U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content







AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window






Chapter XI




If the effects of phosgene on the animal organism were to be described in one phase it would be by the words "pulmonary edema." A study of phosgene poisoning is, therefore, of more than passing medical interest. The work here reported was done in an attempt to get a physiological background for proposed forms of treatment. Various physiological reactions have been followed as closely as possible throughout the course of fatal phosgene poisoning.

The general results of poisoning by lung-irritant gases, including phosgene, have been described in the preceding chapter, and since the experience on this point here reported is identical, such description need not be repeated. The division of phosgene poisoning into three stages (see pp. 351-354) is in accord also with these findings and, although arbitrary in certain cases, as all such classification must be, it is of distinct advantage in locating the various physiological disturbances and in gaining an insight into the condition as a whole.


Dogs were used throughout these studies. The animals were subjected for 30 minutes to air containing 80 to 100 parts per million of phosgene. This was sufficient, with rare exceptions, to produce death in the first 24 hours. The technique of this gas administration was one gradually evolved at the various Chemical Warfare Service laboratories. It consisted of placing the dogs in a 100-liter air-tight glass box through which air was drawn at the rate of 100 liters per minute. The phosgene cylinder was connected to the air inlet tube, the gas outflow being regulated by a needle valve and the rate roughly determined by a flow meter. From the gassing chamber itself air was drawn in order to determine by chemical analysis the exact concentration of phosgene to which the animals were subjected.

All animals on which it was necessary to make incisions were morphinized either before or immediately after being gassed. A series of morphine controls had been carefully studied, and care was taken that none of the effects to be described could be attributed to morphine. In order to make observations rapidly and frequently, the animals were kept tied to operating boards. Since no pressure was exerted anywhere except by the cords on the limbs, and the respiratory passages were entirely unobstructed, this restraint seemed unobjectionable. The animals lay quietly and comfortably until the usual asphyxial stimulations occurred shortly before death.


By attaching a mercury manometer to the femoral artery, arterial blood pressure records were made in the usual way at half-hour intervals. The general course of the blood pressure in a typical case of phosgene poisoning

a The data in this chapter are based, in the main, on the experimental observations made by the detachment of the Medical Division of the Chemical Warfare Service on duty at the University of Wisconsin laboratory, account of which
was published by Maj. W. J. Meek, C. W. S., and Lieut. Col. J. A. E. Eyster, M. C.: Experiments on the Pathological Physiology of Acute Phosgene Poisoning. American Journal of Physiology, Baltimore, Md., 1920, ii, No. 2, 303.


be seen at a glance by referring to Chart XIX. This composite curve, in common with those that are to follow, was made by dividing each of the experiments in the series into 10 equal periods. The animals lived an average of 16 hours after gassing. Each of the 10 intervals, therefore, represents on the average a little over one hour and a half. The data for the same period in all the experiments were averaged and the results plotted as a composite curve.

A few animals showed a slight fall of blood pressure after being taken from the gassing chamber. In most cases this was insignificant, and it did not lower the composite curve during the first period. As a rule the blood pressure gradually rose during the first half of the experiment, increasing some 10 percent above the normal. Beginning with the sixth period, it began to fall slowly,

CHART XIX.- Composite curve, from ten experiments, of the changes in arterial block pressure after acute phosgene gassing.

reaching normal at the beginning of the eighth interval. Once having passed the normal the decline became extremely rapid and continued without intermission until the death of the animal. This break in blood pressure, which occurred at the eighth period, was an extremely striking event in all the animals studied. It made possible a very accurate prediction as to how much longer the animal would survive. Until its significance was fully appreciated, many of the animals died before final observations could be made.

Chart XX is a reproduction of the actual records from a typical experiment. The points just, mentioned may be noted. The first record, at 11.10 a. m., was taken shortly after the gassing. Tie pressure then rose gradually for more than six hours. At 12.10 a. m., 13 hours after gassing, the pressure was still normal, though falling. One hour and twenty minutes later the animal was dead.


