The effect of fibrin glue on intraperitoneal contamination in rats treated with systemic antibiotics

JOURNAL

OF SURGICAL

RESEARCH

52,123-126

(19%)

The Effect of Fibrin Glue on Intraperitoneal Contamination in Rats Treated with Systemic Antibiotics ROBERT J. SCHWARTZ, M.D., TERRY J. DUBROW, M.D., RICHARD A. RIVAL, M.D., SAMUEL E. WILSON, M.D., AND RUSSELL A. WILLIAMS, M.D. Department Presented

at the Annual

Symposium

of Surgery

Harbor-UCLA

of the Association

Medical

of Veterans

Intraperitoneal fibrin sealant lowers septic mortality in a rat model of peritoneal contamination (2 X 10’ organism inoculum) at the cost of increased late intraabdominal abscesses. This study utilized parenteral antimicrobials to determine if the protective effect of intraperitoneal fibrin could be achieved without increasing the late abscess formation rate. One hundred and fiftyfive rats were divided into four groups. Gelatin capsules containing various dilutions of feces (10” CFU/ ml) and barium sulfate were placed into the abdomen in all groups. Group I controls had no antibiotics or fibrin. In group II, the capsule was surrounded by a solution of cryoprecipitate, thrombin, and calcium (fibrin “glue”). Groups III (no fibrin, antibiotics) and IV (fibrin, antibiotics) received a broad-spectrum cephalosporin intramuscularly postoperatively and then daily. Surviving rats were sacrificed on the tenth postoperative day. At a moderate volume of fecal inoculum (0.3 ml), fibrin reduced mortality from 100% in the control group to 0% in treated animals (P < 0.001) that did not receive antibiotics. Abscesses formed in 10% of the surviving fibrin-treated rats which were implanted with 0.1 ml of inoculum. In the 0.2 and 0.3 ml inoculum groups substantially more abscesses occurred (75 and 70%, respectively). The protective effect of fibrin was not manifested in the antibiotic-treated rats since no deaths occurred in either group. At higher and lower inoculum doses, no significant differences between fibrin and control groups were observed in mortality or abscess formation, whether or not antibiotics were given. We conclude that intraabdominal abscess formation among survivors is not intrinsic to fibrin but rather, a consequence of survival in rats that otherwise would have succumbed to early sepsis. Fibrin glue can be used topically in the peritoneum without increased abscess formation, providing antibiotics are given. o iggz Academic Press,

Center, Torrance,

Administration

California

90509

Surgeons, Milwaukee,

Wisconsin,

May 9-11, 1991

This fibrin “glue” has found widespread use in surgery. In gastrointestinal resection, it can serve as an adjuvant to stapled anastomoses or as the primary sealant [l, 21. Fibrin glue-coated polytetrafluorethylene grafts are less likely to leak than the untreated vascular prostheses [3]. Applied to cut tissue surfaces, fibrin glue is an effective hemostatic agent and can both repair and coagulate hepatic and splenic lacerations [4-71. However, because fibrin may promote intraabdominal abscess formation, the safety of fibrin glue in the grossly contaminatedperitoneum is unclear [8]. Moreover, the patient sustaining abdominal trauma may be additionally predisposed to peritonitis because of fecal soiling, multiple transfusions, and multi-organ injury 19, lo]. In an established rat model of peritonitis, we found that fibrin glue lowered early mortality due to sepsis. Surviving animals that received intraperitoneal fibrin glue had a high rate of late, nonlethal abscess formation [ 111. To clarify the abscessogenic potential of fibrin glue, we have now eliminated early septic mortality in controls through the use of systemic, broad-spectrum antimicrobials, so that rates of late abscess formation can be directly compared between control and fibrin-treated groups. MATERIALS

AND

METHODS

Male Wistar rats (Charles River Laboratories, Wilmington, MA) weighing 275 to 300 g were the experimental animals. They were fed rat chow (Teklad Rodent Food, Madison, WI) and water ad libitum preoperatively and postoperatively. Inoculum was prepared from a fresh stool specimen, obtained from one of the experimenters. This was immediately diluted 1:l with brain-heart infusion broth and mixed thoroughly. The suspension was filtered through a single layer of gauze to remove vegetable fibers and large particles. The mixture was divided into 1 ml aliquots and quickly frozen at -70°C. One-milliliter aliquots of the feces-inoculum were diluted lo-fold in prereduced anaerobically sterilized me-

Inc.

