sepsis and organ dysfunction: ...from chaos to rationale

Sepsis and Organ Dysfunction ... from Chaos to Rationale ...

Springer Milano

Berlin

Heidelberg

New York

Barcelona

Hong Kong

London

Paris Singapore

Tokyo

A.E. Baue G. Berlot A. Gullo J.-L. Vincent (Eds)

Sepsis and Organ Dysfunction ... from Chaos to Rationale ...

ORGAN FAILURE ACADEMY

i Springer

A.E. BAVE, M.D. Emeritus Professor, Department of Surgery, S. Louis University, School of Medicine -Fishers Island, New York - USA G. BERLOT, M.D., Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic, Trieste University Medical School, Trieste - Italy A. GULLO, M.D. Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic, Trieste University Medical School, Trieste - Italy J.-L. VINCENT, M.D. Department of Intensive Care, Erasme University Hospital, Free University of Brussels -Belgium

O.EA. - ORGAN FAILURE ACADEMY, VIA BATTISTI, I - 34125 TRIESTE (ITALY)

Steering Committee A.E. Baue, M.D. Emeritus Professor, Department of Surgery, S. Louis University, School of Medicine - Fishers Island, New York - USA

G. Berlot, M.D., Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic, Trieste University Medical School, Trieste - Italy

A. Gullo, M.D., Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic, Trieste University Medical School, Trieste - Italy

L. Silvestri, M.D., Department of Anaesthesia and Intensive Care, Gorizia Hospital, GoriziaItaly

G. Sganga, M.D., Department of Surgery, and C.N.R. Shock Centre, Catholic University of Sacro Cuore, Rome - Italy

© Springer-Verlag Italia, Milano 2002

Springer-Verlag Italia A member of BertelsmannSpringer Science+Business Media GmbH http://www.springer.de ISBN-13: 978-88-470-0178-7 e-ISBN-13: 978-88-470-2213-3 DOI: 10.1007/978-88-470-2213-3

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitations, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplications of this publication or PaJts thereof is permitted only under the provisions of the Italian Copyright Law in its current version, and permission for use must always be obtained from Springer-Verlag. Violations aJ'e liable for prosecution under the Italian Copyright Law.

The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.

Product liability: The publishers cannot gUaJ'antee the accuracy of any information about dosage and application contained in this book. In every individual case the user must check such information by consulting the relevant literature.

Cover design: Simona Colombo, Milan, Italy

SPIN 10859362

Table of Contents

PREFACE G. BERLOT •..............................•.....................•...•.•.. XIn

INTRODUCTORY REMARKS

Sepsis and Organ Dysfunction - Basics, Controversies, Rationale A.E. BAUE .•.....•...................................................... 19

SEPSIS AND ORGAN DYSFUNCTION: FROM CHAOS TO RATIONALE

Alveolar Epithelium in Host Defence: Cytokine Production M.Lw ..•...................•.........•.....•.....•....•.....••....•... 37

Phagocytosis and Lung Injury J.W. BOOTH........ .......... .... ....... .. ..........•.... .•....•.....•... 51

Dual Role of Neutrophil a-Defensins in Lung Inflammation H. ZHANG............................................................... 59

Epithelial Injury in Sepsis and ARDS: Role of Leukocyte-Derived Proteases H. GINZBERG,C.-W. CHow,ANDG.P. DOWNEy................................... 67

Pro- and Anti-Inflammatory Cytokines and Apoptosis in Acute Lung Injury S. UHLIG, AND D. BURDON ........................•.........•.•....•.....• " 77

The Role of Interieukin-lO During Systemic Inflammation and Bacterial Infection EN. LAUW, SJ.H. VAN DEVENTER, ANDT. VANDER POLL.......... .............. ... 95

Cardiovascular Surgery: Modulation of the Inflammatory Reaction P.P. GlOMARELLI, S. SCOLLETTA, AND E. BORRELLI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 103

Microcirculation in Critical Illness D. DE BACKER, M.-J. DUBOIS, AND J. CRETEUR .............•................•... 111

Microbial Translocation: From Myth to Mechanism J.C. MARSHALL. • . . . . . • . . . . . . . . . . . . . . . . . . . • . . . . . . . . . . • . . . . . . . . . . • . . . . . • . .. 121

Empirical Antibiotic Treatment in ICU Patients P. GROSSI •.....•....•...................•.....•...............•......... 131

Rotation of Antibiotics - A New Strategy for Prescription in the Intensive Care Unit D. GRUSON,ANDG. HILBERT ...........................•................•.... 141

Diagnostic Approach to Sepsis - State of the Art EM. BRUNKHORST, AND K. REINHART. . . . • . . . . . . . . . • . • . . . • . . . . • • . . . . . . . . . . • . . .. 151

Septic Shock Therapy R. FuMAGALLI, D. CODAZZI, AND S. CATTANEO. • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 169

Sepsis Trials in Children G. ZOBEL .......•....•....•.....•..............................•........ 177

Sepsis and Clinical Trials: A New Era in Anti-Sepsis Therapies J.-L. VINCENT. . . . . . . . . . . . . . . . • . . . . . . . . . . • . . . • . . . . . . . . . . . . . . . . . . . . . . . . . . .. 189

Evolving Concept and Challenges in Sepsis and MODS A. GULLO .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . . . . . . . . . . . . . .. 197

Index .................................................................. 219

Authors Index

BaueA.E. Emeritus Professor, Department of Surgery, St. Louis University School of Medicine, Fishers Island, New York (U.S.A.)

Berlot G. Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic, Trieste University Medical School, Trieste - Italy

BoothJ.W. Programme in Cell Biology, Hospital for Sick Children, Toronto, Ontario (Canada)

Borrelli E. Department of Thoracic and Cardiovascular Surgery, Siena University, Siena (Italy)

Brunkhorst F.M. Department of Anaesthesiology and Intensive Care Medicine, Friedrich-Schiller-University, Jena (Germany)

BurdonD. Division of Pulmonary Pharmacology, Research Center Borstel, Borstel (Germany)

Cattaneo S. Department of Anaesthesia and Intensive Care, Ospedali Riuniti, Bergamo (Italy)

ChowC.-W. Division of Respirology, Toronto General Hospital Research Institute of the University Health Network, and Department of Medicine, University of Toronto, Toronto, Ontario (Canada)

CodazziD. Department of Anaesthesia and Intensive Care, Ospedali Riuniti, Bergamo (Italy)

Creteur J. Department of Intensive Care, Erasme University Hospital, Free University of Brussels (Belgium)

De BackerD. Department of Intensive Care, Erasme University Hospital, Free University of Brussels (Belgium)

DowneyG.P. Division of Respirology, Toronto General Hospital Research Institute of the University Health Network, Department of Medicine and Clinical Sciences Division, University of Toronto, Toronto, Ontario (Canada)

Dubois M.-J. Department of Intensive Care, Erasme University Hospital, Free University of Brussels (Belgium)

Fumagalli R. Department of Anaesthesia and Intensive Care, Ospedali Riuniti, Bergamo (Italy)

Giomarelli P.P. Department of Thoracic and Cardiovascular Surgery, Siena University, Siena (Italy)

VIII Authors Index

GinzbergH. Division of Respirology, Toronto General Hospital Research Institute of the University Health Network, and Department of Medicine, University of Toronto, Toronto, Ontario (Canada)

Grossi P. Clinic for Infectious and Tropical Diseases, Insubria University, General Hospital and Macchi Foundation, Varese (Italy)

GrusonD. Intensive Care Department of Bordeaux, Pellegrin Hospital, Bordeaux (France)

GulloA. Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic, Trieste University School of Medicine, Trieste (Italy)

HilbertG. Intensive Care Department of Bordeaux, Pellegrin Hospital, Bordeaux (France)

LauwF.N. Laboratory of Experimental Internal Medicine and Department of Infectious Diseases, Academic Medical Center, Amsterdam (The Netherlands)

LiuM. Division of Cellular and Molecular Biology, University Health Network Toronto General Research Institute, and Departments of Surgery, Medicine, Paediatrics and Physiology, Faculty of Medicine, University of Toronto, Toronto, Ontario (Canada)

Marshall J.C. Department of Surgery, Toronto General Hospital, and Interdepartmental Division of Critical Care Medicine, University of Toronto, Toronto, Ontario (Canada)

ReinhartK. Department of Anaesthesiology and Intensive Care Medicine, Friedrich-Schiller-University, lena (Germany)

Scolletta S. Department of Thoracic and Cardiovascular Surgery, Siena University, Siena (Italy)

Uhlig S. Division of Pulmonary Pharmacology, Research Center Borstel, Borstel (Germany)

van der Poll T. Laboratory of Experimental Internal Medicine and Department of Infectious Diseases, Academic Medical Center, Amsterdam (The Netherlands)

van Deventer S.J.H. Laboratory of Experimental Internal Medicine and Department of Gastroenterology, Academic Medical Center, Amsterdam (The Netherlands)

Vincent J.-L. Department of Intensive Care, Erasme University Hospital, Free University of Brussels, Brussels (Belgium)

ZhangH. Departments of Anaesthesia and Critical Care Medicine, St. Michael's Hospital, University of Toronto, Toronto, Ontario (Canada)

Zobel G. Paediatric Intensive Care Unit, Department of Paediatrics, University of Graz, Graz (Austria)

Abbreviations

ACE, angiotensin-converting enzyme ALI, acute lung injury AP-l, activator protein-l APC, activated protein C ARDS, acute respiratory distress syndrome BAL, bronchoalveolar lavage BFA, brefeldin A BPIP, bacterial permeability increasing

proteins C3a, complement 3a COPD, chronic obstructive pulmonary di-

sease CPR, cardio-pulmonary bypass CRP, C-reactive protein c-TnI, cardiac troponin I CytoD, cytochalasin D DIC, disseminated intravascular coagula

tion D02I, oxygen delivery ECMO, extracorporeal membrane oxyge

nation EE, energy expenditure ELSO, Extracorporeal Life Support Orga-

nization ER, endoplasmic reticulum ER, extraction ratio ERK, extracellular signal-regulated kinase ESBL, extended-spectrum beta-Iactamase G-CSF, granulocyte-colony-stimulating

factor GTP, guanosine-triphosphate HA-IA, human IgM monoclonal antibody hBD-l, human ~-defensins-l HNP, human neutrophil peptides ICU, intensive care unit IFN, interferon-y IL-IO, interleukin-lO INH, isonicotinic acid hydrazide JNK, c-Jun N-terminal kinase LBP, lipopolysaccharide binding protein LPS, lipopolysaccharide MAPK, mitogen-activated protein kinases MCPs, monocyte chemoattractant proteins MDF, myocardial depressant factor

MIF, migration inhibitory factor MMPs, matrix metalloproteases MODS, multiple organ dysfunction syn

drome MOF, multiple organ failure MRSA, methicillin-resistant Staphylococ-

cus aureus NF-kB, nuclear factor kappa-B NK, natural killer NOERDF, nitric oxide-endothelial derived

relaxing factor OPS, orthogonal polarization spectral imag-

ing PAl-I, plasminogen activator inhibitor-l PAS, p-aminosalicylic acid PCT, procalcitonin PKC, protein kinase C PLA2, phospholipase A2 POPC, paediatric overall performance ca

tegory PROWESS, Protein C Worldwide Evalua

tion in Severe Sepsis rBPI, recombinant bactericidallpermeabili

ty increasing protein RSV, respiratory synthial virus r-tPA, recombinant tissue plasminogen ac-

tivator Sa02, oxygen haemoglobin saturation SAPK, stress activated protein kinase SDD, selective digestive decontamination SEB, Staphylococcal enterotoxin B SIRS, systemic inflammatory response syn-

drome STS, Society of Thoracic Surgery SVRI, systemic vascular resistance index Th2, T helper 2 TRAM, tris(hydroxymethyl)amino-me-

thane TLR, Toll-like receptor TNF ex, tumour necrosis factor-a TRAIL, tumour necrosis factor-related

apoptosis-inducing ligand VC02, C02 production VISA, vancomycin-intermediate Staphylo

coccus aureus

IPREFACEI

Preface XIII

Sepsis remains a major challenge for intensivists, as virtually all patients admitted to our Intensive Care Units (ICU) can sooner or later develop it. In this review, several leading authors present and discuss both the basic innovations in the field of sepsis and their clinical applications.

As stated by Baue in his introductory chapter, despite the important advances in the knowledge of the basic pathophysiologic mechanism of sepsis, the crude mortality remains high, being largely dependent on the causative factor(s), the appropriateness of care and the underlying conditions. In the last decade, the focus of attention has gradually shifted from the consequences of sepsis, including cardiorespiratory and metabolic derangements, to its causes, that is the biological determinants, and finally to the causes of the causes, that is the genetic milieu.

The lung is frequently, if not always, involved in sepsis, as septic patients present a number of respiratory disturbances ranging from mild hypoxia to full-blown ARDS. The underlying causes include both the sepsis-induced burst of inflammatory mediators and the biological differences existing among the different cells coexisting in the lung. The complex biological interactions induced by the release of inflammatory mediators in the lung is described by Liu, who underscores the difficulties encountered in separating - either in vitro or in vivo - the actions of different cellular lines and investigating the genetic mechanisms prompting these reactions. Also Booth describes the pathophysiologic mechanisms underlying the pulmonary phagocytic activation. The author underlines that, at a certain point of the inflammatory reaction which has been teleologically developed with the aim of containing an infection, a no return point is reached beyond which the inflammatory cascade involves the whole organism. Zhang describes the role played by the defensins, which constitute a family of substances exerting a powerful lung toxicity, as demonstrated by their effect on cultured pulmonary epithelial cells and the inhibitory effect on the neutrophil phagocytic capabilities. These actions must be further explored as (a) these substances have been proposed as a potential antimicrobial therapy and (b) it may be hypothesized that some strategies aimed to counteract their toxic action on the lung cells could be developed. The complex interaction between the alveolar epithelial lining and the activated leukocytes is described by Ginzberg et aI, who provide sound experimental evidence that the proteolytic enzymes produced and released by activated leukocytes favour the transmigration of the inflammatory cells through epithelial cells; it is worthwhile to remark that the same phenomenon can also occur in non pathologic conditions: in these settings however, the leukocytes are not activated and the alveolar wall is not damaged. The possible role played by apoptosis during ALI or ARDS is described by Uhlig et al. who underline, however, some confounding factors:

XIV Preface

this process is actually supposed to clear the inflammatory cells from the lungs and to hasten the resolution of type IT cell hyperplasia; at the same time, there are some experimental data showing apoptosis in acute lung injury in many cells implicated in the ALII ARDS process. However, caspase inhibitors, which are supposed to block the apoptotic process, fail to reduce edema formation. An unequivocal finding from clinical studies is the low rate of apoptosis in extravasated neutrophils, which could represent a key detrimental event in the pathogenesis of ALIIARDS. The complicated relationships between the sepsis-related mediators and the overall response to infection is exemplified by Interleukin 10 (IL-l 0). As demonstrated by Lauw et aI, this cytokine represents a kind of double-edged sword: in overwhelming systemic inflammation, IL-IO exerts a protective effect by inhibiting the release of proinflammatory cytokines and diminishing systemic inflammation and organ failure. On the contrary, during localised infections such as pneumonia, the same cytokines can blunt an appropriate host response to invading micro-organisms.

Moving from the laboratory to the clinical arena, Giomarelli et al. Emphasise how cardio-pulmonary bypass (CPB) represents a trigger of the inflammatory cascade not associated with infections, and that there is a relatively close connection between this priming event and the subsequent activation of the neuroendocrine stress response, which, is association with the effects of the pro-inflammatory mediators, heavily influences the coagulative balance, the immunologic capabilities and the ultimate outcome. This response can be blunted by the heparin coating of the CPB circuits as well as by the administration of steroids in the perioperative phase. Microvascular blood flow alterations are frequent in critically ill patients, and these alterations can be associated with relevant physiopathological and clinical implications. De Backer et al. describe a novel device, the Orthogonal Polarisation Spectral (OPS), that non invasively allows the direct visualisation of the microcirculation especially in regions covered with a thin epithelial layer, such as the sublingual area, the gut and the vagina. These authors demonstrated that in septic shock patients both the sublingual capillary density and the proportion of the perfused capillaries are decreased as compared with healthy subjects and that the severity of these alterations were more marked in nonsurvivors.

The role possibly played by bacterial translocation through the intestinal barrier in the pathogenesis of sepsis and the related consequences remains an extremely controversial issue. Although this process has been described in a number of diseases, its role in this setting is still unclear. Marshall reviews several experimental and clinical studies, concluding that "It would be naive to contend that a single insult can be held responsible for the entire clinical spectrum of a complex disorder, such as the mUltiple organ dysfunction syn-

Preface xv

'drome", but that the data coming from the gut-directed therapeutic or prophylactic strategies, including the Selective Decontamination of the Digestive Tract or early enteral feeding, underline its role as a pathogenetic factor. Another important issue is covered by Grossi who describes the difficulties encountered in deciding an empirical antibiotic treatment in critically ill patients. The occurrence of both gram- and + multi-resistant strains make any decision increasingly difficult. The decision can be guided by two different approaches. First, the knowledge of the more aggressive germs existing in the environment is warranted, in order to maintain the epidemiologic logbook upgraded; second, patients infected with the more aggressive germs must be kept isolated in order to avoid the spread of the infection. Another way to blunt the emergence of resistant strains is to cycle periodically the antibiotics prescribed. This approach has been developed recently, but, as stated by Gruson et aI., despite the sound basis of its rationale, no study has definitely demonstrated the superiority of one type of cycle or rotation over another. As stated by Brunkhorst et aI., an early diagnosis is paramount to reduce the mortality of sepsis. Since blood cultures are positive in only a minority of patients, other markers of inflammation and sepsis have been studied. The ideal diagnostic test I) should be sensitive and specific enough to discriminate between sepsis and other inflammatory states not related to infections. 2) should be inexpensive and readily available and, finally, 3) should be related to the outcome. The authors review the currently available tests, including the monitoring of several pro- and anti-inflammatory mediators, illustrating the advantages and limitations of each. The treatment of sepsis shock patients is reviewed by Fumagalli et aI., who underscore the difficulties associated with hemodynamic management. A thorough overview of both the experimental and current treatments of pediatric patients is provided by Zobel et aI., who describe the common points and differences existing between infants and elderly patients: although the initiating mechanisms are pretty much the same, the functional reserve in elderly patients can be so limited to make futile any attempt to reverse the ongoing inflammatory process. The results deriving from trials with substances aimed to contrast the effects of sepsis mediators is reviewed by Vincent: despite encouraging experimental data, the use of these agents failed to decrease the mortality of sepsis and septic shock patients, even if some limited beneficial effect has been demonstrated in some subgroups of patients. Several causes account for this finding, including the heterogeneity of the patients enrolled in the different clinical trials, the inappropriate timing of intervention, the inter-species differences etc. Only recently, the use of recombinant Activated Protein C (rAPC) has been demonstrated to increase significantly the survival of these patients, and the next logical step consists in the early identification of those patients

XVI Preface

who can draw the maximal advantage from this novel approach. The final remarks are drawn by Gullo, who thoroughly reviews the lights and shadows existing in the field of sepsis and sepsis-induced multiple organ dysfunction syndrome.

G. Berlot

IINTRODUCTORY REMARKSI

Sepsis and Organ Dysfunction - Basics, Controversies, Rationale

A.E. BAUE

Sepsis is not a disease or even a syndrome. For most of us, it indicates an underlying infection that has caused an inflammatory response with systemic manifestations. For some, sepsis indicates a systemic inflammatory response no matter what its cause. Both an infection with an inflammatory response that is no longer localized, as with an abscess or a severe inflammatory problem per se, can produce remote organ dysfunction. This led to the concept of the multiple organ dysfunction syndrome (MODS) and to mUltiple organ failure (MOF) [1-5]. Some still believe that the inflammatory response should be called a systemic inflammatory response syndrome (SIRS). The usefulness of this expression has been questioned.

Some of the causes, intricacies, and inter-relationships of these changes are known in part but not totally understood. They are being studied extensively because only by knowing the mechanisms of damage can we prevent them and support organ function[6, 7]. There are three fundamental changes at work with infection and inflammation. A major one is the circulation - a number of abnormalities of not only cardiac output, blood pressure, peripheral vascular resistance, but also microcirculatory blood flow, distribution of organ flow, and the pathophysiological changes of altered flow on the organs - the kidneys, liver, gut, heart, musculoskeletal system, endocrine, and lymphatic systems. The second major factor is the effect of inflammation through its complex system of mediators on both cell and organ function. The third factor is the complex inter-relationship between mediators and the circulation, how mediators affect endothelium and the microcirculation, and how the circulation influences the production of mediators.

There are many classifications, scores or predictive scales for MODS and MOF, but these do not help us understand the pathophysiology of the process [8-9].

20 AE. Baue

Remote organ injury

Another way to study remote organ injury with infection is by considering (1) functional changes, (2) mediator changes, and (3) system and cell activation changes, with activation of coagulation, priming and activation of neutrophils, and endothelial microcirculatory changes. Factors involved in producing remote organ injury after injury, operation, and sepsis include the following [10-13]. 1. Ischemia - ischemia-reperfusion injury - no flow versus very low flow. In

a study of leukocyte adherence and sequestration comparing shock and total ischemia in rats, Childs et al. [14] found that the microvascular response following hemorrhagic shock is different from total ischemia. Hence it is not possible to extrapolate from one to the other [1].

2. Hypoxia versus anoxia versus dysoxia - Haldane once said that, "anoxia not only stops the machine but wrecks the machinery".

3. Inflammation - may result from injury - a direct injury and the extent, as determined by the injury severity score, leading to infection per se, nonspecific inflammation, and the various inflammatory diseases.

4. Prior organ damage in patients contributes to the problem. For example, with the lungs this could be chronic obstructive pulmonary disease, liver cirrhosis, renal insufficiency, cardiac insufficiency, or cerebrovascular disease, all of which may take their toll.

5. There is the domino effect of organ failure. Gut ischemia produces factors that damage the lung. Liver ischemia does the same. The hepatorenal syndrome is a combination of hepatic failure leading to renal failure and of course white blood cell activation.

6. Other causes of organ dysfunction/failure, as indicated earlier, are ischemia or dysoxia, ischemialreperfusion injury, white blood cell activation with oxidants, and failure of antioxidants such as elastase, the one-hit, two-hit phenomenon of injury leading to infection, overwhelming inflammation called by Erhlich - the horror - autotoxicus. Other causes of organ dysfunction and failure include endotoxin, bacterial translocation from the gut, cytokine activation, inadequate resuscitation, the hyperdynamic, hypermetabolic high oxygen consumption phenomenon of sepsis.

Another contribution of infection to remote organ damage is that an infection requires a hyperdynamic circulation, but in spite of this there is frequently increased venous O2• This means decreased O2 extraction to cells in the periphery, which could produce cell damage. There is decreased microcirculatory blood flow with septic shock, with a low cardiac output and mediator activation. Other factors are white blood cell activation, capillary damage from endothelial changes and coagulation abnormalities with microcirculatory thromboses.

Sepsis and Organ Dysfunction - Basics, Controversies, Rationale 21

There are a number of clinical entities that produce remote organ damage and may not be evident initially. These are inflammatory or infectious problems, which include acute acalculous cholecystitis, bowel ischemia, abscesses in the pelvis, subphrenic or subhepatic region, intraloop abscesses or in the retroperitoneum, and bowel perforation. The setting is frequently fever with an increased white blood count. Remote organ dysfunction in a patient should suggest the possibility of re-operation in the abdomen if there had been a previous abdominal operation. This is no longer necessary as a blind procedure. Ultrasound and computed tomography scans should help identify the problem before multiple organ failure occurs. The evidence and evolution of such a problem should be carefully considered [6].

Mediators produced by infection and promoting inflammation

There are a large number of mediators, factors produced by inflammation, by cells, and elsewhere that promote the inflammatory process and contribute to remote organ and cell damage. These include the cytokines, interleukin (lL-l) and IL-2, tumor necrosis factor (TNF) through to IL-18. There are more being discovered every day. Platelet-activating factor (PAF) is an important mediator. Complement activation occurs and may produce damage. The various kinins contribute, as do the endorphins and histamine. Nitric oxide-endothelial derived relaxing factor (NOERDF) and myocardial depressant factor (MDF), are activated. The cyclo - and lipoxygenase metabolites contribute, as do adhesion molecules, and toxic oxygen radicals in cells and from activated neutrophils that also produce elastase. Free iron contributes, as do the leukotrienes. Nitric oxide is a vasodilator, a bronchodilator, a neurotransmitter, an anticoagulant, antiproliferative agent, an antimicrobial substance, and an endotoxin mediator. It is both good and bad and necessary.

There is the superfamily phenomenon ofIL-l to IL-15, the chemokine superfamily of IL-8, MIP-l, and the IL-6 super family of CNTF. LIP is a factor, as are the enzymes phospholipase Al and A2 and the STAT family of proteins. The cell adhesion molecules / receptors are part of this process. They include the integren superfamily - LFA-l, fibronectin, and platelet glycoprotein, immunoglobulins A and M, cadherens, LEC-CAM-ELAM-l, and lymphocyte homing receptors. Some new and important factors include endothelial factors, such as ELAM and others. IL-8 is a leukocyte adhesion inhibitor. The heat shock proteins are involved when there is necrotic tissue. PAP is criticaL Human granulocyte colony-stimulating factor may be involved, along with epidermal growth factor and natural killer cells.

22 A.E. Baue

The chemokines IL-8, RANTES, CC, and CXC are new mediators. There are new engineered cytokines such as SYNTOKINE - a form of myelopoietin. IL-15 requires IL-15 Ra, I1-I1bb, 2R for its action. Platelets from megakaryocytes are increased by megakaryocyte growth and development factor (MGDF), thrombopoietin, and promegapoietins.

Other new mediators include cytokine-regulating agents HP228, ICE-IL-l~ converting enzymes, and IL-12 therapy, which may be important in oncology and virology. They produce enhanced lytic activity of natural killer (NK) cells, stimulate proliferation of TNK cells, induce secretion of interferon gamma, and promote a Th-l response.

There has been a recent proposal and studies that mesenteric lymph after injury is cytotoxic. Deitch et al. [15] studied this after hemorrhagic shock in animals and found that the lymph was cytotoxic to endothelial cells and it activated neutrophils. Mesenteric lymph was called by Moore "the critical bridge between dysfunctional gut and mUltiple organ failure" [16]. However, Lemairre et al. [17] found that thoracic duct lymph in patients with MOF had no increase in endotoxin, was low in pro-inflammatory cytokines, and had high anti-inflammatory cytokine levels. This raises a question about the clinical applicability of the mesenteric lymph idea.

Translocation of bacteria from the gastrointestinal tract to lymph nodes and to the systemic circulation has been found to occur in animals under adverse circumstances. How often this occurs in people and how important it is clinically has not been established.

There is also the anti-inflammatory network. This response begins shortly after the injury or the infection. It is marked by plasma IL-l RA (receptor antagonist), which increases with trauma and burns, and particularly in critically ill patients. STNFr I(p55) I1(p75) (soluble TNF receptor) increases in experimental endotoxemia in humans and increases in septic patients, but it did not reduce in vitro TNF toxicity. Bacterial permeability increasing proteins (BPIP) increase in volunteers after endotoxin injection and with gram-negative infection but not as much as LBP (lipopolysacharide binding protein). TNF binding protein is also involved, as is sIL-6R and cardizole, CRF, aMSH, IL-4, IL-lO, and IL-13. These anti-inflammatory factors are protective and decrease the likelihood of severe inflammation, but in human infection it may be a matter oftoo little too late [18-20].

Cell and organ injury

In a recent supplement to Critical Care Medicine, Dhainaut et al. [21] hosted


a conference on sepsis: the interface between inflammation, coagulation, and the endothelium. Through a series of presentations and discussions, they review these important factors, which all contribute to remote organ injury with infection. Inflammation, activation of coagulation, and endothelial cell damage are major factors in remote organ injury. Beutler and Poltorak [22] review the innate immune system and the toll-like receptors, of which there are now 10. These receptors both protect against and contribute to septic shock, thus there are virtues and liabilities of this system. Intracellularly the toll-like receptors interact with IL-l and IL-18 receptors, which interact with TNF receptors, which interact with the mitogen system, which is activated through protein kinaselERK kinase-kinase-l through IkB phosphorylation and dissociation from NFkB. Also, MAP kinase, SAP kinase, PI3 kinase, and P38 pathways are activated with MyD88 recruitment. The macrophage migration inhibitory factor (MIF) is part of the innate immune system responding to endotoxin, both gram-negative and gram-positive bacteria and toxic shock factors. They describe a number of mechanisms that contribute to endothelial damage and infection. Neutrophils participate in this by producing elastase and oxygen free radicals. Cytokine-activated NKT cells and cytotoxic T-Iymphocytes injure the endothelium [23, 24]. IL-2 produces trouble as does ischemia / reperfusion injury. Cytokines, complement activation, neutrophils, and adhesion molecules all contribute to ischemia/reperfusion injury.

Endothelial cell dysfunction and coagulation activation are reviewed by Vallett and Wiel [25]. They point out that injured endothelial cells change from anticoagulant properties to procoagulant behavior. Tissue factor synthesis occurs and monocyte activation participates in this. With infection, coagulation is initiated by the extrinsic pathway and then is amplified through the intrinsic pathway [26]. The body's natural anticoagulant mechanisms are significantly decreased.

There are extensive reviews of recombinant human activated protein C and its modulation of vascular function and coagulation activity with severe infection [27, 28]. The recent recombinant human activated protein C (APC) worldwide evaluation in severe sepsis (PROWESS) trial is reviewed in detail [29]. It showed reduced mortality in patients with sepsis treated with recombinant human APC. The question is whether or not the reduction occurs enough to justify giving it to everyone.

Marshall [30] reviews the question of whether infection has direct cytotoxic effects on cells and organs through the effects of bacteria or whether it is the response of the host to infection that produces the major damage. Perhaps both of these are involved. Marshall states, "the elucidation of a complex network of host-derived inflammatory mediators raise the possibility that targeting these individually could improve patient outcomes" [30].

24 A.E. Baue

Finally, Angus and Wax [31] review the epidemiology of sepsis. They review in detail the genetic polymorphism that occurs in patients with sepsis. These genetic differences provide evidence that not all patients respond to a septic challenge in the same way, and that there are many, many differences in the response. These genetic polymorphisms open up an entire field of study, indicating great differences in patient responses and probably also the responses to therapy. Angus and Wax [31] conclude that the composite picture rich in many aspects remains incomplete and emphasizes a heterogeneity of the condition. Genetic predisposition remains to be elucidated, but can not be determined in individual patients when they come to the emergency ward. This may seem like a Star Wars request at the present time.

Why no magic bullets?

Immunotherapy in sepsis is the Bermuda Triangle of the Biotech Industry. Phillip Dellinger, M.D.

Lecture to the Shock Society June 1997

It is not unusual to develop therapeutic agents that seem protective or therapeutic in an experimental animal but are never proven to be worthwhile in patients. Over the years we have studied many such promising agents that never made a difference clinically. These include low molecular weight dextran, which is an anti-sludging agent, Dibenzyline (phenoxybenzamine), an alpha-adrenergic blocking agent to decrease the intense vasoconstriction of shock, 2-3 diphosphoglycerate, which helps red cells unload oxygen in the peripheral circulation, polarizing solutions with homeopathic doses of magnesium, potassium, insulin, glucose, and steroids, white blood, which led to excess Ringer's lactate solution being given, buffers for extra-cellular acidosis, TRAM TRIS(hydroxymethyl)amino-methane, an intracellular buffering agent, excess lactate, and the LIP ratio, which would indicate anaerobiosis and steroids for septic shock. In spite of many positive effects in the experimental laboratory, none of these substances ever came into clinical use for very long. When they were subjected to randomized clinical trials, they failed to improve survival or help patients. Recently, studies of injury, infection, and inflammation have shown many mediators or agents that contribute to illness from such insults. This led to the development of agents that could block the harmful effects of such mediators. These were called magic bullets [32-34]. Many of these agents demonstrated excellent results in experimental animals and suggested great promise for


'clinical use. However, so far none of these agents, when used individually, has had a positive effect in decreasing mortality in prospective randomized placebo-controlled trials in sick patients (Table 1).

Table 1. Clinical trials of agents to control inflammation and infection manifestations (mAbs = monoclonal antibodies, IL-l = interleukin-l, TNF = tumor necrosis factor, PAF = platelet-activating factor, ARDS = acute respiratory distress syndrome, G-CSF = granulocyte colony-stimulating factor, Hb = hemoglobin, NOS = nitric ixide synthase)

No improvement in 28-day mortality with: 2-HA-lA mAbs to endotoxin 2-E-5 mAbs to endotoxin Taurolidine IL-lra Hydrocortisone in hyperdynamic septic shock Pentoxifylline Polyglobin Ibuprofen Lisofylline E-selectin mAb 3-Anti-TNF mAbs STNFr, P-55 and P-75 2 PAF antagonists Bradykinin antagonist (Bradycor) Antithrombin III concentrate Ketoconazole for ARDS G-CSF for pneumonia HMG-mAb Tirilazad mesylate Ng-monoethyl L-arginine Diaspirin cross-linked Hgb N-acetylcysteine NOS inhibitor 45GL88 Aerosolized surfactant Superoxide dismutase

There are a number of reasons why there are no magic bullets as yet and why the trials have been negative in spite of the great molecular biology and technology. The reasons include [32]: 1. The problem of causality of disease - many different complex diseases

(sepsis) are being treated with a single agent. There is also a great variability of severity in similar diseases.

2. There is a great redundancy and overlap of mediators with cross-stimula-tion.

3. No one single factor is the lethal factor or activator. 4. The timing of treatment is a problem. 5. There is the problem of trying to block or blunt an essential biological

function of inflammation. Inflammation is necessary with injury and infection in order to survive and heal.

26 A.E. Baue

6. There is also an anti-inflammatory response to try to control the process before it gets out of hand. The timing and variability of these processes are inconsistent. Some have argued that the end-point should be some improvement in organ function or a better MODS score rather than 28-day mortability. Buchman said "what good is it if a patient improves somewhat but dies anyway?"[35]. Others have argued that better studies of patients before the trial begins would help, such as mediator profile, IL-6 levels, a rapid endotoxin assay, a rapid culture technique. The heterogeneity of mediator activities predicts that this will not help. Where IL-6 levels were used for entry into the study, it did not help.

7. There is also the question of whether there is immune deficiency or immune excess, or both, in different parts of the system.

Thus, there is a great discrepancy between the burgeoning knowledge of molecular biology and the more-limited capability of what we can do for our patients. Our science is powerful, but what we can do for our patients is limited in good part to support of organ function. This has led to consideration of multiple therapeutic agents for patients with infections, disease, or injuries, which stimulate an inflammatory response.

Therapy - multiple therapeutic agents

There are many human diseases in which multiple agents are required for appropriate therapy. These include antituberculous therapy for tuberculosis, immunosuppression for transplanted organs, ionotropes and diuretics for heart failure, multiple antibiotics for polymicrobial peritonitis, cancer chemotherapy, and support of the gastrointestinal tract. Review of several of these diseases illustrates the difficulties and evolution that occurred in therapy with multiple agents.

Development of chemotherapy for tuberculosis serves as an example of the problems even when dealing with a specific disease process and one organism, which may be typical, atypical, or may develop resistance to antibiotics. In 1944 Waksman et al. isolated streptomycin. It was found to be effective against tuberculosis in a small trial in 1945, followed by a large national trial in 1947, demonstrating impressive clinical results. It was immediately apparent that there was a high incidence of relapse and development of resistant organisms. To counteract this, P-aminosalicylic acid (PAS), a drug with mild tuberculostatic activity, was used with streptomycin in a trial in 1948-1949. PAS extended the time during which streptomycin could be administered without developing resistance.


In 1950 a specific program by industry led to the development of an antituberculous agent called isonicotinic acid hydrazide (lNH) or isoniazid. This was found to be very effective in vitro and was strikingly successful in patients in 1952. Other drugs were developed. Presently recommended basic treatment for previously untreated patients with pulmonary tuberculosis includes isoniazid, rifampin, and pyrazinamide given daily for 2 months, followed by 4 months of isoniazid and rifampin. Ethambutol can be added in the initial 2 months if there is any suspicion of resistance, or if the patient is thought to be HIV positive. There has been a steady and continuing evolution of appropriate multiagent chemotherapy for tuberculosis. However, tuberculosis is a single disease with variations in the organism (typical, atypical, and resistant, etc.), which primarily involve the lungs initially. Much of the development of the successful treatment of tuberculosis was done by in vitro studies of the organism in culture and then trial and error clinically. Also, each of the agents used now in combination was effective for some time when used singly.

Cancer therapy is another example of the complexities and difficulty in treating manifestations and causes of human disease. Cancer chemotherapy was initially modeled on the multi agent treatment of tuberculosis. Paul Ehrlich is said to have coined the word "chemotherapy" at the turn of the century. He used rodent models of infectious diseases to develop antibiotics. The era of effective combination chemotherapy began when an array of active drugs from different classes became available for use in combination in the treatment of leukemias and lymphomas [37, 38]. For multi agent cancer chemotherapy, only drugs known to be partially effective against the same tumor when used alone should be selected for use in combination. The least-toxic drug should be used and given in an optimal dose and schedule. The principle of cancer chemotherapy has been clinical trial designed and dominated by the use of alternating cycles of combination chemotherapy. The response to chemotherapy is affected by the biology of tumor growth. All cancers are different. They respond to very different agents. What is effective for one malignancy may do nothing for another. Malignancy is not a common denominator for therapy. Some tumors are hormone dependent, some respond to radiation therapy, some respond to chemotherapy and various combinations, some respond to both, some respond to operation with or without adjuvants. Staging and grading also have a lot to do with this. It is apparent now that cure of malignancy is unusual, and the malignant setting in patients is very important in terms of oncogene influence, genetic mutations, and other factors.

Lessons learned from the treatment of tuberculosis and cancer indicate that specific diseases must be treated by a combination of agents, each of which has been shown to be individually effective in some way, shape, or form. These

28 A.E. Baue

processes of infection and neoplasia are chronic and not immediate, acute, and life threatening problems. Treatment can be carried out over many weeks. Thus, there are many dissimilarities between the use of mUltiple chemotherapy for these diseases and the possibility of using agents for the control of acute inflammation and of SIRS, MODS, and MOE

Therapy for excess inflammation

Therapy for excess inflammation could require control or replenishment of a number of agents [39,40]. Knox et al. [41] used a combined chemotherapeutic regime in bum patients. They gave antioxidants, which included vitamins C, E and glutamine, with an endotoxin binder (parenteral polymixin B), a cyclo- and lipoxygenase inhibitor, ibuprofen, and reconstituted human growth hormone. They believe that this improves mortality but it is based on historical controls.

Kirton et al. [42] used a multi agent approach for patients after trauma. Kilbourn et al. [43] suggest that a combination of approaches that treat vasodilatation, multi organ damage, metabolic dysfunction, and coagulation abnormalities may be needed to treat septic shock.

Because of the many mediators, each of which seems to have a role in the pathogenesis of excessive inflammation, it makes scientific sense to use multiple agents. If we tried to put together an ideal combination of agents for excess inflammation, what would be the components? Certainly early on in the disease process some attempts to block pro-inflammatory mediators (Table 2) might be worthwhile. Soon thereafter, supplemental anti-inflammatory mediators would seem necessary, with control of the many enzyme cascades that are activated by shock, trauma, or infection (Table 2). How many of these are necessary, important, or possible is not known. How do we begin to formulate such an approach? What is the timing? What will be the cost? If the multiagent cocktail becomes beneficial, what ingredients are critical, some may be ineffective. What is the model on which to test such approaches? One is a sheep model. A baboon model could be helpful for multi agent testing [44]. Would that fit the bill? Perhaps so, or do we also need new "multiple models" to cope with a "two or multiple-hit" theory, as suggested. In any case, it will be difficult to prepare a sufficient multidimensional protocol for such a study. We are told that the Food and Drug Administration in the United States would probably not approve a multiagent approach. Perhaps a trial in Europe would help. In the meanwhile, we may learn more from multiple agents in animals.

Sepsis and Organ Dysfunction - Basics, Controversies, Rationale

Table 2. Potential control of factors activated by shock, trauma, or infection

Scavenging of inducers Endotoxin-rBPhl Pro-inflammatory mediator blockade IL-lra STNFr Anti-TNF mab to restore function Supplementation of anti-inflammatory agents IL- \0 to reduce inflammation IL-12, IL- \3 Anti IL- \0 mab to restore immune function RHDL Antioxidants Protease inhibitors Tissue factor pathway inhibitor Cascade control Coagulation - ATIll Complement inhibitor Cyc\o- and lipo-oxygenase inhibition - ibuprofen Histamine antagonist Bradykinin antagonist Control of other factors PAF antagonist Immunomodulators - drugs, diet Anti-adhesion agents

29

Some have suggested that signaling mechanisms such as nuclear factor kappa-B (NF-kappa-B) should be therapeutic targets. NF-kappa-B turns on a number of chemo- and cytokines such as MIP-l, TNF-a., IL-IB, IL-6, and IL-8 and coagulation factors and cell adhesion molecules. Levels of NF-kappa-B are higher in patients that do not survive. There are other important signal transduction molecules.

Therapy - prevention before injury occurs

The best approach to prevent remote organ damage by infection is to prevent specific abnormalities and support organ functions before they fail. For prevention it is necessary to be specific. What is the problem? What is the disease or process you are dealing with? Many treatments, agents, or drugs to support or prevent organ failure may help in some circumstances, but not enough to reduce mortality or help all patients. Our task is to determine in what specific problems and which patients an agent helps or can be combined with other agents to make a difference (multiagent therapy). There are treatments that will help in certain diseases but not in others. Inflammatory bowel disease and rheumatoid arthritis may be helped by monoclonal antibodies to some of the pro-inflammatory mediators. Gut decontamination improved morbidity and mortality in patients with acute

30 A.E. Baue

hemorrhagic pancreatitis, but was no help in general trauma patients [45, 46]. Entero- and immuno-nutrition was a great help in patients having operations

for malignancy, but was no help in patients just having major operations [47]. High-dose steroids in patients with end-stage acute respiratory distress syndrome (ARDS), as used by Meduri et al. [48], were helpful, whereas steroids for septic shock in general did not help. We learned many years ago in cardiac surgery that a patient having a long difficult heart operation did better postoperatively by continuing ventilatory support for 1-2 days to decrease the work of breathing and prevent potential long-term ventilatory support. The same may be true for patients with severe infection who are having ventilatory difficulty.

Prevention of thrombophlebitis, pulmonary embolism, and gastrointestinal stress bleeding and perforation are necessary. Prevention of surgical site infection by asepsis, a clean wound, adequate oxygenation, and warming are important. Prophylactic antibiotics for contaminated or dirty wounds are necessary. Reduction of the stress response by epidural anesthesia, fentanyl, and/or proprofol will preserve organ function. The contributors to organ failure must be diagnosed, such as the abdominal compartment syndrome and hypothermia, coagulopathy and acidosis during abdominal exploration of trauma patients [49]. The importance of a high cardiac output after operation or injury and with infection is critical.

Much has been written about support of organ or system function: the lungs, circulation, liver, kidneys, coagulation, the central nervous system, metabolism, and musculoskeletal system and neuroendocrine system. There is much we can do. I refer the reader to the many specific recommendations in the literature [50-57]. Thangathurai et al. [58] maintained intra-operative tissue perfusion by nitroglycerin and fluids in high-risk patients, and of 155 such patients none developed ARDS. Shoemaker et al. [59] used intra-operative evaluation of tissue perfusion in high-risk patients by invasive and non-invasive hemodynamic monitoring. Blood flow, oxygen delivery and tissue oxygenation of patients who did not survive became inadequate at the end of the operation. This suggests potential intraoperative therapy.

Conclusions

The factors involved in remote organ damage with infection, injury, and inflammation are many. They are complex and they are inter-related. I find these complexities, the multiplicities of factors, and the intricate inter-relationships difficult to understand. Can we ever put them into a system that we can deal with effectively? Now we also learn about genetic polymorphisms in which


every individual is different with respect to susceptibility to infection, response to infection, and the impact and outcome of infection. It is hard to imagine that we will ever be able to understand all the intricacies of these complex mechanisms. It seems that the more we learn the more complex the system and systems become. So far, blockader treatment of a single factor has not helped. Thus, there is no magic bullet as yet. Whether or not APe is truly of great therapeutic value remains to be established. Thus, in the meanwhile we must rely upon those things that we know how to do to help our patients - treat infections with appropriate antibiotics and drainage as needed, support of the circulation, promotion of organ blood flow where possible (kidney, liver, heart, etc.), support organ function if it seems to develop dysfunction, and general support of the patient, including nutrition, the gastrointestinal tract and its flora, ventilation, mobilization, cardiac function, cardiac output, and the best regional blood flow possible [60-64].

References 1. Baue AB (1992) The horror autotoxicus and multiple-organ failure. Arch Surg 127:1451 2. Baue AB. (1994) Multiple organ failure, multiple organ dysfunction syndrome and the systemic

inflammatory response syndrome - where do we stand? Shock 1:385-397 3. Baue AB, Durham R, FaistE (1998) Systemic inflammatory response syndrome (SIRS), multiple

organ dysfunction syndrome (MODS), multiple organ failure (MOF): are we winning the battle? Shock 10:79-80

4. Baue AB (1998) SIRS, MODS, MOF - are we winning the battle? Proceedings of the 7th

International Symposium on Intensive Care Med, Bled, Slovenia 7:59-62 5. Baue AE (2001) Multiple organ failure (MOF) in the next millennium: are we winning the battle?

Can we prevent MOF? Crit Care Shock (in press) 6. Baue AE (1998) The complexities of sepsis and organ dysfunction. In: Baue AB, Berlot G, Gullo

A (eds) Sepsis and organ dysfunction. Springer-Verlag, Berlin Heidelberg New York, pp 145-152 7. Baue AB (1998) Surgical patient issue. Editor: current opinions in crit care med. Rapid Sci Pub

4:229-268 8. Baue AB (1998) An overview to introduce prognostic indexes in MODS. In: Baue AB, Berlot G,

Gullo A (eds) Sepsis and organ dysfunction. Springer-Verlag, Berlin Heidelberg New York, pp 35-44 9. Baue AE, Berlot G, Gullo A (eds) (1998) Sepsis and organ dysfunction- epidemiology and

scoring systems. Pathophysiology and therapy. Springer-Verlag, Milan. 10. Baue AB (1998) Mediators or markers of injury, inflammation and infection, Biologic puzzles

and predictors of disaster. Proceedings of the 7th International Symposium on Intensive Care Medicine, Bled, Slovenia 7:100-103

11. Baue AE, Berlot G, Gullo A, Vincent JL (eds) (2001) Sepsis and organ dysfunction: bad and good news on prevention and management. Organ Failure Academy. Springer, Milan

12. Baue AB, Berlot G, Gullo A, Vincent JL (eds) (2000) Sepsis and organ function. Springer-Verlag, Milan

13. Baue AB (2000) Septic shock, multiple organ dysfunction (MODS) - from the Molecular and cellular level to the whole body. In: Zheng-yao Luo (ed) Thanjon Science Press, Changsha, pp 1-2

14. Childs EW, Woods JG, Smalley DM (1999) Leukocyte adherence and sequestration following hemorrhagic shock and total ischemia in rats. Shock 11 :248

32 A.E. Baue

15. Deitch EA, Adams C, Lu Q, et a1 (2001) A time course study of the protective effect of mesenteric lymph duct ligation on hemorrhagic shock-induced pulmonary injury and the toxic effects of lymph from shocked rats on endotheial cell monolayer permeability. Surgery 129:39-47

16. Botha AJ, Moore FA, Moore EE, et al (1995) Post-injury neutrophils priming and activation states: therapeutic challenges. Shock 3:157-166

17. Lemaire L, van Lanschott J, Stoutenbeek C, et al (1999) Thoracic duct in patients with multiple organ failure: no major route of bacterial translocation. Ann Surg 229: 128-36

18. Baue AE (2001) Sepsis and organ dysfunction: an overview of the new science and new biology. In: Baue, AE, Berlot G, Gullo A, Vincent JL (eds) Sepsis and organ dysfunction. Springer Italia, Milan, pp. 123-132

19. Baue AE (1999) Introduction to sepsis and organ dysfunction. In: Baue AE, Berlot G, Gullo A, Vincent JL (eds) Sepsis and organ dysfunction, Springer-Verlag Italio, Milan

20. Baue AE, Berlot, Gullo A, Vincent JL (eds) (1999) Sepsis and organ dysfunction. Springer-Verlag Italia, Milan pp 13-19

21. Dhainaut JF, Giroir B, Opal S (2001) Introduction to the Second Margaux Conference on critical illness sepsis: interface between inflammation, coagUlation, and the endothelium. Crit Care Med 29:S 1

22. Beutler B, Poltorak A (2001) Sepsis and evolution of the innate immune response. Crit Care Med 29:S2-S7

23. Hack E, Zeerleder S (2001) The endothelium in sepsis: source of and a target for inflammation. Crit Care Med 29: S21-S27

24. Aird WC (2001) Vascular bed-specific hemostasis: role of endothelium in sepsis pathogenesis. Crit Care Med 29:S28-S35

25. Vallet B, Weil E (2001) Endothelial cell dysfunction and coagulation. Crit Care Med 29:536-541 26. Dhainaut JF, Marin N, Mignon A, Vinsonneau C (200 I) Hepatic response to sepsis: Interaction

between coagulation and inflammatory processes. Crit Care Med 29:S42-S47 27. Esmon CT (2001) Protein C anticoagulant pathway and its role in controlling microvascular

thrombosis and inflammation. Crit Care Med 29:S48-S52 28. Grinnell BW, Joyce D (2001) Recombinant human activated protein C: a system modulator of

vascular function for treatment of severe sepsis. Crit Care Med 29:S53-S61 29. Bernard G, Vincent J, Laterre P, et al (2001) Efficacy and safety of recombinant human activated

protein C for severe sepsis. N Engl J Med 344:699-709 30. Marshall JC (2001) Inflammation, coagulopathy and the pathogenesis of multiple organ dysfunc-

tion syndrome. Crit Care Med 29:S99-S 1 06 31. Angus DC, Wax RS (2001) Epidemiology of sepsis - an update. Crit Care Med 29:S109-S116 32. Baue AE (1997) SIRS, MODS, MOF - why no magic bullets? Arch Surg 132: 1-5 33. Baue AE (1997) Multiple organ failure, multiple organ dysfunction syndrome, and systemic

inflammatory response syndrome: why no magic bullets? Arch Surg 132:703-707 34. Baue AE (2001) MOF, MODS, SIRS - why no magic bullets? Proc Assoc Pol Surg (in press) 35. Buchman TG (1996) Physiologic stability and physiologic state. J Trauma 41:599-605 36. Medical Research Council (1950). Treatment of pulmonary tuberculosis with streptomycin and

para-aminosalicylic acid. BJM 2: 1073-1085 37. Marshall EK Jr (1964) Historical perspectives in chemotherapy. In: Goldin A, Hawking IF (eds)

Advances in chemotherapy, vol 1. Academic Press, New York, p 1 38. DeVita VT Jr, Schein PS (1973) Medical progress - the use of drugs in combination for the

treatment of cancer, rationale and results. N Engl J Med 288:998-1006 39. Opal S, Cross AS, Sadoff JC, Fisher CJ Jr (1995) Shock 3 [Suppl):65 40. Faist E (1995) Immunomodulatory approaches in critically ill surgical patients (abstract). Shock

3 [Suppl]:65-66 41. Knox J, Demling R, Wilmore D, et al (1995) Increased survival after major thermal injury: the

effect of growth hormone therapy in adults. J Trauma 39:526-530 42. Kirton 0, Windsor J, Civetta, et al (1996) Persistent uncorrected intramucosal pH in the critically

injured: the impact of splanchnic and antioxidant therapy (abstract). Crit Care Med 24:A82


43. Kilbourn RG, Szabo S, Traber DL (1997) Beneficial versus detrimental effects of nitric oxide synthase inhibitors in circulatory shock: lessons learned from experimental and clinical studies. Shock 7:235-246.

