- Email: [email protected]
The Effect of Sepsis on Wound Healing
Journal of Surgical Research 102, 193–197 (2002) doi:10.1006/jsre.2001.6316, available online at http://www.idealibrary.com on
The Effect of Sepsis on Wound Healing 1 Rebeca M. Rico, M.D., Renato Ripamonti, M.D., Aime L. Burns, M.S., Richard L. Gamelli, M.D., FACS, and Luisa A. DiPietro, D.D.S., Ph.D. 2 Burn and Shock Trauma Institute, Department of Surgery, Loyola University Medical Center, Maywood, Illinois 60153 Submitted for publication October 23, 2000; published online December 6, 2001
Background. Normal wound healing is a regulated sequence of events that successfully restore tissue integrity. Previous studies have suggested that wound healing is impaired in a septic host. The current study examines the effect of sepsis on the inflammatory and proliferative phases of wound healing at a remote site of secondary injury. Methods. Polyvinyl alcohol sponges, either inoculated with a standard dose of Pseudomonas aeruginosa (experimental) or soaked in normal saline (control), were placed subcutaneously in the anterior abdominal region of male B6D2F1 mice. Immediately following sponge placement, full thickness excisional dermal wounds were created on the dorsum. Wound healing was examined at days 3, 5, and 7 postinjury. The infiltration of neutrophils and macrophages into wounds was quantified, and the reepithelialization rate and collagen content were measured. Results. Peripheral neutrophil counts were significantly elevated in infected mice, yet neutrophil content of the remote wound of infected animals was significantly reduced (5% of control, P < 0.05). Wounds of infected mice also showed a 30% reduction in the macrophage content. Wounds of infected animals exhibited delayed reepithelialization (76 ⴞ 3 vs 97 ⴞ 3% at day 5, P < 0.05) and collagen synthesis (55.3 ⴞ 9.5 vs 105 ⴞ 13.0 g/wound, P < 0.05). Conclusion. Systemic infection alters both the inflammatory and the proliferative processes at remote sites of injury. Multiple factors seem likely to contribute to the increased incidence of wound complications in septic patients. © 2001 Elsevier Science 1
This work was supported by the Dr. Ralph and Marion C. Falk Medical Research Trust (L.A.D., R.L.G.) and National Institutes of Health Grants GM55238 (L.A.D.) and GM42577 (R.L.G.). 2 To whom correspondence and reprint requests should be addressed at Loyola University Medical Center, Burn and Shock Trauma Institute, 2160 S. First Avenue, Maywood, IL 60153. Fax: (708) 327-2813. E-mail: [email protected].
Key Words: wound healing; sepsis; neutrophils; macrophages; epithelium; collagen. INTRODUCTION
Sepsis is the most common cause of deaths in the intensive care unit and is the 13th leading cause of death in the United States for persons older that 1 year of age. The care of septic patients accounts for an estimated 5 to 10 billion dollars in medical costs annually [1, 2]. Septic patients are at an increased risk for problems related to poor wound healing, such as anastomotic leaks, fascial dehiscence, and nonhealing wounds. In addition, these patients are more likely to develop secondary infections such as pneumonia and urinary tract infections. These complications result in increased morbidity and mortality in this patient population [2, 3]. The mechanism by which systemic infection alters the inflammatory and reparative response at remote sites of injury is not fully understood [4]. Previous studies have shown that wound bursting strength and collagen content are decreased in animals suffering from systemic infection [3, 5]. However, little is known about the mechanism behind this abnormal wound healing. The current study examines the wound repair process during sepsis and provides unified information about both the inflammatory and the proliferative phases of wound healing in a defined model system. METHODS Animals. All animal protocols used in this study were reviewed and approved by the Loyola University Institutional Animal Care and Use Committee. Male B6D2F1/J mice (The Jackson Laboratory, Bar Harbor, ME) 6 to 8 weeks of age and weighing 22 to 27 g were acclimatized for 1 week before use in an approved animal care facility with a 12-h light/dark cycle. Water was supplied ad libitum. Because previous studies suggest that a nutritional deficit may contribute to the poor healing that is seen with sepsis [5], control and
193
0022-4804/01 $35.00 © 2001 Elsevier Science All rights reserved.
194
JOURNAL OF SURGICAL RESEARCH: VOL. 102, NO. 2, FEBRUARY 2002
experimental animals were pair fed. Standard rodent laboratory chow (5001; Ralston Purina Co., St. Louis, MO) was weighed daily for food consumption measurements. Every morning, control animals were pair fed with an amount of food equal to that consumed by the Pseudomonas- infected animals. Wounding and infectious challenge. Mice were anesthetized by inhalation of methoxyflurane (Metofane; Schering–Plough Animal Health Corp., Union, NJ). A 0.5- to 1.0-cm skin incision was made on the right lower quadrant of the anterior aspect of the abdomen of each mouse. A polyvinyl alcohol sponge disc (8 ⫻ 2 mm) was placed subcutaneously. For all experiments except the assessment of peripheral neutrophil count, the sponges for the infected group (n ⫽ 6) were inoculated with 250,000 –300,000 CFU Pseudomonas aeruginosa, ATCC 19960 (Rockville, MD). This strain of Pseudomonas has been shown to induce septicemia in mice when inoculated at burn wound sites at low doses [6]. Sponges of the control group (n ⫽ 6) contained normal saline. Tissues from the same mice were used for analysis of wound reepithelialization, collagen content, and inflammatory cell content. For the examination of peripheral neutrophil levels, sponges were inoculated with no (n ⫽ 4), 135,000 CFU (n ⫽ 8), or 270,000 CFU (n ⫽ 6) of P. aeruginosa. Following sponge placement, the dorsum of each animal was shaved and briefly washed with 70% isopropanol. A standard dermal biopsy punch (Acuderm, Ft. Lauderdale, FL) was used to place six full-thickness excisional wounds (3 mm in diameter) on the dorsal skin of each mouse. Wounds were covered with Tegaderm (3M Health Care, St. Paul, MN). At 3, 5, or 7 days after injury, the mice were euthanized by halothane inhalation (Halocarbon Laboratories, Riveredge, NJ). For histologic and inflammatory cell analysis, wounds and approximately 1 mm of surrounding tissue were excised with a 5-mmdiameter standard dermal