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 IX







 Since the comparatively recent discovery and isolation of the germs which cause gas bacillus infection, this lesion, which had been grouped among the more virulent varieties of sphacelus or mortification, has become a recognized and distinct surgical entity. Always of great rarity in times of peace, the wide variation in its clinical manifestations resulted in a correspondingly diverse interpretation of this species of infection. With the advent of the World War, however, it became much more frequent, especially on the Western Front, and afforded abundant opportunity for a thorough investigation. It was then expected that the entire subject would be quickly and satisfactorily standardized and an adequate classification of its more or less complex features formulated. Although considerable progress has been made, the problem has not as yet been satisfactorily solved. Unexpected difficulties have been encountered. Thus, the usual association of several varieties of specific organ-isms in the same focus of infection, to which, not infrequently, are added one or more varieties of pathogenic organisms, has greatly increased the complexity of the clinical manifestations. Furthermore, it is difficult to account for the fact that infection by a single species of the specific organism may cause widely different results. It is also quite possible that the list of specific organisms that cause this infection is not yet complete and that certain unexplained features of the infection may be due to an imperfect knowledge of its cause. On the whole, while a certain uniformity has been reached in its classification, the general result of laboratory and clinical investigation during the war led to the conclusion that the subject is one of great complexity and that many of its problems have not yet been satisfactorily solved.


Diversity in language and in methods of investigation led to considerable confusion in the classification of the specific organisms which cause gas bacillus infection. Certain varieties which, from their description by observers in widely separated countries, seemed to be different species, were ultimately found to be identical. Other varieties described in pre-war literature were not identified by any observer during the war period. The confusion was still further increased by the frequent association with the specific organisms of some of the more common pathogenic germs, resulting in a considerable modification of the clinical features of the original infection.
Notwithstanding the confusion arising from these various causes, it has been demonstrated beyond doubt that the majority of cases of gas infection are due to one or more varieties of anaerobic bacilli and that the extent and severity of the infection depends directly upon the capacity of these organisms to secrete toxin. Their growth and development is favored by dead muscle tissue and the liquefaction of this tissue by certain proteolytic ferments, which


many of these organisms possess in variable amounts, provides a medium in which, unquestionably, the secretion of the toxin is promoted. A majority of these same organisms possess also saccharolytic ferments which are largely, if not entirely, responsible for more or less gas production, a clinical feature that long since stamped this infection with its classic title. Less important, as well as less constant, is the occasional possession of certain fat-splitting ferments which cause disintegration of the fatty connective tissue in the invaded area.

In close association with these anaerobic organisms, aerobic varieties are almost always found. These play an entirely secondary r le and are never dangerous when alone. Certain varieties, however, possess proteolytic ferments and by liquefying tissue provide a medium that stimulates the growth and development of the anaerobic varieties. Douglas, Fleming, and Colebrook1 pointed out that further development of the anaerobes is favored by the capacity of the aerobes to absorb oxygen and thereby to diminish the quantity of this element in the area of infection. Closely associated with the specific organisms of gas gangrene and yet nonpathogenic in character, is a third group of bacilli which produce putrid abscesses in animals and to which the foul odor of the discharge in these infected wounds is due.

While tabulations of these various specific and more or less closely affiliated organisms rarely agree, the writer selected the tabulation published in 1919, by Jablons,2 in connection with a paper on the subject of gaseous gangrene, omitting, however, a list of aerobic and anaerobic bacteria which, isolated incases of gaseous gangrene before the war, were not confirmed subsequently by any other observer:

(1) Toxicogenic organisms, anaerobic:
    (a) Those capable of reproducing the disease in animals-
            (1) B. welchii, Gas bacillus, Bacillus aerogenes capsulatus, Bacillus per fringens.
            (2) Vibrion septique, bacillus of malignant edema.
            (3) B. oedematiens, B. gasoedem, B. bellonensis.
            (4) B. fallax.
            (5) B. hystolyticus.
            (6) B. sporogenes.
            (7) B. aerofetidus.  
            (8) Streptococcus anaerobicus.
        (b) Those which do not reproduce the lesions in animals-
            (9) B. bifermentans.
            (10) B. putrificus.
            (11) B. tertius.
            (12) Bacillus V of Ghon and Sachs. '   
    (2) Aerobes capable of reproducing analagous lesions in animals-
        (13) B. mesentericus.
        (14) B. anthracoides.
    (3) Those which are found in association and produce in animals putrid abscesses or are nonpathogenic are-
        (15) B. proteus.
        (16) B. coli.
        (17) B. subtilis.
        (18) B. pyocyaneous.
        (19) B. friedlander.
        (20) B. mycoides.


Of the various anaerobic organisms, Nos. 1, 2, 3, 4, 6, and 7 possess an active saccharolytic ferment, the Bacillus hystolyticus possessing none. No. 1 has a slightly active proteolytic ferment which is more actively present in a, 5, 6, 7, 8, 9 and 10.

Weinberg and Seguin,3 in an analysis of 91 cases, found No. 1 present in77 percent, No. 2 in 13 percent, No. 3 in 34 percent, No. 4 in 16½ percent, No. 5 in 9 percent, No. 6 in 27 percent, No. 7 in 5½ percent, No. 9 in 2 percent, No. 10 in 2 percent, and No. 11 and No. 12 each in 1 percent. Of the 91 cases, 10 were caused by a single variety of anaerobic bacillus, 14 by several varieties and 67 by both aerobes and anaerobes. There was no instance of an infection caused by an aerobe alone. The frequent association of one or more varieties of these specific organisms, as demonstrated in the above table, is interesting also because of the possibility of consequent inhibition or intensification of the virulence of the infection. Thus the toxins of B. oedematiens and of V. septique are practically destroyed by B. sporogenes. On the other hand, B. welchii and B. oedematiens mutually stimulate activity. As similar stimulation occurs when B. welchii and V. septique are associated, while B. welchii alone is stimulated by B. sporogenes.