44. Redl H, Schlag G, Bahrami S, Yao YM (1996) Animal models as the basis of pharmacologic intervention in trauma and sepsis patients. World J Surg 20:487-492

45. Luiten EJT, Wim CJ, Lange JF, Bruining HA (1995) Controlled clinical trial of selective decontamination for the treatment of severe acute pancreatitis. Ann Surg 222:57-65

46. Lingnau W, Berger J, Javorsky F, Benzer H (1997) Selective gut decontamination in multiple trauma patients: a prospective, randomized trial. J Trauma 42:687 -694

47. Braga M, Vignali A, Gianotti L, et al (1995) Benefits of early postoperative enteral feeding in cancer patients. Infusionther Transfunsionsmed 22:280-284

48. Meduri GU, Cinn AJ, Leeper KV, et al (1994) Corticosteroid rescue treatment of progressive fibroproliferation in late ARDS. Patterns of response and predictors of outcome. Chest 105: 1516--1527

49. Baue AB (1998) When to operate or stop operating and to plan re-operation. In: Baue AB, Berlot G, Gullo A (eds) Sepsis and organ dysfunction. Springer-Verlag Milan, pp 131-144

50. Baue AE (1999) Prevention and treatment of sepsis: MODS, MOF - what is wrong? What is right? Present and future problems. In: Baue AE, Berlot G, Gullo A, Vincent JL (eds) Sepsis and organ dysfunction. Springer-Verlag Italia, Milan, pp 69-82

51. Baue AB (1999) What is clinical relevance? Well-controlled experiments in normal animalsclinical studies in diverse sick patients. In: Baue AB, Berlot G, Gullo A, Vincent JL (eds) Sepsis and organ dysfunction. Springer-Verlag Italia, Milan, pp 95-104

52. Baue AE (1999) Can multiple agents which reduce morbidity individually reduce mortality collectively? In: Baue AB, Berlot G, Gullo A, Vincent JL (eds) Sepsis and organ dysfunction. Springer-Verlag Italia, Milan, pp 123-136

53. Baue AB (1999) Sepsis, multi-organ dysfunction syndrome (MODS) and mUltiple organ failure (MOF). Prevention is better than treatment. Minerva AnestesioI65:477-480

54. Baue AE (2000) History of MOF, the concept of limits and the importance of prevention. What is organ failure? In: Baue AB, Faist E, Fry D (eds) Multiple organ failure. Springer-Verlag, Berlin Heidelberg New York, pp. 3-13

55. Baue AE (2000) Problems with magic bullets - future trials and multi-agent therapy. In: Baue AB, Faist E, Fry D (eds) Multiple organ failure. Springer-Verlag, Berlin Heidelberg New York, pp 562-570

56. Baue AE. (2000) Are we making progress in preventing and/or treating MOF? Are we winning the battle? In: Baue AE, Faist E, Fry D. (eds) Multiple organ failure. Springer-Verlag, Berlin Heidelberg New York, pp 656-662

57. Baue AB, Faist E, Fry D (2000) Summary and overview: Prevention is the best answer. In: Baue AB, Faist E, Fry D (eds) Multiple organ failure. Springer-Verlag, Berlin Heidelberg New York, pp 687-691

58. Thangathurai D, Charbonnet C, Wo CCJ, et al (1996) Intraoperative maintenance of tissue perfusion prevents ARDS. New Horiz 4:446-474

59. Shoemaker WC, Thangathurai D, Wo CCJ, et al (1999) Intraoperative evaluation of tissue perfusion in high-risk patients by invasive and noninvasive hemodynamic monitoring. Crit Care Med 27:2147-2152

60. Baue AB (2000) MODS/MOF - a complication of progress in organ support. Shock. 13:7 61. Baue AB (2001) MODS, MOF, and SIRS - are we improving patient care? In: Zheng-yao Luo

(ed) Changsha, PRC 828-845 62. Baue AE (2001) Bad and good news in prevention and management in sepsis and MODS.

Minerva Anestesiol (in press) 63. Baue AE, Redi H (1998) Multiple therapeutic agents - are we making progress? In: Baue AB,

Berlot G, Gullo A (eds) Springer-Verlag, Milan, pp 145-152 64. Baue AB (1999) Injury, inflammation and sepsis - is there a natural, organized and sequential

34 A.E. Baue

progression of neuro-endocrine, metabolic and cytokine mediator events leading to organ system failure? In: Baue AE, Berlot G, Gullo A, Vincent JL (eds) Sepsis and organ dysfunction. Springer-Verlag ltalio, Milan, pp 35-47

SEPSIS AND ORGAN DYSFUNCTION: FROM CHAOS TO RATIONALE

Alveolar Epithelium in Host Defence: Cytokine Production

M.Lru

Infection of the respiratory system is a frequent cause of morbidity and mortality world-wide [1]. Respiratory tract infection, including croup, tracheobronchitis, bronchiolitis and pneumonia, are significant clinical problems. The increasing number of multidrug-resistant microbes has made the treatment of these infections much more difficult [1]. To further improve therapies for respiratory infection, we need to learn more about the host defence in the lung.

Host defence in the lung

The respiratory tract is accessible for potentially infective micro-organisms and noxious substances in the inhaled air. Thus, lung defence mechanisms are crucial for the effective removal of microbes and other debris from the conducting airways and alveoli [2, 3]. Host defence in the respiratory system includes three major components: mechanical (such as cough and mucociliary clearance), humoral (such as secretory immunoglobulins and complement) and cellular (such as alveolar macrophages, lymphocytes, and neutrophils) [4]. Recent studies have shown that the alveolar epithelium is also an important component in the host defence. It functions as a barrier to prevent the invasion of pathogens. Type II pneumocytes produce lung surfactant that can enhance the function of immune cells in the alveoli. Surfactant proteins are also important mediators of host defence. In addition, lung alveolar epithelial cells may also function as sensors for the invasion of micro-organisms and other noxious agents by producing cytokines and chemokines.

Alveolar epithelial cells as a source of cytokines and chemokines

Cytokines are extracellular signalling proteins secreted by cells, which have the ability to modify the behaviour of other adjacent cells. Cytokines are generally

38 M.Liu

divided into pro- and anti-inflammatory mediators, and are important mediators of both innate and acquired immune defences in the lung [1]. Chemokines are chemotactic cytokines for leukocyte recruitment and activation at the sites of infection or tissue injury [2, 3]. The role of chemokines in lung host defence has been a subject of several reviews [1,5]. They are also important mediators in acute lung inflammation [2, 3, 5, 6]. Neutrophil infiltration in the alveolar space is mainly mediated by C-X-C chemokines such as interleukin-8 (lL-8) and its rodent homologue, macrophage inflammatory protein-2 (MIP-2) [6,7]. The C-C chemokine family, such as monocyte chemoattractant proteins (MCPs) and RANTES, activates and/or is chemotactic for macrophages, monocytes and lymphocytes.

Both in vitro and in vivo data suggest that alveolar epithelial cells can produce cytokines such as IL-6, IL-3, interferon ,,(, granulocyte monocyte colony-stimulating factor and tumour necrosis factor-a. (TNFa.) [8]. Alveolar epithelial cells can also produce a variety of chemokines such as IL-8 [9], MIP-2 [10], MCP-l [11, 12], and RANTES [13].

During foetal lung development, the potential airway and alveolar space are filled with amniotic fluid. There are few macrophages and other immune cells in the alveoli. It is unknown how these immune cells are initially recruited and become the residents in the alveolar space after birth. Cytokines and chemokines produced by pulmonary epithelial cells may initiate the establishment of host defence. This is very important for newborns immediately after the birth and for children in their early childhood. Inappropriate recruitment and activation of immune cells in airway and alveoli may contribute to recurrent infections in pediatric lungs.

Cytokine and acute lung injury

The recruited inflammatory cells help remove invading organisms through phagocytic clearance. However, in addition to their defensive role, these immune cells are also involved in acute and chronic injury of the lung. The cytotoxic and proteolytic materials, such as neutrophil elastase, contained in these immune cells may induce lesional changes. Cytokines, especially pro-inflammatory cytokines and some chemokines, also play an important role in acute lung injury, seen in many clinical situations: severe respiratory infection, sepsis, shock, acute respiratory distress syndrome (ARDS), mechanical ventilation-induced lung injury, and ischaemia-reperfusion injury of lung transplants. Immunotherapies have been developed to inhibit pro-inflammatory cytokines [14], such as TNFa. and IL-l [15, 16], and to consequently inhibit

Alveolar Epithelium in Host Defence: Cytokine Production 39

the acute inflammatory response. One of the problems with this approach is that it may disable host defence as well. To overcome the conflict between host defence and acute inflammatory injury in the lung, we need to understand how cytokine production from alveolar epithelial cells and other cellular sources is regulated, and to explore new strategies to control the production of cytokines and chemokines.

The interaction of leukocytes and pulmonary parenchymal cells, including alveolar epithelial cells, via cytokine signalling mediates innate and acquired immunity in lung antimicrobial host defence [1, 17]. Enhanced pro-inflammatory cytokine expression has been attempted as new therapies for lung infection, but the concern is that this strategy may exacerbate acute lung injury.

Regulation of cytokine production

Investigations with macrophages, monocytes, neutrophils, and other immune and non-immune cells have yielded fruitful results regarding the regulation of cytokine production. Given space constraints, we cannot review these exciting studies; instead, several examples are given to illustrate the complexity of regulatory mechanisms of cytokine production at various levels. Recently, many cytokine and chemokine receptors have been characterized at the molecular and cellular level [18, 19]. Soluble TNF receptor [20] and naturally expressed IL-l antagonist have been recognized as potent inhibitors to block the function of these pro-inflammatory cytokines [15, 16]. These molecules have been used in pre-clinical and clinical trials [14]. There have been exciting discoveries elucidating signal transduction pathways initiated by LPS and cytokines leading to nuclear events. The importance of protein phosphorylation, especially tyrosine phosphorylation [21], in cytokine production has been reported. Stress activated protein kinase (SAPK, also called JNK) [22], and p38MAPK [23] have been demonstrated to specifically mediate signals initiated by cytokines and other inflammatory mediators. Pharmacological agents targeting these pathways have been developed for clinical applications. The role of nuclear factor-KappaB (NFKB) as a transcriptional factor in controlling cytokine gene expression has been reported from many cell types under different experimental conditions for several cytokines [24]. In macrophages and other immune cells, cytokine synthesis can be triggered rapidly, and this apparently involves predominantly translation rather than gene transcription [25]. Detailed molecular studies have revealed A-U rich elements in the 3'untranslated region of many cytokine mRNAs [26], which play an important role in controlling cytokine protein synthesis. The intracellular transport and

40 M.Liu

secretion of cytokines is another important regulatory step for cytokine production.

Compared to the inflammatory cells, our knowledge on the regulation of cytokine production from alveolar epithelial cells is much less. In this article, we will use IL-8 and MIP-2 as examples, to discuss the induction of cytokine, the interaction of various cytokines as a network, the transcriptional regulation of cytokine gene expression, and the role of cytoskeleton in regulating cytokine secretion, to elucidate the complexity of cytokine production from alveolar epithelial cells.

Induction of cytokine from alveolar epithelial cells

Production of cytokines from alveolar epithelial cells is an important response towards a variety of stimuli from environmental factors, bacteria, viruses and other stresses. Using IL-8 as an example, many factors can directly induce this cytokine from cultured human lung epithelial cells (Table 1). IL-8 is one ofthe best known C-X-C chemokines to attract and activate polymorphonuclear granulocytes (PMNs) [6]. The biological activities of IL-8 include attracting neutrophils, activating surface adhesion molecules, inducing release of storage enzymes, and stimulating production of reactive oxygen metabolites [27]. IL-8 has been found to be involved in several inflammatory reaction-related diseases in the lung, for example, idiopathic pulmonary fibrosis [28], adult respiratory

Table 1. Induction of IL-8 from lung alveolar epithelial cells

Stimulus

Environmental factors and toxins

Viruses

Environmental particulate, Fibrous particles, Asbestos, Silica, Dust from waster handling facilities, Coal fly ash, House dust, Smoke extract, Ragweed, Fungal allergens, Ozone

Respiratory syncytial virus, Influenza virus, Adenovirus, Rhinovirus

Bacteria and products Gram positive bacteria, Mycobacterium tuberculosis, Thermophilic bacteria, Lipopolysaccharides, Burkholderia cepacia products, Pneumococcal protein, Proteases from Aspergillus jumigatus, Pneumocystis carinii major surface glycoprotein, Pseudomonas nitrite reductase

Cytokines and inflammatory mediators TNFa, IL-J a, IL-l E, Th 2 cytokines, Neutrophil serine proteinases, Defensins, Bradykinin

Other stress Hyperoxia, Anoxia-hyperoxia, Mechanical stretch


distress syndrome [29], and empyema [30]. A monoclonal antibody against IL-8 prevented ischaemia-reperfusion induced lung injury in a rabbit model [31].

Various environmental particulate, fibrous particles, such as asbertos and silica induced IL-8 production, as well as house dust, dust from waster handling facilities, coal fly ash and smoke extract. Fungal allergens, ragweed and other allergens in the air also stimulate alveolar epithelial cells to produce IL-8. Various viruses, such as respiratory synthial virus (RSV), adenovirus, influenza virus and rhinovirus, induced IL-8 production from alveolar epithelial cells. Replication-deficient adenoviral vectors-induced cytokine production from alveolar epithelial cells has drawn increasing attention in gene therapy-related investigations. During bacterial infection, both bacteria (such as Gram positive bacteria and Mycobacterium tuberculosis), and their products (lipopolysaccharides, Burkholderia cepacia products, pneumococcal protein, proteases from Aspergillus jumigatus, Pneumocystis carinii major surface glycoprotein, and Pseudomonas nitrite reductase) induced IL-8 production. Recently it has been demonstrated that primary cultured alveolar epithelial cells isolated from human lung tissues produced IL-8 [32]. Therefore, IL-8, as well as many other cytokines and chemokines, are important messengers for the host defence in the alveolar spaces, produced by alveolar epithelial cells.

IL-8 has structural and biological similarities with MIP-2, which could represent the rodent homologue to IL-8 [6]. MIP-2 was initially purified from a mouse macrophage cell line stimulated with endotoxin [33]. Rat MIP-2 was recently cloned and expressed as a 7.9-kDa peptide [34-36] that showed dose-dependent chemotactic activity for PMNs [34]. This activity ofMIP-2 has been further demonstrated in the lung from several animal models with a variety of pathogens. For examples, increased MIP-2 mRNA and/or protein was observed in the lung, in response to the intra-tracheal instillation of Klebsiella pneumoniae, Pseudomonas aeruginosa, LPS and a-quartz. Intra-peritoneal administration of LPS resulted in an increase in neutrophil influx into the lung, which was at least in part due to increased levels of MIP-2 [37]. LPS induced MIP-2 production from lung explants [38]. LPS also induced MIP-2 production from primary cultured rat lung alveolar epithelial cells, which was regulated at both the transcriptional and post-transcriptional levels [10].

MIP-2 is also involved in lung injury, such as IgG immune complex-induced injury. When MIP-2 trans gene was delivered into the lung with a replicationdefective adenoviral vector through the intratracheal instillation, significant increase in neutrophils and alveolar macrophages was found from the lung lavage fluid [39]. Instillation of recombinant MIP-2 into the alveolar space of rats induced profound neutrophil localization both in the vascular and alveolar space [37]. In these studies, up-regulation of MIP-2 was associated with acute

42 M.Liu

lung injury. Intraperitoneal instillation of anti-MIP-2 antiserum [40] or intrapulmonary instillation of anti-MIP-2 antibodies [39,40] decreased neutrophil influx in the lung and attenuated lung injury.

Interactions between cytokines

Micro-organisms and other pathogen-induced IL-8 and MIP-2 production can be mediated through other cytokines. TNF-a is one of the most important pro-inflammatory cytokines in the cytokine network. It is a very important mediator in host defence and in mediating acute inflammatory reactions in the lung and many other organ systems. Recent studies have demonstrated that, in response to LPS-stimulation, primary cultured rat alveolar epithelial cells produced TNF-a in a dose- and time-dependent manner [8]. TNF-a can induce IL-8 from human lung A549 cells [41-43], or from primary cultured human alveolar epithelial cell [32]. Furthermore, recombinant TNF-a induced MIP-2 production from primary cultured rat lung alveolar epithelial cells [10]. A time-delay between TNF-a and MIP-2 at both mRNA and protein levels was noted upon LPS-stimulation [8,10]. When an antisense oligonucleotide against rat TNF-a was delivered to alveolar epithelial cells, it inhibited not only TNF-a but also MIP-2 release in a dose-dependent fashion. The inhibitory effects on these two molecules were highly correlated [10]. Neutralizing anti-TNF-a antibody also inhibited MIP-2 production [10]. These results suggested that TNF-a released from these cells might function as an alert signal to trigger the production of chemokines such as MIP-2 in rat and IL-8 in human lungs. The latter may recruit neutrophils to the alveoli where the bacteria or other pathogens have invaded. This auto-regulation of the cytokine network may be important for host defence and could be augmented during acute lung injury [10].

IL-l is another cytokine with potent proinflammatory effects. IL-l~ induced IL-8 from human lung alveolar epithelial cells [44]. IL-la-induced neutrophil migration across A549 cell layer is partially mediated through IL-8 [45]. Incubation with neutralizing antibodies against IL-la, IL-l~ and TNF-a showed that IL-la was the predominant soluble mediator that enhanced the mRNA expression and synthesis oflL-8 induced by RSV [46]. IL-l receptor antagonist inhibited IL-8 expression in A549 cells infected in vitro with a replication-deficient recombinant adenovirus vector [47]. These results suggest that IL-l a or IL-l ~ could also function as autocrine regulators to stimulate IL-8 production from alveolar epithelial cells. It is worthwhile to note that it has been observed that TNF-a increased both IL-8 mRNA expression and protein production in isolated human alveolar type II epithelial cells, whereas IL-l~ slightly increased IL-8 release but did not change its mRNA expression [32].


The results from cell lines such as A549 cells need to be interpreted with caution.

Transcriptional regulation of cytokine gene expression in alveolar epithelial cells

The gene expression of cytokine is regulated at the transcriptional level that is mediated via intracellular signal transduction pathways. Activation of nuclear factor NF-l(B is one of the most important regulatory mechanism for IL-8 gene expression induced by RSV, rhinovirus, and human immunodeficiency virus type I protein R in human lung epithelial cells. NF-l(B also plays an essential role in regulation of IL-8 gene expression induced by nitrite reductase from Pseudomonas aeruginosa in respiratory epithelial cells [48]. Asbestos fibers also stimulated DNA binding activity to the regulatory elements in the IL-8 promoter, binding sites ofNF-l(B- and NF-IL-6-like transcription factors [49]. Another important transciptional activation is through Activator protein-l (AP-l), which consisted of Jun and Fos proteins. Although both H20 2 and TNFa can induced IL-8 production in lung epithelial cells, they induce differential binding ofthe redox-responsive transcription factors to the IL-8 promoter. H20 2 activates AP-l but not NF-l(B in A549 cells, whereas TNFa activated both AP-l and NF-l(B [41]. TNFa-induced NF-l(B activation and IL-8 release in A549 cells can be inhibited with the proteasome inhibitor MG-132, which blocks the degradation of NF-l(B complex [43]. AP-l is also the preferred transcription factor for cooperative interaction with NF-l(B in RSV -induced IL-8 gene expression in airway epithelium [50].

The transcriptional activation of IL-8 is mediated through intracellular signal transduction pathways. Asbestos-inducible IL-8 secretion was suppressed by staurosporine, an inhibitor of PKC, and also by inhibitors of tyrosine kinase such as herbimycin A and genistein. The suppression effect paralleled the effect of these inhibitors on asbestos-induced DNA binding to the NF-l(B - and NF-IL-6-like binding sites ofthe IL-8 promoter [49]. Mechanical stretchinduced activation of mitogen-activated protein kinases (MAPK), including c-Jun N-terminal kinase (JNK), p38 and extracellular signal-regulated kinase (ERK) [51]. These MAPK isoforms could be involved in the regulation of transcriptional factors, especially AP-l. IL-8 production in type II alveolar cells is associated with the activation ofJNK [52]. The IL-IS induced JNK activation is through RhoA, a small G protein, whereas H20 2-induced JNK activation is through phosphoinosital-3 kinase and phospholipase A2 pathway [53]. RSV infection results in activation of multiple protein kinase C (PKC) isoforms leading to activation ofMAPK [54]. Activation ofERK2 by RSV in A549 cells

44 M.Liu

is linked to the production ofIL-8 [55]. Further investigation of these pathways may lead to specific regulation of cytokine production from lung alveolar epithelial cells.

Role of cytoskeleton in LPS-induced cytokine secretion from alveolar epithelial cells

The cytoskeletal system of cells is composed of microfilaments, microtubules and intermediate filaments. Recent studies have shown that both microfilaments and microtubules are involved in regulating cytokine production from alveolar epithelial cells. Importantly, these effects appear to be opposite to that in inflammatory cells.

LPS suppressed macrophage phagocytosis by affecting microfilament and microtubule structures [56]. LPS induced a rapid reorganization of F-actin assembly in macrophages [57], increased stiffness and F-actin assembly in monocytes [58] and enhanced a chemotactic factor induced actin polymerization in neutrophils [59]. In contrast, LPS reduced polymerization of microfilaments in primary cultured rat alveolar epithelial cells [60, 61]. Cytochalasin D (CytoD), a microfilament-disrupting agent, blocked LPS-induced TNFa gene expression and/or protein synthesis in macrophages [62]. In contrast, CytoD enhanced LPS-induced TNFa production from rat pneumocytes. A membrane-permeable cyclodepsipeptide, jasplakinolide, can induce actin polymerization and stabilize pre-existing actin filaments [63]. When cells were treated with jasplakinolide, it inhibited LPS-induced TNFa production from rat pneumocytes, but enhanced it from macrophages [60]. The LPS-induced depolymerization of microfilaments has similar effect on LPS-induced MIP-2 production from these cells [61]. Mechanical stretch-induced cytoskeletal deformation enhanced MIP-2 secretion from primary cultured foetal rat lung cells [64].

Disassembly of actin filaments has been found from many non-immune cells that play a significant role in secretion. The cytoskeletal structure endows the cell with a very crowded cytoplasm, and the integrated organization of the cytoskeleton and membrane systems may provide an important barrier to the free diffusion of secretory vesicles [65]. During resting conditions the actin cytoskeleton, localized under the plasma membrane, may prevent secretory granules from reaching their exocytic destination. Upon stimulation, microfilaments may be disassembled or rearranged to allow secretory granules to reach the site of exocytosis [66].

Cytoskeletal elements, particularly micro tubules and their associated motor proteins, are fundamental in facilitating delivery of transport intermediates between spatially segregated organelles and determine the steady-state locali-


zation of the organelles [67]. Eukaryotic cells have highly regulated membrane transport systems that mediates exchange of protein and lipid between distinct membrane-bound compartments of organelles, including the endoplasmic reticulum (ER), Golgi, transport intermediates and others. In higher eukaryotic cells, the ER and Golgi complex are spatially segregated. The ER network with branching membrane tubules extends outward along microtubules throughout the cell [68], while Golgi cisternae are clustered around microtubules near the perinuclear microtubule organizing centre. Transport intermediates arising from peripheral ER sites, thus, often travel considerable distances to reach the Golgi complex [69]. Therefore, microtubules may play an important role in the intracellular transport of cytokine molecules.

MIP-2, as well as most cytokines and chemokines, are synthesized as precursor polypeptides, containing cleavable N-terminal signal or targeting sequences for transport through the ER-Golgi pathway [70]. When cells were incubated with brefeldin A (BFA), which blocks the ER-to-Golgi transportation, LPS-induced MIP-2 production was inhibited in a dose-dependent manner [71]. Microtubules have been recognized as secretory "highways" in the cell [65]. Membranes move along microtubules in both directions between the ER and Golgi, and at the steady state, forward (ER-to-Golgi) and reversed transportation is in balance [65]. Microtubules are also involved in transportation of secretory vesicles from the Golgi to the plasma membrane [67, 69]. Using fluorescent and immunofluorescent staining and confocal microscopy, it was found that LPS reduced polymerization of micro tubules [60,71], whereas LPS increased the number, length, and stability of microtubules in mononuclear phagocytes [72]. When alveolar epithelial cells were pre-incubated with various concentrations of microtubule-disrupting agents, colchicine or nocodazole, LPS-induced-MIP-2 production was further enhanced in a dose-dependent fashion. Alternatively, when cells were stimulated with various concentrations of LPS in the presence of colchicine or nocodazole, both agents increased LPS-induced MIP-2 production over a wide range of LPS concentrations. Taxol, a microtubule-stabilizing agent, partially inhibited LPS-induced MIP-2 production [71].

Although both the anterograde and retrograde traffic depend upon microtubuIes, LPS may selectively block the retrograde transportation from the Golgi back to the ER. Microtubules are very important in maintaining the intracellular localization of the ER and Golgi complex, which are essential in determining the pathways for secretory proteins. When cells were treated with nocodazole, depolymerization of microtubules leads to redistribution of the ER and Golgi [73, 74]. LPS-induced depolymerization of microtubules, especially in the presence of nocodazole or colchicine, may change the distribution of the ER

46 M.Liu

and Golgi in alveolar epithelial cells, which may lead to a microtubule-independent secretion from the ER to plasma membrane. These two mechanisms may be both involved in LPS-induced secretion of MIP-2 as well as other cytokines from alveolar epithelial cells.

Based on the roles of the cytoskeleton in secretion, both microfilaments and microtubules may be involved in regulating cytokine transportation in peumocytes through different mechanisms. The effects ofLPS on the cytoskeleton and the roles of the cytoskeleton in mediating LPS-induced TNFa production in alveolar epithelial cells are opposite to that in immune cells. Selective inhibition of cytokine production from different cell types could be beneficial. For example, ventilation-induced TNFa could be mainly from alveolar epithelial cells. If we block TNFa produced from alveolar epithelial cells, while maintaining the ability of alveolar macrophages to produce the cytokine, lung injury might be ameliorated without compromising host defence.

Conclusions

In this chapter, we described the role of alveolar epithelial cells in the host defence in the lung, as a source of cytokines. Using IL-8 and MIP-2 as examples, it can be see that cytokines can be induced by a variety of environmental factors, bacteria, viruses, and other pathogens. Cytokine production is regulated as a network via autocrine and paracrine mechanisms. The intracellular signal transduction and transcriptional regulation of cytokine production from alveolar epithelial cells are complex. The cytoskeletal system plays an important role in controlling cytokine secretion from alveolar epithelial cells. This effect seems to be opposite between alveolar epithelial cells and other immune cells, which provide an opportunity to selectively control the cytokine production from a different cellular source. This concept may have significant clinical impact to reduce acute inflammatory response, but keep the host defence response intact.

References 1. Standiford TJ, Huffnagle GB (1997) Cytokines in host defense against pneumonia. J Invest Med

45:335-345 2. Baggiolini M, Dewald B, Moser B (1997) Human chemokines: an update. Annu Rev Immuno1

15:675-705 3. Baggiolini M (1998) Chemokines and leukocyte traffic. Nature 392:565-568 4. Simon RH, Paine III R (1995) Participation of pulmonary alveolar epithelial cells in lung

inflammation. J Lab Clin Med 126: 108-118


5. Standiford TJ, Strieter RM, Greenberger MJ, Kunkel SL (1996) Expression and regulation of chemokines in acute bacterial pneumonia. BioI Signals 5:203-208

6. Baggiolini M, Dewald B, Moser B (1994) Interleukin-8 and related chemotactic cytokines--CXC and CC chemokines. Adv Immunol 55:97-179

7. Wolpe SD, Cerami A (1989) Macrophage inflammatory proteins 1 and 2: members of a novel superfamily of cytokines. FASEB J 3:2565-2573

8. McRitchie DI, Isowa N, Edelson JD et al (2000) Production of tumour necrosis factor alpha by primary cultured rat alveolar epithelial cells. Cytokine 12:644-654

9. Standiford TJ, Kunkel SL, Basha MA et al (1990) Interleukin-8 gene expression by a pulmonary epithelial cell line. A model for cytokine networks in the lung. J Clin Invest 86: 1945-1953

10. Xavier AM, Isowa N, Cai L, et al (1999) Tumor necrosis factor-alpha mediates lipopolysaccharide-induced macrophage inflammatory protein-2 release from alveolar epithelial cells. Autoregulation in host defense. Am J Respir Cell Mol Bioi 21:510-520

11. Standiford TJ, Kunkel SL, Phan SH et al (1991) Alveolar macrophage-derived cytokines induce monocyte chemoattractant protein-l expression from human pulmonary type II-like epithelial cells. J BioI Chern 266:9912-9918

12. Paine III R, Rolfe MW, Standiford TJ et al (1993) MCP-l expression by rat type II alveolar epithelial cells in primary culture. J ImmunoI150:4561-4570

13. VanOtteren GM, Strieter RM, Kunkel SL et al (1995) Compartmentalized expression of RANTES in a murine model of endotoxemia. J ImmunoI154:1900-1908

14. Liu M, Slutsky AS (1997) Anti-inflammatory therapies: application of molecular biology techniques in intensive care medicine. Inten Care Med 23:718-731

15. Fisher CJ, Jr., Dhainaut JF, Opal SM et al (1994) Recombinant human interleukin 1 receptor antagonist in the treatment of patients with sepsis syndrome. Results from a randomized, double-blind, placebo-controlled trial. Phase III rhIL-lra Sepsis Syndrome Study Group. JAMA 271: 1836-1843

16. FishereJ, Jr., Slotman GJ, Opal SM et al (1994) Initial evaluation ofhurnan recombinant interleukin-l receptor antagonist in the treatment of sepsis syndrome: a randomized, open- label, placebo-controlled multicenter trial. The IL-IRA Sepsis Syndrome Study Group. Crit Care Med 22: 12-21

17. Standiford TJ, Kunkel SL, Greenberger MJ et al (1996) Expression and regulation of chemokines in bacterial pneumonia. J Leukoc Bioi 59:24-28

18. Baggiolini M, Moser B (1997) Blocking chemokine receptors. J Exp Med 186:1189-1191 19. Bazzoni F, Beutler B (1996) The tumor necrosis factor ligand and receptor families. New Eng J

Med 334:1717-1725 20. Peppel K, Crawford D, Beutler B (1991) A tumor necrosis factor (TNF) receptor-IgG heavy

chain chimeric protein as a bivalent antagonist ofTNF activity. J Exp Med 174:1483-1489 21. Lamer AC, Finbloom DS (1995) Protein tyrosine phosphorylation as a mechanism which

regulates cytokine activation of early response genes. Biochim Biophys Acta 1266:278-287 22. Lee JC, Laydon JT, McDonnell PC et al (1994) A protein kinase involved in the regulation of

inflammatory cytokine biosynthesis. Nature 372:739-745 23. Beyaert R, Cuenda A, Vanden Berghe W et al (1996) The p38/RK mitogen-activated protein

kinase pathway regulates interleukin-6 synthesis response to tumor necrosis factor. EMBO J 15:1914-1923

24. Collart MA, Baeuerle P, Vassalli P (1990) Regulation of tumor necrosis factor alpha transcription in macrophages: involvement of four kappa B-like motifs and of constitutive and inducible forms ofNF-kappa B. Mol Cell Bioi 10:1498-506

25. Han J, Brown T, Beutler B (1990) Endotoxin-responsive sequences control cachectinffNF biosynthesis at the translational level. J Exp Med 171:465-475

26. Shaw G, Kamen R (1986) A conserved AU sequence from the 3' untranslated region of GM -CSF mRNA mediates selective mRNA degradation. Cell 46:659-667

27. Baggiolini M, Walz A, Kunkel SL (1989) Neutrophil-activating peptide-llinterieukin 8, a novel cytokine that activates neutrophils. J Clin Invest 84: 1045-1049

48 M.Liu

28. Carre PC, Mortenson RL, King TE, Jr. et al (1991) Increased expression of the interleukin-8 gene by alveolar macrophages in idiopathic pulmonary fibrosis. A potential mechanism for the recruitment and activation of neutrophils in lung fibrosis. J Clin Invest 88: 1802-1810

29. Miller EJ, Cohen AB, Nagao S et al (1992) Elevated levels ofNAP-lIinterieukin-8 are present in the airspaces of patients with the adult respiratory distress syndrome and are associated with increased mortality. Am Rev Respir Dis 146:427-432

30. Broaddus VC, Hebert CA, Vitangcol RV et al (1992) Interleukin-8 is a major neutrophil chemotactic factor in pleural liquid of patients with empyema. Am Rev Respir Dis 146:825-830

31. Sekido N, Mukaida N, Harada A et al (1993) Prevention of lung reperfusion injury in rabbits by a monoclonal antibody against interleukin-8. Nature 365:654-657

32. Pechkovsky DV, Zissel G, Ziegenhagen MW et al (2000) Effect of proinflammatory cytokines on interleukin-8 mRNA expression and protein production by isolated human alveolar epithelial cells type II in primary culture. Eur Cytokine Netw 11 :618-625

33. Wolpe SD, Davatelis G, Sherry Bet al (1988) Macrophages secrete a novel heparin-binding protein with inflammatory and neutrophil chemokinetic properties. J Exp Med 167:570-581

34. Shanley TP, Schmal H, Friedl HP et al (1995) Role of macrophage inflammatory protein-l alpha (MIP-l alpha) in acute lung injury in rats. J ImmunoI154:4793-4802

35. Feng L, Xia Y, Yoshimura T, Wilson CB (1995) Modulation of neutrophil influx in glomerulonephritis in the rat with anti-macrophage inflammatory protein-2 (MIP-2) antibody. J Clin Invest 95:1009-1017

36. Driscoll KE, Hassenbein DG, Howard BW et al (1995) Cloning, expression, and functional characterization of rat MIP-2: a neutrophil chemoattractant and epithelial cell mitogen. J Leukoc Bioi 58:359-364

37. Gupta S, Feng L, Yoshimura T et al (1996) Intra-alveolar macrophage-inflammatory peptide 2 induces rapid neutrophil localization in the lung. Am J Respir Cell Mol Bioi 15:656-663

38. Zhang H, Kim YK, Govindarajan A et al (1999) Effect of adrenoreceptors on endotoxin-induced cytokines and lipid peroxidation in lung explants. Am J Respir Crit Care Med 160: 1703-1710

39. Foley R, Driscoll K, Wan Y et al( 1996) Adenoviral gene transfer of macrophage inflammatory protein-2 in rat lung. Am J Patho1149: 1395-1403

40. Greenberger MJ, Strieter RM, Kunkel SL et al (1996) Neutralization of macrophage inflammatory protein-2 attenuates neutrophil recruitment and bacterial clearance in murine Klebsiella pneumonia. J Infect Dis 173:159-165

41. Lakshminarayanan V, Drab-Weiss EA, Roebuck KA (1998) H202 and tumornecrosis factor-alpha induce differential binding of the redox -responsive transcription factors AP-l and NF-kappaB to the interleukin-8 promoter in endothelial and epithelial cells. J Bioi Chern 273:32670-32678

42. Brasier AR, Jamaluddin M, Casola A et al (1998) A promoter recruitment mechanism for tumor necrosis factor-alpha-induced interleukin-8 transcription in type II pulmonary epithelial cells. Dependence on nuclear abundance of Rei A, NF-kappaB 1, and c-Rel transcription factors. J Bioi Chern 273:3551-3561

43. Fiedler MA, Wemke-Dollries K, Stark JM (1998) Inhibition ofTNF-alpha-induced NF-kappaB activation and IL-8 release in A549 cells with the proteasome inhibitor MG-132. Am J Respir Cell Mol Bioi 19:259-268

44. Standiford TJ, Kunkel SL, Basha MA et al (1990) Interleukin-8 gene expression by a pulmonary epithelial cell line. A model for cytokine networks in the lung. J Clin Invest 86: 1945-1953

45. Bittleman DB, Casale TB (1995) Interleukin-8 mediates interleukin-l alpha-induced neutrophil transcellular migration. Am J Respir Cell Mol Bioi 13:323-329

46. Patel JA, Jiang Z, Nakajima N, Kunimoto M (1998) Autocrine regulation of interleukin-8 by interleukin-Ialpha in respiratory syncytial virus-infected pulmonary epithelial cells in vitro. Immunology 95:501-506

47. Schwarz YA, Amin RS, Stark JM et al (1999) Interleukin-l receptor antagonist inhibits interleukin-8 expression in A549 respiratory epithelial cells infected in vitro with a replicationdeficient recombinant adenovirus vector. Am J Respir Cell Mol Bioi 21:388-394


48. Mori N, Oishi K, Sar B et al (1999) Essential role of transcription factor nuclear factor-kappaB in regulation of interleukin-8 gene expression by nitrite reductase from Pseudomonas aeruginosa in respiratory epithelial cells. Infect Immun 67:3872-3878

49. Simeonova PP, Luster MI (1996) Asbestos induction of nuclear transcription factors and interleukin-8 gene regulation. Am J Respir Cell Mol Bioi 15:787-795

50. Mastronarde JG, Monick MM, Mukaida N et al (1998) Activator protein-l is the preferred transcription factor for cooperative interaction with nuclear factor-kappaB in respiratory syncytial virus-induced interleukin-8 gene expression in airway epithelium. J Infect Dis 177: 1275-1281

51. Kito H, Chen EL, Wang X et al (2000) Role of mitogen-activated protein kinases in pulmonary endothelial cells exposed to cyclic strain. J Appl PhysioI89:2391-2400

52. Quinn D, Tager A, Joseph PM et al (1999) Stretch-induced mitogen-activated protein kinase activation and interleukin-8 production in type II alveolar cells. Chest 116:89S-90S

53. Roberts ML, Cowsert LM (1998) Interleukin-l beta and reactive oxygen species mediate activation of c-Jun NH2-terrninal kinases, in human epithelial cells, by two independent pathways. Biochem Biophys Res Commun 251: 166-172

54. Monick M, Staber J, Thomas K, Hunninghake G (2001) Respiratory syncytial virus infection results in activation of multiple protein kinase C isoforms leading to activation of mitogen-activated protein kinase. J ImmunoI166:2681-2687

55. Chen W, Monick MM, Carter AB, Hunninghake GW (2000) Activation ofERK2 by respiratory syncytial virus in A549 cells is linked to the production ofinterleukin 8. Exp Lung Res 26:13-26

56. Wondering RS, Ghaffar A, Mayer EP (1996) Lipopolysaccharide-induced suppression of phagocytosis: Effects on the phagocytic machinery. Immunopharmacol ImmunotoxicoI18:267-289

57. Shinji H, Kaiho S, Nakano T, Yoshida T (1991) Reorganization of micro filaments in macrophages after LPS stimulation. Exp Cell Res 193: 127-133

58. Doherty DE, Downey GP, Schwab III B, Elson E, Worthen GS (1994) Lipopolysaccharide-induced monocyte retention in the lung: role of monocyte stiffness, actin assembly, and CD 18-dependent adherence. J ImmunoI153:241-255

59. Howard TH, Wang D, Berkow RL (1990) Lipopolysaccharide modulates chemotactic peptideinduced actin polymerization in neutrophils. J Leukoc Bioi 47:13-24

60. Isowa N, Xavier AM, Dziak E et al (1999) LPS-induced depolymerization of cytoskeleton and its role in TNF-alpha production by rat pneumocytes. Am J PhysioI277:L606-L615

61. Isowa N, Liu M (2001) Role of LPS-induced microfilament depolymerization in MIP-2 production from rat pneumocytes. Am J Physiol Lung Cell Mol Physiol280:L 762-770

62. Shinji H, Akagawa KS, Yoshida T (1993) Cytochalasin D inhibits lipopolysaccharide-induced tumor necrosis factor production in macrophages. J Leukoc Bioi 54:336-342

63. Bubb MR, Senderowicz AM, Sausville EA et al (1994) Jasplakinolide, a cytotoxic natural product, induces actin polymerization and competitively inhibits the binding of phalloidin to F-actin. J BioI Chern 269:14869-14871

64. Mourgeon E, Isowa N, Keshavjee S et al (2000) Mechanical stretch stimulates macrophage inflammatory protein-2 secretion from fetal rat lung cells. Am J Physiol Lung Cell Mol Physiol 279:L699-706

65. Bloom GS, Goldstein LS (1998) Cruising along microtubule highways: how membranes move through the secretory pathway. J Cell Bioi 140:1277-1280

66. Valentijn KM, Gumkowski FD, Jamieson JD (1999) The subapical actin cytoskeleton regulates secretion and membrane retrieval in pancreatic acinar cells. J Cell Sci 112:81-96

67. Cole NB, Lippincott-Schwartz J (1995) Organization of organelles and membrane traffic by microtubules. Curr Opin Cell BioI 7:55-64

68. Saraste J, Kuismanen E (1992) Pathways of protein sorting and membrane traffic between the rough endoplasmic reticulum and the Golgi complex. Semin Cell BioI 3:343-55.

69. Lippincott-Schwartz J (1998) Cytoskeletal proteins and Golgi dynamics. Curr Opin Cell BioI 10:52-59

50 M.Liu

70. Muesch A, Hartmann E, Rohde K et al (1990) A novel pathway for secretory proteins? Trends Biochem Sci 15:86-88

71. Isowa N, Keshavjee SH, Liu M (2000) Role of microtubules in LPS-induced macrophage inflammatory protein-2 production from rat pneumocytes. Am J Physiol :In press.

72. Allen IN, Moore SA, Liao Z, Wewers MD (1997) Changes in mononuclear phagocyte microtubuIes after endotoxin stimulation. I. Changes in microtubule stability. Am J Respir Cell Mol BioI 16:119-126

73. Cole NB, Ellenberg J, Song J et al (1998) Retrograde transport of Golgi-Iocalized proteins to the ER. J Cell Bioi 140:1-15

74. Lippincott-Schwartz J, Donaldson JG, Schweizer A et al (1990) Microtubule-dependent retrograde transport of proteins into the ER in the presence of brefeldin A suggests an ER recycling pathway. Cell 60:821-836

Phagocytosis and Lung Injury

J.W. BOOTH

Phagocytosis of pathogens by macrophages and neutrophils is a key component of innate immunity, with internalization leading in most cases to killing of the pathogen. However, while phagocytosis is a crucial element of host defence, aspects of the phagocytic response can also be associated with host tissue damage. I will review our current understanding of the basic mechanisms of phagocytosis, and potential sources of deleterious effects on the host.

Phagocytosis can be divided conceptually into two stages: interaction of particles with phagocytic receptors, leading to particle engulfment; maturation of the resulting phagosome.

Phagocytic receptors

Particles may be recognized by cell surface receptors on phagocytes either directly or indirectly through coating of the particle with opsonins. Examples of direct recognition include phagocytosis mediated by the mannose receptor, which binds to mannans on the surface of pathogens, and phagocytosis via a variety of scavenger receptors [1]. Opsonins that facilitate phagocytosis include the antibody and the C3bi fragment of complement. While several different receptors can mediate phagocytosis, most studies of the basic mechanisms of phagocytosis have focused on Fc receptors (FcR), which recognize the constant Fc portion of antibodies and allow phagocytosis of antibody-coated particles. Phagocytosis via the complement receptor CR3 (CD18/CDllb) has also been analyzed in some detail, but the function of other receptors has been less well characterized. While there. are basic commonalities between the phagocytic processes mediated by different receptors, it has become clear that there are important differences, as will be discussed further below. In many physiologically relevant situations (e.g., during exposure to new pathogens in the lung), phagocytosis is unlikely to be mediated through specific antibody recognition.

52 J.W. Booth

Thus, an important area of future research will be the investigation of the detailed mechanisms of phagocytosis via receptors other than FcR.

A common feature of phagocytosis is that signalling initiated by engagement of phagocytic receptors leads to a rearrangement of the actin cytoskeleton that drives particle uptake. The signalling steps involved in FcR-mediated phagocytosis have been partially elucidated, and will be discussed as a paradigm for phagocytic signalling. The initiating signal is receptor aggregation induced by particle binding. This leads to phosphorylation of specific tyrosine residues in the cytoplasmic domain of the FcR by src family kinases. The resulting phosphotyrosine residues then recruit the tyrosine kinase Syk, which is crucial for FcR-mediated phagocytosis [2]. Events downstream of Syk include dynamic changes in the phosphoinositides phosphatidylinositol [4, 5] bisphosphate and phosphatidylinositol [3,4,5] trisphosphate due to activation and recruitment of PIC 4 )P5 and PI3 kinases. Accumulation of these phosphoinositides is tightly localized to the sites of particle binding (the "phagocytic cup"). They may contribute to phagocytic signalling by recruiting proteins containing phosphoinositide-binding domains to the phagocytic cup. Generation of second messengers from PIP2 via phopholipase C is also likely important [3,4].

Subsequent remodelling of the actin cytoskeleton involves the GTPases CDC42 and Rac [5]. These control recruitment of the Arp2/3 complex, which may be responsible for nucleating actin filament assembly [6]. Localized accumulation of F-actin occurs at the phagocytic cup. Pseudopods are extended around the particle, with tight apposition of the membrane to the particle mediated by a zippering mechanism. Pseudopods eventually fuse around the particle, leading to its engulfment.

Recently, it has become clear that in addition to cytoskeletal changes, membrane remodelling may also play an important role in maintaining the total surface area of the phagocyte and in pseudopod extension. Focal exocytosis of internal vesicles has been observed at sites of phagosome formation; these vesicles derive at least in part from recycling endosomes [7].

Phagocytosis mediated by the complement receptor CR3 is also actin-dependent, but is distinct in several respects from phagocytosis mediated by FcR [1]. Whereas FcR are constitutively active for phagocytosis, CR3 must be activated by prior stimuli (e.g., phorbol esters or cell adhesion) in order to become competent for phagocytosis. CR3-mediated phagocytosis is also morphologically distinctive; rather than involving elaboration of obvious pseudopods, particles appear to "sink" into the cell. This mode of phagocytosis is not dependent on Syk [2], and depends primarily on Rho GTPase, rather than CDC42 and Rac, to orchestrate actin rearrangements [5].

Phagocytosis and Lung Injury 53

Phagosome maturation

After internalization of a particle into a phagosome, macrophages and neutrophils employ a wide variety of mechanisms to kill and/or degrade the phagosomal contents. In macrophages, the phagosome progressively matures into a phagolysosome, acquiring lysosomal proteases and the NADPH oxidase, and undergoing acidification to create an environment hostile to micro-organisms. This maturation occurs via an ordered series of fusion events with endosomal and lysosomal compartments [8]. In neutrophils, specialized granules fuse with the phagosome to deliver components of the killing machinery. Elements ofthe antimicrobial defence in the neutrophil phagosome include defensins, lactoferrin, bacterial permeability increasing protein, proteases such as elastase and proteinase-3, generation of reactive oxygen intermediates by the NADPH oxidase, and generation of highly toxic HOCI by myeloperoxidase.

The factors that regulate the fusion of endomembranes with the phagosome are largely unknown. Fusion with the phagosome in neutrophils, but not in macrophages, appears to depend on cytosolic calcium transients [9, 10]. Vesicle fusion is likely to be regulated by members of the Rab family of GTPases, as Rabs play a key role in regulating a variety of intracellular membrane fusion events. As with the initial internalization step, phosphoinositides also appear to playa role during phagosome maturation. Accumulation of phosphatidylinositol-3-phosphate has been observed on early phagosomes (Viera et aI., unpublished observations), where it may be involved in specifically recruiting elements of the Rab fusion machinery.

Phagosomal maturation appears to be influenced by the type of phagocytic receptor through which particle uptake occurs. This is suggested by studies indicating differential survival of pathogens within macrophages depending on their route of entry. For example, mycobacteria block phagosome maturation when internalized by nonopsonic pathways, but maturation is not blocked if phagocytosis occurs via FcR in the presence of an opsonizing antibody [11]. Another indication of distinct maturation pathways is provided by studies of antigen presentation by macrophages to T cells [12]. The efficiency of presentation of antigens depends on the signalling characteristics of the FcR mediating antigen uptake, presumably due to differential targeting of antigen to compartments for proteolytic processing and loading on class II major histocompatibility complex (MHC) molecules prior to presentation.

Clinical implications of phagocytosis

Two aspects of phagocytosis that may lead to damage to the host will be

54 J.W. Booth

considered: proinflammatory signalling by phagocytes and release from phagocytes of cytotoxic molecules.

Proinflammatory signalling

Phagocytosis can be accompanied by release of proinflammatory molecules from phagocytes, which can contribute to host tissue damage by inducing an overactive inflammatory response. The exact profile of proinflammatory responses appears to depend on the phagocytic receptor engaged. For instance, FcR phagocytosis is accompanied by increased secretion of reactive oxygen intermediates and arachidonic acid metabolites, whereas CR3-mediated phagocytosis is not [1]. Cross-linking of FcR or CR3 can elicit secretion of proinflammatory cytokines. In contrast, phagocytosis of apoptotic cells mediated through a recently identified phosphatidylserine receptor induces an anti-inflammatory state, with suppression of TNF-a production and increased release of TGF-~ [13]. This is likely important to prevent inappropriate activation of proinflammatory responses during routine phagocytosis of apoptotic corpses.

An important recent advance in understanding proinflammatory signalling has been the identification of the Toll-like receptor (TLR) family of proteins [14]. These receptors initiate proinflammatory signalling in response to pathogen associated molecular patterns (PAMPs) - conserved molecular motifs associated with pathogens that are not found in host cells. Ten mammalian TLRs have been identified to date. Different TLRs are involved in responding to different PAMPs, (e.g., TLR4 mediates the response to bacteriallipopolysaccharide (LPS), while TLR5 recognizes bacterial flagellin) [15]. TLRs may in fact operate via combinatorial mechanisms to recognize a wide spectrum of PAMPs. While the TLRs are not phagocytic receptors per se, they are recruited to phagosomes. There, they may serve to sample the contents of the phagosome in order to initiate appropriate responses to different internalized pathogens [16].

Whether phagocytic receptors and TLRs interact with each other directly, and how signals from these receptors intersect to modulate the overall balance of pro- and anti-inflammatory signals remains unclear. As mentioned above, Fc receptor signalling has generally been thought of as proinflammatory in nature. For instance, cross-linking Fey receptors (which bind the Fc portion of IgG) augments the increase in serum TNF-a levels in response to lipopolysacchari des LPS [17]. On the other hand, it was recently reported that ligating Fey receptors can actually reverse macrophage responses to a number of proinflammatory stimuli including LPS [18]. Coligation of FeyR abrogated interleukin (IL-12) production in response to LPS and induced production of high levels


of the anti-inflammatory cytokine IL-10. In vivo, transfer into mice of macrophages whose FcyR had been ligated could rescue the mice from lethal endotoxernia, suggesting a dominant physiological anti-inflammatory role for FcyR signalling. Thus, the interaction between phagocytic and TLR-mediated proinflammatory signalling remains to be clarified. The identification of distinct receptors transducing phagocytic and pro inflammatory signals suggests that it may be possible to downregulate proinflammatory signalling from phagocytes without impairing their phagocytic function. This may allow for modulation of the host response during infection to decrease tissue damage while maintaining clearance of microbes. Furthermore, the experiments described above with FcyR cross-linking raise the possibility of harnessing phagocytic receptors themselves for anti-inflammatory effects.

Release of toxic effectors during phagocytosis

While essential for pathogen killing, the toxic contents of phagocytes carry the potential for destruction of host tissues. In particular, neutrophil products including elastase and reactive oxygen species are thought to be important in lung injury [19]. Presumably, the principal function of these cytotoxic products is to be delivered to the phagosome. How then, are they released into the extracellular space, where damage to host tissue can ensue? At least two possible mechanisms can be considered: cell lysis and secretion.

Normally, neutrophils are short-lived cells. At the end of their lifespan they undergo apoptosis and are cleared by phagocytosis by macrophages. This clearance prevents release of the neutrophil's toxic contents into the extracellular milieu. A defect in this process could in principle lead to neutrophils undergoing secondary necrosis and lysis, with concomitant release of their contents. Thus, release of neutrophil contents in vivo may occur as an indirect result of deficiencies in phagocytosis by macrophages [20].

Alternatively, neutrophil contents can be secreted directly by fusion of granules with the plasma membrane. Granule secretion is induced during neutrophil activation. Neutrophils possess at least four different types of granules that differ both in their contents and their susceptibility to release upon cell activation. In order of decreasing tendency to be mobilized, there are secretory vesicles, gelatinase (or tertiary) granules, specific (or secondary) granules, and azurophil (or primary) granules. Mobilization of secretory and gelatinase granules facilitates neutrophil migration into and within the perivascular tissue. Specific granules contain lactoferrin and other antimicrobial effectors, while azurophil granules contain highly active antimicrobial factors

56 J.W. Booth

such as myeloperoxidase and elastase. The contents of the azurophil granules, which carry the highest potential for tissue damage, are only minimally released into the extracellular space during stimulation of neutrophils by soluble proinflammatory stimuli [21]. However, they are released to a significant degree during phagocytosis [22]. This seems to be a consequence of very rapid targeting of granules to sites of particle binding, such that fusion of granules with the region of membrane that will pinch off to become the phagosome begins before the phagosome has actually closed [23, 24]. This allows contents of the granule to be released to the extracellular space. The extent of this release may well be increased in situations where phagosome sealing is delayed or prevented (e.g., if cytoskeletal remodelling in the neutrophil is blocked by the action of bacterial toxins, or during attempted phagocytosis of large particles). Thus, during periods of active phagocytosis in sepsis this release may be sufficiently extensive to cause significant host damage.

