antitoxin to neutralize the toxin after it is formed.

Transcription

1 PROPHYLAXIS OF EXPERIMENTAL GAS GANGRENE IN MICE WELTON I. TAYLOR AND MILAN V. NOVAK Department of Bacteriology, University of Illinois College of Medicine, Chicago 1i, Illinois Received for publication December 26, 1950 The severity of clinical gas gangrene and the attendant high mortality have caused these clostridial infections to become potentially fatal sequellae to seemingly innocuous wound infections. The etiological agents involved have been proved to be susceptible to the inhibitory action of penicillin in vitro by Fleming (1941), Abraham et al. (1941), and Hobby et al. (1942). Penicillin susceptibility in vivo for the clostridia has been demonstrated by McIntosh and Selbie (1942), Hac (1944), Hac and Hubert (1943), Dawson et al. (1943), and McKee et al. (1943). The development of procaine penicillin, making possible prolonged therapeutic blood and tissue levels following a single injection, has been reported by Herrell et al. (1947) and Sullivan et al. (1948). The possibility of penicillin action against the toxigenic organisms in tissue during the incubation period of this disease suggested the use of this drug as a prophylactic agent. Penicillin would be especially valuable in minor injuries or in surgery when there would be hesitancy in using antitoxin because of the danger of foreign protein sensitization. The use of penicillin to inhibit growth and multiplication of gas bacilli and thus prevent toxin formation in the tissue would seem more rational than the use of antitoxin to neutralize the toxin after it is formed. The results obtained with mice, as shown in the data presented, demonstrate a definite prophylactic result. The effect of penicillin procaine alone and in combination with polyvalent gas bacillus antitoxin was studied in mice experimentally infected with single strains of gas bacilli. Action against infections caused by more than one organism growing simultaneously in a focus was not investigated. MATERIALS AND METHODS The etiological agents of clinical gas gangrene that were used in this investigation were Clostridium histolyticum (University of Chicago culture collection), and Clostridium bifermentans (sordelli) and Clostridium novyi (University of Indiana, courtesy of Dr. L. S. McClung, strains 143 and 107, respectively). Sporulation of these organisms occurred in 12 days on trypticase soy broth plus 1.5 per cent agar in 8-oz bottles laid flat for maximum surface and incubated in a Brewer anaerobic jar. Spore suspensions were made by washing the surface of the agar with sterile distilled water and heating it in a water bath at 80 C for 30 minutes to destroy toxin and vegetative cells. Dilutions in triplicate were made of the spore suspensions and added to trypticase soy broth agar "deeps" in Fisher's agar slant tubes. A layer of stratifying agar was poured over the solidified "deeps" to ensure anaerobiosis, and the appearance of macroscopic colonies gave the counts of only the viable spores. Spore suspensions were stored at 4 C when not in use. 571

2 572 WELTON I. TAYLOR AND MILAN V. NOVAK [VOL;. 61 Pathogenicity tests were made on white mice with spores and 5 per cent CaCls to determine dosages of LD80 to LD,oo. All subsequent infections were produced with the same number of spores for each species. The effective dosage of C. histolyticum was 42,500 spores; C. bifermentam& was exceedingly pathogenic for mice and only 10 spores were necessary for an LD1oo dose; and C. novyi infections were obtained using 2,700,000 spores. Spore suspensions were diluted with an equal volume of 10 per cent CaCl2, so that the spores were in 5 per cent CaCI2, and 0.1 ml was injected intramuscularly into the flexor surface of the left hind leg of Harlan strain male albino mice of 15 to 20 g in weight. Preliminary trials of the infective dose of spores were made on mice, using spores alone, and the mice remained asymptomatic, proving the absence of a lethal dose of preformed toxin in the spore suspensions. Mice subjected to 0.1 ml of 5 per cent CaCl2 survived, proving a lack of lethal toxicity in the necrotizing agent alone. A series of experiments was conducted with C. bifermentans in which was used aseptically removed mouse hind leg muscle tissue ground in a microblender and injected with the spores to produce necrosis. The results obtained were similar to those obtained with the same organism and CaCl2, proving that the necrosis itself is important, rather than the agent producing it, in experiments of this type. Several penicillin and antitoxin compounds were used since previous investigation had shown that the results varied with the type of penicillin compound used (Taylor and Novak, 1951). An aqueous procaine penicillin G, potassium penicillin compound, in which the ratio of procaine penicillin to potassium penicillin was 3:1, was used in the experiments reported in tables 1, 2, 3, 5, and 6. In table 4 the results of using procaine penicillin G in sesame oil plus 2 per cent aluminum monostearate are shown. The dosages used were based on mouse-toman body weights of a 17.5-g mouse to a 70-kg man. On this basis 150 units of penicillin in the mouse are equivalent to 600,000 units in the man. Dilutions of aqueous penicillin were made with sterile distilled water, and of the sesame oil preparation with a sterile menstruuml of sesame oil plus 2 per cent aluminum monostearate. The use of specific antitoxin was included to investigate the results of combined antitoxin-penicillin prophylaxis and to compare it with penicillin alone. "Gas gangrene antitoxin polyvalent" (Lederle), which contained 26,000 units in 13.8-ml (measured) volume, was used in 0.1-ml amounts, as shown in table 5. Although it is recognized that in such a polyvalent antitoxin the correlation of unitage to volume is inaccurate because of varying titers in the sera from which the mixture is compounded, a theoretical unitage based on the amount used and the total unitage contained for each specific toxin is hereby given. One-tenth ml should contain 21.7 units of C. histolyticum antitoxin and 10.9 units each of C. bifermentans and C. novyi antitoxin. British clinicians (Medical Research Council, 1943) suggested a prophylactic dose of antisera which did not include antitoxin for C. histolyticum or C. bifer- 1 Provided by Abbott Laboratories, North Chicago, Illinois.

