SECTION II
PATHOLOGY OF GAS GANGRENE FOLLOWING WAR WOUNDS
Gas gangrene is a spreading, moist gangrene produced by gas-forming anaerobic bacteria in extensively traumatized tissues. It is characterized by a gaseous infiltration and edema of the part affected, and by changes in the color and contractility of the muscle. The infection may remain strictly localized or it may tend to spread and involve a single muscle group or an entire limb or other part of the body; in the latter type the onset and course are generally rapid and accompanied by profound toxemia and high mortality. The condition is practically always associated with a mixed infection and the clinical picture varies according to the combined characteristic activities of the etiologic agents present. With a certain type of bacterium dominant, we may have, as one extreme, a gaseous infiltration; as another, edema without gas; and as another, rapid digestion and dissolution of tissue with neither edema nor gas. Most frequently the dominance is only partial, and edema, gaseous infiltration, and rapid tissue destruction go hand in hand with profound intoxication. The aerobic bacteria which usually are present are either ancillary, through preparing a more favorable substratum for the anaerobic bacilli which constitute the determining factor in the production of gas gangrene, form a part of the process affecting the wounded tissues, or both.
The biology of the bacteria which are the etiologic agents in this most fatal of war wound infections gives us the key to the peculiarities of distribution and prevalence of gas gangrene during the World War. The anaerobic bacteria here in question are frequently, if not habitually, present in the intestines of man and of many of the lower animals. They thrive on decaying animal and vegetable matter, but not on living tissue, and through their resistant spores are able to retain their vitality through long periods of conditions unfavorable to their multiplication. They grow best in the absence of free oxygen and in their growth produce toxic substances. The type of wounds produced by high explosive shells was peculiarly suited to the development of such anaerobic bacteria, and the bacteria themselves were furnished by a soil, which for centuries had been fertilized with the feces of man and the lower animals.
Of the anaerobic bacteria, B. welchii (Clostridium welchii) claims first place through being the member of the pathogenic group most frequently present. The organism called by Pasteur Vibrion septique (Clostridium septique), commonly, although possibly erroneously, accepted as being identical with the B. Aedematis maligni of Koch, is less frequently present in war wounds than B. welchii, but its presence is of more serious import. B. aedematiens (Clostridium novyi) seems subject to greater variations than the others, but may be found more frequently than Vibrion septique. B. histolyticus (Clostridium histolyticum) was first isolated by Weinberg and Seguin,1 and the frequency of its occur-
408
rence was not noted by other workers. These four anaerobes are recognized as the most important agents in the production of gas gangrene. Their characteristics are as follows:
One of the outstanding characteristics of B. welchii is the large amount of gas it produces. It is one of the most hardy of the anaerobes and one of the most widely distributed. B. welchii produces hemolysins for the blood corpuscles of man, rabbits, dogs, pigeons, white mice, hogs, cattle, sheep, horses, guinea pigs, white rats, and chickens. It dissolves hemoglobin out of the red cells of the guinea pig, rabbit, hen, and pigeon. It destroyes the red cells of the other animals.
It has been possible to produce agglutinating substances by the injection of cultures of washed bacilli into rabbits or horses. McCampbell 2 produced precipitins in rabbits by parenteral injection. Korentchevsky3 claims to have produced precipitins successfully by administering broth culture filtrates to animals per rectum and by feeding them by mouth.
McCampbell 2 claims to have produced bacteriolysins against B. welchii. In 1915, Jablons4 succeeded in producing bacteriolysins in a horse injected serially with small quantities of first killed and then living cultures of B. welchii. This serum had a bacteriolytic titer of 1-100; in doses of 0.5 c.c. it protected against a lethal quantity of a 24-hour culture. McCampbell 2 found complement fixing bodies in the serum of animals injected with B. welchii. Korentchevsky3 claims similarly to have found complement fixing bodies in the serum of dogs fed with culture filtrates.
While Weinberg and Seguin1 claim to have produced toxins in cultures of B. welchii, they state that the problem is surrounded with great difficulties. Korentchevsky3 found that filtrates of broth cultures were toxic for young rabbits, symptoms appearing in from one to three hours after intravenous or intraperitoneal injection. Klose5 claims to have demonstrated toxic substances in subcutaneous exudates from guinea pigs and in filtrates from 5 percent glucose broth cultures after 14 days' incubation. It remained, however, for Bull and Pritchett 6 to produce first a toxin of appreciable potency. This was done by growing a virulent strain of the bacilli in 0.2 or 0.3 percent glucose broth to which fragments of fresh muscle had been added. To obtain the most potent toxin the culture was filtered after 18 to 24 hours' incubation. Tested by injection into the breast muscle of pigeons the lethal dose obtained from several individual strains ranged from 0.3 to 3 c.c. Weinberg and Seguin' at first injected whole cultures in the hope of preparing serums, at the same time anti-infectious and antitoxic. Tested upon animals the serums were shown to have a certain degree of potency. Bull and Pritchett 6 produced antitoxins through the injection of horses with the toxins prepared by their method.
The virulence of B. welchii, in common with all the other anaerobes of gas gangrene, is extremely variable. Strains isolated from fatal cases of gas gangrene are usually very pathogenic for the guinea pig, and are able to retain their virulence over a period of months.
Observations upon animals infected with pure cultures of B. welchii have thrown much light on the pathogenesis of gas gangrene. Formerly it was believed that B. welchii was of itself always capable of producing so-called gas
409
bacillus infection. The injection of large quantities of washed bacilli, however, fail to produce lesions in laboratory animals. Lesions can be produced if either acid or powdered glass is injected with the bacilli, or if a slight injury to muscles precedes the injection. In other words, the Welch bacillus is not a true parasite but a saprophyte that can not grow in healthy tissues; it is able to thrive only if the body cells are first injured by some chemical or mechanical means. The lesions involve the muscles, the blood vessels, and the fat of the subcutaneous connective tissue.