biopsy punch. For biochemical analysis of collagen content, only the central portion of the wound was collected with a 3-mm-diameter standard biopsy punch. Wounds were either flash-frozen in liquid nitrogen or embedded in TBS Tissue Freezing Medium (Triangle Biomedical Sciences, Durham, NC). All samples were stored at ⫺80°C until the time of analysis. Bacterial cultures. At postinjury days 1 and 3, the liver and spleen of each animal were aseptically dissected. The tissue was then cultured in tryptic soy broth for 24 h at 37°C. An aliquot of the culture was then plated onto tryptic soy agar and incubated for a further 24 h. The colonies were then counted and Gram stained with Brown Hopps Gram stain (Newcomer, Middleton, WI) for microbiological analysis. Peripheral neutrophil count. Blood was collected from anesthetized mice in heparin (Elkins–Sinn, Inc., Cherry Hill, NJ)-coated syringes via cardiac puncture. To lyse red cells, an aliquot was mixed with a solution of 3% acetic acid, 0.16% saponin (1:20). Total white cell counts per milliliter were determined using a standard Neubauer hemacytometer. The neutrophil percentage was determined by analysis of a blood smear; a total of 200 white cells per smear were counted and differentiated. The total neutrophil count was determined by multiplying the percentage of neutrophils by the total white cell count. Analysis of wound myeloperoxidase. The measurement of myeloperoxidase (MPO) levels in tissue has been shown to reflect neutrophil content. MPO was measured using previously described methods [7, 8]. Briefly, wounds were homogenized in 2.0 ml of 20 mM phosphate buffer, pH 7.4. Homogenates were centrifuged at 12,000g for 45 min, and the supernatant was decanted. The pellets were resuspended in 1.0 ml of 50 mM phosphate buffer containing 10 mM EDTA and 0.5% hexadecyltrimethylammonium bromide (HTAB). After a freeze–thaw cycle, the samples were sonicated briefly and incubated at 60°C for 2 h. The samples were centrifuged at 500g for 10 min and the supernatant was transferred to 1.5-ml tubes for storage at ⫺20°C. For analysis, samples were thawed, and an MPO standard curve ranging from 0 to 3.0 units/ml was generated. Fiftymicroliter aliquots of samples or standards were placed in 12 ⫻ 75-mm glass tubes with 500 l of assay buffer (0.1 M phosphate
buffer, pH 5.4, 1% HTAB, 0.43 mg/ml 3,3⬘,5,5⬘-tetramethylbenzidine). The reactions were started by addition of 50 l of 15 mM H 2O 2, incubated at 37°C for 15 min, and stopped with 1.0 ml of cold 0.2 M sodium acetate, pH 3.0. The absorbance of each sample was read at 655 nm within 10 min of completion. All samples and standards were tested in duplicate. Quantitation of macrophage infiltration. Sections were fixed in acetone at room temperature for 15 min, pretreated with 3% H 2O 2 in methanol to block endogenous peroxidase, and then blocked for nonspecific binding with normal mouse serum (1:100). Primary antibody, MOMA-2, which recognizes an intracellular antigen on murine macrophages and monocytes (Serotec Laboratories, Oxford, UK; 1:500), was utilized to detect these cells in the wound [9]. Primary and secondary antibody (biotinylated mouse anti-rat IgG; Jackson Laboratories, West Grove, PA; 1:100) incubations were performed for 30 min each, followed by a 30-min incubation with avidin– biotin– horseradish peroxidase complexes (ABC; Vector Laboratories, Burlingame, CA). Color development was performed with 3,3⬘diaminobenzidine and slides were counterstained with Harris hematoxylin. For each section, the number of macrophages within the wound bed was counted in 6 to 10 random high-power fields with the aid of an optical grid. All analyses were performed blinded. Analysis of wound reepithelialization. Reepithelialization of excisional wounds was measured by histomorphometric analysis of tissue sections from the central portion of the wound. Using a standard ocular grid, the distance between the wound edges and the distance that the epithelium had traveled across the wound were measured. The percentage reepithelialization ((distance covered by epithelium/distance between wound edges) ⫻ 100) was calculated for each section. Analysis of wound collagen content. Collagen content was assessed by determining the hydroxyproline content [10]. All reagents for this assay were purchased from Sigma Chemical Company (St. Louis, MO). Frozen wound tissue was hydrolyzed in 2.0 ml of 6 N HCl for 3 h at 130°C or overnight at 110°C. The reaction was neutralized with 2.5 N NaOH and diluted 40-fold with deionized H 2O. One milliliter of a 0.05 M chloramine T solution was added to 2 ml of the neutralized/diluted solution and incubated for 20 min at room temperature. One milliliter of 3.15 M perchloric acid was added and the solution was incubated for 5 more minutes at room temperature. One milliliter of 20% p-dimethylaminobenzaldehyde was then added and the resulting mixture was incubated for 20 min at 60°C. The samples were cooled with cold tap water. Absorbance at 557 nm was measured spectrophotometrically, and the amount of hydroxyproline was determined by comparison to a standard curve. Statistical analysis. Data were analyzed using GraphPad Prism, version 2.01 (GraphPad Software, Inc., San Diego, CA). The means and SEM were calculated for each data set. To compare neutrophil, macrophage, and collagen contents of wounds, a paired t test analysis was used. To compare the degree of reepithelialization over a time course between the two groups, a two-way ANOVA was employed. P values less than 0.05 were considered significant. To compare peripheral neutrophil counts, data were subjected to a one-way ANOVA followed by a Newman–Keuls multiple comparisons test.
RESULTS
Effects of sepsis on mortality and weight loss. The induction of sepsis resulted in a 50% mortality in the infected group, with most deaths occurring at 48 –72 h postinfection. There was one unexplained death in the control group at day 5 (Fig. 1). Bacterial cultures from liver and spleen of both the infected and the control group were obtained at day 1 and day 3 post-sepsis induction. All bacterial cultures from the infected group were positive for growth of gram-negative rods
RICO ET AL.: SEPSIS AND WOUND HEALING
195
FIG. 1. Mortality of control and infected mice. The induction of sepsis by implantation of 250,000 –300,000 CFU of P. aeruginosa resulted in a 50% mortality in the infected group, with most deaths occurring at 48 –72 h.