The signals involved in regulating fusion of granules with the plasma membrane and phagosome remain to be identified. Numerous pathogens have evolved ways to arrest or alter phagosome maturation in order to survive within macrophages (e.g., Mycobacterium, Salmonella and Legionella spp.) Similarly, in neutrophils, fusion of azurophil granules with both the plasma membrane and the phagosome is blocked during phagocytosis of mycobacteria [22]. Uncovering the mechanisms underlying this blockade may shed light on the normal mechanism of phagosome maturation and may suggest targets for therapeutic intervention. If the fusion events leading to release of azurophil granule contents could be specifically retarded without impairing the overall extent of phagocytic uptake, it might be possible to minimize the release of cytotoxic effectors without having an overall negative impact on pathogen clearance. This could provide a useful approach for minimizing tissue damage during sepsis.

References 1. Aderem A, Underhill DM (1999) Mechanisms of phagocytosis in macrophages. Annu Rev

ImmunoI17:593-623 2. Kiefer F, Brumell J, AI-Alawi N et al (1998) The Syk protein tyrosine kinase is essential for

Fcgamma receptor signaling in macrophages and neutrophils. Mol Cell Bioi 18:4209-20 3. Botelho RJ, Teruel M, Dierckman R et al (2000) Localized biphasic changes in phosphatidyli

nositol-4,5-bisphosphate at sites of phagocytosis. J Cell BioI 151: 1353-1368 4. Marshall JO, Booth JW, Stambolic V et al (2001) Restricted accumulation of phosphatidylino

sitol3-kinase products in a plasmalemmal subdomain during Fc receptor-mediated phagocytosis. J Cell BioI (in press)

5. Caron E, Hall A (1998) Identification of two distinct mechanisms of phagocytosis controlled by different Rho OTPases. Science 282: 1717 -1721.


6. May RC, Caron E, Hall A et al (2000) Involvement of the Arp2/3 complex in phagocytosis mediated by FcgammaR or CR3. Nat Cell BioI 2:246-248

7. Bajno L, Peng XR, Schreiber AD et al (2000) Focal exocytosis of VAMP3-containing vesicles at sites of phagosome formation. J Cell BioI 149:697-706

8. Tjelle TE, Lovdal T, Berg T (2000) Phagosome dynamics and function. Bioessays 22:255-263 9. Jaconi ME, Lew DP, Carpentier JL et al (1990) Cytosolic free calcium elevation mediates the

phagosome-lysosome fusion during phagocytosis in human neutrophils. J Cell BioI 110:1555--1564

10. Zimmerli S, Majeed M, Gustavsson M et al (1996) Phagosome-lysosome fusion is a calciumindependent event in macrophages. J Cell Bioi 132:49-61

11. Malik ZA, Denning GM, Kusner DJ (2000) Inhibition of Ca(2+) signaling by Mycobacterium tuberculosis is associated with reduced phagosome-lysosome fusion and increased survival within human macrophages. J Exp Med 191:287-302

12. Shen L, van Egmond M, Siemasko K et al (2001) Presentation of ovalbumin internalized via the immunoglobulin-A Fc receptor is enhanced through Fc receptor gamma-chain signaling. Blood 97:205-213

13. Fadok VA, Bratton DL, Rose DM et al (2000) A receptor for phosphatidylserine-specific clearance of apoptotic cells. Nature 405:85-90

14. Aderem A, Ulevitch RJ (2000) Toll-like receptors in the induction of the innate immune response. Nature 406:782-787

15. Hayashi F, Smith KD, Ozinsky A et al (2001) The innate immune response to bacterial flagellin is mediated by Toll-like receptor 5. Nature 410: 1099-1103

16. Ozinsky A, Underhill DM, Fontenot JD et al (2000) The repertoire for pattern recognition of pathogens by the innate immune system is defined by cooperation between toll-like receptors. Proc Natl Acad Sci USA 97:13766-13771

17. Refici ML, Metzger DW, Arulanandam BP et al (2001) Fcgamma-receptor signaling augments the LPS-stimulated increase in serum tumor necrosis factor-alpha levels. Am J Physiol Regul Integr Comp Physiol 280:Rl 037-1044

18. Gerber JS, Mosser DM (2001) Reversing lipopolysaccharide toxicity by ligating the macrophage fcgamma receptors. J Immunol 166:6861-6868

19. Lee WL, Downey GP (2001) Neutrophil activation and acute lung injury. Curr Opin Crit Care 7:1-7

20. Haslett C (1999) Granulocyte apoptosis and its role in the resolution and control of lung inflammation. Am J Respir Crit Care Med 160:S5-11

21. Sengelov H, Kjeldsen L,Borregaard N (1993) Control of exocytosis in early neutrophil activation. J ImmunolI50:1535-1543

22. N'Diaye EN, Darzacq X, Astarie-Dequeker C et al (1998) Fusion of azurophil granules with phagosomes and activation of the tyrosine kinase Hck are specifically inhibited during phagocytosis of mycobacteria by human neutrophils. J Immunol 161 :4983-4991

23. Tapper H, Grinstein S (1997) Fc receptor-triggered insertion of secretory granules into the plasma membrane of human neutrophils: selective retrieval during phagocytosis. J Immunol 159:409 -418

24. Suzaki E, Kobayashi H, Kodama Y et al (1997) Video-rate dynamics of exocytotic events associated with phagocytosis in neutrophils. Cell Motil Cytoskeleton 38:215-228

Dual Role of Neutrophil a-Defensins in Lung Inflammation

H.ZHANG

The acute respiratory distress syndrome (ARDS) was first described some 3 decades ago [1] as a syndrome of acute respiratory failure complicated by multiple organ failure [2]. The pathophysiological changes of ARDS, and its less-severe form the acute lung injury (ALI), are fairly well documented. There is a breakdown in the endothelial and epithelial barrier and gas exchange function in the lung, with neutrophil migration and sequestration [3]. A number of mediators, including elastase [4], arachidonic acid metabolities [5], reactive oxygen species [6, 7], and cytokines [8], have been implicated as important in ARDSI ALI. Unfortunately, clinical trials using therapy aimed at these mediators have failed to demonstrate benefit in patients with inflammatory lung conditions [2]. A meta-analysis of 101 studies of ARDS/ALI found no reduction in mortality over time [9], although some suggest that mortality has decreased [10, 11]. Our know ledge of the condition is clearly incomplete and future studies are needed to characterize the mediators of the disease process more completely, and to find a more-suitable model for mechanistic studies oflung injury. In this article, I will briefly introduce the role of a relatively new molecule human neutrophil peptide, also known as defensin, in lung inflammation. This review provides some evidence that defensins exert dual actions in host defense against infection and cause lung injury, depending on local concentrations. Defensins may therefore play an important role in modulating lung injury.

Defensins - overview

Antimicrobial peptides that contain six cysteines have been classified as defensins [12]. Defensins form at least three structural groups whose evolutionary relationship is uncertain: the "classical" a-defensins, the ~-defensins, and insect defensins. a-Defensins comprise 6 members [13, 14], four human neutrophil peptides, (HNP)-I, -2, -3 and -4 [15], are located in the azurophilic

60 H. Zhang

granules ofthe neutrophils, and two human defensins, (HD)-5 and HD-6, are present in the secretory granules of the intestinal Paneth' s cells and in epithelial cells of the female reproductive tract [16]. Human ~-defensin-l (hBD-l) is located in epithelial cells of various organs [17-19], and hBD-2 in psoriatic scales [20]. Our studies are concerned specifically with the neutrophil defensins and therefore HD-5-6 and hBD-I-2 will not be discussed further. The present discussion will focus on a-defensins (HNP-I-3), since these account for almost 99% of the total defensin content of neutrophils [12, 21]. In humans, these defensins constitute up to 5% of the total protein content of mature neutrophils and > 50% of the total protein within the azurophilic granules. The composition of azurophilic granule proteins is approximately (in ng/l0 [6] neutrophils): elastase - 1,500; cathepsin G - 2,500; proteinase 3 - 1,000; and defensins - 6,000 [22].

Microbicidal activity of defensins

To date, most defensin studies have been performed in media conducive to microbial growth, typically in low-salt, low-ionic strength media, to test their microbicidal activity. It is known that defensins are active against gram-positive and gram-negative bacteria [23], fungi [24, 25], and herpes simplex virus in vitro [26]. When tested in vitro at concentrations between 10 and 100 mg/ml, purified defensins killed a wide variety of bacteria by permeabilizing both outer and inner lipid membranes of bacteria in a charge- or voltage-dependent manner [13, 26].

We investigated the effect of defensins and lung tissue together on bacterial killing in physiological media. In the presence of cultured lung tissue, the maximal killing capacity of defensins was up to 1,000-fold greater than in its absence, indicating that the antibacterial activity is further enhanced by lung tissue. Theoretically, the lower bacterial number seen in the culture supernatant might be due to an increased bacterial adherence to lung tissue in the presence of defensins. We excluded this possibility by showing that the number of bacteria recovered from the lung homogenates was actually slightly lower in the defensin-treated group than in the Escherichia coli alone group. Our study thus demonstrated that under physiological conditions cationic defensins kill bacteria not only by the well-described mechanisms of increasing membrane permeability and cell lysis directly [12], but also indirectly by producing bactericidal products from lung tissue.

To identify underlying possible mechanisms, we examined the effect of defensins on oxygen burst in the lung, and found that defensins directly induce production of hydrogen peroxide by lung tissue. The generation of oxidant

Dual Role of Neutrophil a-Defensins in Lung Inflammation 61

products by lung tissue following defensin stimulation may play an important role in host defense. It has been shown that hydrogen peroxide dramatically reduces the growth rate of E. coli in culture without cytotoxicity to cultured fibroblasts [27]. There are a number of ways in which hydrogen peroxide may inhibit bacterial growth. For example, bacteria may react to oxidative stress by invoking two distinct oxidant responses, the peroxide stimulon and the superoxide stimulon. The two stimulons each contain genes constituting the OxyR or SoxRS regulon, respectively [28]. Activation of these genes inhibits cell division [27]. In addition, hydrogen peroxide induces DNA damage in E. coli mediated by a Fenton reaction that generates hydroxyl radicals from hydrogen peroxide [29].

To determine the role of oxidant mechanisms in the killing of E. coli by defensins, the effect of diphenyleneiodonium (DPI) on production of hydrogen peroxide and bacterial count was measured. DPI inhibits the NADPH oxidase acting on the flavoprotein in blocking the sequence ofNADPH ~ FAD protein ~ cytochrome b ~ reactive oxygen species, including superoxide, hydrogen peroxide, and hypochlorous acid production [30, 31]. Our data show that the rate of E. coli killing by defensins was reduced by 2,500-fold in the presence of DPI. Our study clearly demonstrates that the generation of oxidants induced by defensins contributes to an enhanced bacterial killing in the lung.

Cytotoxic effects of defensins

To date, all cytotoxic studies of defensins have been conducted using in vitro systems. Defensins are cytotoxic to tumor targets in a concentration- and time-dependent fashion. Optimal lysis was achieved with 25-100 ~g/ml after 6 h in various human and murine tumour cell lines [32]. This effect is not tumor specific, however. Comparable concentrations of human defensins are also cytotoxic to normal mammalian cells [13,25,32], endothelial cells [33], murine thymocytes, and spleen cells [32] in vitro. Defensins may also impair the phagocytic functions of neutrophils [34-36]. Of particular relevance to the present discussion, defensins have been shown to be cytotoxic to the A549 airway epithelial cell line in vitro [33, 37].

It is unclear whether the in vitro cytotoxic data are relevant to clinical settings. Several investigators have measured concentrations of defensins in body fluids of critically ill patients. Ihi et al. [38] demonstrated defensin concentrations in bronchoalveolar lavage (BAL) fluid of patients with bacterial pneumonia (2.0 ±0.9 mg/ml (mean ±SE) that were 5 orders of magnitude greater than those measured in normals (0.000016 ± 0.000015 mg/ml). Pleural

62 H.Zhang

fluid of patients with empyema had defensin concentrations of 13.3 ± 1.9 mg/ml, and cerebrospinal fluid of patients with bacterial meningitis had values of 3.4 ± 1.2 mg/mI. High sputum defensin levels ranging from 0.3 to> 1.6 mg/mI (the upper detection limit in the study) have been reported in patients with cystic fibrosis [39]. Patients with meningitis had plasma defensin levels that were extremely high, ranging from 0.12 Jlg/mI to 170 Jlg/ml, compared with a mean concentration of 0.042 Jlg/ml in healthy blood donors [40]. The mean plasma concentrations of defensins in patients at the onset of bacterial infection, non-bacterial infection, and pulmonary tuberculosis were 4.2, 3.2, and 1.8 times the means for healthy volunteers [38]. Based on the measured amounts of defensins (3-5 !J.g/106 neutrophils) and the known number of defensin-containing granules (I,OOO/neutrophil), high concentrations of neutrophil defensins (1-10 mg/ml) are likely to exist in phagocytic vacuoles containing ingested microbes [12].

Thus high defensin level occurs in several lung disease studies. However, it is not known whether the high concentrations of defensins cause lung injury in vivo. We investigated the direct effect of a wide range of concentrations of purified defensins on the lung of mice in vivo [41]. Intratracheal instillation of defensins, from 5 mg/kg to 30 mg/kg, induced a reduction in oxygen hemoglobin saturation (Sa02) in a dose- and time-dependent manner. Defensins (15 mg/kg) increased lung permeability by threefold estimated by Evans blue dye technique [41], and dose-dependently enhanced lung mitochondrial cytochrome c content [41], a marker of mitochondrial dysfunction and caspase activation [42]. Defensins also increased total cell number in the BAL, particularly increasing the neutrophil population [41]. Taken together, these data suggest that high concentrations of defensins may initiate an inflammatory response and lung injury in addition to their microbicidal activity.

The involvement ofTNF-a in various models of ALI induced by sepsis, acid aspiration, or mechanical ventilation has been well documented, suggesting that cytokines mediate the initiation and maintenance of inflammatory lesions. We measured TNF-a concentrations in BAL fluids and plasma samples from defensin-treated mice. The increased release ofTNF-a found in BAL fluid may reflect the initiation of the lung inflammatory cytokine network, since TNF-a is considered as an early, central inflammatory cytokine. TNF-a itself can also directly increase the permeability of lung endothelial and epithelial barriers.

Conclusions

Defensin levels in body fluids of critically ill patients, including those with


ARDS and sepsis, are elevated [38-40, 43]. Our work explores the novel concept that defensins might be harmful and cause lung injury. The toxicity of defensins to cultured lung epithelial cells [33, 37], the inhibition of neutrophil phagocytic activity by defensins [34-36], and our own in vivo data support this concept. Since defensins have been proposed as a potential antimicrobial therapy [12], it is clearly essential to clarify whether they also have deleterious effects. Furthermore, if defensins are detrimental to lung tissue in high concentrations, mechanisms to inhibit defensins may playa therapeutic role in various inflammatory diseases.

References 1. Ashbaugh DG, Bigelow DB, Petty TL (1967) Acute respiratory distress in adults. Lancet II:

319-323 2. Abraham E, Matthay MA, Dinarello CA, et al (2000) Consensus conference definitions for

sepsis, septic shock, acute lung injury, and acute respiratory distress syndrome: time for a reevaluation. Crit Care Med 28:232-235

3. Weiland JE, Davis WB, Holter JF, et al (1986) Lung neutrophils in the adult respiratory distress syndrome: clinical and pathophysiologic significance. Am Rev Respir Dis 133:218-225

4. Miyazaki Y, Inoue T, Kyi M, et al (1998) Effects of a neutrophil elastase inhibitor (ONO-5046) on acute pulmonary injury induced by tumor necrosis factor alpha (TNFalpha) and activated neutrophils in isolated perfused rabbit lungs. Am J Respir Crit Care Med 157:89-94

5. Abraham E, Baughman R, Fletcher E, et al (1999) Liposomal prostaglandin El (TLC C-53) in acute respiratory distress syndrome: a controlled, randomized, double-blind, multicenter clinical trial. TLC C-53 ARDS Study Group. Crit Care Med 27:1478-1485

6. Blackwell TS, Blackwell TR, Holden EP, et al (1996) In vivo antioxidant treatment suppresses nuclear factor-kappa B activation and neutrophilic lung inflammation. J Immunol 157: 1630- -1637

7. Zhang H, Spapen H, Manikis P, et al (1995) Tirilazad mesylate (U74006F) improves systemic and splanchnic oxygen extraction capabilities following endotoxic shock. Am J Physiol 268:H1847-H1855

8. De Backer WA, Amsel B, Jorens PG, et al (1996) N-acetylcysteine pretreatment of cardiac surgery patients influences plasma neutrophil elastase and neutrophil influx in bronchoalveolar lavage fluid. Intensive Care Med 22:900-908

9. Krafft P, Fridrich P, Pemerstorfer T, et al (1996) The acute respiratory distress syndrome: definitions, severity and clinical outcome. An analysis of 101 clinical investigations. Intensive Care Med 22:519-529

10. Hudson LD, Milberg JA, Anardi D, Maunder RJ (1995) Clinical risks for development of the acute respiratory distress syndrome. Am J Respir Crit Care Med 151 :293-301

11. Milberg JA, Davis DR, Steinberg KP, Hudson LD (1995) Improved survival of patients with acute respiratory distress syndrome (ARDS): 1983-1993. JAMA 273:306-309

12. Ganz T Lehrer RI (1994) Defensins. CUff Opin ImmunoI6:584-589 13. Lehrer RI, Lichtenstein AK, Ganz T (1993) Defensins: antimicrobial and cytotoxic peptides of

mammalian cells. Annu Rev Immunol 11: 105-128 14. Martin E, Ganz T, Lehrer RI (1995) Defensins and other endogenous peptide antibiotics of

vertebrates. J Leukoc Bioi 58: 128-136 15. Selsted ME, Brown DM, DeLange RJ, et al (1985) Primary structures of six antimicrobial

peptides of rabbit peritoneal neutrophils. J BioI Chern 260:4579-4584

64 H.Zhang

16. Quayle AJ, Porter EM, Nussbaum AA, et al (1998) Gene expression, immunolocalization, and secretion of human defensin-5 in human female reproductive tract. Am J Patho1152: 1247-1258

17. Becker MN, Diamond G, Verghese MW, Randell SH (2000) CD14-dependent LPS-induced {beta}-defensin-2 expression in human tracheobronchial epithelium. J Bioi Chern

18. Goldman MJ, Anderson GM, Stolzenberg ED, et al (1997) Human beta-defensin-l is a salt-sensitive antibiotic in lung that is inactivated in cystic fibrosis. Cell 88:553-560

19. Singh PK, Jia HP, Wiles K, et al (1998) Production of beta-de fens ins by human airway epithelia. Proc Nat! Acad Sci USA 95:14961-14966

20. Harder J, Bartels J, Christophers E, Schroder JM (1997) A peptide antibiotic from human skin. Nature 387:861

21. Selsted ME, Harwig SS, Ganz T, et al (1985) Primary structures of three human neutrophil defensins. J Clin Invest 76:1436-1439

22. Gabay JE, Almeida RP (1993) Antibiotic peptides and serine protease homologs in human polymorphonuclear leukocytes: defensins and azurocidin. Curr Opin Immunol 5 :97 -102

23. Selsted ME, Szklarek D, Lehrer RI (1984) Purification and antibacterial activity of antimicrobial peptides of rabbit granulocytes. Infect Immun 45: 150-154

24. Lehrer RI, Daher K, Ganz T, Selsted ME (1985) Direct inactivation of viruses by MCP-l and MCP-2, natural peptide antibiotics from rabbit leukocytes. J ViroI54:467-472

25. Lehrer RI, Szklarek D, Ganz T, Selsted ME (1985) Correlation of binding of rabbit granulocyte peptides to Candida albicans with candidacidal activity. Infect Immun 49:207-211

26. Cociancich S, Ghazi A, Hetru C, et al (1993) Insect defensin, an inducible antibacterial peptide, forms voltage-dependent channels in Micrococcus luteus. J Bioi Chem 268: 19239-19245

27. Hyslop PA, Hinshaw DB, Scraufstatter IU, et al (1995) Hydrogen peroxide as a potent bacteriostatic antibiotic: implications for host defense. Free Radic Bioi Med 19:31-37

28. Farr SB, Kogoma T (1991) Oxidative stress responses in Escherichia coli and Salmonella typhimurium. Microbiol Rev 55:585-561

29. Imlay JA, Linn S. (1988) DNA damage and oxygen radical toxicity. Science 240:1302-1309 30. Ellis JA, Mayer SJ, Jones OT (1988) The effect of the NADPH oxidase inhibitor diphenyleneio

donium on aerobic and anaerobic microbicidal activities of human neutrophils. Biochem J 251:887-891

31. Hampton MB, Winterboum CC (1995) Modification of neutrophil oxidant production with diphenyleneiodonium and its effect on bacterial killing. Free Radic Bioi Med 18:633-639

32. Lichtenstein A (1991) Mechanism of mammalian cell lysis mediated by peptide defensins. Evidence for an initial alteration ofthe plasma membrane. J Clin Invest 88:93-100

33. Okrent DG, Lichtenstein AK, Ganz T (1990) Direct cytotoxicity of polymorphonuclear leukocytes granule proteins to human lung-derived cells and endothelial cells. Am Rev Respir Dis 141: 179-185

34. Kaplan SS, Basford RE, Jeong MH, Simmons RL (1996) Biomaterial-neutrophil interactions: dysregulation of oxidative functions of fresh neutrophils induced by prior neutrophil-biomaterial interaction. J Biomed Mater Res 30:67-75

35. Kaplan SS, Heine RP, Simmons RL (1999) Defensins impair phagocytic killing by neutrophils in biomaterial-related infection. Infect Immun 67: 1640-1645

36. Yomogida S, Nagaoka I, Saito K, Yamashita T (1996) Evaluation of the effects of defensins on neutrophil functions. Inflamm Res 45:62-67

37. Wetering S van, Mannesse-Lazeroms SPG, Van Sterkenburg MAJA, et al (1997) Effect of defensins on interleukin-8 synthesis in airway epithelial cells. Am J PhysioI272:L888-L896

38. Ihi T, Nakazato M, Mukae H, Matsukura S (1997) Elevated concentrations of human neutrophil peptides in plasma, blood, and body fluids from patients with infections. Clin Infec Dis 25:1134-1140

39. Soong LB, Ganz T, Ellison A, Caughey GH (1997) Purification and characterization of defensins from cystic fibrosis sputum. Inflamm Res 46:98-102

40. Panyutich AV, Panyutich EA, Krapivin VA, et al (1993) Plasma defensin concentrations are elevated in patients with septicemia or bacterial meningitis. J Lab Clin Med 122:202-207


41. Zhang H, Porro G, Orzech N, et al (2001) Neutrophil defensins mediate acute inflammatory response and lung dysfunction in dose-related fashion. Am J Physiol Lung Cell Mol Physiol 280:L947-L954.

42. Thress K, Kornbluth S, Smith JJ (1999) Mitochondria at the crossroad of apoptotic cell death. J Bioenerg Biomembr 31 :321-326

43. Ashitani J, Mukae H, Thiboshi H, et al (1996) Defensin in plasma and in bronchoalveolar lavage fluid from patients with acute respiratory distress syndrome. Nippon Kyobu Shikkan Gakkai Zasshi 34:1349-1353

Epithelial Injury in Sepsis and ARDS: Role of Leukocyte-Derived Proteases

H. GINZBERG, C. CHUNG-WAI, G.P. DOWNEY

Research in the last 30 years has demonstrated that acute lung injury and sepsis are largely inflammatory diseases [1, 12]. This is correct whether or not the primary initiating factor directly induces acute inflammation (sepsis, disseminated intravascular coagulation, or pancreatitis) or whether there is direct injury to the lungs (hyperoxia or chemical injury). In the latter circumstances, although the initial injury may involve direct cellular cytotoxicity, the ensuing inflammatory response appears to be responsible for much of the morbidity and mortality associated with these entities [3, 4].

In affected organs, induding the lung, heart, kidneys, gut, and liver, large numbers of leukocytes, primarily neutrophils, are sequestered within the microvasculature [5]. Importantly, leukocytes are adherent to the endothelium and present within the interstitium of these organs. In organs such as the lung, gut, and kidney where there is an epithelium, neutrophils transmigrate through the epithelium into the lumen of the organ (e.g., alveolar space of the lung, lumen of the gut, and tubules of the kidney). In these areas, there is widespread evidence of organ injury, induding endothelial and epithelial damage and loss of the barrier function of the epithelium, with increased permeability to fluid and proteins. In the lungs, this loss of the barrier function is manifest by leakage of protein-rich debris and red cells into the alveolar space, indicating a compromise of the alveolar-capillary membrane. Ultrastructural studies have provided evidence of epithelial loss (denudation) and demonstrated that there are large areas of the basement membrane devoid of epithelial cells, and therefore in direct contact with the alveolar space [5].

As a result of extensive studies from many laboratories, a model for the pathogenesis of inflammatory organ injury has been proposed (Fig. 1). Circulating leukocytes become activated by exposure to a variety of soluble or surface-bound factors, such as lipopolysaccharide (LPS), cytokines (IL-l, TNF), chemokines (lL-8, MIP-2), complement fragments (C5a), lipid mediators (PAF, LTB4), and circulating dotting factors, leading to their sequestration

68 H. Ginzberg, C. Chung-Wai, G. P. Downey

/ Init iat ing Event

Sepsis

Decreased deformability Inc reased adhesiveness

Circulat ng inflammat ory Media ors LPS

L-Select in CD11 blCD18 ? others

• Cytokinesl Chemokines (TNF, IL-8 , IL-1) • lipid Med iators (PAF, LT94)

,---------------------, MICROVASCULAR SEQUESTRATtON

Adhesion and emigration from the vasculat ure Transmigrat ion through epithelium in lung, gut, kid ney Exposure to endo th elial, int ersU lal, and epithelia l ligands

• Defective negat r1e fee~back (signal terminat ion)

I Endothelial and Epithelial Alterat ions/Injury I Disruption of junctional complex (ad herens a nd tight junc t ions) especialt)l of epit helium

• loss of barrier funct io n loss of pump funct ion surfac t ant alterations

~ Expo sure of PM N toad dit ional st imuti

cyt okinesl chemokines pro duced by mac ro phages

Lipid Mediat ors (Ieukotrienes, PAF)

~ Further endot helial and epithelial Injury

~ Acut e Organ Injury

• Lung

GI tract Kidney • Heart

Fig. 1. Model for the pathogenesis of inflammatory lung injury Circulating leukocytes such as neutrophils are sequestered in the microvasculature of the lung and other organs where they adhere to endothelial cells and leave the vascular space (diapedesis). In organs such as the lung, kidney, and gastrointenstinal (GI) tract, leukocytes also transmigrate across the epithelium in a baso-Iateral to apical direction. Leukocyte-derived cytotoxic agents including proteolytic enzymes and oxidants are released in these locations and can injure host cells, leading to organ injury (LPS = lipopolysaccharide, TNF = tumor necrosis factor, IL-8 = interlukin - 8, PAF = platelet-activating factor, LTB4 leukotriene B4)

Epithelial Injury in Sepsis and ARDS: Role of Leukocyte-Derived Proteases 69

in the microvasculature of many organs [6]. The initial consequences of leukocyte activation include a change in their biomechanical properties that is responsible in part for their sequestration within the capillaries and post-capillary venules of the microvasculature [6, 7]. Soon thereafter, adhesive factors take over leading to an increase in the surface expression and affinity of leukocyte integrins. These interact with cognate receptors on the epithelium, primarily ICAM-l, leading to firm adhesion, spreading, and eventually diapedesis of the leukocytes out of the vascular space [8]. In the presence of a chemotactic gradient, the leukocytes move through the interstitium and basement membrane of the epithelium, between the epithelial cells, and finally into the alveolar space of the lung. In the gut, kidney and liver, an analogous series of events occur.

It is important to note that leukocytes do not cause damage while circulating within the bloodstream. Rather, they do so when they are adherent to either endothelial or epithelial cells, or within the confines of the interstitial tissues in contact with connective tissues [3]. In these locations they can be induced to release a variety of cytotoxic compounds, including proteases, cationic proteins, and reactive oxygen and nitrogen species. These products are primarily designed to be microbicidal, but when released in an unregulated manner and into the extracellular space, can cause damage to vicinal cells and tissues. An important point is that in host defense, micro-organisms are usually taken up by the phagocytic cell and encapsulated in membrane-bound compartments, a process termed phagocytosis [9]. This allows the delivery of concentrated cytotoxic compounds to the phagosome and enables the leukocyte to kill the micro-organisms within the confined space, but without being released outside of the phagocyte. Thus, the damage is contained. However, under circumstances where leukocytes are activated in an unregulated manner, these cytotoxic compounds can be released outside the cell and cause damage to host tissues.

Another important concept that has emerged is that epithelial cells of the lung, gut, and kidney are principal targets of leukocyte-derived cytotoxic products, and that interference with their functions (e.g., barrier and transport functions) is responsible for many of the clinical manifestations of sepsis and acute lung injury [10, 11]. In the case of the lung epithelial cells, research has demonstrated that certain epithelial functions are essential for the normal homeostasis of the lung [11]. Importantly, interference with these homeostatic functions, such as fluid and ion transport, are responsible in part for the dysfunction of the lung. Moreover, recovery of these functions determines to a large extent whether or not there is resolution of the pulmonary edema and recovery of the gas exchange function of the lung.

An important (but often overlooked) function of pulmonary epithelia is their


role in host defense [12]. This is mediated in part through production of surfactant and associated proteins, members of the 'collectin' family that can function as opsonins, and in part through direct bacterial killing via production of bactericidal peptides and enzymes [13]. These functions are essential to the pivotal role of the epithelium in the innate immune system. Interference with these host defense functions is a major determinant of the increased susceptibility of the lung to infection (bronchitis and pneumonia) in patients with acute lung injury and sepsis.

In the remainder of this article, three issues will be specifically addressed: (1) whether epithelial injury occurs during neutrophil transmigration and activation, (2) what are the potential mediators of this injury, and (3) what are the mechanisms of this injury. Each of these questions will be addressed with specific focus on acute lung injury and sepsis.

Epithelial injury during leukocyte transmigration

Epithelial cells line the alveolar spaces of the lung, the tubules of the kidney, and the sinusoids of the liver. They make contact with adjacent epithelial cells (intraepithelial cell junctions), as well as with the underlying basement membrane (Fig. 2). In both of these areas, specific molecules mediate cell-cell or cell-substratum interactions. On their lateral border, epithelial cells interact with adjacent epithelial cells to form specific junctional complexes. The most apical of these are the tight junctions comprised of a series of proteins that include occludin and ZO-1 and ZO-2. These tight junctions are responsible for selective ionic permeability of the epithelium [14]. Immediately basal to the tight junctions are the adherens junctions, comprised of the transmembrane protein E-cadherin and associated proteins, including the catenins [15]. Basal to the adherens junctions are the gap junctions [16]. On the basolateral surface linking the epithelium to the underlying connective tissues are the intergrins that form focal adhesions [17]. An important point to bear in mind is that during transepithelial migration of leukocytes from the basolateral to the apical surface of the epithelial cells, leukocytes must disengage or otherwise disrupt these inter-epithelial junctions. This might contribute to epithelial injury and dysfunction. Moreover, it is possible that the inter-epithelial junctional proteins might, by providing signals to the leukocytes, actually modulate leukocyte trafficking in these areas.


Basement m embrane

Cytokines Chemokines

Activated neutrophil

Endothelial DysfunctIOn

/

Epithelial Dysfunc tio n .........

o Neutrophil granules

Pulmonary capi llary tumen

Endothelial cells

• o

o o

00

r-· .. _···_····

Leukocyte Elastase

1. Beneficial .Anh·mlcrobi~ (gram negatIVe ba ctena)

• especially intracell lAar·granule associated elastase

• (?) ReqlJred for cell motl ~ty

• e racecllularoncluding membrane bound elastase

, 2. Detrimental

· • CytotoXiC 10 endolhellal eels

· • Degradall lXl Of cadhenns

· • Increased alveolar-caprlary permeability

, • (?) Reg ulall lXl of wound healing

• (?) Modulahon of inllammalory response

Acute Lung Injury

Fig. 2. Epithelial cell-cell and cell-substratum interactions. Epithelial cells form junctions with other epithelial cells along their lateral borders (tight, adherens, and gap junctions) and to the substratum (focal adhesions)

In vitro model system

Our laboratory has chosen to study these mechanisms in part using an in vitro system employing epithelial cells cultured on the underside of semipermable membranes (Fig. 3). Neutrophils can be added to the upper chamber (baso-Iateral aspect of the epithelial cells) and induced to transmigrate across the epithelium by placement of a chemoattractant in the lower chamber. We have used a variety of epithelial cells, including T84, CAC02, A549, MDCK, and primary cultures of respiratory epithelial cells, with similar results. When neutrophils are induced to transmigrate across the epithelium by the chemoattractant n-formyl-methanyl-Ieucyl-phenylalanine (fMLP), by 20-30 min isolated leukocytes are observed to insinuate themselves between the epithelial cells. Subsequent leukocytes appear to follow the tracks of these "scout" cells, and by 60-90 min, groups or clusters of leukocytes are observed within the epithelium monolayer. At this time, epithelial cells are observed to detach from the monolayer and are present in the bottom chamber in association with clusters


Apical Surface (Lumen)

• ,

\1 • •

Focal adhe sions comprised of epithelia! integrins

• TIght junc lions

tAdherens junctions

o Gap junctions

, • • • • • •

&ooment membrane and extracellular matrix

Fig. 3. Diagram of model system to study the effects of neutrophil transmigration through epithelial monolayers. Epithelial cells are grown on the underside of semi-permeable filters and the inserts are placed in tissue culture plates. Purified human neutrophils are added to the basolateral surface of the mono layers and a chemoattractant is added to the apical side. Neutrophil transmigration is then allowed to occur and the effects on the epithelial cells studied

of neutrophils. The events in this in vitro system thus reflect events in humans with sepsis and acute lung injury based on histological studies of the lung, gut, liver, and kidney where epithelial denudation is a prominent feature. Neutrophil transmigration in this experimental system is largely complete by 6-12 h. At this time, circular defects are present in the monolayer representing areas where epithelial cells have become detached from the filter. These circular areas have been termed areas of 'microinjury' .

Neutrophil transmigration results in alterations in inter-epithelial junctions

During the process of leukocyte transmigration, the inter-epithelial junctions are altered. Several lines of evidence support this assertion. First, during


leukocyte transmigration there is a rapid decrease in the transepithelial electrical resistance. In the case of T84 mono layers, the baseline transepithelial resistance is feater than 1,200 ohm/cm2 and falls rapidly to between 50 and 100 ohm/cm at early stages of neutrophil transmigration. The nadir of this change occurs at approximately 20 min. The transepithelial resistance slowly recovers as the monolayer repairs itself, but can take up to 36 h before it returns to baseline. Both the magnitude of the fall and the rate of recovery are influenced by the number of neutrophils that are induced to transmigrate; the larger the number of neutrophils, the more rapid the fall and the slower the recovery of the transepithelial resistance.

In addition to these functional changes that likely reflect alterations in the tight junctions, there is evidence that there is physical disruption of the adherens junction complexes. In the areas where clusters of neutrophils are observed in the monolayer, there are areas of discontinuity of the adherens junctional proteins. This could be due either to a redistribution of the adherens junction proteins or to their loss due to degradation. Western blot analysis of epithelial proteins was unable to demonstrate a significant loss of total immunoreactive E-cadherin or B-catinen proteins. However, because this is a relatively insensitive method, we investigated whether there could be limited proteolytic degradation of the junctional proteins by collecting and concentrating the apical supernatant from the epithelial monolayers. Proteins were analyzed by SDS-PAGE followed by western blotting. After neutrophil transmigration, a 23 kilodalton degradation fragment of E-cadherin was released into the supernatant. This represents part of the extracellular domain of the E-cadherin that is presumably cleaved by proteolytic activity during leukocyte transmigration [18,19].

Neutrophil elastase is responsible for degradation of E-cadherin

To determine whether elastase was responsible for degradation, neutrophils were pretreated with a specific neutrophil elastase inhibitor, DMP777. Under these conditions, neutrophil transmigration and the decrease in transepithelial resistance were both significantly inhibited. Moreover, the release of the E-cadherin fragment was completely inhibited. The large molecular weight produce inhibitor, alphal-antitrypsin, diminished but did not prevent the fall in transepithelial resistance nor the release of the proteolytic degradation fragment of E-cadherin. We interpret these data to mean that the larger molecular weight inhibitor does not have access to 'protected space' between the neutrophils and the epithelium, which is presumably the site of proteolytic degradation. In contrast, the membrane permeant elastase inhibitor, DMP777 , completely


inactivates both cell associated and extracellular elastase, and therefore abrogates the proteolytic degradation of E-cadherin.

That neutrophil elastase is responsible for the cleavage of E-cadherin was supported by experiments where purified elastase was added directly to the epithelial monolayers. Under conditions where the tight junctions were opened transiently, direct addition of purified elastase induced release of the 23 kilodalton proteolytic fragment of E-cadherin in a manner analogous to that observed during neutrophil transmigration. We conclude that during neutrophil transmigration through epithelial monolayers, there is limited proteolytic cleavage of the junctional complexes. This could be mediated by direct proteolytic cleavage of E-cadherin by neutrophil elastase. Alternatively, it is possible that neutrophil elastase can induce alterations in epithelial cells, leading to activation of epithelial-derived proteases, such as matrix metalloproteases, that could be responsible for the cleavage of epithelial E-cadheren. This area is currently under investigation.

Neutrophil transmigration induces epithelial apoptosis

To investigate further the mechanisms of epithelial injury, we utilized imaging techniques to determine whether or not epithelial cells were undergoing necrosis or apoptosis. These studies indicated that the primary changes in the epithelial cells were apoptosis. Evidence for this was obtained using TUNEL staining that demonstrated that epithelial cells directly adjacent to the areas of microinjury stained positive, indicating apoptosis. An independent assay of apoptosis, loss of mitochondral membrane potential, confirmed that the epithelial cells were undergoing apoptotic death. Epithelial cells that had detached from the monolayer, and were found free in the lower chamber, were also TUNEL positive with evidence of loss of mitochondral integrity.

To determine whether neutrophil-derived elastase was responsible for epithelial apoptosis, purified elastase was added directly to epithelial cells. Under these conditions, epithelial cell apoptosis occurred, as determined by TUNEL staining and loss of mitrochondral membrane potential. Our conclusion from these studies was that leukocyte-derived elastase was responsible for epithelial apoptosis. Preliminary studies examining the signaling pathways regulating these effects have been conducted. We observed that leukocyte transmigration induced a decrease in the phosphorylation status and activation of the epithelial enzyme AKT. This enzyme is known to exert a potent anti-apoptotic effect, and the observed decrease in activity presumably initiates and/or promotes epithelial apoptosis.


Are these changes in the in vitro system relevant to human disease?

Support for the observations outlined in the experimental studies described above comes from patients with inflammatory bowel disease. Large areas of epithelial denudation are observed and in areas of 'crypt abscesses', epithelial apoptosis can be observed adjacent to transmigrating neutrophils. Similar changes in epithelial cells of the gut are observed in patients with sepsis (Shannon, Ginzberg, Downey; unpublished observations).

Concluding statement

In this manuscript, we have provided evidence that epithelial injury can occur during leukocyte transmigration through epithelial cells but only when it is associated with unregulated leukocyte activation and release of proteolytic enzymes. It is important to emphasize that there are well-described examples in vitro, in animal models, and in human studies where leukocyte transmigration can occur without evidence of significant epithelial dysfunction or injury [20]. However if leukocytes become activated during the process of transmigration, leukocyte-derived cytotoxic products can damage the vicinal epithelial cells, leading to changes in their intracellular junctions, including adherence and tight junctions. This we believe contributes to epithelial dysfunction and injury, including loss of the barrier function that is observed clinically in human inflammatory diseases, such as sepsis and acute lung injury.

Reference 1. Abraham E, Matthay MA, Dinarello CA, et a1 (2000). Consensus conference definitions for

sepsis, septic shock, acute lung injury, and acute respiratory distress syndrome: time for a reevaluation. Crit Care Med 28:232-235

2. Ware LB, Matthay MA (2000) The acute respiratory distress syndrome. N Engl J Med 342: 1334--1349

3. Lee WL, Downey GP (2001) Neutrophil activation and acute lung injury. CUff Opin Crit Care 7:1-7

4. Rinaldo IE (1986) Mediation of ARDS by leukocytes. Clinical evidence and implications for therapy. Chest 89:590-593

5. Bachofen M, Weibel ER (1982) Structural alterations of lung parenchyma in the adult respiratory distress syndrome. Clin Chest Med 3:35-56

6. Hogg JC, Doerschuk CM (1995) Leukocyte traffic in the lung. Annu Rev Physiol 57 :97 -114 7. Worthen GS, Schwab B, Elson EL, Downey GP (1989) Mechanics of stimulated neutrophils:

cell stiffening induces retention in capillaries. Science 245: 183-186 8. Wagner JG, Roth RA (2000) Neutrophil migration mechanisms, with an emphasis on the

pulmonary vasculature. Pharmacol Rev 52:349-374


9. May RC, Machesky LM (2001) Phagocytosis and the actin cytoskeleton. J Cell Sci 114:1061--1077

10. Colgan SP, Parkos CA, Delp C, et al (1993) Neutrophil migration across cultured intestinal epithelial monolayers is modulated by epithelial exposure to lPN-gamma in a highly polarized fashion. J Cell Bioi 120:785-798

11. Matthay MA, Folkesson HO, Verkman AS (1996) Salt and water transport across alveolar and distal airway epithelia in the adult lung. Am J PhysioI270:L487-L503

12. Crouch E, Hartshorn K, Ofek 1(2000) Collectins and pulmonary innate immunity. Immunol Rev 173:52-65

13. Oanz T, Weiss J (1997) Antimicrobial peptides of phagocytes and epithelia. Semin Hematol 34:343-354

14. Mitic LL, Van Itallie CM, Anderson JM (2000) Molecular physiology and pathophysiology of tight junctions I. Tight junction structure and function: lessons from mutant animals and proteins. Am J Physiol Oastrointest Liver Physiol 279:0250-0254

15. N athke IS, Hinck LE, Nelson WJ (1993) Epithelial cell adhesion and development of cell surface polarity: possible mechanisms for modulation of cadherin function, organization and distribution. J Cell Sci [Suppl] 17:139-145

16. Borrmann CM, Mertens C, Schmidt A, et al (2000) Molecular diversity of plaques of epithelialadhering junctions. Ann N Y Acad Sci 915:144-150

17. Petit V, Thiery JP (2000) Focal adhesions: structure and dynamics. BioI Cell 92:477-494 18. Oinzberg H, Cherapanov V, Dong Q, et al (2001) Neutrophil-mediated epithelial injury during

transmigration: role of elastase. (abstract) Am J Physiol Oastrointest Liver Physiol (in press) 19. Oinzberg H, Shannon P, Downey OP (2000) Neutrophil products and alterations in epithelial

junctional proteins: prevention of artifactual degradation. J Immunol Methods 239:45-52 20. Martin TR, Pistorese BP, Chi EY, et al (1989) Effects of leukotriene B4 in the human lung.

Recruitment of neutrophils into the alveolar spaces without a change in protein permeability. J Clin Invest 84:1609-1619

Pro- and Anti-inflammatory Cytokines and Apoptosis in Acute Lung Injury

s. UHLIG, D. BURDON

"The human body is composed of approximately 1014 cells, each of which is capable of committing suicide by apoptosis. Not surprisingly this process has inherent weaknesses that can result in inappropriate cell death and disease pathogenesis ,,1

Sepsis is the most frequent cause for the development of acute lung injury (ALI). Acute lung injury is a complex disease that so far has defied mechanistic definitions. However, it is widely accepted that pro-inflammatory mediators and overshooting immune reactions shape the course of the disease. Recently, it has been suggested that apoptosis might play an important role in the pathogenesis of ALI. This has posed the question of how pro- (e.g., TNF, IFN -,,(, IL-I) and anti-inflammatory (e.g., IL-lO, TGFj3, PG&) mediators affect apoptosis in the lung. The beneficial properties of apoptosis during pulmonary development and in the resolution of inflammation are well recognized. However, bacterial endotoxins (lipopolysaccharide, LPS) and infections elicit apoptosis in cells of the lung [2]. An important question therefore is whether pulmonary apoptosis is detrimental under these conditions. It is the scope of the present manuscript to review the current knowledge on the effects of proand anti-inflammatory factors and cytokines on apoptosis in the lung and its possible consequences for the development of ALI. We will not address intracellular signalling mechanisms of apoptosis, a topic for which several excellent reviews are available [5-7]. We will rather focus on pathophysiological implications of apoptosis in the lungs in the context of ALI. After a brief look at apoptosis, we will describe the effect of cytokines on apoptosis in individual cell populations iJ;l the lung, before discussing the consequences for the whole organ.

78 s. Uhlig, D. Burdon

A brief look at apoptosis

Initially, apoptosis was viewed as a beneficial phenomenon, not unlike falling leaves, that allows the organism to dispose of cells during development or in the immune system. Soon, however, it became clear that under certain conditions apoptosis may result in severe tissue destruction, as was illustrated by TNF or Fas-mediated liver injury [8, 9]. Such studies provoked investigations into the role of apoptosis during lung injury as discussed here. Until recently, a sharp distinction was made between apoptosis and necrosis. However, even with the same stimulus cells can switch from apoptosis to necrosis depending on intracellular ATP levels [10]. Moreover, stimuli such as hydroperoxides cause apoptosis at low and necrosis at higher concentrations, whereas at intermediate concentrations cells can die either way [11]. Recently, it was recognized that apoptosis and necrosis represent only extreme forms of cell death, and apoptosis-like as well as necrosis-like programmed cell death have been recognized [5]. All these forms of cell death appear to be mediated by different intracellular signaling pathways [5]. In the future, these distinctions, which also require electron microscopical examination of the nucleus, will certainly be crucial to fully understand tissue injury in the lung.

Many mediators that induce apoptosis do so by stimulation of receptors of the TNF family of surface receptors. Members of this family that have been described in the lung are type 1 TNF-receptor (CDI20a), Fas (CD95) and TRAIL (Tumor necrosis factor-related apoptosis-inducing ligand) receptors, TRAIL-Rl and TRAIL-R2 [12]. The natural ligands for these receptors are TNF, Fas-ligand (FasL) and TRAIL, respectively. Ofthese only TNF and FasL, but not TRAIL, are induced in the lungs by septic shock and cecal ligation and puncture [13]. Receptor ligation initiates cellular signaling cascades that result in the activation of caspases, but also other proteases, that finally execute apoptosis [5, 12]. So far 14 mammalian caspases have been identified, that can be divided in two major subfamilies: those involved in cytokine maturation such as caspase 1, also known as interleukin-l-converting enzyme (ICE, which matures both IL-l and IL-18), and those that participate in initiation or execution of apoptosis [1]. Unspecific caspase inhibitors have provided protection in a number of different models such as ischemia-reperfusion, meningitis, septic shock, apoptotic liver injury and pulmonary fibrosis [1,9, 14, 15]. In addition, a number caspase independent ways of programmed cell death exist [5], but none of them have yet been investigated in the context of ALI.

Pro- and Anti-inflammatory Cytokines and Apoptosis in Acute Lung Injury 79

Endothelial cells

Conceivably, apoptosis of endothelial cells could increase vascular permeability and thereby promote formation of pulmonary edema, a hallmark of ALI. Treatment of pulmonary artery endothelial cells with LPS causes apoptosis [16, 17], which is prevented by activation of integrin receptors through collagen, laminin, fibronectin or integrins themselves [16, 18]. This suggests that LPS will not cause endothelial cell apoptosis as long as the cells have contact with the basement membrane. Therefore, it is important to state that injection ofLPS induces apoptosis in pulmonary endothelial cells within 6 hours [19, 20]. However, rather than a direct effect of LPS, this is the result of tumour necrosis factor (TNF) release and subsequent activation of acid sphingomyelinase and ceramide [20]. Consistent with this finding is the observation that exposure of pulmonary endothelial cells to TNF causes apoptosis and increases the permeability of EC monolayers in culture [21, 22]. However, whereas TNF-induced apoptosis can be blocked by inhibition of myosin light chain kinase or Rho-kinase, this inhibition has no effect on the increased permeability [22]. Thus, apoptosis and vascular permeability appear to be two separate events. This conclusion is further corroborated by in vivo studies, where the caspase inhibitor Z-VAD.fmk prevented apoptosis, but had no effect on pulmonary edema and leukocyte sequestration in endotoxemic mice [14, own unpublished observations]. Rather than by apoptosis, LPS-induced vascular permeability appears to be caused by neutrophils [23] and platelet activating factor [24]. In contrast to a widely held belief, there is almost no direct evidence that TNF contributes to LPS- or E. Coli-induced pulmonary edema [25, 26].

In line with the concept of compartmentalization [27-29], differences exist regarding whether LPS is given systemically or instilled into the airways. While systemic LPS exposure causes endothelial cell apoptosis within hours, local LPS causes endothelial cell apoptosis only after 24 h and even that is not a common finding. Juanita et al. instilled 5 Ilg LPS into male Swiss mice and observed no apoptosis in endothelial cells within 72 h [4], whereas Kitamura et al. found apoptotic endothelial cells I day after instillation of 30 Ilg LPS into ICR mice [30]. In the latter study, lung injury was attenuated by an anti-Fas antibody, yet the specific role of endothelial cell apoptosis for the disease process was not addressed. Notably, instillation of a Fas-activating antibody caused no endothelial cell apoptosis [31].

Taken together, LPS causes pulmonary endothelial cell apoptosis, although this event is probably unrelated to edema formation. The pathophysiological significance of endothelial cell apoptosis in the lung remains to be determined.


Epithelial cells

Leakiness of endothelial cells results in interstitial edema, whereas alveolar edema will develop and persist after airway epithelial injury. Following injection or instillation of LPS, apoptosis occurs in bronchial epithelial cells already after 4 h with a maximum at 24 h [4] and in type II alveolar epithelial cells (ATII) at 24 h [14, 30]. ATII apoptosis at time points later than 24 h or so is thought to be beneficial, because it contributes to resolution of ATII cell hyperplasia, as demonstrated in the case of LPS [32], KGF [33, 34] and ALI patients [35].

The early apoptosis of the bronchial epithelium after local LPS exposure is independent of TNF [4]. In addition, instillation of TNF [4, 36] and chronic interstitial pneumonia [35] failed to cause apoptosis in airway epithelial cells. In agreement with these in vivo findings, Mallampalli et al. found no apoptosis in ATII cells exposed to TNF [36]. However, conflicting data have been provided by several in vitro studies in which TNF-induced apoptosis in bronchial epithelial cells [37, 38] and in ATII cells [39,40], or enhanced apoptosis elicited by activating Fas antibodies [38]. The interpretation of these in vitro findings is not quite clear and, just as with endothelial cells (see above), apoptosis of epithelial cells in culture may depend on the substratum on which the cells are grown. In line with this, it was shown that tissue integrity prevents expression of caspases and apoptosis via integrins [41,42].

Interferon-y (IFN) represents another pro-inflammatory cytokine that has been suggested to induce apoptosis after 48 h within A549 lung epithelial cells [43] and human bronchial epithelial cells (NHBE) [37]. Interestingly, IFN and apoptosis induced by the Fas-activating monoclonal antibody J02 was inhibited by dexamethasone treatment, presumably through induction of the so-called inhibitor of apoptosis (hIAP) [43]. Thus, steroids appear to have dual effects on apoptosis causing apoptosis in some cells (e.g. lymphocytes [44]) and preventing it in others.

Other factors that induce apoptosis in Am cells are Fas-activating antibodies [31,45,46], rh-Fas-ligand [47], andIL-2 [48]. Particular attention has been paid to the Fas system. Fas antigen is expressed in the lungs, and its expression has been localized to alveolar and epithelial cells, Clara cells, alveolar macrophages, and parenchymal cells such as myofibroblasts [47]. Exposure of ATII cells or A549 cells in culture to Fas-activating antibodies causes apoptosis [43, 45]. Furthermore, application of an anti-Fas blocking antibody attenuated lung injury and edema formation that occurred one day after instillation ofLPS [30]. Thus, there is evidence that Fas is involved in LPS-induced acute lung injury. However, recent findings challenge the view that all of these effects of Fas are related to apoptosis, since binding of FasL to Fas can lead to activation of NF-kB and


release of inflammatory cytokines [49-51]. Thus, Fas may cause pulmonary inflammation and tissue injury, independent from causing apoptosis [51,52].

