Prognostic Indicators in Experimental Bacteroides Peritonitis Serum Robert J.
Sharbaugh, PhD,
William M.
Rambo, MD
Serum lysozyme and hemolytic complement (CH100) levels measured in dogs with experimental Bacteroides peritonitis. The CH levels showed little change in surviving animals. Nonsurvivors showed a moderate decrease in complement levels shortly after contamination. Both surviving and nonsurviving animals showed a slight initial decrease in lysozyme levels shortly after contamination. In surviving dogs this was followed by an increase to normal levels. In nonsurvivors, levels continued to increase, reaching a threefold magnification just prior to death. As a result of antibiotic therapy, CH100 levels exhibited no major changes; however, dogs deprived of antibiotic showed noticeable and persistent increases in lysozyme levels while treated animals showed only a mild elevation in lysozyme levels. The changes in the level of serum lysozyme may be a good indicator of antibiotic efficacy and approaching death from Bacteroides peritonitis. were
One
of the most serious consequences of peritonitis is the large number of endotoxin-producing Gram-neg¬ ative bacteria released from the gut into the peritoneal cavity. Though Escherichìa coli is the predominant aerobe isolated in peritonitis, there are much greater numbers of anaerobic bacteria, notably Bacteroides,' that may also gain entry into the peritoneal cavity. Entrance into the circulation of the endotoxin released by these bacteria may produce the same systemic signs seen in classical ex¬ perimental endotoxic shock such as fever, leukopenia, and
hypotension.
Endotoxin has been shown to cause complement inhibi¬ tion.2 ! Also, the transient leukopenia associated with enAccepted for publication Feb 13, 1975. From the Department of Surgery, Medical University of South Carolina,
Charleston.
Reprint requests to Department of Surgery, Medical University of South
Carolina,
80 Barre
St, Charleston, SC 29401 (Dr. Sharbaugh).
dotoxemia may be a result of the complement-endotoxin lysis of circulating neutrophils. If so, such cellular destruc¬ tion would theoretically release numerous lysosomal en¬ zymes into the circulation. This investigation focuses on the effect of experimental anaerobic Gram-negative peri¬ tonitis on the levels of serum complement (CH,„„) and ly¬ sozyme and on their potential use as biologic indicators of uncontrolled sepsis. MATERIALS AND METHODS
Twenty-seven adult mongrel dogs of both sexes were used in study. Body weights ranged from 13 to 18 kg. Throughout the study, all animals were maintained on a standard laboratory diet. Experimental fecal peritonitis was produced using a standard¬ ized sterile fecal suspension and fragilis. The strain of fragilis used was sensitive to clindamycin (2^g) as determined by disk sensitivity. The method employed was a modification of that described by Sharbaugh and Rambo.4 The peritoneal cavity of each dog was contaminated through a 10-cm midline incision with the use of a mixture containing equal quantities of the sterile fe¬ cal suspension and 1 ml/kg body weight of the washed fragilis suspension (lx 10"' per milliliter). During the operation, each dog was given 20 ml/kg body weight of lactated Ringer solution in¬ travenously. One group of nine dogs served as controls and re¬ ceived no treatment. Two hours after contamination, all dogs in the treatment group received thorough peritoneal lavage with 100 ml saline per kilogram of body weight. These animals were then divided into two groups. Eleven dogs received peritoneal lavage only. Seven dogs received clindamycin (40 mg/kg) intravenously at the time of peritoneal lavage and intramuscular injections of 20 mg/kg every eight to ten hours thereafter for a total of four dos¬ this
ages.
Following peritoneal lavage, the incisions were closed with in¬ terrupted sutures and each dog followed up for 48 hours or until death, whichever came first. Serum lysozyme and CH„10 levels were measured using radial immunodiffusion kits.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Auckland User on 05/27/2015
(Fig 2). Initial levels, averaging 28.7 CH,„„ units/ml, de¬ only slightly. There was, however, a more appar¬ ent drop to 24.6 units/ml by the 30th hour, which gradu¬ ally returned to near-normal levels 48 hours after creased
Fig 1.—Efficacy of treatment regimens on survival of dogs with experimental Bacteroides peritonitis. Numbers in parentheses represent survivors
over
total number in group.
