Protective effect of plasma in polymorphonuclear neutrophil-mediated cytotoxicity of endothelial cells in the systemic inflammatory response syndrome.

Polymorphonuclear neutrophil (PMN)-mediated endothelial cell (EC) cytotoxicity is well described in many in vitro systems. These observations have been extended in vivo and suggest that circulating PMNs adherent to endothelial cells can damage these endothelial cells and produce the capillary leak that is central to the evolution of a systemic inflammatory response to multiple organ failure. However, most PMC-EC interactions in the circulation must occur in the presence of plasma, therefore we studied PMN-induced endothelial cell cytotoxicity in the absence and presence of autologous plasma from healthy human volunteers and patients with the systemic inflammatory response syndrome (SIRS) requiring admission to an intensive care unit. PMNs from patients with SIRS had increased endothelial cell adherence compared with controls, but equivalent endothelial cell cytotoxicity. Endothelial cell activation with TNF-alpha and IL-1beta markedly increased PMN adherence to endothelial cells. Plasma had a minimal effect on the adherence of PMNs to endothelial cells at baseline or after endothelial cell activation. In contrast, plasma provided endothelial cells with almost complete protection from PMN-induced cytotoxicity at baseline, as well as after endothelial cell activation in both human volunteers and patients with SIRS. The data suggest that PMNs may cause cytotoxic damage in end organs such as the lung only after they diapedese through the endothelial barrier into the extracellular matrix removed from the protective effect of circulating plasma.

Critically ill anergic patients demonstrate polymorphonuclear neutrophil activation in the intravascular compartment with decreased cell delivery to inflammatory focci.

Skin test anergy, the failure to produce a delayed type hypersensitivity (DTH) response, is associated with an increase in infection-related complications and death usually due to multiple organ failure (MOF). Refractory intravascular activation of polymorphonuclear neutrophils (PMNs) has been implicated in the development of MOF. We studied 20 critically ill surgical patients with life threatening infections to determine if PMN intravascular activation was present and how this affected essential PMN functions such as exudation. The 11 anergic patients had a more intense inflammatory response to their infection. Plasma lactoferrin was 6.1 +/- 0.3 microgram/ml in anergic patients compared to 3.9 +/- 1.5 in reactive P less than 0.05, accompanied by reduced total primary (3.3 +/- 1.9 vs 4.7 +/- 2.1 micrograms/10(6) PMN P less than 0.01) and secondary (2.8 +/- 0.4 vs 5.0 +/- 0.9 microgram/10(6) PMN P less than 0.01) granule content, respectively. In vitro superoxide production following 100 ng/ml PMA stimulation was 0.44 +/- 0.1 in anergics vs 0.36 +/- 0.1 nmol/microgram PMN protein in reactivities, P less than 0.05. PMN chemotaxis was 8.2 +/- 0.6 PMNs/HPF in anergics compared to 10.2 +/- 1.6 PMNs/HPF in reactives P less than 0.05, accompanied by decreased PMN delivery to skin blister windows (3.2 +/- 1.4 vs 4.5 +/- 1.9 x 10(7) PMN/ml, respectively, P less than 0.05). We conclude that critically ill anergic surgical patients have increased intravascular PMN activation, which may contribute to oxygen-derived tissue damage in the vascular space, as well as a deficient delivery of effector cells in areas of bacterial invasion. This may lead to inability to clear the inflammatory signals which set up the vicious circle of MOF leading to death.

Immunological responses to chronic heat exposure and food restriction in rats.

Immunological variables were studied in rats chronically exposed to high environmental temperature (35 degrees C). Responses were compared with those of rats at 25 degrees C both fed ad libitum and pair fed to the decreased intake found in heat-exposed rats. Heat-exposed rats showed slower delayed-type hypersensitivity responses to keyhole limpet hemocyanin. They showed lower counts of peripheral blood total T cells (OX19+) as well as helper T cells (W3/25+) and smaller numbers of splenic T cells. The thymus was decreased in size. Increased levels of serum IgG antitetanus toxoid antibodies were found in heat-exposed rats. [3H]-thymidine incorporation into Concanavalin A (ConA)-stimulated splenic lymphocytes was decreased in pair-fed rats but not significantly altered in heat-exposed rats compared with controls. Heat exposure alters some aspects of both cellular and humoral immune function in a manner different from that induced by comparable food restriction without heat exposure.