An admixture of fibrinogen (cryoprecipitate), thrombin, and calcium polymerizes to form a gelatinous mass that both adheres to tissue and promotes hemostasis. 123

002%4804/92 $1.50 Copyright 0 1992 hy Academic Press, Inc. All rights of reproduction in any form reserved.

124

JOURNAL

OF SURGICAL

RESEARCH:

dia. Aliquots of 0.1 ml were then plated aerobically on MacConkey agar (Scott Laboratories), sheep blood agar (Scott Laboratories), and Columbia CNA plates (Scott Laboratories). These plates were incubated for 1 to 24 hr at 35°C with carbon dioxide present. Anaerobic cultures were performed on O.l-ml aliquots plated onto brainheart infusion agar with and without 20 mg/liter of gentamicin. All anaerobic plates were supplemented with a 3% mixture of hemoglobin digest and yeast extract (Filde’s enrichment, Difco, Detroit, MI) and vitamin K. Aerobe cultures were counted and identified using the Manual of Clinical Microbiology protocol [12]. Virginia Polytechnic Institute methods were used to count and identify anaerobes. Peritoneal fluid and abscess cultures were evaluated in an identical manner. To prepare the fibrin glue, pooled human cryoprecipitate, obtained from a blood bank, was gradually thawed just before surgery and drawn into l-ml syringes. Twenty-thousand units of bovine thrombin (Thrombostat, Parke-Davis, Morris Plains, NJ) was reconstituted with 20 ml of isotonic saline and 5 ml of 10% calcium chloride solution and drawn into l-ml syringes. The cryoprecipitate and thrombin-calcium syringes were connected to a duoflow device (Hemaedics, Pacific Palisades, CA) to ensure thorough mixing. Immediately before surgery, the inoculum was thawed and combined with barium sulfate (10% wt/vol). Using a tuberculin syringe, 0.1 to 0.5 ml of the mixture was placed in a No. 1 gelatin capsule. The capsule was then enclosed in a No. 0 gelatin capsule. Rats anesthetized with ketamine hydrochloride and xylazine hydrochloride instilled intraperitoneally had their abdomens shaved and prepared with povidone-iodine solution. The rats were draped sterilely, and a 1.5-cm lower-midline abdominal incision was made. The gelatin capsule was set in the right lower quadrant. In experimental rats, the capsule was surrounded by 2 ml of fibrin glue introduced through the duoflow device, The abdomen was closed in a single layer with continuous 3-O silk sutures. One-half of both the experimental and control groups received 100 mg of cefotetan (Cefotan, Stuart Pharmaceuticals, Wilmington, DE) intramuscularly at the time of surgery and then once daily until death or sacrifice. The animals were divided into four groups: Group I received neither fibrin glue nor antibiotics. Group II received fibrin glue only. Rats in group III were given antibiotics but no fibrin glue. Both fibrin glue and antibiotics were administered to group IV. Autopsies were performed on animals within 12 hr of death. Survivors were sacrificed on the tenth postoperative day. Gross pathologic appearance of the peritoneum was noted, as were the presence and size of abscesses. Aerobic and anaerobic cultures of peritoneal fluid and abscesses were obtained. Data were analyzed using the x2 test. A P value of less than 0.05 was considered statistically significant.

VOL.