The symptoms following injection of bacilli plus toxin, or following injection of bacilli after injury to muscles, are invariably the same. If the culture is virulent, it is not uncommon to see swelling and induration of the region injected within four to six hours after inoculation. The swelling increases and extends up to the abdomen; in about six to eight hours crackling is perceptible. The animal holds its limb in a flexed position to relieve the tension of the tissues. Swelling is due chiefly to edema, and to gas infiltration. As the bacteria multiply, the symptoms of intoxication become manifest and are the same as those observed when sublethal doses of toxin are injected into the circulation of susceptible animals. The nervous symptoms are tremor, slight convulsion, hiccough, bristling of the hair and paralysis of the bowel. The effect on the circulatory system is evidenced by the extreme rapidity of the pulse.
The injected animal, before death, shows paralysis of its hind legs, it remains hunched up and motionless, and its respirations are markedly increased. Death intervenes by respiratory failure rather than by cardiac paralysis.
Vibrion septique is nonproteolytic, but slightly peptolytic, and very saccharolytic. Specific agglutinins have been demonstrated in the serum of animals inoculated with cultures of V. septique. It produces both a hemolysin and a hemagglutinin. It hemolyzes the blood cells of the guinea pig, rabbit, sheep, goat, and man. The pathogenicity of various strains varies, so that it is sometimes necessary to inject fairly large quantities of culture to produce characteristic lesions.
The lesions in rabbits and guinea pigs are similar. The swelling and induration develop fairly rapidly but are never so extensive as those produced by B. welchii. Crepitation is very slight as the bacillus produces only a small amount of gas. The edema extends up the abdomen but is soft and compressible. Small blisters appear on the skin from which serosanguinolent fluid exudes. The hair does not fall out as it does in B. welchii infection. With some strains the swelling is slight but there is stiffness of the injected limb. Intravenous injection into a rabbit usually is followed by the same train of symptoms. Paralysis of the fore limbs is particularly to be noted.
B. aedematiens is saccharolytic and digests gelatin. Agglutinating serum has been produced. This organism, however, has a tendency to agglutinate spontaneously, and satisfactory suspensions can be made only with very young cultures. It produces hemolysis which destroys, in vitro, the red cells of man, sheep, and guinea pigs. Toxins fatal to guinea pigs in doses of 1/10 to 1/40 c.c. have been produced. The animals died within 48 hours following intravenous injection. There seems to have been little difficulty in obtaining antitoxins of a potency comparable to that of diphtheria antitoxin. According to the tests reported, the antitoxin likewise had anti-infectious power.
410
The lesions caused by different strains of B. aedematiens are not always the same. If a toxin-producing strain is used, the intramuscular inoculation is followed in a very short time by an appreciable swelling of the limb and this increases hour by hour. The thigh becomes cold, the skin pale and bluish, a firm elastic edema extends up the abdomen, often to the sternum. At autopsy, the muscles at the site of injection are congested and infiltrated with very fine gas bubbles. There is no putrid odor present. Bacteria are found seldom in the fluid of the wound, rarely in the blood stream, and with difficulty in the peritoneal fluid. Blood cultures made before or immediately after death are usually negative.
If the strain used is virulent but not very toxigenic it produces a different type of lesion. The thigh becomes swollen, the skin red and moist and a fine crepitation is noticeable, extending up the edematous abdomen. The edema on incision is gelatinous and deep pink or strawberry red, and is infiltrated with a few, relatively large, gas bubbles. The edematous fluid infiltrating between and around the muscles is usually hemorrhagic and is visible beneath the semitrans- parent edema. The muscles are markedly congested, deep violet-red in color, and infiltrated with large gas bubbles. No gangrenous necrosis and no putrid odor are present. A large number of Gram-positive bacilli are present, arranged in pairs, both in muscle fluid, peritoneal exudate, and on the liver surface. Blood cultures made before death are usually positive.
Subcutaneous injections produce the same type of lesion but require larger amounts. There is usually a very thick, lardlike edema infiltrating the subcutaneous connective tissue for a depth of from 3 to 4 cm. The lesions are never putrid. The abdominal muscles are congested but never gangrenous.
B. histolyticus is the most proteolytic of all the anaerobes. It is pathogenic for all laboratory animals (guinea pig, rabbit, white mouse, and rat). The pathogenicity varies with the strains studied. The intravenous injection of 1 to 2 c. c. of a 24-hour culture in Martin broth will kill a guinea pig in several minutes. The intraperitoneal injection is lethal in the same dose. Intramuscular injection into the thigh of a whole culture produces lesions characteristic of this microorganism. Immediately after injection of 1 or 2 c.c. the animal shows violent tetany of the muscles of the head and neck and cries out plaintively. After several hours, the thigh is swollen, the skin is violet-red, and a slight edema, which is soft and compressible, invades a part of the abdomen. If one kills the guinea pig at this time, incision of the lesion shows the presence of a hemorrhagic layer in the subcutaneous connective tissue, containing large red, cherrylike clots. The subcutaneous and perimuscular connective tissues are attacked and seem to be undergoing rapid digestion. The lesion is accompanied by a hemorrhagic edema, which invades the connective tissue of the peritoneal wall. If the process continues for several hours, the digestion of the perimuscular connective tissue is more evident, then the muscles and the muscular bundles become dissociated and liquefied.