RESULTS The survival of dogs in the various treatment groups are shown in Fig 1. Dogs deprived of both lavage and clin¬ damycin showed a low survival of only 33% with three of nine dogs surviving contamination. Dogs receiving only peritoneal lavage showed an apparent increase in survival with eight of 11 animals (73%) surviving, while dogs treated with both peritoneal lavage and clindamycin dis¬ played 100% survival. Chi-square analysis showed that treatment with lavage and clindamycin produced a signif¬ icant improval in survival over those for the controls
(P < .05). Although dogs receiving only peritoneal lavage appeared to exhibit much higher survival than those re¬ ceiving no treatment, the difference was not statistically significant. Following the induction of peritonitis, circulating levels of serum hemolytic complement showed little change for the first 24 hours in animals surviving peritoneal insult
contamination. In contrast, nonsurvivors showed a statis¬ tically significant (P < .02) decrease in serum complement levels shortly after contamination, with levels dropping to 21.7 CH1I10 units/ml eight hours after insult. This initial drop was followed by a gradual rise to near-normal levels by the 24th hour, only to decrease again within a few hours of death. Both surviving and nonsurviving animals displayed a slight decrease in circulating lysozyme levels following the onset of peritonitis (Fig 2). Lysozyme levels of approx¬ imately ^g/ml to l^g/ml dropped to between 8µg/ml and 9µg/ml eight hours after contamination. In surviving dogs this was followed by an increase to 14^g/ml by the 24th hour and a return to near-normal levels by 48 hours. However, following the slight initial decrease in serum ly¬ sozyme levels of nonsurviving dogs, enzyme levels contin¬ ued to increase rapidly, reaching in excess of 32ug/ml just prior to death. This increase was significant at the 5% level. The effectiveness of ongoing antibiotic therapy was readily apparent as evidenced by serum lysozyme levels postcontamination (Fig 3). Dogs in the group receiving only peritoneal lavage displayed very noticeable increases in serum lysozyme levels. Initial levels of ll^g/ml rose to an average of ^g/ml and 25^g/ml by the 24th and 30th hour, respectively. Statistical analysis showed these levels to be significant at the 1% level. Although enzyme activity then began to slowly recede, highly elevated lev¬ els averaging 22µg/ml continued to be present 48 hours after contamination. In contrast, dogs in the group receiv¬ ing clindamycin in addition to peritoneal lavage showed an initial decrease in activity followed by a mild elevation to l^g/ml 24 hours after contamination. By the 30th hour, enzyme levels had returned to normal, where they remained until the animal was killed. Serum complement levels failed to exhibit any major changes as a result of antibiotic therapy (Fig 3). While dogs receiving only lavage displayed an observable de¬ crease in complement titer to 23.4 units/ml 30 hours after insult, the drop was very gradual and was not statistically significant. Animals receiving lavage and clindamycin showed a similar decrease of small magnitude, with titers dropping to 24.8 units/ml. However, while complement lev¬ els had returned to normal in this group by 48 hours, dogs in the lavage only group continued to display a mild, yet definite, decrease in serum complement activity. COMMENT
Despite recent advances in the therapy of bacterial peritonitis, management of the septic complications asso¬ ciated with gastrointestinal trauma often proves difficult. Recently, Füllen et al·' reported septic complications to be a major cause of deaths associated with abdominal vis¬ ceral injuries. Virtually all of the Gram-negative bacilli commonly in-
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Auckland User on 05/27/2015
f
Ê 30-
30
•
yo
X Ü
„__
Ü
•—·
_30I
!»-
Lavage only Lavage + Clindamycin
30
Survivors •—· Non Survivors .--.
/
—·
3 20-
€>
•
>» N
>»
E
E
2
S
lo-r -1-
8
24 30 Hours After Contamination
Ib
Fig 2.—Serum complement and lysozyme levels in surviving nonsurviving dogs with experimental Bacteroides peritonitis.