52, NO. 2, FEBRUARY

1992

TABLE Qualitative

1

Bacterial Analysis of Fecal Inoculum (1 X 10” CFU/ml) Aerobes Staphylococcus hominis Streptococcus intermedium Staphylococcus epidermidis Enterococcus faecium Streptococcus mitis Streptococcus bouis Streptococcus salivarius Enterococcus durans Klebsiella pneumoniae Escherichia coli Enterobacter cloacae Viridans streptococcus Anaerobes Eubacterium lentum Eubacterium sabureum Eubacterium aeraofaciens Eubacterium limosum Clostridium ramosum Bacteroides vulgatus Bacteroides thetaiotaomicron Bacteroides loeseheii/denticola Bifidobacterium animalis

RESULTS

Bacteriologic

Analysis

Qualitative bacterial analysis of the fecal inoculum yielded gram-negative enterobacteriaceae, anaerobes, and gram-positive cocci (Table 1). The overall bacterial count for the encapsulated inoculum was lOlo colonyforming units (CFUs)/ml. Multiple organisms were found in the peritoneal fluid of animals who suffered early deaths from sepsis. Cultures from one of these rats grew mostly Clostridium subterminale. This animal had received neither fibrin glue nor antibiotics. In all other cases, Eschericia coli predominated in peritoneal fluid. Cultures of late-forming abscesses grew mainly Bacteroides fragilis. Mortality Antibiotic-treated. Virtually all animals treated with antimicrobials survived the experimental period whether or not they received fibrin glue. Ten percent of the rats that received the largest fecal inoculum (5 X 10’ CFU) died of peritoneal sepsis on the first post-operative day. This mortality rate was unchanged by fibrin glue (Table 2). Untreated. In the absence of antibiotics, the mortality of fibrin glue-treated rats did not differ significantly from controls at the highest (5 X 10’ CFU) or lowest (1 X 10’ CFU) inoculum doses. In the former group, the inoculum dose was uniformly lethal; in the latter no deaths occurred.

SCHWARTZ

ET AL.:

INTRAPERITONEAL

TABLE Mortality Inocuhnn

in Rats Receiving

size (ml) 0.1 0.2 0.3 0.5

o/10 6/10 lO/lO 515

Fibrin 0% 60% 100% 100%

125

GLUE

2

Different

Control

FIBRIN

Doses of Fecal glue 0% 0% 0% 100%

o/10 o/lot o/10* lO/lO

Inoculum Fibrin glue, antibiotic

Antibiotic o/10 o/10 o/10 l/10

0% 0% 0% 10%

o/10 o/10 o/10 l/10

0% 0% 0% 10%

t Effect of fibringlue, P < 0.05, significant compared with control. * Effect of fibrin glue, P < 0.001, significant compared with control.

At an intermediate inoculum (3 X 10’ CFU), however, all control rats died if not treated with antibiotics. Fibrin glue significantly reduced mortality in this group; all treated animals survived (P < 0.001). One fibrin-treated rat that received 2 X 10’ CFU inoculum died less than 2 hr postoperatively. Autopsy revealed intraperitoneal hemorrhage and this animal was excluded from the study due to technical error. No other fibrin glue-treated rats in this group died compared to 60% mortality in controls (Table 2). Abscess Formation Antibiotic-treated. The abscess formation rate was the same in the fibrin glue and control groups. The rate of abscess formation rose in proportion to the amount of fecal inoculum implanted (Table 3). Untreated. Few abscesses formed in rats receiving the smallest fecal inoculum (1 X 10’ CFU) and this process was not altered by fibrin glue. At 3 X 10’ CFU, 70% of the fibrin glue group-formed abscesses. More frequent abscess formation also ocsignificurred at 2 X 10’ CFU but was not statistically cant (Table 3). DISCUSSION Bacterial contaminants are eliminated from the peritoneal cavity by three established mechanisms [8, 13, 141. First, bacterial numbers are rapidly reduced by me-

