Effects of granulocyte colony-stimulating factor on ... - Springer Link

2 downloads 0 Views 431KB Size Report
Su EavTooAv. 9 Springer-Vcrlag 1997. Effects of Granulocyte Colony-Stimulating Factor on Bacterial. Translocation Due to Burn Wound Sepsis. ORHAN YAL~IN ...
Surg Today Jpn J Surg (1997) 27:154-158

Su EavTooAv 9 Springer-Vcrlag 1997

Effects of Granulocyte Colony-Stimulating Factor on Bacterial Translocation Due to Burn Wound Sepsis ORHAN YAL~IN, I GORSEL SOYBIR, I FERDA KOKSOY,I HAKKI KOSE, I RECEP (~ZTURK, 2 and BAKI ~OKNE~ELI 3 Departments of ~Surgery and ~Plastic and Reconstructive Surgery, Istanbul Taksim Hospital, Straseluilere cad. 80060 Beyo~ Istanbul Turkey. 2Department of Microbiology, Cerrahpasa Medical School, Istanbul University, Kocamustafa papa lstanbul, Turkey. Abstract: The presence of certain defects in both cellular and humoral immunity after thermal injury has been established. Likewise, the translocation of enteric bacteria to the mesenteric lymph nodes and to distant organs has also been observed following serious thermal injury. The effects of granulocyte colony-stimulating factor (G-CSF) on bacterial translocation, the small bowel mucosa, and cecal bacterial content were investigated in a rat model of burn wound sepsis in which albino Wistar rats were scalded over 30% of their bodies, after which the lesions were infected by 1 x 10s colony-forming units (cfu) Pseudomonas aeruginosa. The control group was treated with 5% dextrose solution subcutaneously starting 2 days preburn, while the treatment group received 100 ~tg/kg human G-CSF subcutaneously. On the 4th day post burn all animals were killed to examine the bowel and culture of the mesenteric lymph nodes (MLN), livers, and spleens. No significant differences were observed between the groups regarding the cecal bacterial content and small bowel; however, a difference was seen in the ratio of translocation in the MLN liver and spleen and quantitative MLN cultures. Based on these findings, G-CSF was thus found to be significantly effective in reducing bacterial translocation due to burn wound sepsis. Key Words: bacterial translocation, burn wound sepsis,

G-CSF

Introduction

In recent years mortality in burn injury patients has been greatly reduced thanks to advances in the treatm e n t provided by m o d e r n intensive care units. Postburn sepsis and related multiple organ failure, however, still

Reprint requests to: O. Yal~in, Feyzullah Efendi Sok. No:8/1 Daire:4 Fatih 34260, Istanbul, Turkey (Received for publication on Dec. 18, 1995; accepted on July 4, 1996)

remain as important causes of death? These factors are responsible for 7 5 % - 8 0 % of all cases of thermal injury mortality. Because these patients have a diminished cellular and humoral i m m u n e response, sepsis cannot be prevented despite extensive antibiotic treatment. 2 In particular, neutrophil function disturbances are often detected in burn patients? In healthy persons, bacteria do not grow in the mesenteric lymph nodes. In the 1980s the passage of viable bacteria from the intact gastrointestinal tract to the mesenteric lymph nodes and beyond was initially described by Berg et al. 4 and Deitch et al., s and termed "bacterial translocation." Bacterial translocation has been observed after various types of stress including thermal i n j u r y y ' significant t r a u m a ] and immunosuppression. 4 In experimental studies, a relationship between bacterial translocation and thermal injury mortality has been reported?' If this is true, then the prevention of bacterial translocation could thus potentially reduce the n u m b e r of deaths due to thermal injury. Colony-stimulating factors are a group of glycoproteins that stimulate various cell lines. Granulocyte colony-stimulating factor (G-CSF) stimulates the growth of granulocyte precursors, and activates mature neutrophils, s It is known that thermal injury can both cause immunosuppression and bacterial translocation. We therefore a t t e m p t e d to determine whether or not the immunostimulating factor G - C S F would be of value in decreasing experimental bacterial translocation due to thermal injury and burn wound sepsis.

