key: cord-0009501-nn0wkl40
authors: Romaschin, Alexander D.; Foster, Debra M.; Walker, Paul M.; Marshall, John C.
title: Let the Cells Speak: Neutrophils as Biologic Markers of the Inflammatory Response
date: 1998
journal: Sepsis (Boston)
DOI: 10.1023/a:1009769923763
sha: 86cb8a7651e9c75856faad6b4a5bb29683216586
doc_id: 9501
cord_uid: nn0wkl40

nan

Clinical manifestations of an activated systemic in_ammatory response are both common and non-speci~c in critically ill patients. Timely and effective therapy directed against the microbial triggers of sepsis, or against its host-derived mediators, remains unsatisfactory because of our inability to answer two questions de~nitively and rapidly:

1. In a given patient, is the clinical syndrome a result of invasive infection, ie tissue invasion by microorganisms or their products? 2. What is the biologic nature of the host response at a particular point in time?

These questions are inherently complex, since the septic process involves the concerted interplay of multiple dynamic biochemical and physiological cascades that manifest in an unpredictable and highly variable manner. Simpli~ed models of the host immune response in critical illness have been proposed [1] [2] [3] [4] , but not tested or validated in a manner that permits them to be of use in clinical decision-making. The value of any proposed model is to provide mechanistic insights and to identify targets for intervention that can be manipulated in a predictable manner to achieve clinical bene~t. Trauma has been a popular model of the in_ammatory response since the onset of injury is well-de~ned, and the inciting insult readily identi~able. Patrick et al. [1] have described a two hit model of an initial dysfunctional in_ammatory response following major trauma, which contributes to the development of multiple organ dysfunction ( Figure 1 ). Sequential neutrophil priming plays a pivotal role in the genesis of remote organ injury. Faist et al. [2] have proposed a model relating monocyte/macrophage and T-cell interactions to the development of an anergic state following trauma; disruption of the normal balance between TH1 and TH2 lymphocyte subsets which serves as both a marker, and a potential mechanism for developing anergy (Figure 2 ). Both models describe early and late phases of the immune response: an early phase of systemic in_ammation that can be exacerbated by a secondary insult, with resultant progression to a later and potentially lethal immunosuppressed or anergic state. An important as-sumption of these models is that the exaggerated or prolonged early pro-in_ammatory phase triggers a compensatory anti-in_ammatory response which may shift the immune balance to a suppressed phenotype [5] . This anergic state not only predisposes the host to nosocomial infection, but previously immunocompetent effector cells such as granulocytes become injurious due to super activation [6] [7] [8] , uncoordinated function [9] [10] [11] and/or delayed apoptosis [12] .

Models of trauma which delineate the course of immune competence, provide useful parallels in other clinical scenarios such as ischemia-reperfusion, or systemic infection. A logical prediction from these models is that anti-in_ammatory therapies which blunt the hysteresis of the pro-in_ammatory cytokinemic phase would be bene~cial at an early stage, but potentially lethal during the anergic phase. Similarly, interventions designed to augment the immune response in the anergic phase could exacerbate the cytokinemic pro-in_ammatory state and contribute to a deleterious outcome for the patient. De~ning the immune status prior to therapeutic intervention may therefore be of critical importance to selection of therapy.

There is a recognized need for markers which can rapidly identify: 1) the inciting insults, 2) the immunological staging of the patient and 3) the immune response to therapy. Circulating levels of in_ammatory mediators of acute in_ammation such as interleukin 6 [13] , procalcitonin [14] , or C reactive protein [15] may provide clinically useful information about the state of activation of the host septic response. Previous studies have shown that the measurement of sustained plasma pro-in_ammatory cytokines such as TNF-a and IL-6 rather than their peak concentrations identify those patients who develop multiple organ dysfunction and death [16] . An alternative apporach to the use of levels of in_ammatory mediators as diagnostic markers is to analyze the responses of the cellular effectors of an in_ammatory response-to let the cells speak for themselves. As the foot soldiers of acute in_ammation, neutrophils are ideally suited to ful~ll this role.