Pathogenic organisms frequently coexisting in wounds infected by one or more varieties of gas bacilli include streptococci, diplococci, staphylococci, tetanus bacilli, diptheroid bacilli, and others. . Thus Ivens4 reports 464 cases in 59 of which a virulent streptococcus was isolated and in 15, the bacillus of tetanus. Weinberg and Séguin,3 in their analysis of 91 cases state that streptococcus was found in 40 per cent, diplococcus in 33 percent, staphylococcus being slightly less frequent.

Douglas, Fleming and Colebrook1 discuss certain interesting features of symbiosis, stating that streptococcus, staphylococcus and diptheroids stimulate the growth of B. welchii and that with both staphylococcus and streptococcus the stimulation is mutual. Furthermore that streptococcus stimulates the growth of other anaerobic bacilli.

The predisposing and local conditions which favor the development of this infection may be grouped as follows: Predisposing--(1) Systemic, as in the more common types of infection; causes that diminish the powers of resistance, such as fatigue, loss of sleep, lack of nourishment. (2) Loss of blood with consequent loss of antigen bodies, as well as of other elements that ordi- narily inhibit infection. (3) Shock, which causes a suspension of nervous activity and regulation. Local conditions--(1) Atmospheric conditions of heat and continued moisture or humidity. (2) The character of the soil. Thus, soil that has been repeatedly fertilized contains large numbers of the specific organisms of gas gangrene. Fragments of clothing and of other material so contaminated and driven into the wound, together with similarly infected shell fragments, frequently cause gas gangrene. (3) Wounds involving the large intestine. Such wounds provide a means of exit for the specific organisms which are indigenous in this part of the intestinal canal. (4) Length of exposure. It is quite obvious that this infection is more common in neglected wounds than in those in which d bridement is promptly done.


As to the character of the wound, lacerated and contused wounds caused by irregular high-explosive fragments of low velocity, with a relatively small wound of entrance into which muscle substance may prolapse, create favorable conditions for the development of this infection. The more extensively muscle tissue is damaged the greater the likelihood of infection. Contusion, without laceration of muscle tissue, as, for example, from the pressure of broken bony fragments, predisposes to its death by directly compressing its blood supply. A similar though more destructive condition arises from a cutting off of the main arterial supply or from interference with the venous return. The importance of an intact circulation in preventing the development of this infection is seen in those cases in which the infection appears in wounds of several weeks' duration, subsequent to the ligation of a nutrient artery for aneurysm or for secondary hemorrhage.

Conditions relieved by prompt debridement well illustrate competent predisposing causes for gas gangrene. Thus, irregular pockets from which proper drainage is impossible, hematomas from uncontrolled bleeding, foreign bodies, fragments of clothing, loose bone fragments, damaged and lacerated tissues, all favor, if neglected, the development of this infection. The need of providing free and unrestricted drainage of all discharge from every part of the wound can not be too strongly emphasized.

While gas bacillus infection is unquestionably due to one or more varieties of specific organisms, of which a list has been given, the presence alone of the specific organism in the wound would not necessarily result in infection, for wounds from the discharge of which the specific organism of gas gangrene has unquestionably been cultured have been observed repeatedly to heal by primary or secondary union without the slightest evidence of infection. It is only in wounds in which muscle is so traumatized that it undergoes necrosis and where the damaged tissue is so deeply placed that access of oxygen is difficult or impossible that this justly dreaded infection is likely to develop.


Lesions caused by gas bacillus infection vary according to the virulency of the specific organisms; the presence or absence of associated germs, such as streptococci, staphylococci, the bacillus of tetanus, and the bacilli producing putrid changes; and especially the extent of the death of muscle tissue due to the violence of the trauma or to the destruction of the main or collateral circulation of the part involved. It is therefore impossible to define any single pathological picture that is typical of the infection as a whole.

Reference has been made above to the toxicogenic qualities of anaerobic bacilli and to the proteolytic and saccharolytic ferments which they possess invariable amounts. To their capacity to secrete toxin the greatest importance is attached, for not only do these toxins cause tissue changes leading ultimately to gangrene at the point of infection, but, with their rapid entry into the general circulation, patients quickly succumb to the intensity of the virus. On the other hand, in the absence of toxin secretion, no serious local or constitutional changes are observed.




For anaerobic bacilli to secrete toxins, dead muscle is necessary. Such a suitable medium is usually provided by the force of the trauma, either directly through the impact of a displaced bony fragment or through the crushing of the intima of nutrient vessels, with resulting thrombosis. Where death is due to circulatory interference the color of the muscle becomes purple and on microscopical examination the striae are intact. On the other hand, when death is directly the result of the action of gas bacilli the color is either brick-red or mahogany, depending upon the degree of heinolysis, and the striation is lost. This change in the color of muscle tissue as well as its loss of contractility and the fact that muscle so invaded remains dry without sign of blood when divided are positive indications of the presence of gas gangrene.

Microscopical examination of the affected muscle shows at first a dilatation of the capillaries and small blood vessels, followed either by a rupture of their walls with small ecchymotic extravasations or, because of the paralysis of the muscular part of the media by the local action of the toxin, multiple aneurysmal dilatations appear. In both, the hlmina of the vessels are thrombosed, a condition that still further contributes to the spread of the gangrene.

In necrosis due to ordinary pyogenic infection, nature usually endeavors to restrict the infection by phagocytic concentration and other defensive means. In gas bacillus infection, however, mobilization of nature's protective forces does not take place. On the contrary, extension along the course of the muscle or muscles originally involved is very rapid, and, in the event of the shutting off of the main blood supply, the entire extremity becomes quickly necrotic.
The bacilli are found in the early stages of infection between individual muscle fibers. These become swollen, surrounded with an edematous exudate, and finally, losing their structure, are invaded directly by the infecting organism.