and
volved in the infectious process possess endotoxins of varying potency. Of the aerobes and anaerobes, E coli and fragilis, respectively, are the two most often isolated in cases of bacterial peritonitis with, as pointed out by No¬ bles,1 the Bacteroides comprising the greater majority of all colon bacteria. In addition, the Bacteroides have just recently gained recognition for their serious and often fa¬ tal pathogenic effects. The potentially serious nature of Gram-negative infec¬ tions, especially those occuring as a result of perforation of the gastrointestinal tract, points to the need for addi¬ tional indicators of oncoming or uncontrolled sepsis. In the present investigation, two such potential indicators, serum hemolytic complement and lysozyme levels, were examined and the effect that fragilis-inaxiced peri¬ tonitis had on each was observed. The inhibition of serum complement by purified lipopolysaccharide (LPS) has been known for some time. Stud¬ ies such as those by Gilbert and Braude- and Spink et al6 have shown experimental endotoxemia to cause a rapid and significant fall in serum hemolytic complement levels. However, all of these studies employed the injection of pu¬ rified LPS into the circulation, a condition unlike that which occurs in the clinical course of peritonitis. In the present study dogs were subjected to experimental Bacte¬ roides peritonitis, which, under the conditions employed, is quite similar to clinical Gram-negative peritonitis. Ani¬ mals destined to survive the infection exhibited little change in serum CH„,„ levels for the first 48 hours. How¬ ever, shortly after contamination there was a pronounced drop of nearly 25% in the complement levels of nonsurvi-
io
0
8
24 30 Hours After Contamination
48
Fig 3.—Effect of antibiotic treatment on serum complement and lysozyme levels in dogs with experimental Bacteroides peritonitis. vors
that continued to be
expressed until death of the ani¬
mal, indicating apparent continuous state of endotoxemia. This decrease was not as great as that reported by From et al7 or by Spink and his co-workers.6 However, their studies employed lethal intravenous injections of pu¬ rified E coli endotoxin, thus producing a severe endotoxemia in a matter of minutes. In our studies, the endotoxin produced by fragilis was probably not as potent as E coli LPS. In addition, it gained access to the circulation over a much longer period of time. The present data offer no explanation for the source of increased lysozyme in septic animals and we can only as¬ sume that this increase in activity is due, either directly or indirectly, to endotoxemia. However, early studies by Ribble8 showed intravenous administration of purified en¬ dotoxin in rabbits caused a rise in plasma lysozyme pre¬ sumably released from damaged or ruptured granulocytes. Also, similar studies by Janoff et al·' linked endotoxin ad¬ ministration with marked increases in circulating acid phosphatase and /3-glucuronidase. More recent studies in man also provide good evidence that plasma lysozyme is primarily a product of disintigrating neutrophils.1" Such evidence, plus the recent report by Goldstein et al" that endotoxin-activated complement can interact with human neutrophils to stimulate the release of lysosomal enzymes, would indicate that the majority of lysozyme activity ob¬ served in our studies probably stems from disrupted gran¬ ulocytes. One other possibility would involve a reduced rate of lysozyme catabolism due to renal failure as a result of sepsis. While the present studies were not designed to evaluate renal function, previous studies have shown that an
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Auckland User on 05/27/2015
and hymassive outpouring of plasma from the povolemia vascular compartment to the peritoneal cavity.1- Thus, re¬ duced renal perfusion with subsequent decreased ly¬ sozyme catabolism in the kidney must also be considered as a possible contributor to the increased lysozyme levels reported in this study. In addition, however, to the neutrophil, the lungs, spleen, kidneys, and intestines also possess significant ly¬ sozyme activity.l0 Also, Abe et al" have reported the re¬ lease of lysozomal enzymes into the circulation as a result of intestinal injury. Thus the traumatized gastrointestinal tract often affiliated with cases of bacterial peritonitis may also be a significant contributor to the elevated ly¬ sozyme levels observed in the present study. The strain of Bacteroides used in this study was sensi¬ tive to the antibiotic clindamycin. Regardless of antibiotic therapy, however, the subtle changes in CH1IHI levels after contamination were such that no effective evaluation of antibiotic efficacy could be made. In contrast, the steep rise in serum lysozyme levels among animals deprived of antibiotic therapy was readily apparent and presumably indicated continuous sepsis and endotoxemia. In conclusion, experimental Bacteroides peritonitis is accompanied by changes in the levels of both serum hemolytic complement and serum lysozyme. While the change in CH10(I was small and offered no prognostic value, the rapid rise in serum lysozyme proved a good indicator of antibiotic efficacy and approaching death of the animal.
Nonproprietary Name
experimental peritonitis produced hypotension with
a
and Trademark of
Drug
Clindamycin-CfeoCTn. This
investigation
was
supported by
the
Upjohn Co, Kalamazoo, Mich.