chanical clearance through diaphragmatic lymphatics. Leukocytes then opsonize and phagocytize bacteria in situ. Initially, this task is performed by local peritoneal macrophages. Neutrophils are later recruited as part of the acute inflammatory response. Bacterial spread is further hindered through physical containment by omental, visceral, and parietal compartmentalization. The role of fibrin in the latter two mechanisms was demonstrated in a series of related experiments. Ahrenholz and Simmons and Dunn and Simmons used preformed bovine fibrin clots containing various bacterial strains to show that fibrin can nonspecifically trap bacteria preventing early death due to sepsis [8, 151. This trapping, however leads to late intraperitoneal abscess formation with an attendant mortality equal to the early mortality of peritonitis. Hau and Simmons showed that heparin significantly reduces mortality from peritoneal sepsis in a dog model by preventing further fibrin apposition [ 161. Moreover, Rotstein et al. proved that physiologic concentrations of fibrin inhibit phagocytosis of radio-labeled Escherichia coli and that this inhibition is dose-dependent [17]. They concluded that fibrin clot could thus serve as a nidus for abscess formation. Aerosolized fibrin solutions are increasingly used in intra-abdominal procedures, including trauma surgery. Their safety in the contaminated abdomen, though, has been unclear. In a previous study, using an established rat model of sepsis, we showed that aerosolized fibrin solutions decrease early mortality but resulted in late

TABLE Abscess Inoculum

size (ml) 0.1 0.2 0.3 0.5

Formation

Rate in Rats Receiving

Control z/10 114

Fibrin 20% 25%

t Effect of fibrin glue, P = NS compared with control. * Effect of fibrin glue, P = NS compared with antibiotic

l/lot

61% 7/10

alone.

3 Different glue 10% 75% 70%

Doses of Fecal Antibiotic l/10 5/10 7/10 m

10% 50% 70% 89%

Inoculum Fibrin glue, antibiotic o/10* 4/10* 3/10* 7/9*

0% 40% 30% 78%

126

JOURNAL

OF SURGICAL

RESEARCH:

abscess formation in surviving rats [ 111. At a critical concentration of bacteria, death due to peritonitis was significantly reduced after treatment with fibrin glue. We postulated that this effect was due to the bacterial trapping properties of fibrin allowing protection from intraperitoneal dissemination and hence lower early septic mortality. All survivors treated with fibrin glue formed abscesses which were small and well circumscribed. We felt that the “abscessogenic” effect of fibrin glue was due to bacterial sequestration and augmented growth in a fibrin-rich environment. This unanticipated and potentially therapeutic effect of fibrin glue must be tempered by the deviation of the model from the clinical setting in which intravenous antibiotics are used pre- and postoperatively in patients with intraperitoneal contamination. To simulate the clinical practice of adjunctive antimicrobial therapy, this study used a broad spectrum cephalosporin designed to reduce early septic mortality. Cefotetan was chosen because it is an effective single agent against both the aerobic and anaerobic human bowel flora of the inoculum [ 181. The first part of the study (without antibiotics) confirmed our previous assertion that fibrin glue alone reduces early mortality at the expense of late abscess formation. At 3 X 10’ CFU, fibrin glue reduced mortality from 100 to 0% with 70% of survivors forming well-circumscribed abscesses. In animals treated with antibiotics alone, early mortality was almost eliminated for all inoculum sizes. The rate of late abscess formation, however, was directly proportional to inoculum size. These findings suggest that fibrin glue and antibiotics used alone act similarly to reduce early mortality at the expense of late abscess formation. Used in combination, fibrin glue and antibiotics eliminate early mortality due to sepsis but, similarly to either used alone, result in abscess formation proportional to inoculum size. The use of fibrin glue in antibiotic treated rats, however, does not significantly alter the frequency of abscess formation at any given inoculum size. These results indicate that, in combination with antibiotics, fibrin glue does not promote abscess formation and should not be considered “abscessogenic”. We conclude that if the appropriate antibiotic regimens for intraabdominal sepsis are employed, fibrin glue has neither a therapeutic nor a deleterious effect even in major fecal contamination. Further studies with the use of fibrin solutions in patients with intraabdominal contamination are clearly needed, the animal data indicate

VOL.