Edema develops early and causes the initial swelling. At first a serous exudate, it soon becomes discolored at the point of infection. Remaining clear at the outskirts of the infection, it extends along the areolar planes and in the subcutaneous tissues, preceding the extension of the infectious process in the muscle planes. Occasionally it spreads with such speed as to deserve the title of "malignant" and under these circumstances it completely outstrips and overshadows the phase of tissue necrosis.

All anaerobic bacilli, with the exception of hystoliticus, though in different degree, possess saccharolytic ferments. Upon this content seems to depend the production of gas. At first deeply seated, it infiltrates individual muscle fibers between which the bacilli have penetrated and, as it increases in amount, extends along the course of the wound to the subcutaneous tissue and so at times over the entire body. Animal experimentation has shown that this gas has no toxic properties. By the pressure it exerts in the deeper planes it probably facilitates the extension of the infection, and by compressing the blood supply increases the degree of necrosis. It can not be too strongly emphasized that it does not usually develop in sufficient amount to be detected in the early stage of the infection and that when it can be recognized the infection has reached a stage in which radical treatment is demanded.


Proteolytic changes include the conversion of necrotic tissue into a mushy shapeless mass of extremely foul odor which, in localized processes, give rise to the formation of putrid abscesses, and in the rapidly extending varieties increase the intensity of the general toxemia. Some of the bacilli of gas infection secrete fat-splitting ferments, with the result that fatty connective tissue is attacked and partially digested.

It is both interesting and important to consider the pathological changes that mark the line of demarcation between healthy tissue and the invaded area. Such a line may be either regular or extremely irregular, due to the unequal involvement of the different planes. In the muscle tissue an area of congestion with small hemorrhages generally is observed, on the proximal side of which muscle fiber still retains its normal contractility, color, and blood supply. If a single muscle only is involved, the infectious process may be limited to that structure extending up and down along its longitudinal axis, for adjacent muscles are well protctel by intermuscular aponeurosis acting as a barrier unless there is a shutting off of the main arterial supply, in which event massive gangrene of the entire extremity rapidly develops. The associated serous exudate in the zone of muscle necrosis is usually irritating and infectious, while beyond the line of demarcation it is probably innocuous as long as it remains clear and straw colored and is free from any suspicion of odor.

Lesions found in distant organs are chiefly due to toxemia. The occasional development of metastasis, however, and the recovery of the specific organism from the circulation during life, clearly demonstrate that under certain conditions the infection may become generalized. This is borne out by the observations of Weinberg and Séguin 3 who, in their study of 91 cases, recovered the Bacillus welchii in 4 cases before death and in 11 of 13 after death. The Vibrion septique was recovered in 3 out of 4 cases and the B.oedematiens in two only before death and in 5 afterwards. Further proof of generalization is furnished by the observation of Mullally and McNee 5 in which the bacillus of malignant edema was recovered at the site of three needle punctures foul days after the patient was wounded; an amputation of the arm was performed six days subsequent to the receipt of the injury. This patient recovered from the infection only to succumb to an attack of pneumonia one month later. In metastatic gas gangrene the organism is frequently recovered from the secondary focus. These metastases usually occur in the buttock or shoulder, regions ordinarily subjected to pressure in the recumbent posture, with consequent diminution in the local blood supply.

To the region attacked by metastasis the infecting organism travels through either the blood or lymph. While lymphatic transmission is probably the rule, the development of metastasis when the evident path of infection is directly opposed to the course of the lymphatic stream demonstrates the fact that the bacillus must be occasionilly transmitted bv the blood current. The occurrence of metastasis in an extrenity from a plriimar focus on the trunk may be mentioned in illustration of infection bv this route. On the other hand. in the much more frequent metastasis in the buttock or shoulder from a primary focus in the corresponding extremity, the specific organism is probably conveyed through lymphatic channels. While anaerobic bacilli unquestionably


increase rapidly in the occluding throinbi at the point of infection and easily pass into the adjacent blood stream, their further development must be greatly impeded by the oxygen content of the red cell. This probably accounts for the difficulty in recovering these bacilli from blood smears. To be sure, in the later stages of the infection when the patient becomes moribund, resistance to the entrance into the general circulation greatly diminishes and at that stage as well as immediately after death, they are more frequently recovered from the blood, as Weinberg has demonstrated, than during the active course of the infection. On the contrary, the absence of oxygen carriers in the lymphatic stream facilitates the transmission of the bacilli by this route, although the lymphatic ganglia may more or less effectively retard their progress. The writer knows of no attempt to determine the presence or absence of the specific organism in the glands which directly drain the primary focus of infection. Investigation of this part of the subject might lead to very interesting results.


Changes in the abdominal organs vary according to the type of infection. When associated with pyogenic organisms, the liver, spleen, and kidneys show acute degeneration such as is ordinarily seen in septic processes. In gas gangrene alone, these changes are not so marked. In their place these same organs appear swollen and spongy and on section show a frothy or foam like infiltration due to the presence of gas. Although similar changes can be produced in laboratory animal experimentation, it is not at all certain that in man they are not post-mortem in character. At all events, there is no record of any such observation during the life of the patient.
Adrenal capsule
.-The yellow cortical substance becomes grayish white, due to the disappearance of the lipoid tissue. The fasciculated cells show degenerative and inflammatory changes similar to those observed in peritonitis but more rapid in development. In the presence of these advanced changes in the cortex, the interior or pulpy portion of the glands remains normal. These changes are of interest on account of their possible relation to the changes observed in the blood pressure of patients suffering from gas bacillus infection.
Brain.-This organ is usually pale and somewhat edematous. The fluid in both the subarachnoid space and ventricular cavity may be increased. In delayed death, the basal vessels show gaseous infiltration.
Heart.-The muscle fiber is pale and shows cloudy swelling. There may be subendocardial ecehymotic extravasations.
Lungs.-There are often patches of bronchopneumonia, in which occasionally the anaerobic bacillus may be found. There may be small subpleural hemorrhages.
Stomach and intestines.-These organs usually are distended, although exceptionally they may be empty.