References 1. Nobles ER: Bacteroides infection. Ann Surg 177:601-606, 1973. 2. Gilbert VE, Braude AI: Reduction of serum complement in rabbits after injection of endotoxin. J Exp Med 116:477-490, 1962. 3. Spink WW, Potter R: Endotoxin shock and immunologic factors: Studies on complement. Fed Proc 22:628, 1963. 4. Sharbaugh RJ, Rambo WM: A new model for producing experimental fecal peritonitis. Surg Gynecol Obstet 133:843-845, 1971. 5. Fullen WD, Hunt J, Altmeier WA: Prophylactic antibiotics in penetrating wounds of the abdomen. B Trauma 12:282-289, 1972. 6. Spink WW, Davis RB, Potter R, et al: The initial stage of canine endotoxin shock as an expression of anaplylactic shock: Studies on complement titers and plasma histamine concentrations. J Clin Invest 43:696-704,1964. 7. From AHL, Gewurz H, Gruninger RP, et al: Complement in endotoxin shock: Effect of complement depletion on the early hypotensive phase. Infec Immun 2:38-41, 1970. 8. Ribble JC: Increase of plasma lysozyme following injections of typhoid vaccine. Proc Soc Exp Biol Med 107:597-600, 1961. 9. Janoff A, Weissmann G, Zweifach W, et al: Pathogenesis of experimental shock: IV. Studies on lysozomes in normal and tolerant animals subjected to lethal trauma and endotoxemia. J Exp Med 16:451-466, 1962. 10. Hansen NE, Karle H, Andersen V, et al: Lysozyme turnover in man. J Clin Invest 51:1146-1155, 1972. 11. Goldstein IM, Brai M, Osler AG, et al: Lysosomal enzyme release from human leukocytes: Mediation by the alternative pathway of complement activation. J Immunol 111:33-37, 1973. 12. Sharbaugh RJ, Rambo WM: Cephalothin and peritoneal lavage in the treatment of experimental peritonitis. Surg Gynecol Obstet 139:211-214, 1974. 13. Abe H, Carballo J, Appert HE, et al: The release and fate of the intestinal lysosomal enzymes after acute ischemic injury of the intestine. Surg Gynecol Obstet 135:581-585, 1972.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Auckland User on 05/27/2015
Sharbaugh, PhD,
William M.
Rambo, MD
Serum lysozyme and hemolytic complement (CH100) levels measured in dogs with experimental Bacteroides peritonitis. The CH levels showed little change in surviving animals. Nonsurvivors showed a moderate decrease in complement levels shortly after contamination. Both surviving and nonsurviving animals showed a slight initial decrease in lysozyme levels shortly after contamination. In surviving dogs this was followed by an increase to normal levels. In nonsurvivors, levels continued to increase, reaching a threefold magnification just prior to death. As a result of antibiotic therapy, CH100 levels exhibited no major changes; however, dogs deprived of antibiotic showed noticeable and persistent increases in lysozyme levels while treated animals showed only a mild elevation in lysozyme levels. The changes in the level of serum lysozyme may be a good indicator of antibiotic efficacy and approaching death from Bacteroides peritonitis. were
One
of the most serious consequences of peritonitis is the large number of endotoxin-producing Gram-neg¬ ative bacteria released from the gut into the peritoneal cavity. Though Escherichìa coli is the predominant aerobe isolated in peritonitis, there are much greater numbers of anaerobic bacteria, notably Bacteroides,' that may also gain entry into the peritoneal cavity. Entrance into the circulation of the endotoxin released by these bacteria may produce the same systemic signs seen in classical ex¬ perimental endotoxic shock such as fever, leukopenia, and
hypotension.
Endotoxin has been shown to cause complement inhibi¬ tion.2 ! Also, the transient leukopenia associated with enAccepted for publication Feb 13, 1975. From the Department of Surgery, Medical University of South Carolina,
Charleston.