Recently angiotensin II, which traditionally is not viewed as a cytokine, has been implicated in ATII cell apoptosis [53]. Apparently high local concentrations of this peptide exist, and fibroblasts from fibrotic human or rat lungs release angiotensinogen and its product angiotensin which are capable of inducing ATII cell apoptosis [54]. Even more intriguing, activation of both Fas and TNF receptors on these cells resulted in increased transcription and release of angiotensiogen [40,55]. Additionally, ATII cell apoptosis was inhibited by blockade of angiotensin II receptors or inhibition of the angiotensin-converting enzyme (ACE) [40]. Therefore, pharmacological interventions of the angiotensin pathway might be of therapeutic value in the treatment of ALI, given that apoptosis contributes to its development. Notably, the ACE-inhibitor captopril (as well as a caspase inhibitor) was already shown to attenuate bleomycin-induced fibrosis in mice [56].

Alveolar macrophages

Alveolar macrophages (AM<I» do not only defend the alveolar air space against pathogens, but also appear to play an important role in regulating apoptosis in the alveolus. As such they have the capability to down-regulate apoptosis in other cells [57] and play an important part in removing apoptotic cells, primarily invaded neutrophils [58]. Hence, apoptosis of AM<I> themselves may favour secondary necrosis of apoptotic cells in the alveolus followed by inflammation. The uptake of apoptotic cells suppresses the secretion of pro-inflammatory mediators from activated macrophages [59]. Phagocytosis of apoptotic neutrophils has anti-inflammatory effects on macrophages linked to the production of TGF-~? PGE2 and IL-l 0, and a decreased production of chemokines and TNF [60,61]. These effects are mediated via the recently cloned phosphatidylserine receptor. In contrast, necrotic neutrophils release proteases that inactivate the phosphatidylserine receptor and then lead to a pro-inflammatory response upon engulfment of the corpses [61]. Thus inhibition of apoptosis, e.g., by nitric oxide, ATP-depletion, oxidative stress and a conversion of death to necrosis has a profound effect on the persistence of inflammation. The effect of apoptotic cells on the macrophage system might also explain some of the beneficial effects of steroids in pulmonary inflammation. Apart from their direct anti-inflammatory role, steroids highly increase the capacity for the uptake of apoptotic cells and thereby foster the resolution of inflammation by natural pathways [62,63].

82 S. Uhlig, D. Burdon

Exposure of alveolar macrophages (AM<I» to high concentrations of LPS (10 J..lg/ml) causes apoptosis within 24 h, an effect that was not seen by a variety of different cytokines [64], but that was enhanced by IFNy and mitigated by IL-lO and TGF~. In vivo, AM<I> apoptosis was observed within 24 h after instillation of 30 J..lg LPS [30], but not of 5 J..lg LPS [4]. Notably, AM<I> may produce factors such as GM-CSF and IL-8 which may delay apoptosis of other cells such as ATll cells and PMN s [65]. Whether AM <I> apoptosis is detrimental or whether it is simply a sign of resolving inflammation in the alveolus remains to be established.

Polymorphonuclear neutrophil granulocytes (PMN)

Persisting tissue granulocytes and other leukocytes in the alveolar space form the cellular elements of inflammation in the lung, hence, leading to an ALI. For the past decades, inflammation was regarded as an entirely beneficial host defence response although it has been recognized for centuries that there is the potential both for complete resolution of inflammation and for persistence and harmful progression. Perhaps, the best examined example for a beneficial inflammation is the response to the invasion by Streptococcus pneumoniae in the lung with a successive lobar pneumonia. A massive migration of neutrophil granulocytes into the local air space is characteristic for this disease. Lobar pneumonia illustrates the potential of severe inflammation to resolve completely with no destructive sequelae. Comparing this disease with other forms of infection or inflammation suggests that the outcome is determined by the balance or imbalance of mechanisms that cause and amplify injury versus the mechanisms that protect tissue and promote resolution.

Therefore, ALI could either be a result of uncontrolled pro-inflammatory mechanisms or the failure, or inefficiency, of normal resolution processes. The neutrophil granulocyte is a typical inflammatory cell and frequently the first to migrate to an inflamed site. This is flanked by a number of subsequent events including monocyte migration and edema [66, 67]. The neutrophil has been directly implicated in the pathogenesis of ALI and ARDS. It contains a large number of agents responsible not only for tissue injury but also the capacity to generate further chemotactic agents [68, 69]. Originally, it was assumed that extravasated granulocytes underwent necrosis in the inflamed site and that their fragments were cleared by macrophages [70]. More recently, however, it was shown that neutrophils isolated from peripheral blood undergo apoptosis and that this process determines the rapid clearance of intact senescent neutrophils by macrophages [71]. A range of inflammatory mediators including IL-2,


G-CSF, GM-CSF, IL-8, other CXC chemokines, IL-6, and endotoxin inhibit apoptosis, whereas TNFa and IL-lO seem to promote granulocyte apoptosis [72-76]. TNFa is frequently found in BAL fluid of ARDS, and experimental studies have shown that the interaction between TNFa and the p55 receptor is crucial for the development of ALI [77]. While PMN s cultured with TNFa undergo rapid apoptosis within the first 4-10 h, TNFa markedly inhibits apoptosis in surviving PMNs after 24 h. This is probably a result of the activation of the p38-mitogen-activated protein kinase and the successive release of IL-8 [74].

There now exist numerous data that support removal of apoptotic neutrophils in the process of resolving acute inflammation in the lung [78]. During apoptosis, neutrophil membranes remain intact and macrophages can clear apoptotic cells without leakage of potentially injurious neutrophil contents into alveolar space [79]. Additionally, neutrophils normally degranulate their injurious contents after external stimulation with inflammatory mediators, such as bacterial toxins. During apoptosis, however, neutrophils lose this ability presumably by a down-regulation of associated receptors such as the IgG receptor FcR III or CDl6 [80-82]. Therefore, in contrast to necrosis, apoptosis provides a granulocyte clearance mechanism that limits tissue injury and promotes resolution rather than persistence of inflammation [83].

Experimental studies on apoptosis and ALI

ALI and acute respiratory distress syndrome (ARDS) are characterized by non-cardiogenic pulmonary edema, infiltration of predominantly neutrophilic granulocytes, and fibrosis at later stages of the disease. In this chapter we will discuss the possible role of apoptosis for the development of these pathological alterations.

Many studies are of a descriptive nature and only few have investigated the role of apoptosis for the pathophysiology of ALI/ARDS (Table 1). A number of studies have shown that LPS, FasL or activating anti-Fas antibodies cause apoptosis in a number oflung cells (see above) and acute pulmonary inflammation [2, 47, 84]. In contrast to these acute effects within 24 h, repeated inhalation of J02 caused pulmonary fibrosis in mice [85]. Because pulmonary fibrosis may occur in the late phase of ARDS, these and other findings raise the hypothesis that the fibrosis in these patients is preceded by apoptosis of presumably alveolar epithelial cells [86]. However, in sepsis models that show pulmonary fibrosis, such as the generalized zymosan-induced inflammation [77], the role of apoptosis has not been studied. It is of interest, though, that

00

T

able

1. S

tudi

es o

n a

popt

osis

in

expe

rim

enta

l se

psis

mod

els

~

Spe

cies

M

odel

C

ells

wit

h A

popt

osis

In

terv

enti

on

Com

men

t R

efer

ence

Mou

se

J02,

rh-

Fas

L:

Inst

illa

tion

A

TII

pu

lmon

ary

infl

amm

mat

ion

wit

h 4

5,4

6

infi

ltra

tion

of

PM

N a

nd o

edem

a 3

1,4

7

J02:

Chr

onic

aer

osol

A

TII

, br

onch

ial

epit

hel

fibr

osis

85

Mou

se

TN

Fi.

t.

Abs

ent

(AT

ll)

4

Rat

T

NF

i.t.

A

bsen

t 36

Mou

se

LP

S L

v. (

20m

gikg

) P

ulm

onar

y E

C Z

.VA

D.f

mk

redu

ced

pulm

onar

y ap

opto

sis,

19

(C

aspa

se i

nhib

itor

) bu

t ef

fect

no

effe

ct o

n ac

ute

lung

inj

ury.

in

crea

sed

surv

i val

L

PS L

p.

Not

spe

cifi

ed

from

6-24

h 2

LP

S i

.p.

Abs

ent

3 L

PS

i.p

. P

ulm

onar

y E

C A

SM

ase

k.o"

A

popt

. is

med

iate

d by

TN

F a

nd c

eram

i de

20

FG

F:

less

apo

ptos

is

LP

S.

Lp.

PM

N

IL-I

k.o

: ap

opto

sis

PM

Ns

not

resp

onsi

ble

for

apop

tosi

s II

I un

affe

cted

L

PS

Lt.

PM

N

IL-6

k.o

.: ap

opto

sis

Ant

i-F

as a

tten

uate

s ed

ema

112

unaf

fect

ed

LP

S i

.t. (

5 ?g

) on

ly in

bro

nchi

al e

pith

el

Ant

i-T

NF

: ap

opto

sis

apop

tosi

s to

res

olve

LP

S-i

nduc

ed l

ung

4 un

affe

cted

ep

ithe

lial

cel

l pr

olif

erat

ion

LP

S i

.t. (

30 ?

g)

in A

TII

, A

M?,

EC

A

nti-

Fas

: le

ss a

popt

osis

A

M?

engu

lf a

popt

otic

PM

Ns

30

LP

Si.

t.

AT

l!

32

LP

S i

.t.

PM

N

79

Rat

L

PS

i.p

. N

ot s

peci

fied

Z

n2 +

: le

ss a

popt

osis

P

rote

ctio

n by

HS

P70

? 11

3 L

PS

i.t.

P

MN

s IL

-IO

: le

ss P

MN

s IL

-IO

pro

mot

es c

lear

ance

of

PM

Ns

114

~

LPS

inh

ibit

s P

MN

apo

ptos

is

115

c::::

Mou

se

Cec

al l

igat

ion

+ p

unct

ure

Not

spe

cifi

ed

Sim

ilar

in C

h3/H

eN a

nd H

eJ m

ice

3 g:

Mou

se

Hae

mor

rhag

e P

MN

IL

-I k

.o:

apop

tosi

s II

I ciQ

' un

affe

cted

~

Rab

bit

Inst

illa

tion

of

HC

I N

ot s

peci

fied

O

verv

enti

lati

on:

100

to

!::

less

apo

ptos

i ..., 0.

. 0 ::s

Pro- and Anti-inflammatory Cytokines and Apoptosis in Acute Luug Injury 85

caspase inhibitors prevented fibrosis caused by bleomycin [87], a model in which the Fas-system is not involved [88].

Only limited data are available that address the crucial question of how prevention of apoptosis impacts lung injury and survival. In such studies it was noted that pre-treatment with neutralizing anti-Fas antibodies reduced epithelial cell apoptosis and pulmonary edema one day after LPS instillation [30]. However, because Fas-activation may be pro-inflammatory independent from apoptosis [49-51], it remains to be established whether there is a link between apoptosis and edema. Such a link appears doubtful, because prevention of apoptosis in endotoxemic mice by neutralizing TNF or with caspase inhibitors prevented neither pulmonary edema nor leukocyte sequestration [14, 20, own unpublished observation]. Interestingly, however, these treatments improved survival [14,20], bringing to attention the still unsettled debate on the significance of edema formation for the outcome of ARDS. How inhibition of apoptosis would enhance survival, and whether this is related to the lung at all, remains to be determined.

As a note of caution, we should add that mechanisms other than apoptosis cannot be completely excluded to explain the increased survival in the studies mentioned above. For example, the caspase inhibitor used (Z-VAD) also inhibits interleukin-l-converting enzyme (caspase-l) and therefore production of IL-l [89], a critical mediator of septic shock [90]. It is also known that mice deficient in caspase-l are resistant to endotoxic shock [91]. Moreover, Z-VAD inhibits a variety of other proteases, such as cathepsin B, that also have been associated with inflammation [5]. Further complications arise from the fact that apoptosis is difficult to measure, which is why supposedly related endpoints, such as the TUNEL assay, are widely used. However, agents like ZVAD may inhibit these particular endpoints, but not necessarily programmed cell death. In fact, programmed cell death frequently occurs caspase-independent [5]. Another observation in those LPS shock studies was that endothelial cell apoptosis and survival were related to reduced ceramide production (e.g., by using mice deficient in acid sphingomyelinase) [20]. However, ceramide may not only cause apoptosis, but is also an integral component of rafts and as such is involved in many different signaling events [92, 93].

Most patients with acute lung injury have to be mechanically ventilated. Mechanical ventilation, however, may have profound side effects [94-96]. Recently, it was shown that mortality in ARDS is reduced by 21 % if patients are ventilated with reduced tidal volumes [97]. Experimental studies have demonstrated that ventilation with increased tidal volumes or pressure (termed overventilation [98] results in the release of a number of different mediators such as TNF that are capable of inducing apoptosis [96, 98]. For these reasons,


it is important to understand the effects of mechanical ventilation on apoptosis. In vitro studies have shown that cyclic stretch activates caspases and induces apoptosis in ATII cells [99]. The first in vivo study in this area was performed in the ARDS-model of hydrochloric acid instilled rabbits. Somewhat surprisingly, it was found that overventilation (with low PEEP) reduced the incidence of apoptotic cells [100]. On a mechanistic basis this may be explained by the fact that overventilation also induces NF-KB [98], which is a known anti-apoptotic factor [101]. Clearly, further studies, also in healthy lungs, are needed to clarify the effect of ventilation on apoptosis. Notably, in that study it was revealed that overventilation increases apoptosis in renal tubular epithelial cells, suggesting that ventilation-induced apoptosis might contribute to multiple systemic organ failure during ARDS.

Clinical studies on apoptosis and ALI

A number of studies have evaluated apoptosis in patients with ALI or ARDS (Table 2). Most of these studies focused on the cells that are retrieved by BAL. The proportion of apoptotic neutrophils in the BAL recovered from ARDS patients is significantly lower than in normal BAL neutrophils [102] and the lavage of such patients delays PMN apoptosis [102]. Delayed apoptosis of PMNs is a general observation in ARDS and different studies have named slightly different cytokine patterns to explain this. Matute-Bello et al. identified G-CSF and GM-CSF [102, 103], Aggarwal et al. G-CSF and IL-8 rather than GM-CSF [75] and Lesur et al. IL-2 and GM-CSF [73].

Table 2. Studies on apoptosis in ALI and ARDS

Apoptotic cell Mediators found in BAL

ATII

(ATII)

PMN from BAL IL-2. GM-CSF

PMN from BAL G-CSF, GM-CSF

PMN from BAL G-CSF and IL-8, little

(BAL)

(BAL)

FasL

perforin, granzyme A+B, FasL, Fas

BAL, Bronchoalveolar lavage

Comment Reference

Resolution of ATII cell hyperplasia 35

Fas(L) staining of alv. Epithel 110

PMN apoptosis is lower in ARDS 73

BAL is anti-apoptotic for PMNs by G(M)-CSF 102, 103

GM-CSFGM-CSF higher in survivors 75

BAL causes Apo of lung epithelial cells 109

76


While numerous data on the influence of these mediators on granulocytes are available from in vitro experiments, the in vivo situation seems to be more complex. For example, high levels of IL-2 in the BAL fluid of ARDS were associated with increased survival [73]. In contrast, elevated levels of G-CSF in BAL fluid in ARDS correlated with higher percentages of non-survival [75]. Especially, G-CSF and GM-CSF may exacerbate lung injury [104] and occasionally induce ALI [l05, 106]. However, experimental studies show contrasting properties of G-CSF and GM -CSF, demonstrating protection by G-CSF [107] and aggravation by GM-CSF [108] treatment in murine endotoxic shock models. Endogenously produced GM-CSF may playa different role, as is suggested by the higher levels of GM -CSF in survivors of ARDS compared to non-survivors [103]. To explain this, one could hypothesize that higher levels of GM -CSF result in a faster differentiation of migrating monocytes to alveolar macrophages, hence, enhancing the phagocytosis system.

These data could suggest that the course of ARDS is determined by persistent fully functional neutrophils owing to the lack of apoptosis. Yet, the following facts have to be taken into consideration. As described above, the ideal pathway is the non-inflammatory clearance of apoptotic neutrophils by macrophages, because otherwise apoptotic neutrophils undergo secondary necrosis followed by inflammation [59]. One could therefore speculate that the main pathogenetic feature in ARDS is not the persistence of neutrophils, but rather an ineffective clearance by the alveolar macrophages [83].

At any rate, low levels of apoptotic neutrophils are found in the lungs of patients suffering an ALI or ARDS and the down-regulation of apoptosis in these cells seems to be an important feature of the disease process. The control of granulocyte apoptosis underlies a complex balance of pro- and anti-apoptotic mediators. These regulation mechanisms are still poorly understood. To target PMN apoptosis in a therapeutic intervention, one will have to consider how this might affect death of other lung cells and if this is really beneficial for the resolution of inflammation. Additionally, the massive induction ofPMN apoptosis may overwhelm the capacity of alveolar macrophages, resulting in secondary necrosis and prolonged inflammation.

In contrast to its anti-apoptotic effects on PMNs, the BAL liquid of ARDS patients, but not at risk patients, caused apoptosis of distal lung epithelial cells in a Fas-dependent manner [109]. However, since the BAL contained much lower concentrations of FasL than needed to induce apoptosis of distal lung epithelial cells, the authors concluded that a cofactor must be present in the BAL of ARDS patients [109]. Therefore it is interesting that in addition to Fas and FasL, the BAL of these patients contained elevated concentrations of perforin, granzyme A and B [76, 109]. Recently, Albertine and colleagues


observed intense Fas staining in alveolar epithelial cells in patients who died of ARDS [110]. In addition, they detected FasL on sloughed epithelial cells. All these findings suggest a prominent role of the Fas system in ARDS. However, because Fas must also be considered as a pro-inflammatory mediator, the role of apoptosis versus inflammation remains to be determined. Clearly, more studies are needed to define the role of apoptosis for the pathogenesis of ARDS.

Conclusion

Like any of the about 1014 cells of which the human body is composed, also lung cells can undergo apoptosis. In the adult lung, this process is beneficial to the clearance of inflammatory cells from the lungs and to the resolution of type II cell hyperplasia. Potentially harmful consequences of apoptosis are suggested by experimental data showing apoptosis in acute lung injury in many cells, of which endothelial and ATII cells have received the most attention. However, the hypothesis that this might promote pulmonary edema is questioned by the failure of caspase inhibitors to reduce edema formation in acute lung injury. These findings, however, do not discount an important role for the Fas system, which is highly expressed in lungs from ARDS patients. The interpretations of experiments in this area are confounded by the unexpected finding that FasL may be another pro-inflammatory mediator. An unequivocal finding from the clinical studies is the low proportion of apoptotic neutrophils. If the persistence of neutrophils in the alveolar space of patients represents a key detrimental event in ARDS or ALI has yet to be elucidated.

Acknowledgements

This work was supported by the Deutsche Forschungsgemeinschaft grants SFB 367/A9 and DFG Uh 8812-4 to S.U.

References 1. Nicholson DW (2000) From bench to clinic with apoptosis-based therapeutic agents. Nature

407:810-816 2. Bohlinger I, Leist M, Gantner F, et al (1996) DNA fragmentation in mouse organs during

endotoxic shock. Am J Patho1149: 1381-1393 3. Hiramatsu M, Hotchkiss RS, Karl IE, Buchman TG (1997) Cecal ligation and puncture (CLP)

induces apoptosis in thymus, spleen, lung, and gut by an endotoxin and TNF-independent pathway. Shock 7:247-253


4. Vernooy JR, Dentener MA, van Suylen RJ, et al (200l) Intratracheal instillation oflipopolysaccharide in mice induces apoptosis in bronchial epithelial cells: no role for tumor necrosis factor-alpha and infiltrating neutrophils. Am J Respir Cell Mol Bio 24:569-576

5. Leist M, Jiiiitteia M (2001) Four deaths and funeral: from caspases to alternative mechanisms. Nat Rev Mol Cell BioI 2: 1-10

6. Krammer PH (2000) CD95's deadly mission in the immune system. Nature 407:789-795 7. Kuwano K, Hara N (2000) Signal transduction pathways of apoptosis and inflammation induced

by the tumor necrosis factor receptor family. Am J Respir Cell Mol Bioi 22: 147-149 8. Leist M, Gantner F, Bohlinger!, et al (1995) Thmornecrosis factor-induced hepatocyte apoptosis

precedes liver failure in experimental murine shock models. Am J PathoI146:1220-1234 9. Kiinstle G, Leist M, Uhlig S, et al (1997) ICE-protease inhibitors block murine liver injury and

apoptosis caused by CD95 or by TNF-alpha. Immunol Lett 55:5-10 10. Leist M, Single B, Castoldi AF, et al (1997) Intracellular adenosine triphosphate (ATP) concen

tration: a switch in the decision between apoptosis and necrosis. J Exp Med 185:1481-1486 11. Weisser M, Tiegs G, Wendel A, Uhlig S (1998) Quantification of apoptotic and lytic cell death

by video microscopy in combination with artificial neural networks. Cytometry 31 :20-28 12. Screaton G, Xu X-N (2000) T cell life and death signalling via TNF receptor family members.

CUff Opin ImmunoI12:316-322 13. Tannahill CL, Fukuzuka K, Marum T, et al (1999) Discordant tumor necrosis factor-alpha

superfamily gene expression in bacterial peritonitis and endotoxemic shock. Surgery 126:349-357 14. Kawasaki M, Kuwano K, Hagimoto N, et al (2000) Protection from lethal apoptosis in lipopoly

saccharide-induced acute lung injury in mice by a caspase inhibitor. Am J PathoI157:597-603 15. Kuwano K, Kunitake R, Maeyama T, et al (2001) Attenuation of bleomycin-induced pneumo

pathy in mice by a caspase inhibitor. Am J Physiol Lung Cell Mol PhysioI280:L316-L325 16. Hoyt DG, Mannix RJ, Rusnak JM, et al (1995) Collagen is a survival factor against LPS-induced

apoptosis in cultured sheep pulmonary artery endothelial cells. Am J PhysioI269:L171-L177 17. Maeda K, Abello PA, Abraham MR, et al (1995) Endotoxin induces organ-specific endothelial

cell injury. Shock 3:46-50 18. Hoyt DG, Mannix RJ, Gerritsen ME, et al (1996) Integrins inhibit LPS-induced DNA strand

breakage in cultured lung endothelial cells. Am J Physiol 270:L689-L694 19. Fujita M, Kuwano K, Kunitake R, et al (1998) Endothelial cell apoptosis in lipopolysaccharide

induced lung injury in mice. Int Arch Allergy Immunol 117 :202-208 20. Haimovitz-Friedman A, Cordon-Cardo C, Bayoumy S. et al (1997) Lipopolysaccharide induces

disseminated endothelial apoptosis requiring ceramide generation. J Exp Med 186: 1831-1841 21. Polunovsky VA, Wendt CH, Ingbar DH, et al (1994) Induction of endothelial cell apoptosis by

TNFa: Modulation by inhibitors of protein synthesis. Exp Cell Res 214:584-594 22. Petrache I, Verin AD, Crow MT, et al (2001) Differential effect of MLC kinase in TNF-alpha

induced endothelial cell apoptosis and barrier dysfunction. Am J Physiol Lung Cell Mol Physiol 280:L1168-L1178

23. Seydel U, Ulmer AJ, Uhlig S, Rietschel ET (2000) Lipopolysaccharide, a membrane-forming and inflammation-inducing bacterial macromolecule. In: Zimmer G (ed) Membrane Structure in Disease and Drug Therapy. Marcel Dekker Inc., New York, pp 217-252.

24. Falk S, Goggel R, Heydasch U, et al (1999) Quinolines attenuate PAF-induced pulmonary pressor responses and edema formation. Am J Respir Crit Care Med 160:1734-1742

25. Gatti S, Faggioni R, Echtenacher B, Ghezzi P (1993) Role of tumour necrosis factor and reactive oxygen intermediates in lipopolysaccharide-induced pulmonary oedema and lethality. Clin Exp ImmunoI91:456-461

26. Mizgerd JP, Peschon JJ, Doerschuk CM (2000) Roles of tumor necrosis factor receptor signaling during murine Eschericia Coli pneumonia. Am J Respir Cell Mol Bioi 22:85-91

27. Nelson S, Bagby OJ, Bainton BG, et al (1989) Compartirnentalization of intraalveolar and systemic lipopolysaccharide-induced tumor necrosis factor and the pulmonary inflammatory response. JInfectDis 159:189-194


28. Dehoux MS, Boutten A, Ostinelli J, et al (1994) Compartimentalized cytokine production within the human lung in unilateral pneumonia. Am J Respir Crit Care Med 150:710-716

29. Haitsma 11, Uhlig S, Goggel R et al (2000) Ventilator-induced lung injury leads to loss of alveolar and systemic compartmentalization ofTNF-a.. Intens Care Med 26:1515-1522

30. Kitamura Y, Hashimoto S, Mizuta N, et al (2001) Fas/FasL-dependent apoptosis of alveolar cells after lipopolysaccharide-induced lung injury in mice. Am J Respir Crit Care Med 163:762-769

31. Matute-Bello G, Winn RK, Jonas M, et al (2001) Fas (CD95) induces alveolar epithelial cell apoptosis in vivo: implications for acute pulmonary inflammation. Am J PathoI158:153-161

32. Tesfaigzi J, Wood MB, Johnson NF, Nikula KJ (1998) Apoptosis is a pathway responsible for the resolution of endotoxin-induced alveolar type II cell hyperplasia in the rat. Int J Exp Pathol 79:303-311

33. Fehrenbach H, Kasper M, Tschernig T, et al (1999) Keratinocyte growth factor-induced hyperplasia of rat alveolar type II cells in vivo is resolved by differentiation into type I cells and by apoptosis. Eur Respir J 14:534-544

34. Fehrenbach H, Kasper M, Koslowski R, et al (2000) Alveolar epithelial type II cell apoptosis in vivo during resolution of keratinocyte growth factor-induced hyperplasia in the rat. Histochem Cell Bioi 114:49-61

35. Bardales RH, Xie SS, Schaefer RF, Hsu SM (1996) Apoptosis is a major pathway responsible for the resolution of type II pneumocytes in acute lung injury. Am J Pathol 149:845-852

36. Mallampalli RK, Peterson EJ, Brent Carter A, et al (1999) TNF-a. increases ceramide without inducing apoptosis in alveolar type II epithelial cells. Am J Respir Cell Mol BioI20:L481-L490

37. Kampf C, Relova AJ, Sandler S, Roomans GM (1999) Effects of TNF-a IPN-g and IL-lb on normal human bronchial epithelial cells. Eur Respir J 14:84-91

38. Trifilieff A, Futjitani Y, Coyle AJ, Bertrand C (1999) Fas-induced death of a murine pulmonary epithelial cell line: modulation by inflammatory cytokines. Fundam Clin Pharmacol13:656-661

39. Liu AN, Mohammed AZ, Rice WR, et al (1999) Perforin-independent CD8(+) T-cell-mediated cytotoxicity of alveolar epithelial cells is preferentially mediated by tumor necrosis factor-alpha: relative insensitivity to Fas ligand. Am J Respir Cell Mol BioI 20:849-858

40. Wang R, Alam G, Zagariya A, et al (2000) Apoptosis of lung epithelial cells in response to TNF-alpha requires angiotensin II generation de novo. J Cell PhysioI185:253-259

41. Boudreau N, Sympson CJ, Werb Z, Bissell MJ (1995) Suppression of ICE and apoptosis in mammary epithelial cells by extracellular matrix. Science 267:891-893

42. Boudreau N, Werb Z, Bissell MJ (1996) Suppression of apoptosis by basement membrane requires three-dimensional tissue organization and withdrawal from the cell cycle. Proc Natl Acad Sci USA 93:3509-3513

43. Wen LP, Madani K, Fahrni JA, et al (1997) Dexamethasone inhibits lung epithelial cell apoptosis induced by lPN-gamma and Fas. Am J PhysioI273:L921-L929

44. Thompson EB (1999) Mechanisms of T-cell apoptosis induced by glucocorticoids. Trends Endocrin Metabolism 10:353-358

45. Fine A, Andersonm NL, Rothstein TL, et al (1997) Fas expression in pulmonary alveolar type II cells. Am J PhysioI273:L64-L71

46. Gochuico BR, Williams MC, Fine A (1997) Simultaneous in situ hybridization and TUNEL to identify cells undergoing apoptosis. Histochem J 29:413-418

47. Matute-Bello G, Liles WC, Frevert CW, et al (2001) Recombinant human Fas ligand induces alveolar epithelial cell apoptosis and lung injury in rabbits. Am J Physiol Lung Cell Mol Physiol 281:L328-L335

48. Zhang J, Wenthold RJ, Yu ZX, et al (1995) Characterization of the pulmonary lesions induced in rats by human recombinant interleukin-2. Toxicol PathoI23:653-666

49. Rensing-Ehl A, Hess S, Ziegler-Heitbrock HW, et al (1995) Fas/Apo-l activates nuclear factor kappa B and induces interleukin-6 production. J Inflamm 45:161-174

50. Ponton A, Clement MY, Stamenkovic I (1996) The CD95 (APO-l/Fas) receptor activates NF-kappaB independently of its cytotoxic function. J BioI Chern 271 :8991-8995


51. Miwa K, Asano M, Horai R, et al (1998) Caspase I-independent IL-1beta release and inflammation induced by the apoptosis inducer Fas ligand. Nat Med 4:1287-1292

52. Chen JJ, Sun Y, Nabe1 GJ (1998) Regulation of the pro inflammatory effects of Fas ligand (CD95L). Science 282:1714-1717

53. Filippatos G, Tilak M, Piniloss H, Dhal BD (2001) Regulation of apoptosis by angiotensin II in the heart and lungs. Int J Mol Med 7:273-280

54. Wang R, Ramos C, Joshi I, et al (1999) Human lung myofibroblast-derived inducers of alveolar epithelial apoptosis identified as angiotensin peptides. Am J Physiol 277:L1l58-L1164

55. Wang R, Zagariya A, Ang E, et al (1999) Fas-induced apoptosis of alveolar epithelial cells requires ANG II generation and receptor interaction. Am J PhysioI277:L1245-L1250

56. Wang R, Ibarra-Sunga 0, Verlinski L, et al (2000) Abrogation of bleomycin-induced epithelial apoptosis and lung fibrosis by captopril or by a caspase inhibitor. Am J Physiol Lung Cell Mol PhysioI279:L143-L151

57. Edwards YS, Sutherland LM, Murray AW (2000) NO protects alveolar type II cells from stretch-induced apoptosis. A novel role for macrophages in the lung. Am J Physiol Lung Cell Mol PhysioI279:L1236-L1242

58. Yamamoto T, Ebe Y, Hasegawa G, et al (1999) Expression of scavenger receptor class A and CD14 in lipopolysaccharide-induced lung injury. Pathol Intern 49:983-992

59. Savill J, Fadok V (2000) Corpse clearance defines the meaning of cell death. Nature 407 :784-788 60. Henson PM, Bratton DL, Fadok VA (2001) The phosphatidylserine receptor: a crucial molecular

switch? Nat Rev Mol Cell BioI 2:627-633 61. Fadok VA, Bratton DL, Guthrie L, Henson PM (2001) Differential effects of apoptotic versus

lysed cells on macrophage production of cytokines: role of proteases. J Immuno1166:684 7 -6854 62. Liu Y, Cousin JM, Hughes J, et al (1999) Glucocorticoids promote nonphlogistic phagocytosis

of apoptotic leukocytes. J ImmunoI162:3639-3646 63. Giles KM, Ross K, Rossi AG (2001) Glucocorticoid augmentation of macrophage capacity for

phagocytosis of apoptotic cells is associated with reduced p130Cas expression, loss of paxillinlpyk2 phosphorylation, and high levels of active rac. J InununoI167:976-986

64. Bingisser R, Stey C, Weller M, et al (1996) Apoptosis in human alveolar macrophages is induced by endotoxin and is modulated by cytokines. Am J Respir Cell Mol BioI 15:64-70

65. Herlihy JP, Vermeulen MW, Hales CA (1996) Human alveolar macrophages prevent apoptosis in polymorphonuclear leukocytes. Am J Physiol271 :L681-L687

66. Wedmore CV, Williams TJ (1981) Control of vascular permeability by polymorphonuclear leukocytes in inflammation. Nature 289:646-650

67. Doherty DE, Downey GP, Worthen GS, et al (1988) Monocyte retention and migration in pulmonary inflammation. Requirement for neutrophils. Lab Invest 59:200-213

68. Spitznagel JK (1990) Antibiotic proteins of human neutrophils. J Clin Invest 86: 1381-1386 69. Haslett C, Jose PJ, Giclas PC et al (1989) Cessation of neutrophil influx in C5a-induced acute

experimental arthritis is associated with loss of chemoattractant activity from the joint space. J ImmunoI142:3510-3517

70. Hurley JV (1983) Termination of acute inflammation: 1. Resolution. In: Hurley JV (ed) Acute inflammation 2nd edn, Churchill Livingston, London.

71. Savill JS, Wyllie AH, Henson JE, et al (1989) Macrophage phagocytosis of aging neutrophils in inflammation. Programmed cell death in the neutrophil leads to its recognition by macrophages. J Clin Invest 83:865-875

72. Haslett C (1999) Granulocyte apoptosis and its role in the resolution and control of lung inflammation. Am J Respir Crit Care Med 160:S5-S 11

73. Lesur 0, Kokis A, Hermans C, et al (2000) Interleukin-2 involvement in early acute respiratory distress syndrome: relationship with polymorphonuclear neutrophil apoptosis and patient survival. Crit Care Med 28:3814-3822

74. Dunican A, Leuenroth SJ, Grutkoski P, et al (2000) TNFalpha-induced suppression of PMN apoptosis is mediated through interleukin-8 production. Shock 14:284-288


75. Aggarwal A, Baker CS, Evans TW, Haslam PL (2000) G-CSF and IL-8 but not GM-CSF correlate with severity of pulmonary neutrophilia in acute respiratory distress syndrome. Eur Respir J 15:895-901

76. Hashimoto S, Kobayashi A, Kooguchi K, et al (2000) Upregulation of two death pathways of perforin/granzyme and FasLlFas in septic acute respiratory distress syndrome. Am J Respir Crit Care Med 161:237-243

77. Burdon D, Tiedje T, Pfeffer K et al (2000) The role oftumor necrosis factor in the development of multiple organ failure in a mmine model. Crit Care Med 28: 1962-1967

78. Grigg JM, Savill JS, San"af C, et al (1991) Neutrophil apoptosis and clearance from neonatal lungs. Lancet 338:720-722

79. Cox G, Crossley J, Xing Z (1995) Macrophage engulfment of apoptotic neutrophils contributes to the resolution of acute pulmonary inflammation in vivo. Am J Respir Cell Mol BioI 12:232-237

80. Savill J, Haslett C (1995) Granulocyte clearance by apoptosis in the resolution of inflammation. Semin Cell BioI 6:385-393

81. Whyte MK, Meagher LC, MacDermot J, Haslett C (1993) Impairment of function in aging neutrophils is associated with apoptosis. J ImmunoI150:5124-5134

82. Meagher LC, Savill JS, Baker A, et al (1992) Phagocytosis of apoptotic neutrophils does not induce macrophage release of thromboxane B2. J Leukoc BioI 52:269-273

83. Dransfield I, Stocks SC, Haslett C (1995) Regulation of cell adhesion molecule expression and function associated with neutrophil apoptosis. Blood 85:3264-3273

84. Carson WE, Yu H, Dierksheide J, Pfeffer K, et al (1999) A fatal cytokine-induced systemic inflammatory response reveals a critical role for NK cells. J ImmunoI162:4943-4951

85. Hagimoto N, Kuwano K, Miyazaki H, et al (1997) Induction of apoptosis and pulmonary fibrosis in mice in response to ligation of Fas antigen. Am J Respir Cell Mol BioI 17:272-278

86. Chapman HA (1999) A Fas pathway to pulmonary fibrosis. J Clin Invest 104: 1-2 87. Kuwano K, Hagimoto N, Kawasaki M, et al (1999) Essential roles of the Fas-Fas ligand pathway

in the development of pulmonary fibrosis. J Clin Invest 104: 13-19 88. Aoshiba K, Yasui S, Tamaoki J, Nagai A (2000) The FaslFas-ligand system is not required for

bleomycin-induced pulmonary fibrosis in mice. Am J Respir Crit Care Med 162:695-700 89. Oberholzer A, Harter L, Feilner A, et al (2000) Differential effect of caspase inhibition on

proinflammatory cytokine release in septic patients. Shock 14:253-257 90. Karima R, Matsumoto S, Higashi H, Matsushima K (1999) The molecular pathogenesis of

endotoxic shock and organ failure. Mol Med Today 5:123-132 91. Li. P., Allen H, Banerjee S, et al (1995) Mice deficient in IL-l beta-converting enzyme are

defective in production of mature IL-l beta and resistant to endotoxic shock. Cell 80:401-411 92. Liu P, Anderson RGW (1995) Compartmentalized production of ceramide at the cell surface. J

BioI Chern 45:27129-27185 93. Zundel W, Swiersz LM, Giaccia A (2000) Caveolin I-mediated regulation of receptor tyrosine

kinase-associated phosphatidylinositol 3-kina~e activity by cerantide. Mol Cell BioI 20: 1507-1514 94. Dreyfuss D, Saumon G (1998) Ventilator-induced lung injury. Lessons from experimental

studies. Am J Respir Crit Care Med 157:294-323 95. Slutsky AS (1999) Lung injury caused by mechanical ventilation. Chest 116:9S-15S 96. Uhlig S, Uhlig U (2001) Molecular mechanisms of pro-inflammatory responses in overventilated

lungs. Recent Res Devel Resp Critical Care Med 1 :49-58 97. The Acute Respiratory Distress Syndrome Network (2000) Ventilation with lower tidal volumes

as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 342: 1301-1308

98. Held H-D, Boettcher S, Hamann L, Uhlig S (2001) Ventilation-induced chemokine and cytokine release is associated with activation of NFkB and is blocked by steroids. Am J Respir Crit Care Med 163:711-716

99. Edwards YS (2001) Stretch stimulation: its effects on alveolar type II cell function in the lung. Comp Biochem Physiol A Mol Integr Physiol 129:245-260


100.Imai Y, Kajikawa 0, Frevert C, et al (2001) Injurious ventilation strategies enhance organ apoptosis in rabbits. Am J Respir Crit Care Med 163:A677

101. Barkett M, Gilmore TD (1999) Control of apoptosis by Rel/NF-kB transcription factors. Oncogene 18:6910-6924

102. Matute-Bello G, Liles WC, Radella F, et al (1997) Neutrophil apoptosis in the acute respiratory distress syndrome. Am J Respir Crit Care Med 156: 1969-1977

103. Matute-Bello G, Liles WC, Radella F, et al (2000) Modulation of neutrophil apoptosis by granulocyte colony-stimulating factor and granulocyte/macrophage colony-stimulating factor during the course of acute respiratory distress syndrome. Crit Care Med 28: 1-7

104. Wollin L, Uhlig S, Nusing R, Wendel A (2001) Granulocyte-macrophage colony-stimulating factor amplifies lipopolysaccharide-induced bronchoconstriction by a neutrophil- and cyclooxygenase 2-dependent mechanism. Am J Respir Crit Care Med 163:443-450

105. Verhoef G, Boogaerts M (1991) Treatment with granulocyte-macrophage colony stimulating factor and the adult respiratory distress syndrome. Am J HematoI36:285-287

106. Demuynck H, Zachee P, Verhoef GE, et al (1995) Risks of rhG-CSF treatment in drug-induced agranulocytosis. Ann Hematol 70: 143-147

107. Gorgen I, Hartung T, Leist M, et al (1992) G-CSF treatment protects rodents against LPS-induced toxicity via suppression of systemic TNF-a. J ImmunoI1992:918-924

108. Tiegs G, Barsig J, Matiba B, et al (1994) Potentiation by granulocyte macrophage colony-stimulating factor of lipopolysaccharide toxicity in mice. J Clin Invest 93:2616-2424

109. Matute-Bello G, Liles WC, Steinberg KP, et al (1999) Soluble Fas ligand induces epithelial cell apoptosis in humans with acute lung injury (ARDS). J ImmunoI163:2217-2225

110. Albertine KH, Wang Z, McIntyre TM, Zimmermann GA (2001) Fas/Fas ligand interaction is associated with epithelial cell injury in ARDS. Am J Respir Crit Care Med 163:A306

111. Parsey MV, Kaneko D, Shenkar R, Abraham E (1999) Neutrophil apoptosis in the lung after hemorrhage or endotoxemia: apoptosis and migration are independent of interleukin-1 b. Chest 116:67S-68S

112. Xing Z, Gauldie J, Cox G, et al (1998) IL-6 is an antiinflammatory cytokine required for controlling local or systemic acute inflammatory responses. J Clin Invest 101:311-320

113. Klosterhalfen B, Hauptmann S, Offner FA, et al (1997) Induction of heat shock protein 70 by zinc-bis-(DL-hydrogenaspartate) reduces cytokine liberation, apoptosis, and mortality rate in a rat model of LD100 endotoxemia. Shock 7:254-262

114. Cox G (1996) IL-lO enhances resolution of pulmonary inflammation in vivo by promoting apoptosis of neutrophils. Am J Physiol 271 :L566-L571

115. Duffy AJ, Nolan B, Sheth K, et al (2000) Inhibition of alveolar neutrophil immigration in endotoxemia is macrophage inflammatory protein 2 independent. J Surg Res 90:51-57

The Role of Interleukin-10 During Systemic Inflammation and Bacterial Infection

F. N. LAUW, S.J. H. VAN DEVENTER, T. VANDER POLL

Sepsis is a clinical syndrome that results from a systemic response of the host to an infection. Activation of several host inflammatory mediator systems, including the cytokine network, is considered to play an important role in the pathogenesis of sepsis. Cytokines are a family of small proteins which function in a complex network in which they can influence each other's production and activity. Pro-inflammatory cytokines, of which tumour necrosis factor-a(TNF) and interleukin-l (lL-l) are studied most extensively, stimulate inflammatory processes and facilitate the immune response against invading pathogens. However, excessive systemic release of pro-inflammatory cytokines during sepsis syndrome has been found to contribute to the development of tissue damage. The production of these pro-inflammatory cytokines can be inhibited by so-called anti-inflammatory cytokines. The prototype of this group of mediators is IL-IO. In this article we will discuss the role of IL-IO in the pathogenesis of sepsis and severe bacterial infections.

Production of interleukin-lO

IL-IO was first identified as a protein produced by mouse T helper 2 (Th2) lymphocytes that suppressed the production of cytokines by Thl lymphocytes [1, 2]. IL-IO is secreted as a 18 kD homodimer by a variety of cells including T cells, B cells, monocytes and macrophages [3]. Stimuli that can induce IL-l 0 production are diverse and include bacteria, bacterial products (e.g., endotoxin), parasites, fungi and viruses. In addition, several cytokines can enhance IL-IO synthesis, including TNF-a, IL-l, IL-6 and IL-12.

Elevated plasma levels of IL-IO have been found in patients with sepsis [4-7]. While in healthy individuals, IL-I0 is not detectable in the circulation, detectable plasma concentrations of IL-l 0 are found in 80%-100% of patients with septic shock. Studies have reported that IL-I0 serum concentrations were

96 F. N. Lauw, S. J.H. van Deventer, T. van der Poll

related to the severity of sepsis and organ failure [7]. In patients with meningococcal septic shock higher IL-l 0 concentrations were found in non-survivors compared to patients who survived [5]. Also, during sequential measurements in patients with clinical sepsis, plasma IL-lO remained high in nonsurviving patients while in survivors, plasma IL-I0 significantly decreased [6].

Endotoxin, a lipopolysaccharide (LPS), is part of the outer membrane of all gram-negative bacteria. Endotoxin has potent proinflammatory properties and is capable of activating multiple inflammatory pathways, and is therefore considered to play a key role in the toxic sequelae of gram-negative sepsis. Intravenous injection of low dose endotoxin has been used as a human model of systemic inflammation [8]. In humans and nonhuman primates injected with endotoxin, the appearance of IL-l 0 in the circulation follows that ofTNF-a, and peaks after two to three hours [8, 9]. At least part of endotoxin-induced IL-lO release is dependent on TNF-a since injection of recombinant TNF-a into healthy humans induced a modest rise in plasma IL-lO concentrations and neutralization of TNF-a activity in endotoxin-treated chimpanzees by simultaneous infusion of an anti-TNF monoclonal antibody attenuated IL-lO secretion [9]. Other endogenous factors that may control IL-l 0 synthesis include prostaglandins, cortisol, catecholamines, and, as mentioned above, other cytokines [10-12].

Effects of IL-IO

The cellular targets of IL-lO include T-Iymphocytes, B-Iymphocytes, monocytes, macrophages, dendritic cells and epithelial cells. These effects may be exerted during cell-to-cell contact (i.e., in the case of regulatory T -lymphocytes, or macrophage-lymphocytes interactions) or at long-distance (i.e., during systemic inflammatory responses in sepsis). Many of the biological effects of IL-lO result in inhibition of T lymphocyte and monocyte-initiated inflammatory responses (Table 1). In vitro, IL-I0 is a potent inhibitor of production of TNF-a, IL-l, IL-6, IL-8, IL-12 and other cytokines [13, 14]. In addition, IL-lO impairs the antigen-presenting capacity of monocytes and macrophages, by decreasing expression of major histocompatibility complex (MHC) class II molecules as well as co-stimulatory molecules, and by decreasing release of

Table 1. Anti-inflammatory and immunosuppressive effects of interleukin-l 0 (IL-l 0)

Inhibition of cytokine production by macrophages, T lymphocytes and granulocytes

Inhibition of class II MHC expression by monocytes Inhibition of monocyte-dependent T helper cell proliferation Suppression of procoagulant activity Inhibition of production of reactive nitrogen oxides by macrophages Inhibition of arachidonic acid metabolites production Inhibition of killing of parasites and intracellular bacteria by macrophages

The Role of Interleukin-l 0 During Systemic Inflammation and Bacterial Infection 97

IL-12 [15, 16]. Furthermore, IL-lO is capable of attenuating the expression or production of a number of other mediators of inflammation, including tissue factor, and nitric oxide [17, IS].

Several studies have demonstrated the anti-inflammatory potential ofIL-lO in vivo. Administration of recombinant IL-l 0 prior to injection of a lethal dose of endotoxin to mice markedly suppressed TNF-a release, and prevented lethality [19, 20]. In healthy humans, recombinant human IL-lO, given as a single dose of25 J.1g/kg immediately prior to endotoxin, reduced the rise in body temperature and in plasma TNF-a, IL-6 and IL-S concentrations [21]. In addition, IL-I0 treatment diminished granulocyte degranulation and prevented endotoxin-induced granulocyte accumulation in lungs. Further, IL-I0 also inhibited the activation of the fibrinolytic system and the coagulation system [22].

Recent studies have suggested that under certain conditions IL-l 0 also has immunostimulatory properties. In vitro IL-lO was shown to have stimulatory effects on CD4+, CDS+ T cells and/or NK cells resulting in enhanced IFN-y production [23, 24]. During experimental human endotoxemia it was found that IL-lO, especially when administered 1 h after the injection of endotoxin, enhanced endotoxin-induced IFN-y release, as well as the release of the IFN-y-dependent chemokines IP-lO and Mig, while inhibiting or not influencing the production ofIFN-y-inducing cytokines [25]. Administration ofIL-lO did also result in increased endotoxin-induced release of soluble granzymes suggesting that IL-I0 treatment results in enhanced activation of cytotoxic lymphocytes leading to increased IFN -y production. Therefore, although IL-l 0 is well-known forits anti-inflammatory activities it should be taken into account that under certain conditions, IL-lO may have immunostimulatory effects.

Role of IL-IO during systemic inflammation

Neutralization of endogenously produced IL-l 0 in endotoxemic mice, resulted in an increased production of several proinflammatory cytokines, including TNF-a, and an enhanced mortality [26, 27]. Similarly, IL-lO gene deficient mice showed an increased mortality after endotoxin administration together with elevated levels of TNF-a, IL-l, IL-6, IL-12, IFN-g and nitrate [2S]. A number of cytokines may contribute to the increased endotoxin-induced lethality in the absence of endogenous IL-lO, since it can be prevented in part by antibodies directed against TNF-a, IFN -g or macrophage inflammatory protein-2. Hence, during systemic inflammation induced by bolus administration of endotoxin, endogenous IL-I0 represents an important autoregulatory mechanism controlling the production of proinflammatory cytokines and endotoxin

98 F. N. Lauw, S. J.R. van Deventer, T. van der Poll

toxicity in vivo. In addition, IL-IO was found to play an important immunoregulatory role during superantigen (SAg)-induced pathology. Injection of antiIL-IO prior to administration of Staphylococcal enterotoxin B (SEB), a bacterial SAg, resulted in enhanced serum levels of SEB-induced IFN-y and IL-2, which was associated with increased lethality [29]. Likewise, the inflammatory response was more pronounced in IL-I O-deficient mice and these mice were more susceptible to SEB-induced lethal shock compared to wild type mice [30].

However, it is important to note that intravenous injection of endotoxin or live bacteria results in an acute syndrome, unlike many cases of sepsis in which a subacute or intermittent course is noted. Furthermore, clinical sepsis almost invariably is the result of an infection that was initially localized in an organ or body cavity. When discussing the role of cytokines in the pathogenesis of sepsis, it is important to consider the function of these mediators during localized infections.

Role of IL-IO in experimental models of localized infection

Mouse models indicate that proinflammatory cytokines (e.g., TNF-a, IL-6), produced at the site of infection, playa crucial role in the innate immunity to bacterial infection. Neutralization of IL-I 0 has been found to augment bacterial clearance from the lungs and to improve survival during murine pneumonia with Klebsiella pneumoniae or Streptococcus pneumoniae [31, 32]. Thus in contrast with its anti-inflammatory and protective effects in systemic models of inflammation, IL-I 0 hampers an adequate proinflammatory response crucial for effective clearance of an infectious agent in a localized infection (Fig. 1). A model of murine Escherichia coli peritonitis has provided insight into the seemingly paradoxical role ofIL-lO in severe bacterial infection. Intraperitoneal administration of live Escherichia coli in mice results in a subacute syndrome characterized by local production of cytokines in the peritoneal cavity, influx of granulocytes to the primary site of infection, and eventually in systemic inflammation, multiple organ failure and death. IL-IO gene deficient mice demonstrated an enhanced bacterial clearance from the abdominal cavity and a diminished dissemination of the infection to distant organs in this model of septic peritonitis (Fig. 2A) [33]. Despite these findings, which are strictly in line with the role of endogenous IL-IO in pneumonia, systemic inflammation, including enhanced TNF-a release, and multiple organ failure were more prominent in IL-IO knockout mice, and lethality was increased. Treatment of IL-IO deficient animals with anti-TNF-a protected against the enhanced lethality seen in IL-IO deficient mice treated with a control antibody (Fig. 2B). Hence, endogenous IL-I 0 protected against lethality during abdominal sepsis,

The Role of Interleukin-lO During Systemic Inflammation and Bacterial Infection 99

Fig. 1. The role of endogenous interleukin-10 (IL-10) in the pathogenesis of endotoxin shock and pneumococcal pneumonia in mice. Neutralization of IL-lO during sublethal endotoxemia in mice results in a 60% lethality (A), whereas administration of IL-lO during murine pneumococcal pneumonia protects against lethality (B). (From Marchant A, Bruyns C, Vandenabeele P, et al. IL-lO controls IFN-y and TNF-a production during experimental endotoxemia. Eur J Immunol 1994;24:1167-1171 and Van derPoll T, Marchant A, Buurman WA, et al. Endogenous interleukin 10 protects mice from death during septic peritonitis. J Immunol1995;155:5397-540l with permission)

100 F. N. Lauw, S. I.H. van Deventer, T. van der Poll

A Peritoneal fluid Blood

SOOOO 30000

..J 10000 E SOO -:::J LL ()

B

400

300 200 100

6h

P< O.OS

• • •

20h

I • IL-10+1+

SOu • 600

400 • o o

200 ........ · ", ...c~ ••• •• -~. -... (I-, ~~r.

u 6h 20h

o IL-10-1-

100._------------. ---..... •• --------e - 7S ~ 0

.. ,--co > SO .s: I- - .. -IL-1 0+1+ ,co mAb :::s

P < O.OS

UJ -o-IL-10-1-,anti-TNF 2S -0- IL-1 O-I-,co mAb

---_ ... ,

O+---~~--_r----~--~~--_r--~

0.0 O.S 1.0 1.S 2.0 2.S 3.0

Days after E. coli challenge

Fig. 2. The role of endogenous interleukin-l 0 (IL-I0) during E. coli peritonitis in mice. IL-l 0 gene deficient (IL-l 0-1-) mice demonstrate an enhanced bacterial clearance from the abdominal cavity and a diminished dissemination of the infection to distant organs (A). However, lethality during E. coli peritonitis is increased in IL-l 0-1- mice, which is associated with enhanced local and systemic tumor necrosis factor (TNF) production, and reversed when IL-l 0-1- mice are treated with anti-TNF-a (B). (From Sewnath ME, Olszyna DP, Birjmohun R, et al. (2001) IL-IO-deficient mice demonstrate multiple organ failure and increased mortality during Escherichia coli peritonitis despite an accelerated bacterial clearence. J ImmunoI166:6323-6331 with permission)

The Role of Interleukin-l 0 During Systemic Inflammation and Bacterial Infection 101

in spite of hampering antibacterial effector mechanisms, by a mechanism that involved inhibition ofTNF production, and hereby controlling systemic inflammation and organ failure.