Variations in the type and number of the specific organisms, as well as their frequent association with one or more varieties of common pyogenic organisms, mentioned above in the consideration of the pathology of this


subject, make it necessary to observe a rather wide latitude in any description of the clinical manifestations of gas gangrene. The subject perhaps is preferably approached by a detailed description of individual symptoms followed by their collection into groups according to the severity of the infection.


This is usually from one to four days, according to the malignancy of the specific organism and to the extent of the necrosis of muscle tissue. It is noteworthy, however, that this period may be prolonged for weeks or even months. In fact the specific organism may remain quiescent during the entire reparative process and continue harmless for months in the cicatrix, only to be released in the course of an operation then undertaken to remove a piece of shell fragment or other foreign body or to correct a bony deformity resulting from the original trauma, or it may even follow the removal of the cicatrix for plastic purposes. A similar outbreak of the infection occasionally develops in cases of appendicitis in which a secondary abscess may occur months or years after the original operation. In gas infection, however, such tardy activity is not, as in colon infection, necessarily mild but may prove most serious and even may terminate fatally.


The invasion is usually insidious. If the infection is not associated with pyogenic organisms, pain, referred to or below the point of infection, is not infrequent. It is rather a more or less sudden intensification of the pain previously existing due to the trauma of the penetrating wound from which it must be carefully distinguished. Under similar conditions (absence of pyogenic organisms) swelling is quickly added to the pain, due to the incipient edema of the subcutaneous tissue. This edema gives to the skin above and below the wound a whitened appearance followed by a creamy tint, and makes the surface veins more distinct and dilated. Also it imparts a slightly tense feeling of elasticity, though there is no pitting on pressure, to the subcutaneous tissue. Of the four cardinal symptoms of inflammation these two only are present, redness and heat being conspicuously absent. The frequency with which the streptococcus or staphylococcus is associated with the specific organism, however, accounts for numerous exceptions to this general rule. In that event all four symptoms of inflammation may be elicited. In addition to the paili and swelling, local symptoms include the appearance of the wound and its immediate environment, the type and character of the edema, the formation, location, and behavior of the gas.



At first, as in a healthy wound, the immediate discharge is serosanguinolent. However, the serous element quickly diminishes in amount, becomes pinkish and discolored, and shortly afterward assumes a dirty brown color. The discharge is irritating to the surrounding parts. The wound rapidly becomes


unhealthy and its sloughy surface is apt to be covered with a gelatinous discharge which, as gas is produced, may contain air bubbles. These same bubbles can frequentlvy be expressed from the wound. The discharge, moreover, usually develops a more or less typically foul odor in the early stages of the infection due to rapidly increasing putrid bacteria in tissues already swarming with the specific organisms. The edges of the wound become necrotic and ragged. With the approach of the gas toward the surface, the skin adjacent to the wound assumes a brown, bronzed, or orange color; in

FIG. 157.- Gas gangrene of arm before operation. (Courtesy of Maj. Benjamin Jablons, M. C.)

the more malignant types the color is blue or violet. In this discolored area, coalescing vesicles appear. These vesicles are numerous in the vicinity of the wound, and occasionally they show a tendency to encircle irregularly the circumference of the extremity. In other cases they seem to follow the course of the superficial veins. If their contents are light in color these vesicles are believed to be due to lymphatic occlusion; if dark in color, to thrombosis of the smaller superficial blood vessels. Later, as the condition progresses, the fluid contents may contain air. The segment of discolored skin containing


these vesicles gradually becomes leatherlike and may eventually be discharged as a slough. Such necrosis is rapid and certain in the more malignant type where the discoloration is blue or violet and is the forerunner of death of the entire extremity below the level of the infection.


This is due to serous exudate both in the involved muscle planes and, afterwards, in the subcutaneous tissue. In the zone of infection it is tinged with blood and swarms with anaerobic germs and, later, with those responsible for the foul odor of the discharge. Receding from the point of primary infection the specific organisms decrease in number until, at the line of demarcation, the serum, although still possessing toxigenic qualities, is entirely free from organisms of alny kind. In this area it is considered by some to be relatively harmless, if not actually beneficial. It is thus said to be analogous to the straw-colored exudate in the early stages of infectious peritonitis, which, at

FIG. 158.- Gas gangrene of arm, colored man, after amputation (Courtesy of Maj . Benjamin Jablons, M. C.)

first free from contaminating organisms, speedily becomes invaded by the spreading infection. The edema extends more or less rapidly along the lymphatic spaces in the neurovascular sheaths and in the subcutaneous tissue, and while at times it indicates the level of muscle necrosis, it occasionally outstrips this particular phase of the lesion and spreads with frightful rapidity over the entire body and neck. The edema is at times much less compressible than is ordinarily the case, and the line of demarcation is sharply delineated by a distinctly raised wall which permits an accurate estimate of its progress. Such a type of edema seems to be associated with the more malignant forms of gas bacillus infection. especially with those dlue to the bacillus edematiens.