The classic criteria for SIRS include either neutropenia or neutrophilia [17] . Neutropenia occuring a a consequence of malignancy or chemotherapy carries an increased risk of mortality, and identi~es a clinical state that can be treated with recombinant G-CSF [18] . The risk of ICU [19] or hospital [20] mortality also increases with increasing degrees of neutrophilia, although neutrophilia is neither speci~c for infection, nor diagnostic of a particular pattern of in_ammatory mediator response.

Although elevation of the white blood cell count is commonly interpreted as evidence of possible infection, neutrophilia is neither a sensitive nor speci~c marker of infection. The neutrophil count may be elevated, for example following upper gastrointestinal hemorrhage [21] or blood transfusion [22] . Moreover the diagnostic value of a left shift, re_ecting release of newly-formed white cells from the bone marrow, is similarly limited as a marker of systemic in_ammation [23] .

While quanti~cation of the number of circulating neutrophils provides little information about the nature or evolution of an in_ammatory response, evaluation of their functional status may be more informative, both as a measure of the degree of activation or deactivation through the process of programmed cell death, or as a means for the dynamic assay of antigens such as bacterial lipopolysaccharide in the blood. We present here a novel method that uses the cellular response repertoire of neutrophils, as measured by whole blood chemiluminescence, to identify bacterial products and to assess the state of neutrophil activation within the milieu of the patients immunologic pro~le. Measurement of bacterial products (endotoxin) provides a measure of a major stimulus for the septic in_ammatory response. The immune status of the patient may be re_ected in the degree of neutrophil activation and their ability to respond to agonists and immune complexes. Repeated measures can be used to observe the in_uence of therapeutic intervention on neutrophil activity as a surrogate measure of ef~cacy. Further bene~ts of this approach include rapidity and simplicity of analysis using an analytical protocol which does not require cell puri~cation and mimics the in vivo cellular milieu. The approach is simple and minimizes factitious cell activation during cell puri~cation and possible contamination from exogenous LPS.

The measurement of neutrophil chemiluminescence has been pioneered by Allen [24, 25] . Neutrophil respiratory burst activity can be measured in the presence of a lumiphor (light emitting reporter molecule) by the production of NADPH oxidase dependent oxidants such as H 2 O 2 , O 2 -(using lucigenin as the lumiphor) and myeloperoxidase dependent HOCL production (using luminol as the lumiphor). Other lumiphors such as pholasin allow detection of chloramines produced secondary to HOCL release [26] . The respiratory burst activity of normal unstimulated non-in_ammatory neutrophils is minimal and its induction requires stimulation with particulate or soluble agonists such as zymosan or phorbol myristate acetate (PMA). Using particulate stimulants such as zymosan, the choice of ligand dependent stimulation of respiratory burst activity and degranulation in PMN's can be achieved by varying the ligands bound to zymosan. Zymosan A alone stimulates both degranulation and respiratory burst activity via interaction with the glucan binding site on CD11b/CD18 (CR3) [27] . Opsonization of Zymosan A with complement fragments such as iC3b and C3b, leads to synergistic engagement of CR1 (CD35) and CR3 receptors. When IgG molecules are bound to zymosan A, engagement of neutrophil Fc receptors with stimulation of oxidative burst and degranulation occurs. Low concentrations of the soluble diglyceride analogue PMA (10 pmol/assay) can be used to activate the NADPH oxidase complex and cause degranulation of secondary granules while high doses (5 nmol/tube) activate the NADPH oxidase and trigger degranulation of both primary and secondary granules.