To obtain a good toxin, it is necessary to use very young cultures. Beyond 24 hours at 37 º C., both toxic power and virulence rapidly diminish and finally disappear. This toxin produces in vitro, as well as in vivo, all of the phenomena observed with the culture. The filtrate of B. histolyticus liquefies gelatin rapidly, just as does the bacterium itself. The guinea pig, the rabbit, and white mouse
411
injected with toxin present very extensive hemorrhagic lesions. Similar lesions have been observed in horses used for the preparation of an anti-histolytic serum. The B. histolyticus does not secrete hemolysins for the red cells of man, sheep, and guinea pigs. Horses immunized with the bacterial bodies injected intravenously and deprived of their toxin by centrifugalization produce a serum which strongly agglutinates homologous strains.
The activity of spores in soil has been made the subject of a special study by Joffe and Conn.7 From their results it is easy to see how in ground under constant cultivation, and especially terrain over which severe fighting had recently occurred, or where many men had been living, the activities of spore- forming bacteria would be enormously stimulated. The conclusions of Joffe and Conn, in part, are as follows:
There is reason to believe, therefore, that the role of spore-forming bacteria in soil is, under ordinary conditions, one of watchful waiting. They seize upon bits of organic matter and utilize periods of high moisture content to grow occasionally for a few generations before going back into the spore state, and thus maintain their numbers. When special conditions furnish them with some unusually available organic matter in the presence of considerable moisture, they immediately germinate in large numbers and carry on the initial stages of the decomposition of the organic matter more rapidly than the nonspore formers, which, under ordinary field conditions, are the active soil organisms.
The presence of the group of bacteria which produces gas gangrene upon the clothing and skin of the soldier always depends upon the bacterial flora of the geographic area in which he is and upon the character of fighting going on at the time of his wounding. These factors are also influenced by moisture, temperature, and the character of the terrain, an area with a stone or chalk subsoil being less favorable to the maintenance of this flora than a soft subsoil. The intensity of manure tillage and the recency of this cultivation are important factors. If the terrain is such that dugouts and bombproofs cannot be built in the ground and the trenches are practically built on top of it, the chances of the occurrence of gas gangrene are greater. The intensity of the fighting is another important factor, as the uncovering of buried bodies by constant shelling and the consequent inability to provide the area with proper sanitary conveniences also operate to increase the liability to this condition. The soldier cannot keep his uniform in a condition even approaching proper cleanliness, and infrequently has opportunity for bathing. Scabies is particularly important in this regard because of the habit of the female of boring into the skin to deposit her eggs. These borings, filled up behind her with dirt and bacteria, are veritable culture tubes that, carried deep into the muscle by high explosive shell, afford the presence of the exciting factor in an ideal situation for the production of the infection. Simple bathing cannot eradicate such foci.
These are the most important of the external causes of the production of this lesion. The same factors also influence the internal causes. Lack of water and proper food and loss of sleep produce those physiological conditions which increase the liability to shock by lowering blood pressure. This makes local vascular blocking easier.
Of the 224,080 officers and men of our Army who were wounded in battle in France, 13,691 died as a result of their wounds, the total death rate there-
412
fore being 6.11 percent.8 The wounds complicated by gas gangrene may be divided among those who sustained injuries of the soft parts only and those whose wounds were complicated by bone fracture. Injuries to the soft parts as recorded here do not include those due to chemical-warfare gases. There were, with these exceptions, 128,265 wounds of the soft parts with 9,719 deaths; of the wounded in this group, 1,389 developed gas gangrene, which amounts to only a little more than 1 percent (1.08). The death rate among those who received wounds of the soft parts which became complicated with gas gangrene was 48.52 percent, the actual number of deaths being 674.
Among the 25,272 whose wounds included bone fracture there were 2,751 deaths.9 The incidence of gas gangrene among the bone-fracture cases was much higher than among those who sustained wounds of the soft parts only, the total being 1,329 with 593 deaths. The incidence in this group of the wounded was therefore 6.26 percent and the case mortality rate 44.62 percent.
A study was made at Base Hospital No. 15, A. E. F., of a series of 276 cases of gas gangrene with 73 deaths, a death rate of 26.45 percent.9 The figures showed that the activity of the gas-gangrene group was self-limiting and practically confined to the first week after the wound was received, showing a drop of the anaerobes during the first 7 days from 38 to 7 percent. It was shown, further, that as the self-limiting process of the gangrene went on, the aerobes, especially the common pyogens, streptococcus and staphylococcus, accumulated rapidly in the wound. The cases were charted according to the influence of the symbiosis of the various aerobes and anaerobes, and a summary of these charts indicates that the various organisms did not seem to alter the killing action of the gas-gangrene group. Although no deductions were made from the statistics because of the small totals, it was believed that the association of anaerobes with streptococcus produced a more virulent infection than with anaerobes alone, and that the association of staphylococcus seemed to produce less virulent infections than with anaerobes alone.
The following table gives a striking example of the etiology of gas infection. Of 16 fatal cases 9 showed gas infection and 4 showed the presence of staphylococcus, streptococcus, and anaerobes. Seven without clinical evidence of gas infection showed the same bacteriological incidence. It is possible that eight of these deaths were due to pyogenic septicemia, and the deaths which occurred after the fifth day were probably all due to pyogens. Among the nonfatal cases of this group, 18 of the cases without clinical gas infection showed the presence of B. welchii and other anaerobes, practically always accompanied by the pyogens, and four cases of the nine with gas infection showed the virulent gas bacilli, whereas six of these nine also showed the presence of the pyogens. The important feature of this analysis is that in 18 cases there were present the anaerobes concerned in the production of gas gangrene without the infection ensuing. This shows that the condition is not a concrete bacteriological entity but a pathological complex depending on numerous tissue factors.
413
Bacteriology of gas infections 9
CHART
From this study it appeared that anaerobes were prominent in the early bacterial picture of wounds, but began to disappear by the end of the fourth day, when the aerobes became progressively more prominent. Streptococcus had a fairly constant incidence, and was very persistent.