One need not attempt to interpret the arterial blood-pressure curve of acute phosgene poisoning until all the other data have been presented. Its resemblance to an asphyxial vasoconstriction, however, is obvious. That the rise may have been due to vasoconstriction and the fall in part to paralysis of the vasomotor center is suggested by the large vasomotor waves which so often occurred late in the blood-pressure tracings. (See Chart XX, the record at 12.45 a. m.)


In the preceding series of 10 experiments venous pressure observations were made simultaneously with the arterial. These were secured by inserting a sound into the femoral vein. The sound was connected to a manometer and a reservoir of Ringer's solution. The pressure in the system was raised above what the reading was likely to be. On removing a clip the blood pressure was balanced against that of the fluid in the manometer. The reading could be made before there was any tendency to clot, and the fluid added to the bloodstream at each observation was negligible.

CHART XX.- Arterial blood pressure record from a case of acute phosgene poisoning

Venous pressure was found to be rather variable, conforming, on the whole, however, to what one might expect from the arterial. During the long period of increased arterial pressure, venous pressure was either normal or slightly below. In the terminal stages, however, it often rose markedly. In 2 of the 10 experiments there was a noticeable increase in venous pressure immediately after gassing. These were very severe cases, death occurring within nine hours. It would seem probable that in these animals the initial injury to the lungs was so great that the pulmonary circulation was obstructed and venous pressure therefore forced to rise.


Shortly after gassing the pulse rate fell in practically all cases. This occurred in animals morphined before gassing, as, well as in those that received none of the drug. It was therefore an expression of the action of the poison itself. The decrease in rate brought the pulse from an average of 95 to 70 beats per


minute. By the time the experiment was half over the heart rate had returned to normal, and following this it was very markedly accelerated. Chart XXI presents a composite curve from 14 experiments in which the pulse rate was carefully followed. The final determination of the curve is the average of the highest rates obtained in the tenth period. As death became imminent, the heart rate became irregular and the rate then, of course, decreased. Electrocardiograms taken at this time showed various kinds of blocks, dropped beats and extra systoles, features which characterize most records taken during death by asphyxiation.

The pulse rate offered an excellent means of following the condition of the animal. Two of the dogs in one series recovered. The pulse curves from these

CHART XXI.- Composite curve from fourteen experiments, of the changes in the heart rate after acute phosgene gassing

animals were of special interest. There was the initial fall and the subsequent rise, but the latter never exceeded 120 beats per minute. A large number of observations have confirmed the opinion that a fatal outcome is to be expected if the heart rate continues to rise above 125 or 130 beats.


It was soon realized that there were very significant changes in the hemoglobin content of the blood in dogs suffering from phosgene poisoning. At death the blood was viscous, even tarry in consistency, and the hemoglobin readings showed an almost unbelievable concentration. Underhill 1 first pointed out that this stage of concentration was preceded by an initial one in which the hemoglobin content of the blood was decreased.


The characteristic hemoglobin changes during the course of the poisoning may be seen in Chart XXII. In an average experiment the hemoglobin readings were below normal during the first four periods of the experiment. This constitutes Underhill's first stage of phosgene poisoning. Concentration then began, and in the period preceding death the hemoglobin readings average over 150 percent. The periods of concentration make up Underhill's second stage.

The maximum decrease in hemoglobin occurred anywhere during the first four periods; in other words, during the first five or six hours of the experiment. Since the maximum decrease did not occur in the same period for all the experiments, the composite curve does not show the lowest limit reached in hemoglobin concentration. In the 16 cases reported, it actually averaged

CHART XXII.- Composite curve from sixteen experiments, of the changes in hemoglobin concentration after acute phosgene gassing

11 percent; that is, a hemoglobin reading of 89. The lowest reading noted was 85, although in a treatment series not here reported there were readings as low as 80 and one of 78. The maximum decrease lasted a very brief time and often there was difficulty in making desired observations at exactly the proper moment.

