Intensive Care Med (2000) 26: 883±892 Ó Springer-Verlag 2000
A. C. Muller Kobold J. E. Tulleken J. G. Zijlstra W. Sluiter J. Hermans C. G. M. Kallenberg J. W. Cohen Tervaert
Received: 4 October 1999 Final revision received: 29 February 2000 Accepted: 4 April 2000 This study was performed with the aid of departmental funds.
)
A. C. Muller Kobold ( ) ´ C. G. M. Kallenberg ´ J. W. Cohen Tervaert Clinical Immunology Division, Department of Internal Medicine, University Hospital Groningen, 9700 RB Groningen, The Netherlands A. C. Muller Kobold ´ J. E. Tulleken ´ J. G. Zijlstra Intensive and Respiratory Care Unit, Department of Internal Medicine, University Hospital Groningen, 9700 RB Groningen, The Netherlands W. Sluiter Department of Internal Medicine, University Hospital Groningen, 9700 RB Groningen, The Netherlands J. Hermans Department of Medical Statistics, Leiden University Medical Center, Leiden, The Netherlands Correspondence address: A. C. Muller Kobold Department of Pathology, University Hospital Groningen, PO Box 30001, 9700 RB Groningen, The Netherlands, e-mail: a.
[email protected], Tel.: + 31-50-3 61 61 61, Fax: + 31-50-3 63 25 10
O R I GI N A L
Leukocyte activation in sepsis; correlations with disease state and mortality
Abstract Objective: The immune response in sepsis shows a bimodal pattern consisting of an early, frequently exaggerated inflammatory response followed by a state of hyporesponsiveness often referred to as the compensatory anti-inflammatory response syndrome (CARS). Insight into the disease state may be helpful in deciding whether to choose immune stimulatory or antiinflammatory therapy in these patients and may determine clinical outcome. We hypothesized that poor outcome in patients with sepsis is related to the severity of CARS, as reflected in the degree of leukocyte activation. Design: Prospective study. Setting: Intensive and respiratory care unit at a university hospital. Patients: Twenty consecutive patients with sepsis and 20 healthy age-matched volunteers. Interventions: None. Measurements and results: Analysis of surface expression of HLA-DR, CD11 b, ICAM-1, CD66 b, CD63 and CD64 on neutrophils and monocytes by flow cytometry and determination of plasma concentrations of lactoferrin, interleukin 6 and neopterin by ELISA at the time of diagnosis. Patient data were related to those of controls; moreover patient data between survivors and non-survivors were compared. Increased expression of all markers, except HLA-DR, was ob-
served on both neutrophils and monocytes from patients compared to healthy controls. HLA-DR expression on monocytes was significantly decreased in patients with sepsis (p < 0.01). Expression of CD11 b and HLE on neutrophils, and ICAM-1 on monocytes, were lower in patients who died compared to those who survived (p < 0.05). Conclusion: In sepsis, both neutrophils and monocytes are activated compared to healthy controls. Poor prognosis is associated with a lower expression of activation markers on monocytes and neutrophils, suggesting that poor outcome in these patients may be due to the compensatory anti-inflammatory response. Key words Sepsis ´ Systemic inflammatory response syndrome (SIRS) ´ Compensatory antiinflammatory response syndrome (CARS) ´ Immunoparalysis ´ Cell activation ´ Monocyte ´ Neutrophil ´ Flow cytometry
884
Introduction Sepsis is a syndrome that complicates other disease states. Mortality associated with sepsis is still unsatisfactorily high, despite modern intensive medicine and antibiotic therapy. Its pathophysiology is, as yet, not completely understood. Sepsis, however, is known to result from dysregulated immune and metabolic responses of the host to injury or infection. In normal situations, pro-inflammatory mediators are released in response to infection, injury and/or ischemia, to eliminate pathogens and to promote wound healing. This response is then downregulated by the release of anti-inflammatory mediators, resulting in the restoration of homeostasis. In sepsis, however, local defense mechanisms are insufficient in eliminating the infectious agent, and the inflammatory response becomes systemic. Moreover, this overwhelming systemic proinflammatory reaction is frequently followed by an overactive compensatory anti-inflammatory response [1]. The balance between pro- and anti-inflammatory responses is frequently lost. This state of immunologic imbalance ranges from persistent massive inflammation to ongoing immune suppression [2]. So far, most therapeutic studies in patients with sepsis have focused on inhibiting pro-inflammatory mediators [3]. Overall clinical responses were, however, disappointing and, in some studies, even increased mortality was observed [1, 4]. Immune stimulation with interferon g (IFN g) has been suggested as a therapeutic strategy to improve monocyte function in selected sepsis patients with impaired immune function [5, 6]. Thus, better understanding of the pro- or anti-inflammatory state of the disease may contribute to a more rational use of different therapeutic regimens in patients with sepsis [3]. In this respect, Volk and co-workers showed that decreased expression of HLA-DR on monocytes from patients with sepsis has predictive value for a fatal outcome [7]. Furthermore, they demonstrated that treatment with IFN g restores monocyte function, as manifested by an increase in HLA-DR expression on monocytes [6]. Most studies have focused on the state of activation of monocytes, whereas little is known about the state of activation of neutrophils from septic patients during the anti-inflammatory response. As monitoring of monocyte HLA-DR expression appears promising in sepsis, we wondered whether membrane expression of activation markers on monocytes and neutrophils reflect the state of pro- or anti-inflammatory immune response in this condition. Since identification of the proor anti-inflammatory response is essential in defining the disease state of sepsis, we investigated the extent of activation of both monocytes and neutrophils in a cohort of patients with sepsis and related our findings to clinical outcome. In order to address this issue, we questioned whether circulating leukocytes in patients with
sepsis were activated compared to circulating leukocytes in healthy controls. More important, we questioned whether the extent of leukocyte activation at the time of diagnosis differed between patients who died due to sepsis and those who survived this disease. Cell activation was measured by flow cytometry, using a non-activating whole blood method and a panel of monocyte and neutrophil activation markers [8, 9].