Reprint requests to Department of Surgery, Medical University of South
Carolina,
80 Barre
St, Charleston, SC 29401 (Dr. Sharbaugh).
dotoxemia may be a result of the complement-endotoxin lysis of circulating neutrophils. If so, such cellular destruc¬ tion would theoretically release numerous lysosomal en¬ zymes into the circulation. This investigation focuses on the effect of experimental anaerobic Gram-negative peri¬ tonitis on the levels of serum complement (CH,„„) and ly¬ sozyme and on their potential use as biologic indicators of uncontrolled sepsis. MATERIALS AND METHODS
Twenty-seven adult mongrel dogs of both sexes were used in study. Body weights ranged from 13 to 18 kg. Throughout the study, all animals were maintained on a standard laboratory diet. Experimental fecal peritonitis was produced using a standard¬ ized sterile fecal suspension and fragilis. The strain of fragilis used was sensitive to clindamycin (2^g) as determined by disk sensitivity. The method employed was a modification of that described by Sharbaugh and Rambo.4 The peritoneal cavity of each dog was contaminated through a 10-cm midline incision with the use of a mixture containing equal quantities of the sterile fe¬ cal suspension and 1 ml/kg body weight of the washed fragilis suspension (lx 10"' per milliliter). During the operation, each dog was given 20 ml/kg body weight of lactated Ringer solution in¬ travenously. One group of nine dogs served as controls and re¬ ceived no treatment. Two hours after contamination, all dogs in the treatment group received thorough peritoneal lavage with 100 ml saline per kilogram of body weight. These animals were then divided into two groups. Eleven dogs received peritoneal lavage only. Seven dogs received clindamycin (40 mg/kg) intravenously at the time of peritoneal lavage and intramuscular injections of 20 mg/kg every eight to ten hours thereafter for a total of four dos¬ this
ages.
Following peritoneal lavage, the incisions were closed with in¬ terrupted sutures and each dog followed up for 48 hours or until death, whichever came first. Serum lysozyme and CH„10 levels were measured using radial immunodiffusion kits.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Auckland User on 05/27/2015
(Fig 2). Initial levels, averaging 28.7 CH,„„ units/ml, de¬ only slightly. There was, however, a more appar¬ ent drop to 24.6 units/ml by the 30th hour, which gradu¬ ally returned to near-normal levels 48 hours after creased
Fig 1.—Efficacy of treatment regimens on survival of dogs with experimental Bacteroides peritonitis. Numbers in parentheses represent survivors
over
total number in group.
RESULTS The survival of dogs in the various treatment groups are shown in Fig 1. Dogs deprived of both lavage and clin¬ damycin showed a low survival of only 33% with three of nine dogs surviving contamination. Dogs receiving only peritoneal lavage showed an apparent increase in survival with eight of 11 animals (73%) surviving, while dogs treated with both peritoneal lavage and clindamycin dis¬ played 100% survival. Chi-square analysis showed that treatment with lavage and clindamycin produced a signif¬ icant improval in survival over those for the controls
(P < .05). Although dogs receiving only peritoneal lavage appeared to exhibit much higher survival than those re¬ ceiving no treatment, the difference was not statistically significant. Following the induction of peritonitis, circulating levels of serum hemolytic complement showed little change for the first 24 hours in animals surviving peritoneal insult
contamination. In contrast, nonsurvivors showed a statis¬ tically significant (P < .02) decrease in serum complement levels shortly after contamination, with levels dropping to 21.7 CH1I10 units/ml eight hours after insult. This initial drop was followed by a gradual rise to near-normal levels by the 24th hour, only to decrease again within a few hours of death. Both surviving and nonsurviving animals displayed a slight decrease in circulating lysozyme levels following the onset of peritonitis (Fig 2). Lysozyme levels of approx¬ imately ^g/ml to l^g/ml dropped to between 8µg/ml and 9µg/ml eight hours after contamination. In surviving dogs this was followed by an increase to 14^g/ml by the 24th hour and a return to near-normal levels by 48 hours. However, following the slight initial decrease in serum ly¬ sozyme levels of nonsurviving dogs, enzyme levels contin¬ ued to increase rapidly, reaching in excess of 32ug/ml just prior to death. This increase was significant at the 5% level. The effectiveness of ongoing antibiotic therapy was readily apparent as evidenced by serum lysozyme levels postcontamination (Fig 3). Dogs in the group receiving only peritoneal lavage displayed very noticeable increases in serum lysozyme levels. Initial levels of ll^g/ml rose to an average of ^g/ml and 25^g/ml by the 24th and 30th hour, respectively. Statistical analysis showed these levels to be significant at the 1% level. Although enzyme activity then began to slowly recede, highly elevated lev¬ els averaging 22µg/ml continued to be present 48 hours after contamination. In contrast, dogs in the group receiv¬ ing clindamycin in addition to peritoneal lavage showed an initial decrease in activity followed by a mild elevation to l^g/ml 24 hours after contamination. By the 30th hour, enzyme levels had returned to normal, where they remained until the animal was killed. Serum complement levels failed to exhibit any major changes as a result of antibiotic therapy (Fig 3). While dogs receiving only lavage displayed an observable de¬ crease in complement titer to 23.4 units/ml 30 hours after insult, the drop was very gradual and was not statistically significant. Animals receiving lavage and clindamycin showed a similar decrease of small magnitude, with titers dropping to 24.8 units/ml. However, while complement lev¬ els had returned to normal in this group by 48 hours, dogs in the lavage only group continued to display a mild, yet definite, decrease in serum complement activity. COMMENT
Despite recent advances in the therapy of bacterial peritonitis, management of the septic complications asso¬ ciated with gastrointestinal trauma often proves difficult. Recently, Füllen et al·' reported septic complications to be a major cause of deaths associated with abdominal vis¬ ceral injuries. Virtually all of the Gram-negative bacilli commonly in-
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Auckland User on 05/27/2015
f
Ê 30-
30
•
yo
X Ü
„__
Ü
•—·
_30I
!»-
Lavage only Lavage + Clindamycin
30
Survivors •—· Non Survivors .--.