The degree of subsequent hemoglobin concentration always bore a definite relation to the severity of the poisoning and made possible a rather accurate prognosis. One hundred and twenty-five per cent may be said to represent a critical point. Any animal exceeding this figure was pretty sure to die in the course of the next five or six hours. If an animal was not gassed enough to reach this concentration, or if by any means a 125 percent concentration could be prevented, there was an excellent chance for recovery.


Hemoglobin readings are usually interpreted in terms of the fluid content of the blood, or blood volume. If plasma or water has left the blood stream there is of course a concentration of hemoglobin, and if fluid has entered the blood stream from the tissues or elsewhere there is naturally a lowered hemoglobin content. While this is the general rule, there may be exceptions. Hemolysis or stagnation of red blood cells at any point might very greatly modify hemoglobin determinations and yet the blood volume would be entirely unchanged.

The natural interpretation of the hemoglobin curve for phosgene poisoning would be that in stage 1 there is an increase of blood volume and in stage 2 a marked decrease. That there is a real decrease in stage 2 is borne out by the fact that the lungs are now full of fluid which must, of course, have come from the blood. There is, however, no equally obvious explanation for an increase of blood volume in stage 1. Underhill 1 also found in this stage a decrease in the blood chlorides, but the excess at the time in the urine and fluid of the lungs might account for this decrease.


The blood volume was determined directly in eight animals during stage 1 and in three animals during stage 2. The technique used was the acacia method 2 which has been developed in the laboratory of the University of Wisconsin. In Table 26 may be seen the results.

TABLE 26.- Blood volume in phosgene poisoning

The data presented give no evidence of a blood volume increase in stage I. In a large series of normal dogs the volume was frequently 10 and 11 percent of body weight, with an average of 9.7 percent. The eight animals here investigated averaged, then, within 2 percent of normal. Furthermore, of the three animals having the greatest hemoglobin dilutions, only one had a volume above the average.


In stage 2 only three determinations were made; unfortunately no hemoglobin readings were made simultaneously. Red blood-cell counts, which always paralleled the hemoglobin readings, indicated, however, that the latter were in the neighborhood of 135 percent. Since the technique required the removal of 20 c.c. of blood, the determinations were made some time before the anticipated death of the animal. The figures, particularly the first and third, show that these animals had a marked decrease in blood volume. Even 8.1 percent body weight is a lower blood volume than we have ever found in a normal dog.

Stage 1 of phosgene poisoning, on the grounds of blood-volume data, is believed to represent an actual decrease in the total hemoglobin content of the blood, a point which will be discussed later under histological examination

CHART XXIII.- Composite curve, from nine experiments, of the changes in the red blood cell counts after acute phosgene

of the lungs. In stage 2 there can be no doubt that the blood volume is actually greatly decreased.


Little need be said concerning the red blood-cell counts other than that they uniformly paralleled the hemoglobin determinations. A composite curve from nine experiments may be seen in Chart XXIII. As in the hemoglobin curve, stages 1 and 2 are evident, and they occupy the same relative positions.


Careful histological examinations of lung tissue were made in a series of poisoned dogs b Special methods for fixing and staining pulmonary tissues.

b These examinations were made by Dr. W.S.Miller, of the Department of Anatomy, University of Wisconsin.


were used, rather than the routine technique of general pathology. These examinations showed that the injury from phosgene was almost exclusively in the lower respiratory passages. There was constriction or spasm of the small bronchioles, with the accompanying atelectasis and emphysema, and edema of the connective tissue. The alveoli were irregular, their membranes were injured, and in many cases they contained exudate. Very important from the point of view of these studies was the extensive clogging of the capillaries with red blood cells. Even small veins were solidly plugged. In many cases these masses in the veins had shrunk slightly and the surrounding clear areas were filled with serum. That these changes in the lungs were not post-mortem is substantiated by the fact that they were characteristic of all the early stages of poisoning and that they were found after every attempt to avoid post-mortem clot

Furthermore, if phosgene in a dilution as great as 1 to 20,000-that is, 0.222 milligrams per liter-was bubbled through a 2 percent suspension of defibrinated dog's blood, there was in 20 minutes a marked agglomeration of the red corpuscles. This is direct evidence that the gas has the power of doing just what the histological picture shows.