Patients, materials and methods Patient selection and design of the study The patient group consisted of consecutive patients with sepsis admitted to the intensive care unit (ICU) of our hospital. Inclusion and exclusion criteria for sepsis were identical to those described by Ziegler et al. [10]. Septic shock was defined as sepsis with hypotension, despite adequate fluid resuscitation, along with the presence of perfusion abnormalities that may include lactic acidosis, oliguria or an acute alteration in mental status. Multiple organ failure (MOF) was defined as having at least two signs of organ dysfunction (deterioration in mental state, hypoxemia, oliguria, thrombocytopenia) unrelated to the primary septic focus and not explained by any underlying chronic disease [11]. Patients were entered in the study as soon as they fulfilled the inclusion criteria [10], and received standard treatment. Patients who received immunosuppressive drugs or were immune compromised prior to the development of sepsis were excluded from this study. Informed consent was obtained from each patient or an appropriate family member for collection of blood samples and clinical data in accordance with the human experimentation guidelines for clinical research of our institute. Within 12 h after a patient's admittance to the ICU and fulfilling the inclusion criteria, blood was collected for the analysis of flow cytometric parameters and measurement of plasma concentrations of lactoferrin (LF), neopterin and interleukin 6 (IL-6). Disease severity was scored with the Acute Physiology and Chronic Health Evaluation (APACHE) II scoring system [12] the same day blood samples were taken. During follow-up, length of stay on the ICU was noted. Mortality due to sepsis was defined as death occurring within 28 days after diagnosis. For each patient one healthy age-matched laboratory person served as normal control. Flow cytometric analysis of surface marker expression To investigate the extent of leukocyte activation, markers specific for different stages of leukocyte activation were analyzed on both neutrophils and monocytes from patients and healthy controls. An overview of these markers is given in Table 1. To avoid in vitro activation of leukocytes we used a whole blood method [13]. EDTA-anticoagulated blood was kept on ice until sample preparation. Sample preparation was started always within 5 min after blood sampling. All steps were performed in Hanks' balanced salt solution (HBSS) without calcium and magnesium (Gibco, Life Technologies, Paisley, Scotland, UK), supplemented with 1 % bovine serum albumin (BSA, Boseral, Organon Teknika, Boxtel, Netherlands). Cells were fixed with 1 % paraformaldehyde in phosphate buffered saline (PBS) for 10 min on ice, washed, followed by 2 times erythrocyte lysis with lysis buffer (155 mM NH4Cl, 10 mM KHCO3, 0.1 mM Na2EDTA.H2O) for
885
Table 1 Activation markers on neutrophils and monocytes (Pr3 proteinase 3, MPO myeloperoxydase, HLE human leukocyte elastase, CLB Central Laboratory for the Blood transfusion Service, Amsterdam, The Netherlands, Dako Dakopatts Dako Dakopatts, Glostrup, Denmark, BD Becton Dickinson, Mountain View, California, USA, Immunotech Immunotech, Marseilles, France, Medarex Medarex, Annandale, USA) Marker Markers for priming Pr3 MPO HLE Adhesion molecules CD11b ICAM-1 (CD54) Activation markers HLA-DR CD66b FcgRI (CD64) CD63
Monoclonal antibody
Source
12.8 4.15 NP57
CLB CLB Dako Dakopatts
2LPM19c 84H10
Dako Dakopatts Immunotech
L243 CLB-B13.9 22 CLB-gran/12,435
BD CLB Medarex CLB
5 min at 37 C. A panel of monoclonal antibodies to leukocyte surface antigens was used for the analysis of leukocyte activation (Table 1) [8, 9, 14, 15, 16]. The first antibody was incubated for 1 h at 4 C. After washing, the cells were incubated with a goat antimouse Ig polyclonal antibody conjugated with phycoerythrin (Southern Biotechnology Associates, Birmingham, USA), supplemented with 5 % normal goat serum and 5 % normal human serum, diluted 1:20, for 1 h at 4 C in the dark. Subsequently, the cells were washed and stored until flow cytometric analysis was performed. Analysis of surface marker expression was performed on a Coulter Epics ELITE flow cytometer (Coulter, Hiaelea, Florida, USA), the same day or, in some cases, the next day (always within 18 h). Blood from a healthy age-matched volunteer was analyzed in parallel to every patient sample. When the cell pellet contained erythrocytes, the intercalating dye, LDS751 (Exiton Chemical, Dayton, Ohio, USA) was added before flow cytometry measurement. Erythrocytes could successfully be excluded from the leukocyte population in the LDS751/forward scatter dot plot, when combined with a life gate. Neutrophils and monocytes were identified by forward and sideward