A combination of Streptococcus hemolyticus and Staphylococcus aureus, was the most prominent of all the bacterial associations in wounds at Base Hospital No. 15.9 This association was particularly prominent in fatal wounds. Streptococcus hemolyticus, Staphylococcus aureus, Bacillus welchii, Bacillus coli communis, Bacillus proteus, and, to a lesser extent, Staphylococcus albus, were the bacteria appearing most prominently in both the fatal and nonfatal cases.
The prognosis was good in cases showing Staphylococcus aureus or other aerobes excepting streptococcus. An association with Staphylococcus aureus seemed to lower the virulence of infection with anaerobes.
The presence of the frank pus producers had an important influence on checking the progress of gas gangrene. This seemed to be especially true of staphylococcus. In all probability the positive chemotaxis of these organisms for leucocytes may well account for the reverence with which the surgeons of the Civil, and other earlier wars, spoke of "laudable" pus in connection with gangrenous wounds.
Infections with anaerobes showed a high death rate, but a short period of danger to life, unless the anaerobes were in association with streptococcus.
Deaths from streptococcus infections were numerous, and occured at least up to the end of the fourth month. The mortality was even higher, where streptococcus was associated with Staphylococcus aureus or with anaerobes. On the other hand, the association of streptococcus with both Straphylococcus aureus and anaerobes gave a lower death rate than the simple association with either Staphylococcus aureus or anaerobes. Moreover, for infections by the association of all three, the period of danger to life seemed to be shorter than that for infections, by streptococcus with either Staphylococcus aureus or anaerobes alone.
414
At Base Hospital No. 15, a streptococcus bacteriemia was by far the most important cause of death in cases of war wound.9 This was common where patients lived beyond the first week, the self-limiting period of the gangrenous process, and many deaths attributed to the anaerobes were in reality deaths due to streptococcus in the process of replacing them.
An important study was made at Evacuation Hospital No. 8, A. E. F., of cultures from wounds in an investigation into the causes of gas gangrene.10 In the fall of 1918, this hospital was near the forward areas, served a sector on the heights of the Meuse, which, because of the hinging operation in the retreat of the German forces, remained fairly stationary, and was practically confined to trenches and dugouts until the end of hostilities. It was an area that had been fought over throughout the war, and its soil was rich with the bacterial flora found in gas gangrene. Because of the forward position of this hospital, the evacuation of patients was very rapid and gas gangrene cases were encountered in their earliest stages. Between September 10 and November 13, 1918, 4,741 wounded were admitted to this hospital. Of these, 4,683 required surgical treatment. Two hundred and six of the wounded required amputation, 96 of these being performed for gas gangrene. Ninety-three amputations were done because of severity of injury, many probably being done as a prophylactic procedure for the prevention of gangrene. The following table shows the distri- bution of these amputations:
CHART
During this period the total number of deaths was 363. These deaths were due to the following causes:
CHART
It is thus seen that gas gangrene was the second most important cause of death. Of the 4,377 patients admitted to the hospital that had received wounds in action, 221, or 5 percent, had gas gangrene. This incidence is high but since this hospital received only the seriously wounded, and since it was only approximately 10 hours from the front line, this is not unusual. This agrees with the carefully worked out figures of German and French authorities.
415
Of the 157 cases of gangrene in which satisfactory records could be obtained, the lesion was well developed on admission to the hospital in 141 cases, or 90 percent. In 16 cases, or 10 percent, the condition developed after operation. The average length of time between injury and operation in 162 cases of gangrene with satisfactory record, was 41.8 hours. In 349 cases of similar wounded which were nontransportable, and which did not develop gas gangrene, the average length of time between injury and operation was 24.67 hours. This indicates that many of those operations performed for extent of injury and the prevention of shock probably had the double effect of preventing, as well, the onset of gas gangrene.
The following table shows the relation between the interval before operation and the development of gas gangrene:
CHART
It is seen from this table that the delay in the modern surgical application of debridement had a marked influence in increasing the incidence of this condition.
The majority of these lesions occurred in the heavily muscled areas of the body and in one-third of the cases in this series they occurred in the lower leg. The anatomical reason for this high incidence is contained in the fact that the posterior tibial artery in a distance of about 1½ inches gives off the majority of its muscular branches to the calf. Machine gun and rifle bullets which fracture the tibia in this vascular area, necessitated the blocking of the bulk of the blood supply of the calf and presented the typical conditions of the failure of circulation, so important as the causative factor of this condition. Because of this vascular mechanics this is one area where bullets are almost as important factors as high explosive shell fragments in the production of gas gangrene.
Of 890 wound cultures, 478 or 53 percent contained anaerobic bacilli. Of these 478 wounds, 321 or 67 percent at no time showed clinical evidences of gas gangrene infection.
Of the remainder of the gas gangrene cases, 16 or 3 percent developed gas gangrene after debridement, while in 141 cases, or 29 percent, gas gangrene was clinically evident at the time the wound culture was made. It is thus seen that more than two-thirds of severe nontransportable wounds contaminated by anaerobic bacilli failed to develop gangrene, all of these cases being under observation at least five days, and many of them as long as two weeks. This fact involves a very important pathological principle. These bacteria are incapable of producing gas gangrene by their presence alone, and must be accompanied by the failure of circulation, the extensive cellular damage of large quantities of muscle, and in all probability by a constantly progressive increase of this series of factors.