Neutrophils from patients with circulating pro-in_ammatory mediators such as TNF-a, PAF, LTB4, C5a, and IL-8 are primed for enhanced respiratory burst activity and also manifest increased oxidant production under quiescent conditions. Pro-in_ammatory molecule priming can result in enhancement of oxidant production which can exceed basal levels by more than 50 fold when measured by complement opsonized zymosan stimulated chemiluminescence. Such priming is associated with an increased surface expression of CR1 and CR3 receptors and conformational activation of CR3 into a higher af~nity "avid" state [28] . The extent of surface CR1 and CR3 receptor expression can be estimated by comparing luminescence curves in the presence and absence of a maximal stimulatory dose of C5a (20 pmol/tube). The ratio of luminescence areas (without C5a divided by maximal C5a during the acceleration phase of neutrophil activation with complement opsonized zymosan), when subtracted from 1 gives the fraction of opsonin receptor reserve which is remaining on the neutrophil surface. In patients with persistent highly activated neutrophils this reserve approaches zero.

Stevens et al. [29] examined the utility of neutrophil chemiluminescence to diagnose and stage infection. Functional analysis of phagocyte activity was determined in healthy volunteers, patients with acute infections of varying severity, patients with diabetes mellitus, and those with human immunode~ciency virus (HIV) infection. The authors were able to discriminate these patient groups on the basis of de~ned neutrophil chemiluminescence parameters using multiple discriminate function analysis. These parameters included basal and PMA stimulated oxidase activity, oxidase driven myeloperoxidase activity, circulating and exo-genously primed opsonin receptor dependent dioxygenation of luminol. Longitudinal studies in patients with soft tissue infection and HIV indicated that chemiluminescence parameters correlated with the degree of disease progression or stage of infection.

In a study of 22 patients following elective repair of an abdominal aortic aneurysm (AAA), and 15 patients with a ruptured AAA [30] , we demonstrated that neutrophils from patients with a ruptured AAA were signi~cantly primed in response to both PMA and zymosan by chemiluminescence analysis; these neutrophils were signi~cantly primed before surgery and maintained a high level of activation for up to 4 days following aneurysm repair. In contrast, neutrophils from patients who had undergone elective AAA repair demonstrated only a modest increase in priming with regard to NADPH oxidase and myeloperoxidase activity when stimulated with high dose PMA. This increase occurred 24 hours post-surgery but also persisted for up to 4 days. These studies indicate that neutrophils from patients with ruptured AAA's are highly primed prior to surgery, presumably as a consequence of hypotension and associated gut ischemia. The mortality in this patient population often exceeds 40% and is associated with multiple organ failure. The high level of neutrophil priming which persists for days post surgery in these patients makes them vulnerable to a "second hit which could provoke dysregulated and injurious neutrophil activity, and perhaps contribute to the development of organ failure.

We have also used whole blood chemiluminescence to identify foci of in_ammation by evaluating neutrophil priming across various tissue beds. By comparing neutrophil priming across the lung in blood taken from the right atrium (mixed venous) and relating the values with arterial (radial arterial line) and femoral venous blood (percutaneous stab) we have been able to distinguish ARDS, or in_ammatory sources above the diaphragm from those with intraabdominal sources of infection and lower limb ischemia.

The rapid detection of lipopolysaccharide (endotoxin) in whole blood has been a daunting analytical problem and a major diagnostic limitation in the a priori selection of patients for treatment with anti-endotoxin agents. Many clinical studies [31] [32] [33] [34] [35] base patient selection predominantly on the suspicion of infection and the presence of physiological signs of the systemic in_ammatory response syndrome (SIRS). Retrospective subset analysis of data from a failed clinical trial using an anti-endotoxin agent [36] showed statistically signi~cant mortality bene~t for patients who were endotoxemic at the time of randomization and received the study drug when compared with endotoxemic patients who received placebo. In contrast, a planned con~rmatory phase III trial of an anti-endotoxin mono-clonal antibody was terminated prematurely because of evidence of increased mortality in patients with Gram positive infections [34] . A major obstacle to the successful implementation of clinical trials of immunologic therapy for patients with sepsis has been lack of diagnostic markers for the activity of the components of the in_ammatory cascade that are targetted by the experimental intervention [37] .