scatter. Eosinophils were excluded from the neutrophil population by their high autofluorescence. Data were analyzed using Immuno-4 software [17]. Initially, the flow cytometer was calibrated by using QC3 beads (Flow Cytometry Standards, Leiden, The Netherlands). During the study these beads appeared to be unstable and demonstrated significant batch-to-batch variation. To reduce the influence of the day-to-day variation of the flow cytometry measurements, in parallel to every patient sample, a blood sample from a healthy age-matched control was analyzed simultaneously. To address the questions mentioned in the introduction and to reduce the influence of day-to-day variation, these measurements were expressed as follows: the expression of surface markers was calculated as a percentage decrease or increase compared to the expression (mean fluorescence intensity, MFI) on monocytes and neutrophils from matched healthy controls, corrected for non-specific binding of the conjugate (NSB) and the percentage of positive cells (pos%), according to the following formula: exp ressionratio =
MFI NSBpatient pos%patient 100 %
MFI NSBcontrol pos%control
If a patient marker value equals the marker value from its matched control the patient's expression ratio will be 100 %. Expression ratio values of a patient above or below 100 % indicate increased or decreased expression compared to the expression on cells from the age-matched healthy control [18, 19]. Sequential analysis of leukocyte activation during follow-up For the sequential analysis of surface expression of activation markers, blood from two patients with sepsis and one healthy volunteer was drawn every day at 9.00 a. m. for the first week and every other day the second week. Subsequently, blood was analyzed for cell activation as described above. Data are expressed as expression ratios according to the above-mentioned formula. C-reactive protein (CRP), plasma lactoferrin (LF), neopterin, interleukin 6 (IL-6) C-reactive protein concentrations were measured by using a particle-enhanced nephelometric method and NA latex CRP reagents (Behring, Marburg, Germany). Measurement of LF concentrations in the plasma samples, as a marker for neutrophil degranulation, was performed as described previously [19]. Neopterin concentrations, a plasma marker for monocyte activation [20, 21], were measured by using a commercially available ELISA (Brahms diagnostica, Germany) as previously described [15]. IL-6 concentrations were analyzed by ELISA, according to Helle et al. [22] with modifications [15]. Statistical analysis To address the question of whether the expression of activation markers on cells from patients with sepsis differed from the expression on cells from healthy controls, the expression ratios from patients were evaluated by calculating the 95 % confidence intervals for the means of the patient population. In cases where the confidence interval did not cover the value of 100 %, the mean value of the expression ratios was considered to be significantly increased or decreased compared to 100 % (one sample t-test). Differences in plasma concentrations (LF, neopterin and IL-6) between groups were analyzed by the Mann-Whitney test. Correlation between parameters was analyzed by the Spearman rank correlation test. Differences in leukocyte activation between survivors and non-survivors were analyzed with the Mann-Whitney test. A two tailed p value of 0.05 or less was considered to indicate statistical significance.
Results Patients Twenty consecutive patients with sepsis (12 male, 8 female, median age 69 years, range 29±87) were included in this study. On entry, all patients were clinically suspected of infection that justified the initiation of parental antibiotics. Gram positive bacteria were cultured in nine patients and gram negative bacteria in eight. Additionally, in nine of these patients blood cultures were positive. Seventeen patients presented with septic
886
Table 2 Patient characteristics (m male, f female, CRP C-reactive protein, WBC white blood cell count, APACHE Acute Physiology and Chronic Health Evaluation II score, Survival survival exceeding (yes) or not exceeding (no) 28 days after admission, nt: not tested) Age (years)
Sex
Sepsis due to:
Underlying disease
CRP (mg/l)
WBC 109/l
APACHE
Survival
1 2
64 48
m f
Skin infection Cholangitis
201 19
4.9 12.3
12 22
Yes Yes
3
74
m
Pneumonia
Autoimmune hepatitis Ankylosing spondylitis, amyloidosis None
193
17.2
22
No
4
71
m
Cholangitis
Diabetes mellitus, choledocholithiasis
244
23.7
22
Yes
5
29
f
Peritonitis due to endometritis
None
181
22.4
21
Yes
6
77
m
Urinary tract infection
Myotonic dystrophia
193
14.8
34
Yes
7 8
61 46
m m
Pneumonia Peritonitis
Alcohol abuse Alcohol abuse
203 121
7.8 8.5
31 34
Yes No
Patient no.