The presence of gas causes a distinct crepitation which can be detected by gently stroking the overlving skin with a smooth, flat instrument, and, as it becomes more abundant, bv pressure with the palmar surface of the fingers. It indicates by its location the muscle or muscles involved and extends in all






directions until the affected muscles are completely necrotic. Intermuscular septa ordinarily check its advance until the entire extremity becomes gangrenous, when gas infiltration is rapidly generalized. It can be detected in the early stages by auscultatory percussion and shortly afterwards by ordinary percussion. X-ray exposures show at first longitudinal streaks which, extending, become gradually more distinct and then quickly develop into air bubbles-an extension which can be readily noted in successive X-ray plates.

As gas production increases, bubbles appear in the discharge or may be expressed from the wound. The length of time elapsing before gas can be recognized is approximately longer when the affected muscle is deeply seated than when it is superficial. When deeply seated, the infection may have gained great headway before this symnptomll appears. In general, it must be emphasized that the presence of gas is not manifest in the early staves of the infection. The subcutaneous tissue becomes infiltrated with the gas when it passes along the interstices of the wound toward the surface of the body. It is at this stage that it either appears in the discharge or may be expressed from the wound. The possibilitv of the primary involvement of the subcutaneous tissue has been much discussed. That gas infection develops originally in necrotic muscle is now generally conceded, and where the gas is entirely superficial to the deep fascia, involvement of a small portion of the under-lying muscle has been assumed. The strongest argument that has been advanced in favor of the development of this infection without damage to muscle substance rests upon those cases in which gas bacillus infection has followed the subcutaneous injection of some medicinal agent, such as camphor or digitalis, or where it has developed after the infiltration of subcutaneous tissue with saline solution in hypodermoclvsis. As a matter of fact, however, it is in patients already suffering from gas gangrene at some distant point that this unusual condition has been noted, and it is quite possible that it maybe accounted for as a metastatic manifestation due to the presence of the specific organism in the general circulation.


The initial whitening of the skin due to the incipient edema has been noted. With the formation of gas and the continued edema, the swelling increases, involving the extremity more rapidly below the point of infection than nearer the trunk. The gradual extension of the swelling may be demonstrated by successive measurements of corresponding portions of the two extremities, the line of demarcation being at that point where no increase in the circumference of the affected extremity can be detected.

It can not be too strongly emphasized that treatment should not be delayed until the characteristic symptomls of crepitation and odor appear. In the early part of the war radical and thorough débridement of the wound had not been developed, and treatment frequently was delaved until symptoms of infection had appeared, with most unfortunate results. Even in the latter part of the war, because of unavoidable delay in the succor and transportation of the wounded, or because of the virulency of the infection, patients not


infrequently were admitted to hospitals at the front in such a condition that no radical measure could be of benefit. These unfortunate patients afforded abundant opportunity for the observation and study of the later clinical manifestations of the infection.


Constitutional symptoms are due to the toxemia as well as to the absorption of chemical products resulting from the destruction andl decomposition of the invaded tissues in the infected area.

.-Acceleration of the pulse is one of the earliest symptoms. In a few hours after the invasion the pulse reaches 130 or higher and is small in quality. At first a temporary rise of pressure is noted, possibly due to suprarenal changes; a rapid fall quickly ensues and, in fatal cases, continues to the end.

-The temperature is almost always relatively low, usually not exceeding 101 0 to 103 0 F.; it is entirely out of proportion to the quality and rapidity of the pulse. This peculiar combination, namely, a low temperature with an unusually rapid pulse, should immediately excite suspicion of the presence of this serious infection.

-Respiration is decidedly increased in frequency due to the rapid decrease in red cells, and other changes that greatly diminish the oxygen content of the blood. In the more malignant cases dyspnea appears, although usually the skin remains pale, without sign of cyanosis, to the end.

Surface of the body.-With the development of the circulatory disturbance and the progressive anemia marked pallor appears. This continues to the end. Jaundice is not infrequent, giving, together with the pallor, a dusky hue to the skin. The surface of the body is usually moist and may be bathed in perspiration.

-There is usuallv constipation; occasionally diarrhea.

General nervous system.-Patients are usually apathetic and fail to appreciate the seriousness of their condition. Restlessness is not uncommon; delirium rarely appears; consciousness is preserved to the end. Death results from paralysis of the important basal nerve centers.


Since gas infection varies widely in virulency, and consequently in the extent and character of both local and constitutional symptoms, it is customary to somewhat arbitrarily assign all cases to one of three groups: (a) Those of mild character; (b) those of a malignant type; (c) those intermediate in severity. It is quite obvious that the line of demarcation between these different groups is not always sharply defined. Thus, an infection which appears at first mild and localized, may suddenly become virulent and spreading, while infections of moderate severity may rapidly develop malignant manifestations. Unfortunately infections, originally serious, rarely if ever become less virulent. In these malignant cases only prompt treatment can save the patient's life.



In view of the fact that one or more varieties of anaerobic organisms frequently can be cultured from the discharge of wounds which heal without complication, it is not at all surprising that gas gangrene may remain localized. This fortunate result is most likely to ensue when the damage to muscle is superficial, or slight in extent, and the muscle exposed to air; or when the associated anaerobic bacilli are mutually restrictive. Similarly, when a single muscle only is invaded, the infection, although spreading in both directions along its longitudinal axis, may yet be confined by the intermuscular barrier to the muscle originally involved.
In these cases of localized gangrene the infection not infrequently appears after transportation of the patient to a base hospital. The relatively long period of incubation is followed by a mild grade of infection, with the result that the mortality is very much less than in a field or evacuation hospital. In this group of cases the local symptoms differ from those previously enumerated in their extent and frequently resemble the local symptoms of an abscess due to one or more pathogenic gas-producing bacteria. In fact, the symptoms of crepitation common to both types of infection has probably led to erroneous diagnoses. By the inspection of the abscess cavity in the course of treatment the two conditions are readily differentiated. In infection due to pathogenic organisms leucocytic infiltration is extensive, and muscular necrosis, limited to the immediate abscess wall, is manifested by irregularly sloughing strings. The red-brick color, so characteristic of gas gangrene, is absent. The constitutional symptoms in the two conditions present a sharp contrast. In abscess due to pathogenic gas-producing organisms the temperature is relatively high, the pulse bounding and of good tension, the face decidedly red, and, when exceptionally the infection becomes diffuse, delirium with metastatic multiple abscesses may occur, a condition never developing in uncomplicated gas bacillus infection.