Using neutrophil dependent whole blood chemiluminescence as a platform, we have developed a rapid whole blood assay for endotoxin [38] . The speci~city of the assay is dependent upon the binding of the lipid A moiety of endotoxin with a monoclonal antibody against this antigenic determinant. The assay sensitivity is dependent on the differential priming of neutrophils by LPS-anti-LPS antibody complexes opsonized with complement. Each patient assay is calibrated with an internal LPS standard to compensate for inter-individual differences in neutrophil concentration and oxidative activity allowing standardized endotoxin measurement using the patients own neutrophils and complement proteins as reagents.

An essential element of the assay technique for the quantitation of endotoxin by chemiluminescent assay is measurement of the priming of the patient's neutrophils to a maximal stimulatory concentration of endotoxin. This component of the assay constitutes the determination of the "maximal gain" of the neutrophil priming response to maximal stimulation with LPSanti-LPS antibody complexes. This parameter, that we have termed "responsiveness," is measured by calculating the light integral difference between the control tube (with maximal exogenous LPS) and the antibody containing tube (with maximal exogenous LPS), and re_ects the ability of neutrophils to respond to opsonized immune complexes. Another parameter relating the mathematical product of neutrophil priming and neutrophil concentration termed the "maximum chemiluminescence" (Clmax) provides a measure of the destructive potential of neutrophils in the circulation. As demonstrated in Table 1 a combination of "responsiveness" and "Clmax" can be used to distinguish septic from non-septic patients when combined with LPS measurement. All of these parameters are derived from the chemiluminescent endotoxin assay, and can be measured within the 20 minutes required to perform the assay.

We have evaluated the chemiluminescent endotoxin assay as a screening tool to rule out Gram negative infection in a population of ICU patients including all admissions (study 1, nϭ74) and patients with suspected sepsis (study 2, nϭ104) Table 2 . In a subgroup of 52 septic patients with pneumonia, neither Clmax nor endotoxin levels alone were predictive of mortality. Re-sponsiveness, however, was able to stratify patients with respect to outcome. Those patients whose neutrophils were not signi~cantly primed by maximally stimulatory concentrations of immune complexes were at statistically higher risk of mortality, perhaps because of an inability to resolve invasive infections. Patients who manifested low neutrophil responsiveness and elevated Clmax were at higher risk of multi-organ failure and mortality (unpublished observations).

Neutrophil activation in vivo is re_ected not only in augmented respiratory burst activity, but also in alterations in the process through which neutrophils die. Under normal circumstances, neutrophils are continuously formed from precursor cells in bone marrow stores, and released into the systemic circulation. Here their survival is brief. The life span of the quiescent neutrophil in vivo is no more than~ve or six hours, terminated through the activation of a constitutive All measured values are given as a mean Ϯ 1 SD. For septic and non-septic patients all chemiluminescent measurement were made within 12 hours of ICU admission. Statistical signi~cance was evaluated by student "t" test. Pred; positive predictive, Ϫ Pred; negative predictive. In study 1 all admissions to the ICU were studied over a period of 6 weeks (n ϭ 74). Patients had LPS levels measured twice consecutively over the~rst two days of admission. LPS results were correlated to bacterial cultures drawn within a 24 hour window of the LPS assay. In study 2, 104 patients with suspected sepsis were studied longitudinally, until discharge from the ICU or death. LPS assays were correlated to cultures drawn within 24 hours of chemiluminescent assay. process of programmed cell death or apoptosis [39] . Apoptosis is a highly regulated intracellular process, mediated through the activation of a cascade of cysteine proteases known as caspases, that results in the degradation of nuclear DNA and key cytostructural proteins, and ultimately in the controlled elimination of the cell [40] . DNA fragmentation during apoptosis is readily evident as reduced uptake of the nuclear dye, propidium iodide, by _ow cytometric analysis, or through the characteristic nuclear changes seen by light microscopy. Changes at the cell surface include decreased expression of CD16, and exteriorization of phosphatidyl serine, detected as increased binding of the dye, Annexin V. Augmented expression of phosphatidyl serine permits apoptotic neutrophils to be recognized by~xed tissue macrophages of the reticuloendothelial system, and removed from the circulation without evoking an in_ammatory response [41] .