9
36
f
Urinary tract infection
Urolithiasis
239
24.8
8
Yes
10 11
80 47
m f
Fecal peritonitis Pancreatitis
None Choledocholithiasis
14 141
4.2 11.7
14 15
Yes Yes
12
67
f
Pneumonia
Diabetes mellitus, breast carcinoma
208
33.1
42
No
13
68
m
Urinary tract infection
Crohn's disease, urolithiasis, diabetes mellitus
nt
27.8
28
Yes
14
87
f
Pneumonia
None
234
17.4
25
No
15 16
70 79
m f
Urinary tract infection Urinary tract infection
None Urolithiasis
130 206
3.1 6.6
31 29
Yes Yes
17
73
f
Cholangitis
None
148
35.0
35
No
18
76
m
Pneumonia
292
22.4
23
No
19
78
m
Septic arthritis
Chronic obstructive pulmonary disease None
216
5.3
28
No
20
38
m
Aspiration pneumonia
None
181
15.5
32
Yes
Table 3 95 % confidence intervals of activation marker expression ratios on monocytes and neutrophils based on patients with sepsis (Pr3 proteinase 3, MPO myeloperoxydase, HLE human leukocyte elastase, n number of patients analyzed) Marker
n
Monocyte
Neutrophil
Pr3 MPO HLE CD11b ICAM-1 HLA-DR CD66b CD64 CD63
20 20 20 20 20 17 20 19 19
137.3±799.42 149.6±614.1 39.1±679.37 185.1±322.1 157.4±406.7 20.5±67.6 not expressed 151.4±351.1 251.1±463.5
532.4±2347.8 220.6±755.1 ±233.6±1567.8 115.0±625 189.0±757.5 not expressed 255.5±735.4 877.1±3568.5 208.4±463.5
shock. MOF was found in ten patients. In none of the patients was therapy withdrawn or withheld. At the time of diagnosis, patients had a median APACHE II score of 26, range 8±42. Median length of stay on the ICU was 16 days, ranging from 4 to 57 days.
Seven patients died within 28 days after diagnosis. All of these patients died due to refractory shock and MOF due to sepsis. The median period between diagnosis and death was 7 days (range 1±11 days). In all 20 patients blood sampling was performed within 12 h after the patients had fulfilled the inclusion criteria. Patient characteristics are given in Table 2. Leukocyte activation Analysis of surface expression ratios showed that all individual membrane markers for leukocyte activation, except HLE, had a mean value significantly different from 100 %. All these membrane markers, besides HLA-DR, were expressed at higher levels on cells from patients with sepsis compared to healthy controls (95 % confidence intervals all exceeded 100 %, Table 3). In contrast, the expression of HLA-DR on monocytes from patients was significantly lower than that in healthy controls (Table 3) (Figs. 1, 2). None of the neutro-
887
Fig. 1 Box and whisker plots indicating the overall range (extreme bars), 25±75 % range (boxes) and median value (horizontal lines) of the expression of activation markers on monocytes from 20 patients with sepsis. Markers are grouped as markers for priming [surface expression of proteinase 3 (Pr3), myeloperoxidase (MPO) and elastase (HLE)], adhesion (surface expression of CD11 b and ICAM-1), and activation [surface expression of HLA-DR (not expressed by neutrophils), CD66 b (not expressed by monocytes), CD64 and CD63]. A patient marker value identical to its matched healthy control leads to an expression ratio of 100 % (102). For all markers, except HLE, the mean values of the expression ratios are significantly different from 100 %
Fig. 2 Box and whisker plots indicating the overall range (extreme bars), 25±75 % range (boxes) and median value (horizontal lines) of the expression of activation markers on neutrophils from 20 patients with sepsis. (Please see Fig. 1 for details)
phil or monocyte membrane markers of cell activation correlated with disease severity as measured by the APACHE II score, CRP values, or white blood cell count, or with the individual neutrophil, lymphocyte or monocyte counts. Soluble markers for leukocyte activation Lactoferrin, neopterin and IL-6 plasma concentrations were higher in patients with sepsis than in healthy controls (p < 0.05, p < 0.0001, p < 0.01, respectively) (Fig. 3). Neopterin concentrations tended to correlate with disease severity, as expressed by the APACHE II score (r = 0.43, p = 0.07).
Fig. 3 Box and whisker plots indicating the overall range (extreme bars), 25±75 % range (boxes), and median value (horizontal lines) of soluble products of cell activation (lactoferrin, neopterin and IL-6) in plasma from 20 patients with sepsis compared to plasma from healthy individuals (S sepsis, HC healthy controls) (p < 0.05 for all markers)
Relation with survival
CD11 b and leukocyte elastase on neutrophils (p = 0.05 and p < 0.05, respectively), and ICAM-1 on monocytes (p < 0.05; Table 4 and Fig. 4). The expression of other markers on neutrophils and monocytes tended to be lower in patients who died than in those who survived. Plasma concentrations of LF, IL-6 and neopterin did not differ between patients who died and those who survived.