Skin window chambers, a novel method for recovering the cells involved in delayed-type hypersensitivity.

A method is described for the recovery of cells involved in the delayed-type hypersensitivity (DTH) reaction in man. A small plastic chamber was placed over DTH sites from which the stratum corneum had been removed by tape stripping. The chambers were then filled with serum. Analysis of cells harvested at 24 and 48 h has shown an antigen-dependent increment in mononuclear and polymorphonuclear cells in donors with a DTH response. T cell clones were established from the recovered cells; these will provide a useful adjunct to a functional analysis of cells participating in the DTH reaction.

Iron absorption and therapy after gastric bypass.

BACKGROUND: Iron deficiency anemia is a common complication of gastric bypass. The authors assessed the value of taking vitamin C with oral iron in correcting deficiencies in iron stores and anemia postoperatively. MATERIALS AND METHODS: Iron absorption tests were performed on 55 patients 3.2+/-2.0 years after isolated gastric bypass to identify those at higher risk for the late development of anemia. Twenty-nine of this group agreed to a therapeutic trial of iron alone or with vitamin C over a 2-month period. All 55 patients were followed up for 27.1+/-1.0 months following the study. RESULTS: The iron absorption test identified patients with low iron stores, as indicated by low serum ferritin, and those with sufficient absorption surface to benefit from oral iron. The addition of vitamin C appears to enhance the therapeutic effect of iron by correcting ferritin deficits (P < 0.01) and anemia (P < 0.05). Differences in intestine length bypassed by the operation (10 vs. 100 cm) did not affect late ferritin and hemoglobin values. CONCLUSION: This study suggests but does not prove that the addition of vitamin C to iron therapy after gastric bypass is more effective in restoring ferritin and hemoglobin than iron alone. These results are in contrast with the outcome 22.8 months later, when approximately 50% of study patients were again anemic. Closer follow-up of patients is urgently needed.

Activation state of polymorphonuclear leukocytes in surgical patients: characterization of surface receptor expression.

BACKGROUND: We studied the expression of chemotactic, opsonic, and adherent receptors on the membranes of polymorphonuclear neutrophils (PMNs) from surgical patients because modulation of these receptors has been suggested in the control of critical PMN functions that directly or indirectly influence patient outcome. METHODS: Healthy subjects who had PMN adherence, chemotaxis, and receptor measurements were compared with preoperative surgical patients and those who were within 48 hours of an admission to an intensive care unit (surgical intensive care unit) because of an acute illness. RESULTS: The following receptor/cell pattern was found in control subjects, preoperative patients, and patients in the surgical intensive care unit, respectively: formyl peptide (13,000 vs 18,000 vs 22,000), CR3 (59,000 vs 105,000 vs 121,000), fibronectin (21,000 vs 20,000 vs 35,000), FcII gamma (9,000 vs 20,000 vs 25,000), C5a (347 vs 265 vs 250 mean channel number), and FcIIIR (64,000 vs 75,000 vs 26,000). This receptor pattern was partly correlated directly or indirectly with the acute-phase response, neutrophil adherence, and PMN chemotaxis from these subjects. CONCLUSIONS: The data suggest that there are alterations in the expression and modulation of PMN surface membrane receptors in patients with a "stable" disease process compared with those with an acute illness, which may affect critical PMN functions needed to combat bacterial infections.

Neutrophil function in surgical patients: in vitro correlation of abnormal neutrophil chemotaxis by Levamisole.

Cutaneous anergy to recall skin test antigens is associated with decreased polymorphonuclear neutrophil (PMN) chemotaxis (CTX). This decreased PMN chemotaxis is mediated by factors circulating in the sera (AS) from patients with anergy. Levamisole hydrochloride will correct the chemotactic defect of neutrophils from anergic patients, in vitro, from 96.2 +/- 1.2 to 125.1 +/- 1.7 microns at concentrations of 10(-3) M to 10(-18) M. Pretreatment of normal PMN with Levamisole at 10(-4) M will protect them from the chemotactic inhibiting effect of AS. Normal PMN migrating in the normal range, 128.1+/- 2.7 microns, can be made to behave like anergic PMN by treatment with AS. These PMN which now migrate in the anergic range 92.1 +/- 1.7 microns can be converted back to normal by Levamisole treatment at 10(-3) M to 10(-18) M. In 35 surgical patients who demonstrated the spectrum of decreased PMN CTX, the majority toward the anergy level, Levamisole improved the PMN CTX toward normal levels in every instance, while not affecting the CTX of the normally migrating PMN.