/
—·
3 20-
€>
•
>» N
>»
E
E
2
S
lo-r -1-
8
24 30 Hours After Contamination
Ib
Fig 2.—Serum complement and lysozyme levels in surviving nonsurviving dogs with experimental Bacteroides peritonitis.
and
volved in the infectious process possess endotoxins of varying potency. Of the aerobes and anaerobes, E coli and fragilis, respectively, are the two most often isolated in cases of bacterial peritonitis with, as pointed out by No¬ bles,1 the Bacteroides comprising the greater majority of all colon bacteria. In addition, the Bacteroides have just recently gained recognition for their serious and often fa¬ tal pathogenic effects. The potentially serious nature of Gram-negative infec¬ tions, especially those occuring as a result of perforation of the gastrointestinal tract, points to the need for addi¬ tional indicators of oncoming or uncontrolled sepsis. In the present investigation, two such potential indicators, serum hemolytic complement and lysozyme levels, were examined and the effect that fragilis-inaxiced peri¬ tonitis had on each was observed. The inhibition of serum complement by purified lipopolysaccharide (LPS) has been known for some time. Stud¬ ies such as those by Gilbert and Braude- and Spink et al6 have shown experimental endotoxemia to cause a rapid and significant fall in serum hemolytic complement levels. However, all of these studies employed the injection of pu¬ rified LPS into the circulation, a condition unlike that which occurs in the clinical course of peritonitis. In the present study dogs were subjected to experimental Bacte¬ roides peritonitis, which, under the conditions employed, is quite similar to clinical Gram-negative peritonitis. Ani¬ mals destined to survive the infection exhibited little change in serum CH„,„ levels for the first 48 hours. How¬ ever, shortly after contamination there was a pronounced drop of nearly 25% in the complement levels of nonsurvi-
io
0
8
24 30 Hours After Contamination
48
Fig 3.—Effect of antibiotic treatment on serum complement and lysozyme levels in dogs with experimental Bacteroides peritonitis. vors
that continued to be
expressed until death of the ani¬
mal, indicating apparent continuous state of endotoxemia. This decrease was not as great as that reported by From et al7 or by Spink and his co-workers.6 However, their studies employed lethal intravenous injections of pu¬ rified E coli endotoxin, thus producing a severe endotoxemia in a matter of minutes. In our studies, the endotoxin produced by fragilis was probably not as potent as E coli LPS. In addition, it gained access to the circulation over a much longer period of time. The present data offer no explanation for the source of increased lysozyme in septic animals and we can only as¬ sume that this increase in activity is due, either directly or indirectly, to endotoxemia. However, early studies by Ribble8 showed intravenous administration of purified en¬ dotoxin in rabbits caused a rise in plasma lysozyme pre¬ sumably released from damaged or ruptured granulocytes. Also, similar studies by Janoff et al·' linked endotoxin ad¬ ministration with marked increases in circulating acid phosphatase and /3-glucuronidase. More recent studies in man also provide good evidence that plasma lysozyme is primarily a product of disintigrating neutrophils.1" Such evidence, plus the recent report by Goldstein et al" that endotoxin-activated complement can interact with human neutrophils to stimulate the release of lysosomal enzymes, would indicate that the majority of lysozyme activity ob¬ served in our studies probably stems from disrupted gran¬ ulocytes. One other possibility would involve a reduced rate of lysozyme catabolism due to renal failure as a result of sepsis. While the present studies were not designed to evaluate renal function, previous studies have shown that an