416
Only a very small percentage of cases developed gas gangrene after operation. One hundred fifty-seven cases of gas gangrene were examined bacteriologically. In 18 of these the cultures were made at a distance from the wound, incisions above the wound, and from the amputation stump. In 139 cases in which cultures were made directly from the wounds, the relation of the B. welchii to the other anaerobes was as follows:
B. welchii was the only anaerobe in 20 cases, or 14 per cent; other anaerobes were found in 30 cases, or 21 percent, and B. welchii and other anaerobes were found in 89 cases, or 65 percent. In 139 cases of gangrene, streptococci were found in 44 cases, or 31.6 percent. In 321 cases without gas gangrene, streptococci and anaerobes Were found together in 97 cases, or 30 percent. It is thus seen that the streptococcus was no more frequent in the gangrenous wounds than in those that failed to develop gas gangrene.
Previous wars had not prepared the medical departments of the various armies for the high incidence of gas gangrene which characterized the evolution of gunshot wounds received in France, during the World War, for the essential causative factors were heretofore not all cooperating so perfectly. Peace-time injuries so rarely produce the conditions which are favorable to the development of gas gangrene that our knowledge of it at the outbreak of the war in 1914 was both fragmentary and inaccurate. The chaotic state of our knowledge of the anaerobic bacteria was due very largely to the difficulty of obtaining and of maintaining them in pure culture.
In all probability the most important single factor in the production of gas gangrene is the intensive damage to muscle at a distance from the surface, under conditions that prevent the free access of oxygen. The next most important factor is the local ischemia produced by the damage to the circulation. There is histological evidence of immediate injury to muscles with a zone of disintegrated muscle around the missile tract, which closely resembles a freshly made beef emulsion, and makes an ideal culture medium for this group of bacteria. The lack of definite lymphatics within the mass of skeletal muscles, making a whole muscle bundle a true lymph space, determines the first actual interference to circulation. When a large muscle bundle is broken into, this single lymph space is destroyed, and the ebb and flow of nourishment is disrupted. Although the muscle in close relation to the missile tract is killed immediately it does not differ essentially in appearance from the muscle at a distance from this point, where its death is caused by the internal pressure of edema and gas gangrene, and the resulting interference to the circulation. The center of large muscles under these conditions, receives the greatest injury, since the muscles just beneath fascias get their lymph from lymphatics in these structures. A classical vicious circle is produced under these conditions. The traumatic opening of this main lymph space, produces an imbalance in the circulation and results in edema and the flow of serum into the damaged areas. The invasion of these collections of serum, of edema along dense fascial planes and of neighboring muscles by the members of this group of bacteria produces gas gangrene. The gas elaborated by the bacteria increases the internal pressure within the injured area. This pressure increases until it equals that in the lymphatics when it closes this avenue of release to the general circulation.
417
With the return lymph flow thus occluded the pressure mounts and approaches that of the pulse pressure. When this occurs the return of the circulation through the veins is impeded. At this point muscle degenerates from lack of nutrition, and the infection progresses rapidly so that the pressure may rise and approach first the level of the diastolic, then of the systolic level of the circulation. This completes the mechanics of the vicious circle and opens the malignant phase of the progress of gas gangrene into neighboring tissue. The pressure may be so great that reverse flow occurs in the lymphatics of neighboring muscles, and the bacteria spread back along perivascular lymph channels to involve more distant muscle groups.
Bashford 11 noted the spread of these bacteria at a point 5½ inches from the wound tract in muscle fibers and capillaries. He found that when cultures were injected into the back of the thigh of a rabbit the organisms appeared halfway up the anterior abdominal wall 3½ hours after injection, under favorable conditions for infection.
The muscles at a distance from the wound rapidly undergo cloudy swelling, fatty degeneration, necrosis, and the accompanying changes which alter their color and reactions to stains. They are penetrated by the spreading gases, and the mechanics of this vicious circle is so pyramided, that the infection can be well advanced in two hours, and death from massive gas gangrene of an entire limb can occur in less than 30 hours.
The reaction to these bacteria varies widely. Following the traumatism there is a definite period of delay before any increase in the number of bacteria present in the exudate. This period is rarely less than four hours and often not more than six. The only changes in the wound that may be noted during this period are due directly to the injury. There is a positive leucocytic reaction when the quantity of bacteria is small enough and the damage to the circulation is not great enough to prevent the constant renewal of these cells and the liquid elements of exudate that contain antibodies. With the development of a gangrenous process and the accompanying multiplication of the anaerobic group of microorganisms, the leucocytes diminish rapidly in number. The few that remain are quickly destroyed and converted into a homogeneous granular detritus which may bear no resemblance to white blood cells.
With favorable evolution of the wound considerable phagocytosis will be demonstrable. The leucocytes will be found to contain relatively large numbers of bacteria. This is usually a favorable sign and indicates increasing resistance. As the infection diminishes in intensity the polynuclear leucocytes decrease in number to be replaced by a corresponding increase in the mononuclear elements which are also phagocytic.
When, however, the evolution of the wound does not progress favorably and gangrenous changes begin, the cytologic picture is quite different. This difference concerns especially the leucocytes and may be revealed by vital staining and staining with Sudan III. Smears made at about the second hour show that out of 96 leucocytes only 8 show any evidences of fat content. After the tenth hour at least half the leucocytes show fatty changes.
After this period, with unfavorable evolution of the wound and with the development of gas gangrene, the bacterial content of the exudate and of the
418
surrounding tissues holds the center of the stage. The bacteria are aided in producing damage by the hydraulic injury to the walls of blood vessels produced by the peculiar dynamics of high explosive shells. Ruptures if surface veins have been found more than 3 inches below the site of injury. (Fig. 168) Large muscles, especially those under tension, because of the closed lymphatic space which they represent, produce an ideal situation for dissemination of injury to a distance because of the hydraulic action of this semifluid tissue within dense fascial envelopes. The force which ruptures smaller vessels tends to damage the intima, produces shredding of the inner coats of larger vessels and results in secondary or delayed gas gangrene, and many of the dramatic secondary hemorrhages which occurred in these conditions. Shredding of the inner coats accounts for the linear tears from which these massive hemorrhages occurred.