In vitro polymorphonuclear neutrophil function in surgical patients does not correlate with anergy but with "activating" processes such as sepsis or trauma.

We studied 199 preoperative patients admitted for esophagogastric, gastric, colonic, or rectal resections, 132 patients with severe blunt trauma, 180 surgical intensive care unit patients with major sepsis, and 95 laboratory controls in order to clarify the role of polymorphonuclear neutrophil (PMN) adherence and chemotaxis to outcome. Patients were also stratified by the delayed-type hypersensitivity response to five ubiquitous antigens. PMN adherence and PMN chemotaxis were not different in preoperative reactive or anergic patients and were equal to the control values, whereas both reactive patients and anergic patients showed altered PMN function after trauma or sepsis. There was no difference in PMN adherence or chemotaxis between patients who died and those who lived. Multiple logistic regression analysis showed that patient age, delayed-type hypersensitivity, and admission serum albumin level, not PMN adherence or chemotaxis, were significantly related to septic mortality. We concluded that altered circulating PMN adherence and chemotaxis is seen in all patients after an "activation" event such as trauma or sepsis. This is a nonspecific immune alteration not related to specific immune events such as delayed-type hypersensitivity; it does not correlate with patient outcome and should not be used as a predictive variable.

Small-bowel bacterial overgrowth and systemic immunosuppression in experimental peritonitis.

The effect of intraperitoneal infection on small-bowel flora and on systemic immunity was studied in a rat model, with use of the delayed-type hypersensitivity (DTH) response to keyhole limpet hemocyanin (KLH) as a measure of global immunologic integrity. Twenty-four hours after the induction of peritonitis by cecal ligation and puncture, concentrations of Escherichia coli in the proximal gastrointestinal tract increased from fewer than 10(3) colony-forming units (CFU)/ml to more than 10(9) CFU/ml, and the DTH response decreased from 10.0 +/- 0.2 to 2.1 +/- 0.4 mm. To assess the contribution of this altered luminal flora to the observed suppression of DTH scores, cecal ligation without puncture was performed in a group of animals whose endogenous flora had been suppressed by administration of oral neomycin. Oral administration of live antibiotic-resistant E. coli to the study animals resulted in significant DTH depression compared with controls given saline solution (2.7 +/- 0.4 vs 4.4 +/- 0.4 mm, p less than 0.005), even though the gastrointestinal tract was anatomically intact. Similar depression was seen if the challenge was limited to the small bowel as a result of the prior performance of an ileostomy and occurred in the absence of significant systemic or portal levels of viable bacteria. The results suggest that gut endotoxin plays a role in the immunosuppression associated with peritonitis.

The influence of long-term protein deprivation on in vivo phagocytic cell delivery to inflammatory lesions.

The effect of long-term protein deprivation and refeeding was assessed on the in vivo delivery of phagocytic leukocytes (PHAGS) to delayed-type hypersensitivity (DTH) reaction and a bacterial abscess. Male inbred Lewis rats sensitized to keyhole limpet hemocyanin (KLH) were fed either a normal diet or a 2% protein diet for 1, 6, and 10 weeks. Two additional groups were fed a 2% protein diet for 10 weeks but were refed with a normal diet for 1 or 4 weeks. At the end of each diet period rats were injected intradermally with KLH, Staphylococcus aureus 502A, and saline solution at different sites at from 1 to 24 hours. Technetium-99m-colloid labeled PHAGS (99PHAG) were injected intravenously and used to assess in vivo PHAG cell delivery. In normally fed rats the peak influx of 99PHAG was at 2 to 4 hours. After 1 week of protein-deficient diet there was a significant drop in early (2 to 4 hours) 99PHAG influx to both the DTH and bacterial reactions. After 10 weeks of protein deprivation (severe malnutrition) there was a further drop and a delay in the peak 99PHAG influx (from 2 to 4 hours, to 8 hours). A return to normal 99PHAG influx occurred only after 4 weeks of refeeding, and it coincided with a return to normal body weight and a normal DTH reaction. There was a direct correlation between total 99PHAG delivery to a DTH reaction and a bacterial abscess (rs = 0.87, Spearman rank; p less than 0.001). We conclude that both moderate and severe protein deprivation is associated with reduced in vivo phagocytic cell delivery to both a DTH reaction and a bacterial skin abscess, which can be restored with refeeding.