3 1951] PROPHYLAXS OF EXPERIMENTAL GAS GANGRENE IN MICE 573 mentans, but which contained 3,000 international units of C. novyi antitoxin. The suggested dosage for prophylaxis with the polyvalent antitoxin used in these experiments was 4 vials.2 A lack of unanimity as to what constitutes adequate antitoxin prophylaxis exists in the literatu.re, so the dosage actually used in these experiments was based on the suggested dosage of the antitoxin used. On a mouse-to-man weight basis, mice should have received 1.5 units of antitoxin for C. novyi and C. bifermentan,s and 3 units of C. histolyticum antitoxin; however, the larger volume was given in order to minimize the possibility of injecting less than the suggested prophylactic dose. The time elapsing between the infective dose of spores and the administration of prophylactic agents has been shown by other investigators to affect mortalities (Novak, Goldin, and Taylor, 1949; Chain et al., 1940; McIntosh and Selbie, 1943), so that the start of prophylaxis was varied from immediately to 3-, 6-, and 24-hour intervals after infection. Prophylaxis consisted of a single injection, except as is noted in group G, table 2, in which the animals often died in less than 24 hours if they were not treated. This necessitated a multiple injection. As the experiments shown in tables 1 to 3 progressed, it was apparent that group F, receiving prophylaxis at 6 hours after the infective injection, was of sufficiently low mortality to give a comparison with the control group A in all cases, and thereafter only groups A and F were used for each organism to determine the effectiveness of the other compounds used, as shown in tables 4 to 6. The relative inefficiency of systemic penicillin against infections that are localized in necrotic tissue has been theorized upon by other authors (Jeffrey, 1943; Rammelkamp and Keefer, 1943; Smith, 1949), and the data shown in group C, tables 1 to 3, illustrate the effect when the prophylactic dose is administered at a site other than that of the initial infection. Group D, tables 1 to 3, shows the effects of increasing the dosage 2-fold on these infections. As a result of the difference in the rates of excretion of penicillin from the mouse as compared to man (Eagle et al., 1949) a 20-fold increase in penicillin dosage was investigated, and the data are shown in table 6. The mice employed in these experiments were routinely held for 20 days to determine survival rates and times. The only exception to this was the first of the three series comprising the data shown in table 1, in which only a 15-day holding period was observed. Experimental animals were done in several series at different times to reduce unusual variations in results. All data were statistically analyzed (Snedecor, 1946) for the significance in lowered mortalities and for increased time to death. The P value (likelihood of difference arising through chance alone) of the mortality of each group as compared with controls (group A) was obtained from a chi-square test of independence fourfold table including Yates's correction. P values of 0.05 or less are significant. The analysis of difference in time to death in days between each group and controls includes the mean time to death in days and the P value obtained from Fisher's "t" table. P values of 0.05 or less are significant. 2 Brochure, Lederle: polyvalent gas gangrene antitoxin 50 BPC 2/47.