Within 28 days after admittance to the ICU 7 patients died of sepsis, whereas 13 patients survived. HLA-DR expression on monocytes from non-survivors was lower than that on cells from survivors, but this difference did not reach statistical significance. Individual markers that were significantly lower in patients who died were
888
Leukocyte activation during the follow-up of two patients with sepsis
Fig. 4 Box and whisker plots indicating the overall range (extreme bars), 25±75 % range (boxes) and median value (horizontal lines) of ICAM-1 expression on monocytes from patients who survived (survivors) compared to patients who died (non-survivors) (p < 0.05)
To investigate fluctuations of leukocyte activation during the course of sepsis, we measured the expression of HLA-DR on monocytes and CD66 b on neutrophils from two patients during a period of 12 days (Fig. 5). Monocytes from patient 1 showed HLA-DR expression of less than 30 % of healthy control values at the onset of sepsis until day 8, which thereafter gradually increased to normal values. CD66 b expression on neutrophils from this patient was increased compared to healthy control values, but decreased gradually from day 3 to day 8. On day 8 the increase in expression of CD66 b on neutrophils paralleled the increase in HLA-DR expression on monocytes. HLA-DR expression on monocytes from patient 2 on day 1 did not differ significantly from the healthy control values, but decreased during the stay on the ICU, although not below the level of 30 % of controls. CD66 b expression on neutrophils from patient 2 did not change significantly during that period. In line with the decrease in leukocyte activation markers in patient 1, a protracted recovery was seen in this patient, necessitating a stay of 15 days on the ICU, whereas patient 2 recovered fast and left the ICU on day 7.
Table 4 Comparison of survivors versus non-survivors (95 % CI 95 % confidence interval, NS not significant) Non-survivors Neutrophils (ratio) CD11b ICAM-1 CD63 CD64 CD66b MPO Pr3 HLE Monocyte (ratio) CD11b ICAM-1 CD63 CD64 HLA-DR MPO Pr3 HLE Soluble markers Neutrophils LF (mg/ml) Monocyte IL-6 (pg/ml) Neopterin (nmol/l) a
Mann-Whitney U test
p valuea
Survivors
Median
95 %CI
n
Median
95 %CI
n
192 178.5 248.5 1374 365 149.5 650.5 100
141±280 6.7±567.5 142±387 397±3322 196±494 78.5±527 ±29.5±2116 19±260
7 7 7 6 7 7 7 7
296 274 277.5 824 415.5 286.5 628.5 186.5
29±943 138±1057 174±529 246±4661 190±1000 173±1050 198±3295 100±877
13 13 12 13 13 13 13 13
0.05 NS NS NS NS NS NS 0.047
168.5 84.5 228.5 204.5 20.5 404.5 100 100
85±339 ±32±295 80±433 79±355 ±6.6±75.6 134±677 ±19±308 63±134
7 7 7 6 6 7 7 7
320 245 373.5 178.5 29 192 100 100
185±377 197±529 275±560 120±435 15.4±855 5.6±732 100±1144 0±1002
13 13 12 13 11 13 13 13
NS 0.015 NS NS NS NS NS NS
1.16
±2.3±5.9
7
2.01
0.6±5.1
13
NS
1.16 114.5
±2.3±5.9 ±1.6±566
7 7
2.01 34
0.6±5.1 ±6.3±398
13 13
NS NS
889
Fig. 5 Leukocyte activation during follow-up of two patients with sepsis and a healthy individual. The course of expression is shown of HLA-DR on monocytes and CD66 b on neutrophils. Arrows indicate the length of stay on the ICU for patient 1 and patient 2. The 30 % line indicates 30 % of expression of HLA-DR on monocytes from healthy controls (n = 5)
Discussion In this study we demonstrate that both neutrophils and monocytes are activated in patients with sepsis. Furthermore, we show that the extent of activation of neutro-
phils and monocytes at diagnosis correlates with clinical outcome, since cells from patients who died of sepsis had a decreased expression of activation markers compared to patients who survived. This may indicate that extensive cell activation, probably resulting from stimulation by pro-inflammatory cytokines, is favorable for survival in sepsis. Neutrophils play a pivotal role in inflammation since they form the first line of defense against bacterial infections. Neutrophils are capable of adhering to (activated) endothelium via adhesion molecules, migrating to areas of bacterial invasion, and, once there, phagocytosing and killing invading pathogens. The process of transmigration and pathogen-specific phagocytosis is subject to multiple biologic regulations that serve to maximize the defense capacity of neutrophils and to limit damage to surrounding cells. To be able to perform these tasks optimally neutrophils have to be activated by pro-inflammatory cytokines or such bacterial products as endotoxin. Indeed, we demonstrated that circulating neutrophils in patients with sepsis are activated in the circulation, since they expressed increased levels of activation markers, including adhesion molecules, on their cell surfaces. Since activated neutrophils adhere to the endothelium, these adherent cells, in general, can not be detected anymore in the circulation. The increased expression of adhesion molecules on circulating cells, as has been demonstrated in this study, may, therefore, result from a dysregulated interaction with endothelial cells. In septic patients, neutrophil adherence and transmigration is impaired, as shown by the incapacity of neutrophils to transmigrate the endothelial barrier [23, 24]. The increased expression of adhesion molecules on circulating (activated) neutrophils may also be the result of the immense cytokine activation often seen in these patients. This `cytokine storm' may lead to intravascular leukocyte activation. If activated neutrophils interact improperly with endothelial cells and prematurely degranulate, releasing lytic enzymes and other toxic products such as oxygen radicals, they can cause endothelial cell dysfunction [25]. This, subsequently, results in fluid leakage and vessel obstruction, which ultimately results in organ dysfunction. Additionally, the increased expression of adhesion molecules on circulating leukocytes may facilitate cell-cell interactions between neutrophils or other circulating cells, which may, eventually, lead to the formation of microaggregates or clumping of cells and occlusion of capillaries. During sepsis, neutrophils are rapidly mobilized from the bone marrow. These relatively immature neutrophils, however, express lower levels of activation markers [26, 27] and are functionally less active [28, 29, 30]. Therefore, the mobilization of immature neutrophils may account for the neutrophil dysfunction reported by others [23, 24, 31, 32, 33, 34, 35, 36].