In the severe or malignant types of gas gangrene, the intensity of the infection is measured chiefly by the rapidity of extension of the local symptoms and the quickness with which patients become acutely toxic, the latter being indicated especially by extreme circulatory weakness and a condition of collapse. While the evidence of grave constitutional disturbance shows no essential variation, the local symptoms, especially the crepitation, the edema, and the character of the discharge, may vary very considerably. Thus, theoretically, the invasion of dead muscle by anaerobic bacilli alone results in a condition of intense toxemia with little edema and, if the bacilli do not possess the saccharolytic ferment, in little or no gas production. In infection due to Vibrion septique edema may be the predominant symptom. As has previously been noted the type and rapidity of extension of the edema may also vary. Thus in infection due to the Bacillus edematiens the edema is much less compressible than is the case in infection due to the Vibrion septique and, advancing with a sharp line of demarcation in the formation of a raised wall, it may extend much more rapidly over the main part of the entire body. Again the foul odor of the discharge


develops only when one or more putrid bacteria are added to the specific organisms. As a matter of fact, however, such association is rarely, if ever, absent. Considerable variation in the character and arrangement of the bullke and in the color of the skin near the point of infection is not uncommon, a darker color indicating more extensive vascular changes. Finally the frequent association of the specific organism with pathogenic bacteria may greatly modify the character of the local symptoms. In a general way, gas production, edema, and the development of a foul discharge may he considered fairly constant local symptoms while, accordling to the nature of the specific bacillus, either gas production or ed ema may predominate.



Observers differ materially in regard to the behavior of the white cell in gas gangrene infection. The larger number state that they are diminished. All agree that, in the tissue invaded by the specific organism, the barrier caused by the massing of leueocvtes is either defective or absent. This discrepancy in the effect of the infection upon the general and polymorphonuclear count is perhaps to be ascribed to the fact that, in the early stages of infection, as well is in cases of mixed infection in whichl pathogenic organisms are present, the leucocyte count, although ultimately diminished, may at first be increased. It is also quite possible that the increased leucocvtosis, reported by some observers is essentially a relative increase made possible by the rapid and progressive decrease in the number of red cells. More striking is the condition of the red cell itself. There is a marked anemia, even in the early stages, the number of red cells diminishing to perhaps less than one million per cubic centimeter. In an interesting and exhaustive essay on this subject, Jablons2 has called attention to anisocytosis, stating that the macrocytes predominate at the beginning, while the microcytes predominate toward the end of the infection. He further states that polkiiocytosis is constant. In 10 cases there was a noticeable polychromatophilia, nucleated red cells being present in 4 cases. The blood serum in fatal cases shows a definite hemolysis.

Examination of the urine, especially in cases of mixed infection, reveals the presence of albumin and casts.



Débridement of all contaminated wounds, especially of those due to explosive fragments at short range, at the earliest possible moment, is unquestionably the most efficient means of forestalling the development of gas gangrene. Wounds so treated should be left in such condition as to permit free exit of all discharge and should be allowed to heal by granulation. secondary closure being done, if at all, after the danger of infection has passed. This procedure, adopted generally in the latter part of the war, greatly diminished the incidence of gas gangrene afnd lessened its virulence.



As the knowledge of the bacteriology of gas gangrene increased it was generally expected that serum administered both as a prophylactic and also after the infection had appeared, would prove as successful as serum therapy had proved in tetanus and diphtheria. It was found, however, that while some animals could be rendered immune for various lengths of time in the laboratory, the results achieved in man were somewhat disappointing. This was probably due to several factors. In the first place the spread of the infection was frequently so rapid that the patient became either moribund or died before the specific organism could be isolated. In the second place the infection, usually due to more than one variety of the bacillus, was associated with some form of pathogenic organism, such as the streptococcus or staphylococcus, in either of which conditions a serum prepared from a single species would naturally prove of little or no value. This complex nature of the infection led Weinberg and Séguin 6 to replace serum derived from a single organism with one composed of a mixture of the sera of several of the most frequent organisms of gas gangrene. A similar polyvalent serum was introduced by Léclainche and Vallée.7 The use of this "hit or miss " method was more encouraging. Thus Frances Ivens reported 10 cases treated by the mixture of Weinberg and Séguin with a successful result in 5. The fatal cases were septicemic when the serum was given. M. Weinberg 8 stated that the polyvalent serum of Léclainche and Vallée seemed to have given particularly good results as a prophylactic, especially in wounds in which the specific organism was associated with strepto-coccus. He further emphasized the fact that the best autovaccine is one prepared from all the organisms both anaerobic and aerobic, found in the wound. Several injections are made daily or every two days and "in certain cases the effect is indisputable.” Weinberg and Séguin referred to 6 cases of infection due to Bacillus oedematiens which recovered after treatment by anti-oedematiens serum "after the infection had reached an alarming stage." Frances Ivens 9 describes the results of the prophylactic treatment of wounds with anti-gangrenous serum. This writer reports 433 cases, many of which presented clinical signs of gas gangrene and divided them into three groups, as follows: (1) 222 cases (126 fractures) treated by 10 c.c. each of anti-Welch, anti-Vibrion septique and anti-oedematiens serum of Weinberg. (2) 154 cases (110 fractures) treated bv 30 c.c. of L clainche and Vallee polyvalent serum. (3) 57 cases (34 fractures) treated by 30 c.c. Weinberg and 10 c.c. of L clainche. In each case the serum was given subcutaneously in one pint of saline at the time of operation.