4 574 WELTON I. TAYLOR AND MILAN V. NOVAK [VOL.. 61 EXPERIMENTAL RESULTS AND DISCUSSION The results obtained with penicillin prophylaxis of C. histolyticum infection (table 1) showed a slight decrease in mortality provided the penicillin was given in 6 hours or less after infection. None of these mortalities was significant, however, although a significant increase in time to death is noted in groups B, D, and F. This highly proteolytic organism produced the infection most difficult to prevent of any of the organisms used, the animals being practically liquefied internally. The prolongation of life in some groups shows that some effectiveness may be ascribed to penicillin, however. Penicillin prophylaxis of C. bifermentans infections (table 2) showed a significant decrease in mortalities when prophylaxis was begun at 3 hours or later. It is possible that the rate of excretion of the penicillin leaves the tissue level too low to combat the toxin-producing vegetative cells when the penicillin is given GU GROUP TABLE 1 Mortality in penicillin-treated mice infected with C. histolyticum PENICILLIN MORTALITY TIME TO DEATH ~~~~~~~ANIMALS ~~~~~~~TOTAL Units Hours before % P values*t Mean days P value Ot A _ 1.9 B > <0.05 C > >0.30 D > <0.05 E > >0.05 F > <0.05 G >0.50 * Groups compared with group A. t P values taken from chi-square 4-fold table including Yates's corrections; value less than 0.05 indicates significant difference. t P values taken from Fisher's "t" table; value less than 0.05 significant. Penicillin injected IM in opposite leg rather than infected leg. immediately after the infective dose of spores. The spores can be assumed to be resistant to penicillin (Curran and Evans, 1945), whereas, if allowed to germinate (Russell, 1927) and then subjected to the penicillin administered, the high level is available to inhibit the penicillin-susceptible cells, which are in their active growth stage. The time to death of all but group G is significant, again indicating a probable inhibition of toxigenic organisms. In practice one might find this would allow for the regeneration of new tissue following necrosis to the extent that the anaerobic environment necessary for the proliferation of these organisms was no longer present at the time tissue levels of penicillin had dropped below inhibitory levels. Because of the massive dosages of organisms given experimentally and with the severe and large areas of necrosis produced by the CaCI2, the survival of animals must indicate a diminution of organisms to almost none, if not actual sterility.

5 1951] PROPHYLAXIS OF EXPERIMENTAL GAS GANGRENE IN MICE In group G, the multiple injection produced the lowest mortality, but the mean time to death indicated the all-or-nothing nature of this infection in that the animals either were saved or died with no significant lengthening of time to death. TABLE 2 Mortality in penicillin-treated mice infected with C. bifermentane PENICILLIN MORTALITY TIME TO DEATH TOTAL ANIMALS - _- _ Units Hours before % P values*t Mean days P value*t A B > <0.01 C > <0.02 D > <0.01 E < <0.01 F < <0.01 Gil < >0.05 * Groups compared with group A. t P values taken from chi-square 4-fold table including Yates's corrections; value less than 0.05 indicates significant difference. $ P values taken from Fisher's "t" table; value less than 0.05 significant. Penicillin injected IM in opposite leg rather than infected leg. I Because deaths occur during the first 24 hours, multiple injections were used in this group, 150 unit dosages being given at 6 and at 48 hours after the infective dose of spores. GROUP TABLE 3 Mortality in penicillin-treated mice infected with C. novyi PENICILLIN MORTALIY TM TO DEATH Units Hours before TOTAL ANIMALS' % P valuet Mean days P value*t A B > >0.1 CQ > >0.5 D > >0.1 E > >0.3 F < >0.3 G < >0.2 * Groups compared with group A. t P values taken from chi-square 4-fold table including Yates's corrections; value less than 0.05 indicates significant difference. $ P values taken from Fisher's "t" table; value less than 0.05 significant. Penicillin injected IM in opposite leg rather than infected leg. 575 C. novyi (table 3) produced an infection that illustrated graphically the effect of time lapse in prophylaxis. The control animals (group A) had a mean time to death of 6 days, considerably longer than the time that the therapeutic blood and tissue levels of aqueous procaine penicillin and potassium penicillin could be expected to be maintained. The higher mortalities in groups B to E tend to