while those of the control group were negative (n ⫽ 6) (data not shown). When wound sections from day 3 postinjury were Gram stained, no histologic evidence of Pseudomonas was seen. Therefore, although the experimental group was systemically infected, no evidence of infection at the wound site was observed. The effect of systemic infection on weight loss was also examined. Both groups of animals were weighed at day 0 and at days 3, 5, and 7 postinjury. No significant difference in weight loss between the two groups was observed at any time point (Table 1). Although the control mice appeared to have lost more weight than the infected group, this difference was not statistically significant. Peripheral white blood cell counts. Peripheral total white blood cell counts and neutrophil counts were determined at day 3 postwounding in mice infected with either 135,000 (n ⫽ 8) or 270,000 CFU (n ⫽ 6) and were compared to those of control uninfected mice (n ⫽ 4). Infected mice from both groups exhibited a significant elevation in both peripheral total WBC number and peripheral neutrophil levels (Fig. 2). Effects of sepsis on wound inflammation. To evaluate the effects of sepsis on the inflammatory phase of wound healing, neutrophil and macrophage infiltration into the wound was examined at day 3 postinjury. The relative levels of neutrophils between the wounds of the infected and the control mice were determined by measuring myeloperoxidase levels. Surprisingly, the neutrophil content in the wounds of the infected animals was only 5% of the control group value (P ⬍ 0.05, Fig. 3). In addition, a 30% reduction in the number of wound mac-
FIG. 2. Peripheral white blood cell counts in control mice (n ⫽ 4) and those receiving either 135,000 (n ⫽ 8) or 270,000 CFU (n ⫽ 6) of P. aeruginosa. Peripheral white counts were taken at day 3 after infection. (A) Total peripheral white blood cell counts. (B) Total neutrophil counts. Bars, means; lines, SEM. Data were compared by one-way ANOVA and Newman–Keuls post hoc multiple comparisons test. *P ⬍ 0.01, **P ⬍ 0.05, compared to control.
rophages in the infected animals versus the control group was observed (16 ⫾ 2 vs 24 ⫾ 3 cells/HPF at day 3, Fig. 4). Thus, infected mice exhibited significant changes in the inflammatory response at the wound site. Neutrophilic infiltrate was also examined histologically at days 3 and 5 following injury. Little to modest neutrophil infiltration
TABLE 1 Average Weight Loss of Control versus Infected Mice Day postinjury
Control ( ⫾ SEM)
Infected ( ⫾ SEM)
3 5 7
2.30 ⫾ 0.46 2.37 ⫾ 0.45 2.06 ⫾ 0.18
1.68 ⫾ 1.42 1.20 ⫾ 0.73 1.46 ⫾ 1.30
FIG. 3. The effect of sepsis on wound neutrophil content. Neutrophil content of dermal wounds was assessed by measurement of myeloperoxidase, a neutrophil enzyme marker. Myeloperoxidase concentration per wound (⫾SE) was determined at day 3 after injury for both control and infected mice receiving 250,000 –300,000 CFU of P. aeruginosa, n ⫽ 6. Bars, means; lines, SEM. *P ⬍ 0.05 compared to control.
196
JOURNAL OF SURGICAL RESEARCH: VOL. 102, NO. 2, FEBRUARY 2002
FIG. 4. Effects of sepsis on wound macrophage content. The number of macrophages per high-power field (⫾SE) was determined by immunohistochemical analysis for wounds at day 3 after injury. Infected mice received 250,000 –300,000 CFU of P. aeruginosa, n ⫽ 6 for each group. Bars, means; lines, SE.
was observed in wounds of both experimental and control mice at day 5, and no difference was observed between the groups (data not shown). This observation is supported by our previous studies in this wound model. Quantification of neutrophils in this model has shown that the day 5 time point is well past peak neutrophil infiltration into the wound [11]. Effects of sepsis on the proliferative response in the healing wound. To evaluate the proliferative phase of wound healing, the amount of reepithelialization and the collagen content of each wound were measured. Compared to the control group, reepithelialization in the infected animals was significantly inhibited, and infected mice showed reduced reepithelialization at each of three time points that were examined (Fig. 5). Whereas wounds of the control group exhibited nearly complete closure at day 5 postinjury, wounds of infected mice were only 76% reepithelialized (P ⬍ 0.05, Fig. 5). Similarly, the collagen content of the wounds was also reduced in the infected animals versus the control group (Fig. 6). At day 7, hydroxyproline levels in the wounds of infected mice were only about 50% those of control mice (55.3 ⫾ 9.5 vs 105 ⫾ 13.0 g/ wound, P ⬍ 0.05). Thus, the proliferative aspects of excisional wound repair were significantly delayed by systemic infection.
FIG. 5. Time course of reepithelialization in control versus infected mice. Percentage reepithelialization (⫾SEM) is shown, n ⫽ 6 (days 3, 5) and n ⫽ 4 (day 7). Infected mice received 250,000 – 300,000 CFU of P. aeruginosa. *P ⬍ 0.05 between groups by twoway ANOVA.
FIG. 6. Effects of sepsis on wound collagen content. Hydroxyproline content per wound (⫾SEM) was determined at day 7 after injury. Infected mice received 250,000 –300,000 CFU of P. aeruginosa, n ⫽ 4 for each group. Bars, means; lines, SEM. *P ⬍ 0.05 compared to control.