Materials and M e t h o d s

This experimental study was carried out at the Taksim Hospital Experimental Research Laboratory. Male Wistar rats (250 _+ 20g) were divided into (a) control and (b) treatment groups. All rats were housed in individual cages, in a temperature-controlled r o o m with al-

O. Yalqin et al.: G-CSF on Bacterial Translocation Due to Burns in Rats ternating 12-h light-dark cycles, and the animals were acclimated for a minimum of 3 days before the study. All rats were provided with standard laboratory chow ad libitum. Experimental burn injury was induced according to the method of Walker and MasonY Briefly, the animals were anesthetized with ether, their hair was r e m o v e d from the dorsal surfaces, and the dorsal surfaces were then exposed to 95~ water for 10s through a template designed to produce a 30% total body surface area (TBSA) burn. After receiving 6ml lactated Ringer's solution intraperitoneally, all animals were infected using l ml of a suspension containing 1 • 10 ~ colonyforming units (cfu) per ml of Pseudomonas aeruginosa. The suspension was dripped onto the burn wound, and then rubbed with a fingertip covered with a sterile glove. R e c o m b i n a n t h u m a n granulocyte colony-stimulating factor (rhG-CSF) (Neupogen 30MU, Roche, Zurich, Switzerland) was diluted with 5% dextrose at a ratio of 1 to 4. The treatment group animals received r h G - C S F 100 ~tg/kg subcutaneously daily. T r e a t m e n t started 48 h before thermal injury and continued until the 4th day post burn. The control groups received equal volumes of 5% dextrose daily. All animals were killed by an overdose of ether anesthesia on the 4th day post burn. Full thickness burn wound biopsies (1 cm 2) were collected for culture and placed in a 5-ml broth. Each animal was placed on a sterile field, and the a b d o m e n was soaked with 70% alcohol. Under sterile conditions the skin and peritoneum were opened. The mesenteric lymph nodes were excised, while sections of the liver and spleen were also collected. All tissue specimens were weighed and homogenized with manual glass grinding rods in a 5-ml broth. Next, 0.5ml of the cecal contents were diluted in 4.5 ml broth, and serial dilutions were p e r f o r m e d in sterile saline. Then, 200-~tl aliquots from each broth were plated on McConkeys agar and sheep blood plates. All cultures were incubated and examined at 24 and 48h for presence of growth. In quantitative mesenteric lymph node (MLN) cultures, sheep blood plates were used to determine the values of translocations. McConkeys agar plates were used for the identification of any Gram-negative microorganisms. The identification of bacterial species was performed by standard microbiologic methods. Growth in the mesenteric lymph nodes was calculated as colony-forming units in I g tissue (cfu/g). Before they were killed, blood samples from all animals were obtained from their tail veins to determine the in vivo activity of G-CSF. The entire small bowel from the ligament of Treitz to the cecum was excised. The bowel contents were removed by saline irrigation. A sample from the distal edge of the intact bowel was taken. Samples were fixed with 10%

155

formalin and stained with hematoxylin and eosin for light microscopy. The total small bowel was weighed and the mucosal layer of the bowel was separated according to the method of Carter et alY ~Briefly, the bowel was opened and the mucosa was scraped with a glass slide across a porcelain plate, and the contents were collected and weighed. Burn wound biopsies were also taken and examined for any microscopic evidence of invasive burn wound infection. The results were analyzed by Fisher's exact test and the Mann-Whitney U test.

Results There were no deaths during the experiment. On physical examination, no difference between the two groups was observed, while in all animals the infection was limited to within the burn wound. In the control groups, all rats had bacterial translocation in each specimen. For both groups the results of the cultures are shown in Table 1. G - C S F treatment a u g m e n t e d the white blood cell (WBC) counts significantly (Table 2). The quantitative results of the mesenteric lymph node cultures were 986 • 349cfu/g (mean _+ SD) in the control group, and 371 • 151cfu/g (mean • SD) in the treatment group, the difference being significant (P < 0.05). In evaluating translocating organisms, Escherichia coil was the predominant organism. Proteus was also found in smaller quantities. No Pseudornonas were found in any viscera cultures. The intact small bowel mass, mucosal mass, and ileal mucosal thickness in groups are shown in Table 3. No significant difference between the values was seen (P > 0.05). Quantitative cultures of the burn wound biopsies and cecal contents are shown in Table 4. No significant differences in the values were observed between the

Table 1. Bacterial translocation of enteric organisms after thermal injury and infection Mesenteric lymph nodes Control Treatment

12/12 5/12"

Liver

Spleen

12/12 6/12"

12/12 6/12"

* Control group versus treatment group; P < 0.01 (Fisher's exact test)

Table 2. White blood cell counts White blood cells/ram 3 Before burn injury After burn injury