6 576 WELTON I. TAYLOR AND MILAN V. NOVAK [VOL. 61 verify this, as when the penicillin is given at 6 or 24 hours (groups F and G) the mortality rates become significantly lower than in the controls. The increase in time to death was not significant in this infection. The effectiveness of procaine penicillin G in sesame oil plus 2 per cent aluminum monostearate for the three clostridia is shown in table 4. C. histolyticum was not significantly inhibited, but all species showed a prolonged time to death. For both C. bifermentans and C. novyi infections mortalities were much lower than with the aqueous procaine penicillin, and the time to death for animals infected with TABLE 4 Prophylaxis of gas gangrene with procaine penicillin in oil PENICILLIN MORTALITY TIME TO DEATH ORGANISM GROUP TOTAL Uis Hours ANIMALS Mean Pvau Units [before % P value days C. histolyticum A F > <0.05 C. bifermentans A F < <0.01 C. novyi A F < <0.05 TABLE 5 -Prophylaxis of gas gangrene with antitoxin and penicillin PENICILLIN AND ANTITOXIN MORTALITY TIME TO DEATH ORGANISM GRO'UP TOTAL Hours ANIMALS Ma Units before % P value Man P value C. histolyticum A F ml < <0.01 C. bifermentans A F ml < <0.01 C. novyi A F ml < >0.5 C. novyi is significantly greater with this compound. The penicillin in oil is reputed to provide a therapeutic level of longer duration than that of the aqueous compound, and this may well be responsible for the lowered mortalities. The combined effect of aqueous procaine penicillin and polyvalent antitoxin is shown in table 5. This combination secured the lowest mortality for C. histolyticum infections. The prophylactic dose consisted of 0.2 ml of a mixture of antitoxin and penicillin of equal volume. A pronounced delay in time to death was observed. Mice infected with C. novyi and C. bifermentans were observed to have

7 19511 PROPHYLAXIS OF EXPERIMENTAL GAS GANGRENE IN MICE a significant reduction in mortality, but the increased time to death of mice infected with C. novyi was not significant. The rapid rate of excretion of penicillin by the mouse as compared to man, reported by Eagle, Fleischman, and Musselman (1949), prompted an experiment involving a 20-fold increase in penicillin, which should prolong the blood and tissue level for an additional time. The results of this increased dosage are shown in table 6. The mortality for C. histolyticum was reduced, but the time to death was not significantly delayed. The mortality for C. bifermentans was the same as in table 2, but was not significant here, probably because of the reduction in the number of experimental animals used. The time to death was delayed. C. novyi infections showed a lower mortality but no appreciable delay in time to death. From these experiments it is obvious that many variables affect the degree to which penicillin prophylaxis is efficient in preventing the onset of these clostridial diseases. C. histolyticum infections were most responsive to a mixture of TABLE 6 Prophylaxis of gas gangrene with massive penicillin dosages PENICIILIN MORTALITY TME TO DEATH ORGANISU GROUP TOTAL Units beoure ANILS P value Mdeays P value C. histolyticum A F 3, < >0.05 C. bifermentans A F 3, > <0.03 C. novyi A F 3, < >0. 1 penicillin and antitoxin (table 5), or a massive dose of penicillin alone secured satisfactory results (table 6). C. bifermentans infections were well controlled by the use of penicillin in oil (table 4) or with the antitoxin-penicillin mixture (table 5). Penicillin in oil was most effective in infections caused by C. novyi (table 4). The failure of systemic penicillin given at a site other than that of the infection has been noted in previous experiments on animals by the authors and by Hac'(1944). The lack of complete protection afforded by penicillin in the therapy of clinical cases of gas gangrene has been explained as a matter of nonpenetration of systemically administered penicillin into local necrotic areas not being serviced by the blood (Smith, 1949). The advantage of the protection afforded by prophylaxis with penicillin can be demonstrated, however, when one considers the report by Langley and Winkelstein (1945) of 96 gas gangrene cases, 75 per cent of which showed a reaction to horse serum after both skin and ophthalmic tests for horse serum sensitivity had proved negative and in which one death from anaphylactic shock occurred as a result of the administration of polyvalent antitoxin. 577