DISCUSSION
Studies dating back to those performed by Dr. Alexis Carrel in 1924 have demonstrated that the rate of wound contraction is inhibited by infection and that a distant abscess impairs healing [12, 13]. More recent studies have documented that both wound strength and collagen content are diminished in a systemically infected animal [5]. Other studies have shown that both a systemic transient bacteremia and a distant inflammation from intradermal croton oil result in impaired wound healing [3, 14]. These studies, as well as the present one, support the concept that inflammatory responses at sites distant from the wound bed can profoundly influence the pace of local tissue inflammation and repair. Aberrant healing occurs in numerous circumstances, including diabetes, peripheral vascular disease, and cancer. Not surprisingly, one of the common denominators of poor wound healing in these conditions is the alteration of the inflammatory response [15, 16]. Previous studies have demonstrated the importance of leukocytes in orchestrating the proliferative response in wound healing [17–19]. As the first leukocyte to enter the wound, neutrophils provide the initial line of defense against infection. In the face of systemic infection, the ability of these leukocytes to perform critical functions such as chemotaxis and cytokine production may be impaired [20]. Likewise, the depletion of monocytes and macrophages results in impaired healing, including decreased fibroblast proliferation, and immature fibrosis [17]. Macrophages are known to produce factors that stimulate both angiogenesis and collagen synthesis [17, 18]. In the case of systemic infection, the recruitment of leukocytes is notably affected, and several previous studies have documented a decrease in peripheral leukocyte infiltration in the septic condition. In a murine cecal ligation and puncture-induced peritonitis model, neutrophilic recruitment to a remote site of secondary injury (skin) was shown to be significantly reduced in septic mice [21]. In addition, a 72% reduction in neutrophil influx into skin wound blisters was seen in septic patients compared to healthy subjects [22]. In contrast, alterations in macro-
RICO ET AL.: SEPSIS AND WOUND HEALING
phage content of wounds of septic animals are not yet as well documented. The significance of a decrease in wound macrophages to the delayed wound proliferative response seen in septic animals will require further investigation. The mechanism for the decrease in neutrophils and macrophages in the wounds of the infected animal is not yet completely understood. One postulate is that a finite number of neutrophils are available for delivery to one or more sites of injury and that neutrophils are concentrated in the site of the predominant injury [21]. However, the dramatic decrease in neutrophils at the wound site occurs in the face of an increase in the number of circulating neutrophils in the periphery, suggesting that other mechanisms come into play. In burn injury, and burn sepsis, the chemotactic response of neutrophils has been shown to be impaired [20]. The mechanism of decreased chemotaxis is probably multifactorial and may involve either changes in surface receptor expression and ligand engagement or alterations in intracellular signaling [22]. A further consideration is that alterations in the responsiveness of peripheral leukocytes may derive from an altered bone marrow environment in the septic host. Significant perturbations in both myelopoiesis and monocytopoiesis have been documented to arise during sepsis [23]. These central changes might influence responsiveness in the periphery and might provide some explanation for the observed effect of sepsis on inflammation at a site of secondary injury. Finally, elevated systemic levels of proinflammatory cytokines are well documented in sepsis. High levels of proinflammatory cytokines, particularly TNF␣, have been shown to inhibit peripheral leukocyte function in wounds [24]. The current study has interesting implications that extend beyond the healing wound. Clinically, septic patients are predisposed not only to wound complications but also to secondary infections, such pneumonia. The described deviations in the local inflammatory response may play a significant role in the increased incidence of secondary infections in this patient population. The effect of systemic infection on the function and recruitment of leukocytes requires additional rigorous examination. Ultimately, such mechanistic information may contribute to the development of therapeutic treatments to counteract this disordered inflammatory response. REFERENCES Darville, T., Giroir, B., and Jacobs, R. The systemic inflammatory response syndrome (SIRS): Immunology and potential immunotherapy. Infection 21: 279, 1993. 2. Barriere, S. L., and Lowry, S. F. An overview of mortality risk prediction in sepsis. Crit. Care Med. 23: 376, 1995. 3. De Haan, B. B., Ellis, H., and Wilks, M. The role of infection on wound healing. Surg. Gynecol. Obstet. 138: 693, 1974.
4. 5.
6.
7.
8.
9.
10.
11.
12. 13. 14.
15.
16. 17.
18.
19.
20.
21.
22.
1.
23.
24.
197
Thornton, F. J., Schaffer, M. R., and Barbul, A. Wound healing in sepsis and trauma. Shock 8: 391, 1997. Greenhalgh, D. G., and Gamelli, R. L. Is impaired wound healing caused by infection or nutritional depletion? Surgery 102: 306, 1987. Rosenthal, S. M. A procedure for measurement of wound healing, with special reference to burns. Proc. Soc. Exp. Biol. Med. 123: 347, 1966. Suzuki, K., Ota, H., Sasagawa, S., Sakatani, T., and Fujikura, T. Assay method for myeloperoxidase in human polymorphonuclear leukocytes. Anal. Biochem. 132: 345, 1983. Schierwagen, C., Bylund-Fellenius, A. C., and Lundberg, C. Improved method for quantification of tissue PMN accumulation measured by myeloperoxidase activity. J. Pharmacol. Methods 23: 179, 1990. Kraal, G., Rep, M., and Janse, M. Macrophages in T and B cell compartments and other tissue macrophages recognized by monoclonal antibody MOMA-2. An immunohistochemical study. Scand. J. Immunol. 26: 653, 1987. Woessner, J. F. The determination of hydroxyproline in tissue and protein samples containing small proportions of the imino acid. Arch. Biochem. Biophys. 93: 440, 1961. Swift, M. E., Burns, A. L., Gray, K. L., and DiPietro, L. A. Altered inflammatory response in wounds of aged mice. J. Invest. Dermatol. 117: 1027, 2001. Carrel, A. Cicatrization of wounds. XII. Factors initiating regeneration. J. Exp. Med. 24: 429, 1916. Carrel, A. Effect d’un abcesa distance sur la cicatrization d’una plaque aseptique. C. R. Soc. Biol. 90: 333, 1924. Houck, J. C., and Jacob, R. Connective tissue. II. Distant dermal response to local inflammation. Proc. Soc. Exp. Biol. Med. 106: 145, 1961. Fahey, T. J. D., Sadaty, A., Jones, W. G. D., Barber, A., Smoller, B., and Shires, G. T. Diabetes impairs the late inflammatory response to wound healing. J. Surg. Res. 50: 308, 1991. Mast, B. A. Wound Healing: Biochemical and Clinical Aspects. Philadelphia: Saunders, 1992. Leibovich, S. J., and Ross, R. The role of the macrophage in wound repair: A study with hydrocortisone and antimacrophage serum. Am. J. Pathol. 78: 71, 1975. Hunt, T. K., Knighton, D. R., Thakral, K. K., Goodson, W. H. D., and Andrews, W. S. Studies on inflammation and wound healing: Angiogenesis and collagen synthesis stimulated in vivo by resident and activated wound macrophages. Surgery 96: 48, 1984. DiPietro, L. A., Burdick, M., Low, Q. E., Kunkel, S. L., and Strieter, R. M. MIP-1alpha as a critical macrophage chemoattractant in murine wound repair. J. Clin. Invest. 101: 1693, 1998. Sartorelli, K. H., Silver, G. M., and Gamelli, R. L. The effect of granulocyte colony-stimulating factor (G-CSF) upon burn-induced defective neutrophil chemotaxis. J. Trauma 31: 523, 1991. Swartz, D. E., Seely, A. J., Huang, F., Giannias, B., and Christou, N. V. Polymorphonuclear neutrophil (PMN) recruitment in a 2-front murine injury model: Triage of PMNs to competing stimuli of recruitment. Surgery 127: 323, 2000. Ahmed, N. A., McGill, S., Yee, J., Hu, F., Michel, R. P., and Christou, N. V. Mechanisms for the diminished neutrophil exudation to secondary inflammatory sites in infected patients with a systemic inflammatory response (sepsis). Crit. Care Med. 27: 2459, 1999. Newsome, T. W., and Eurenius, K. Suppression of granulocyte and platelet production by Pseudomonas burn wound infection. Surg. Gynecol. Obstet. 136: 375, 1973. Cooney, W. P. Somatic versus sympathetic mediated chronic limb pain. Experience and treatment options. Hand Clin. 13: 355, 1997.