Conclusion

During overwhelming systemic inflammation such as after injection of endotoxin or bacteria or during meningococcal septic shock, endogenously produced IL-IO exerts a protective effect by inhibiting the release of proinflammatory cytokines into the circulation and diminishing systemic inflammation and organ failure. In contrast, during localized infections such as pneumonia and peritonitis, endogenously produced IL-IO hampers an appropriate host response to invading micro-organisms. Hence, the effects of IL-l 0 during bacterial infections are complex and likely determined by the source of the infection and the functional balance between proinflammatory and anti-inflammatory forces within the cytokine network.

References 1. Fiorentino DF, Bond MW, Mosmann TR (1989) Two types of mouse helper T cell. IV. Th2 clones

secrete a factor that inhibits cytokine production by Thl clones. J Exp Med 170:2081-2095 2. Y ssel H, De Waal Malefyt R, Roncarolo MG, et al (1992) IL-1 0 is produced by subsets of human

CD4+ T cell clones and peripheral blood T cells. J Immunol149:2378-2384 3. Moore KW, de Waal MalefytR, Coffman RL, et al (2001) Interleukin-l0 and the Interleukin-l0

receptor. Annu Rev Immunol19:683-765 4. Marchant A, Deviere J, Byl B, et al (1994) Interleukin-1 0 production during septicaemia. Lancet

434:707-708 5. Derkx B, Marchant A, Goldman M, et al (1995) High levels of interleukin-l0 during the initial

phase of fulminant meningococcal septic shock. J Infect Dis 171 :229-232 6. van der Poll T, de Waal Malefyt R, Coyle SM, et al (1997) Antiinflammatory cytokine responses

during clinical sepsis and experimental endotoxemia: sequential measurements of plasma soluble interleukin-1 (IL )-1 receptor type II, IL-lO and IL-13 concentrations. J Infect Dis 175: 118-122

7. Friedman G, Jankowski S, Marchant A, et al (1997) Blood interleukin-lO levels parallel the severity of septic shock. J Crit Care 12: 183-187

8. van der Poll T, van Deventer SJH (1999) Endotoxemia in healthy subjects as a human model of inflammation. In: Cohen J, Marshall J (ed) The Immune Response in the Critically III :335-357

9. van der Poll T, Jansen J, Levi M, et al (1994) Regulation of interleukin 10 release by tumor necrosis factor in humans and chimpanzees. J Exp Med 180:1985-1988

10. van der Poll T, Coyle SM, Barbosa K, et al (1996) Epinephrine inhibits tumor necrosis factor-alpha and potentiates interleukin-l 0 production during human endotoxemia. J Clin Invest 97:713-719

11. van der Poll T, Barber AE, Coyle SM, et al (1996) Hypercortisolemia increases plasma interleukin-lO concentrations during human endotoxemia clinical research center study. J Clin Endocrin Metab 81:3604-3606

12. van der Pouw Kraan TCTM, Boeije LCM, Smeenk RJT, et al (1995) Prostaglandin-E2 is a potent inhibitor of human interleukin-12 production. J Exp Med 181:775-779

102 F. N. Lauw, S. J.H. van Deventer, T. van der Poll

13. de Waal Malefyt R, Abrams J, Bennett B, et al (1991) Interleukin 10 (IL-lO) inhibits cytokine synthesis by human monocytes: an autoregulatory role of IL-l 0 produced by monocytes. J Exp Med 174:1209-1220

14. D. Andrea A, Aste-Arnezaga M, Valiante NM, et al (1993) Interleukin 10 (IL-I0) inhibits human lymphocyte interferon gamma-production by suppressing natural killer cell stimulatory factorlIL-12 synthesis in accessory cells. J Exp Med 178:1041-1048

15. de Waal MalefytR, HaanenJ, Spits H, et al (1991) Interleukin 10 (IL-lO) and viral IL-lO strongly reduce antigen-specific human T cell proliferation by diminishing the antigen-presenting capacity of monocytes via downregulation of class II major histocompatibility complex expression. J Exp Med 174:915-924

16. Willems F, Marchant A, Delville JP, et al (1994) Interleukin-lO inhibits B7 and intracellular adhesion molecule-l expression on human monocytes. Eur J Immunol 24: 1007 -1009

17. Pradier 0, Gerard C, Delvaux A, et al (1993) Interleukin-lO inhibits the induction of monocyte procoagulant activity by bacterial lipopolysaccharide. Eur J Immunol 23:2700-2703

18. Gazzinelli RT, Oswald IP, James SL, et al (1992) IL-lO inhibits parasite killing and nitrogen oxide production by IFN-g activated macrophages. J Immunol 148: 1792-1796

19. Gerard C, Bruyns C, Marchant A, et al (1993) Interleukin 10 reduces the release oftumornecrosis factor and prevents lethality in experimental endotoxemia. J Exp Med 177:547-550

20. Howard M, Muchamuel T, Andrade S, et al (1993) Interleukin 10 protects mice from lethal endotoxemia. J Exp Med 177: 1205-1208

21. Pajkrt D, Camoglio L, Tiel-van Buul MC, et al (1997) Attenuation of proinflammatory response by recombinant human IL-IO in human endotoxemia: effect of timing of recombinant human IL-lO administration. J ImmunoI158:3971-3977

22. Pajkrt D, van der Poll T, Levi M, et al (1997) Interleukin-lO inhibits activation of coagulation and fibrinolysis during human endotoxemia. Blood 89:2701-2705

23. Groux H, Bigler M, de Vries JE, et al (1998) Inhibitory and stimulatory effects ofIL-l 0 on human CD8+ T cells. J ImmunoI160:3188-3193

24. Shibata Y, Foster LA, Kurimoto M, et al (1998) Immunoregulatory roles of IL-lO in innate immunity: IL-lO inhibits macrophage production ofIFN-gamma-inducing factors but enhances NK cell production ofIFN-gamma. J ImmunoI161:4283-4288

25. Lauw FN, Pajkrt D, Hack CE, et al (2000) Proinflammatory effects of IL-lO during human endotoxemia. J ImmunoI165:2783-2789

26. Marchant A, Bruyns C, Vandenabeele P, et al (1994) IL-lO controls IFN-? and TNF production during experimental endotoxemia. Eur J Immunol 24: 1167 -1171

27. Standiford TJ, Strieter RM, Lukacs NW, et al (1995) Neutralization of IL-10 increases lethality in endotoxemia. Cooperative effects of macrophage inflammatory protein-2 and tumor necrosis factor. J Immunol 155

28. Berg DJ, KUhn R, Rajewsky K, et al (1995) Interleukin-l 0 is a central regulator of the response to LPS in murine models of endotoxic shock and the Schwartzman reaction but not endotoxin tolerance. J CLin Invest 96:2339-2347

29. Florquin S, Amraoui Z, Abramowicz D, et al (1994) Systemic release and protective role of IL-I0 in staphylococcal enterotoxin B-induced shock in mice. J Immunol 153:2618-2623

30. Hasko G, Virag L, Egnaczyk G, et al (1998) The crucial role of IL-lO in the suppression of the immunological response in mice exposed to enterotoxin B. Eur J ImmunoI28:1417-1425

31. Greenberger MJ, Strieter RM, Kunkel SL, et al (1995) Neutralization ofIL-lO increases survival in a murine model of Klebsiella pneumonia. J ImmunoI55:722-729

32. van der Poll T, Marchant A, Keogh CV, et al (1996) Interleukin-l 0 impairs host defense in murine pneumococcal pneumonia. J Infect Dis 174:994-1000

33. Sewnath ME, Olszyna DP, Birjmohun R, et al (2001) IL-l O-deficient mice demonstrate multiple organ failure and increased mortality during Escherichia coli peritonitis despite an accelerated bacterial clearence. J ImmunoI166:6323-6331

Cardiovascular Surgery. Modulation of the Inflammatory Reaction

P. GIOMARELLI, S. SCOLLETIA, E. BORELLI

Incomplete myocardial protection and occasional damaging effects of the cardiopulmonary bypass (CPB) are the causes of most adverse reactions in cardiac surgery. Incomplete myocardial protection is the cause of ischaemic complications and/or low cardiac output. CPB is responsible for systemic inflammatory response syndrome (SIRS), which can lead to multi-organ failure (MOF) [1].

The incidence of MOF in 174,806 patients after coronary arterial bypass grafts in 1997 was 1.4 in patients with a predicted risk of mortality from 5 to 10% and 6.9 in those with a predicted risk of mortality from 30 to 50%. The source of this data is the STS (Society of Thoracic Surgery USA) National Cardiac Surgery Data Base.

Systemic metabolic and inflammatory response

During cardio-pulmonary bypass, the metabolic response or rather the EE (energy expenditure) are obviously decreased. In the hours following the bypass the V02 (oxygen consumption) and VC02 (C02 production) and consequently the EE, show a marked increase which, however, return to normal after the 6th

to 9th hour in the intensive care unit (lCU) [2]. During hypermetabolic response, the heart rate, body temperature and haematological markers signify a systemic inflammatory response.

In a study conducted on 90 patients using a cluster analysis three groups were identified. Those with an increased VC02 index in which the VAC02D ranges from normal to incremented show a clear hypermetabolic response associated with either a normal or insufficient cardiovascular response. It is these patients that have a more marked risk of SIRS [3].

During extra-corporeal circulation we studied the correlation between the V021 and the DO~ (oxygen delivery) and observed a slight correlation in the group 1 patients which is characterized by an extraction ratio (ER) < 0.4.

104 P. Giomarelli, S. Scolletta, E. Borrelli

The second group, characterized by an ER > 0.4, the correlation during extracorporeal circulation between the VOzI and the DOzI presents an "r" of 0.71 and a "p" ofless than 0.001. The dependency of the oxygen consumption on oxygen delivery leads one to think that these patients, during the extracorporeal circulation, may risk tissue damage related to alterations of perfusion [4]. This might be attributable to the presence of a continual flow instead of a pulsating flow.

Indeed, medical literature shows that the limit of the extraction ratio is 0.7, while during extracorporeal circulation, Dantzer has demonstrated that beyond 0.4 a pathological condition exists [5].

Oudemans studied the oxygen consumption in twenty-one consecutive male patients undergoing elective coronary artery bypass surgery and separated these patients into two distinct groups: the first being characterized by a slight variation of O2 consumption and the second characterized by elevated variations of O2 consumption in the first hours after admission to ICU [6].

It is interesting to note how circulating levels of endotoxin, tumor necrois factor (TNF) and Interleukin 6 (IL-6) in patients with elevated variations of O2

consumption are significantly superior with respect to those of the group with inferior levels of O2 consumption. The endotoxin is considered by some authors to be the catalyzing element of the inflammatory cascade and the initial increase is attributed to contact between blood, extraneous surfaces and priming liquids. TNF and IL-6, which increase afterwards, are considered pro-inflammatory citokines that induce responses in the organism commonly known as SIRS.

Song Wan studied forty-four patients selected for off-pump coronary artery bypass grafting. Twenty-six were operated on using CPB, eighteen were operated without CPB. The patients that were operated on using CPB presented plasma levels superior to the others. IL-8 is statistically significant in 4 out of 7 intervals while IL-IO is significant only at the end of the operation. IL-8 is also considered to be a pro-inflammatory citokine, while IL-IO is an anti-inflammatory citokine.

No statistical differences were found with regards to CK-MB in patients undergoing coronary arterial bypass surgery with CPB or without. The increase at 24 hours after surgery was normal and the same in both groups [7].

Even more interesting is the behaviour of the cardiac troponin I (c-Tnl), which is a sensitive marker for cardiac ischaemia that reaches higher significant levels in patients who have undergone CPB. This means that those who underwent CBP have a higher risk of ischaemic lesions than those without. This does not usually correspond to daily practice because the surgeon's work is facilitated during CPB by a non-beating heart.

Song Wan's findings reveal a good correlation between levels of IL-8 and

Cardiovascular Surgery. Modulation of the Inflammatory Reaction 105

those of c-Tn! four hours after coronary arterial bypass grafting with and without CPB. According to this data the inflammatory response appears to be the responsible factor for myocardial damage.

The post-bypass response is therefore exaggerated, confusing and complex, and characterized by large fluid shifts, temperature changes, coagulation disturbances and increased concentrations of catecholamines and stress hormones. In this model the adhesion of neutrophil-endothelial cells is central to postischaemic reperfusion. Pro-inflammatory cytokines such as IL-l beta and TNF alfa, and chemokines such as IL-8, largely within endothelial cells, in reperfused organs within hours, and induce increased expression of adhesion molecules on endothelial cells and other tissue-specific cells (for example: myocytes, pulmonary alveolar lining cells, glomerular or renal tubular epithelium). In tum, this results in neutrophil attachment, transmigration of neutrophils into the interstitial space, and the release of large amounts of free radicals [8].

All these events play a certain role in the complex inflammatory cascade which may further lead to post-operative complications [9].

The studies that have been done throughout the years on perioperative inflammatory response can be categorized in to four stages: 1) observational studies demonstrating complement activation, the release of cytokines and the increased concentrations of neutrophils, endotoxins and elastases; 2) the determination of relationship between inflammatory markers and adverse clinical outcome; 3) the identification oftherapeutic agents that might mitigate inflammatory response; 4) clinical trials involving larger populations [13].

Cytokine production from isolated and stimulated mononuclear cells during cardiac operation

To assess whether heparin coating of bypass circuits affected cytokine release, IL-I0, IL-6 and IL-2 production was measured on the supernatants ofPHA-stimulated PBMC after 24 hours of incubation. A significant increase is evident in the production of IL-6 in both patient groups. However, 24 hours after admission in the postoperative intensive care unit, patients with heparin-coated circuits showed a lower production of IL-6 when compared with patients with uncoated circuits (p<O.05). IL-6 released by phytohemoagglutinin (PHA) stimulated perpheral blood mononuclear cells (PBMC) in patients with heparincoated circuits and with conventional circuits was 11.223±3.912 pg/ml and 18.129± 7.179 pg/ml (p=O.002), respectively. A more significant difference was noted for IL-IO production in patients with heparin-coated CPB. As expected, a significant increase in the production of IL-l 0 was observed in patients with


uncoated circuits. On the other hard, patients with heparin-coated CPB circuits did not show any increase in the production of IL-lO (p<O.05) 24 hours after ICU admission. IL-lO released by PHA-stimulated PBMC in patients with heparin-coated circuits and with conventional circuits was respectively 404±178 pg/ml and 1.403±897 pg/ml (p=O.008). Results regarding IL-6 and IL-10 release are presented in Figure 1.

To see whether the release of a Th 1-type cytokine was affected by heparincoated circuits, IL-2 production was evaluated in PHA-stimulated PBMC. IL-2

* 25000

I 20000

15000

3 10000

5000

0 BeforeCPB EndCPB 24Hrs

2500 * =- 2000

i 1500 -0 1000 '\""

~ 500

0 BeforeCPB EndCPB 24Hrs

Fig. 1. Effect of CPB on IL-6 and IL-IO production from PBMe. PBMC were cultured at concentration of 2x105 cell/well with PHA (5 ug/ml) and cell free supernatant was obtained after 24 hours of culture. Data are presented as mean ± standard deviation of IL-6 present in the supernatants as determined by ELISA. Asterisk indicates statistically significant differences (p<O.05) between group 1 (coated circuits) and group 2 (uncoated circuits) 24 hours after admission to the ICU.


production decreased in both patient groups; however, no significant difference was observed in the patients with heparin-coated circuits as compared with uncoated CPB. IL-2 released by PHA-stimulated PBMC in patients of group 1 and group 2 was 72±44 pglml and 113±78 (p<0.21) respectively. These results suggest that heparin-coated circuits did not affect Thl-type responses. To confirm that CPB effects were transient, IL-I0, IL-6 and IL-2 were repeated one week after CPB. As expected, after one week IL-I0, IL-6 and IL-2levels were no longer significantly modified with respect to preoperative values [10, 11]. In a prospective, randomized study 886 high-risk patients who had cardiac operations with CPB were enrolled in an Italian multicentre study. They were randomly allocated to have cardio-pulmonary bypass with heparin-coated circuits (Duraflo II) or conventional circuits.

Heparin-coated circuits are associated with a shorter ICU and postoperative hospital stay and with a lower rate of patients having a severely impaired clinical outcome (SICa, stay in intensive care unit for more than 5 days or until death) (relative risk 0.66, p= 0.045) [12].

Steroids and fast-track management

Our institution is currently participating in a Fast-Track management trial in conjunction with six other Italian Centres all of which are part of the study group called Gruppo di Studio per l' Anestesia in Cardiochirugia. The aim of this study is to evaluate the advantages of the use of peri operative steroids in Fast-Track management. As the steroid itself has an anti-inflammatory action, it may on one hand, have a positive effect in modulating the inflammatory response, or it may reveal a reduction of immuno-competency.

In a group of 20 patients that we evaluated according to the Fast-Track protocol, [13, 14, 15, 16] 10 were given steroids and the other 10 no steroids. We compared changes in the following parameters: PCR, Protein C, TNF, IL-6, IL-8, and IL-lO in the following moments: T I preoperative, T 2 at the end of extra-corporeal circulation, T 3 upon arrival in ICU, and T 4 through T 7 during the stay in ICU from the fourth to the seventh hour.

In Protein C Reactive there is a significant difference in untreated patients at T 6 and T 7, which confirms the presence of an accentuated inflammatory response.

The TNF in the control group, while superior to the treated group, only assumes statistical significance at T 6 and T 7. The lack of significant difference, at T 4 and T 5, might be attributed to the time variable of response in the individual patient at the time of evaluation.


IL-6 assumes statistical significance at T 4 and T 6. In the patients without cortisone treatment, inflammatory response is accentuated.

IL-8, the third most important pro-inflammatory interleukin, confirms the aforementioned results for IL-6.

IL-lO, the typical anti-inflammatory interleukin, shows values that are considerably higher in the cortisone group (Fig. 2).

It is interesting to note that the O2 arterial venous differences (a-v02) are

100

80 e

~ 60 a. --CD 40 ~

20

0 T1 T2

350

e 300

~ 250 a. 200 --0 .... 150 ~

100

50

0 T1 T2

** **

T3 T4 T5

** **

**

T3 T4 T5

• Control

o Cortisone

T6 T7

• Control

o Cortisone

T6 T7

Fig. 2. lL-6 and lL-IO concentration in plasma before, after CPB and during the stay in lCU. Data are presented as mean ± standard error. Asterisks indicate statistically significant differences (lL-6: *p<0.05; **p<O.O 1) (lL-l 0: *p<O.O 1; **p<O.OO 1) between the control group and the cortisone group.


notably significant at T 1,2 hours after admission into the ICU. The group that was not treated with cortisone has a value of about 6 mI, while the treated group has a value of 5 ml. This indicates that the cardiovascular system's capacity to satisfy the body's energy requirements remains within the physiological limits more often in the patients treated with cortisone. The values in the later time intervals, however, run parallel.

With regards to the oxygen extraction ratio at the first time interval, the group treated with cortisone has a level of 35% while the group that was not treated exhibits a value of 40%. The 35% is certainly a physiological value whereas the 40%, in those patients who underwent a CPB is, according to some authors, pathological, thereby indicating an increased effort on the part of the cardiovascular system to achieve the body's energy requirements [5].

Conclusion

• CPB constitutes a model of inflammatory cascade independent of infections. • The stress response measured in inflammation, immunity, coagulation, and

regeneration demonstrated a close relationship between cytokine and organ failure.

• Heparin coating of extracorporeal circuits inhibits cytokyne release from mononuclear cells during cardiopulmonary bypass and reduce the rate of patients having severe clinical outcame

• The corticosteroids are able to modulate this response, which reduces the presence of pro-inflammatory or immuno-competent cytokines.

References

1. Kirklin JK (1991) Prospect of understanding the deleterious effects of CPB. Ann Thorac Surg 51:523-529.

2. Chiara 0, Giomarelli PP, Biagioli B et al (1987) Hypermetabolic responce after hypothermic cardiopulmonary bypass. Crit Care Med 15-985/1000.

3. Giomarelli PP, Chiara 0, Rosi R (1988) Hemodynamic and Metabolic Responce after cardiopulmonary bypass: Continuous Monitoring of C02 Production. Update in Intensive Care Emergency Medicine JL Vincent Editor Springer - Verlag Bruxelles 341-356.

4. Giomarelli PP, Borrelli E, Marchetti L et al 02 Extration Ratio and cytokine production during cardiopulmonary bypass Supp 27:201.

5. Dantzker DR (1993) Adequacy of tissue oxygenation. Crit Care Med 21:540-43. 6. Ondemans-van Straaten HM, Jansen PG et al (1996) Increased oxygen consumption after cardiac

surgery is associated with inflanunatory response to endotoxemia. Intensive Care Med 22:294-300 7. Whan S, Irrat MB (1999) Avoiding cardiopulmonary bypass in Multivessel CABG reduces

cytokine response and myocardial Injury. Ann Thorac Surg 68:52-7.


8. Herskowitz A, Mangano DT (1996) Inflammatory cascade a final common Pathway for periooperative Injury? Anesthesiology 85:957-60.

9. Whan S, LeClercJ R, Vincent JL (1997) Cytokine responser to cardiopulmonary bypass in year book of Int. Care and Em. Med Vincent JL Ed Springer Berlin.

10. Giomarelli PP, Naldini A, Biagioli B et al (2000) Heparin coating of extracorporeal circuits inhibits cytokine release from mononuclear cells during cardiac operations. The International Journal of Artificial Organs. 23:250-255.

11. Naldini A, Borrelli E, Carraro F et al (1999) Interleukin 10 production in patients undergoing cardiopulmonary bypass: evidence of inhibition of Th I-type responses. Cytokine 11:74-79.

12. Ranucci M, Mazzucco A, Pessotto Ret al (1999) Heparin-coated circuits for high-risk patients: a multicenter, prospective, randomized trial. Ann Thorac Surg 67:994-1000.

13. Cheng DC (1998) Fast Track Cardiac Surgery Pathways. Anesthesiology 88:1429-33. 14. Cheng DCH, Karski J, Peniston C et al (1996) Early tracheal extubation after coronary artery

bypass graft surgery reduces costs and improves resource use: a prospective randomized controlled trial. Anesthesiology 85: 1300-10.

15. London MJ, Shroyer AL, Shoyer JR et al (1998) Early extubation following cardiac surgery in a veterans population. Anesthesiology 88:1447-58.

16. Tabardel Y, Duchateau J, Schmartz D (1996) Corticosteroids increase blood interleukin 10 levels during cardiopulmonary bypass in men. Surgery 119:76-80.

Microcirculation in Critical Illness

D. DE BACKER, M.-1. DUBOIS, 1. CRETEUR

Multiple organ failure is frequently observed in critically ill patients, despite the restoration of whole-body haemodynamics. Alterations in microvascular blood flow may playa crucial role in the development of mUltiple organ failure in these patients. These alterations can have important implications. In rats submitted to 60 min of severe haemorrhage with subsequent restauration of blood volume, Zhao et al. [I] observed that microvascular alterations were more severe in rats that will subsequently die compared to survivors, despite similar whole-body haemodynamics.

Evidence for microcirculatory alterations in experimental studies

Numerous experimental studies reported that microvascular blood flow is altered in various conditions, especially in sepsis [2-8]. Baker et al. [3] observed that endotoxin in rats elicited a severe arteriolar and venular vasocostriction. Similarly, Drazenovic et al.[9] reported that capillary density decreased after endotoxin administration in dogs. In a normodynarnic sepsis model obtained by cecal ligation and perforation in rats, Lam et al. [4] observed in striated muscles that the perfused capillary density decreased and that the number of capillaries with stopped-flow increased. They also reported that the heterogeneity of the spatial distribution of perfused capillaries increased. In the same model, Farquhwar et al. [7] subsequently reported a decrease in the number of perfused capillaries in the small bowel mucosa.

Microvascular blood flow alterations may be responsible for alterations in tissular metabolism, but one can also consider that flow is matching metabolism, with direct metabolic. alterations. It is difficult to separate these two opposite alternatives, but various animal studies have reported that microvascular alterations lead to major physiopathological implications. First, the coexistence of well perfused and non-perfused capillaries will lead to a marked heterogeneity in blood flow which may be responsible for the decrease in

112 D. De Backer, M.-J. Dubois, J. Creteur

oxygen extraction capabilities that is observed in sepsis [9-11]. Second, microvascular alterations are associated with zones of decreased intravascular P02

[12, 13], which is not compatible with primary metabolic alterations. Finally, the transient flow observed in some capillaries may lead to focal areas with ischaemia / reperfusion injury.

Multiple causes can be evoked to explain these microvascular alterations. First, various inflammatory mediators may be involved. In rats, Vicaut et al. [14] elicited a decrease in microvascular blood flow by the administration of tumor necrosis factor (TNF), a central mediator of sepsis. Also, endothelin, a potent vasoconstrictor often found to be elevated in patients with sepsis [15], can cause microvascular vasoconstriction. On the contrary, nitric oxide seems to have a protective role since mice lacking inducible nitric oxide presented less severe microvasular alterations after cecal ligation and puncture than normal mice [16]. Second, microthrombi can transiently occlude microvessels, and microthrombi formation is facilitated in septic conditions [17, 18]. This mechanism is strongly highlighted by the results of a recent study demonstrating that the administration of activated protein C significantly improved survival in patients with severe sepsis [19]. Third, sepsis impairs the deformation of leukocyte [20] and erythrocyte [21, 22] and promotes adhesion of leukocytes to endothelial cells [22, 23]. Fourth, interstitial oedema may compress small vessels. However, Piper et al. [6] observed that the increase in erythrocyte flow heterogeneity after cecal ligation and perforation in rats was not related to tissue oedema as measured by wet-to-dry ratio and albumin flux. Similarly, DIOrio et al. [24] reported that fluid loading improved oxygen extraction capabilities in endotoxic animals. Hence, it is likely that many of these mechanism contributed to the microvascular alterations, with vasoconstriction in some capillaries playing a predominant role.

Methods of investigating the microcirculation in critically ill patients

Most of the experimental studies were performed using intravital microscopy, the gold standard technique for studying the microcirculation. Unfortunately, this technique cannot be used in humans, as large microscopes are generally applied on a fixed tissue preparation while fluorescent dyes are infused. Alternative methods have been used in humans, including videomicroscopy of the nailfold area and laser Doppler technique. An extended review of the available techniques can be found elsewhere [25]. Due to the intrisic limitations of these techniques, human data are more scarce. Using videomicroscopy of the nailfold area, Freedlander et al. (1922) reported that capillary stasis occurred [26]. However these observations are quite old, and the definition of shock state,

Microcirculation in Critcal Illness 113

although lethal, may be questioned in the absence of cardiovascular and respiratory support. However, these observations can be easily reproduced in the modern area (Fig. 1 and 2). However, the nailfold area is probably not representative as it may be subjected to vasoconstriction during changes in external temperature or during chills, so that this area is of limited interest in critically ill patients. More recently, various investigators [22, 27] used laser Doppler to investigate skin and muscle microvascular blood flow and observed that basal blood flow may be decreased or increased compared to healthy volunteers. In addition, the increase in microvascular blood flow was blunted after partial occlusion [28]. Nevertheless, the laser Doppler technique does not take into account the heterogeneity of microvascular blood flow, the measured parameter representing the average of the velocities in all the vessels included in the investigated volume.

Fig.I. Representative example of nailfold microcirculation in a human healthy volunteer

114 D. De Backer, M.-J. Dubois, 1. Creteur

Fig. 2. Representative example of nailfold microcirculation in a patient with septic shock. Note the decreased capillary density, and stagnant flow in various capillaries

Orthogonal Polarization Spectral (OPS) imaging is a newly developed non-invasive technique that allows the direct visualization ofthe microcirculation [29]. The device composed of a small camera and a few lenses is small and can easily be used at the bedside (Fig. 3). Briefly, polarized light is used to illuminate the area of interest. The light is scattered by the tissue and collected by the objective lens. A polarization filter (analyser), oriented orthogonal to the initial plane of the illumination light, is placed in front of the imaging camera and eliminates the reflected light scattered at or near the surface of the tissue that retains its original polarization (glare). Depolarized light scattered deeper within the tissues passes through the analyser. High contrast images of the microcirculation are formed by absorbing structures (e.g., blood vessels) close to the surface that are illuminated by the depolarized light coming from deeper structures. Due to its specific characteristics, this device is particularly convenient for studying the tissues protected by a thin epithelial layer, such as mucosal surface. In critically ill patients, the sublingual area is the most easily investigated mucosal surface. Other mucosal surfaces include rectal and vaginal surfaces which are of limited accessibility, and ileal or colic mucosa in patients with enterostomies. Images can also be generated in eyelids and in the nailfold [30] (Fig. 1 and 2).

Microcirculation in Critcal Illness llS

Fig. 3. Cytoscan ARII device for Orthogonal Polarization Spectral imaging technique

Fig. 4. Representative example of sublingual microcirculation in a human healthy volunteer. Note the dense venular (large vessels) and capillary (small vessels) network


Fig. S. Representative example of nailfold circulation in a human healthy volunteer. Note the decreased capillary (small vessels) network while the venular (large vessels) network is preserved

The use of OPS imaging techniques to visualize the microcirculation has been validated. Using OPS imaging and standard intravital fluorescence videomicroscopy in a hamster dorsal skinfold chamber, Groner et al. [29] observed that vessel diameters and functional capillary density were similar with both techniques. Similar results were reported by Langer et al. [31] applying both techniques on the surface of solid organs in rats. In addition, Laemmel et al. [32] observed, in mouse skin flaps and cremaster muscle preparations, that the velocity in straight vessels was similar with both techniques. Similar results were reported by Harris et al. [33] in a model of pressure-induced ischaemia in the dorsal skinfold chamber in hamsters. Recently, Mathura et al. [30] applied OPS imaging and capillaroscopy on the nailfold area in human healthy volunteers and observed an excellent agreement with both techniques in the measurement of capillary density and red blood cell velocity. However, this quantitative approach cannot be used for observations of the sublingual microcirculation in humans, due to the movement of artifacts generated by small tongue or respiratory movements. Hence, we [34] developed a semi-quantitative method to determine capillary density and the proportion of perfused capillaries.


Microvascular blood flow is altered in critically ill patients

Using the OPS technique in the sublingual area of patients in shock states, we [34] recently observed that microcirculatory alterations are frequent in shock states. Compared to healthy volunteers, patients septic shock presented a decrease in capillary density and a decrease in the proportion of the perfused capillaries. The decrease in capillary perfusion was due to an increase in the number of capillaries with stagnant flow and in the number of capillaries with intermittent flow. The severity of these alterations were more pronounced in non survivors. Microcirculatory alterations can also be observed in other conditions than sepsis. We [34] observed that the proportion of perfused capillaries was decreased in patients with cardiogenic shock. In 16 patients submitted to cardiac surgery, we observed that the proportion of perfused capillaries decreased after cardiopulmonary bypass, and remained altered during the first hours of admission in the intensive care unit, and normalized the day after surgery (unpublished data). However, these alterations were less pronounced than in patients with septic or cardiogenic shock.

Interestingly, the alterations observed in patients with septic shock were fully reversible after topical application of acetylcholine [35], suggesting that the microcirculation can be manipulated. Current studies are ongoing to determine the effects of various interventions on the microcirculation in humans.

Alterations observed in the sublingual area may not be representative of other areas. The splanchnic circulation may be altered earlier, and recover, later than other parts of the body. Interestingly, the sublingual mucosa which share a similar embryologic origin with the digestive mucosa and may also be of interest. Weil and coworkers [36, 37] observed that sublingual PC02 was increased in various shock states, reflected the severity of shock states and was related with outcome. They also reported that sublingual capnometry and gastric tonometry revealed parallel alterations, suggesting that both areas can be similarly and simultaneously affected [38]. In addition, sublingual PC02 was inversely related with changes in the tongue, splanchnic and renal blood flows [39]. Hence, the sublingual region may be similarly affected as other areas, including the splanchnic area.

Conclusions

Microvascular blood flow alterations are frequent in critically ill patients, and these alterations can have important physiopathological implications. Using OPS imaging techniques allows the direct visualization of the microcirculation


in critically ill patients, opening a new area for the investigation of the processes involved in the haemodynamic alterations of shock states.

References 1. Zhao KS. Junker D, Delano FA, et al (1985) Microvascular adjustments during irreversible

hemorrhagic shock in rat skeletal muscle. Microvasc Res 30: 143-153 2. Cryer HM, Garrison RN, Kaebnick HW et al (1987) Skeletal microcirculatory responses to

hyperdynamic Escherichia coli sepsis in unanesthetized rats. Arch Surg 122:86-92 3. Baker CH, Wilmoth FR (1984) Microvascular responses to E. coli endotoxin with altered

adrenergic activity. Circ Shock 12:165-176 4. Lam CJ, Tyml K, Martin CM et al (1994) Microvascular perfusion is impaired in a rat model of

normotensive sepsis. J Clin Invest 94:2077-2083 5. Piper RD, Pitt-Hyde ML, Anderson LA et al (1998) Leukocyte activation and flow behavior in

rat skeletal muscle in sepsis. Am J Respir Crit Care Med 157: 129-134 6. Piper RD, Pitt-Hyde M, Li F et al (1996) Microcirculatory changes in rat skeletal muscle in

sepsis. Am J Respir Crit Care Med 154:931-937 7. Farquhar 1, Martin CM, Lam C et al (1996) Decreased capillary density in vivo in bowel mucosa

of rats with normotensive sepsis. J Surg Res 61: 190-196 8. McCuskey RS, Urbaschek R, Urbaschek B (1996) The microcirculation during endotoxemia.

Cardiovasc Res 32:752-763 9. Drazenovic R, Samsel RW, Wylam ME et al (1992) Regulation of perfused capillary density in

canine intestinal mucosa during endotoxemia. J Appl Physiol 72:259-265 10. Walley KR (1996) Heterogeneity of oxygen delivery impairs oxygen extraction by peripheral

tissues: theory. J Appl Physiol 81:885-894 11. Humer MF, Phang PT, Friesen BP et al (1996) Heterogeneity of gut capillary transit times and

impaired gut oxygen extraction in endotoxemic pigs. J Appl Physiol 81:895-904 12. Ince C, Sinaasappel M (1999) Microcirculatory oxygenation and shunting in sepsis and shock.

Crit Care Med 27: 1369-1377 13. Zuurbier CJ, van Iterson M, Ince C (1999) Functional heterogeneity of oxygen supply-consump

tion ratio in the heart. Cardiovasc Res 44:488-497 14. Vicaut E, Hou X, Payen D et al (1991) Acute effects of tumor necrosis factor on the microcir

culation in rat cremaster muscle. J Clin Invest 87: 1537-1540 15. Groeneveld AB, Hartemink KJ, de Groot MC et al (1999) Circulating endothelin and nitrate-ni

trite relate to hemodynamic and metabolic variables in human septic shock. Shock 11: 160-166 16. Hollenberg SM, Broussard M, Osman J et al (2000) Increased microvascular reactivity and

improved mortality in septic mice lacking inducible nitric oxide synthase. Circ Res 86:774-778 17. Diaz NL, Finol HJ, Ton'es SH et al (1998) Histochemical and ultrastructural study of skeletal

muscle in patients with sepsis and multiple organ failure syndrome (MOFS). Histol Histopathol 13:121-128

18. Schneider J (1993) Fibrin-specific lysis of microthrombosis in endotoxemic rats by saruplase. Thromb Res 72:71-82

19. Bernard GR, Vincent J-L, Laterre PF et al (2001) Eft1cacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 344:699-709

20. Drost EM, Kassabian G, Meiselman HJ et al (1999) Increased rigidity and priming of polymorphonuclear leukocytes in sepsis. Am J Respir Crit Care Med 159: 1696-1702

21. Astiz ME, DeGent GE, Lin RY et al (1995) Microvascular function and rheologic changes in hyperdynamic sepsis. Crit Care Med 23:265-271

22. Kirschenbaum LA, Astiz ME, Rackow EC et al (2000) Microvascular response in patients with cardiogenic shock. Crit Care Med 28:1290-1294


23. Eichelbronner 0, Sielenkamper A, Cepinskas G et al (2000) Endotoxin promotes adhesion of human erythrocytes to human vascular endothelial cells under conditions of flow. Crit Care Med 28:1865-1870

24. D'orio V, Mendes P, Carlier P et a1 (1991) Lung fluid dynamics and supply dependency of oxygen uptake during experimental endotoxic shock and volume resuscitation. Crit Care Med 19:955-962

25. De Backer D, Dubois MJ (2001) Assessment of the microcirculatory flow in patients in the intensive care unit. Curr Opin Crit Care 7:200-203

26. Freedlander SO, Lenhart CH (1922) Clinical observations on the capillary circulation. Arch Intern Med 29: 12-32

27. Young JD, Cameron EM (1995) Dynamics of skin blood flow in human sepsis. Intensive Care Med 21:669-674

28. Neviere R, Mathieu D, Chagnon JL et al (1996) Skeletal muscle microvascular blood flow and oxygen transport in patients with severe sepsis. Am] Respir Crit Care Med 153: 191-195

29. Groner W, Winkelman JW, Hanis AG et al (1999) 011hogonal polarization spectral imaging: a new method for study of the microcirculation. Nat Med 5:1209-1212

30. Mathura KR, Vollebregt KC, Boer K et al (2001) Comparison of OPS imaging and conventional capillary microscopy to study the human microcirculation. J Appl Physiol 91 :74-78

31. Langer S, von Dobschuetz E, Harris AG et al (2000) Validation of the orthogonal polarization spectral imaging technique on solid organs. In Messmer K (ed): Orthogonal polarization spectral imaging. Basel, Karger, pp 32-46

32. Laemmel E, Tadayoni R, Sinitsina I et al (2000) Using orthogonal polarization spectral imaging for the experimental study of microcirculation: comparison with intravital microscopy. In Messmer K (ed): Orthogonal Polarization Spectral Imaging. Basel, Karger, pp 50-60

33. Harris AG, Sinitsina I, Messmer K (2000) The Cytoscan(TM) Model E-II, a new reflectance microscope for intravital microscopy: Comparison with the standard fluorescence method. J Vasc Res 37:469-476

34. De Backer D, Creteur J, Vincent J-L (2000) Microcirculatory alterations in cardiogenic and septic shock. Intensive Care Med 26:S334(Abstract)

35. De Backer D, Preiser J-C, Creteur Jet al (2001) Alterations in microvascular blood flow in septic patients can be reversed by acetylcholine. Am J Respir Crit Care Med 163:AI37(Abstract)

36. Nakagawa Y, Weil MH, Tang W et al (1998) Sublingual capnometry for diagnosis and quantitation of circulatory shock. Am J Respir Crit Care Med 157: 1838-1843

37. Weil MH, Nakagawa Y, Tang W et al (1999) Sublingual capnometry: a new noninvasive measurement for diagnosis and quantitation of severity of circulatory shock. Crit Care Med 27: 1225-1229

38. Povoas HP, Weil MH, Tang W et al (2000) Comparisons between sublingual and gastric tonometry during hemonhagic shock. Chest 118: 1127-1132

39. Jin X, Weil MH, Sun S et al (1998) Decreases in organ blood flows associated with increases in sublingual PC02 during hemorrhagic shock. ] Appl PhysioI85:2360-2364

Microbial Translocation: from Myth to Mechanism

J. C. MARSHALL

The microbial flora of the normal human gastrointestinal (GI) tract is enormously complex. The number of organisms that normally reside along the mucosal surface stretching from the oropharynx to the rectum exceeds the total number of cells in the human body by a factor of ten [1]. Careful studies of the nature of this flora have shown that it consists of upwards of 600 separate microbial species, each tending to associate with a specific anatomical niche in the GI tract, yet the total population is remarkably stable over time [2]. Under normal circumstances, this flora exists in a symbiotic relationship with the adjacent epithelial cells of the GI mucosa, and remains located on the luminal surface where it prevents colonization by potentially pathogenic organisms [3] and even modulates the immunological activity of epithelial cells of the host [4]. However a variety of pathological conditions can disrupt the relationship between the indigenous flora of the GI tract and the gut epithelium, and organisms can gain entrance to deeper tissues; this process has been termed "bacterial (or more appropriately "microbial") translocation".

The discovery that viable bacteria can pass from the gut lumen into the body dates back to the late nineteenth century, and gave rise to the concept of "intestinal autointoxication" [5]. This theory held that afflictions ranging from puerperal fever to rheumatoid arthritis to premature ageing were caused by microbes and other toxins that had entered the body from the gut; so pervasive was its appeal that it caught the attention of the playwright, George Bernard Shaw, who lampooned the theory in his play The Doctors' Dilemma. Jacob Fine demonstrated the translocation of radiolabelled Escherichia coli from the gut lumen into the peritoneal cavity in the late 1940s, and proposed that an intestinal factor, later identified as endotoxin, was responsible for the morbidity of irreversible shock [6]. Contemporary interest in microbial translocation was stimulated by the work of Rodney Berg, who defined the critical role of the normal flora in limiting translocation [7], and by his protege, Edwin Deitch, whose experimental studies showed that translocation is common in a variety of models of acute illness [8].

122 J.e. Marshall

The experimental literature on translocation is large; an expanding body of studies in humans indicates that similar insults evoke translocation in the clinical setting. It is probable that low-level translocation is common, and its occurrence during relatively minor physiological stresses suggests that it normally plays an adaptive and beneficial role. In critical illness, where multiple stimuli capable of inducing translocation are often present simultaneously, its contribution to attendant physiological derangements may be substantial [9].

This review will summarize the experimental and human evidence supporting the hypothesis that microbial translocation is common during acute illness, and consider the mechanisms through which it might contribute to the morbidity of critical illness.

Does translocation occur: experimental studies

There is ample evidence in the experimental literature that translocation of both viable bacteria, and dead bacteria or their products, is a common event in a number of experimental disease states [10]. In general, translocation occurs when the composition of the indigenous flora is altered, permitting overgrowth of one or more microbial species, when intestinal permeability is increased, or when host immune defences are impaired; commonly more than one of these processes is present simultaneously [11].

Translocation occurs by several distinct mechanisms. Bacteria can enter from the gut when there is a physical breach of the anatomical integrity of the GI tract, as occurs, for example, with perforated diverticulitis, however the term is generally reserved for the passage of bacteria through a grossly intact GI mucosa. Paracellular translocation occurs if the epithelium is disrupted, or if the tight junctions between epithelial cells are compromised, and likely represents the dominant mechanism associated with altered splanchnic blood flow or epithelial atrophy associated with prolonged fasting. However micro-organisms can also cross the gut epithelial barrier by a transcellular route, surviving intact within the cellular cytoplasm. Viable bacteria have been identified within a variety of cells of the gut mucosa, including macrophages [12], M cells [13], and enterocytes [14]. Their entry into host cells is a complex process. Shigella, for example, induces rearrangement of the epithelial cell cytoskeleton, promoting its uptake by macropinocytosis; once within the cell, it is able to modulate actin arrangement, and to induce cellular protrusions that facilitate its entry into adjacent cells [15]. A surface protein of Listeria, intemalin, binds to E-cadherin on epithelial cells, facilitating microbial invasion [16]. The receptor for IgE is

Microbial Translocation: From Myth to Mechanism 123

yet another example of a receptor on host cells that promotes bacterial uptake and translocation to regional nodes [17].

In health, the composition of the indigenous flora of the GI tract is remarkably constant over time. Its stability depends on a variety of factors including competition for nutrients and mucosal binding sites, and antimicrobial defences of both host and bacterial origin [18]. Microbial interactions can inhibit the translocation of micro-organisms in in vitro models [19]. In vivo, anaerobes reside in close association with the mucosa of the large intestine. By occupying epithelial cell binding sites, they prevent the binding of other potential pathogens, and so prevent local proliferation and tissue invasion. This inhibitory influence of the anaerobic flora on pathological gut colonization has been termed "colonization resistance" [20]. Ablation of the anaerobic flora of the rodent GI tract by oral antibiotics with activity against anaerobes results in a five log increase in the number of Gram-negative aerobes in cecal contents. Whereas the mesenteric lymph nodes of the normal mouse are sterile, these same Gram-negative organisms can be cultured from the nodes of all antibiotic-treated animals [7].

Other regulatory influences on the composition of the gut flora include bile, intestinal production of IgA, and antimicrobial substances produced by other components of the gut flora [18]. Disruption of bile flow by bile duct ligation, for example, results in cecal overgrowth with Gram-negative organisms, and bacterial translocation to mesenteric lymph nodes [21,22].

Yet another factor limiting gut microbial overgrowth is the influence of normal GI motility. A reduction in gut motility, as a result of ileus or mechanical obstruction, is associated with microbial overgrowth and translocation. The presence of enteral nutrients exerts mUltiple influences on gut structure and function, promoting epithelial cell growth and mucus production, and stimulating peristalsis. Parenterally fed rats have high rates of bacteria translocation to mesenteric lymph nodes; rates are lower in animals fed parenteral formulae by mouth, but lowest in those receiving a normal diet [23].

Finally, a variety of pathological insults promote bacterial translocation in experimental animals, including endotoxemia, thermal injury, hemorrhagic shock, peritonitis, pneumonia, pancreatitis, and intestinal transplantation. Splanchnic hypoperfusion and GI ileus with microbial overgrowth may represent common mechanisms through which these divergent insults promote microbial translocation. Local intestinal inflammation also appears to be implicated, as the extent of translocation following experimental bum injury can be reduced by prior neutrophil depletion [24] or by the activation of a heat shock response [25].

Microbial translocation in animal models is largely limited to aerobic

124 J.C. Marshall

organisms; entry of anaerobic organisms into the body requires an anatomical breach of the gut. While translocating micro-organisms can be detected in a number of different tissues, they are most numerous in mesenteric lymph nodes, and less commonly found in peritoneal fluid or portal venous blood, suggesting that lymphatic absorption may be the predominant route of their dissemination. Studies using radiolabelled bacteria have demonstrated that both viable and non-viable organisms are capable of translocation [26]. Moreover microbial products such as endotoxin or microbial DNA [27] are much more readily detectable than viable bacteria, suggesting that the culture of viable organisms significantly under-represents the burden of microbial translocation.

While microbial translocation has been best studied in the small bowel and colon, cervical lymph nodes from patients with oropharyngeal cancer have also been shown to harbor bacteria [28], suggesting that the process can occur across any colonized epithelial surface.

Microbial translocation in human disease

Microbial translocation can be detected directly in the animal model, by simultaneous culture of gut contents and homogenates of mesenteric lymph nodes or portal venous blood. These sites are generally not readily accessible in humans, and inferences about the extent to which translocation occurs in humans must often be drawn indirectly. Nonetheless, there is ample evidence to suggest that translocation occurs in humans under many of the same conditions that induce it in the experimental animal (Table 1), and that it accounts for a significant percentage of nosocomial infections, particularly in the critically ill patient.

Perhaps the most compelling evidence for microbial translocation resulting from alterations in the composition of the gut flora in humans derives from a bold experiment undertaken by a German surgeon 3 decades ago. This intrepid investigator consumed a suspension of viable Candida albicans, containing approximately 1012 organisms; his published case report documents that he became ill several hours later, and that candidemia and candiduria were detected [29]. Fortunately he survived.

Other evidence is less direct. Mesenteric lymph nodes cultured from patients undergoing elective laparotomy, or abdominal surgery for trauma, intestinal obstruction, or inflammatory bowel disease yield viable organisms in about 10% of patients; the frequency of culture positivity increases with the severity of the disease process, and correlates with susceptibility to post-operative infection. For example, O'Boyle et al. [30] cultured bacteria from mesenteric

Microbial Translocation: From Myth to Mechanism

Table 1. Microbial translocation in human illness (TPN = total parenteral nutrition)

Insult

Altered gut flora

Altered intestinal physiology

Intestinal ischemia

Intestinal inflammation

Miscellaneous

Clinical setting

Candida ingestion Cirrhosis Sort bowel syndrome Critical illness

Intestinal obstruction Obstructive jaundice Aortic vascular disease Cardiopulmonary bypass Cardiac arrest Inflammatory bowel disease Pancreatitis Trauma Small bowel transplant Elective laparotomy HomeTPN

125

lymph nodes of 15.4% of a cohort of 448 patients undergoing laparotomy; post-operative infections developed in 41 % of these patients, in comparison with only 14% of those whose cultures were sterile (P = .001). Similarly, a study of 51 patients undergoing repair of an abdominal aortic aneurysm documented evidence of bacterial translocation in 5 patients, 4 of whom developed postoperati ve infectious complications, including an aorto-enteric fistula [31].

We provided indirect evidence of a role for bacterial translocation in the pathogenesis of intensive care unit (ICU)-acquired infections in a study that found significant proximal gut overgrowth with the predominant species isolated from cases of nosocomial ICU-acquired infections being Candida, coagulase-negative Staphylococci, Pseudomonas, and the enterococcus (Table 2). Patients colonized with these species had a significantly higher probability of

Table 2. Proximal gut colonisation and ICU-acquired infection (CFU = colony-forming units)

Organism Number of patients MeanloglO Nosocomial infection colonized CFU/ml colonized not colonized P.

Candida 19 4.3 ± 1.6 15/19 8/22 <0.01 Enterococcus 12 6.8±0.8 6/12 11/29 NS Staphylococcus 10 5.7 ± 1.6 8/10 8/31 <0.01 epidermidis Pseudomonas 10 6.9 ± 1.1 9/10 6/31 <0.001

From [32]

126 J.e. Marshall

developing leU-acquired infection with the same organism; patients in whom the proximal gut was colonised with Pseudomonas experienced a significantly higher leU mortality rate. The infections involved included not only pneumonia (39% of all patients), but recurrent peritoniti s (27% of all patients), urinary tract infection (27% of all patients), and even bacteremia (27% of all patients) [32]. Similar data were reported by MacFie et al. in a cohort of patients undergoing laparotomy. Proximal gu t overgrowth was evident in 69% of 279 patients and predi sposed to remote infection. The same microbial species was found in the nasogastric aspirate and septic focus of 30% of patients, in the nasogastric aspirate and a mesenteric lymph node in 3 1 % of patients, and in a mesenteric lymph node and a focus of infection in 45% of patients [33].

Microbial products such as endotox in can also enter the body from the GI tract. Endotoxemia is a universal find ing during elective su rgical procedures associated with reduced splanchnic blood flow, particu larly cardiopulmonary bypass r34] or repair of an abdominal aortic aneurysm [35). Major abdominal surgery resu lts in increased concentrations of circul ating endotoxin, and a decreased plasma endotox in-neutralizi ng capacity that, in turn, correlate with an activated acu te-phase response [36]. Endotoxemia has also been observed in patients with major burns [37] and uncontroll ed congestive heart failure [38] .

Is bacterial translocation clinically relevant?

The complexity of the normal interactions between gut bacteria and cell s of the intestinal mucosa, and the frequency with which bacterial translocation or endotoxin absorption can be documented fo llowing relatively trivial insults, suggests that low- level entry of bacteria across the gut mucosa is a common, perhaps event physiologicall y beneficial, process. Translocation is associated with increased numbers of plasma cells, and with increased concentrations of IgA and IgM {39), and with increased expression of interleukin (IL)-6, fL -1 0, and soluble CD14 [40] , suggesti ng that translocation promotes the local development of antibacterial immun ity. However, translocati on of larger numbers of bacteria, particularly in the face of compromised host defenses, may well con tribute to illness.