In slightly later stages than the one figured, there was evidence that compensatory paths were being opened up for the blood stream. Capillaries which were not plugged were widened and others had been dilated sufficiently to allow a flow of fluid around the obstructions.

The importance of these histological findings on the physiological conception of phosgene poisoning is at once apparent. The plugged capillaries and veins at first must have greatly increased pulmonary resistance and the work of the right heart. Later there was relief by the development of compensatory passages. Just how the heart reacted to this will be seen in the following section. Furthermore the injuries to the alveolar walls must have decreased the exchange of gases between the blood and the alveolar air.


In a large series of experiments, stereoscopic examinations and X-ray photographs of the thorax were made at frequent intervals. At first the outline of the heart was sketched in with a grease pencil on the glass cover of the fluoroscope. Later stereoscopic plates were made and these, of course, proved much more reliable than the former method. To be sure that the animal was in the same position for each photograph, a lead cross was sewed to the chest and the center of this brought under a plumb bob at each exposure. Exposures were always made during the same phase of respiration, preferably inspiration, and the flashes were long enough to insure that it was the diastolic size of the heart which was secured.

These observations showed two interesting and important changes in heart size during acute phosgene poisoning. There was, first, immediately after removal from the gas chamber, an increase in the size of the heart which varied considerably in degree, but which was always associated with a relative enlargement of the right auricle and ventricle. This condition persisted for several hours and might even increase for an hour or more.


By the beginning of the fourth period, assuming that the experiment had been divided into 10 equal intervals, a second change appeared which was a gradual reduction in heart size. This seemed to appear first in the left ventricle, but soon the whole heart became distinctly smaller. It assumed a pendular shape which was apparently identical with that following severe hemorrhage, as determined in control experiments. This decrease in size continued during the development of the extensive pulmonary edema. During the period of asphyxial death the heart began to enlarge, particularly the right side, and after the death plates invariably showed a dilated heart with relative increase of the right side.

FIG. 45.- Superimposed outlines of three X-ray photographs taken at intervals during phosgene poisoning to show changes in shape of the heart. The solid line indicates the normal. The dotted line shows the right-sided dilatation 39 minutes after gassing. The broken line is from a photograph taken 11hours and 53 minutes after exposure to the gas. The heart had then become pendular in shape and much reduced in size

Figure 45 illustrates strikingly the stages of increase and decrease in heart size. The area of the normal heart shadow as determined by the planimeter was 51.2 sq. cm. Thirty-nine minutes after gassing, the area had increased to 55.4 sq. cm. Eleven hours and fifty-three minutes after exposure the heart had decreased to 44.5 sq. cm. This was at a time when the pulmonary edema had become very marked.


The condition of the lungs was determined in a large number of animals by means of physical, fluoroscopic, and stereoscopic X-ray examinations. Immediately after ramoval from the gas clwamber there was noted a diffuse, clouding of the hings, usually most marked in the middle and upper lobes, but


sometimes involving the lower lobes also. This was accompanied by an increase in density and number of the streaky shadows cast by the larger bronchi and vessels at the roots of the lungs. For a few hours the cloudy appearance of the lungs generally increased slowly without being associated, however, with any constant physical signs. Occasionally transitory fine, dry, crackling rales were heard during expiration, and occasionally there was a slight roughening of the normal respiratory sounds.

After these few hours the lungs often appeared somewhat clearer. This improvement, however, was transient and gave way during the latter third of the experiment to a streaky, mottled appearance, at first marked near the roots of the lungs but later involving the whole of both lungs. Often the heart outline was in part lost or obscured. It was at this time that numerous rales of all varieties made their appearance, medium moist predominating. These were heard in both inspiration and expiration. In brief, the clinical signs were now those of extensive pulmonary edema and passive congestion.