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Auckland User on 05/27/2015
and hymassive outpouring of plasma from the povolemia vascular compartment to the peritoneal cavity.1- Thus, re¬ duced renal perfusion with subsequent decreased ly¬ sozyme catabolism in the kidney must also be considered as a possible contributor to the increased lysozyme levels reported in this study. In addition, however, to the neutrophil, the lungs, spleen, kidneys, and intestines also possess significant ly¬ sozyme activity.l0 Also, Abe et al" have reported the re¬ lease of lysozomal enzymes into the circulation as a result of intestinal injury. Thus the traumatized gastrointestinal tract often affiliated with cases of bacterial peritonitis may also be a significant contributor to the elevated ly¬ sozyme levels observed in the present study. The strain of Bacteroides used in this study was sensi¬ tive to the antibiotic clindamycin. Regardless of antibiotic therapy, however, the subtle changes in CH1IHI levels after contamination were such that no effective evaluation of antibiotic efficacy could be made. In contrast, the steep rise in serum lysozyme levels among animals deprived of antibiotic therapy was readily apparent and presumably indicated continuous sepsis and endotoxemia. In conclusion, experimental Bacteroides peritonitis is accompanied by changes in the levels of both serum hemolytic complement and serum lysozyme. While the change in CH10(I was small and offered no prognostic value, the rapid rise in serum lysozyme proved a good indicator of antibiotic efficacy and approaching death of the animal.
Nonproprietary Name
experimental peritonitis produced hypotension with
a
and Trademark of
Drug
Clindamycin-CfeoCTn. This
investigation
was
supported by
the
Upjohn Co, Kalamazoo, Mich.
References 1. Nobles ER: Bacteroides infection. Ann Surg 177:601-606, 1973. 2. Gilbert VE, Braude AI: Reduction of serum complement in rabbits after injection of endotoxin. J Exp Med 116:477-490, 1962. 3. Spink WW, Potter R: Endotoxin shock and immunologic factors: Studies on complement. Fed Proc 22:628, 1963. 4. Sharbaugh RJ, Rambo WM: A new model for producing experimental fecal peritonitis. Surg Gynecol Obstet 133:843-845, 1971. 5. Fullen WD, Hunt J, Altmeier WA: Prophylactic antibiotics in penetrating wounds of the abdomen. B Trauma 12:282-289, 1972. 6. Spink WW, Davis RB, Potter R, et al: The initial stage of canine endotoxin shock as an expression of anaplylactic shock: Studies on complement titers and plasma histamine concentrations. J Clin Invest 43:696-704,1964. 7. From AHL, Gewurz H, Gruninger RP, et al: Complement in endotoxin shock: Effect of complement depletion on the early hypotensive phase. Infec Immun 2:38-41, 1970. 8. Ribble JC: Increase of plasma lysozyme following injections of typhoid vaccine. Proc Soc Exp Biol Med 107:597-600, 1961. 9. Janoff A, Weissmann G, Zweifach W, et al: Pathogenesis of experimental shock: IV. Studies on lysozomes in normal and tolerant animals subjected to lethal trauma and endotoxemia. J Exp Med 16:451-466, 1962. 10. Hansen NE, Karle H, Andersen V, et al: Lysozyme turnover in man. J Clin Invest 51:1146-1155, 1972. 11. Goldstein IM, Brai M, Osler AG, et al: Lysosomal enzyme release from human leukocytes: Mediation by the alternative pathway of complement activation. J Immunol 111:33-37, 1973. 12. Sharbaugh RJ, Rambo WM: Cephalothin and peritoneal lavage in the treatment of experimental peritonitis. Surg Gynecol Obstet 139:211-214, 1974. 13. Abe H, Carballo J, Appert HE, et al: The release and fate of the intestinal lysosomal enzymes after acute ischemic injury of the intestine. Surg Gynecol Obstet 135:581-585, 1972.
Downloaded From: http://archsurg.jamanetwork.com/ by a University of Auckland User on 05/27/2015