Inferential evidence such as that described above suggests that bacterial translocation plays a significant role in the pathogenesis of ICU-acquired infection and its clinical sequelae. Perhaps the strongest direct evidence that translocation is responsible for morbidity in critically ill patients comes from clinical stud ies of select ive digestive tract decontamination (SOD) in critically ill patients. SOD is a technique for in fection prophylaxis that consists of the topical application of antibiotics active against aerobic Gram-negative orga-


nisms (polymyxin and tobramycin) and fungi (amphotericin B), with a short course of systemic cefotaxime administered until the effect of the topical agents on the gut flora has become stabilized [41]. The anaerobic flora is spared to preserve colonization resistance. More than three dozen randomized trials of SDD have been performed, and meta-analyses of these indicate that the intervention not only reduces rates of nosocomial infection, but also results in a statistically significant increase in survival [42]. The beneficial effects are particularly pronounced in surgical patients, in whom the use of the full SDD regimen is associated with a 50% reduction in rates of bacteremia, and a 30% reduction in mortality [43]. Further data in support of an important role of bacterial translocation corne from studies of enteral nutrition in trauma, showing a significant reduction in secondary infections in patients receiving nutritional support via the enteral route [44], and from an evolving body of studies demonstrating that enteral feeding in severe pancreatitis reduces the frequency of infectious complications, and attenuates the inflammatory response [45]. Finally, data from a phase II study of an endogenous endotoxin-neutralizing compound, recombinant bactericidal/permeability increasing protein (rBPI), administered to patients following multiple trauma, showed a reduction in rates of pneumonia and acute respiratory distress syndrome in the treated group, and a trend towards an improved clinical outcome [46].

Conclusions

The evidence that microbial translocation is a common event following a variety of acute insults is compelling. Yet controversy persists regarding its clinical importance [47]. Much of this controversy arises because translocation cannot be documented in patients having, or at risk for, organ failure [48], or because patients in whom translocation can be demonstrated often experience a relatively benign clinical course [40]. It would be naive to contend that a single insult can be held responsible for the entire clinical spectrum of a complex disorder such as the multiple organ dysfunction syndrome. Just as not all patients with organ dysfunction have pneumonia, and not all patients with pneumonia develop organ dysfunction, so bacterial translocation is simply one of a number of processes through which micro-organisms can induce disease in the human host. A compelling body of evidence showing benefit for gut-directed prophylactic measures, such as selective decontamination of the digestive tract or early enteral feeding, suggests that its importance cannot be ignored.

128 J.C. Marshall

References 1. Savage DC (1977) Microbial ecology of the gastrointestinal tract. Annu Rev Med 31:107-133 2. Lee A (1985) Neglected niches. The microbial ecology of the gastrointestinal tract. Adv

Microbial EcoI8:115-162 3. Van Der Waaij D (1989) The ecology of the human intestine and its consequences for overgrowth

by pathogens such as Clostridium difficile. Annu Rev MicrobioI43:69-87 4. Neish AS, Gewirtz AT, Zeng H, et al (2000) Prokaryotic regulation of eptihelial responses by

inhibition ofIkBa ubiquitination. Science; 289:1560-1563 5. Adami JG (1914) Chronic intestinal stasis: autointoxication and subinfection. B M J 1: 177-183 6. Fine J, Frank ED, Ravin HA, et al (1959) The bacterial factor in traumatic shock. N Engl J Med

260:214-220 7. Berg RD (1981) Promotion of the translocation of enteric bacteria from the gastrointestinal

tracts of mice by oral treatment with penicillin, clindamycin, or metronidazole. Infect Immun 33:854-861

8. Deitch EA (1990) The role of intestinal barrier failure and bacterial translocation in the development of systemic infection and multiple organ failure. Arch Surg 125:403-404

9. Carrico CJ, Meakins JL, Marshall JC et al (1986) MUltiple organ failure syndrome. The gastrointestinal tract: The 'motor' ofMOF. Arch Surg 121:196-208

10. Wells CL, Maddaus MA, Simmons RL (1988) Proposed mechanisms for the translocation of intestinal bacteria. Rev Infect Dis 10:958-979

11. Berg RD (1999) Bacterial translocation from the gastrointestinal tract. Adv Exp Med Bioi 473:11-30

12. Wells CL, Maddaus MA, Erlandsen SL, Simmons RL (1988) Evidence for the phagocytic transport of intestinal particles in dogs and rats. Infect Immun 56:278-282

13. Kucharzik T, Lugering N, Rautenberg K, et al (2000) Role ofM cells in intestinal barrier function. Ann NY AcadSci915:171-183

14. Wells CL, Westerlo EM van de, Jechorek RP, Erlandsen SL (1996) Intracellular survival of enteric bacteria in cultured human enterocytes. Shock 6:27-34

15. Sansonetti PJ (2001) Microbes and microbial toxins: paradigms for microbial-mucosal interactions III. Shigelosis: from symptoms to molecular pathogenesis. Am J Physiol Gastrointest Liver Physiol 280:G319-G323

16. Lecuit M, Vandormael-Pournin S, Lefort J, et al (2001) A transgenic model for listeriosis: role of internalin in crossing the intestinal barrier. Science 292: 1722-1725

17. Dombrowicz D, Nutten S, Desreumaux P, et al (2001) Role of the high affinity immunoglobulin E recetor in bacterial translocation and intestinal inflammation. J Exp Med 193:25-34

18. Marshall JC (1999) The gastrointestinal flora and its alterations in critical illness. CUff Opin Crit Care 5:119-125

19. Mattar AF, Drongowski RA, Coran AG, Harmon CM (2001) Effect ofprobiotics on enterocyte bacterial translocation in vitro. Pediatr Surg Internat 17:265-268

20. Van Der Waaij D, Berghuis De Vries JM, Lekkerkerk Van Der Wees JEC (1971) Colonization resistance of the digestive tract in conventional and antibiotic treated mice. J Hyg Camb 69:405-411

21. Deitch EA, Sittig K, Li M, et al (1990) Obstructive jaundice promotes bacterial translocation from the gut. Am J Surg 159:79-84

22. Kuzu MA, Kale IT, Col C, et al (1999) Obstructive jaundice promotes bacterial translocation in humans. Hepato-gastroenterology 46:2159-2164

23. Alverdy JC, Aoys E, Moss GS (1988) Total parenteral nutrition promotes bacterial translocation from the gut. Surgery 104:185-190

24. Fazal N, Shamim M, Khan SS, et al (2000) Neutrophil depletion in rats reduces burn injury-induced intestinal bacterial translocation. Crit Care Med 28: 1550-1555

25. Eaves-Pyles T, Wong HR, Alexander JW (2000) Sodium arsenite induces the stress response in


the gut and decreases bacterial translocation in a burned mouse model with gut-derived sepsis. Shock 13:314-319

26. Alexander JW, Boyce ST, Babcock GF, et al (1990) The process of microbial translocation. Ann Surg 212:496-512

27. Kane TD, Alexander JW, Johannigman JA (1998) The detection of microbial DNA in the blood: a sensitive method for diagnosing bacteremia and/or bacterial translocation in surgical patients. Ann. Surg 227:1-11

28. Sakamoto H, Naito H, Ohta Y, et al (1999) Isolation of bacteria from cervical lymph nodes in patients with oral cancer. Arch Oral Bioi 44:789-793

29. Krause W, Matheis H, Wulf K (1969) Fungaemia and funguria after oral administration of Candida albicans. Lancet 1:598-599

30. O'Boyle CJ, MacFie J, Mitchell CJ, et al (1998) Microbiology of bacterial translocation in humans. Gut 42:29-35

31. Woodcock NP, Sudheer V, EI-Barghouti N, et al (2000) Bacterial translocation in patients undergoing abdominal aortic aneurysm repair. Br J Surg 87:439-442

32. MarshalIJC, Christou NY, Meakins JL (1993) The gastrointestinal tract. The "undrained abscess" of multiple organ failure. Ann Surg 218:111-119

33. MacFie J, O'Boyle C, Mitchell CJ, et al (1999) Gut origin of sepsis: a prospective study investigating associations between bacterial translocation, gastric microflora, and septic morbidity. Gut 45:223-228

34. Andersen LW, Landow L, Baek L, et al (1993) Association between gastric intramucosal pH and splanchnic endotoxin, antibody to endotoxin, and tumor necrosis factor-a concentrations in patients undergoing cardiopulmonary bypass. Crit Care Med 21 :210-217

35. Soong CV, Blair PHB, Halliday ML, et al (1993) Endotoxemia, the generation of cytokines, and their relationship to intramucosal acidosis of the sigmoid colon in elective abdominal aortic aneurysm repair. Eur J Vasc Surg 7:534-539

36. Buttenschoen K, Buttenschoen DC, Berger D, et al (2001) Endotoxemia and acute-phase proteins in major abdominal surgery. Am J Surg 181:36-43

37. Winchurch RA, Thupari IN, Munster AM (1987) Endotoxemia in burn patients: Levels of circulating endotoxins are related to bum size. Surgery 102:808-812

38. Niebauer J, Yolk HD, Kemp M, et al (1999) Endotoxin and immune activation in chronic heart failure: a prospective cohort study. Lancet 353: 1838-1842

39. Woodcock NP, Robertson J, Morgan DR, et al (2001) Bacterial translocation and immunohistochemical measurement of gut immune function. J Clin Pathol 54:619-623

40. Schoeffel U, Pelz K, Haring RU, et al (2000) Inflammatory consequences of the translocation of bacteria and endotoxin to mesenteric lymph nodes. Am J Surg 180:65-72

41. Baxby D, van Saene HKF, Stoutenbeek CP, Zandstra DF (1996) Selective decontamination of the digestive tract: 13 years on, what it is and what it is not. Intensive Care Med 22:699-706

42. D'Amico R, Pifferi S, Leonett C, et al (1998) Effectiveness of antibiotic prophylaxis in critically ill adult patients: systematic review of randomized controlled trials. B M J 316: 1275-1285

43. Nathens AB, Marshall JC (1999) Selective decontamination of the digestive tract (SDD) in surgical patients. Arch Surg 134: 170-176

44. Kudsk KA, Croce MA, Fabian TC, et al (1992) Enteral versus parenteral feeding. Effects on septic morbidity after blunt and penetrating abdominal trauma. Ann Surg 215:503-513

45. McGregor CS, Marshall JC (2001) Enteral feeding in acute pancreatitis: just do it. Curr Opin Crit Care; 7:89-91

46. Demetriades D, Smith JS, Jacobsen LE, et al (1999) Bactericidal/permeability-increasing protein (rBPI21) in patients with hemorrhage due to trauma: results of a multicenter phase II clinical trial. J Trauma Injury Infect Crit Care; 46:667-677

47. Lemaire LC, Lanschot JJ van, Stoutenbeek CP, et al (1997) Bacterial translocation in multiple organ failure: cause or epiphenomenon still unproven. Br J Surg 84: 1340-1350

48. Moore FA, Moore EE, Poggetti R, et al (1991) Gut bacterial translocation via the portal vein: a clinical perspective with major torso trauma. J Trauma 31 :629-638

Empirical Antibiotic Treatment in leu Patients

P.GROSSI

The increasing resistance of more and more bacteria to the antibiotics originally designed to treat them has created a unique challenge for the medical profession and the pharmaceuticals industry.

Potent broad-spectrum antimicrobials in current use will retain their potency in serious infections only if used appropriately. This means identifying the right patient and indication and employing the right antibiotic at its optimal dosage regimen from the outset, rather than waiting until the failure of other regimens. Inadequate initial antimicrobial treatment of infections among critically ill patients has been shown to be a risk factor for hospital mortality and, in the survivors, for increased length of stay in intensice care unit (lCU) and in the duration of mechanical ventilation [1-3].

Of great concern is the increasing incidence of serious nosocomial infections caused by Gram-positive bacteria with acquired multidrug resistance. For more than 30 years, vancomycin has been a reliable treatment for gram-positive bacterial infections. The emergence of enterococci with resistance to vancomycin (VRE) [4], seen predominantly in the species Enterococcus jaecium, has been followed by an increase in the frequency with which this species is recovered [5-7]. Worldwide methicillin-resistant staphylococci, both Staphylococcus aureus and coagulase-negative strains, are numerically the greatest problem. Since 1996, vancomycin-intermediate S.aureus (VISA; ~ancomycin minimum inhibitory concentration = 8-16 Jlg/ml) has been identified in Europe, Asia and United States [8-10]. To date, no resistant S.aureus isolate (vancomycin MIC ~ 32 Jlg/ml) has been reported, however it seems likely that vancomycin-resistant S.aureus (VRSA) will emerge as a nosocomial pathogen with disastrous consequences if widespread nosocomial transmission occurs. In the United States Gram-positive bacteria are responsible for over 60% of nosocomial bacteremias [11] and in our country the prevalence of MRSA in ICU was reported to be 81 % [12].

Among Gram-negative aerobic bacilli, changes in resistance are also being observed worldwide. A study of susceptibility of these organisms isolated from

132 Paolo Grossi

blood culture in 24 European hospitals during 1997 and 1998 identified the potential presence of beta-Iactamase enzymes in 0.3% of Escherichia coli and 16.7% of Klebsiella pneumoniae strains studied [13]. Similar findings of recent increasing resistance have come from New Jersey and Oklahoma in the United States [14,15].

How we ensure that the power of important antimicrobials is preserved over the coming years depends on attention to diagnostic testing, drug use, and appropriate antibiotic choice. The most-important factors are summarized below.

Correct Diagnosis

A study in the United States recently documented that only 83% of surveyed laboratories were using correct procedures to screen for extended-spectrum beta-Iactamase (ESBL)-producing organisms and only 17% were using proper procedures to confirm ESBL presence [16]. In addition, in some hospital microbiology laboratories, antimicrobial susceptibility testing of enterococcal isolates from urine or non sterile body sites (e.g., wounds) is not performed routinely; thus identification of nosocomial VRE colonization or infection in hospitalized patients may be delayed. Therefore, in hospitals where VRE have not yet been detected, implementing special measures can promote early detection of VRE.

Measures aimed at minimizing nosocomial transmission of multidrug-resistant organisms

The prevention and control of drug-resistant infections requires measures to promote the prudent use of antimicrobial drugs and prevenrthe transmission of infections (whether drug resistant or not). There is a need for continuous education of healthcare professionals and the public in this regard. Hand washing is considered the single most-important measure for preventing the spread of infection in hospitals [17, 18]. Hands should be washed between patient contacts, after contact with potentially infectious material (e.g., blood, body fluids, patient-care items), and after removal of examination gloves. The implementation of rigorous infection control measures has been recommended as a first response at any occurrence of VRE [19]. Recent guidelines have stressed the need for strict barrier precautions (including isolation of transmission and the routine collection of surveillance cultures from patients in high-

Empirical Antibiotic Treatment in leu Patients 133

-risk wards as part of colonized or infected patients, use of gowns and gloves, and disinfection of surfaces and equipment) combined with the education of healthcare workers concerning the risk of a package aimed at preventing the spread of VRE. The same guidelines have also called for restriction on the use of oral vancomycin, as means of reducing any selective pressure that might be acting in favor of the resistant strains. Because the use of second- and third-generation cephalosporins might promote the development of VRE [20, 21], restricting or replacing the use of certain antibiotics has sometimes been successful in controlling outbreaks of nosocomial infections [22]. Healthcare facilities should adopt appropriate guidelines to identify and manage MRSA outbreaks [23]. The recovery of Staphylococcus au reus with reduced susceptibility to vancomycin (e.g., MIC ~ 4 Jlg/ml) should be reported promptly to local and health state departments, infection control precautions should be implemented [24], and an epidemiological investigation should be conducted.

Use of currently available antimicrobial agents

The use of antibiotics should be under strict control and they should only be prescribed for accepted indications. In the United States vancomycin is the only antibiotic approved for use against MRSA, whereas in Europe, teicoplanin, a compound of the same chemical class (the glycopeptide antibiotics), is available. For treatment of life-threatening infections, such as endocarditis or sepsis caused by MRSA, and serious infections caused by MRSE and other methicillin-resistant coagulase-negative staphylococci, especially those involving prosthetic valve devices, vancomycin and teicoplanin are still powerful weapons. Vancomycin is currently administered by intermittent infusion to produce specific peak and trough concentrations for efficacy and to avoid potential serum drug concentration-related side effects. However, vancomycin demonstrates concentration-independent killing, which is maximized ar concentrations of four or five times the MIC for the organism. Administration of vancomycin as a continuous infusion and maintaining a constant concentration in serum of four or five times the MIC for the infecting organisms may be the ideal way to deliver this antibiotic for serious infections [25].

Combinations of vancomycin and ~-lactams have shown synergistic effect in the treatment of intermediate susceptible S.aureus [26], although many isolates of vancomycin-resistant enterococci in the United States are not susceptible to the combination of vancomycin and ~-lactams [27]. Teicoplanin is active in vitro against most VanB and VanC types of enterococci; however VanA-type resistance, characterized by resistance to teicoplanin, is more com-

134 Paolo Grossi

mon [7]. Finally the use of high doses of ampicillin or ampicillin and sulbactam (18-30 g I day) may be effective in vivo against ampicillin-resistant enterococci that are inhibited by ;::: 64 Jlg/ml of ampicillin, although the toxicity of these doses is largely unknown.

Gram-negative bacilli are frequently associated with nosocomial infections in leu patients, particularly ventilator-associated pneumonia and catheter-associated urinary tract infections [28]. Of particular concern is the nosocomial infection caused by enterobacteria-producing extended-spectrum j3-lactamases (ESBLs), particularly Klebsiella pneumoniae. Organisms that possess these enzymes are usually resistant to multiple antimicrobials and hydrolyze thirdgeneration cephalosporins and aztreonam, rendering these potent antibacterial agents useless [29]. In some cases, ceftriaxone or cefotaxime may test susceptible or intermediate to ESBL-producing K. pneumoniae, but the clinical utility of these agents against such isolates is uncertain, and clinical failures have been reported [29, 30]. Evaluation of data from NNIS hospitals shows a dramatic increase in the proportion of K. pneumoniae resistant to ceftriaxone, cefotaxime, or ceftazidime over the past decade, with a much greater increase among isolates recovered from leu patients. The prevalence of ESBL-producing strains is easily underestimated because resistance to j3-lactam agents, although increased, may fail to reach currently specified resistance breakpoints [31]. Tracking resistance, however, provides us with a rough estimate of the growing magnitude of this troublesome pathogen. Duration of stay in the hospital, especially the leU, has been associated with acquisition of ESBL-producing K. pneumoniae [32-34] and has been implicated in inter-facility transmission within a geographic region [35]. There is strong evidence that antimicrobial exposure has an impact on the acquisition of ESBL-producing K. pneumoniae. One study [29] demonstrated that preferential use of a specific j3-lactamlj3-lactamase inhibitor combination (i.e., piperacillinltazobactam) rather than ceftazidime was associated with a decrease in rates of isolating these organisms in the leu. Another [36] demonstrated that patients exposeg. to any j3-lactamlj3-lactamase inhibitor combination (Le., amoxicillinlclavulanic acid, ampicillinlsulbactam, ticarcillinlclavulanic acid, or piperacillinltazobactam) appeared to be at decreased risk of colonization or infection with ESBL-producing K. pneumoniae in multivariate analysis. This suggests that preferential use of j3-lactamlj3-lactamase inhibitor combinations may be an important control measure, along with hand-washing and infection control precautions, to help control out-breaks of ESBL-producing K. pneumoniae.

Other common antimicrobial-resistant pathogens encountered among leu patients include Pseudomonas aeruginosa resistant to imipenem and P. aeruginosa or Enterobacter spp. resistant to third-generation cephalosporins, such

Empirical Antibiotic Treatment in leU Patients 135

as cefotaxime, ceftriaxone, or ceftazidime. Examination of data from NNIS hospitals shows that rates of resistance among these pathogens again appear higher among isolates from ICU patients than non-ICU patients. The rates of resistance have been relatively stable over the past decade, however. Ampicillin-resistant E. coli is of less concern to the ICU clinician because alternative therapy is readily available, and these patients are commonly on a broad-spectrum agent to which the organism is susceptible. Finally, the rate of fluoroquinolone resistance (Le., resistance to ofloxacin or ci-profloxacin) among P. aeruginosa reported to NNIS has increased rapidly over the past decade. In contrast to all the other pathogens discussed so far, however, quinolone-resistant P. aeruginosa is not more prevalent among ICU patients than non-ICU patients. There are probably many reasons for this. Contributing factors may include the large amounts of quinolones used by patients outside the ICU, or the development of fluoroquinolone resistance among P. aeruginosa unrelated to the ICU setting [37].

Candida albicans is the seventh most common pathogen associated with nosocomial infection in ICU patients. In general, resistance to antifungal agents among Candida spp. is rare. Susceptibility testing for C. albicans is difficult and not routinely performed in most hospitals, however, so data on the frequency of fluconazole-resistant C. albicans tend to be limited to research scenarios [36]. Therapeutic options to treat patients with C. albicans infection include polyenes (amphotericin), imidazoles, and triazoles. The emergence of antimicrobial-resistant fungal pathogens limits the few therapeutic options. Some acquired immunodeficiency syndrome patients, particularly those with greater exposure to azole therapy or low CD4 counts, have developed azole-resistant C. albicans infections [38, 39]. Resistance to azoles has not been well documented in human immunodeficiency virus-negative patients.

The appearance of azole-resistant C. albicans infection in AIDS patients portends resistance in other immunocompromised patient populations. Data suggest that increasing use of prophylactic antifungal therapy ill patients at highest risk for endogenous Candida spp. Infection may lead to the increasing frequency of infections with fungi such as C. krusei, which have intrinsic azole resistance, or the azole-resistant C. glabrata or C. albicans [38-41]. Consequently, issues relating to azole-resistant Candida spp. will usually be limited to the specialized care unit exclusively treating patients with a severely compromised immune system. Of concern are data from a recent multicenter study of 50 United States medical centers that documented that 10% of C. albicans isolates from the bloodstream of hospitalized patients were resistant to fluconazole [42]. The resistance rate ranged from 5% to 15%, depending on the region of the United States, suggesting that local

136 Paolo Grossi

factors, such as amount of azole usage, may playa role in the relative frequency of azole-resistant C. albicans infections.

Optimal dosing

A team approach to making sure that correct dose, correct duration of treatment, etc. are being used for ICU patients has been shown to pay clear dividends.

Following appropriate indications

Making sure that the right drug is being delivered at the right time is of worldwide concern [43]; campaigns to deal with this issue in the United States are matched in many other countries.

Adopting global concern but local approaches

Several countries are taking steps to deal with resistance on a national basis [44]. However, as stressed in a recent statement by the Alliance for Prudent Use of Antibiotics, "resistance is a local phenomenon with cumulative global presence and consequences" [45].

Carrying out these important steps will require different activities in different areas and in different countries, as resources and local problems vary. Dramatic differences in antimicrobial resistance exist within individual hospitals, and may depend on both antimicrobial use and infection control practices. Thus, a single approach to the global spread of resistance is not possible, and approaches to control will have to be individualized to be effective. Preserving effective antimicrobials will be a global battle for many yeal's to come, and will require careful attention to several of the areas emphasized here. No strategy for controlling resistance or optimizing antimicrobial use will be successful unless the entire healthcare delivery system views this problem as vital. A multidisciplinary, systems-oriented approach involving the hospital leadership is required to succeed in combating the growing problem of antimicrobial resistance in ICUs [46].


References 1. Kollef MH, Sherman G, Ward S, Fraser VJ (1999) Inadequate antimicrobial treatment of

infections: a risk factor for hospital mortality among critically ill patients. Chest 115:462-474 2. Dupont H, Mentec H, Sollet JP, Bleichner G (200 1) Impact of appropriateness of initial antibiotic

therapy on the outcome ofventilator associated pneumonia. Intensive Care Med 27:355-362 3. Ibrahim EH, Sherman G, Ward S, et al (2000) The influence of inadequate antimicrobial

treatment of bloodstream infections on patients outcome in the ICU setting. Chest 118:146-155 4. Handwerger S, Raucher B, Altarac D, et al (1993) Nosocomial outbreak due to Enterococcus

faecium highly resistant to vancomycin, penicillin and gentamicin. Clin Infect Dis 16:750-755 5. Nosocomial enterococci resistant to vancomycin-United States, 1989-1993 (1993) MMWR

42:597-599 6. Iwen PC, Kelly DM, Linder J, et al (1997) Change in prevalence and antibiotic resistance of

Enterococcus species isolated from blood cultures over an 8-year period. Antimicrob Agents Chemother 41 :494-495

7. Murray BE (2000) Vancomycin-resistant enterococcal infections. N Engl J Med 342:710-721 8. Hiramatsu K, Hanaki H, Ino T, et al (1997) Methicillin-resistant Staphylococcus aureus clinical

strain with reduced vancomycin susceptibility. J Antimicrob Chemother 40: 135-136 9. Hiramatsu K, Aritaka N, Hanaki H, et al (1997) Dissemination in Japanese hospitals of strains

of Staphylococcus aureus heterogeneously resistant to vancomycin. Lancet 350: 1670-1673 10. Rotun SS, McMath V, Schoonmaker DJ, et al (1999) Staphylococcus aureus with reduced suscepti

bility to vancomycin isolated from a patient with fatal bacteremia. Emerg Infect Dis 5:147-149 11. Edmond MB, Wallace SE, McClish DK, et al (1999) Nosocomial bloodstream infections in the

United States hospitals: a three-year analysis. Clin Infect Dis 29:239-244 12. Vincent J-L, Bihari DJ, Suter PM, et al (1995) The prevalence of nosocomial infection in

intensive care units in Europe. JAMA 274:639-644 13. Fluit AC, Jones ME, Schmidt F-J, et al SENTRY Participants Group (2000) Antimicrobial

susceptibility and frequency of occurrence of clinical blood isolates in Europe from the SENTRY antimicrobial surveillance program, 1997 and 1998. Clin Infect Dis 30:454-460

14. Flournoy 01, Reinert RL, Bell-Dixon C, Gentry CA (2000) Increasing antimicrobial resistance in gram-negative bacilli isolated from patients in intensive care units. Infect Control Hosp EpidemioI28:244-250

15. Wu LC, Brook JH (2000) Multiple antibiotic resistant bacteria in New Jersey hospitals, 1992-1998 (abstract). International Conference on Emerging Infectious Diseases. Atlanta, Georgia, USA, July 16-19,2000. Centers for Disease Control and Prevention, Georgia, p 75

16. Centers for Disease Control and Prevention (2000) Laboratory capacity to detect antimicrobial resistance, 1998. MMWR Morb Mortal Wkly Report 48: 1167 -1171

17. Gamer JS, Hospital Infection Control Practices Advisory Committee (1996) Guideline for isolation precautions in hospitals. I. Evolution of isolation practices. Am J Infect Control 24:24-31

18. Larson EL, APIC Guidelines Committee (1995) APIC guideline for handwashing and hand antisepsis in health care settings. Am J Infect Control 23:251-269

19. Hospital Infection Control Practice Advisory Committee (HICPAC) (1995) Recommendations for preventing the spread of vancomycin resistance. Infect Control Hosp Epidemiol 16: 1 05-113

20. Quale J, Landman D, Atwood E, et al (1996) Experience with a hospital-wide outbreak of vancomycin-resistant enterococci. Am J Infect Control 24:372-379

21. Moulin F, Dumontier S, Saulnier P (1996) Surveillance of intestinal colonization and of infection by vancomycin-resistant enterococci in hospitalized cancer patients. Clin Microbiol Infect 34:751-752

22. Bradley SJ, Wilson ALT, Allen MC, et al (1999) The control of hyperendemic glycopeptide-resistant Enterococcus spp. on a haematology unit by changing antibiotic usage. J Antimicrob Chemother 43 :261-266

23. Wenzel RP, Reagan DR, Bertino JS, et al (1998) Methicillin-resistant Staphylococcus aureus

138 Paolo Grossi

outbreak: a consensus panel's definition and management guidelines. Am J Infect Control 26:102-110

24. CDC interim guidelines for prevention and control of staphylococcal infections associated with reduced susceptibility to vancomycin. (2000) MMWR 48: 1167 -1171

25. James JK, Palmer SM, Levine DP, Rybak MJ (1999) Comparison of conventional dosing versus continuous infusion vancomycin therapy for patients with suspected or documented Gram positive infections. Antimicrob Agents Chemother 40:696-700

26. Climo M, Patron RL, Archer G (1999) Combinations of vancomycin and beta-lactams are synergistic in the treatment of vancomycin intermediate susceptible S.aureus Antimicrob Agents Chemother 43(7):1747-1753

27. Cercenado E, Eliopoulos GM, Wennerstern CB, Moellering RC Jr. (1992) Absence of synergistic activity between ampicillin and vancomycin against highly vancomycin-resistant enterococci. Antimicrob Agents Chemother 36:2201-2203

28. Fridkin SK, Weibel SF, Weinstein RA (1996) Magnitude and prevention of nosocomial infections in the intensive care unit. Infect Dis Clin North Am 11:479--496

29. Rice LB, Eckstein EC, DeVente J, et al (1996) Ceftazidime-resistant Klebsiella pneumoniae isolates recovered at the Cleveland Department of Veterans Affairs Medical Center. Clin Infect Dis 23:118-124

30. Karas JA, Pillay DG, Muckart D, et al (1996) Treatment failure due to extended spectrum ~-lactamase. J Antimicrob Chemother 37:203

31. Jacoby GA, Han P. (1996) Detection of extended-spectrum ~-Iactamases in clinical isolates of Klebsiella pneumoniae and Escherichia coli. J Clin Microbiol 34:908-911

32. Burwen DR, Banerjee SN, Gaynes RP, et al (1994) Ceftazidime resistance among selected nosocomial gram-negative bacilli in the United States. J Infect Dis 170: 1622-1625

33. Piroth L, Aube H, Doise J, et al (1998) Spread of extended-spectrum-Iactamase-producing Klebsiella pneumoniae: are ~-lactamase inhibitors of therapeutic value? Clin Infect Dis 27:76-80

34. Schiappa DA, Hayden MK, Matushek MG, et al (1996) Ceftazidime-resistant Klebsiella pneumoniae and Escherichia coli bloodstream infection: a case-control and molecular epidemiologic investigation. J Infect Dis 174:529-536

35. Monnet D, Biddle JW, Edwards JR, et al (1997) Evidence of interhospital transmission of extended-spectrum-Iactam-resistant Klebisella pneumoniae in the United States, 1986-1993. Infect Control Hosp EpidemioI18:492--498

36. Rex JH, Pfaller MA, Rinaldi MG, et al (1993) Antifungal susceptibility testing. Clin Microbiol Rev 6:367-381

37. McCaig LF, Hughes JM (1995) Trends in antimicrobial drug prescribing among office-based physicians in the United States. JAMA 273:214-219

38. Johnson EM, Warnock OW, Luker J, et al (1995) Emergence of azole drug resistance in Candida species from HIV-infected patients receiving prolonged fluconazole therapy for oral candidosis. J Antimicrob Chemother 35: 103-114 ~

39. Maenza JR, Keruly JC, Moore RD, et al (1996) Risk factors for fluconazole-resistant candidiasis in human immunodeficiencyvirus-infected patients. J Infect Dis 173 :219-225

40. Abi-Said D, Anaissie E, Uzon 0, et al (1997) The epidemiology of hematogenous candidiasis caused by different Candida species. Clin Infect Dis 24: 1122-1128

41. Fridkin SK, Jarvis WR (1996) Epidemiology of nosocomial fungal infections. Clin Microbiol Rev 9:499-511

42. Pfaller MA, Jones RN, Messer SA, et al (1998) National surveillance of nosocomial blood stream infection due to Candida albicans: frequency of occurrence and antifungal susceptibility in the SCOPE Program. Oiagn Microbiol Infect Dis 31 :327-332

43. World Health Organization. (2000) Overcoming Antimicrobial Resistance -World Health Report on Infectious Diseases 2000. Available at http://www.who.intlinfectious-disease-reportl2000Iindex.html. World Health Organization, Geneva

44. Interagency Task Force on Antimicrobial Resistance (2000) Draft public health action plan to combat


antimicrobial resistance. I Domestic issues. Available at website: http://www.cdc.gov/drugresistance!actionplanlindex.htm Atlanta: Centers for Disease Control and Prevention

45. Association for the Prudent Use of Antibiotics. (2000) Response to the Draft Public Health Action Plan on Antimicrobial Resistance. Available at APUA website: http://www.healthsci.tufts.edu/apua/apua.htrnl

46. Goldmann DA, Weinstein RA, Wenzel RP, et al (1996) Strategies to prevent and control the emergence and spread of antimicrobial-resistant microorganisms in hospitals. JAMA 275:234-240

Rotation of Antibiotics - A New Strategy for Prescription in the Intensive Care Unit

D. GRUSON, G. HILBERT

Bacterial resistance, encountered most often in intensive care units (ICUs) increases with time. This resistance is associated with a high death rate and a significant rise in cost. It is now recognized that resistance to antibiotics in the ICU is a problem, that influences the patients' outcome [1]. This negative outcome could be due to a delay in prescribing a well-adapted and effective antibiotic.

Several "pressures" are responsible for the rise in incidences of resistant germs found in the ICU. One example is the incorrect use of antibiotics in IC, which increases the pressure of the selection of mutant-resistant bacteria [2]. An unsuitable geography and a lack of space in ICU bedrooms, a shortage of care staff, and an increase in the workload can also explain the pressure of selection and the occurrence of resistance.

In a recent consensus conference, the CDC suggested that the problem could be better checked by optimizing the prophylactic, empirical, and therapeutic prescriptions of antibacterial agents [3]. An education in the "right" prescription and a regular supervision protocol of the ecology in the ICU would seem necessary. Strategies that enable us to limit this emergence should be "dogmatically" used in ICUs. The principal strategies are as follows: 1. Use of protocols which optimize antibiotic prescriptions ,; 2. Limitation of antibiotic prescriptions (empirical, prophylactic, and thera

peutic) that are unnecessary and not clinically indicated 3. To consider that using an antibiotic prescription strategy can limit the

occurrence of resistance 4. To consider that a "heterogeneous" choice of antibiotics can contribute to

a potentially effective strategy against the occurrence of resistance 5. Prescription of an antibiotic that is, a priori, effective, according to the

ecology at the time 6. The use of hygiene protocols optimizes infection control 7. Isolation of patients who are colonized or infected by resistant bacteria

142 D. Gruson, G. Hilbert

In 20% - 50% of patients with severe infections the empirical choice of antibiotics is inappropriate to the identified germs. In addition, in 15-25% of other cases, the empirical antibiotics chosen have too-wide a spectrum for the identified germ [4]. Seeking help from theoretical antibiograms or from the literature seems difficult. In fact, the theoretical antibiogram of Staphylococcus au reus does not correspond to a nosocomial Staphylococcus aureus in an ICU ward.

Ibrahim et al. [5] recently reported the results of a protocol aimed at strictly adhering to a policy of antibiotic prescription and its positive impact on reducing unsuitable prescriptions in cases of nosocomial infections, and on the total duration of antibiotic treatments.

Computer-assisted diagnosis and prescription has been reported in studies since the 1990s [6, 7]. The advantages of a cyclic or rotating use of antibiotics had been demonstrated by a team from Salt Lake City [6, 8]. This system makes use of data on the patient and information on the previous use of antibiotics, on the positive bacterial cultures, and on established antibiograms. The resulting information is therefore the "best selection" of empirical antibiotics. This automatic prescription was successfully used to minimize the side-effects of drugs and to optimize the choice of a rapidly effective antimicrobial therapy. This technique stabilized the sensitivity of isolated germs [7, 9]. The cyclic or rotating use of antibiotics could correspond to point no. 4 of the strategy "dogmas" cited above. Such a strategy is, of course, only possible when there is a sufficient pool of antibiotics to choose from. Nevertheless, simply observing the ecology and comparing it with the prescribed antibiotics constitutes an excellent method for reducing the occurrence of resistance [10]. Chow et aL [11] demonstrated the relationship between a previous use of cephalosporins and emergence of resistance to this class of antibiotics in the Enterobacter species.

Restriction in prescription of a class of antibiotics has been used with a view to avoiding an outbreak of resistance. A strict limitation slrould not be used in the case of cyclic or rotating strategies when prescribing antibiotics. In fact, a predetermined prescription cycle of antibiotherapy could consequently lead to a secondary restriction in another type of antibiotherapy. Reducing the use of one class of antibiotics is not always advantageous. This policy has not always proved its efficacy in reducing pharmaceutical expenditure [12, 13]. Certain antibiotic restrictions may be associated with the introduction of new resistant strains [14]. Indeed, antibiotic restrictions are applied to a class of antibiotics that give in rapidly to resistance (third-generation cephalosporins), or to a large spectrum (imipenem), or to potentially toxic antibiotics (aminosides). At present, it is difficult to assert that such a strategy is really beneficial to the ecology

Rotation of Antibiotics - A New Strategy for Prescription in the Intensive Care Unit 143

and the non-appearance of resistant mutants. However, this method remains effective in curbing an epidemic or a particular germ [13, 15]. Quale et aL [16] reported that, despite a very intensive prevention program for nosocomial infections, almost 50% of patients in their hospital had a gastrointestinal colonization of vanco-resistant Enterococcus.

In order to eradicate this epidemic, the decision was made to reduce the prescription of vancomycine and third-generation cephalosporins. In 6 months, the mean usage of ceftazidime and vancomycin was decreased by 55% and 34%, respectively. This was accompanied by a significant reduction in the prevalence of colonization by resistant enterococcus from 47% to 15 % [16,17]. Rahal et al. [14] and Meyer et al. [18] demonstrated that by diminishing the use of third-generation cephalosporins, they reduced the incidence of infection by resistant Klebsiella. Nonetheless, this was associated with a more-frequent prescription of imipenem, which consequently led to the emergence of imipenem-resistant Acinetobacter. Our experiment on strategic antibiotic restriction was also beneficial to our ecology. For us, the context was not that of an epidemic, but an overuse of ceftazidime and ciprofloxacine, and we decided to limit to its maximum the prescription of these antibiotics during a 2-year period [19]. The sensitivity of potentially resistant germs increased after a year of limitation: 61 % versus 79%, and 63% versus 72% in sensitivity of Pseudomonas aeruginosa to ciproflaxocine and ceftazidime, respectively. This favorable result should be considered with precaution, given that restriction was not the only strategy used in this study.

Restricting the use of a certain class of antibiotic is not be the only strategy that decreases the occurrence of resistance in an ICU, especially if the prescription of other antibiotics is not controlled!

Strategic restriction may be used in the case of an epidemic outbreak due to a germ that resists one class of antibiotics. The same strategy could also be useful in the case of a characterization of the occurrence of a resistant germ during an ecology check in an ICU. ""

The notion of restriction as a "strategic arm", either in eradicating an epidemic or in increasing the sensitivity of bacteria to a restricted antibiotic, is, without doubt, at the origin of rotating or cyclic prescription of antimicrobial therapy in the ICUs.

Antibiotic rotation seems to be an effective new approach to the question of reducing the incidence of severe nosocomial infections in ICU wards, notably those due to resistant Gram-negative bacteria. This method involves limiting the prescription of an antibiotic or a class of antibiotic and re-introducing it at a later stage. The main objective of rotation is to decrease, or at least stabilize, resistance to an antibiotic, during the period in which it is not administered. It


has already been shown that using a greater number of antibiotics has a beneficial impact on the emergence of resistance. A limited choice of antibiotics could be harmful to the ecology of the lCU [20]. Recent work has suggested that a "heterogeneous" use of antibiotics could be carried out in a different manner. The different classes of antibiotics could be used in alternation during a pre-defined period, or else prescribed in no particular order in rotation for an undetermined period [21].

The first team to employ this method in the United States was that of Gerding et al. [22]. The rotation concerned gentamycin and amikacin. The rotation was introduced when a high resistance of Gram-negative bacilli to gentamycin appeared. The authors used prescription cycles of 12 - 51 months. The resistance to gentamycin decreased significantly during the period in which amikacin was administered. Secondary re-introduction of gentamycin did not provoke a resistance to this antibiotic. This new strategy would seem to suggest that the rotation of antibiotics within the same class can, in certain circumstances, effectively check the occurrence of resistant bacteria. This said, the consumption of aminosides was reduced in parallel during the 10-year study. These combined factors (rotation and decrease in prescription) seem to be effective in controlling resistance. Other authors have since employed the same technique. Kollef et al. [23] proposed a protocol of antibiotic prescription in a cardiac surgery ward, in the case of an outbreak of Pseudomonas and Klebsiella resistant to third-generation cephalosporins. An initial 6-month period, in which ceftazidime was prescribed for the empirical treatment of nosocomial pneumonia, was followed by a second 6-month period in which ciprofloxacine was administered. The incidence of nosocomial pneumonia was significantly reduced between the two periods, coupled with a lower occurrence rate of pneumonia due to resistant Gram-negative bacilli (Fig. I). Kollef et al. [24] recently reported a single-center study of a cyclic prescription of empirical antibiotics. Three consecutive periods of about 6 months each were analyzed. Ceftazidime was the antibiotic chosen for the first period, which was considered as the base period. The antibiotherapy selected for the second and third periods was ciprofloxacine and cefepime, respectively. These two cycles were compared with the first. The study was carried out on more than 3,500 consecutive patients in surgical and medical lCU wards. As a result of the prescription cycle employed, the authors were able to reduce the number of inappropriate empirical antibiotherapies, and equally the hospital mortality rate (Table 1). Figure 2 represents the percentage of patients receiving inappropriate antibiotic treatments during the two periods of the study of Kollef et al. [23]. However, the authors reported a higher occurrence of Gram-negative bacteria resistant to the antibiotics used in their respective cycles. Kollef et al. [24] suggest that close


Yt of YAP between the 2 period.

" ,. ,.

Period 1 Poriod2

I

I

-all VAP _VAP with resistant

bacteria

Fig.1. Percentage of diagnosed pneumonia during the two study periods by Kollef et al. [23]

GNB = Gram-negative bacilli, GPB = Gram-positive bacteria, VAP = ventilator-associated pneumonia

Patients ('!o) with unsuitable treatment

-

- I--

;---

-I--

... ,- "Ne

~-

-

-,

-, _2 0>0<I0d3

Fig. 2. Percentage of patients receiving unsuitable empirical treatment, in the case of nosocomial infections. Comparison between 3 successive periods of 6 months. Results of Kollef et al. [24] Only the difference between the different cycles for nosocomial infections due to GNB is significant (P <O.OOlfor nosocomial infections)


Table 1. Classification of inadequate antibiotic treatments (%) according to the cycle periods. Results of Kollef et aI .. [24]

Classification Period I n= 1,323

GNB resistant to third-generation cephalosporins 35 (2.6)

GNB resistant to ciprofloxacine 2 (0.2)

GNB resistant to cefepiroe 0

GNB = Gram negative bacteria * : P 0.05 patients in period 2 vs. period I ** : P 0.05 patients in period 3 vs. period I *** : P 0.05 patients in period 3 vs. period 2

Period 2 n=I,243

4 (0.3) *

14 (1.1)*

2 (0.2)

Period 3 n=I,102

8 (0.7)**

4 (0.4)***

10 (0.9)**, ***

monitoring of the evolution of resistance is necessary during a rotation strategy. The same method has been used in a hematology ward [25]. It had no influence on the efficacy of antibiotics in this leukopenic population. The occurrence of Gram-negative infections diminished during the rotation period. On the other hand, the rate of Gram-positive infections rose, notably those due to the Enterococcus species.

Our team used the rotation system for antimicrobial therapy prescriptions in the treatment of ventilator-associated pneumonia, following a worsening ecology [19]. We compared the ecology and the sensitivity of the bacteria responsible for the pneumonia in two periods. The first period, which lasted 2 years, was the "control" period, before the institution of a new antibiotic prescription strategy in the ICU. The technique consisted of a monthly rotation of each class of antibiotics, according to the date of the appearance of a nosocomial pneumonia (late or early stage). Contrary to other studies, the cycle existed for all the antibiotic classes, since the treatment of nosocomial pneumonia in our unit consists of associating two antibiotics: a beta-Iactam associated most often with either an aminoside or a fluoroquilwlone. The cyclic prescription was used for all the antibiotics available in our hospital pharmacy. When comparing the two periods, we observed a drop in the global incidence of pneumonia, in particular, early stage pneumonia. A fall was also seen in the occurrence of pneumonia due to Pseudomonas aeruginosa, Burkholderia cepacia, Stenotrophomonas maltophilia, and Acinotobacter baumanii (Tables 2 and 3). The sensitivity of Gram-negative bacilli improved significantly during the second period (Table 4).

Our study [19] and that of Kollef et al. [24] confirm that a different antibiotic prescription strategy can have a direct effect on the outbreak of mechanical ventilator-associated pneumonia; patients appear to be predisposed to infec-


Table 2. Classification of patients admitted during our study and the number of nosocomial pneumonia (ICU == intensive care unit, YAP == ventilator-associated pneumonia

1st period 2nd period P (2 years) (2 years)

Patients admitted to ICU (n) 1,737 1,718 NS

Patients requiring mechanical ventilation>48 h, n (%) 1,004 (57.8) 1,029 (59.8) NS

Clinical suspicion ofVAP, n patients (%) 294 (28) 203 (19.8). <0.01

Microbiologically documented YAP, n patients (%) 231 (22.1) 161 (15.7). <0.01

Table 3. Description of isolated organisms in bronchoalveolar lavage [19]

1st period 2nd period P

Episodes of polymicrobial VAP, n (%) 75 (32.5) 49 (30.4) NS

Total number of isolated bacteria, n 332 229

Resistant Gram-negati ve bacilli, n (%) 140 (42.2) 79 (34.5) 0.06

Pseudomonas aeruginosa, n (%) 62 (18.7) 47 (20.5) NS

Burkholderia cepacia, n (%) 39 (11.7) 17 (7.4) 0.09

Stenotrophomonas maltophilia, n (%) 19 (5.7) 8 (3.5) NS

Acinetobacter bau111anii, n (%) 20 (6) 7 (3) NS

Other Gram-negative, n (%) 105 (31.6) 74 (32.3) NS

Staphylococcus aureus, n (%) 67 (20.2) 54 (23.6) NS

Methi-sensitive Staphylococcus aurcus, n (% of SA) 27 (40.3) 34 (63) 0.013

"..

tions by resistant bacteria, and the regimen also seems to contribute directly to the pathogenesis of pneumonia.

Raymond et al. [26] very recently studied the impact of a pre-determined empirical antimicrobial therapy prescription program. The authors did a before/after single-center descriptive study, over two I-year periods. During the first period, no protocol was established. This was the base period, which was compared with a second I-year period involving a cyclic empirical antimicrobial therapy for pneumonia. The cycle schedule was as follows: ciprofloxacine plus clindamycin for 3 months, piperacillin/tazobactam for the next 3 months, then carbapenem plus clindamycin for 3 months, and cefepime plus clindamy-


Table 4. Resistant Gram-negative bacilli sensitivity to the antimicrobial treatment [19]

P. aeruginosa B. cepacia S. maltophilia A.haumanii

before after before after before after before after

n=62 n=47 n=39 n=17 n=19 n=8 n=20 n=7

Ticarcillin+ac clavulanic 41.9 44.7 5.1 11.8 94.7 87.5 75 85.7

Pipericillin+tazobactarn 62.9 72.3 87.2 82.4 0 25 50*** 100***

Aztreonam 45.1 55.3 0 5.9 5.3 0 5 0

Cefotaxime 8.1 0 33.3 58.8 0 12.5 5 0

Ceftazidime 62.9 74.5 84.6 76.5 36.8 37.5 10 14.3

Cefepime 53.2* 74.5* 7.7* 41.2* 26.3 25 15 14.3

Cefpirome 38.7* 72.3* 0 41.2 0 0 20 14.3

Imipenem 69.3 76.6 0 5.9 0 0 100 100

Gentamycin 46.8* 74.5* 0 5.9 15.8 12.5 40 28.6

Amikacin 74.2* 91.5* 2.6 5.9 21 12.5 40 85.7

Tobramycin 80.6* 95.7* 0 5.9 10.5 0 45 42.9

Ciprofloxacin 61.3** 78.7** 5.1* 29.4* 42.1 50 10* 50*·

.* P -;; 0.05; ** P = 0.051; *** P = 0.057.

cin for the last 3 months. In approximately 1,500 consecutive patients, 540 episodes of infection were treated. The study revealed a reduction in the rate of infections caused by resistant Gram-negative cocci. (7.8 infections/IOO admission vs. 14.6 infections /100 admissions, P < 0.0001). Infections due to resistant Gram-negative bacilli were also significantly reduced (2.5/ 100 admissions vs. 7.7/ 100 admissions, P < 0.0001). This type of rotation was certainly beneficial in terms of hospital mortality in Ie due to infection (2.9 deaths/IOO admissions vs. 9.6/100 admissions, P <0.0001)

A regular change in the class of antibiotics prescribed in the treatment of nosocomial infections seems useful. No study, however, seems capable of resolving the question of the period length: 3 or 6 months? More or less? Similarly, no study has resolved the question of the best time to re-introduce a


class of antibiotics with a restricted prescription. Certain studies propose a monotherapy treatment for nosocomial infections. Any changes made in this case are from one class to another. For example, Kollef et al. [24] recently switched from a prescription of ceftazidime to ciproflaxocine between period I and period 2. No mention is made in his study of a treatment combing two antibiotics, and therefore a rotation or change in the framework of a bi-therapy. No study has compared, in a rotation perspective, a change within one class of antibiotic, for example a third-generation cephalosporin. The different classes of antibiotic can be used in alternation within a pre-defined period, or prescribed in rotation, in no particular order and for an undetermined period. To our knowledge, ours is the only study dealing with a real cyclic rotation of antimicobial therapy, since the notion of a cycle was by definition, strictly adhered to. In fact, each antibiotic was used in several successive cycles, and was therefore periodically re-introduced, the study period being sufficiently long to enable us to do this.

If the reasons for the success of this new strategy are clearly defined, this method would appear to be effective, and could be considered within a global policy of "reasonable" prescription of antibiotics. No study has demonstrated the superiority of one type of cycle or rotation over another. Rotation seems to be an effective arm, to be used by each according to the ecology of his unit and the antibiotics available.

References 1. Casewell MW, Hill RLR (1986) The carrier state: methicillin-resistant Staphylococcus aureus.