Apoptosis proceeds more slowly in neutrophils studied in vitro, however by eighteen hours of culture, between 30 and 40% of cells will manifest reduced nuclear uptake of propidium iodide, a hallmark of apoptosis. The relative ease with which these changes can be detected, either by light microscopy or by _ow cytometry, suggests that neutrophil survival in vitro may serve as a marker of cellular activation state, and of the kinetics of the resolution of in_ammation.

The mechanism through which neutrophils are programmed to undergo spontaneous cell death is not well understood. Since neutrophils are known to express both a cellular receptor capable of triggering apoptosis (Fas), and its agonistic ligand, Fas ligand, on the cell surface, it has been suggested that autocrine interactions between these two receptors may result in the expression of the apoptotic program [42] . Such a model would explain why apoptosis is accelerated in neutrophils that have phagocytosed bacteria [43] . In contrast, a wide variety of activational stimuli including bacterial products such as endotoxin [44] , host-derived in_ammatory mediators such as GM-CSF and IL-1b [45] , and the process of transmigration into an in_ammatory focus or the engagement of adhesion molecules of the b2 integrin family [46] , can result in inhibition of the expression of apoptosis, resulting in prolonged neutrophil survival.

Clinical studies have demonstrated prolonged neutrophil survival, consequent to inhibition of apoptosis in neutrophils harvested from patients with thermal injury [47] , following major surgical trauma, or in patients meeting clinical criteria for the systemic in_ammatory esponse syndrome (SIRS) [48] (Figure 3 ). Delayed apoptosis is associated with evidence of increased cellular activation, and can be induced by circulating factors in patient plasma, including the cytokine GM-CSF. In vitro studies show that inhibition of apoptosis following exposure to either LPS or GM-CSF is an active process, dependent on the synthesis and processing of IL-1b by the neutrophil, through the activity of the interleukin 1b converting enzyme [45] . Thus delayed neutrophil apoptosis is both a marker of cellular activation, and a re_ection of a biologic effect of the cytokine, IL-1b, acting in an autocrine or paracrine manner. The persistence of activated neutrophils in the circulation and at in_ammatory foci may contribute to microvascular and parenchymal cell injury.

Neutrophil chemiluminescence measured in whole blood can be used to detect bacterial lipopolysaccharide, and hence to identify infection or translocation from the gastrointestinal tract. By examining unstimulated (native) and stimulated (PMA or zymosan) neutrophil oxidant burst activity and opsonin receptor reserve it is possible to gauge the extent of neutrophil activation by proin_ammatory mediators such as C5a, PAF, TNF, IL-8 and LTB4, a potential re_ection of the pro-in_ammatory or cytokinemic phase of immune cell activation. By measuring responsiveness as an index of immune effector cell competence it may be possible to determine whether the patient is progressing to an anergic phenotype. Whole blood neutrophil dependent chemiluminescence may therefore serve as a rapid and informative tool for the staging of immune effector cell status and may be useful as an intermediate biological marker to monitor the ef~cacy of therapeutic intervention. Conversely monitoring the kinetics of neutrophil apoptosis provides a marker of the persistence or resolution of an activated in_ammatory response. Fig. 3 . Critically ill patients meeting clinical criteria for SIRS show signi~cant inhibition of constitutive neutrophil apoptosis, a state that is persistent, and associated with evidence of enhanced activation, re_ected in the generation of reactive oxygen intermediates. From [48] ; used with permission.

Neutrophil priming and activation in the pathogenesis of post injury multiple organ failure

Update on the mechanisms of immune suppression of injury and immune modulation

Sepsis: A new hypothesis for pathogenesis of the disease process

Multiple organ failure. Generalized autodestructive in-_ammation?