Control

Treatment

4560 _+ 972 5032 _+ 1 021

4492 +_ 829 9274 _+ 1 228*

*White blood cell counts of control group versus treatment group after burn injury~ P < 0.05 (Mann-Whitney U test). Values are expressed as mean + SD

156

O. Yalqin et al.: G-CSF on Bacterial Translocation Due to Burns in Rats

Table 3. Small bowel mass and ileal mucosal thickness

Control Treatment

Intact small bowel

Mucosa

Ileal mucosal thickness

242 + 0.16 2.35 _+ 0.14"

0.85 _+ 0.11 0.83 .+ 0.10"

522 .+ 23 507 + 19"

Data for bowel and mucosal mass = mean + SD g/100gbody weight; for mucosal thickness - mean -+ SD, thickness in [am. * P > 0.05 vs control (Mann-Whitney U test) Table 4. Quantitative cecal cultures and burn wound cultures

Control Treatment

Cecal cultures

Wound cultures

8400 _+ 7300 x 10~ 8200 _+ 7400 • l04.

4.9 _+ 0.8 X 10~) 4.7 + 0.9 • 10'~*

Data for cecal cultures mean -+ SD of colony-formingunit gramnegative facultative anaerobes per cc cecal contents; for wound cultures = mean _+SD of colony-formingunits/cm2tissue. *P > 0.05 vs control (Mann-Whitney U test) two groups. Histologic examinations of the burn wounds suggested wound infection. Sections of the distal ileum showed blunted villae in both groups; however, there was no difference in the appearances of the two groups.

Discussion

Infection plays a major role in thermal injury mortality. In these patients, overcoming infection is difficult because they are immunosuppressed. In severely burned patients, several neutrophil function deficiencies including chemotaxis, phagocytosis, intracellular killing, and oxidative burst have been found to occur. ~ ~4 In addition, a disordered regulation of granulopoiesis demonstrated after burn injury worsens due to coexisting bacterial infection. ~5.~' In the last decade, several studies have recognized bacterial translocation. Bacterial translocation may occur due to three pathophysiological conditions: (a) an alteration of the host defense, (b) a loss of the mucosal barrier, and (c) an overgrowth of the intestinal flora, jJ,J7 I'~ Sometimes more than one factor is responsible for translocation and every factor has some additional effect on translocation. 4 Bacterial translocation has been shown following extensive burn injury in experimental studies. The ratio of translocation is high in the early period; however, this ratio diminishes thereafter. However, the presence of infection results in continued translocation. Jones et al. 2" used the P s e u d o m o n a s burn wound sepsis model to evaluate the effects of ongoing infection after thermal injury on bacterial translocation. He found that the presence of a source of infection distant from the gastrointestinal tract after thermal injury was associated with a prolonged trans-

location to the mesenteric lymphatics. In uncomplicated burn injury, early bacterial translocation usually subsides within 4 days. For this reason we also used the burn wound sepsis model to evaluate late bacterial translocation. In our study we investigated the protective effect of G-CSF on bacterial translocation due to burn wound sepsis. A standard model of P s e u d o m o n a s burn wound sepsis in the rat was used to evaluate the effects of ongoing infection after burn injury on bacterial translocation, z~ With this treatment, the bacterial translocation rates and M L N quantitative culture values decreased significantly. We also investigated the effects of G-CSF on these three mechanisms of bacterial translocation. G-CSF treatment did not change the cecal bacterial content significantly, while, in addition, no significant change was seen in the mucosal barrier. Therefore, GCSF possibly shows its effect by reversing immunosuppression due to burn wound sepsis. The exact mechanism of protection of G-CSF from bacterial translocation has yet to be clarified. There are two components of the barrier that forms against bacterial translocation: one, a nonspecific mechanical and chemical barrier; the other, an immunological barrier. G-CSF has a glycoprotein structure and is found in the human body. This substancc augments granulopoiesis, affects the surviwd of neutrophils, and primes mature cells for enhanced functional activity in response to physiological s t i m u l i . 22,2:~ R h G - C S F produced with E. coil has in vitro and in vivo stimulating activities similar to those of native human G-CSF and is active in mice, rats, hamsters, nonhuman primates, and humans. 24 G-CSF prolongs the lifespan of granulocytes by suppressing apoptosis. -'5 Therel'ore, this factor could directly improve resistance to infections in critically ill patients either by reversing defects in the neutrophil ['unction or improving their production. The therapeutic effects of G-CSF on the bacterial resistance of bacterial translocation could thus be mainly caused by the induction of bactericidal or bacteriostatic mechanisms, by the stimulation of an enhanced rate of phagocytosis, or by the stimulation of an increased production of neutrophils by bone marrow. It has been revealed that this factor can increase the capacity of granulocytes to phagocyte and thus kill bacteria, while also participating in antibody-dependent, cell-mediated cytotoxicityY' Besides the immunological barrier, G-CSF may also give some help to the nonspecific barrier. A close correlation has been shown between the production of colony-stimulating factors and prostaglandin E (PGE). z7 Colony-stimulating factors may promote new P G E production. P G E is also known to decrease bacterial translocation after thermal injury) This protective effect has been attributed to various effects