The Effect of Sepsis on Wound Healing 1 Rebeca M. Rico, M.D., Renato Ripamonti, M.D., Aime L. Burns, M.S., Richard L. Gamelli, M.D., FACS, and Luisa A. DiPietro, D.D.S., Ph.D. 2 Burn and Shock Trauma Institute, Department of Surgery, Loyola University Medical Center, Maywood, Illinois 60153 Submitted for publication October 23, 2000; published online December 6, 2001
Background. Normal wound healing is a regulated sequence of events that successfully restore tissue integrity. Previous studies have suggested that wound healing is impaired in a septic host. The current study examines the effect of sepsis on the inflammatory and proliferative phases of wound healing at a remote site of secondary injury. Methods. Polyvinyl alcohol sponges, either inoculated with a standard dose of Pseudomonas aeruginosa (experimental) or soaked in normal saline (control), were placed subcutaneously in the anterior abdominal region of male B6D2F1 mice. Immediately following sponge placement, full thickness excisional dermal wounds were created on the dorsum. Wound healing was examined at days 3, 5, and 7 postinjury. The infiltration of neutrophils and macrophages into wounds was quantified, and the reepithelialization rate and collagen content were measured. Results. Peripheral neutrophil counts were significantly elevated in infected mice, yet neutrophil content of the remote wound of infected animals was significantly reduced (5% of control, P < 0.05). Wounds of infected mice also showed a 30% reduction in the macrophage content. Wounds of infected animals exhibited delayed reepithelialization (76 ⴞ 3 vs 97 ⴞ 3% at day 5, P < 0.05) and collagen synthesis (55.3 ⴞ 9.5 vs 105 ⴞ 13.0 g/wound, P < 0.05). Conclusion. Systemic infection alters both the inflammatory and the proliferative processes at remote sites of injury. Multiple factors seem likely to contribute to the increased incidence of wound complications in septic patients. © 2001 Elsevier Science 1
This work was supported by the Dr. Ralph and Marion C. Falk Medical Research Trust (L.A.D., R.L.G.) and National Institutes of Health Grants GM55238 (L.A.D.) and GM42577 (R.L.G.). 2 To whom correspondence and reprint requests should be addressed at Loyola University Medical Center, Burn and Shock Trauma Institute, 2160 S. First Avenue, Maywood, IL 60153. Fax: (708) 327-2813. E-mail: [email protected].
Key Words: wound healing; sepsis; neutrophils; macrophages; epithelium; collagen. INTRODUCTION
Sepsis is the most common cause of deaths in the intensive care unit and is the 13th leading cause of death in the United States for persons older that 1 year of age. The care of septic patients accounts for an estimated 5 to 10 billion dollars in medical costs annually [1, 2]. Septic patients are at an increased risk for problems related to poor wound healing, such as anastomotic leaks, fascial dehiscence, and nonhealing wounds. In addition, these patients are more likely to develop secondary infections such as pneumonia and urinary tract infections. These complications result in increased morbidity and mortality in this patient population [2, 3]. The mechanism by which systemic infection alters the inflammatory and reparative response at remote sites of injury is not fully understood [4]. Previous studies have shown that wound bursting strength and collagen content are decreased in animals suffering from systemic infection [3, 5]. However, little is known about the mechanism behind this abnormal wound healing. The current study examines the wound repair process during sepsis and provides unified information about both the inflammatory and the proliferative phases of wound healing in a defined model system. METHODS Animals. All animal protocols used in this study were reviewed and approved by the Loyola University Institutional Animal Care and Use Committee. Male B6D2F1/J mice (The Jackson Laboratory, Bar Harbor, ME) 6 to 8 weeks of age and weighing 22 to 27 g were acclimatized for 1 week before use in an approved animal care facility with a 12-h light/dark cycle. Water was supplied ad libitum. Because previous studies suggest that a nutritional deficit may contribute to the poor healing that is seen with sepsis [5], control and
193
0022-4804/01 $35.00 © 2001 Elsevier Science All rights reserved.
194
JOURNAL OF SURGICAL RESEARCH: VOL. 102, NO. 2, FEBRUARY 2002
experimental animals were pair fed. Standard rodent laboratory chow (5001; Ralston Purina Co., St. Louis, MO) was weighed daily for food consumption measurements. Every morning, control animals were pair fed with an amount of food equal to that consumed by the Pseudomonas- infected animals. Wounding and infectious challenge. Mice were anesthetized by inhalation of methoxyflurane (Metofane; Schering–Plough Animal Health Corp., Union, NJ). A 0.5- to 1.0-cm skin incision was made on the right lower quadrant of the anterior aspect of the abdomen of each mouse. A polyvinyl alcohol sponge disc (8 ⫻ 2 mm) was placed subcutaneously. For all experiments except the assessment of peripheral neutrophil count, the sponges for the infected group (n ⫽ 6) were inoculated with 250,000 –300,000 CFU Pseudomonas aeruginosa, ATCC 19960 (Rockville, MD). This strain of Pseudomonas has been shown to induce septicemia in mice when inoculated at burn wound sites at low doses [6]. Sponges of the control group (n ⫽ 6) contained normal saline. Tissues from the same mice were used for analysis of wound reepithelialization, collagen content, and inflammatory cell content. For the examination of peripheral neutrophil levels, sponges were inoculated with no (n ⫽ 4), 135,000 CFU (n ⫽ 8), or 270,000 CFU (n ⫽ 6) of P. aeruginosa. Following sponge placement, the dorsum of each animal was shaved and briefly washed with 70% isopropanol. A standard dermal biopsy punch (Acuderm, Ft. Lauderdale, FL) was used to place six full-thickness excisional wounds (3 mm in diameter) on the dorsal skin of each mouse. Wounds were covered with Tegaderm (3M Health Care, St. Paul, MN). At 3, 5, or 7 days after injury, the mice were euthanized by halothane inhalation (Halocarbon Laboratories, Riveredge, NJ). For histologic and inflammatory cell analysis, wounds and approximately 1 mm of surrounding tissue were excised with a 5-mmdiameter standard dermal biopsy punch. For biochemical analysis of collagen content, only the central portion of the