The non-speci~c nature of endotoxin tolerance

Increased f Met-Leu-Phe receptor expression and altered superoxide production of neutrophil granulocytes in septic and post traumatic patients

Role of the neutrophil in adult respiratory distress syndrome

Neutrophil-mediated vascular injury in shock and multiple organ failure

Polymorphonuclear leukocyte dysregulation in patients with gram-negative septicemia assessed by _ow cytometry

Longitudinal analysis of neutrophil superoxide anion generation in patients with septic shock

Impaired neutrophil adherence as an early marker of systemic in_ammatory response syndrome and severe sepsis

Apoptosis in the resolution of systemic in_ammation

Assessment of the safety and ef~cacy of the monoclonal anti-tumor necrosis factor antibody-fragment, MAK 195F, in patients with sepsis and septic shock: A multicenter, randomized, placebo-controlled, dose-ranging study

A new marker of the systemic in_ammatory response syndrome: Procalcitonin (PCT)

In_ammatory mediators in relation to the development of multiple organ failure in patients after severe blunt trauma

Serum cytokine levels in human septic shock: Relation to multiple-system organ failure and mortality

ACCP-SCCM Consensus Conference: De~nitions of sepsis and organ failure and guidelines for the use of innovative therapies in sepsis

Management of fever in patients with cancer and treatment induced neutropenia

Microbial infection and the septic response in critical surgical illness: Sepsis, not infection, determines outcome

Prognostic signi~cance of marked leukocytosis in hospitalized patients

The prevalence and signi~cance of leukocytosis in upper gastrointestinal bleeding

Blood transfusion as a cause of leukocytosis in critically ill patients

The diagnostic value of the neutrophil left shift in predicting in_ammatory and infectious disease

Phagocytic leukocyte oxygenation activities and chemiluminescence: A kinetic approach to analysis

Evidence for generation of an electronic excitation state(s) in human polymorphonuclear leukocytes and its participation in bactericidal activity

Pholasin: A new chemiluminescent probe for the detection of chloramines derived from human phagocytes

Analysis of the sugar speci~city and molecular location of the b glucan binding lectin site of complement receptor type 3

Density and avidity changes of CD11b on circulating polymorphonuclear leukocytes (PMN) in systemic in_ammatory response syndrome (SIRS)

Analysis of circulating phagocyte activity measured by whole blood luminescence: Correlation with clinical status

Rupture of an abdominal aortic aneurysm causes priming of phagocytic oxidative burst

Observations using antiendotoxin antibody (E5) as adjunctive therapy in humans with suspected, serious, gram-negative sepsis

A controlled clinical trial of E5 murine monoclonal IgM antibody to endotoxin in the treatment of Gram negative sepsis

A second large controlled clinical study of E5, a monoclonal antibody to endotoxin: Results of a prospective, multicenter, randomized, controlled trial

Treatment of septic shock with human monoclonal antibody HA-1A

Treatment of gramnegative bacteremia and septic shock with HA-1A human monoclonal antibody against endotoxin

Effectiveness of a human monoclonal anti-endotoxin antibody (HA-1A) in gram-negative sepsis: Relationship to endotoxin and cytokine levels

Executive Summary of an American College of Chest Physicians, National Institute of Allergy and infectious Disease, and National Heart, Lung, and Blood Institute Workshop. From the bench to the bedside: The future of sepsis research

A rapid assay of endotoxin in whole blood using autologous neutrophil dependent chemiluminescence

Resolution of acute in_ammation and the role of apoptosis in the tissue fate of granulocytes

Protease activation during apoptosis: Death by a thousand cuts

Macrophage phagocytosis of aging neutrophils in in_amma-tion

Differential expression of Fas (CD95) and Fas ligand on normal human phagocytes: Implications for the regulation of apoptosis in neutrophils

Neutrophils undergo apoptosis following ingestion of Escherichia coli

Inhibition of apoptosis and prolongation of neutrophil functional longevity by in_amma-tory mediators

The interleukin-1 beta converting enzyme (caspase-1) inhibits neutrophil apoptosis through processing of interleukin-1

Marshall JC. Neutrophil apoptosis is modulated by endothelial transmigration and adhesion molecule engagement

Inhibition of apoptosis in polymorphonuclear neutrophils from burn patients

Dysregulated expression of neutrophil apoptosis in the systemic in_amma-tory response syndrome (SIRS)