Group (1): Mortality, (a) where the serum was given at or before the first operation, no case died of gas gangrene; (b) amputation, of 14, 2 died after a fortnight from streptococcal septicemia; (c) conservative treatment, serum therapy has permitted "conservative treatment instead of amputation in a large number of cases." The result in 10 cases was not ascertained owing to early evacuation of the patients. Group (2): (a) Mortality, 19 fatal cases, of which 3 were due to gas gangrene and 3 to gas gangrene with concurrent septicemia; (b) amputation, 4 of 15 were fatal, 2 with concurrent septicemia.


In the majority of cases no severe streptococcal infection occurred during the period of preventive administration with the Léclainche polyvalent serum. Group (3): (a) Mortality, 2 cases, one of massive gangrene; (b) amputation, 3 cases, 1 for streptococcal infection, 2 for secondary hemorrhage. Gangrene was present at the beginning of treatment in 10 cases. Of these, massive gangrene developed only in 1, 15 days after the preventive dose of serum had been given. Frances Ivens 9 concludes that a powerful antigangrenous serum is of real value in preventing gas gangrene; used in sufficient quantities it is of great value as a disintoxicating agent in cases of advanced infection; that the Léclainche and Vallée polyvalent serum has a marked effect in cases with concurrent streptococcic infection; that anaphylactic phenomena were frequently averted through the dilution of the serum with normal saline solution. Before secondary operation, a further fractional dose of the serums should be administered.

While these interesting results seem to indicate that serum administration is a valuable method of treatment, it is perhaps unfortunate that no attempt was made to compare them with those of a fourth group in which surgical measures only were employed. Unquestionably the virulence of gas gangrene, irrespective of the special etiological organism, varied according to the location of the battle line, the question of easy or difficult succor and transportation, the presence of heat or moisture, and other well-known conditions. While the 433 cases reported by Ivens occurred between March 21 and September 6, 1918, in patients "recently wounded, arriving at the hospital for primary operation," it is quite possible that conditions favored a relatively mild grade of infection. It is believed, without belittling in any way the results obtained in these 433 cases, that in other parts of the battle line the mortality might have proved much greater in an equal number of cases treated by similar methods. At least during the Meuse-Argonne operation, patients were admlitted for primary treatment to an evacuation hospital not infrequently with the infection so far advanced that no form of serum therapy could possibly have been of any avail. In this connection it is interesting to compare the results obtained bv serum therapy within the German lines with those of Frances Ivens.9 Herman Coenen10 reported 1,180 wounded injected immediately with polyvalent serum. Of these only 8, of whom 4 died, developed gas gangrene. Of  75 wounded not injected. 8, of whom also 4 died, developed gas gangrene, at very much higher percentage. The serum was given in doses of from 20 to 40 c.c. and later in larger doses intraveneously. Anaphylactic complications were rare and wvere manifested by dyspnea, heart weakness, and coma, resulting rarely in death. In 1 91 7, Aschoff 11 reported 2,356 wounded, of whom 223 were injected with polyvalent serum, of whom 98 died (43 percent), while of those not injected 68 percent died.


Gas gangrene, irrespective of the efficiency of prophylactic serum therapy, demands prompt surgical attention. The shorter the period of incubation, the more prompt should be the treatment. Especially where the infection shows a malignant tendency the greatest precaution should be exercised to detect it in


its incipiency. Treatment consists either in the excision of the infected area, together with all dead tissue, or of amputation. On the trunk and buttocks, obviously excision only is available. In the extremities the question of amputation must receive due consideration. Excision is justified where the period of incubation has been relatively long, where the infection is near the tip of an extremity, where the infection is localized or very slowly spreading, where the main circulatory channels are intact, where no fracture exists, where no large joint has been opened, and, irrespective of its virulence, where the infection involves the trunk. On the other hand, amputation is indicated in extensive laceration of the soft parts, where several groups of muscles are invaded by the infection, where there is an extensive comminuted fracture with or without opening into a large joint, where gangrene is self-evident, where the main vascular channels are divided, and where the symptoms of general toxemia develop early. To advise delay in the presence of one or more of these conditions until the question of the rapidity of extension can be determined, is to waste valuable time and to jeopardize the chances of recovery. A third group comprises infection on the border line. While prompt operation is as essential in these as in the more serious types of infection, the question of excision or of amputation depends upon the individual judgment of the surgeon. Many lives have undoubtedly been sacrificed by conservatism which might have been saved at the expense of the loss of a limb. In cases of doubt, amputation is the operation of choice.

In either excision or amputation, the narcosis, for obvious reasons, should be as short as possible, and all constriction during and after the operation, should be avoided, for the pressure so exerted still further decreases a local blood supply already severely impaired, and mechanically forces poisonous infectious products through lymphatic and vascular channels into the general circulation. The object of excision should be to remove foreign bodies, including shell fragments, clothing, and all dead tissue, especially necrotic muscle, stopping only when the divided muscle tissue both bleeds and contracts; to remove all hematoma; to check all bleeding; and to leave the operative field free from pockets so that all discharge shall pass without possibility of retention into the enveloping dressing. Large joints, if involved, may require excision, although in these cases, amputation, as already stated, is usually preferable. All fractures should be so treated as to avoid constriction by any form of retentive apparatus.