In the later stages there was frequently an impairment of the percussion notes, most evident in the pendant portions, and an extension of deep cardiac dullness, especially on the right side. X-ray plates showed, however, that this was not due to cardiac enlargement but probably to better transmission of heart dullness by the edematous lung.

There were thus three more or less distinct changes to be made out by means of the X ray in the lungs of animals fatally poisoned with phosgene. First a diffuse cloudiness due probably to the initial epithelial injury and to the agglomeration of the corpuscles in the capillaries. Second, an improvement, a decrease in the cloudiness, accounted for by a reopening of many of the capillary passages. Third, a marked increase in the density and extent of all shadows cast by bronchi and blood vessels and an extension of the mottled appearance to all parts of the lungs. It is worthy of mention that it was only during the third stage that the classical clinical signs of pulmonary edema developed. The X ray proved a much more delicate method of following lung lesions than the older methods of percussion and auscultation.


One of the earliest results of phosgene gassing was an increase in the respiratory rate. By the end of the fourth period, that is stage 1, the rate had usually increased from an average of 30 to 45 per minute. This increase continued through stage 2 until the death period itself, when the respirations, of course, became irregular and gradually less rapid.

Respiratory rate, particularly in the dog where there may be much panting, gives a poor idea of the amount of air actually passing in and out of the lungs. To secure such data a series of five experiments was run, in which the dogs were placed in an air-tight rigid chamber that inclosed the entire animal except the head. An inflated rubber collar secured an air-tight fit around the neck. The box was connected by tubing to a piston recorder which not only made a record of the respiratory rate but on calibration gave an accurate measure of the air passing in and out of the lungs.

The data thus secured, as shown in Table 27, indicate in all cases a final marked increase in pulmonary aeration. In all cases the amount of air respired was at least doubled in the latter periods of the poisoning. Even after the


break in arterial pressure when respirations often slowed down, the increase in aeration was maintained. In three cases the amount respired immediately after gassing was definitely lower than normal.

TABLE 27.- Pulmonary aeration during phosgene poisoning


The only constant change in temperature was a gradual fall as blood concentration increased and death became imminent. This amounted to as much as 20 in many cases. Very frequently there was a slight initial rise in temperature during stage 1.


Our studies on alkaline reserve were very incomplete. Determinations taken at irregular intervals in 11 experiments seem, however, to justify the statement that there was no change of particular significance until the latter periods of the experiment. At about the time blood pressure fell so markedly there was a decided decrease in the alkaline reserve. The tissues at that time were undoubtedly suffering from oxygen want and the decrease in carbonate was due to the formation of fixed acids.


The various pathological physiology studies just reported are of interest chiefly in giving a conception of phosgene poisoning as a whole. In fatal cases, and these studies were made on such, two rather well-marked stages were apparent. The first of these was characterized by the nervous reflexes due to the irritation of the gas in the respiratory passages and by the direct chemical action of the gas or its decomposition products on the blood. The second stage was characterized by well-developed pulmonary edema and its natural consequences. The whole subject, indeed, might well be termed a study of pulmonary edema induced by phosgene.

The first effect of the gas was to injure the linings of the deep respiratory passages. Spasm of bronchiole musculature was evidence of the stimulation produced by the fumes. As a result of this there was a reflex cardiac inhibition, very characteristic of many stimulations of the respiratory surfaces. The composite curve of heart rates shows a decrease during the first half of the poisoning. Another reflex from the same cause was a vasomotor one which brought about peripheral constriction, with a rise in blood pressure. The X ray gave evidence of pulmonary injury at this time although physical signs were usually entirely absent. More important than these nervous phenomena, however, is the direct action of the gas on the blood in the pulmonary capillaries. Here the red corpuscles were agglomerate(l into masses which largely


filled and blocked the capillary passages. Bubbling gas through blood showed that it might have just this effect. The results of this plugging of the capillaries were twofold. In the first place pulmonary resistance was increased and a load thrown upon the right heart. Evidence of this was seen in the right cardiac dilatation found in the X-ray plates. A second result was the removal of red cells from the circulation, which resulted in a decreased hemoglobin content of the blood. This first stage of phosgene poisoning, as shown by Underhill, is most easily determined by following the hemoglobin, and it may be spoken of as the stage of decreased hemoglobin concentration.