890
In this study, we also investigated monocyte activation in patients with sepsis. We found evidence that monocyte activation was disturbed in these patients, since the expression of monocyte activation markers was increased, whereas HLA-DR expression on monocytes was decreased, as found in previous studies [37, 38]. Similar to what has been reported for neutrophil function in patients with sepsis, monocyte function is also impaired, as demonstrated by decreased cytokine production [39], decreased HLA-DR surface expression [7, 37, 40] and associated decreased capacity to present antigens [41]. In previous studies HLA-DR expression of less than 30 % of normal values was considered to result from a compensatory anti-inflammatory response [7, 42]. According to this definition, 67 % of our patients could be defined as having CARS (compensatory anti-inflammatory response syndrome) (data not shown). In our study HLADR expression was lower on the monocytes from nonsurvivors than on those from survivors. This difference, however, was not significant and, probably due to the limited numbers of patients in our study, we could not find a difference in survival between patients with a HLA-DR expression of below 30 % compared to those who had an expression of more than 30 %. Several studies have demonstrated that plasma levels of IL-6 have predictive value for clinical outcome [43, 44, 45]. IL-6 is an anti-inflammatory cytokine [46] and levels of this cytokine may therefore be increased during a state of CARS. In our study IL-6 levels were also higher in non-survivors than in survivors (Table 4), but due to the limited number of patients and the huge differences of plasma concentrations in these patients, this was not statistically significant. Our data on IL-6 could therefore not confirm the prognostic value of this cytokine in patients with sepsis. Although plasma levels of neopterin are indicative for monocyte/macrophage activation both in the circulation and in the tissues [20, 21], plasma levels of neopterin in septic patients who survived were not statistically different from plasma levels in patients who died. We therefore evaluated whether markers of activation on circulating neutrophils and monocytes correlated with clinical outcome as defined by the 28 day survival. The expression of CD11 b and elastase on neutrophils and ICAM-1 on monocytes was higher on cells
from survivors than on cells from non-survivors. CD11 b is the a subunit of Mac-1, an adhesion molecule expressed on both neutrophils and monocytes, but also involved in opsonization of microorganisms [47]. ICAM-1 is also involved in cell adhesion, but has an additional role as co-stimulatory molecule in antigen presentation by antigen-presenting cells, like monocytes [48]. A diminished expression of these markers may indicate a state of CARS, since cell functions such as opsonization, adhesion and antigen presentation are suppressed [23, 24, 41, 49]. Indeed, a prolonged state of CARS is unfavorable for survival [1]. Finally, to illustrate that CARS is associated with decreased monocyte and neutrophil activation we followed two patients with sepsis during a period of 2 weeks following diagnosis. A state of CARS, defined as less than 30 % HLA-DR expression compared to control values, could be detected in one patient. A temporary, relative decrease of CD66 b expression was observed on neutrophils in this patient. By contrast, no change in CD66 b expression occurred in the second patient, who only showed a minor temporary decrease in HLA-DR expression. These longitudinal data, although limited to only two patients, may illustrate that CARS seems to be associated with an impaired activation of monocytes and neutrophils. More extensive studies are needed to confirm these observations. In conclusion, in patients with sepsis systemic immune activation is observed, as demonstrated by the increased expression of activation markers both on monocytes and neutrophils. The expression of HLA-DR on monocytes, however, was decreased compared to healthy controls, suggesting that monocyte function is impaired. Furthermore, the expression of other markers on monocytes (ICAM-1) and neutrophils (CD11 b and HLE) was lower in non-survivors than in survivors, indicating that neutrophil function is also affected during CARS. Preliminary data show that the analysis of activation markers can be used for assessing prognosis. A more extensive study, however, is needed to determine the exact predictive value of these markers for survival. Acknowledgements The authors wish to thank M. G. Huitema, W. W. Oost-Kort, I.Bouwman and G. Mesander for their technical assistance and Dr. P. C. Limburg for his valuable advice.