Antibiotic prophylaxis for surgery in morbidly obese patients.

The rate of wound infections in morbidly obese patients who underwent gastroplasty surgery at our institution was 16.5% compared with a rate of 2.5% in normal-weight patients who underwent clean-contaminated surgery. Both groups received 1 gm of cefazolin intramuscularly before surgery was performed. We hypothesized that this regimen of prophylaxis did not provide adequate tissue levels in the morbidly obese. Morbidly obese patients who were undergoing gastroplasty were randomly selected to receive 1 gm cefazolin in the buttock fat, buttock muscle, or by intravenous injection. A fourth group of morbidly obese patients received 2 gm of cefazolin intravenously. Normal-weight patients who were undergoing upper abdominal surgery received 1 gm of cefazolin intravenously. At incision and closure, both blood and tissue levels of cefazolin were significantly (p less than 0.001) lower for all morbidly obese patients who received 1 gm cefazolin when compared with the blood and tissue levels of the drug found in normal-weight patients. The cefazolin levels obtained were below the minimal inhibitory concentrations of greater than 2 micrograms/ml for gram-positive cocci and of greater than 4 micrograms/ml for gram-negative rods. Only when the morbidly obese patient received 2 gm cefazolin were both the serum and adipose tissue levels adequate. For a 4-month period, all morbidly obese patients received 2 gm cefazolin prophylaxis, and the wound infection rate dropped to 5.6% compared with the previous rate of 16.5% (p less than 0.03). We conclude that antibiotic prophylaxis must be specially tailored to the needs of these obese patients.

In vivo and in vitro humoral immunity in surgical patients: antibody response to pneumococcal polysaccharide.

We have previously shown that the antibody response to a T cell-dependent protein antigen, tetanus toxoid is reduced in patients having surgery. Because most bacterial antigens are not protein but polysaccharide, we studied in vivo and in vitro antibody responses to a relatively T cell-independent polysaccharide antigen, 23 valent pneumococcal polysaccharide (Pneumovax) (PPS). Subjects were classified by skin testing with five standard antigens as reactive patients (R, positive response to any antigen, N = 5), anergic patients (A, no response, N = 7) or laboratory controls (C, reactive personnel, N = 8). Blood lymphocytes taken before and after immunization with 0.5 ml PPS were cultured in vitro. Quantities of total and anti-PPS IgG, IgM, and IgA in culture supernatants and serum were measured by radioimmunoassay. There was no difference in in vivo anti-PPS production in the three groups (p greater than 0.05 for all comparisons, Wilcoxon rank-sum test). Positive response rates (greater than twofold increase) for all classes of immunoglobulin were also similar in the three groups (X2(2) = 0.36, 0.36, and 0.81 for IgG, IgM, and IgA, p greater than 0.05). In in vitro studies, peak quantities of IgA anti-PPS produced by A were significantly less than C (0.64 X divided by 0.41 versus 2.03 X divided by 0.6, p less than 0.05, Wilcoxon rank-sum test). Synthesis of all other classes of Ig anti-PPS and simultaneous measurement of total Ig (nanograms per culture) produced in vitro were not significantly different among all groups (p greater than 0.05, Wilcoxon rank-sum test). In C, R, and A, peak in vitro isotype-specific anti-PPS production correlated with the magnitude of the in vivo serum response (Spearman rank correlation = 0.53, 0.60, and 0.59 for IgG, IgM, and IgA, p less than 0.05). We conclude that these data show normal in vivo antibody responses to a relatively T cell-independent bacterial polysaccharide antigen in surgical patients and a good correlation of in vivo- to in vitro-specific antibody responses. The data imply that a T cell defect is responsible for reduced humoral immunity to protein antigens. Because most bacterial antigens are not protein but polysaccharides, active immunization of patients with bacterial vaccines may produce effective immunity.

Polymorphonuclear neutrophil (PMN) recruitment in a 2-front murine injury model: triage of PMNs to competing stimuli of recruitment.