J Antimicrob Chemother 18:A1-A12 2. Fagon JY, Chastre J, Domart Y, et al (1989) Nosocomial pneumonia in patients receiving

continuous mechanical ventilation: prospective analysis of 52 episodes with use of a protected specimen brush and quantitative culture techniques. Am Rev Respir Dis 139:877-884

3. Goldman DA, Weinstein RA, Wenzel RP, et al (1996) Strategies to prevent wd control the emergence and spread of antimicrobial resistant microorganisms in hospitals : a challenge to hospital leadership. JAMA 275:234-240

4. Leibovici L, Gitelman V, Yehezkelli Y, et al (1997) Improving antibiotic empirical antibiotic treatment : prospective, non intervention testing of a decision support system. J Intern Med 242:395-400

5. Ibrahim EH, Ward S, Sherman G, et al (2001) Experience with a clinical guideline for the treatment of ventilator-associated pneumonia. Crit Care Med 29: 1109-1115

6. Pestonik SL, Classen DC, Evans RS, et al (1996) Implementing antibiotic practice guidelines through computer - assisted decision support: clinical and financial outcomes. Ann Intern Med 124:884-890

7. Pestonik SL, Evans RS, Burke JP, et al (1990) Therapeutic antibiotic monitoring surveillance using a computerized expert system. Am J Med 88:43-48

8. Evans RS, Pestonik SL, Classen DC, et al (1998) A computer assisted management program for antibiotics and other antiinfective agents. N Engl J Med 338:232-238


9. Burke JP (1998) Antibiotic resistance - squeezing the balloon? JAMA 280:1270-1271 10. Fridkin SK, Steward CD, Edwards JR, et al (1999) Surveillance of antimicrobial use and

antimicrobial resistance in United States hospitals: project intensive care antimicrobial resistance epidemiology (ICARE) hospitals. Phase 2. Clin Infect Dis 29:245-252

11. Chow JW, Fine MJ, Shlaes OM, et al (1991) Enterobacter bacteremia: clinical features and emergence of antibiotic resistance during therapy. Ann Intern Med 115:585-590

12. McGowan IE Jr, Gerding ON (1996) Does antibiotic restriction prevent resistance? New Horiz 4:370-376

13. Rifenburg RP, Paladino JA, Hanson SC, et al (1996) Benchmark analysis of strategies hospitals use to control antimicrobial expenditures. Am J Health Syst Pharm 23:2054-2062

14. Rahal JJ, Urban C, Hom 0, et al (1998) Class restriction of cephalosporin use to control total cephalosporin resistance in nosocomial Klebsiella. JAMA 280:1233-1237

15. Giamarellou H, Antoniadou A (1997) The effect of monitoring of antibiotic use on decreasing antibiotic resistance in the hospital. In : Symposium on antibiotic resistance: origins, evolution, selection, and spread. Ciba Foundation, London, pp 76-86

16. Quale J, Landman 0, Saurina G, et al (1996) Manipulation of a hospital antimicrobial fonnulary to control an outbreak of vancomycin-resistant enterococci. Clin Infect Dis 23: 1020-1025

17. Quale J, Landman 0, Atwood E, et al (1996) Experience with a hospital- wide outbreak of vancomycin-resistant enterococci. Am J Infect Control 24:372-379

18. Meyer KS, Urban C, Eagan JA, et al (1993) Nosocomial outbreak of Klebsiella infection resistant to late generation cephalosporins. Ann Intern Med 119:353-358

19. Groson 0, G Hilbert, F Vargas, et al (2000) Rotation and restricted use of antibiotics in a medical intensive care unit. Impact on the incidence of ventilator associated pneumonia caused by antibiotic resistant Gram negative bacteria. Am J Respir Crit Care Med 162:837-843

20. KollefMH, Fraser VJ (2001) Antibiotic resistance in the intensive care unit setting. Ann Intern Med 134:298-314

21. Sanders WE Jr, Sanders CC (1997) Circumventing antibiotic resistance in specialized hospital units. Clin Microbiol Infect 3:272-273

22. Gerding ON, Larson TA, Hughes RA, et al (1991) Aminoglycoside resistance and aminoglycoside usage: ten years of experience in one hospital. Antimicrob Agents Chemother 35: 1284-1290

23. Kollef MH, Vlasnik J, Sharpless Let al (1997) Scheduled rotation of antibiotics classes. A strategy to decrease the incidence of ventilator - associated pneumonia due to antibiotic resistant gram negative bacteria. Am J Respir Crit Care Med 156: 1040-1048

24. Kollef MH, Wrad S, Shennan G, et al (2000) Inadequate treatment of nosocomial infections is associated with certain empiric antibiotic choices. Crit Care Med 28:3456-3464

25. Dominguez EA, Smith TL, Reed E, et al (2000) A pilot study of antibiotic cycling in a hematologic-oncology unit. Infect Control Hosp Epidemiol21 :S4-S8

26. Raymond DP, Pelletier SJ, Crabtree TD, et al (2001) Impact of a rotating empiric antibiotic schedule on infectious mortality in an intensive care unit. Crit Care Mea 29:1101-1108

Diagnostic Approach to Sepsis - State of the Art

EM. BRUNKHORST, K. REINHART

Early diagnosis of the different severities of septic inflammation is important for early implementation of specific therapies. Sepsis and severe sepsis are accompanied by clinical and laboratory signs of systemic inflammation. However, patients suffering from non-infectious inflammation may present with similar signs and symptoms, making it difficult to diagnose infection based on clinical findings alone. Bacteriological evidence of sepsis, although definitive and specific, may not be obtainable, is time consuming, and may not occur concurrently with clinical signs of sepsis. It is therefore important to identify markers, which, by enabling an early diagnosis of sepsis and organ dysfunction, would allow early specific therapeutic interventions. Whereas C-reactive protein (CRP) is a more-sensitive parameter for the diagnosis of non-systemic infections, procalcitonin (PCT) seems to be a useful parameter to improve the diagnosis and monitoring of therapy in patients with severe sepsis and septic shock.

Inflammatory response is a common response of the organism and plays a key role in the healing process. The goal of the inflammatory reaction, which consists of humoural, cellular, and molecular factors, is protection from bacterial infection, restriction of tissue destruction, and survival of the host. Changes in body temperature, hypothermia, leukocytosis, leukopenia, tachycardia, hypotension, and hyperventilation are clinical signs of systematic inflammation. However, they may also be non-infectious in origin and are neither specific nor sensitive for sepsis (Table 1).

Blood cultures are positive in only 30-40% of cases [1], and are not specific, as they may be also found in patients without sepsis. The microbiological proof of infection is expensive and may be difficult, especially in the patient already receiving antibiotic treatment [2]. Positive results may be due to colonization or contamination that are without pathophysiological relevance. Despite obvious presence of a focus and clinical signs of sepsis, sepsis cannot be proven microbiologically in 35% of cases [2]. The mortality of 'clinical' sepsis increases when infection remains undocumented [3]. A generalized inflammation with organ dysfunction (SIRS) is caused by many non-infectious diseases,

152 P.M. Brunkhorst, K. Reinhatt

like pancreatitis, major trauma, and burns. Beyond their epidemiological meaning, the criteria from the American College of Chest Physicians/Society of Critical Care Medicine [4] are not helpful for the clinical decisions at the bedside [5]. Furthermore, in order to discover organ dysfunctions associated with severe sepsis, a tight 24-h monitoring of the patient's organ functions is necessary, which may be impossible even in some intensive care units. It is however of tremendous importance to diagnose sepsis in an early stage, to reduce the high mortality (40-60%) and morbidity of this disease.

In the past, biochemical and immunological markers have not been helpful for the diagnosis of sepsis. An ideal marker should be of high sensitivity and specificity, easy to handle, and of low cost for daily routine use. The marker should have a long half-life, and the plasma concentration should correlate with the stage of sepsis, as well as the outcome of the patient.

Humoral and cellular host response

Conventional and new markers for the diagnosis of sepsis are derived from the complex host response to the infectious stimulus. This host response consists of an activation of plasmatic and cellular systems. The activation of plasmatic (complement system, coagulation cascade, kallikrein-kinin-system, eicosanoids) and cellular elements (granulocytes, thrombocytes, macrophages, enthothelial cells) leads to the release of different mediators and molecules (cytokines, chemokines, acute-phase proteins) that regulate the inflammatory response. Although the cellular system reacts differently to gram-negative and gram-positive bacteria, the end is a common pathway.

Cytokines as markers of the septic host response

Increased plasma levels of cytokines are the first response of the host's inflammatory reaction. Cytokines are glycoproteins released by different cells (macrophages, monocytes, lymphocytes, endothelial cells). Their binding to specific receptors leads to defined reactions. Proinflammatory cytokines are elevated in sepsis, although anti-inflammatory cytokines may also be detected, depending on the stage of sepsis. Despite their outstanding importance in the frame of sepsis, they only playa minor role in the clinical routine. Cytokines have a short half-life of a few minutes. The release of cortisol prevents the production of further cytokines and, once in the plasma, numerous cytokines bind to specific receptor antagonists leading to low plasma levels. Additionally,

Diagnostic Approach to Sepsis - State of the Art 153

cytokines are also induced by many non-infectious causes such as surgical treatment or autoimmune disorders, which restricts their specificity. Other problems are the high costs of the assays and the long time until the results are available.

Tumour necrosis factor (TNF)-a, interleukin (IL )-1, IL-6, IL-8, and IL-IO are the most important cytokines associated with sepsis. Since TNF-a plasma levels differ individually, it is of minor diagnostic importance [6]. Measurement of soluble TNF receptors with a long half-life (TNF-R55, TNF-R p75) is inconvenient for the clinical routine because of accumulation. These substances are also elevated in patients suffering from renal dysfunction and chronic heart failure [7]. IL-6 and IL-8 are most closely related to the severity of the physiological response to infection and systematic inflammation [6,8]. IL-6 has shown to be elevated by up to 1,000 times in patients with sepsis. Recently, patients with IL-6 values of greater than 1,000 pg/ml were selected to undergo adjuvant therapy with monoclonal antibodies [9]. The MONARCS trial - a treatment study with monoclonal anti-TNF-a in patients with severe sepsis -showed a tight correlation between the plasma levels of IL-6 and the severity of organ dysfunction and outcome (mortality 47.7% at IL6 > 1,000 pg/ml vs. 28.6% at IL6 < 1,000 pg/ml) in the placebo group [10]. In contrast to TNF-a, the measurement ofIL-6 plasma levels was not influenced by soluble receptors [11]. However, the interindividual differences of IL-6 release into the plasma of septic patients are tremendous, due to the stage-dependent activation of anti-inflammatory cytokines. IL-6 measurements are therefore only meaningful in the assessment of the disease's time course. Persistently elevated IL6leveis are most likely not due to release by monocytes but from non-immunological cells (i.e., endothelial cells). This may be due to peripheral hypoxia caused by malperfusion [12]. In neutropenic patients, IL-6 and IL-8 were able to distinguish between patients with proof of infection and patients with fever of unclear origin, while CRP was not different [13]. In neonatal sepsis, combined measurement ofIL-6 and IL-8 plasma levels can predict an early onset of sepsis with high sensitivity and specificity [14]. However, IL-8 had a low sensitivity in detecting an infected necrosis in patients with necrotic pancreatitis [15]. High levels of IL-6 and IL-8 are also found in patients after major surgery [16], severe trauma [17], autoimmune disorders [18,19], viral infection [20, 21], and after transplant rejection [22]. In summary, it seems that IL-6 and IL-8 are from minor importance in the diagnosis of sepsis and also in the indication of the patient's outcome. There is no systematic study ofIL-6 that clearly shows the importance of IL-6 as a parameter in the early stage of sepsis. Nevertheless, IL-6 may be helpful to detect complications in neonatal sepsis.

154 P.M. Brunkhorst, K. Reinhart

C-reactive protein as a marker of sepsis

CRP is an acute-phase protein released by hepatic cells after stimulation by inflammatory mediators like IL-6 and IL-8 [23]. CRP has both pro- and anti-inflammatory properties. It activates the complement system after binding bacterial polysaccharides or fragments of cell membranes released during infection or trauma. CRP prevents the adhesion of granulocytes on endothelial cells and the synthesis of superoxides, and stimulates the production of IL-l receptor antagonists. CRP is probably the most frequently applied marker to measure the presence and severity of inflammatory reactions. Some studies positively evaluated high plasma levels of CRP in patients with infection and sepsis [13]. CRP was able to discriminate between patients with pneumonia and with tracheobronchitis [24], as well as between bacterial and viral infections [25]. It also correlates with the severity of sepsis [26]. Finally, CRP may help to diagnose appendicitis [27]. However, other studies questioned the meaning of elevated CRP levels for the diagnosis of infection and sepsis, or the assessment of sepsis severity. In contrast to cytokines and PCT, plasma levels of CRP reach their peak after 24 h [28,29] (Fig. 1).

Plas mac once ntrati on

0

0

0

fE

1

fE

TNF

\ill

2

8 0

0

181 0

/'

fE 0

181 fE fE fE

6 12 24 48 72

Time (h)

Fig. 1. Plasma kinetics of various markers of the inflammatory host response to infection [37] (CRP = C-reactive protein, IL-6 = interlukin-6, TNF = tumor necrosis factor, PCT = procalcitonin)


Secondly, plasma levels of CRP also increase during minor infections and do not adequately correlate with the severity of host response or differentiate between survivors and non-survivors of sepsis [30,31]. Thirdly, CRP plasma levels remain elevated up to several days even when the focus of infection is eliminated [32]. Lastly, CRP is found in many states of non-infectious origin, such as autoimmune and rheumatic disorders [33,34], acute coronary syndromes [35], malignant tumors, and after surgery [36].

In summary, the predictive value of CRP for the diagnosis of sepsis is poor, and the correlation with the severity of sepsis remains unproven. Nevertheless, CRP plays an important role for the control of antibiotic therapy in localized infection.

Procalcitonin

PCT is a 13-kilodalton propeptide of calcitonin and is normally produced in the C-ce1ls of the thyroid gland. The hormone calcitonin is secreted into the blood stream after endopeptic cleavage. In healthy individuals, the levels of PCT are below 0.1 ng/ml. In patients with sepsis, PCT levels may increase up to 5,000-10,000 times [37,38], while the levels of calcitonin are still in the normal range [39]. In contrast to the short half-life of calcitonin (10 min), the half-life of PCT approximates to 24 h [28,40].

The physiological role of PCT and the site of production are not completely understood. It has been proven that during sepsis, PCT can be produced by almost all extrathyroid tissues. After stimulation with endotoxin, moncytes showed a 100 fold rise in PCT mRNA [41]. However, the expression of PCT mRNA in almost all extrathyroid tissues in septic hamsters has recently been described [42]. The injection of recombinant PCT reduced the survival rate of these animals, but was without effects in healthy animals. The injected PCT could be antagonized by an antiserum [43]. It is still unclear whether PCT plays a role as a mediator. During clinical observations, very high PCT levels themselves did not seem to affect the patient's health. Bacterial endotoxins are a major stimulus for PCT induction [40], but gram-positive infections may also induce PCT release. During severe fungal infections, PCT induction was described in some patients, whereas a case report described a lack of PCT elevation [44]. Various stimuli other than bacterial infection may induce a PCT increase under certain clinical conditions, including major surgery, severe trauma, or bums. However, plasma levels observed under these conditions are not as high as in patients with severe sepsis or septic shock. PCT elevation is recognized 2 h after the microbiological impact, which is faster than CRP but

156 F.M. Brunkhorst, K. Reinhart

slower than the cytokines [28, 40]. PCT is very stable ex vivo: 90% of its concentration is still detectable after 12 h of storage at room temperature [45]. A number of studies support PCT as a marker of severe infections and sepsis [46]. Patients with PCT levels below or equal to 0.5 ng/ml are unlikely to have severe sepsis or septic shock, whereas levels above a threshold of 5.0 ng/ml identify patients at high risk [47, 48]. PCT concentrations exceeding 10 ng/rnl occur usually in patients with organ failure remote to the site of infection [29, 32]. A localized focus of bacterial infection without systemic inflammation is not accompanied by an increased PCT, but by increased levels of CRP [49]. Similarly, in patients with severe infection, appropriate therapy may reduce PCT to very low levels without indicating the complete eradication of infection. In these patients, further antibiotic or surgical focus control may be required despite normal or low PCT levels [50, 51]. When PCT is released non-specifically due to major surgery or severe trauma, daily monitoring may be helpful to early detect septic complications [52]. For instance, during the first 48 h of life a 20 fold increase ofPCT concentrations had been reported. By establishing 95% hour-specific reference ranges, studies demonstrated a sensitivity of 92.6% and a specificity of 97.5% for detecting an early onset of sepsis in neonates [53]. In prolonged cardiogenic shock, increased PCT levels were detected, associated with elevated plasma levels of TNF-a, IL-6, and soluble CD 14, which is the soluble part of the monocyte membrane receptor for endotoxin. The authors explained these effects with a translocation of endotoxin due to an impaired perfusion of the mucosa [54, 55]. These elevated PCT levels dropped to normal in surviving patients who had received a biventricular assist device due to end-stage cardiac failure. PCT levels remained elevated or increased even more in patients dying of septic complications [56]

The diagnosis of severe sepsis correlates strongly with a significant increase in PCT concentrations. This relationship does not exist in patients only fulfilling the criteria of the systemic inflammatory response syndrome (SIRS) (Fig. 2) [32, 49].

One study reported a greater predictive value for CRP in the prediction of septic complications [31]. However, this study did not include patients with severe sepsis and, thus, organ dysfunction remote to the site of infection.

In several studies investigating outcome prediction in critically ill patients, peT proved to be superior to TNF-a, IL-6, and CRP [47, 57]. Initially elevated PCT levels in poly traumatized patients were indicative of the risk for developing septic complications and multiple organ failure [58]. This was found to be similar in patients after abdominal surgery [52]. Patients after cardiac surgery, who presented elevated PCT levels 5.0 ng/ml I day after cardiopulmonary bypass, had a higher in-hospital mortality (Fig. 3) [59].

Diagnostic Approach to Sepsis - State of the Art

3)

18

16 j

14 1

~12 j c:

;:10 I

(.) 0-

8 ,

61

4 ,

2 ·

0 -

25

10

SIRS

SIRS

,- "" --I

sepsis

r- NS --,

sepsls

seWI9 sepsis

~ NS --,

157

Fig. 2. C-reactive protein (CRP) and procalcitonin (PCT) concentrations (mean±STD) in 185 patients with systemic infammatory response syndrome (SIRS) (N=l?). sepsis (=severe infection with;::: 2 SIRS Kriterien, n=6l), severe sepsis (= remote organ dysfunction, N=68), and septic shock (N=39), * P< 0,005, ** P< 0,001 [49]

158

- 500 ... G) .c 400 E :::::s 300 c -(/) 200 -r::::: G)

100 :;= co a..

0

F.M. Brunkhorst, K. Reinhart

Survivors Non-su rVivors

I P<O.03 II P<O.002 I

(1.8%) (5.9%) (21.8%)

<1.0 1.0-5.0 >5.0

PCT (ng/ml) on day 1 after CPS

Fig.3. PCT plasma levels in 691 patients on day 1 following cardiopulmonary bypass (CPB) in open heart surgery. An increase of PCT I ng/ml was highly predictive of all-cause hospital mortality (in parentheses) [59]

PCT may also be helpful in the differentiation of bacterial and viral infections. Children with bacterial meningitis had significantly higher levels of PCT than those suffering from viral meningitis [20]. The differentiation of infectious and non-infectious causes of the acute respiratory distress syndrome by PCT measurement has also been described. Interestingly, the extent of hypoxemia did not contribute to the increase of the PCT level [60]. Likewise, in patients after liver, heart, and kidney transplantation, systemic fungal or bacterial infections may be differentiated from graft rejection [61-63]. PCT is also elevated in septic patients with chemotherapy induced neutropenia suggesting that white blood cells are not the only source of PCT [64]. In patients with necrotizing pancreatitis, PCT was the best predictor of infection of the pancreatic necrosis when compared with CRP and IL-8, with a predictive power almost equal to the gold standard: the fine needle biopsy [15]. Lastly, PCT was able to distinguish between presence and absence of systemic infections in patients with highly active autoimmune diseases [33]. However, this observation was undertaken with only a few cases and needs proof in a study with a larger sample size.

In summary, PCT is a new marker that correlates very well with the onset of organ dysfunction remote to the site of infection. A differentiation between


infectious and non-infectious etiology of systemic inflammation seems to be possible, which may have fundamental therapeutic consequences at the bedside. Beside the earlier diagnosis, the progress of focus control and the efficacy of antibiotic treatment can be monitored. However, one should bear in mind that clinical situations, where hemodynamic failure might cause translocation of endotoxin, also induce an increase of PCT levels.

Complement 3a

C3a is a proinflarnmatory mediator, derived from the a-chain of C3 after activation of the classical and alternative pathway of the complement cascade. The assay for C3a by column chromatography is complicated and expensive. It is known that systematic inflammation leads to 40 times increased levels of C3a [65]. However, studies on critically ill patients with a large sample size have not been undertaken. A recently published study showed a significantly higher plasma concentration of C3a in 22 septic patients compared with 11 patients with SIRS [66]. The same result was found with PCT and IL-6, but not with CRP and elastase. However, the different stages of sepsis severity were not assessed in this study. The importance of this parameter in the diagnosis of sepsis remains therefore unclear, especially considering the low concentration range of C3a.

Neutrophil elastase

Neutrophil elastase is a serine-proteinase released from neutrophil granUlocytes. Increased levels occur both during sepsis and after trauma [67-69]. Elastase concentrations correlate only poorly with the severity of inflammation. Thus, elastase measurements have not gained much importance as a marker of sepsis.

Neopterin

Neopterin is substance of low molecular weight which is synthesized by GTP. Interferon is the central stimulation for the synthesis of neopterin. The function of neopterin is probably associated with the cytotoxic reactivity of activated macrophages [70, 71]. Since elevated plasma levels are found in all infectious and non-infectious inflammations as well as in malignant diseases, the specificity of neopterin as a marker of systemic infection is limited [27]. Another


disadvantage is the long time of induction (>24 h) and the accumulation in patients with renal failure.

HLA-DR

HLA-DR is a surface antigen expressed on monocytes. Sepsis and severe infections suppress HLA-DR expression. This status of immune paralysis characterizes a high-risk patient population. The degree of suppression of the monocytic HLA-DR expression correlates with the severity and the outcome of sepsis [72, 73]. However, this only occurs in a subgroup of septic patients and may also occur in patients after major surgery [74, 75]. Beside its value for outcome prediction, HLA-DR measurement does not playa role in the diagnosis of sepsis. Additionally, the method is complicated and the results hard to repeat.

Phosopholipase A2

Phospholipases are localized in all tissue cells and inflammatory cells. Phospholipase A2 (PLA2) generates arachidonic acid from phospholipids, which then is metabolized to prostaglandins via the cyclooxygenase pathway, and to leukotrienes via the lipooxygenase pathway. PLA2 associated enzymes playa key role in the inflammatory response. Although PLA2 (especially type II) is associated with severity and outcome of sepsis [76, 77], it has not been investigated as a diagnostic tool for this disease. PLA2 consists of a family of four different enzymes and it may be difficult to distinguish them [78]. Which PLA2 enzyme is measured depends on the assay. It is unknown which PLA2 enzyme should be measured in sepsis.

Endotoxin (lipopolysaccharide)

Endotoxin is an essential structure of the outer cell membrane of gram-negative bacteria. Methodological problems of the currently used LAL-bioassay, with variations of more than 50%, low specificity due to differences in the endotoxin structure in several gram-negative bacteria, interactions with plasma proteins and antibiotics, may explain the different results of the currently available studies. Endotoxin measurement cannot be recommended for the clinical setting [79]. A new chemoluminescence assay might improve sensitivity and specificity of endotoxin measurement in the future [80].


Lipopolysaccharide-binding protein

Lipopolysaccharide-binding protein (LBP) was first described in 1990 [81]. It is a 58-kilodalton class 1 acute-phase protein that mediates the endotoxin-induced activation of monocytes by the CD 14-receptor and, thus, the production ofIL-6. Recently it was shown that the same mechanism is responsible for peptidoglycan release in gram-positive bacteria. Therefore, sepsis by gram-positive bacteria might be identified by elevated LBP levels [82]. LBP is mainly synthesized by hepatocytes, intestinal and pulmonal epithelial cells. The normal plasma level for LBP ranges from 5 to 15 J.lg/ml. A few reports described levels up to 30 times of the normal value [83]. The time of induction (36 h) is slow compared with PCT and CRP [84]. Valid studies ofLBP are still unavailable.

Protein C

Protein C is a vitamin K-dependent zymogen from a serinprotease. It has anticoagulant properties via the inactivation of factors Va and VIlla. Depending on the applied assay, the plasma activity varies from 75% to 140%. Recently, a reduction of the 28-day mortality was demonstrated in a study with recombinant activated protein C in patients with severe sepsis [85]. Interestingly, patients with normal levels of activated protein C also had a benefit from this therapy. However, the correlation between low plasma concentration of protein C and the severity of inflammation has never been proven in a study with a large number of critically ill patients [86]. Since protein C is secreted by hepatic cells, decreased levels of protein C can be found in patients with hepatic failure and cancer. Additionally, any kind of coagulation activation (pulmonary embolism, venous thrombosis) causes protein C levels to drop. The relatively short half-life of protein C (10 h) should cause a low concentration after any shock-induced hepatic hypoperfusion. Currently, a recommendation for the measurement of protein C in the clinical routine cannot be given.

Other parameters

Other mediators, metabolites, and proteins such as granulocyte-colony-stimulating factor, soluble adhesion molecules, or products of nitric oxide metabolism are altered in different manners during sepsis. Beyond the scope of research, these parameters have no application in clinical practice.


Conclusion

The differential diagnosis of severe systemic inflammatory reactions in routine clinical practice is difficult, expensive and time consuming. The onset of organ dysfunction remote to the site of infection (severe sepsis) delays the diagnosis of sepsis and results in a high mortality of 35-70% or, depending on comorbidities, to a longer hospital stay. An early diagnosis with sensitive and specific biochemical and immunological markers may contribute to reduce this high morbidity and mortality. Clinical signs such as body temperature, leukocytosis, and CRP are unable to reflect the patient's immune status or the inflammatory host response (Table 1).

Table 1. Comparison of various markers in clinical use for the diagnosis of sepsis

Specific Sensitive response for to infection int1ammation

PCT 4+ +

CRP 2+ 2+

Cytokines + 3+

Leukocytes +

Temperature + 4+

Clinical use: advantages

rapid induction (2 h)

high biostability

half-life 24 h wide biological range not expensive

high sensitivity

rapid induction (minutes)

simple method high sensitivity simple method high sensitivity

disadvantages

low sensitivity for local infections

high specificity for severe sepsis, septic shock release in "severe SIRS" settings expensive low specificity slow induction time (peak> 24 hrs) low biological range no correlation with seveJity of int1ammation short half-life in plasma (minutes) high variability low biostability low correlation with severity of int1ammation

expensive very low specificity

very low specificity

Proinflammatory cytokines such as IL-6 and TNF-a are scientific parameters than valid diagnostic tools at the bedside because of their short half-life and the lack of standardization of the available assays. PCT has numerous advantages over other parameters due to the wide biological range (increases up to 5,000 times), short time of induction after the bacterial stimulus (2 h), long half-life (24 h), and the simple sample preparation. Furthermore, the plasma


levels ofPCT correlate with the severity of organ dysfunction remote to the site of infection and, therefore, with the outcome. PCT plasma levels also correlate with the success of treatment. PCT predicts elevated IL-6 and TNF-a levels in patients with sepsis [57]. The characteristic behavior ofthis parameter, i.e., not rising during severe local reactions (sepsis according to the consensus criteria), needs special attention. A lack of PCT elevation should not cause the physician to avoid a search for a focus or induction of antibiotic treatment. On the other hand, the outcome of these special patients is characterized by the switch from a local to a systemic infection, which is difficult to detect by clinical signs only. This emphasizes the importance of a daily monitoring of PCT values. As CRP levels are also elevated during local infections, and as the treatment success correlates with a drop in CRP levels, the monitoring of CRP is superior to PCT measurement in this situation. However, CRP is of little importance for the identification of infection due to its low sensitivity and specificity.

The mortality of patients suffering from sepsis correlates with the severity of the inflammatory host response [2]. Therefore, the physician needs a marker to early discriminate patients with infections and organ dysfunction (severe sepsis), for the early induction of adequate treatment. In this context, the measurement of PCT plasma levels represents important progress in the diagnosis of sepsis [87].

References 1. Reimer LG, Wilson ML, Weinstein MP (1997) Update on detection of bacteremia and fungemia.

Clin Microbiol Rev 10:444-465 2. Rangel-Frausto MS, Pittet D, Costigan M, et al (1995) The natural history of the systemic

inflammatory response syndrome (SIRS). A prospective study. JAMA 273:117-123 3. Reyes WJ, Brimioulle S, Vincent J-L (1999) Septic shock without documented infection: an

uncommon entity with a high mortality. Intensive Care Med 25:1267-1270 4. American College of Chest Physicians-Society of Critical Care Medicine Consensus Conference

(1992) Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 328:864-875

5. Abraham E, Matthay MA, Dinarello CA, et al (2000) Consensus conference definitions for sepsis, septic shock, acute lung injury, and acute respiratory distress syndrome: time for a reevaluation. Crit Care Med 28:232-235

6. Wakefield CH, Barclay GR, Fearon KC, et al (1998) Proinflammatory mediator activity, endogenous antagonists and the systemic inflammatory response in intra-abdominal sepsis. Scottish Sepsis Intervention Group. Br J Surg 85:818-825

7. Paulus WJ (1999) How are cytokines activated in heart failure? Eur J Heart Fail 309-312 8. Hamano K, Gohra H, Noda H, et al (1998) Increased serum interleukin-8: correlation with poor

prognosis in patients with postoperative multiple organ failure. World J Surg 22:1077-1081 9. Reinhart K, Wiegand-Lohnert C, Grimminger F, et al (1996) Assessment of the safety and

efficacy 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. Crit Care Med 24:733-742


10. Panacek E, Albertson T, Johnson D, et al (2000) Septest, a rapid, imrnunostrip test for serum IL-6 concentrations, prospectively defined a population of patients with sepsis at greater risk for mortality in the MONARCS (monoclonal anti-TNF: a randomized controlled sepsis) trial. Poster presented at the 40th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, Canada

II. Hack CE, De-Groot ER, Felt-Bersma RJ, et al (1989) Increased plasma levels of interleukin-6 in sepsis. Blood 74: 1704-1710

12. Hasper D, Hummel M, Kleber FX, et al (1998) Systemic inflammation in heart failure. Eur Heart J 19:761-765

13. Engel A, Mack E, Kern P, Kern WV (1998) An analysis of interleukin-8, interleukin-6 and C-reactive protein serum concentrations to predict fever, gram-negative bacteremia and complicated infection in neutropenic cancer patients. Infection 26:213-221

14. Berner R, Niemeyer CM, Leititis JU, et al (1998) Plasma level and gene expression of granulocyte colony-stimulating factor, tumor necrosis factor-alpha, interleukin (IL)-1 beta, IL-6, IL-8, and soluble intercellular adhesion molecule-l in neonatal early onset sepsis. Pediatr Res 44:469-477

15. Rau B, Steinbach G, Gansauge F, et al (1997) The role of pro calcitonin and interleukin-8 in the prediction of infected necrosis in acute pancreatitis. Gut 41 :832-840

16. Shenkin A, Fraser WD, Series J, et al. (1989) The serum interleukin-6 response to elective surgery. Lymphokine Res 42:446-454

17. Nast-Kolb D, Waydhas C, Gippner-Steppert C, et al (1997) Indicators of the posttraumatic inflammatory response correlate with organ failure in patients with multiple injuries. J Trauma 42:446-454

18. Kutukculer N, Caglayan S, Aydogdu F. (1998) Study of pro-inflammatory (TNF-alpha, ILl alpha, IL-6) and T-cell-derived (IL-2, IL-4) cytokines in plasma and synovial fluid of patients with juvenile chronic arthritis: correlations with clinical and laboratory parameters. Clin Rheumatol 17:288-292

19. Robak T, Glaalska A, Stepien H, Robak E. (1998) Serum levels of interleukin-6 type cytokines and soluble interleukin-6 receptor in patients with rheumatoid arthritis. Mediators Inflamm 7:347-353

20. Gendrel D, Raymond J, Assicot M, et al (1997) Measurement of procalcitonin levels in children with bacterial or viral meningitis. Clin Infect Dis 24: 1240-1242

21. Malaguarnera M, Di Fazio I, Romeo MA, et al (1997) Elevation of interleukin-6 in patients with chronic hepatitis due to hepatitis C virus. J GastroenteroI32:211-215

22. Kita Y, Iwaki Y, Noguchi K, et al (1996) Daily serum interleukin-6 monitoring in multiple organ transplantation with or without liver allografts. Transplant Proc 28: 1229-1234

23. Gabay C, Kushner I (1999) Acute-phase proteins and other systemic responses to inflammation. N Engl J Med 340:448-454

24. Dev D, Wallace E, Sankaran R, et al (1998) Value of C-reactive protein measurements in exacerbations of chronic obstructive pulmonary disease. Respir Med 92:664-667

25. Shaw AC (1991) Serum C-reactive protein and neopterin concentrations in patients with viral or bacterial infections. J Clin PathoI44:596-599

26. Povoa P, Almeida E, Moreira P, et al (1998) C-reactive protein as an indicator of sepsis. Intensive Care Med 24: 1052-1056

27. Erikson S, Granstrom L, Olander B, Wretlind B, (1995) Sensitivity of interleukin-6 and C-reactive protein concentrations in the diagnosis of acute appendicitis. Eur J Surg; 161 :41-45

28. Brunkhorst FM, Heinz U, Forcki ZF (1998) Kinetics of procalcitonin in iatrogenic sepsis. Intensive Care Med 24:888-889

29. Monneret G, Labaune JM, Isaac C, et al (1997) Procalcitonin and C-reactive protein levels in neonatal infections. Acta Paediatr 86:209-212

30. Meisner M, Tschaikowsky K, Palmaers T, Schmidt J (1999) Comparison of procaIcitonin (PCT) and C-reactive protein (CRP) plasma concentrations at different SOFA scores during the course of sepsis and MODS. Crit Care 3:45-55


31. Ugarte H, Silva E, Mercan D, et al (1999) Procalcitonin used as a marker in the intensive care unit. Crit Care Med 27:498-504

32. Meisner M, Tschaikowsky K, Palmaers T, Schmidt J (1999) Comparison of procalcitonin (PCT) and C-reactive protein (CRP) plasma concentrations at different SOFA scores during the course of sepsis and MODS. Crit Care 3:45-55

33. Eberhard OK, Haubitz M, Brunkhorst FM, et al (1997) Usefulness of procalcitonin for differentiation between activity of systemic autoimmune disease (systemic lupus erythematosus/systemic antineutrophil cytoplasmatic antibody-associated vasculitis) and invasive bacterial infection. Arthritis Rheum 40:1250-1256

34. Schwener V, Sis J, Breitbart A, Andrassy K (1998) CRP levels in autoimmune disease can be specified by measurement of procalcitonin. Infection 26:274-276

35. Lindahl B, Toss H, Siegbahn A, et al. for the FRISC study group (2000) Markers of myocardial damage and inflammation in relation to long-term mortality in unstable coronary artery disease. N Engl J Med 343:1139-1147

36. Meisner M, Tschaikowsky K, Hutzler A, et al (1998) Postoperative plasma concentrations of procalcitonin after different types of surgery. Intensive Care Med 24:680-684

37. Meisner M, Tschaikowsky K, Beier W, Schuttler J (1996) Procalcitonin (PCT) - ein neuer Parameter zur Diagnostik und Verlaufskontrolle von bakteriellen Infektionen und Sepsis. Anasth Intensivmedizin 10:529-539

38. Gramm HJ, Dollinger P, Beier W (1995) Procalcitonin: ein neuer Marker der inflammatorischen Wirtsantwort. Longitudinalstudien bei Patienten mit Sepsis und Peritonitis. Chir Gastroenterol 11 [Suppl 2]:51-54

39. Assicot M, Gendrel D, Carsin H, et al (1993) High serum procalcitonin concentrations in patients with sepsis and infection. Lancet 341:515-518

40. Dandona P, Nix D, Wilson MF, et al (1994) Procalcitonin increase after endotoxin injection in normal subjects. J Clin Endocrinol Metab 79:1605-1608

41. Oberhoffer M, Stonans I, Russwurm S, et al (1999) Procalcitonin expression in human peripheral blood mononuclear cells and its modulation by lipopolysaccharides and sepsis related cytokines in vitro. J Lab Clin Med 134:49-55

42. Muller B, White JC, Nylen ES, et al. (2001) Ubiquitous expression of the calcitonin-I gene in multiple tissues in response to sepsis. J Clin Endocrinol Metab 86:396-404

43. Nylen ES, Whang KT, Snider Jr RH, et al (1998) Mortality is increased by procalcitonin and decreased by an antiserum reactive to procalcitonin in experimental sepsis. Crit Care Med 26: 1001-1006

44. Beaune G, Bienvenue C, Pondarre G, et al (1998) Serum procalcitonin rise is only slight in two cases of disseminated aspergillosis. Infection 26:168-169

45. Meisner M, Tschaikowsky K, Schnabel S, et al (1997) Procalcitonin: influence of temperature, storage, anticoagulation and arterial or venous asservation of blood samples on procalcitonin concentrations. Eur J Clin Chern Clin Biochem 35:597-601

46. Meisner M. (2000) Procalcitonin. A new, innovative infection parameter, 3rd edn, Thieme Verlag Stuttgart

47. Werra I de , Jaccard C, Corradin SB, et al (1997) Cytokines, nitrite/nitrate, soluble tumor necrosis factor receptors, and procalcitonin concentrations: comparisons in patients with septic shock, cardiogenic shock, and bacterial pneumonia. Crit Care Med 25:607-603

48. Muller B, Becker KL, Schachinger H, et al (2000) Calcitonin precursors are reliable markers of sepsis in a medical intensive care unit. Crit Care Med 28:977-983

49. Brunkhorst FM, Wegscheider K, Forycki ZF, Brunkhorst R (2000) Procalcitonin for early diagnosis and differentiation of SIRS, sepsis, severe sepsis, and septic shock. Intensive Care Med 26 [Suppl]:SI48-S152

50. AI-Nawas B, Krammer I, Shah PM (1996) Procalcitonin in diagnosis of severe infections. Eur J Med Res 1:331-333

51. Gramm HJ, Hannemann L (1996) Activity markers for the inflammatory host response and early criteria of sepsis. Clin Intensive Care 7 [Suppll]:1-320-321


52. Reith HB, Mittelkotter U, Debus ES, et al (1998) Procalcitonin in early detection of postoperative complications. Dig Surg 5:260-265

53. Chiesa C, Panero A, Rossi N, et al (1998) Reliability of procalcitonin concentrations for the diagnosis of sepsis in critically ill neonates. Clin Infect Dis 26:664-672

54. Brunkhorst FM, Clark AL, Forycki ZF, Anker SD. (1999) Pyrexia, procalcitonin, immune activation and survival in cardiogenic shock: the potential importance of bacterial translocation. Int J Cardiol 72:3-10

55. Brunkhorst FM, Niebauer J, Anker SD (1999) Endotoxin and immunactivation in chronic heart failure (letter). Lancet 354: 599-600

56. Loebe M, Lodziewski S, Brunkhorst FM, et al. (2000) Procalcitonin (PCT) in cardiac surgery. In: Bakut D, Krian A (eds.) Current perspectives of the extracorporeal circulation. Springer Berlin Heidelberg New York, pp 127-135

57. Oberhoffer M, Karzai W, Meier-Hellmann A, et al (1999) Sensitivity and specifity of various markers of inflammation for the prediction of tumor necrosis factor alpha and interleukin-6 in patients with sepsis. Crit Care Med 27:1814-1818

58. Wanner GA, Keel M, Steckholzer U, et al. (2000) Relationship between procalcitonin plasma levels and severity of injury, sepsis, organ failure, and mortality in injured patients. Crit Care Med 28:950-957

59. Loebe M, Lociewski S, Brunkhorst FM, et al (2000) Procalcitonin in patients undergoing cardiopulmonary bypass in open heart surgery-first results of the Procalcitonin in Heart Surgery study (ProHearts). Intensive Care Med 26 [Supp 2]:193-198

60. Brunkhorst FM, Eberhard OK, Brunkhorst R (1999) Discrimination of infectious and noninfectious causes of early acute respiratory distress syndrome by procalcitonin. Crit Care Med 27:2172-2176

61. Eberhard OK, Langefeld I, Kuse E, et al (1998) Procalcitonin in the early phase after renal transplantation: will it add to diagnostic accuracy? Clin Transplant 12:206-211

62. Kuse E-R, Langefeld I, Jager K, Kiilpmann WR (2000) Procalcitonin in fever of unknown origin after liver transplantation: a variable to differentiate acute rejection from infection. Crit Care Med 28:555-559

63. Hammer S, Meisner F, Dirschedl P, et al (1998) Procalcitonin: a new marker for diagnosis of acute rejection and bacterial infection in patients after heart and lung transplantation.Transpl ImmunoI6:235-241

64. AI-Nawas B, Shah PM (1996) Procalcitonin in patients with and without immunosuppression and sepsis. Infection 24:434-436

65. Hack CE, Nuijens JH, Felt-Bersma RJ, et al (1989) Elevated plasma concentrations of the anaphylatoxins C3a and C4a are associated with a fatal ourtcome in sepsis. Am J Med 86:20-26

66. Selberg 0, Hecker H, Martin M, et al (2000) Discrimination of sepsis and systemic inflammatory response syndrome by determination of circulating plasma concentrations of procalcitonin, protein complement 3a, and interleukin-6. Crit Care Med 28:2793-2798

67. Pacher R, Redl H, Fraas M, et al (1989) Relationship between neopterin and granuolcyte elastase plasma levels and the severity of multiple organ failure. Crit Care Med 17:221-226

68. Janoff A (1985) Elastase in tissue injury. Annu Rev Med 36:207-216 69. Gardinali M, Padalino P, Suffredini A, et al (1992) Complement activation and polymorphonu

clear neutrophil leucocyte elastase in sepsis. Arch Surg 127: 1219-1224 70. Fuchs D (1992) The role of neopterin as a monitor of cellular immune activation in transplanta

tion, inflammatory, infectious, and malignant diseases. Crit Rev Clin Lab Sci 29:307-341 71. Fuchs D, Hausen A, Reibnegger G, et al (1998) Neopterin as a marker for activated cell-mediated

immunity: application in HIV infection. Immunol Today 9:150-155 72. Docke WD, Randow F, Syrbe U, et al (1997) Monocyte deactivation in septic patients: restoration

by IFN-gamma treatment. Nat Med 3:678-681 73. Yolk HD, Reinke P, Krausch D, et al (1996) Monocyte deactivation-rationale for a new

therapeutic strategy in sepsis. Intensive Care Med 22[Suppl]4:S474-S481


74. Haupt W, Riese J, Mehler C, et al (1998) Monocyte function before and after surgical trauma. Dig Surg 15:102-104

75. Asadullah K, Woiciechowsky C, Docke WD, et al (1995) Immunodepression following neurosurgical procedures. Crit Care Med 23:1976-1983

76. Uhl W, Beger HG, Hoffmann G, et al (1995) A multicenter stndy of phospholipase A2 in patients in intensive care units. J Am ColI Surg 180:323-331

77. Rinalta EM, Nevalainen TJ (1993) Group II phospholipases in sera of febrile patients with microbiologically or clinically documented infections. Clin Infect Dis 269:278-288

78. Yang HC, Mosior M, Johnson CA, et al (1999) Group-specific assays that distinguish between the four major types of mammalian phospolipase A2. Ann Biochem 269:278-288

79. Cohen J (2000) The detection and interpretation of endotoxaemia. Intensive Care Med;26:S51--S56

80. Romaschin AD, Harris DM, Ribeiro MB, et al (1998) A rapid assay of endotoxin in whole blood using autologous neutrophil-dependent chemiluminescence. J Immunol Methods 212:169-185

81. Schumann RR, Leong SR, Flaggs GW, et al (1990) Structure and function oflipopolysaccaride binding protein. Science 249: 1429-1431

82. Pugin J, Heumann D, Tomasz A et al (1994) CD14 is a pattern recognition receptor. Immunity 1:509-516

83. Opal SM, Scannon PJ, Vincent J-L, et al (1999) Relationship between plasma levels of lipopolysaccaride (LPS) and LPS-binding protein in patients with severe sepsis and septic shock. IInfectDis 180:1584-1589

84. Erwin PJ, Lewis H, Dolan S, et al (2000) Lipopolysaccaride binding protein in acute pancreatitis. Crit Care Med 28: 104-109

85. Bernard GR, Vincent J-L, Laterre P-F, et al (2001) For the Recombinant Human Activated Protein C Worldwide Evaluation in Severe Sepsis (PROWESS) Stndy Group. Efficacy and safety for recombinant human activated protein C for severe sepsis. N Engl J Med 344:699-709

86. Fisher CJ, Yan B (2000) Protein C levels as a prognostic indicator of outcome in sepsis and related diseases. Crit Care Med 28 [Suppl]:S49-S56

87. Matot I, Sprung CL (2001) Definition of sepsis. Intensive Care Med 27 [Suppl]:S3-S9

Septic Shock Therapy

R. FUMAGALLI, D. CODAZZI, S. CATIANEO

Sepsis represents one of the pathological conditions that can develop during an intensive care unit (lCU) stay or can be the cause of admission to the ICU. In the past the lack of a commonly accepted definition of this pathology led to a common misinterpretation of morbility and mortality [1-3].

Table 1. Incidence of severe sepsis and septic shock

Author Severe sepsis Septic shock

Rangel-Fausto et al. 1995 [1] 18 4

Pittet et al. 1995 [2] 16 7

Salvo et al. 1995 [3] 2 3

Saez-Llorens et al. 1995 [4] 61 18

Proulx et al. 1996 [5] 4 2

Jones and Lowes 1996 [6] 5 3

Bassink et al. 1998 [7] 14 20

From Definition of sepsis I. Most Intensive Care Med (2001) 27:53-59

According to this poorly defined description of sepsis, its incidence has varied between 2 and 11 % of the patients admitted to the ICU, with a mortality ranging between 25% and 80% (Tables 1,2) [4-6].

One of the aims of ICU researchers was to obtain a uniform definition of sepsis. For this reason, ACCP and SCCM provided a consensus in order to better define patients with infection and the systemic response to it.

Systemic inflammatory response syndrome is defined by the simultaneous presence of the following:

170 R. Fumagalli, D. Codazzi, S. Cattaneo

Table 2. Mortality of severe sepsis and septic shock

Author Severe sepsis (%) Septic shock (%)

Rangel-Fausto et al. 1995 [I) 20 46

Pittet et al. 1995 [2] 35 58

Salvo et aI. 1995 [3] 52 82

Saez-Llorens et al. 1995 [4] 40 62

Jones and Lowes 1996 [6] 38 56

Bassink et al. 1998 [7] 18 53

From: Definition of sepsis I. Most Intensive Care Med (2001) 27:53-59

Core temperature> 38.S oC or < 36°C Heart rate >90 beats/min (bpm) (in the absence of drugs interfering with the heart conduction system) Respiratory rate> 20 bpm or paC02 < 32 mmHg or on mechanical ventilation White blood cells> 12,000 or < 4,000/mm3 or band forms> 10% Lactic acidosis

According to the previous indications, sepsis represents the systemic reactions of the host to infection [7,8]. Cytokines are involved in the pathogenesis of the alteration of organ perfusion; the progression from dysfunction to the complete failure of the organ functions accounts for the high mortality of this syndrome.

Besides diagnosis, which represents a cornerstone for the correct management, therapy of sepsis recognizes the following issues: Therapy against infection (antibiotics, drainage, surgical operation, etc) Hemodynamic support (vasoactive drugs, therapy of the occult adreno-cortical insufficiency) Modulation of anti-inflammatory response (anti-tumor necrosis factor, steroids) Management of endothelial damage and of microthrombosis associated with sepsis Support of organ function (ventilatory support, renal replacement therapy)

Therapy against infection

The first aim is the identification of the source of infection. As known by the

Septic Shock Therapy 171

Greeks, the removal of the source of infection, if possible, either an abscess or a perforated viscus, remains the basic "conditio sine qua non". Waiting until the source of infection is identified and the injudicious use of broad-spectrum antibiotics with a high tissue penetration should be avoided. Few studies have dealt with this type of issue and no randomized studies had been performed to show the difference between conservative and aggressive strategies. A systematic study of the effect of different source control strategies appears to be quite complicated: too many variables need to be taken into account to obtain comparable data. It is, however, the opinion of the authors that even if the patient's conditions appear complicated, the delay in removing the source of infection can be detrimental [9].

Some cohort studies, dealing with new anti-infective agents, have shown that failure to identify the source of infection is associated with a higher mortality. At present, there is a substantial agreement on the mechanical removal of the septic source.

Antibiotic therapy

The removal of the source of infection, if possible, is mandatory; the use of antibiotics to decrease the chance of spread of the infection follows the same rule [10]. To choose the right antibiotic it is essential to know the site of infection, the pharmacological properties of the drug and, most of all, the pathogen responsible for the infection and its antibiotic resistance (Table 3).

It has been shown that the use of appropriate antibiotic treatment is associated with a lower mortality [11]. Although this appears obvious, the practical consequences of these observations are substantial. The systematic collection of organic material, allowing the detection of the pathological agent, should be

Table 3. Mortality with and without appropriate antibiotics

Mortality with Mortality without appropriate antibiotics appropriate antibiotics

Category of underlying disease n % n %

Four studies combined Rapidly fatal 82/98 84 34/40 85 Ultimately fatal 124/289 42 64196 67 Nonfatal 501506 10 441152 2 Total 256/902 28 1421288 49

From: Pierre Yves Bochem Intensive Care Med (2001) 27 S33-S48

p

NS <0.001 <0.001 <0.001


pursued. Cooperation with the microbiological department in order to obtain more-detailed and accurate information is mandatory [10].

Hemodynamic therapy

Sepsis is the most-frequent cause of vasodilatory shock. Vasoactive agents, secretions, and the renin-angiotensin system are fully activated during vasodilatory shock. Despite the presence of these mediators responsible for the contraction of smooth muscles, blood pressure is low. The possible causes of this effect are summarized in Fig. 1 [12].

The alteration of tissue perfusion due to blind vasodilation is the rationale for the use of drugs able to increase vascular tone and, eventually, to increase tissue perfusion. Norepinephrine, because of its vasoconstrictive effect, is the drug of choice in vasodilatory shock. In the pathophysiology of septic shock, it is important to emphasize that the capacitance of the vascular bed, because of the vasodilation, is markedly increased and the patients require volume expansion.

Colloid or crystalloid infusion during septic shock has been shown to improve cardiac output and increase oxygen transport, and sometimes, has been effective in normalizing blood pressure.

Fig. 1. Possible causes of low blood pressure in sepsis

I Sepsis or tissue hypoxia with lactic acidosis I 1£ .J. '>0.

l' Nitric oxide synthase

l' Nitric 0 xide

~ l' cO M P

.J. A T P ,.J. H + l' Lac ta te in vascular sm ooth muscle

OpenKATP

/ w Phosphorylated myosin

V &so dilatation

From: N Engl J Med 345: 588-595

l' Vasopressin secretion

w V asoprcsin Stores

1 Plasm a vasopressin


The events leading to initial vasodilation in septic shock are associated with a direct effect on myocardial muscle. This combination can further decrease tissue perfusion. For this reason the association of vasoconstrictors and inotropes has been proposed.

Vasopressin is first increased during septic shock but, because of the rapid volume expansion and the inability to release more than 20-30% of the stores, it progressively decreases [13].

Several investigations have proposed its use to reduce catecholamine dosage or to support hypotension refractory to the usual dosage of cathecholamine after cardiopulmonary bypass. Its systematic use requires, however, further investigation.

Septic shock is associated with an activation of the hypothalamus-hypophysis axis, mainly due to a neurogenic reflex [14]. Circulating cytokines are able to activate this in a synergistic way.

The alteration of the cortico-adrenal response had been clearly shown in septic patients [15]. In the study by Annane [14] of 189 patients with septic shock, the plasma cortisol was below the normal range in more than 50% of the population studied. The mortality of this group with a low cortisol level was 75%. Hydrocortisone administration (100 mg 3 times per day) was effective in reducing cathecolamine dosage, supporting the hypothesis of an "occult cortico- -adrenal insufficiency".

Inflammatory and endothelial damage therapy

Several multicenter studies using several drugs have tried to show an impact on mortality and morbidity. The clinical manifestations of severe sepsis and septic shock are mainly due to the exaggerated host response to the infectious agent. For this reason, together with the etiological therapy against the pathogen and the supportive therapy to prevent the complete derangement of homeostasis leading to an irreversible condition, several therapies have been proposed that interact with the pathophysiological mechanism responsible for the tissue damage.

Among all the therapies utilized, and not proven effective to date, a new promising proposal involves modulation of the inflammatory response in conjunction with the prevention of microthrombosis (Table 4).

Activated protein C (drotrecogin alpha) has been effective in reducing the mortality of patients with severe sepsis [16].

Several reasons have been proposed to explain this type of effect: septic patients have a protein C deficit


sepsis is associated with a decrease of activated factor V in sepsis there is a reduction of plasminogen activator inhibition factor it has direct anti-inflammatory effect

Some indirect data support the proposed mechanisms. During drotrecogin alpha infusion, plasma interleukin-6 and D-dimer were lower in the treated than in the control group. When the drug was stopped the two markers showed an overshoot.

Table 4. Mortalities in studies of different interventions for sepsis (PAF = platelet-activating factor, PO = prostaglandin, IL-l-ra = interleukin-l-receptor antagonist, TNF = tumor necrosis factor; mAb = monoclo-nal antibody, SR = soluble receptor, NO = nitric oxide, APe = activated protein C)

Death rates (%) Intervention Number of Number of Control Experimental P value

studies patients

High-dose steroid ~9 1,300 35 39 <0.05 Anti-bradykinin 2 755 39 39 Anti-PAF 2 870 50 45 Anti-PO (ibuprofen) 3 508 40 38 IL-l-ra 3 1,898 35 31 Anti-TNF mAb 8 4,139 36 35 TNF soluble receptor p75-SR 1 141 30 45 <0.05 p55-SR phase IIIIII 1 444 39 34 p55-SR phase III 1 1,340 28 27 NO synthase inhibitor 2 1,059 50 56 APC 1 1,690 30.8 24.7 <0.05

Supportive therapy

The modalities of organ support during sepsis are beyond the scope of this chapter and will be discussed in other sections.

Conclusions

Sepsis represents a frequent pathological event in ICU patients. It carries a high mortality and morbility. Actual therapy is based on infective source control and removal, organ function support and manipulation of inflammatory and prothrombotic activity. Activated protein C is effective in reducing the mortality of patients with severe sepsis.