wound was collected with a 3-mm-diameter standard biopsy punch. Wounds were either flash-frozen in liquid nitrogen or embedded in TBS Tissue Freezing Medium (Triangle Biomedical Sciences, Durham, NC). All samples were stored at ⫺80°C until the time of analysis. Bacterial cultures. At postinjury days 1 and 3, the liver and spleen of each animal were aseptically dissected. The tissue was then cultured in tryptic soy broth for 24 h at 37°C. An aliquot of the culture was then plated onto tryptic soy agar and incubated for a further 24 h. The colonies were then counted and Gram stained with Brown Hopps Gram stain (Newcomer, Middleton, WI) for microbiological analysis. Peripheral neutrophil count. Blood was collected from anesthetized mice in heparin (Elkins–Sinn, Inc., Cherry Hill, NJ)-coated syringes via cardiac puncture. To lyse red cells, an aliquot was mixed with a solution of 3% acetic acid, 0.16% saponin (1:20). Total white cell counts per milliliter were determined using a standard Neubauer hemacytometer. The neutrophil percentage was determined by analysis of a blood smear; a total of 200 white cells per smear were counted and differentiated. The total neutrophil count was determined by multiplying the percentage of neutrophils by the total white cell count. Analysis of wound myeloperoxidase. The measurement of myeloperoxidase (MPO) levels in tissue has been shown to reflect neutrophil content. MPO was measured using previously described methods [7, 8]. Briefly, wounds were homogenized in 2.0 ml of 20 mM phosphate buffer, pH 7.4. Homogenates were centrifuged at 12,000g for 45 min, and the supernatant was decanted. The pellets were resuspended in 1.0 ml of 50 mM phosphate buffer containing 10 mM EDTA and 0.5% hexadecyltrimethylammonium bromide (HTAB). After a freeze–thaw cycle, the samples were sonicated briefly and incubated at 60°C for 2 h. The samples were centrifuged at 500g for 10 min and the supernatant was transferred to 1.5-ml tubes for storage at ⫺20°C. For analysis, samples were thawed, and an MPO standard curve ranging from 0 to 3.0 units/ml was generated. Fiftymicroliter aliquots of samples or standards were placed in 12 ⫻ 75-mm glass tubes with 500 l of assay buffer (0.1 M phosphate
buffer, pH 5.4, 1% HTAB, 0.43 mg/ml 3,3⬘,5,5⬘-tetramethylbenzidine). The reactions were started by addition of 50 l of 15 mM H 2O 2, incubated at 37°C for 15 min, and stopped with 1.0 ml of cold 0.2 M sodium acetate, pH 3.0. The absorbance of each sample was read at 655 nm within 10 min of completion. All samples and standards were tested in duplicate. Quantitation of macrophage infiltration. Sections were fixed in acetone at room temperature for 15 min, pretreated with 3% H 2O 2 in methanol to block endogenous peroxidase, and then blocked for nonspecific binding with normal mouse serum (1:100). Primary antibody, MOMA-2, which recognizes an intracellular antigen on murine macrophages and monocytes (Serotec Laboratories, Oxford, UK; 1:500), was utilized to detect these cells in the wound [9]. Primary and secondary antibody (biotinylated mouse anti-rat IgG; Jackson Laboratories, West Grove, PA; 1:100) incubations were performed for 30 min each, followed by a 30-min incubation with avidin– biotin– horseradish peroxidase complexes (ABC; Vector Laboratories, Burlingame, CA). Color development was performed with 3,3⬘diaminobenzidine and slides were counterstained with Harris hematoxylin. For each section, the number of macrophages within the wound bed was counted in 6 to 10 random high-power fields with the aid of an optical grid. All analyses were performed blinded. Analysis of wound reepithelialization. Reepithelialization of excisional wounds was measured by histomorphometric analysis of tissue sections from the central portion of the wound. Using a standard ocular grid, the distance between the wound edges and the distance that the epithelium had traveled across the wound were measured. The percentage reepithelialization ((distance covered by epithelium/distance between wound edges) ⫻ 100) was calculated for each section. Analysis of wound collagen content. Collagen content was assessed by determining the hydroxyproline content [10]. All reagents for this assay were purchased from Sigma Chemical Company (St. Louis, MO). Frozen wound tissue was hydrolyzed in 2.0 ml of 6 N HCl for 3 h at 130°C or overnight at 110°C. The reaction was neutralized with 2.5 N NaOH and diluted 40-fold with deionized H 2O. One milliliter of a 0.05 M chloramine T solution was added to 2 ml of the neutralized/diluted solution and incubated for 20 min at room temperature. One milliliter of 3.15 M perchloric acid was added and the solution was incubated for 5 more minutes at room temperature. One milliliter of 20% p-dimethylaminobenzaldehyde was then added and the resulting mixture was incubated for 20 min at 60°C. The samples were cooled with cold tap water. Absorbance at 557 nm was measured spectrophotometrically, and the amount of hydroxyproline was determined by comparison to a standard curve. Statistical analysis. Data were analyzed using GraphPad Prism, version 2.01 (GraphPad Software, Inc., San Diego, CA). The means and SEM were calculated for each data set. To compare neutrophil, macrophage, and collagen contents of wounds, a paired t test analysis was used. To compare the degree of reepithelialization over a time course between the two groups, a two-way ANOVA was employed. P values less than 0.05 were considered significant. To compare peripheral neutrophil counts, data were subjected to a one-way ANOVA followed by a Newman–Keuls multiple comparisons test.
RESULTS
Effects of sepsis on mortality and weight loss. The induction of sepsis resulted in a 50% mortality in the infected group, with most deaths occurring at 48 –72 h postinfection. There was one unexplained death in the control group at day 5 (Fig. 1). Bacterial cultures from liver and spleen of both the infected and the control group were obtained at day 1 and day 3 post-sepsis induction. All bacterial cultures from the infected group were positive for growth of gram-negative rods
RICO ET AL.: SEPSIS AND WOUND HEALING
195
FIG. 1. Mortality of control and infected mice. The induction of sepsis by implantation of 250,000 –300,000 CFU of P. aeruginosa resulted in a 50% mortality in the infected group, with most deaths occurring at 48 –72 h.