References 1. Bone RC (1996) Sir Isaac Newton, sepsis, SIRS and CARS. Crit Care Med 24: 1125±1128
2. Bone RC (1996) Immunologic dissonance: a continuing evolution in our understanding of the systemic inflammatory response syndrome (SIRS) and the multiple organ dysfunction syndrome (MODS). Ann Intern Med 125: 680±687
3. Natanson C (1997) Anti-inflammatory therapies to treat sepsis and septic shock: a reassessment (editorial). Crit Care Med 25: 1095±1100 4. Cohen J, Heumann D, Glauser MP (1995) Do monoclonal antibodies and anticytokines still have a future in infectious diseases? Am J Med 99: 45S±52 S
891
5. Polk HC, Cheadle WG, Livingston DH, Rodriguez JL, Starko KM, Izu AE, Jaffe HS, Sonnenfeld G (1992) A randomized prospective clinical trial to determine the efficacy of interferon gamma in severely injured patients. Am J Surg 163: 191±196 6. Docke WD, Randow F, Syrbe U, Krausch D, Asadullah K, Reinke P, Volk HD, Kox W (1997) Monocyte deactivation in septic patients: restoration by IFN gamma treatment. Nat Med 3: 678±681 7. Volk HD, Thieme M, Heym S, Docke WD, Ruppe U, Tausch W, Manger D, Zuckermann S, Golosubow A, Nieter B, et al. (1991) Alterations in function and phenotype of monocytes from patients with septic disease ± predictive value and new therapeutic strategies. Behring Inst Mitt 88: 208±215 8. Liote F, Boval-Boizard B, Weill D, Kuntz D, Wautier JL (1996) Blood monocyte activation in rheumatoid arthritis: increased monocyte adhesiveness, integrin expression and cytokine release. Clin Exp Immunol 106: 13±19 9. De Haas M, Kerst JM, Van der Schoot CE, Calafat J, Hack CE, Nuijens JH, Roos D, Van Oers RH, Von dem Borne AE (1994) Granulocyte colony-stimulating factor administration to healthy volunteers: analysis of the immediate activating effects on circulating neutrophils. Blood 84: 3885±3894 10. Ziegler EJ, Fisher CJ Jr, Sprung CL, Straube RC, Sadoff JC, Foulke GE, Wortel CH, Fink MP, Dellinger RP, Teng NN, et al. (1991) Treatment of gram-negative bacteremia and septic shock with HA-1 A human monoclonal antibody against endotoxin. A randomized, double-blind, placebo-controlled trial. The HA-1 A Sepsis Study Group. N Engl J Med 324: 429±436 11. Bone RC, Sibbald WJ, Sprung CL (1998) The ACCP-SCCM consensus conference on sepsis and organ failure. Chest 101: 1481±1483 12. Knaus WA, Draper EA, Wagner DP, Zimmerman JE (1985) APACHE II: a severity of disease classification system. Crit Care Med 13: 818±829 13. Kuijpers TW, Tool AT, Van der Schoot CE, Ginsel LA, Onderwater JJ, Roos D, Verhoeven AJ (1991) Membrane surface antigen expression on neutrophils: a reappraisal of the use of surface markers for neutrophil activation. Blood 78: 1105±1111
14. Halbwachs Mecarelli L, Bessou G, Lesavre P, Lopez S, Witko Sarsat V (1995) Bimodal distribution of proteinase 3 (PR3) surface expression reflects a constitutive heterogeneity in the polymorphonuclear neutrophil pool. FEBS Lett 374: 29±33 15. Muller Kobold AC, Kallenberg CG, Cohen Tervaert JW (1999) Monocyte activation in patients with Wegener's granulomatosis. Ann Rheum Dis 58: 237±245 16. Muller Kobold AC, Kallenberg CG, Cohen Tervaert JW (1998) Leukocyte membrane expression of proteinase 3 correlates with disease activity in patients with Wegener's granulomatosis. Br J Rheumatol 37: 901±907 17. Sladek TL, Jacobberger JW (1993) Flow cytometric titration of retroviral expression vectors: comparison of methods for analysis of immunofluorescence histograms derived from cells expressing low antigen levels. Cytometry 14: 23±31 18. Muller Kobold AC, Mesander G, Stegeman CA, Kallenberg CG, Cohen Tervaert JW (1998) Are circulating neutrophils intravascularly activated in patients with anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitides? Clin Exp Immunol 114: 491±499 19. Muller Kobold AC, Zijlstra JG, Koene HR, De-Haas M, Kallenberg CG, Cohen Tervaert JW (1998) Levels of soluble Fc gammaRIII correlate with disease severity in sepsis. Clin Exp Immunol 114: 220±227 20. Fuchs D, Weiss G, Wachter H (1993) Neopterin, biochemistry and clinical use as a marker for cellular immune reactions. Int Arch Allergy Immunol 101: 1±6 21. Fuchs D, Weiss G, Reibnegger G, Wachter H (1992) The role of neopterin as a monitor of cellular immune activation in transplantation, inflammatory, infectious and malignant diseases. Crit Rev Clin Lab Sci 29: 307±341 22. Helle M, Boeije L, De Groot E, De Vos A, Aarden L (1991) Sensitive ELISA for interleukin-6. Detection of IL-6 in biological fluids: synovial fluids and sera. J Immunol Methods 138: 47±56 23. Tellado JM, Christou NV (1991) Critically ill anergic patients demonstrate polymorphonuclear neutrophil activation in the intravascular compartment with decreased cell delivery to inflammatory focci. J Leukoc Biol 50: 547±553 24. Terregino CA, Lubkin CL, Thom SR (1997) Impaired neutrophil adherence as an early marker of systemic inflammatory response syndrome and severe sepsis. Ann Emerg Med 29: 400±403