Hematomas were very important factors in the production of gas gangrene in the World War wounded. These hematomas, encircling the wound, checked hemorrhages by their own pressure within confining fascias. They interfered physically with lymphatic drainage, and augmented internal pressure and its influence on the progress of this condition. Fresh blood is an unfavorable medium for the growth of gas gangrene group but as the clot ages it becomes an ideal culture medium for these organisms. (FIG. 169) The formation of acides of this group of bacteria
FIG. 168.- Posterior aspect of left leg amputated for a mixed type of gas gangrene, with extensive putrefactive blackening of the inner and posterior aspect of the upper two-thirds of the calf. There are several high-explosive wounds on the posterior surface, around which there has been superficial dissection of the skin in the search for injuries to vessels and subcutaneous hemorrhages responsible for this zone of gangrene. Seven and one-half centimeters above the uppermost wound is a rupture of the long saphenous vein, probably produced by the missile through hydraulic transmission of its force. Arrow points to rupture. Accession number 4720. Army Medical Museum. Negative number 30237
419
forces them into the spore stage. This self-limiting action is neutralized by the alkalinity of the exudate. With complete blocking of the circulation the increased acidity stops proliferation, and gas gangrene no longer progresses.
The Welch bacillus forms spores in the presence of coagulable proteins, and produces quantities of gas and organic acids in the presence of glucose. These acids, which consist chiefly of the butyric group are inimical to the growth of this organism and are formed readily in muscle, on account of its glucose con tent. This, and the fact that pus is in part a coagulable protein and forces the organism out of the vegetative and into the sporulating stage, are important factors in the self-limiting character of this infection. It must do its damage in the first few days of its activity or be superseded by pus-producing bacteria.
Emery12 used powdered chalk in his cultures in order to overcome this inhibition of growth by acids. The chalk neutralizes the acid as it forms and allows a sufficiently luxuriant growth for the study of pathogenicity. The protective power of serum and leucocytes is maintained only as long as they are fresh and constantly renewed, and fails when the circulation no longer furnishes the needed replenishment of these elements as determined by the extent of the infection. A preponderance of the bacilli and their products over the leucocytic response may result in a negative chemotaxis for these cells and they no longer phagocytose and destroy the organisms. It is readily seen then how important the blood supply is to a part subjected to the factors which produce gas gangrene. The normal blood supply may prevent infection but when this has already been started a much greater supply must be furnished.
As the gas gangrene group decreases, the more saprophytic organisms take their place. When this occurs the tissues of the muscle or the limb go on to ordinary gangrene and slough away. At those zones where the circulation is established, the more virulent members of the group are limited by local leucocytosis and give way to the members of the pyogenic group, especially hemolytic streptococcus and staphylococcus. Inflammatory reactions due to these organisms then replace the gangrenous process which does not reappear unless there occurs a thrombosis or other obstruction to circulation which renews the degenerative process in muscles, and again affords a suitable medium for the multiplication of these organisms. Many of the deaths attributed to gas gangrene which occurred after the fifth day of wounding probably should be credited to a blood invasion by streptococcus.
The clinical classification of Weinberg and Seguin1 offers a convenient one in describing these lesions.
(a) Virulent gas gangrenes.
1. Emphysematous type.
2. Toxic or edematous type.
3. Mixed form.
4. Putrefactive type.
(b) Avirulent gas gangrenes.
421
The Welch bacillus is the greatest producer of gas, and the infections caused by it are usually the classical or emphysematous type. B. aedematiens and Vibrion septique produce less gas and more edema, and predominate in the edematous lesions. The mixed forms are caused by any combination of the above, while the putrefactive lesions may show any and all combinations of aerobes and anaerobes, including the ordinary pyogenic bacteria.
During the World War the emphysematous type of gas gangrene was the most virulent of this group. It usually showed a preponderance of B. welchii in all cultures of the wound, and occurred in relative purity in the depths of the muscles involved. It was responsible for the large collections of gas bubbles in muscle and the most extensive swelling of limbs occurred in this type of lesion. The type of wound producing the emphysematous form of gangrene was usually that of a small opening with massive internal injury of a well muscled portion of the limb. The small wound of entrance and exit served to keep the infection in relative purity in the deeper zones, and maintained the anaerobic conditions in these areas throughout the process. The small wound of entrance and exit tended to decrease the liability to contamination with the putrefactive members of this group of bacteria. The surface wound was dry, its edges were fairly clean, and the intense pressure produced by the gas and edema in the deeper portions forced the walls of the wound tract to protrude some distance above the level of the limb. The following cases illustrate the clinical and pathological characteristics of the virulent emphysematous type of gas gangrene.
CASE 1.- Soldier was wounded July 21, 1918, at Chȃteau Thierry, and was admitted to American Red Cross Military Hospital No. 5 on July 22, 1918, with a diagnosis of gunshot wound of the right thigh and of the left knee. There is a gaping wound on the anterior surface of the lower third of the right thigh and another wound on the external surface of the same thigh, just above the knee joint, with a compound comminuted fracture of the right femur in its lower third and a well advanced gas gangrene of the right thigh. Above the external condyle of the left femur is a superficial furrowed wound whose base is covered with a shaggy exudate. This specimen consists of the right thigh including the articular end of the femur. On both medial and lateral surfaces there are two wounds. The external wound measures 10 by 6 cm. and the internal one is approximately the same size. Both wounds have been debrided, the skin edges having been trimmed, and part of the muscle removed. Denuded bone fragments can be seen in the external wound. All muscles are infiltrated by gas. The X ray shows a large amount of comminution of the fracture of the femur with most of the muscle bundles separated by gas bubbles. (See Pls. XV and XVI and figs. 170 to 180, inclusive.)