52, NO. 2, FEBRUARY

1992

that fibrin glue does not act as a nidus for infection the presence of parenteral antimicrobials.

in

REFERENCES 1.

Rendaro, F., Coca, G., Quarto, G., Rinaldo, M., Castellitto, D., and Sarno, A. Valutazione sperimentale dell’effeto di rinforzo della colla di fibrina sulle anastomosi intestinali con un metodo originale in vivo. Boll. Sot. Ital. Biol. Sper. 61: 1336, 1985. 2. Harrison, R. C., and Oka, H. Rectal anastomosis: Sutures vs. staples and glue. Contemp. Surg. 21: 17, 1982.

3. Wilson, S. E., and Tsuchida, H. Unpublished data. 4. Blocker, S. H., and Ternberg, J. Traumatic liver laceration in the newborn: repair with fibrin glue. J. Pediutr. Surg. 21: 369,1986. I., 5. Jacob, H., Campbell, C. D., Stemberger, A., Wriedt-Ltibbe, Blumel, G., and Replogle, R. Combined application of heterologous collagen and fibrin sealant for liver injuries. J. Surg. Res.

36: 571,1984. 6. Kram, H. B., Hino, S. T., Harley,

7. 8.

9.

10.

D. P., Fleming, A. W., and Shoemaker, W. C. Use of concentrated fibrinogen in experimental splenic trauma. J. Biomed. Mater. Res. 20: 547, 1986. Scheele, J., Gentsch, H. H., and Matteson, E. Splenic repair by fibrin tissue adhesive and collagen fleece. Surgery 95: 6, 1984. Ahrenholz, D. H., and Simmons, R. L. Fibrin in peritonitis. I. beneficial and adverse effects of fibrin in experimental E. coli peritonitis. Surgery 88: 41, 1980. Thadepalli, H., Gorbach, S. L., Braido, P. W., Norsen, J., and Nyhus, L. Abdominal trauma, anaerobes and antibiotics. Surg. Gynecol. Obstet. 137: 270, 1973. Nichols, R. L., Smith, J. W., Klein, D. B., et al. Risk of infection after penetrating abdominal trauma. N. Engl. J. Med. 311: 1065,

1984. 11.

Dubrow, T., Schwartz, R. J., McKissock, J., and Wilson, S. E. Effect of aerosolized fibrin solution on intraperitoneal contamination. Arch. Surg. 126: 80, 1991. 12. Coyle, M. B., Morello, J. A., and Smith, P. B. Aerobic bacteria. In E. H. Lennette, A. Balows, W. J. Jr. Hausler, and H. J. Shadomy (Eds.), Manual of Clinical Microbiology, 4th ed. Washington, DC: Am. Sot. Microbial. 1985. Pp. 143-411. 13. Dunn, D. L., Barke, R. A., Ahrenholz, D. H., Humphrey, E. W., and Simmons, R. L. The adjuvant effect of peritoneal fluid in experimental peritonitis: mechanism and clinical implications. Ann. Surg. 199: 37, 1984. 14. Cohn, Z. A. Determinants of infection in the peritoneal cavity, I: response to and fate of Staphylococcus aureus and Staphylococcus albus in the mouse. Yale J. Biol. Med. 36: 12, 1962. III. the 15. Dunn, D. L., and Simmons, R. L. Fibrin in peritonitis. mechanism of bacterial trapping by polymerizing fibrin. Surgery 92: 513, 1982. 16. Hau, T., and Simmons, R. L. Heparin in the treatment of experimental peritonitis. Ann. Surg. 187: 294, 1978. 17. Rotstein, 0. D., Pruett, T. L., and Simmons, R. L. Fibrin inhibits phagocytic killing of Escherichia coli by human polymorphonuclear leukocytes. Ann. Surg. 203: 413, 1988. 18. Wilson, S. E., Boswick, J. A., Jr., Duma, R. J., et al. Cephalosporin therapy in intraabdominal infections. Am. J. Surg. 155: 61,

1988.