrastructural methods, and also on the basis of the gas content and acid-base balance of the blood, haemodynamic data, and gravimetric results.
All the patients underwent chest radiography on days 1, 3, 5 and 7 days post-injury.

Our research shows that even in cases of minimal burn area (potentially favourable prognosis) the presence of IT played a highly negative role in the final outcome (71% lethal). On the basis of the prognosis, a lethal outcome was to be expected in 17 cases, while in reality there were 27 fatalities.
RDS was detected radiologically in 18.7% of the burned patients with IT. In 50% of the cases different stages of cardiac pulmonary oedema (interstitial, alveolar) were found. In 21.8% of the cases severe pneumonia was observed. A combination of RDS and HPO was detected in 74.2% of the cases, of RDS and severe pneumonia in 12.9%, and of HPO and severe pneumonia in 12.9% (Table II).

In severe pneumonia, focus infiltration of high intensity with fuzzy contours was detected with reactive changes of the roots, and increased pulmonary vascular pattern. There was no correlation between the development of pneumonia and area of deep skin lesions - 25% of patients affected had developed pneumonia by day I post-trauma, 44% by day 3, and 30% by day 5. Changes due to pneumonia were detected radiologically not earlier than days 23 post-burn. A bilateral process in the basal segments was detected more often.
With the development of RDS, a strengthening of the pulmonary vascular pattern with fuzzy contours and network deformation due to the vascular component was detected (Table III). In 54% of the cases enlargement and fuzziness of the bronchial walls were detected.

Nonsignificant malfunctioning in the haemodynamics of the pre-capillary part of the pulmonary network of the blood circulation was noted. This was expressed by enlargement of the diameter of the artery of the third segment (up to 5.5 mm) and by widening of the descending branch of the right pulmonary artery (up to 18 mm). All subsequent stages of the pathological process in the lungs were characterized by the appearance of focal infiltrative shadows. A typical observation in the majority of cases was the localization of the changes on the periphery of the pulmonary fields as well as in the gravitation-dependent segments (60%) (Figs. 1-3). The appearance of focal and infiltrative shadows was followed by the breach of the pulmonary root structure in 93% of the observations and by the appearance of subpleural oedema along the horizontal interlobular border (36%). There was an increase in the changes in the bronchial walls.
The diminishing of the process was characterized by a decrease in the quantity and intensity of focal and infiltrative changes in 13 patients. These patients had dominant discrete shadows that started to disappear on days 20-24.

Fig. 1 - Alveolar pulmonary oedema. Fig. 2 - Respiratory distress syndrome. Fig. 3 - Acute pneumonia and alveolar pulmonary oedema

The multiplicity of the causes of RDS gives the radiological changes specific characteristics. Thus, in patients with acute surgical abdominal pathology, the radiological signs - even in the early stages of RDS - are noncharacteristic disc collapses, pleural effusion, and high diaphragm position. This indicates a catastrophe in the abdominal cavity and makes it difficult to evaluate the changes. In cases of mixed trauma the radiological interpretation become more difficult owing to various traumatic radiological characteristics, such as emphysema, fractures, pneumothorax, hydrothorax, etc. Differential diagnostics of the pulmonary injury and RDS is not so difficult because of the local nature of the traumatic changes. In patients with pulmonary complications due to blood loss and massive haemotransfusions, radiological findings show evidence of increasing volume blood circulation and the development of right ventricle deficiency (significant increase in size of right atrium). These features are not common in cases of RDS. A similar picture was detected in burn patients.
The radiological pulmonary picture in burned patients with haemodynamic pulmonary oedema was different from that seen in RDS patients. Results of a comparative analysis of the radiological picture in cases of RDS and of HPO are summarized in Table IV.

Changes in the lung roots were the earliest signs of interstitial pulmonary oedema of haemodynamic origin. In cases of RDS these were unchanged in almost 28% of the cases. An analysis of the pulmonary picture showed that venous hypertension was typical of interstitial pulmonary oedema, but not of RDS. Kerley lines were detected only in cases of interstitial pulmonary oedema.
Progressive haemodynamic damage led to alveolar pulmonary oedema. The most pronounced changes were detected in near-roof pulmonary segments. The pulmonary roots were markedly changed and undifferentiated in 12.5% of the cases. Total or infiltrative change with large shadows was characteristic of haemodynamic pulmonary oedema. Liquid evacuation from mobile parts of the lung caused the appearance of a band-like pneumatized parenchyma above the diaphragm. This symptom was observed mostly in patients with haemodynamic pulmonary oedema (30%). The picture of air bronchography was not typical in these conditions. In haemodynamic pulmonary oedema there was a faster dynamic in cases of progressive change as also in cases of resolution of the process after adequate intensive therapy.
The data obtained from animal models and clinical observations support the difference in radiological expression of haemodynamic pulmonary oedema and RDS.
C-ray investigations in experimental toxic oedema showed that vasodilatation and an increased number of focal shadows in cases of hyperhydration and hypoxaemia were the earliest signs of pulmonary vessel permeability lesions.
Optical image post-processing made it possible to perform a detailed analysis of the radiological picture (Table V). RDS was characterized by diffuse strengthening and deformation of the pulmonary picture and vascular pattern focal shadows, which were localized in the periphery of the pulmonary fields. The presence of large unstructured zones correlated with the focus of the oedema, with central localization, was typical of haemodynamic pulmonary oedema. A similar but relatively localized picture was observed in cases of pneumonia. In general, implementation of optical image post-processing made it possible to increase precision of diagnostics to 90.4%, specificity to 87.5%, and sensitivity to 92.4%.