8 578 WELTON I. TAYLOR AND M1LAN V. NOVAK [VOL. 61 The experiments presented here have been designed to afford the maximum opportunity for the toxigenic organisms to find an environment suitable for proliferation and toxin production, and a sufficient nuimber of organisms were injected to produce a high mortality rate. The use of a single dose of penicillin, without debridement or other surgical measures, without the aid of the defense mechanisms of the tissue in the destruction of the organisms because of the extreme trauma produced, and with the increased rate of excretion of the drug from the mouse as compared to man, lends added weight to the statistically significant reduction in mortalities and prolonged time to death obtained. SUNMMARY The prophylactic use of procaine penicillin compounds caused a significant reduction in mortality in experimental clostridial infections in mice. Systemic penicillin afforded less protection than did local penicillin. The time at which the prophylaxis was begun after the initiation of the infection was an important variable in determining the degree of protection afforded by penicillin. The use of a mixture of penicillin and polyvalent antitoxin was more effective in the prophylaxis of infections caused by Clostridium histolyticum, but no more effective than penicillin alone in infections caused by Clostridium bifermentans and Clostridium novyi. The maintenance of inhibitory tissue levels either by the administration of massive doses of aqueous procaine penicillin or by penicillin in oil caused a marked decrease in the mortalities of test animals. REFERENCES ABRAHAM, E. P., CHAIN, E., FLETCER, C. M., GARDNER, A. D., HEATLEY, N. G., JENNINGS, M. A., AND FLOREY, H. W Further observations on penicillin. Lancet, 2, CHAIN, E., FLOREY, H. W., GADNER, A. D., JENNINGS, M. A., ORR-EWING, J., SANDERS, A. F., AND HEATLEY, N. G Penicillin as a chemotherapeutic agent. Lancet 2, CURRAN, H. R., AND EVANS, F. R Penicillin as a sporicidal agent. Proc. Soc. Exptl. Biol. Med., 58, DAWSON, M. H., HOBBY, G. L., MEYER, K., AND CHAFFEE, E Penicillin as a chemotherapeutic agent. Ann. Internal Med., 19, EAGLE, H., FLEISCHMAN, R., AND MUSSELmAN, A. D The serum concentration of penicillin G in mice, rabbits, and men after its intramuscular injection in aqueous solution. J. Bact., 57, FLEMING, A Mode of action of chemotherapeutic agents. Lancet, 2, 761. HAC, L. R Experimental Clostridium welchii infection. IV. Penicillin therapy. J. Infectious Diseases, 74, HAC, L. R., AND HUBERT, A. C Penicillin in treatment of experimental Clostridium welchii infection. Proc. Soc. Exptl. Biol. Med., 53, HERRELL, W. E., NICHOLS, D. R., AND HEILMAN, F. R Procaine penicillin G (duracillin): a new salt of penicillin which prolongs action of penicillin. Proc. Staff Meetings, Mayo Clinic, 22, HOBBY, G. L., MEYER, K., AND CHAFFEE, E Activity of penicillin in vitro. Proc. Soc. Exptl. Biol. Med., 50, JEFFREY, J. S Penicillin: discussion at RSM. Brit. Med. J., 2, 656.

9 19511 PROPHYLAXIS OF EXPERIMENTAL GAS GANGRENE IN MICE 579 LANGLEY, F. H., AND WINKELSTEIN, L. B Gas gangrene. J. Am. Med. Assoc., 128, MCINTOSH, J., AND SELBIE, F. R Zinc peroxide, proflavine, and penicillin in experimental Cl. welchii infections. Lancet, 2, MCINTOSH, J., AND SELBIE, F. R Combined action of antitoxin and local chemotherapy on Cl. welchii infection in mice. Lancet, 2, McKEs, C. M., HAMRE, D. M., AND RAKz, G The action of antibiotics on organisms producing gas gangrene. Proc. Soc. Exptl. Biol. Med., 54, Medical Research Council War Wounds Committee and Committee of London Sector Pathologists 1943 Notes on gas gangrene. Prevention, diagnosis, treatment. Med. Research Council Brit., War Memo. Revised, 2, NOVAK, M., GOLDIN, M., AND TAYLOR, W. I Tetanus prophylaxis with penicillinprocaine G. Proc. Soc. Exptl. Biol. Med., 70, RAMMHLKAMP, E. H., AND KEEFER, C. S Absorption, excretion and distribution of penicillin. J. Clin. Investigation, 22, RUSSELL, D. S Local fate of tetanus spores inoculated into guinea-pigs. Brit. J. Exptl. Path., 8, SMITH, L. DES Clostridia in gas gangrene. Bact. Revs., 13, SNEDECOR, G. W Statistical methods. Iowa State College Press, Ames, Iowa. SULLIVAN, N. P., SYMMES, A. T., MILLER, H. C., AND RHODEHAMEL, H. W., JR New penicillin for prolonged blood levels. Science, 107, TAYLOR, W. I., AND NOVAK, M. V Prophylaxis of tetanus with procaine-penicillin. Ann. Surg. In press. Downloaded from on October 3, 2018 by guest