CASE 2.- Amputation just above knee joint for gas gangrene following a resection of the head of the tibia for gunshot wound. Advanced gas gangrene of the muscles with putre- factive changes in the tissues in and about a superficial hemorrhage into the skin and sub- cutaneous tissues. (See figs. 181 to 188, inclusive.)
CASE 3.- Soldier was wounded July 17, 1918. Gunshot wound of left thigh and left shoulder. Left thigh amputated July 22, 1918. Secondary hemorrhage occurred July 31 at 10 a. m., because of which the femoral artery was ligated 2 inches above the stump. Patient died three hours after ligation. The amputated limb showed a compound comminuted fracture of the middle of the shaft of the femur. The surrounding tissues were infiltrated with gas, and the muscles were a characteristic brick red color, except in the immediate vicinity of the wound where there was some darker discoloration due to putrefactive bacteria.
At necropsy there was gas gangrene of the stump and gas in small quantities was seen in the liver. (See figs. 189 to 195, inclusive.)
422
CASE 4.- Soldier was wounded on July 17, 1918. He entered the hospital, July 20, with multiple wounds, from fragments of high explosive shell, of the right thigh, middle finger of left hand, neck and left thigh. The wound of the left thigh was extensive but did not fracture the femur. There were large debrided wounds of exit and entrance exposing dark colored, partially necrotic muscles. July 26, the left leg was amputated because of gas gangrene.
The amputated limb shows a diffuse infiltration with gas, the base of the wounds being covered by a greyish-green slough.
At necropsy the stump was seen to be involved in the gangrenous process. There was a purulent arthritis of the right knee. There was a perforating wound of the larynx just below the vocal cord, with a shell fragment the size of a pea in the cartilagenous tissue of the left side. This wound was accompanied by edema of the larynx and glottis and extensive peritrachial hemorrhage. Bacterial stains of sections of the wound of the left thigh, superficial portions, show numerous streptococci. (See figs. 196 to 203, inclusive.)
CASE 5.- Soldier wounded October 16, 1918, by fragments of high explosive shell. October 18, 1918, the wound was debrided, and a free dissection of all the muscles of the calf was done. October 21, there was present an extremely emphysematous gas infection. The leg was amputated just above the knee on October 25, because of extension of the em- physematous gangrene, and secondary infection with pyogenic organisms.
Amputated leg shows a free dissection of the posterior muscle bundles which appear attached only by small bands of tissue bearing their blood supply. The muscles show extensive fragmentation as the result of gas emphysema, the gas eing more abundant in the muscle tissue than
FIG 170.- Vertical midsection through the femur. There is a massive comminution with putrefactive involvement of the narrow cavity. The muscles on the right of the bone show the same putrefactive reaction. Those to the left of it show almost no putrefactive changes but contain massive collections of gas bubbles. Accession number 2879, Army Medical Museum. Negative number 30890.
PLATE XV
DEBRIDED WOUND OF ANTERIOR SURFACE OF THIGH.
Muscles extruded by pressure from below. Tension and pigmentation of skin from gangrenous process in muscles beneath.
Accession 2879, Army Medical Museum. Colored photograph.
PLATE XVI
GAS GANGRENE.
Vertical section behind femur of thigh shown in Plate XV. Muscle about wound shows the dark color of putrefactive gangrene. Surrounding this area the muscles are the brick red color of gas gangrene. Muscle fibers separated by gas.
Accession 2379, Army Medical Museum. Colored photograph.
433
between the muscles and in the subcutaneous tissues. Cross section above the dissected area shows the freedom of the muscles from marked emphysema but there is a marked superficial and deep lymphangitis. (See figs. 204 to 209, inclusive.)
CASE No. 6. - Soldier wounded June 7, 1918. Admitted to hospital the same date with bullet wound of shoulder. Operated June 8. Findings: Penetrating wound of the arm at about the level of the fold of the axilla. Almost complete severing of the brachial artery. Contusion of the brachial vein with thrombosis. Hemorrhage necessitated tying of the subclavian artery.
The tissues became infected with gas-producing bacilli, extended over the chest and the bacteria entered the circulation. Death occurred June 13.
Necropsy was performed 14 hours after death. The tissues of the arm and of the upper chest on the left are discolored, and present several incisions. The tissues of the upper arm are almost black. The forearm is swollen and tense but not discolored. Subcutaneous gas can be detected over the left arm and forearm, the left side of the neck and beneath both pectoral muscles. Necrosis and liquefaction are present in the left pectoral muscles. Bubbles of gas numerous in the tissues of the lungs, liver, and spleen. (See figs. 210 to 214, inclusive.)
CASE 7. - The following case is a typical one to show the influence that traumatized muscle has in the production of gas gangrene, and the speed with which this can develop. This soldier was hit by a 37 mm. shell which exploded on contact. The wound occurred about 4 a. m., October 27, 1918, and he was seen in the admitting room of Evacuation Hospital No. 6 at 8.30 p. m., walking about the room, smoking a cigarette and talking rather volubly. His diagnosis tag read " Gunshot wound right thigh, compound, comminuted fracture." A tight cotton cloth tourniquet had been applied and was still in position and tight enough to check hemorrhage. His ability to walk about with complete crushing of the lower thigh and fracture of the femur and tibia led the examiner to suspect that this soldier was in the excited stage of shock. He was sent immediately to the shock room and amputation was performed, after four and one-half hours of shock treatment, at 1 a. m., October 28, 1918, about 21 hours after being wounded. He answered very readily to shock treatment and it was thought by the members of the shock team, that the tight tourniquet had saved him from the shock and toxic products which the pulped muscles might have thrown into the circulation but for it. In all probability such shock as he had, was due to insult to the central nervous system, and the tourniquet had been applied quickly enough to prevent augmentation of this shock, first from hemorrhage and second from the toxic products of autolysized muscles.