O. Yalqin et al.: G-CSF on Bacterial Translocation Due to Burns in Rats of PGE on the gastrointestinal mucosa, for example an enhancement of mucus secretion, a gastrointestinal blood flow increase, the promotion of bicarbonate secretion, and cyclic adenosine m o n o p h o s p h a t e ( A M P ) production. 2~ In our study, G-CSF increased the W B C counts nearly 1.5-fold. However, the W B C increase is not the only cause of a decrease in bacterial translocation. There have been several reports showing GCSF efficacy under immunosuppressive conditions. Steinshamn 2') showed that G-CSF in 100-~tg/kg doses is more effective than 20-btg/kg doses in the treatment of mice with a Candida albicans infection. It has been found that G-CSF is also effective in nonneutropenic hosts. In experimental pneumococcal pneumonia, 3~ this efficacy has also been proven. O'Reilly et al. 31 have examined the G-CSF effect on an intraabdominal sepsis model. In a model of cecal ligation and puncture in mice, G-CSF was also highly efficacious in increasing the survival rate. In the m o u s e model, G-CSF corrected decreased neutrophil function disorders due to burn injury within 72 h of burn injury. 32 In a study conducted by M o o n e y et al., 33 G-CSF treatment improved survival in 15% burn and Pseudomonas infection of mice. However, this model was only a survival model and translocation was not investigated. In conclusion, in a rat thermal injury and burn wound sepsis model, G-CSF was found to reduce bacterial translocation signilicantly; thus, G-CSF may possibly reduce the Gram-negative bacteriemia caused by enteric microorganisms in thermal injury.

References 1. Sinig K, l)eitch AE (1988) F~ffcctof bactcrcn]ia on mortality after thernlal injury. Arch Surg 123:1367 1370 2. Strana RJ, Warden (}11, Ninncmann ,I1,, Sat'lle JR (1986) hnmunologic paramclcrs in burned palicnts. Effect of thcrapculic interventions. J Trauma 21:7-17 3. Alcxandcr JW, Mcakins JL (1972) A physiologic basis for lhc development of opportunistic mfcclions in man. Surgery 176:273 287 4. Berg RD, Wommack E, l)citch EA (1988) hmnunosupprcssion and intestinal bacterial overgrowth synergistically promote bacterial translocation. Arch Surg 123:1359-1364 5. l)citch EA, Wintertorn ,I, Berg RD (1984) Pronlotion by burn stress of tile translocation of bacteria from the gastrointestinal tracts of mice. Arch Surg 119:166-172 6. Fukushima R, Giannotti L, Alexander JW, Pyles T (1992)The degree of bacterial translocation is a determinant factor fc.r n]ortality al:ter burn injury and is improved by prostaghmdin analogs Ann Surg 216:438-445 7. Sahnan FT, Buyruk MN, Gfirler N, (.'elik A (1992) The effect of surgical trauma on the bacterial translocation from the gut. J Pediatr Surg 27(7):802-8(t4 8. Sieff CA (1987) Hematopoietic growth factors. J Clin Invest 79:1549-1557 9. Walker HL, Mason AD (1968) A standard animal burn. J Trauma 8:1049-1054