When amputation is indicated, no delay is justifiable. He who hesitates will frequently lose his patient. Only the plane, not the time, of amputation is to be determined. It should be at a level sufficiently high to permit of the division of healthy contractile muscle and, if possible, on the proximal side of the area of edema. As in the case of excision, no closure of the operative wound should he attempted. Circular or lateral flaps of skin and subcutaneous tissue are advocated. These may be easily and quickly fashioned and eventually brought over the end of the stump when all danger of infection has passed. In this way secondary amputation, involving further sacrifice in the length of the extremity may frequently be avoided. Such a procedure is far preferable to the alternative measure, the so-called guillotine amputation, in which the


skin, muscle, and bone are all divided at the same level. The dressing after either excision or amputation should be loosely applied and arranged so as to avoid pressure or constriction. Antiseptic wet dressings have been advocated to prevent the reappearance of the infection in the wound, such as Dakin's solution, hypertonic salt solution, peroxide of hydrogen, and weak solutions of sulphate of quinine. Some prefer to leave the wound freely exposed to the air. This, at least, has the advantage of facilitating dressings without the excessive pain of which patients almost invariably suffer when the compresses, partially dry and closely adherent to denuded tissues, are removed. Of all local applications, dichloramine-T is preferred, as it combines a certain antiseptic value with the possibility of removal of compresses previously moistened by it, with the minimum discomfort to the patient. The "open air" method also is of valtue, especially if combined with some form of cage protection arranged so as to avoid pressure or constriction of the stump.

While metastasis in pyogenic infection is usually the precursor of a fatal termination, it must not be so regarded in gas gangrene. This rare complication, occurring usually in parts subjected to pressure, such as the buttocks or shoulders, must be attacked in the same way as the original focus of infection, namely, through the excision of the necrotic tissue, the wound being left open. A considerable number of recoveries after operation for metastasis have been reported, probably to be explained by the difficulty the specific organism encounters in its effort to flourish in the general circulation.

While radical surgical treatment is invariably indicated in all cases of gas gangrene in which the patient is not moribund, the mention of other measures less frequently practised, should not be omitted. Chief of these is treatment designed to increase the blood supply in the infected area. This may be accomplished either by cataplasm or by enveloping the limb with continued hot applications, always taking care to avoid constriction. By increasing the blood supply and therefore the oxygen content, the activity of anaerobic organisms is correspondingly curtailed. The Bier method of constriction, well on the proximal side of the infection, has also had at limited trial, chiefly by its author and a few others in German hospitals, with a certain amount of success. In still other cases, through special apparatus, devised in German hospitals, intermittent or rhythmic constriction is applied to the affected extremity with more satisfactory results than when the constriction is stationary. The writer has failed to find any mention of the application of constriction to the treatment of gas gangrene by any member of the Allies. For this reason the actual value of this method is more difficult to estimate than the value of other methods of treatment which were in general use bv friend and foe alike.

Theoretically the insufflation or injection of oyxgen into the infected field ought to inhibit, if not actually paralyze, anaerobic activity. Because of the impossibility, however, of oxygen so injected coming into direct contact with all specific organisms, this method proved of little value. Furthermore patients so treated are exposed to the danger of the entrance of gas directly into venous channels with, at times, a fatal result.

During the war, transfusion was occasionally given and while its value was not generally admitted, favorable reports were cited in a limited number of cases.



The prognosis of gas gangrene depends upon the nature of the specific organisms, their association with pyogenic organisms, and the stimulating or inhibitive effect of each upon the others' activities. Local conditions and the type of a warfare also influence the prognosis. In the discussion of the pathology and symptoms of gas gangrene, these factors received due consideration. The question of the location of the infection, irrespective of its type, modifies the prognosis. Cases in which the trunk or buttocks are primarily involved, having a mortality of about 30 per cent, are more serious than infections of the extremities in which the prognosis becomes more favorable as the tip of the extremity is approached.


(1) Douglas, S. R., Fleming, A., and Colebrook, L.: Studies in Wound Infections; on the question of Bacterial Symbiosis in Wound Infections. Lancet, London, April 21, 1917, I, 604.
(2) Jablons, Benjamin: Gas Gangrene. New York Medical Journal, 1919, cx, December 20, 1914.
(3) Weinberg, M. and Séguin, P.: Ètude sur la gangrène gazeuse. Annales de l'Institut Pasteur, Paris, 1917, xxxi, No. 9, 442.
(4) Ivens, Frances: A Clinical Study of Anaerobic Wound Infection with an Analysis of 107 Cases of Gas Gangrene. British Medical Journal, London, December 23, 1916, ii, 872.
(5) Mullally, G. T. and McNee, .J. W.: A Case of Gas Gangrene Exhibiting Unusual Proofs of Blood Infection. British Medical Journal, London, April 1, 1916, i, 478.
(6),Weinberg, M. and Sèguin, P.: Essais de sérothérapie de la gangrène gazeuse chez I'homme. Comptes rendus des sèances de l'académie des sciences, Paris, 1917, clxv, No. 5, 199.
(7) Léclainche, E. and Valée, H.: The Specific Serum Treatment of Wounds. Journal of Comparative Pathology and Therapeutics, Edinburgh and London, 1916, xxix, No. 4, 283.
(8) Weinberg, M.: Bacteriological and Experimental Researches on Gas Gangrene. Proceedings of the Royal Society of Medicine, London, 1915-16, ix, Occasional Lecture, March 10, 1916, 119.
(9) Ivens, Frances: The Preventive and Curative Treatment of Gas Gangrene by Mixed Serums. British Medical Journal, London, October 19, 1918, ii, 425.
(l0) Coenen, H.: Ein Rückblock auf 20 Monate feldärztlicher Tatigkeit, mit besonderer berücksictigung der gasplegmane. Beiträdge zur klinischen Chirurgie, Tübingen, 1916, ciii, 397, 463.
(11) Aschoff, L.: Ueber bakteriologische Befunde bei den Gasoedemen. Deutsche medizinische Wochenschrift, Leipzig and Berlin, February 14, 1918, xliv, 172.