BACKGROUND: Intensive care unit patients as a group have the highest rate of nosocomial infections, such as pneumonia, urinary tract infections, and wound infections. The triage of polymorphonuclear neutrophils (PMNs) during an acute inflammatory response was investigated to determine if the severity of injury or infection contributes to PMN delivery. METHODS: A murine cecal ligation and puncture-induced peritonitis model with polyvinyl sponge discs were used to collect the PMNs in the abdomen (primary site) and in the subcutaneous tissue of the dorsum (remote site). Eighty CD1 male mice--20 in each of 4 groups--were assigned to the following: cecal ligation and puncture (CLP), sham laparotomy with cecal manipulation (CM), polyvinyl sponge placement in the abdomen and back only (SP), and sponge placement in the back alone (CON [control]). After 24 hours, the sponges were harvested, and the PMNs were collected and counted on a hemocytometer. RESULTS: These data, reported as mean PMN cells x 10(5) +/- SEM, demonstrated that back sponges contained significantly fewer PMNs in the CLP group (3.29 +/- 1.1) than in the CM group (7.77 +/- 1.61, P = .04), the SP group (8.69 +/- 1.67, P = .01), and the CON group (11.04 +/- 1.91, P < .001). CONCLUSIONS: These results demonstrate that PMN delivery to sites of secondary injury are inversely correlated to the severity of the primary injury or peritonitis.

Induction of an immune response to keyhole-limpet hemocyanin in surgical patients with anergy.

Groups of surgical patients, classified as reactive or anergic on the basis of delayed type hypersensitivity skin testing with five recall antigens, were immunized with keyhole-limpet hemocyanin (KLH) alone or KLH together with mediators derived from leukocytes of a KLH immune donor cultured with antigen. Patients with anergy injected with KLH alone did not generate an immune response as judged by a T cell proliferative reaction performed 14 days after immunization. In contrast, leukocytes of patients with anergy immunized with KLH together with the mediators reacted to KLH in vitro in similar numbers and with a magnitude comparable to that given by reactive, hospitalized patients without anergy immunized with KLH alone. These results confirm and extend our previous observations showing that anergy defined as a lack of cell-mediated immunity to recall antigens such as purified protein derivative extends to the generation of a systemic immune response to a neoantigen such as KLH and mediators that could restore a state of delayed hypersensitivity to purified protein derivative could also be instrumental in inducing cell-mediated immunity de novo when injected together with the antigen.

Malnutrition and humoral immunity: short-term acute nutritional deprivation.

The effects of short-term acute nutritional deprivation and refeeding on immune function was investigated in rats. Animals previously sensitized to keyhole-limpet hemocyanin were starved for 72 hours and refed for 7 days. Recall skin testing with keyhole-limpet hemocyanin and immunization with tetanus toxoid (TT) were used to assess delayed-type hypersensitivity (DTH) and humoral immune responses. DTH was maximally depressed late, after refeeding had begun. Anti-TT responses were depressed early during starvation. Neither DTH nor anti-TT responses had returned to normal after a period of refeeding sufficient to restore weight. The data indicate that short-term acute nutritional deprivation may contribute to acquired immunodeficiency in patients undergoing surgery.

Host defense mechanisms of surgical patients. Friend or foe?

Surgeons and members of this society commonly deal with 2 types of infections of great concern in hospitalized patients. These are hospital-acquired pneumonia and intra-abdominal infections. Both of these infections have the potential for severe morbidity and mortality. We have learned how to classify the types of intra-abdominal infections into primary peritonitis, localized abscess with or without peritonitis, diffuse suppurative peritonitis, or combinations of these classifications. Each of these conditions carries a different mortality risk proportional to its severity. We have also learned how to diagnose surgical infections by properly taking medical history and performing physical examination, appropriate laboratory testing, and sophisticated imaging techniques. The treatment of intra-abdominal infections has become fairly standardized and includes surgical or percutaneous drainage of the infected material, correction of the underlying pathologic symptoms, and broad-spectrum empirical antibiotic therapy.

The local role of tumor necrosis factor alpha in the modulation of neutrophil function at sites of inflammation.