while those of the control group were negative (n ⫽ 6) (data not shown). When wound sections from day 3 postinjury were Gram stained, no histologic evidence of Pseudomonas was seen. Therefore, although the experimental group was systemically infected, no evidence of infection at the wound site was observed. The effect of systemic infection on weight loss was also examined. Both groups of animals were weighed at day 0 and at days 3, 5, and 7 postinjury. No significant difference in weight loss between the two groups was observed at any time point (Table 1). Although the control mice appeared to have lost more weight than the infected group, this difference was not statistically significant. Peripheral white blood cell counts. Peripheral total white blood cell counts and neutrophil counts were determined at day 3 postwounding in mice infected with either 135,000 (n ⫽ 8) or 270,000 CFU (n ⫽ 6) and were compared to those of control uninfected mice (n ⫽ 4). Infected mice from both groups exhibited a significant elevation in both peripheral total WBC number and peripheral neutrophil levels (Fig. 2). Effects of sepsis on wound inflammation. To evaluate the effects of sepsis on the inflammatory phase of wound healing, neutrophil and macrophage infiltration into the wound was examined at day 3 postinjury. The relative levels of neutrophils between the wounds of the infected and the control mice were determined by measuring myeloperoxidase levels. Surprisingly, the neutrophil content in the wounds of the infected animals was only 5% of the control group value (P ⬍ 0.05, Fig. 3). In addition, a 30% reduction in the number of wound mac-
FIG. 2. Peripheral white blood cell counts in control mice (n ⫽ 4) and those receiving either 135,000 (n ⫽ 8) or 270,000 CFU (n ⫽ 6) of P. aeruginosa. Peripheral white counts were taken at day 3 after infection. (A) Total peripheral white blood cell counts. (B) Total neutrophil counts. Bars, means; lines, SEM. Data were compared by one-way ANOVA and Newman–Keuls post hoc multiple comparisons test. *P ⬍ 0.01, **P ⬍ 0.05, compared to control.
rophages in the infected animals versus the control group was observed (16 ⫾ 2 vs 24 ⫾ 3 cells/HPF at day 3, Fig. 4). Thus, infected mice exhibited significant changes in the inflammatory response at the wound site. Neutrophilic infiltrate was also examined histologically at days 3 and 5 following injury. Little to modest neutrophil infiltration
TABLE 1 Average Weight Loss of Control versus Infected Mice Day postinjury
Control ( ⫾ SEM)
Infected ( ⫾ SEM)
3 5 7
2.30 ⫾ 0.46 2.37 ⫾ 0.45 2.06 ⫾ 0.18
1.68 ⫾ 1.42 1.20 ⫾ 0.73 1.46 ⫾ 1.30
FIG. 3. The effect of sepsis on wound neutrophil content. Neutrophil content of dermal wounds was assessed by measurement of myeloperoxidase, a neutrophil enzyme marker. Myeloperoxidase concentration per wound (⫾SE) was determined at day 3 after injury for both control and infected mice receiving 250,000 –300,000 CFU of P. aeruginosa, n ⫽ 6. Bars, means; lines, SEM. *P ⬍ 0.05 compared to control.
196
JOURNAL OF SURGICAL RESEARCH: VOL. 102, NO. 2, FEBRUARY 2002
FIG. 4. Effects of sepsis on wound macrophage content. The number of macrophages per high-power field (⫾SE) was determined by immunohistochemical analysis for wounds at day 3 after injury. Infected mice received 250,000 –300,000 CFU of P. aeruginosa, n ⫽ 6 for each group. Bars, means; lines, SE.
was observed in wounds of both experimental and control mice at day 5, and no difference was observed between the groups (data not shown). This observation is supported by our previous studies in this wound model. Quantification of neutrophils in this model has shown that the day 5 time point is well past peak neutrophil infiltration into the wound [11]. Effects of sepsis on the proliferative response in the healing wound. To evaluate the proliferative phase of wound healing, the amount of reepithelialization and the collagen content of each wound were measured. Compared to the control group, reepithelialization in the infected animals was significantly inhibited, and infected mice showed reduced reepithelialization at each of three time points that were examined (Fig. 5). Whereas wounds of the control group exhibited nearly complete closure at day 5 postinjury, wounds of infected mice were only 76% reepithelialized (P ⬍ 0.05, Fig. 5). Similarly, the collagen content of the wounds was also reduced in the infected animals versus the control group (Fig. 6). At day 7, hydroxyproline levels in the wounds of infected mice were only about 50% those of control mice (55.3 ⫾ 9.5 vs 105 ⫾ 13.0 g/ wound, P ⬍ 0.05). Thus, the proliferative aspects of excisional wound repair were significantly delayed by systemic infection.
FIG. 5. Time course of reepithelialization in control versus infected mice. Percentage reepithelialization (⫾SEM) is shown, n ⫽ 6 (days 3, 5) and n ⫽ 4 (day 7). Infected mice received 250,000 – 300,000 CFU of P. aeruginosa. *P ⬍ 0.05 between groups by twoway ANOVA.
FIG. 6. Effects of sepsis on wound collagen content. Hydroxyproline content per wound (⫾SEM) was determined at day 7 after injury. Infected mice received 250,000 –300,000 CFU of P. aeruginosa, n ⫽ 4 for each group. Bars, means; lines, SEM. *P ⬍ 0.05 compared to control.