25. Weiss SJ (1989) Tissue destruction by neutrophils . N Engl J Med 320: 365±376 26. Lund Johansen F, Terstappen LW (1993) Differential surface expression of cell adhesion molecules during granulocyte maturation. J Leukoc Biol 54: 47±55 27. Terstappen LW, Safford M, Loken MR (1990) Flow cytometric analysis of human bone marrow. III. Neutrophil maturation. Leukemia 4: 657±663 28. Glasser L, Fiederlein RL (1987) Functional differentiation of normal human neutrophils. Blood 69: 937±944 29. Macey MG, Sangster J, Kelsey SM, Newland AC (1993) Pilot study: effects of G-CSF on neutrophil ex-vivo function post bone marrow transplantation. Clin Lab Haematol 15: 79±85 30. Bass JD (1984) Flow cytometric studies on phagocyte function in bacterial infections. Acta Pathol Microbiol Immunol Scand 92: 167±171 31. Simms HH, D'Amico R (1994) Polymorphonuclear leukocyte dysregulation during the systemic inflammatory response syndrome. Blood 83: 1398±1407 32. Lanser ME, Brown GE, Mora R, Coleman W, Siegel JH (1986) Trauma serum suppresses superoxide production by normal neutrophils. Arch Surg 121: 157±162 33. Antrum RM, Solomkin JS (1986) Monocyte dysfunction in severe trauma: evidence for the role of C5 a in deactivation. Surgery 100: 29±37 34. Lanser ME, Mao P, Brown G, Coleman B, Siegel JH (1985) Serum-mediated depression of neutrophil chemiluminescence following blunt trauma. Ann Surg 202: 111±118 35. Solomkin JS, Jenkins MK, Nelson RD, Chenoweth D, Simmons RL (1981) Neutrophil dysfunction in sepsis. II. Evidence for the role of complement activation products in cellular deactivation. Surgery 90: 319±327 36. Solomkin JS, Bauman MP, Nelson RD, Simmons RL (1981) Neutrophils dysfunction during the course of intra-abdominal infection. Ann Surg 194: 9±17 37. Lin RY, Astiz ME, Saxon JC, Saha DC, Rackow EC (1994) Relationships between plasma cytokine concentrations and leukocyte functional antigen expression in patients with sepsis. Crit Care Med 22: 1595±1602 38. Lin RY, Astiz ME, Saxon JC, Rackow EC (1993) Altered leukocyte immunophenotypes in septic shock. Studies of HLA-DR, CD11 b, CD14 and IL-2R expression. Chest 104: 847±853
892
39. Kremer JP, Jarrar D, Steckholzer U, Ertel W (1996) Interleukin-1, ±6 and tumor necrosis factor-alpha release is down-regulated in whole blood from septic patients. Acta Haematol 95: 268±273 40. Kono K, Sekikawa T, Matsumoto Y (1995) Influence of surgical stress on monocytes and complications of infection in patients with esophageal cancer ± monocyte HLA-DR antigen expression and respiratory burst capacity. J Surg Res 58: 275±280 41. Ayala A, Ertel W, Chaudry IH (1996) Trauma-induced suppression of antigen presentation and expression of major histocompatibility class II antigen complex in leukocytes (editorial). Shock 5: 79±90 42. Reinke P, Volk HD (1992) Diagnostic and predictive value of an immune monitoring program for complications after kidney transplantation. Urol Int 49: 69±75
43. Pinsky MR, Vincent JL, Deviere J, Alegre M, Kahn RJD, Dupont E (1993) Serum cytokine levels in human septic shock. Relation to multiple-system organ failure and mortality. Chest 103: 565±575 44. Friedland JS, Porter JC, Daryanani S, et al. (1996) Plasma proinflammatory cytokine concentrations, acute physiology and chronic health evaluation (APACHE III) scores and survival in patients in an intensive care unit. Crit Care Med 24: 1775±1781 45. Casey LC, Balk RA, Bone RC (1993) Plasma cytokine and endotoxin levels correlate with survival in patients with the sepsis syndrome. Ann Intern Med 119: 771±778 46. Tilg H, Dinarello CA, Mier JW (1997) IL-6 and APPS: anti-inflammatory and immunosuppressive mediators. Immunol Today 18: 428±432
47. Vetvicka V, Thornton BP, Ross GD (1996) Soluble beta-glucan polysaccharide binding to the lectin site of neutrophil or natural killer cell complement receptor type 3 (CD11 b/CD18) generates a primed state of the receptor capable of mediating cytotoxicity of iC3bopsonized target cells. J Clin Invest 98: 50±61 48. Semnani RT, Nutman TB, Hochman P, Shaw S, Van Seventer GA (1994) Costimulation by purified intercellular adhesion molecule 1 and lymphocyte function-associated antigen 3 induces distinct proliferation, cytokine and cell surface antigen profiles in human ªnaiveº and ªmemoryº CD4 + T cells. J Exp Med 180: 2125±2135 49. Rothe G, Kellermann W, Valet G (1990) Flow cytometric parameters of neutrophil function as early indicators of sepsis- or trauma-related pulmonary or cardiovascular organ failure. J Lab Clin Med 115: 52±61