The results of the research are summarized in the proposed differential diagnostic algorithm. Effective differential diagnostics is only possible if based on the stages of process of the development in the lungs. Thus, two large groups can be highlighted on the basis of the presence or absence of focal and infiltrative changes in the lungs. Diagnostics is based on the detection of haemodynamic malfunctioning or on increasing permeability of the pulmonary capillaries. These become the major differential diagnostic criteria in cases of absence of focal infiltrative pulmonary changes. Kerley lines of the pulmonary roots and heart signs are additional criteria.

Conclusions

1. Radiological investigation of patients in the early stages post-injury proved to be highly informative.
   Pathological pulmonary changes appeared as early as 24 h post-injury. The earliest indication of the development of the pathological process in the lungs was the increase and deformation of the pulmonary pattern in root segments as a result of vascular hyperaemia, which was caused by hypertransfusion.
2. Diagnostics of pulmonary complications (e.g., RDS, cardial pulmonary oedema, and severe pneumonia in the pre-clinical period) was possible through evaluation of the dynamics of the radiological changes. This allowed early adequate intensive therapy.
3. Respiratory distress syndrome and haemodynamic pulmonary oedema detected in burn patients with IT did not have specific diagnostic features in this group of patients compared with other patients with pulmonary complications. Most of the difficulties arose in the diagnostics and differential diagnostics of RDS.
4. Comparative analysis of clinical cases and data from animal experimental studies showed the difference in radiological features, which depended on the type of injury to the vascular pulmonary channel. This difference permitted differential diagnostics between haemodynamic pulmonary oedema, RDS, and pneumonia. RDS was characterized by diffuse strengthening and deformation of the pulmonary vascular pattern and focal changes mostly located in the peripheral pulmonary fields. In cases of HPO the presence of large central unstructured zones was typical.

RESUME. Les Auteurs considèrent les brûlures légères avec un prognostic potentiellement favorable, dont les effets ont été notamment aggravés par la présence de traumatismes causés par le feu. Dans le groupe de patients pris en examen, les dix-sept fatalités prévues ont été en réalité 27. Les investigations radiologiques ont fourni beaucoup d'informations très importantes, parce qu'il était possible de prédire les complications polmonaires en évaluant l'évolution de la dynamique des modifications radiologiques. En cette manière, les premières phases de la thérapie intensive ont été facilitées.

1. Pruitt B.A., Flemma R.J., Di Vincenti F.C., Foley F.D., Mason A.D.: Pulmonary complications in burn patients: a comparative study of 697 patients. J. Thorac. Cardiovasc. Surg., 59: 7-20, 1970.
2. Schatzki R.: Roentgenologic report of the pulmonary lesions. Ann. Surg., 117: 841-64, 1947.
3. Pruitt B.A., Erickson DR., Morris A.: Progressive pulmonary insufficiency and other pulmonary complications of thermal injury. J. Trauma, 15: 369-79, 1975.
4. Chi-Shing Chu: New concepts of pulmonary burn injury. J. Trauma, 21: 958-61, 1981.
5. Horovitz, J.H.: Diagnostic tools for use in smoke inhalation. J. Trauma, 21: 717-9, 1981.
6. Munster A.M.: The early management of thermal burns. Surgery,88: 29-40, 1980.
7. Teixidor H.S., Rubin E., Novic G.S., Alonso D.R.: Smoke inhalation: radiologic manifestations. Radiology, 149: 383-7, 1983.
8. Demling R.H.: Burns. N. Eng. J. Med., 313: 1389-98, 1985.
9. Edward F., Haponik M.D. et al.: Increased vascular pedical width preceding burn-related pulmonary edema. Chest, 90: 681-5, 1986.
10. Herndom D.N. et al.: Pulmonary injury in burned patients. Surg.Clin. N. Amer., 67: 31-46, 1987.