DISCUSSION
Studies dating back to those performed by Dr. Alexis Carrel in 1924 have demonstrated that the rate of wound contraction is inhibited by infection and that a distant abscess impairs healing [12, 13]. More recent studies have documented that both wound strength and collagen content are diminished in a systemically infected animal [5]. Other studies have shown that both a systemic transient bacteremia and a distant inflammation from intradermal croton oil result in impaired wound healing [3, 14]. These studies, as well as the present one, support the concept that inflammatory responses at sites distant from the wound bed can profoundly influence the pace of local tissue inflammation and repair. Aberrant healing occurs in numerous circumstances, including diabetes, peripheral vascular disease, and cancer. Not surprisingly, one of the common denominators of poor wound healing in these conditions is the alteration of the inflammatory response [15, 16]. Previous studies have demonstrated the importance of leukocytes in orchestrating the proliferative response in wound healing [17–19]. As the first leukocyte to enter the wound, neutrophils provide the initial line of defense against infection. In the face of systemic infection, the ability of these leukocytes to perform critical functions such as chemotaxis and cytokine production may be impaired [20]. Likewise, the depletion of monocytes and macrophages results in impaired healing, including decreased fibroblast proliferation, and immature fibrosis [17]. Macrophages are known to produce factors that stimulate both angiogenesis and collagen synthesis [17, 18]. In the case of systemic infection, the recruitment of leukocytes is notably affected, and several previous studies have documented a decrease in peripheral leukocyte infiltration in the septic condition. In a murine cecal ligation and puncture-induced peritonitis model, neutrophilic recruitment to a remote site of secondary injury (skin) was shown to be significantly reduced in septic mice [21]. In addition, a 72% reduction in neutrophil influx into skin wound blisters was seen in septic patients compared to healthy subjects [22]. In contrast, alterations in macro-
RICO ET AL.: SEPSIS AND WOUND HEALING
phage content of wounds of septic animals are not yet as well documented. The significance of a decrease in wound macrophages to the delayed wound proliferative response seen in septic animals will require further investigation. The mechanism for the decrease in neutrophils and macrophages in the wounds of the infected animal is not yet completely understood. One postulate is that a finite number of neutrophils are available for delivery to one or more sites of injury and that neutrophils are concentrated in the site of the predominant injury [21]. However, the dramatic decrease in neutrophils at the wound site occurs in the face of an increase in the number of circulating neutrophils in the periphery, suggesting that other mechanisms come into play. In burn injury, and burn sepsis, the chemotactic response of neutrophils has been shown to be impaired [20]. The mechanism of decreased chemotaxis is probably multifactorial and may involve either changes in surface receptor expression and ligand engagement or alterations in intracellular signaling [22]. A further consideration is that alterations in the responsiveness of peripheral leukocytes may derive from an altered bone marrow environment in the septic host. Significant perturbations in both myelopoiesis and monocytopoiesis have been documented to arise during sepsis [23]. These central changes might influence responsiveness in the periphery and might provide some explanation for the observed effect of sepsis on inflammation at a site of secondary injury. Finally, elevated systemic levels of proinflammatory cytokines are well documented in sepsis. High levels of proinflammatory cytokines, particularly TNF␣, have been shown to inhibit peripheral leukocyte function in wounds [24]. The current study has interesting implications that extend beyond the healing wound. Clinically, septic patients are predisposed not only to wound complications but also to secondary infections, such pneumonia. The described deviations in the local inflammatory response may play a significant role in the increased incidence of secondary infections in this patient population. The effect of systemic infection on the function and recruitment of leukocytes requires additional rigorous examination. Ultimately, such mechanistic information may contribute to the development of therapeutic treatments to counteract this disordered inflammatory response. REFERENCES Darville, T., Giroir, B., and Jacobs, R. The systemic inflammatory response syndrome (SIRS): Immunology and potential immunotherapy. Infection 21: 279, 1993. 2. Barriere, S. L., and Lowry, S. F. An overview of mortality risk prediction in sepsis. Crit. Care Med. 23: 376, 1995. 3. De Haan, B. B., Ellis, H., and Wilks, M. The role of infection on wound healing. Surg. Gynecol. Obstet. 138: 693, 1974.
4. 5.
6.
7.
8.
9.
10.
11.
12. 13. 14.
15.
16. 17.
18.
19.
20.
21.
22.
1.
23.
24.
197
Thornton, F. J., Schaffer, M. R., and Barbul, A. Wound healing in sepsis and trauma. Shock 8: 391, 1997. Greenhalgh, D. G., and Gamelli, R. L. Is impaired wound healing caused by infection or nutritional depletion? Surgery 102: 306, 1987. Rosenthal, S. M. A procedure for measurement of wound healing, with special reference to burns. Proc. Soc. Exp. Biol. Med. 123: 347, 1966. Suzuki, K., Ota, H., Sasagawa, S., Sakatani, T., and Fujikura, T. Assay method for myeloperoxidase in human polymorphonuclear leukocytes. Anal. Biochem. 132: 345, 1983. Schierwagen, C., Bylund-Fellenius, A. C., and Lundberg, C. Improved method for quantification of tissue PMN accumulation measured by myeloperoxidase activity. J. Pharmacol. Methods 23: 179, 1990. Kraal, G., Rep, M., and Janse, M. Macrophages in T and B cell compartments and other tissue macrophages recognized by monoclonal antibody MOMA-2. An immunohistochemical study. Scand. J. Immunol. 26: 653, 1987. Woessner, J. F. The determination of hydroxyproline in tissue and protein samples containing small proportions of the imino acid. Arch. Biochem. Biophys. 93: 440, 1961. Swift, M. E., Burns, A. L., Gray, K. L., and DiPietro, L. A. Altered inflammatory response in wounds of aged mice. J. Invest. Dermatol. 117: 1027, 2001. Carrel, A. Cicatrization of wounds. XII. Factors initiating regeneration. J. Exp. Med. 24: 429, 1916. Carrel, A. Effect d’un abcesa distance sur la cicatrization d’una plaque aseptique. C. R. Soc. Biol. 90: 333, 1924. Houck, J. C., and Jacob, R. Connective tissue. II. Distant dermal response to local inflammation. Proc. Soc. Exp. Biol. Med. 106: 145, 1961. Fahey, T. J. D., Sadaty, A., Jones, W. G. D., Barber, A., Smoller, B., and Shires, G. T. Diabetes impairs the late inflammatory response to wound healing. J. Surg. Res. 50: 308, 1991. Mast, B. A. Wound Healing: Biochemical and Clinical Aspects. Philadelphia: Saunders, 1992. Leibovich, S. J., and Ross, R. The role of the macrophage in wound repair: A study with hydrocortisone and antimacrophage serum. Am. J. Pathol. 78: 71, 1975. Hunt, T. K., Knighton, D. R., Thakral, K. K., Goodson, W. H. D., and Andrews, W. S. Studies on inflammation and wound healing: Angiogenesis and collagen synthesis stimulated in vivo by resident and activated wound macrophages. Surgery 96: 48, 1984. DiPietro, L. A., Burdick, M., Low, Q. E., Kunkel, S. L., and Strieter, R. M. MIP-1alpha as a critical macrophage chemoattractant in murine wound repair. J. Clin. Invest. 101: 1693, 1998. Sartorelli, K. H., Silver, G. M., and Gamelli, R. L. The effect of granulocyte colony-stimulating factor (G-CSF) upon burn-induced defective neutrophil chemotaxis. J. Trauma 31: 523, 1991. Swartz, D. E., Seely, A. J., Huang, F., Giannias, B., and Christou, N. V. Polymorphonuclear neutrophil (PMN) recruitment in a 2-front murine injury model: Triage of PMNs to competing stimuli of recruitment. Surgery 127: 323, 2000. Ahmed, N. A., McGill, S., Yee, J., Hu, F., Michel, R. P., and Christou, N. V. Mechanisms for the diminished neutrophil exudation to secondary inflammatory sites in infected patients with a systemic inflammatory response (sepsis). Crit. Care Med. 27: 2459, 1999. Newsome, T. W., and Eurenius, K. Suppression of granulocyte and platelet production by Pseudomonas burn wound infection. Surg. Gynecol. Obstet. 136: 375, 1973. Cooney, W. P. Somatic versus sympathetic mediated chronic limb pain. Experience and treatment options. Hand Clin. 13: 355, 1997.