OBJECTIVE: To examine the hypothesis that tumor necrosis factor alpha (TNF-alpha) is an important local modulator of neutrophil function in the inflammatory microenvironment. DESIGN: In vitro studies of host defense. PATIENTS: A volunteer sample of healthy subjects. INTERVENTION: Exudative neutrophils were collected from skin-blister chambers and functionally compared with blood neutrophils. METHODS: Tumor necrosis factor alpha levels at sites of inflammation and neutrophil exudation were determined and compared with serum concentrations. Flow cytometry was used to evaluate neutrophil microbicidal activity and N-formyl-methionyl-leucyl-phenylalanine-induced changes in intracellular calcium and superoxide production. In vitro TNF-alpha was used to evaluate the nature and dose response of TNF-alpha-induced changes in neutrophil function. RESULTS: Exudative neutrophils have an increased responsiveness to subsequent N-formyl-methionyl-leucyl-phenylalanine stimulation, as determined by changes in intracellular calcium. Microbicidal activity and superoxide production are also up-regulated compared with circulating neutrophils. The exudative microenvironment contains TNF-alpha at local levels that are capable of significantly enhancing neutrophil host defense. CONCLUSIONS: Tumor necrosis factor alpha may serve to enhance neutrophil function at sites of inflammation. Neutrophils become more cytotoxic and have an enhanced ability to respond to weak environmental signals.

Circulating and exudative polymorphonuclear neutrophil priming and oxidative capacity in anergic surgical patients.

OBJECTIVES: To examine the oxidative capacity of circulating and exudate polymorphonuclear neutrophils from reactive patients and anergic patients before surgery to determine why anergic patients have increased sepsis-related mortality once an infectious complication develops. DESIGN: Prospective in vitro patient study. SETTING: Tertiary care, major university teaching hospital. PARTICIPANTS: Surgical patients admitted for major elective gastrointestinal surgery. MAIN OUTCOME AND MEASURES: We used flow cytometry and the dye 2-7-dichlorofluorescein diacetate to measure hydrogen peroxide production of circulating and exudate polymorphonuclear neutrophils at baseline and after stimulation with Staphylococcus epidermidis. RESULTS: We found that polymorphonuclear neutrophils were primed in the intravascular space as evidenced by increased numbers of formyl-methionyl-leucyl-phenylalanine receptors, increased hydrogen peroxide production at baseline, and increased hydrogen peroxide production with stimulation. These results were more evident in the anergic patient. After exudation, anergic polymorphonuclear neutrophils lost most of their capacity to produce additional hydrogen peroxide. CONCLUSIONS: The data suggest that this intravascular priming adversely affected polymorphonuclear neutrophils during exudation, more marked in the anergic patient, and may contribute to the sepsis-related mortality of the anergic patients.

Relative contribution of endothelial cell and polymorphonuclear neutrophil activation in their interactions in systemic inflammatory response syndrome.

OBJECTIVE: To examine the relative contribution of polymorphonuclear neutrophil (PMN) vs endothelial cell (EC) activation on the adherence and subsequent killing of ECs by PMNs. DESIGN: In vitro comparative studies of PMN-EC adherence and cytotoxicity. SETTING: Research laboratory and the surgical intensive care unit of a tertiary-level university hospital. PATIENTS: Patients with systemic inflammatory response syndrome admitted to the surgical intensive care unit and hospitalized preoperative noninfected surgical patients. INTERVENTION: None. METHODS: Polymorphonuclear neutrophils were isolated from 21 healthy volunteers, 22 preoperative patients, and 30 patients from the surgical intensive care unit with systemic inflammatory response syndrome. The PMNs were activated with lipopolysaccharide, 100 ng/mL (Escherichia coli 0111:b4), for 40 minutes at 37 degrees C before the adherence and cytotoxicity assays. Human umbilical vein endothelial monolayers were stimulated with tumor necrosis factor alpha, 25 ng/mL, and interleukin 1 beta, 15 U/mL, for 3 hours. The PMNs or EC cells were labeled with sodium chromate Cr 51 and used in a standard adherence or killing assay as required. RESULTS: Control and preoperative patient PMN treatment with lipopolysaccharide produced a modest increase in adherence. The PMNs from patients with systemic inflammatory response syndrome showed moderately increased human umbilical vein endothelial cell adherence, and this could not be augmented further with lipopolysaccharide stimulation. There was a marked increase in PMN adherence to EC after EC activation in all study groups (P < .001). Similar to the adherence data, human umbilical vein endothelial cell cytotoxicity was significantly increased in all groups after human umbilical vein endothelial cell activation (P < .01) but not after PMN stimulation with lipopolysaccharide. CONCLUSION: These data suggest that stimulation of ECs is far more important in producing increased adherence and cytotoxicity of EC than PMN stimulation with lipopolysaccharide in all study groups. Therapeutic efforts in patients with systemic inflammatory response syndrome should be focused on the EC.