157

111. Carter EA, Jung W, Ehrlich P, Ovellerre A (1988) Thermal trauma and gastrointestinal function: Ill. Effect ol' hot and cold burn tramna on small intestinal weight and mucosal mass of mice. J Burn Care Rehab 9:351-353 11. Deitch EA, Gclder F, Mcdonald JC (1984) Sequential prospective analysis of the non specific host defense system after thermal injury. Arch Surg 119:83-89 12. Nelson R, Hasslen SR, Ahrenholz DH, Solem LD (1987) Mechanism of loss of human neutrophil chemotaxis following thermal injury. J Burn Care Rehab 8:496-502 13. Dobke MK, Deitch EA, Harner TJ, Baxter CR (1989) Oxidative activity of polymorphonuclear lcukocytes after thermal injury. Arch Surg 124:856-859 14. Peterson VM, Robinson WA, Wallner SF, Rundus C, Hansbrough JF (1985) Granulocyte stem cells are decreased in humans with fatal burns. J Trauma 25:413-418 I5. McEven DD, Ogavar M, Evrenius K (1977) Myelopoesis in the infected burn J Lab Clin Med 89:540-543 16. Newsome JW, Evrenius K (1973) Suppression of granulocyte and platelet production by Pseudomonas burn wound infection. Surg Gynecol Obstet 136:375-379 17. Deitch EA, Berg RD (1987) Endotoxin but not malnutrition promotes bacterial translocation of the gut florae in burned mice. J Trauma 27:161-169 18. Deitch EA, Berg RD (1987) Bacterial translocation from the gastrointestinal tract, a mechanism of infection. J Burn Care Rchab 8:475 482 19. Deitch EA, Berg RD, Specian R (1987) Endotoxin promotes tile translocation from tile gut. Arch Sure 122:185-192 21). Jones WG, Minei JP, Barber AE, Rayburn JL, Fahey TJ, Shires Ill GT, Shires GT (199(I) Bacterial translocation and intestinal atrophy al'ter thermal injury and burn wound sepsis. Ann Surg 211:399-4(/5 21. Marano MA, Moldawcr LL, Fong Y, Wci H, Minci J, Yurt R, Cerami A, Lowry SF" (1988) Cacheclin/TNF production in experimenial burns and Psetulomonas infection. Arch Surg 123:13831388

22. Nelson S (1994) Role of granulocytc cohmy-stimulating factor in the immune i-cspc.nse 10 ~.ICUICbacterial infection in tile nonnet, tropenic host: An overview. ('lit] Infect Dis IglSuppl]:197-204 23. Lang Ctt, 13agby (;,1. l)obrcscu (', Nelson S, Spitzcr JJ (19921 Effects of granulocytc cohmy-stinmlating factor on sepsis-induced changes in neutrophil accumulalJon and organ glucose uptake. J Infect l)is 166:336 343 24. Tanaka H, Kancko ( 1991 ) Pharnu~kinctics of recombinanl human granulocyte colony-stimulating factor in file rat. l)rug Mctab l)ispos 19:200-204 25. Williams GT, Smith (!A, Spooncer E, Dexter TM, Taylor DR (1990) l tcmopoiclic colony stimulating factors promote tell survival by suppressing apoptosis. Naturc 343:76-79 26. lx~pez AF, Nicola NA, 13m-gesAW, Metcall' 11, 13attye FL, Sewcl WA, Vadas M (1983) Activation of granulocyte cytotoxic function by purilied mouse colony stimulating factors. J hmnunol 131:2982-2988

27 Kurland,ll, Pelus I~M, Ralph P, Bockman RS, Moore M A. (1979) Induction ot' prostaglandin E synthesis in normal and neoplastic macrophagcs. Role for cohmy stimulating factors distinct from effects on myeloid progenitor cell prolifcration. Proc Nall Acad Sci USA 76:2326-2331) 28. Miller TA (1983) Protective effect of prostaglandins against gastric mucosal damage. Current knowledge and proposed mechanisms. Am J Physiol 245:601-602 29. Steinshamn S, Bergh K, Waage KA (1993) Effects of stem cell factor and granulocyte cohmy stimulating factor granulocyte recovery and Candida albicans infection in granulocytopenic mice. J Infect Dis 168:1444-1448 31). Lister PD, Gentry M,I, Preheim LC (1993) Granulocyte colony stimulating factor protects control rats but nol ethanol fed rats from pneumococcal pneumonia. J Infect Dis 168:922-926

158

O. Yalqin et al.: G - C S F on Bacterial T r a n s l o c a t i o n D u e to B u r n s in Rats

31. O'Reilly M, Silver GM, Greenhalgh DG, Gamelli RL, Davis JH, Hebert JC (1992) Treatment of intraabdominal infection with granulocyte colony-stimulating factor. J Trauma 33:679-682 32. Sartorelli KH, Silver GM, Gamelli RL (1991) The effect of granulocyte colony-stimulating factor (G-CSF) upon burn-

induced defective neutrophil chemotaxis. J Trauma 31:523530 33. Mooney DP, Gamelli RL, O'Reilly M, Hebert JC (1988) Recombinant human granulocyte colony-stimulating factor and pseudomonas burn wound sepsis. Arch Surg 123:1353-1357