Underhill explained the decreased hemoglobin concentration on the basis of blood dilution by body fluids. Just how or why blood volume should be increased at this time is not clear. That this interpretation is probably not sufficient is shown by direct determinations which indicate no increase in blood volume, and by the histological examination which show the red cells agglomerated in the capillaries.

Long before the end of the first stage pulmonary edema was under way. The direct cause of this was undoubtedly the increased permeability of alveolar and capillary walls, due to direct injury from the gaseous fumes. The increased pulmonary blood pressure resulting from the capillary plugging greatly favored the condition.

The second stage of acute phosgene poisoning was characterized by rapid development of the pulmonary edema with all its physical signs. This resulted in a decreased blood volume and increased hemoglobin concentration and a smaller diastolic size of the heart. The essential thing was the greatly reduced blood volume, which is almost entirely accounted for by the increased fluid in the lungs.

Death under such conditions obviously may be accounted for in either one of two ways. The edematous condition of the lungs may interfere with the gaseous exchanges to such an extent that the animal asphyxiates, or the blood volume may be so reduced that even though the hemoglobin is oxygenated there is not enough fluid to secure its proper distribution to and circulation in the tissues. So far as the tissues themselves were concerned the result was the same. They died of oxygen starvation. Death was probably due to a combination of the two causes. This belief is based on a series of experiments in which gassed animals were immediately placed in chambers containing 40 to 60 percent oxygen. The carbon dioxide content was of course kept within physiological limits and oxygen supplied automatically so as to keep the amount constant. The majority of these animals lived from 48 to 72 hours instead of the average 16 and seemed to be recovering. On being released the usual occurrence was for each dog to walk across the room and fall into an asphyxial convulsion, which quickly terminated in death. Several of the animals were hurried back into the oxygen chamber and resuscitated. These animals were edematous with reduced blood volumes, but in an atmosphere of 40 percent oxygen life was preserved, one is tempted to believe, because of complete hemoglobin saturation and physical absorption of oxygen. That decreased blood volume is a cardinal part of the syndrome, and possibly by far the most important part, need not be questioned.



1. A study of the pathological physiology of acute phosgene poisoning shows a well-marked succession of events which finally results in typical pulmonary edema. The microscope and the X ray both show an early injury to the linings of the deep respiratory passages. Irritation from this results in a certain amount of reflex cardiac inhibition and vasoconstriction. Coincident with these changes there is a direct action of the gas on the red blood cells, which causes them to agglomerate and obstruct the pulmonary capillaries.The removal of red blood cells from the active circulation in this way results in a decreased hemoglobin percentage. The plugging of the capillaries throws a strain on the right heart and a right-sided cardiac dilatation is apparent. These are the chief characteristics of stage 1.

2. Even during stage 1 the injury to the alveolar membranes and the increased pressure have initiated the transfusion of fluid from the blood into the tissue spaces and later into the air passages of the lungs. The rapid development of the edema is the chief characteristic of stage 2. It results in hemoglobin concentration, reduction in blood volume and decrease in heart size, all three of which proceed to extreme degrees. Death ultimately results from decreased oxygenation of the pulmonary blood and from oxygen starvation of the tissues due to decreased blood volume, the latter being probably the more important.


(1) Underhill, Frank P.: The Lethal War Gases. New Haven, Conn., Yale University Press, 1920.
(2) Meek, W. J., Gasser, H. S., and Erlanger, J.: Studies in Secondary Traumatic Shock American Journal of Physiology, Baltimore, Md., October, 1919, 1, No. 1, 31.