The amputation specimen consists of the right leg which had been removed by the guillotine operation at the middle of the thigh. The lower half of the leg at the lower third of the soleus muscle has been cut away as it showed no lesion. The cotton cloth tourniquet was still in place just below the line of amputation. All of the popliteal vessels have been ruptured by this direct explosion. There is a piece of 37 mm. shell casing measuring 1 by 2 by 5 cm. lying just under the skin over the patella. There is a compound, comminuted fracture of the tibia and fibula, at the middle of the upper third. The fracture through the femur was by secondary violence. The wound on the upper surface of the right calf measured 11 cm. in diameter. It is ragged and roughly circular in outline and the muscle is completely pulped throughout its extent. (Fig. 215.) It has driven the tibia through the skin on the opposite side of the leg, making a wound through the skin 7 by 3 c. m. in extent. (See figs. 215 to 225, inclusive. P1. XVII.)
CASE 8. - Soldier wounded by high explosive shell July 19, 1918. Admitted to hospital July 20, 1918. On July 21, amputation was done through upper third right humerus for comminuted fracture of the femur in its middle third, complicated by extensive gas gangrene.
Right leg was amputated through the upper third of the femur for gas gangrene com- plicating two shell wounds of the right thigh. The gas extended above Poupart's ligament.
X-ray report: Gunshot wound of right thigh with numerous foreign bodies of various sizes, large cavity in the muscle with loss of substance, some separation of bundles of muscle with gas. (See figs. 226 to 235, inclusive.)
434
CASE 9 .- Soldier received a shell wound of the right foot July 20, 1918. Admitted to hospital July 21.
On July 22, foot was amputated above the right ankle because of gas gangrene. The amputated limb shows a shell wound on the outer side of the instep of the right foot caused by a piece of shell 4 by 4 by 4 cm. The wound was infected with gas producing bacilli, the gas infection involving the plantar muscles. There was beginning gangrene of the fifth toe due to destruction of the vessel. (See figs. 236 to 242, inclusive.)
CASE 10. - Soldier wounded by high explosive shell fragments August 28, 1918. On this date an extensive wound of the left leg, with compound comminuted fracture of the upper half of the tibia, was debrided and 4 inches of tibia removed. There was considerable serous discharge at the time of operation. Length of time after injury not given. There were little dirt or fragments of clothing in the wound. September 6, 1918, amputation was performed in the lower third of the left thigh for gas gangrene, and severe secondary infection by pyogenic bacteria, which included the tibia. Hemolytic streptococci and Welch bacilli were recovered in cultures taken on this date. (See figs. 243 to 250, inclusive.)
CASE 11. - Soldier was wounded June 6, 1918. Admitted to hospital June 9, with gunshot wounds of both thighs. There was a compound comminuted fracture of the upper third of the left femur and a simple gunshot wound of the right thigh. The wound of the left thigh was debrided and bone sequestra were removed. The bone was infected with streptococci. Patient died August 5, 1918, the intervening history being unknown.
The necropsy, 16 hours after death, showed a partially healed wound of the left thigh, 9 by 5 cm., communicating with necrotic bone. There is another large open wound on the undersurface close to the perineum. The leg is markedly swollen. There are small perforating wounds on each side of the left knee, the outer one extending to the joint capsule. There is a large amount of exudate in the tissues. The tissues of the left thigh are distended with exudate and gas. The muscles are the characteristic brick red color of gas gangrene. Considerable young callus has been thrown out about the fractured bones. The edges of the wound show beginning epithelization. The lungs, spleen, liver and kidneys show infection with gas producing organisms, in the form of gas bubbles. The wound in the right thigh is nearly healed. (See figs. 251 to 255.)
The toxic or edematous type of virulent gas gangrene was usually caused by B. aedematiens and Vibrion septique, and showed less virulence than the emphysematous form. There was marked edema beneath fascias between muscle bundles, and even surrounding individual fibers. There was little or no gas in these lesions. The muscle was pale and moist instead of dry, and had a homogeneous, textureless character. As in the gaseous form, the fascias were swollen with a glairy, gelatinous exudate and they and the collections of edema beneath them were often stained a cherry red. Hemolysis of escaped blood was a prominent feature and most of the collections of exudate were stained by it. These cases showed abundant evidence of hemolysis in the wound, which was usually confined to the local focus. Jaundice and anemia from hemolysis rarely occurred. These cases died with toxic symptoms, often showing extremely high temperatures, and those respiratory failures which indicate toxicity for the heat regulating and respiratory centers in the brain. The following cases illustrate this form of the virulent type of gas gangrene:
CASE 12. - Soldier was wounded in the right arm by a machine-gun bullet which produced a markedly comminuted fracture of the humerus. An extensive debridement was done. The wound measures 15 by 18 cm. and occupies the inner aspect of the upper arm. The musculospiral nerve is severed in the upper portion of the wound. The brachial vein is severed in its middle portion and the brachial artery has been injured. Dissection of the wound shows a fragment of the humerus which has pushed the musculospiral nerve before it without breaking the sheath into which there is a moderate amount of hemorrhage. (See figs. 256 to 265, inclusive, and Pls. XVIII and XIX.)
PLATE XVII
GAS GANGRENE OF THE CALF MUSCLES.
Beginning formation of gas bubbles and some putrefactive change in the clot around the tibia, Typical color of muscle of gas gangrene without secondary infection. Fragment of shell under skin below patella.
Accession 4707, Army Medical Museum. Autochrome.
473
FIG. 219.- Gas gangrene of the calf muscles. Putrefactive change in the clot around the upper end of the fractured tibia. Accession number 4707, Army Medical Museum. Negative number 30233