References 1. Rangel-Fausto MS, Pittet D, Costigan M, et al (1995) The natural history of the systemic

inflammatory response syndrome (SIRS). JAMA 273:117-123 2. Pittet D, Rangel-Frausto S, Li N, et al (1995) Systemic inflammatory response syndrome, sepsis,

severe sepsis, and septic shock; incidence, morbidities and outcomes in surgical ICU patients. Intensive Care Med 21: 302-309

3. Salvo I, Cian W de, Musicco M, et al (1995) The Sepsis Study Group. The Italian sepsis study: preliminary results on the incidence and evolution of SIRS, sepsis, severe sepsis, and septic shock. Intensive Care Med 21:S244-S249

4. Saez-Llorens X, Vargas S, Guerra F, Coronado L (1995) Application of new sepsis definitions to evaluate outcome of pediatric patients with severe systemic infections. Pediatr Infect Dis J 14:557-561

5. Proulx F, Fayon M, Farrell CA, et al (1996) Epidemiology of sepsis and multiple organ dysfunction syndrome in children. Chest 109: 1033-1037

6. Jones GR, Lowes JA (1966) The systemic inflammatory response syndrome as a predictor of bacteraemia and outcome from sepsis. Q J Med 89:515-522

7. Bossink AWJ, Groeneveld J, Hack CE, Thjjs LG (1998) Prediction of mortality in febrile medical patients. How useful are systemic inflammatory response syndrome and sepsis criteria? Chest 113:1533-1541

8. Bone RC, Blak RA, Cerra FB, et al (1992) American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: definition for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 20: 864-874

9. Sugerman HJ (1971) Physiologic management of septicemic shock in man. Surg Forum 22: 3-5 10. Bochud P-Y (2001) Antibiotics in sepsis. Intensive Care Med 9 [Supp 1]: S36 11. Radetsky M (1994) The timing of antimicrobial therapy and outcome in serious bacterial

infection. Curr Opin Infect Dis 7:341-344 12. Landry DW (2001) The pathogenesis of va so dilatory shock. N Engl J Med 345:588-595 13. Landry DW (1997) Vasopressin pressor hypersensitivity in vasodilatory shock. Crit Care 25:

1279-1282 14. Annane D (2000) A 3 level prognostic classification in septic shock based on cortisol levels and

cortisol response to corticotropin. JAMA 283:1038-1045 15. Johnston CA, Greinsman SE (1984) Endotoxemia induced by antibiotic therapy: a mechanism

for adrenal corticosteroid protection in gram-negative sepsis. Trans Assoc Am Physicians 97:172-181

16. Bernard GR, Vincent J-L, Laterre P-F, et al (2001) Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 344:699-709

Sepsis Trials in Children

G.ZOBEL

Sepsis remains an important cause of morbidity and mortality in children, despite the availability of antibiotic therapy, specialized transport of critically ill children, and advances in paediatric intensive care [1]. Though mortality from paediatric sepsis has declined substantially from the pre-antibiotic era, reported mortality remains high, ranging from 10 to 40% [2-4].

Therapy for paediatric septic shock remains primarily supportive. Awareness that septic shock represents a pathophysiologic host response to infection has prompted investigation of immune mediators and coagulation factors as potential targets for anti-sepsis therapies.

Children with sepsis provide a unique opportunity for investigation. The majority of children with sepsis are previously healthy and thus present a more homogenous population for clinical trials. Meningococcal septic shock is an ideal model for the study of immunotherapy in sepsis because its rapid onset and characteristic skin haemorrhages allow bedside diagnosis [5].

Definitions

American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference Committee (ACCP/SCCM) Sepsis terminology (Adapted to children by Hayden) [6, 7].

Infection

Microbial phenomenon characterized by an inflammatory response to the presence of micro-organisms (bacteria, viruses, parasites) or the invasion of normally sterile host tissue by those organisms.

178 G. Zobel

Bacteria

The presence of viable bacteria in the blood.

Systemic inflammatory response syndrome (SIRS)

The systemic response to a variety of severe clinical insults (infection, trauma, bums). The response is manifested by two or more of the following:

Temperature >38 0 or <36o C Heart rate> 2SD above normal age Respiratory rate> 2SD above normal age Leukocyte count> 12,000 cells/mm3, < 4,000 cells/mm3, or >10% band

forms.

Sepsis (SIRS plus infection)

The systemic response to infection.

Severe sepsis

Sepsis associated with organ dysfunction, hypoperfusion, or hypotension.

Septic shock

Sepsis associated with hypotension despite adequate fluid resuscitation, lactic acidosis, oliguria, acute alteration in mental status.

Refractory septic shock

Septic shock with hypotension that lasts for more than 1 hour and does not respond to i.v. fluids or pharmacological intervention and requires vasopressor support.

Hypotension

A systolic blood pressure reading more than 2SD below the mean for age.

Sepsis Trials in Children 179

Multiple organ dysfunction syndrome

The presence of altered organ function in an acutely ill patient so severe that homeostasis cannot be maintained without intervention.

Prevention

Prevention remains an important strategy in decreasing the mortality from paediatric septic shock. Important components of this strategy are continuing efforts to identify infants and children at risk for infection, improvements in vaccination efficacy, provision of antibiotic prophylaxis to target populations, and education of parents and physicians. Appropriate preventive techniques in the area of hospitalized patients deserve continued attention, including hand washing, isolation from other patients, correct use of antibiotics, nutrition, and limitation in the use of invasive devices.

A potential novel therapy for the prevention of sepsis is the use of granulocyte-stimulating factors in neutropenic neonates or cancer patients undergoing chemotherapy [8,9].

Supportive measures

Recently published protocols stress recognition of early signs and symptoms of paediatric shock, the importance of aggressive fluid resuscitation and early intubation, and continued monitoring and resuscitation during transport [1]. Advances in paediatric intensive care unit (lCU) management include improved monitoring capabilities, advances in support of paediatric respiratory failure and renal insufficiency, and increased emphasis on nutrition.

Endotoxins, tumour necrosis factor-a. (TNF- a. ), interleukin (IL)-l and 2, arachidonic acid metabolites, complement, platelet activating factor, low molecular weight peptides, and nitric oxide have all been demonstrated to have cardiac depressant effects in experimental models of septic shock or on myocardial tissue or myocytes in vitro [10]. It is conceivable that a combination of these mediators or complex interactions between them are responsible for myocardial dysfunction in patients with septic shock.

Septic shock in children is often associated with capillary leak syndrome, and vasodilatation requiring aggressive initial fluid resuscitation to avoid severe hypovolemia. Carcillo et al. reported that rapid fluid resuscitation in excess of 40 ml/kg in the first hour following emergency department presentation was associated with improved survival, decreased occurrence of persistent

180 G. Zobel

hypovolemia, and no increase in the risk of cardiogenic pulmonary oedema or adult respiratory distress syndrome in this group of paediatric patients [11]. However, fluid-refractory shock frequently occurs because sepsis also impairs cardiac and vascular function.

Ceneviva et al. examined haemodynamic variable-directed inotrope, vasopressor, and vasodilator therapy in fifty consecutive children with fluid-refractory septic shock after a minimum of 60 ml/kg volume resuscitation with a pulmonary artery catheter within 6 hours of resuscitation [12]. Patients were categorized according to haemodynamic state and use of inotrope, vasopressor, and/or vasodilator therapy to maintain cardiac index (CI) 3.3 Llminlm2 and systemic vascular resistance index (SVRI) 800 dyneosec/cm5/m2 to reverse shock. After fluid resuscitation, 58% of the children had a low CI and responded to inotropic therapy with or without a vasodilator, 20% had a high CI and a low SVRI and responded to vasopressor therapy, and 22% had both vascular and cardiac dysfunction and responded to combined vasopressor and inotropic therapy. Shock persisted in 36% of children. Four children showed a complete change in haemodynamic state and responded to a switch from inotrope to vasopressor therapy or vice versa. The overall 28-day mortality was 20%. The authors concluded that unlike adults, children with fluid refractory shock are frequently hypodynamic and respond to inotrope and vasodilator therapy. Because haemodynamic states are heterogeneous and change with time, an incorrect cardiovascular therapeutic regimen should be suspected in any child with persistent shock. Limitations of this study include experimental design. As prospective haemodynamic evaluation and effectiveness of different classes of cardiovascular therapy was not allowed by the local ethical committee, the authors used an observational case series design in which existing therapies were directed to abnormal haemodynamic variables.

Barton et al. studied in a prospective, double-blinded, randomized, placebocontrolled, descriptive, interventional study in 12 paediatric patients with nonhyperdynamic septic shock the haemodynamic effects of i.v. milrinone lactate [13]. The authors hypothesized that i. v. milrinone would increase cardiac index by 20% and decrease SVRI by 20% during a 2 hour study period. Patients were randomized into two groups; group A received a loading dose of 50 !Lg/kg i.v. milrinone followed by a continuous infusion of 0.5 !Lg/kg/min while group B received an equal volume of placebo.

Milrinone, when used in addition to catecholamines in paediatric patients with nonhyperdynamic septic shock, significantly increased CI, stroke volume index, right and left ventricular stroke work index, and oxygen delivery while significantly decreasing systemic and pulmonary vascular resistance indices, and mean pulmonary artery pressure. There was no significant change in heart


rate, systemic arterial pressures, or PCWP during milrinone and placebo treatment. No adverse effects such as hypotension, tachycardia, or supraventricular or ventricular arrhythmias were documented during milrinone infusion.

The host inflammatory response

Current data suggest that early in the course of septic shock inflammation is pathologically excessive. The inflammatory cascade results in endothelial injury and disruption causing capillary leak, myocardial depression and microvascular thrombosis, which may culminate in mUltiple organ injury and cardiovascular collapse.

Inflammatory mediators of sepsis

Several inflammatory molecules are elevated in neonates and children with sepsis [14-17]. Prime examples are TNF-a, interleukins IL-l, IL-6, IL-8, and the complement complexes C3b/c and C3-CRP as well as anti-inflammatory molecules including soluble TNF receptors, IL-l receptor antagonist, and IL-I0. Elevated levels of these molecules correlate with the development of MOSF and poor outcome [14, 15]. Metabolites of nitric oxide are also elevated and correlate with the outcome [15]. Investigations showed a decreased fibronection and increased nitric oxide metabolites in children with meningococcal disease [18]. Markers of endothelial injury are present in children with sepsis (intracellular adhesion molecules and E-selectin, thrombomodulin) [19]. Coagulation abnormalities in paediatric sepsis include low concentrations of protein S, antithrombin III, and markedly reduced levels of protein C [20]. Concentrations of plasminogen activator inhibitor-l (PAl-I) are very high in children with meningococcal sepsis and correlate with the disease severity [21].

Immunotherapies for sepsis

Clinical trials to block endotoxin or specific cytokines in patients with sepsis have failed to demonstrate any benefit of this therapeutic approach. Even with nearly 20 years of clinical trial experience, it remains unclear which processes should be suppressed or augmented.

182 G. Zobel

Corticosteroids

The effectiveness of corticosteroids in septic shock is controversial. Most human trials investigating steroid therapy for adult patients with septic shock have failed to demonstrate a beneficial effect from steroids [22, 23]. In children with bacterial meningitis low to moderate dosages of dexamethasone, if administered before antibiotic-induced release of endotoxin, were adequate to decrease cytokine production in the cerebrospinal fluid and to improve subsequent neurologic outcome [24, 25].

Slusher et al. conducted a randomized, double blinded, placebo-controlled trial of dexamethasone or placebo in 72 children with sepsis [26]. Dexamethasone (0.2 mg/kg b/w.) was administered 5-10 min before the first dose of antibiotic and then every 8 hours for 2 days. Primary outcome variables were determined before analysis of data. All patients were well matched with respect to age, weight, sex, percentage of positive cultures and other objective signs of infection. Treatment with dexamethasone was not associated with improved outcome. Survival to discharge was 83% in the dexamethasone group and 89% in the placebo group. In addition, haemodynamic stability at 48 hours after starting therapy did not show any significant difference between both groups. In addition, the proportion of afebrile patients at 48-72 hours was not different between both groups and there was no significant difference in the duration of stay in the hospital and the proportion of normal children at discharge from the hospital or at follow-up. The authors concluded that moderate doses of dexamethasone given before antibiotic therapy did not significantly influence the outcome for paediatric patients with sepsis syndrome or shock. Therefore, steroid treatment cannot be routinely recommended for this patient population.

Anti-endotoxin therapies

Endotoxin, the lipopolysaccharide (LPS) component of the gram-negative bacterial cell wall, is considered to be the most important bacterial factor in the pathogenesis of systemic meningococcal infections. Initial plasma endotoxin levels correlate closely with morbidity and mortality. The toxic moiety of endotoxin is lipid A. Endotoxin is not only a marker for gram-negative infections, but also a potential target for therapy.

This led to investigations of different antibodies directed against the lipid A moiety of endotoxin. A study in children with purpura fulminans showed that antiserum to J5 did not significantly alter the clinical course or mortality [27].


Human JgM monoclonal antibody

Derkx et al. used a human IgM monoclonal antibody (HA-1A) that binds to the lipid A domain of endotoxin [28]. HA-1A has been shown to bind N. meningitidis LPS. Clinical studies reported no severe side effects and antibodies to HA-1A were not detected in any patient.

The aim of this randomized, placebo-controlled trial of HA-1A was to evaluate the efficacy of a single dose ofHA-1A in children with meningococcal septic shock and in patient subgroups defined by N. meningitidis culture and antigen status. The secondary objective was to assess the safety of HA-1A. During a 4 year period 269 infants and children with clinical evidence supporting a presumptive clinical diagnosis of fulminant meningococcal septic shock were either assigned to a single dose of HA-1A (6 mg/kg b.w.i.v.=1.2 mllkg b/w., maximum 100 mg, diluted with 3.5 g of albumin) or to placebo consisting of 3.5 g of human serum albumin.

Of the 267 treated patients, 137 received placebo and 130 received HA-1A. The patients were well matched for demographics, clinical variables, severity scoring, biochemical data and initial endotoxin levels. In this trial no significant benefit of HA-1A could be demonstrated. The 28-day mortality rates in the treatment and placebo groups were 18% and 28%, respectively. However, this 33% reduction in mortality did not reach statistical significance (p = 0.11). Nearly twice as many patients survived with sequelae in the HA-1A treated group (14.6%) than in the placebo-treated group (7.3%). Once again, this finding did not reach statistical significance. The authors speculated, that HA-IA is not a highly active endotoxin-neutralizing agent and that the failure to detect a clinical beneficial effect should not indicate a failure of the principle of endotoxin-antibody therapy but may indicate that HA-1A was not the best anti-endotoxin agent to use in clinical trials.

Bactericidiallpermeability increasing protein

Bactericidiallpermeability increasing protein (BPI) was discovered in the late 1970's as a human defence protein, found in neutrophils, that has the property of killing gram-negative bacteria. A modified bactericidallpermeability- increasing protein (rBPIzI) was introduced with the ability not only to kill gram-negative bacteria, but also to bind to and neutralize endotoxin, and to block endotoxin-mediated events, both in vitro and in vivo.

Levin et al. used a recombinant 21-kDa modified N-terminal fragment of human BPI (rBPIzI). In this study children aged 2 weeks to 18 years of age presenting a clinical picture of severe meningococcal sepsis were randomly

184 G. Zobel

assigned either to rBPIzI (2 mg/kg b/w. over 30 min followed by 2 mg/kg over 24 h) or placebo treatment (0.2 mg/ml human albumin solution) in addition to conventional treatment [29]. Primary outcome variables were mortality, amputation, and change in paediatric overall performance category (POPC) from before illness the sixteen day. Prospectively defined exploratory endpoints included: duration of stay in the paediatric ICU, duration of the hospital stay, time on the ventilator, and the use of blood products.

Of the 1,287 infants and children that were screened, 892 were excluded because they failed to meet the inclusion criteria according to the Glasgow Meningococcal Septicaemia Prognostic Score (GMSPS score) (n=825) or they died before enrolment or met criteria for imminent death [30]. A total of 395 patients were enrolled, 191 were assigned to rBPhl and 204 to placebo treatment. One patient in each group died before receiving any study drug and was excluded from the data analysis. Mortality rate in the control and rBPIzI groups was 9.5% and 7.4%, respectively. The proportion of patients requiring severe amputations was lower in the rBPIzI group (3.2%) compared to the placebo group (7.4%, pC). After 60 days, the distribution of patients who returned to pre-illness POPC, or who had a scale deterioration of one or two or more was significantly better for the treatment group (p = 0.019) than for the placebo group. Duration of respiratory support, ICU and hospital stay showed a tendency toward a shorter duration of mechanical ventilation, a shorter stay in the ICU and the hospital in the rBPIzI group compared to the control group.

Any anti-enotoxin agent is likely to be most beneficial when given very early in the course of the disease. This is especially true in meningococcal sepsis, where the rate of disease progression is so fast. The authors speculated that rBPIzI given simultaneously with antibiotics immediately on diagnosis of the disease would result in a much greater benefit as in this trial where the mean time to receive rBPIzI was 5.9 hours because of the necessity to transfer each patient to a teriary center, to complete screening for the study, and to obtain an informed consent.

Modulation of the clotting cascade

Alterations in the coagulation system have been demonstrated in paediatric septic shock, particularly in cases of purpura fulminans [20]. This has led to consideration of the potential therapeutic benefit of various mediators of coagulation such as protein C, recombinant tissue plasminogen activator (rtPA), and tissue factor pathway inhibitor. Though there have been reports of the successful use of protein C and r-tPA in decreasing morbidity or mortality from


paediatric sepsis, no controlled studies have been performed and these therapies remain unproven [31, 32]. In a recent open-label prospective study White et al. assessed the efficacy of protein C replacement therapy in 36 infants, children, and adult patients with fulminant meningococcemia [32]. The Protein C concentrate was initially administered intravenously as a test dose of 10 IV/kg over 10 minutes, followed by a loading dose of 100 IV/kg and a continuous infusion of 10 IU/kg/h. Thereafter, the dose was adjusted on a daily basis with the aim of maintaining a plasma protein C level of 80-120 IV/ml. In these patients with a predicted mortality and amputation rate of 50% and 30%, respectively, the authors observed a mortality rate of 8%, and an amputation rate of 12%. The authors concluded that the beneficial effect of protein C replacement might reflect both the anticoagulant and anti-inflammatory properties of the protein C pathway. A multinational trial of protein C in meningococcal sepsis is ongoing. In addition, a large multicenter trial of recombinant activated protein C for the adjunctive treatment of paediatric sepsis has been initiated.

Continuous plasma filtration

Most of the clinical manifestations of severe infections are caused by an intense, generalized inflammatory response in the host mediated by a multitude of interrelated cellular and humoural factors. Moderate modulation of this response may hold the key to improved survival.

Plasma exchange has been used as salvage therapy in severe infection, especially meningococcemia. Controlled animal studies of plasma exchange in sepsis have yielded conflicting results.

The aim of a multicenter randomized controlled trial conducted by the Plasmafiltration Sepsis Study Group was to investigate the effect of plasmafiltration on survival, number of organs failing and the humoral inflammatory response in 30 patients (22 adults, 8 children) affected by sepsis syndrome [33]. Patients in the plasmafiltration group received continuous plasmafiltration for 34 hours with a total volume of plasma exchange of 250 ml/kg. Plasma filtrate was replaced with a mixture of fresh frozen plasma, albumin and electrolyte solution.

Eight out of the 14 patients (57 % ) in the plasmafiltration group and 8 out of the 16 controls (50%) survived for 14 days (p=0.73). There was no difference in the mean number of organs failing in the first 7 days (2.57 in the plasmafiltration group vs. 2.94 in controls, p=0.37). Plasmafiltration did not influence mean concentrations of IL-6, granulocyte colony-stimulating factor, thromboxane B2, total white cell count, neutrophil count, or platelet count, but it was

186 G. Zobel

associated with significant reductions of alpha-I-antitrypsin, haptoglobin, Creactive protein, and complement fragment C3 in the first 6 hours (p=O.05). No significant difference in the plasma concentration of any individual inflammatory mediator was found between survivors and nonsurvivors. The sieving coefficients for all inflammatory mediators approached one. Substances with a large volume distribution exist predominantly outside the circulation, where they are unavailable for removal by extracorporeal techniques. Alternatively, there may be a proinfiammatory feedback mechanism for some mediators that increase their production in the face of increased clearance through the plasma filter.

The authors concluded that the predominant immunomodulatory effect of continuous plasmafiltration in sepsis was an attenuation of the acute-phase response without any effect on the cytokines response; whether this translates into a survival benefit remains unknown. Therefore, continuous plasmafiltration should be considered experimental in patients with sepsis.

Extracorporeal membrane oxygenation

Goldman et al. reported on extracorporeal membrane oxygenation (ECMO) in 12 infants and children with intractable cardiorespiratory failure due to meningococcal disease [34]. In this retrospective analysis 7 patients underwent ECMO support due to intractable shock within 36 hours of admission to intensive care and in 5 patients ECMO was indicated for intractable acute respiratory failure later in the disease. Survival rate in these patients was 66%. Two patients developed brain death and in 2 patients ECMO was discontinued after 72 hours of support at the request of the families in view of four limb amputations for severe limb ischaemia. Data from the Extracorporeal Life Support Organization (ELSO) registry in 1997 showed that 76 (11.6%) out of 655 children with acute respiratory failure undergoing ECMO had sepsis as the primary disease [35]. The survival rate in septic children was lower than in children without sepsis (36.8% vs. 51.6%, p=O.02). However, the authors concluded that systemic sepsis does not independently influence survival in paediatric ECMO and that this therapy should not be withheld solely because of sepsis, although neurologic complications may occur more frequently.

Genetic influences

Genetic factors may substantially influence the host response to infection [36]. It is conceivable that in the future patients with septic shock will undergo


rapid genetic screening, perhaps on bedside DNA microarrays, and that the results will be utilized to guide therapeutic use of novel immunomodulators.

Conclusion

Despite advances in antibiotic therapy and paediatric critical care, paediatric sepsis remains an important cause of morbidity and mortality. Increased appreciation and understanding of the pathophysiologic host response to sepsis is the current focus of research. Immunotherapy, currently in its early stages of development requires further well designed and controlled studies.

References 1. Pollard AJ, Britto J, Nadel S et al (1999) Emergency management of meningococcal disease.

Arch Dis Child 80:290-296 2. Jafari HS, McCracken GH (1992) Sepsis and septic shock: a review for clinicians. Pediatr Infect

Dis J 11:739-748 3. Anderson MR, Blumer JL (1997) Advances in the therapy for sepsis in children. Pediatr Clin

North Am 44:179-205 4. Butt W (2001) Septic shock. Pediatr Clin North Am 48:601-625 5. De Kleijn ED, Haze1zet J A, Kornelisse RF et al (1998) Pathophysiology of meningococcal sepsis

in children. Eur J Pediatr 157: 869-880 6. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference

Committee (1992) Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 20:864-874

7. Hayden WR (1994) Sepsis terminology in pediatrics. J Pediatr 124:657-658 8. Schibler KR, Osborne KA, Leung LY et al (1998) A randomized, placebo-controlled trial of

granulocyte colony-stimulating factor administration to newborn infants with neutropenia and clinical signs of early-onset sepsis. Pediatrics 102:6-13

9. Garcia-Cabonero R, Mayordomo n, Tornamira MY et al (2001) Granulocyte colony-stimulating factor in the treatment of high-risk febrile neutropenia: a multicenter randomized trial. J Natl Cancer Inst 93:31-38

10. Parker MM (1998) Pathophysiology of cardiovascular dysfunction in septic shock. New Horizons 6:130-138

11. Carcillo JA, Davis AL, Zaritsky A (1991) Role of early fluid resuscitation in pediatric septic shock. JAMA 266:1242-1245

12. Ceneviva G, Paschall A, Maffei F et al (1998) Hemodynamic support in fluid-refractory pediatric septic shock. Pediatrics 102:e19

13. Barton P, Garcia J, Kouatli A et al (1996) Hemodynamic effects of i.v. milrinone lactate in pediatric patients with septic shock. A prospective, double-blinded, randomized, placebo-controlled, intervention study. Chest 109: 1302-1312

14. Hatherill M, Tibby SM, Turner Ch et al (2000) Procalcitonin and cytokine levels: Relationship to organ failure and mortality in pediatric septic shock. Crit Care Med 28:2591-2594

15. Doughty LA, Kaplan SS, Carcillo JA (1996) Inflammatory cytokine and nitric oxide responses in pediatric sepsis and organ failure. Crit Care Med 24:1137-1143

16. Doughty LA, Kaplan SS, Carcillo J A (1998) Plasma nitrite and nitrate concentrations in pediatric sepsis. Crit Care Med 26:157-162

188 G. Zobel

17. Carcillo JA (1999) Nitric oxide production in neonatal and pediatric sepsis. Crit Care Med 27:1063-1065

18. Riordan FA, Bestwick K, Thomson APJ et al (1999) Plasma fibronectin levels in meningococcal disease. Eur J Pediatr 158:81-82

19. Baines PB, Marzouk 0, Thomson APJ et al (1999) Endothelial cell adhesion molecules in meningococcal disease. Arch Dis Child 80:74-76

20. Brandtzaeg P, Sandset PM, J 00 GB et al (1989) The quantitative association of plasma endotoxin, antithrombin, protein C, extrinsic pathway inhibitor and fibropeptide A in systemic meningococcal disease. Thromb Res 55:459-470

21. Komelisse RF, Hazelzet JA, Savelkoul HFJ et al (1996) The relationship between plasminogen activator inhibitor-l and proinflammatory and counterinflamrnatory mediators in children with meningococcal septic shock. J Infect Dis 173:1148-1156

22. Bone RC, Fisher CJ, ClemmerTP et al (1987) A controlled trial of high-dose methylprednisolone in the treatment of severe sepsis and septic shock. N Engl J Med 317: 653-658

23. Hinshwa L, Peduzzi P, Young E et al (1987) Effect of high-dose glucocorticoid therapy on mortality in patients with clinical signs of systemic sepsis. N Engl J Med 317:659-665

24.0dio CM, Faingezicht I, Paris M et al (1991) The beneficial effects of early dexamethasone administration in infants and children with bacterial meningitis. N Engl J Med 324: 1525-1531

25. Mustafa MM, Ramilo 0, Saez-Lorens X et al (1990) Cerebrospinal fluid prostaglandins, interleukin I-beta, and tumor necrosis factor in bacterial meningitis. Am J Dis Child 144: 883-887

26. Slusher T, Gbadero D, Howard C et al (1996) Randomized, placebo-controlled, double blinded trial of dexamethasone in African children with sepsis. Pediatr Infect Dis J 15:579-583

27. J 5 Study Group (1992) Treatment of severe infectious purpura in children with a human plasma from donors immunized with Escherichia coli J5: a prospective double-blind study. J Infect Dis 165:695-701

28. Derx B, Wittes J, McCloskey R, and the European Pediatric Meningococcal Septic Shock Trial Study Group (1999) Randomized, placebo-controlled trial of HA-1A, a human monoclonal antibody to endotoxin, in children with meningococcal septic shock. Clin Infect Dis 28:770-777

29. Levin M, Quint PA, Goldstein Bet al (2000) Recombinant bactericidal/permeability-increasing protein (rBPI21) as adjunctive treatment for children with severe meningococcal sepsis: a randomized trial. Lancet 356:961-967

30. Thomson APJ, Sills JA, Hart CA (1991) Validation of the Glasgow Meningococcal Septicemia Prognostic Score: a year retrospective survey. Crit Care Med 19:26-30

31. Zenz W, Muntean W, Gallistl S et al (1995) Recombinant tissue plasminogen activator treatment in two infants with fulminant meningococcemia. Pediatrics 96:144-148

32. White B, Livingstoone W, Murphy C et al (2000) An open-label study of the role of adjuvant hemostatic support with protein C replacement therapy in purpura fulminans-associated meningococcemia. Blood 96:3719-3724

33. Reeves JH, Butt WW, Shann F et al (1999) Continuous plasmafiltration in sepsis syndrome. Plasmafiltration in Sepsis Study Group. Crit Care Med 27:2096-2104

34. Goldman AP, Kerr SJ, Butt W et al (1997) Extracorporeal support for intractable cardiorespiratory failure due to meningococcal disease. Lancet 349:466-469

35. Meyer DM, Jesson ME, and the Extracorporeal Life Support Organization (1997) Results of extracorporeal membrane oxygenation in children with sepsis. Ann Thorac Surg 63:756-761

36. Hermans PWM, Hibberd ML, Booy R et al (1999) 4G/5G promotor polymorphism in the plasminogen-activator-inhibitor-l gene and outcome of meningococcal disease. Lancet 354:556-560

Sepsis and Clinical Trials: a New Era in Anti-Sepsis Therapies

J.-L. VINCENT

Over the last decade, we have made great progress in our knowledge and understanding of the pathophysiolgy of sepsis, the mediators involved, and the underlying mechanisms, and yet, until very recently, little advance had been made in the field of sepsis therapeutics. Achievements in basic science and cellular research have not been matched by clinical success, and mortality rates from this disease process have remained virtually unaltered over the past 50 years [1]. Indeed, despite the development of many dozens of new so-called immunomodulatory agents, poor results from clinical trials mean that the treatment of sepsis remains antibiotic therapy, source removal, and organ support. In this chapter we will briefly discuss the reasons behind these 'failed' clinical trials before focussing on the recent and exciting results of activated protein C (APC), a drug aimed at the coagulation system, which has been shown to improve mortality rates in patients with severe sepsis.

Clinical trials: why have they failed?

There are many possible reasons why, until very recently, we had failed to see positive results from clinical trials of sepsis 'immunomodulatory' therapies (Table 1) [2-5]. Importantly, there was no one reason behind the persistently negative results, but rather a combination of factors was involved, and despite

Table 1. Some of the reasons why clinical trials of anti-sepsis therapies may have failed

Inadequate preclinical testing Ineffective agents Inadequate doses Inappropriate timing of intervention Patient populations too heterogeneous Inadequate characterization of the sepsis response Single therapies too simplistic

190 J.-L. Vincent

recent 'successful' studies, these aspects must be remembered in the design and interpretation of future clinical trials in this field.

The experimental agent under investigation was ineffective

In retrospect, there are certainly examples where the agents being tested were simply not able to do what they were supposed to. For example, the anti-endotoxin agents, HA-IA and E5, were supposed to bind to the lipid A portion of endotoxin and hence neutralize endotoxin activity; however, testing in vitro shows that neither of these compounds are able to limit endotoxin activity, or to reduce the release of interleukin (IL)-l or tumor necrosis factor (TNF) [6]. In addition, while all agents tested clinically will, of course, have undergone successful preclinical trials, animal physiology is not always identical to that of the human, and animal models of sepsis are but limited imitators of the true-life clinical situation. The fact that a particular drug works in an animal model does not mean that it will automatically work in the clinical arena. However, in general, the activity of the agents tested clinically are supported by adequate preclinical, experimental data.

The timing of drug administration or therapeutic intervention was inadequate

Patients often arrive in the intensive care unit (leU) already with well-established sepsis, and clinical trials therefore include patients at different stages in the disease process. It is well established that the degree of inflammatory response varies among patients, and over time in the same patient. While some patients may thus have a predominantly pro-inflammatory response at the time of clinical trial inclusion and would likely benefit from an anti-inflammatory therapy, others may be in a state of 'immune paralysis' [7] and the same anti-inflammatory therapy may do more harm than good. Unfortunately, our methods of characterizing the degree of immune response are rather inadequate, and predicting which patient will likely benefit from which type of treatment is fraught with difficulty, especially as different therapies may be most beneficial at different stages of the disease process in the same patient.

In addition, by the time sepsis has become established such that clinical signs and symptoms are apparent, it may already be too late for some therapies to have any effect. Although we know which patients are more likely to develop sepsis, we are, as yet, unable to actually predict its development. Yet it is at this stage, or the very early stages of sepsis, that many therapies are most likely to have a beneficial effect. TNF and IL-l, two of the key mediators of sepsis, are

Sepsis and Clinical Trials: a New Era in Anti-Sepsis Therapies 191

released early in the course of the disease, and there is thus a 'narrow window of opportunity' for effective treatment. Pretreatment of patients with louse-borne relapsing fever with murine anti-TNF Fab reduces the penicillin-induced Jarisch-Herxheimer reaction and the associated increases in plasma IL-6 and IL-8levels [8]. However, in other forms of sepsis, such pretreatment schedules are not so easily applicable. Clark et al. [9] studied the effects of treatment with a chimeric human/mouse monoclonal anti-TNF antibody administered within 12 h of diagnosis of severe sepsis, but, as the authors noted, even at this early stage overwhelming cytokine activation will have occurred. Treatments targeted against more-distal mediators that are released later in the sepsis cascade may be more effective in patients with established sepsis.

The dose used was inadequate

Dosing is to a degree a matter of trial and error. While some idea of likely effective doses can be gained from animal models, the lack of a satisfactory means of monitoring the effects of therapy on the immune response in humans makes it difficult to establish dose-response curves, and extrapolating from the animal model is not always appropriate. Phase I or II clinical trials often include a dose-escalation study, and the dose to be used in the phase III trial is then chosen, usually the dose which had the best benefit/side-effect profile.

The patients studied were too heterogeneous

Sepsis affects patients of all ages, genders, and underlying health status, and is caused by different organisms invading from various sites and sources. In addition, genetic differences may alter the degree to which patients mount their immune responses [10]. Problems in defining sepsis added to these 'innate' population differences, leading to the use of non-specific terms such as the systemic inflammatory response syndrome (SIRS) as entry criteria, further compounding the heterogeneous quality of the study groups [11]. The degree of immune response varies among patients and in the same patient over time [12]. Thus, in studies including patients with varying degrees of response, a single anti-inflammatory therapy, for example, will cause benefit in those patients with a predominantly pro-inflammatory response, but this may be negated by the harm done to patients with a predominantly anti-inflammatory response who may rather have benefited from a pro-inflammatory therapy. This reasoning is supported by the fact that many of the studies of immunomodulatory agents have shown beneficial effects in certain subgroups or in retrospective analyses [13-19].

192 J.-L. Vincent

While various markers of infection, including C-reactive protein [20] and procalcitonin [21], have been suggested to facilitate diagnosis, none are specific for sepsis. IL-6 levels have also been used to select patients for inclusion in clinical trials [22,23]. Importantly, one of these studies [23] involving 2,634 patients reported a 10% decrease in relative mortality risk in the anti-TNF antibody group.

The use of a single therapy may be ineffective

The immune response to infection is extremely complex, and it is perhaps rather simplistic to imagine that a single therapy could be adequate; more likely combinations of several agents, as for the treatment of cancer, will be necessary.

Clinical trials: recent successes

Having witnessed years of frustratingly negative results, the last year has seen perhaps the most-exciting breakthrough in the history of sepsis trials, with the publication of the results from a randomized, controlled clinical trial investigating the effects of APC [24]. While the multiple mediators of sepsis have now been relatively well defined, the means by which these lead to severe sepsis with its high mortality rates are still unclear. However, it has become apparent that the mechanisms that regulate inflammation and those that control the coagulation system are closely linked. Indeed, disseminated intravascular coagulation (DIC), as is commonly seen in sepsis, may represent a common final pathway of organ dysfunction [25]. Sepsis is associated with a procoagulant state with thrombin generation and fibrin deposition. The protein C pathway plays a key role in maintaining normal coagulation homeostasis; however, in sepsis, protein C levels fall and, in addition, endothelial damage reduces protein C activation and function [26]. Reduced protein C levels, reported in more than 80% of patients with sepsis [24], have been related to increased mortality from sepsis and septic shock [27-29]. Importantly, in addition to its role within the coagulation system, APC also has anti-inflammatory properties. Several mechanisms have been proposed to account for the ability of APC to modulate inflammation [30,31], including its ability to block nuclear-factor kappa B (NF-KB) nuclear translocation [32,33], a key mechanism in the generation of cytokines by mononuclear cells and the endothelium. APC also appears to modulate anti-apoptosis and cell survival pathways, thus limiting endothelial damage [34].

Early clinical studies of recombinant APC suggested a treatment benefit in


patients with severe sepsis [35], which led to the initiation of a large, multicenter, phase III, randomized, controlled trial - the PROWESS (Protein C Worldwide Evaluation in Severe Sepsis) study. This study [24] was designed to include 2,280 patients, but was terminated prematurely at the second interim analysis at 1,520 patients, because of a significantly reduced mortality in the patients treated with APC. As patients continued to be enrolled during this interim analysis, a total of 1,690 patients with severe sepsis were included, from 164 centers in 11 countries. Patients were randomized to receive an infusion of either APC (drotrecogin alfa) at a dose of 24 J.tglkg per hour or placebo for 96 h. Indeed, at final analysis, the 28-day mortality rates were 30.8% in the placebo group versus 24.7% in the APC group (P = 0.005), giving a 19.4% relative risk reduction, corresponding to an absolute reduction in the risk of death of 6.1 %, i.e., 16 patients need to be treated to save 1 life (Fig. 1). Importantly, this positive effect on mortality was present in all subgroup analyses, regardless of patient age, gender, site of infection, type of infection, severity of disease, or protein

l00r.~~.-~--------------------------------------~

.\ ......

70

'\ ... , '-.

.... ,\ ..... .. , ...... .....

..... ..............

Drotrecogin alfa (activated) (n=850)

........ "' ... Placebo '.,., .. , ..... .

(n=840) , ... , .............. , .................... , ........ , ........... +

p=O.006 (stratified log-rank test) o ~------------------------------------______ ~

o 7 14 28 Days from start of Infusion to death

Fig. 1. Kaplan-Meier estimate of survival among 850 patients with severe sepsis treated with drotrecogin alfa activated (activated protein C) and 840 patients with severe sepsis given placebo. Treatment with drotrecogin alfa activated was associated with a significantly higher rate of survival (P = 0.006 by the stratified log-rank test). Reproduced from [24] with permission

194 J.-L. Vincent

C levels. As expected, there was a greater incidence of severe bleeding in the treated group (P =0.06), but this occurred primarily in patients predisposed to bleeding, such as those with gastrointestinal ulceration, trauma to a major blood vessel or vascular organ, etc. APC was otherwise well tolerated, with no worrying side effects and no increase in the incidence of secondary infections.

These results are indeed exciting and encouraging, but the fight against sepsis is far from over. APC reduces mortality, this seems certain, but other questions remain unanswered. If a patient failed to respond, was it because the dose was inadequate? Should the dose sometimes be increased? If a patient responds, should the dose be prolonged beyond the 96 h used in this trial? In the future, will other treatments be found to be beneficial in addition to APC? The anti-TNF antibody, afelimomab, has recently been shown to reduce mortality [23]; when should a patient receive this treatment in addition to APe?

Conclusion

A new era has begun in the field of immunomodulation in sepsis. Years of frustration have finally been rewarded by positive clinical results with several anti-sepsis drugs. APC is certainly an effective therapy to reduce mortality in patients with sepsis, and should be included as a routine part of treatment regimes for patients with severe sepsis. Other treatments are still needed and will need to be combined with APe. The challenge is to identify those populations that will respond best to which therapy (ies), to determine optimal dosing regimes, and to evaluate the effects and place of combination therapies.

References 1. Friedman G, Silva E, Vincent JL (1998) Has the mortality of septic shock changed with time?

Crit Care Med 26: 2078-2086 2. Bone RC (1996) Why sepsis trials fail. JAMA 276: 565-566 3. Abraham E (1999) Why immunomodulatory therapies have not worked in sepsis. Intensive Care

Med25: 556-566 4. Cohen J (1999) Adjunctive therapy in sepsis: a critical analysis of the clinical trial programme.

Br Med Bull 55: 212-225 5. Dellinger RP (1999) Severe sepsis trials: why have they failed? Minerva Anestesiol65: 340-345 6. Warren HS, Amato SF, Fitting C, Black KM, Loiselle PM, Pastemack MS, Cavaillon JM (1993)

Assessment of ability of murine and human anti-lipid A monoclonal antibodies to bind and neutralize lipopolysacchmide. 1 Exp Med 177: 89-97

7. Volk HD, Reinke P, Krausch D, Zuckermann H, Asadullah K, MUlier 1M, Docke WD, Kox WI (1996) Monocyte deactivation - rationale for a new therapeutic strategy in sepsis. Intensive Care Med 22: S474-S481

8. Fekade D, Knox K, Hussein K, Melka A, Lalloo DG, Coxon RE, Wan'ell DA (1996) Prevention


of Jarisch-Herxheimer reactions by treatment with antibodies against tumor necrosis factor a. N EnglJ Med 335: 311-315

9. Clark MA, Plank LD, Connolly AB, Streat SJ, Hill AA, Gupta R, Monk DN, Shenkin A, Hill GL (1998) Effect of a chimeric antibody to tumor necrosis factor-a on cytokine and physiologic responses in patients with severe sepsis - a randomized, clinical trial. Crit Care Med 26: 1650-1659

10. Appoloni 0, Dupont E, Andrien M, Duchateau J, Vincent JL (2001) Association of TNF2, a TNFa promoter polymorphism, with plasma TNFa levels and mortality in septic shock. Am J Med 110: 486-488

11. Vincent JL (1997) Dear Sirs, I'm sorry to say that I don't like you .. Crit Care Med 25: 372-374 12. Damas P, Carnivet JL, De Groote D, Vrindts Y, Albert A, Franchimont P, Lamy M (1997) Sepsis

and serum cytokine concentrations. Crit Care Med 25: 405-412 13. Greenman RL, Schein RMH, Martin MA, Wenzel RP, MacIntyre NR, Emmanuel G, Chme! H,

Kohler RB, McCarthy M, Plouffe J, Russell JA (1991) A controlled clinical trial of E5 murine monoclonal IgM antibody to endotoxin in the treatment of gram-negative sepsis. JAMA 266: 1097-1102

14. Reinhart K, Wiegand-U:ihnert C, Grimminger F, Kaul M, Withington S, Treacher D, Eckert J, Willatts S, Bouza C, Krausch D, Stockenhuber F, Eiselstein J, Daum L, Kempeni J (1996) Assessment of the safety and efficacy of the monoclonal anti-tumor necrosis factor antibody fragment, MAKI95F, in patients with sepsis and septic shock: a multicenter, randomized, placebo-controlled, dose ranging study. Crit Care Med 24: 733-742

15. Abraham E, Glauser MP, Butler T, Garbino J, Gelmont D, Laterre PF, Kudsk K, Bruining HA, Otto C, Tobin E, Zwingelstein C, Lesslauer W, Leighton A, Ro 45-2081 Study Group. (1997) p55 tumor necrosis factor receptor fusion protein in the treatment of patients with severe sepsis and septic shock. JAMA 277: 1531-1538

16. Fisher CJ, Dhainaut JF, Opal SM, Pribble JP, Balk RA, Slotman GJ, Therti TJ, Rackow EC, Shapiro MJ, Greenman RL, Reines HD, Shelly MP, Thompson BW, LaBrecque JF, Catalano MA, Knaus WA, Sadoff JC (1994) Recombinant human interleukin 1 receptor antagonist in the treatment of patients with sepsis syndrome. JAMA 271: 1836-1843

17. Dhainaut JF, Tenaillon A, Le Tulzo Y, Schlemmer B, Solet JP, Wolff M, Holzapfel L, Zeni F, Dreyfuss D, Mira JP, et al (1994) Platelet-activating factor receptor antagonist BN 52021 in the treatment of severe sepsis: a randomized, double-blind, p1acebo- controlled, multicenter clinical trial. BN 52021 Sepsis Study Group. Crit Care Med 22: 1720-1728

18. Fein AM, Bernard GR, Criner GJ, Fletcher EC, Good JT, Knaus WA, Levy H, Matuschak GM, Shanies HM, Taylor RW Jr, Rodell TC (1997) Treatment of severe systemic inflammatory response syndrome and sepsis with a nove! bradykinin antagonist, deltibant (CP-0127). JAMA 277: 482-487

19. Baudo F, Caimi TM, deCataldo E, Ravizza A, Arlati S, Casella G, Carugo D, Palareti G, Legnani C, Ridolfi L, Rossi R, D'Angelo A, Crippa L, Giudica D, Gallioli G, Wolfler A, Calori G (1998) Antithrombin III (ATIII) replacement therapy in patients with sepsis and/or postsurgical complications: a controlled, double-blind, randomized, multicenter study. Intensive Care Med 24: 336-342

20. Matson A, Soni N, Sheldon J (1991) C-reactive protein as a diagnostic test of sepsis in the critically ill. Anaesth Intensive Care 19: 182-186

21. Ugarte H, Silva E, Mercan D, de Mendon<;a A, Vincent JL (1999) Procalcitonin as a marker of infection in the intensive care unit. Crit Care Med 27: 498-504

22. Reinhart K, RAMSES Study Group (1998) Treatment of severe sepsis in patients with highly elevated IL-6 levels with anti-TNF monoclonal antibody MAK 195F: The RAMSES study (abstract). Crit Care 2: P18

23. Panacek EA, Marshall J, Fischkoff S, Barchuk W, Leah T (2000) Neutralization of TNF by a monoclonal antibody improves survival and reduces organ dysfunction in human sepsis: results of the MONARCS trial (abstract). Chest 118: 88S

196 J.-L. Vincent

24. Bernard GR, Vrncent JL, Laterre PF, LaRosa SP, Dhainaut JF, Lopez-Rodriguez A, Steingrub JS, Garber GE, Helterbrand JD, Ely EW, Fisher CJ, Jr (2001) Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 344: 699-709

25. Marshall JC (2001) Inflammation, coagulopathy, and the pathogenesis of multiple organ dysfunction syndrome. Crit Care Med 29: S99-S106

26. Grinnell BW, Joyce D (2001) Recombinant human activated protein C: a system modulator of vascular function for treatment of severe sepsis. Crit Care Med 29: S53-S61

27. Fourrier F, Chopin C, Goudemand J, Hendrycx S, Caron C, Rime A, Marey A, Lestavel P (1992) Septic shock, mUltiple organ failure, and disseminated intravascular coagulation. Compared patterns of antithrombin III, protein C, and protein S deficiencies. Chest 101: 816-823

28. Fijnvandraat K, Derkx B, Peters M, Bijlmer R, Sturk A, Prins MR, Van Deventer SJ, Cate JW ten (1995) Coagulation activation and tissue necrosis in meningococcal septic shock: severely reduced protein C levels predict a high mortality. Thromb Haemost 73: 15-20

29. Fisher CJ Jr, Yan SB (2000) Protein C levels as a prognostic indicator of outcome in sepsis and related diseases. Crit Care Med 28: S49-S56

30. Grey ST, Tsuchida A, Hau H, Orthner CL, Salem HH, Hancock WW (1994) Selective inhibitory effects of the anticoagulant activated protein C on the responses of human mononuclear phagocytes to LPS, IFN-gamma, or phorbol ester. J Immunol153: 3664-3672

31. Schmidt-Supprian M, Murphy C, While B, Lawler M, Kapurniotu A, Voelter W, Smith 0, Bernhagen J (2000) Activated protein C inhibits tumor necrosis factor and macrophage migration inhibitory factor production in monocytes. Eur Cytokine Netw 11: 407-413

32. Murakami K, Okajima K, Uchiba M, Johno M, Nakagaki T, Okabe H, Takatsuki K (1997) Activated protein C prevents LPS-induced pulmonary vascular injury by inhibiting cytokine production. Am J PhysioI272: Ll97-L202

33. White B, Schmidt M, Murphy C, Livingstone W, O'Toole D, Lawler M, O'Neill L, Kelleher D, Schwarz HP, Smith OP (2000) Activated protein C inhibits lipopolysaccharide-induced nuclear translocation of nuclear factor kappaB (NF-kappaB) and tumour necrosis factor alpha (TNF-alpha) production in the THP-l monocytic cell line. Br J Haematol110: 130-134

34. Joyce DE, Gelbert L, Ciaccia A, DeHoff B, Grinnell BW (2001) Gene expression profile of antithrombotic protein C defines new mechanisms modulating inflammation and apoptosis. J BioI Chern 276: 11199-11203

35. Bernard GR, Hartman DL, Helterbrand JD, et al.(1999) Recombinant human activated protein C produces a trend toward improvement in morbidity and 28 day survival in patients with severe sepsis (abstract). Crit Care Med 27: S4

Evolving Concepts and Challenges in Sepsis and MODS ... a brief history ... from chaos to order ...

A. GULLO

Sepsis is a life-threatening condition in patients admitted to critical care units. Sepsis consumes considerable healthcare resources in acute care hospitals. Considering the rapid evolution of knowledge,few studies on sepsis epidemiology are available [1, 2]; the sepsis-related mortality rate has persisted despite recent advances in understanding the complex pathophysiology of sepsis and significant improvements in the clinician's ability to monitor and provide high technological function supports and management for critically ill patients in today's intensive care unit (ICU) (Tables 1 and 2).

Table 1. Epidemiology of sepsis

Patients enrolled from 991CU: 1101

SIRS5 2%

Sepsis 4.5%

Severe sepsis 2.1%

Septic shock 3%

From Intensive Care Med (1995) 26:254-259

Table 2. Mortality due to sepsis

Patients enrolled from 991CU: 1101

SIRS 26.5%

Infections 24%

Severe sepsis 2.1%

Septic shock 3%

From Intensive Care Med (1995) 26:254-259

198 A. Gullo

Morbidity and mortality have remained high despite improvement in both supportive and anti-microbial therapies; although new and sophisticated diagnostic procedures and several pharmacological treatments with monoclonal antibody, tumor necrosis factor (TNF) receptor antagonist, interleukin-1 (IL-1) ra [3], anti-platelet-activating factor (PAF), ibuprofen [4] are now available, patient outcome is unchanged. According to the data processing by the Centers for Disease Control (CDC) [5], mortality rates vary from 40% for uncomplicated sepsis to 80% in patients suffering septic shock and multi organ dysfunction/failure (MODS); it is estimated that there were 450,000 cases of sepsis per year in the United States, with more than 100,000 deaths.

The CDC warned that the incidence was increasing, citing different variables such as age, functional reserve, and the type of organism and situations able to increase the prevalence of human patients suffering from septic shock and multiorgan dysfunction. In 1990, the CDC considered immunodeficiency virus (HIV) infection as a contributing factor. However, the CDC study counted cases of sepsis, not severe sepsis, which occurs in patients without positive blood cultures [6]. During the last decade studies with the purpose of reaching a consensus on sepsis definitions [7], elucidating the mechanisms, considering prevention strategies available, and efficacy of treatment have been published.

Angus et al. [8] conducted a large, nationally representative study to estimate the incidence, associated costs, and outcome of severe sepsis in the United States; they concluded that severe sepsis is a common, frequently fatal, and expensive condition. It is especially common in the elderly and the mortality rate increases with age; mortality was explained by differences in underlying disease and as a consequence of significant reduction of functional reserve; it is also dependent on the severity and the site of infection. The inflammatory response, involving different mediators, plays a key role; different immunological mechanisms playa pivotal role in determining patients' death.

The average cost per case was U.S. $22.100, with annual total costs of U.S. $16.7 billion nationally. Costs were higher in neonates, non-survivors, in patients admitted to the ICU, in trauma patients, and patients with more organ dysfunction/failure. The incidence was projected to increase by 1.5% per annum. Several mechanisms involved in this process are now based on evidence-based methodology devised primarily for therapeutic trials, including definitions, epidemiology of infection, and experimental therapy [9]. Sprung et al. [10] focused on the importance of human studies. Animal trials have been mentioned when relevant, but it is difficult to compare results with clinical trials. In order to move from the chaos it is essential to understand the process of sepsis; the host inflammatory response in sepsis is a complicated reaction about which much more still needs to be learned. Only a greater understanding

Evolving Concepts and Challenges in Sepsis and MODS 199

of the network of immune, inflammatory, and hematological mediators will allow the development of rational and novel therapies.

What is sepsis?

Defining sepsis is clearly a problem, since several mechanisms are involved in this process. "The presence of various pus-forming and other pathogenic organisms, or their toxins, in the blood or tissues", as given in Stedman s Medical Dictionary [11], seems a true but static definition.

Different definitions of sepsis, using terms such as bacteremia, septicemia, sepsis syndrome, septic shock, and refractory septic shock, were used for patients enrolment into clinical trials, making comparison of results difficult or impossible; nevertheless, the increasing blood level of lactic acid and the high level of oxygen delivery (D02) are considered useful but generic indications for the diagnosis of sepsis.

In 1992 in a special article on the definition of sepsis, Bone et al. [7] pointed out that there is, as yet, no standardized terminology for sepsis. The Canadian group also pointed out that sepsis syndrome, septic shock and refractory septic shock are not separate entities, but simply reflect increasing degrees of illness severity. In the above-mentioned article, the American College of Chest Physicians established a series of definitions of sepsis in an attempt to improve the early detection of sepsis at the bedside and to enable comparison between studies of sepsis by standardizing assessment criteria.

The consensus conference was also motivated by the desire of industry to insure that criteria used in large expensive trials would meet approval of the academic community. The discovery that the catastrophic host responses to overwhelming bacteremia or endotoxemia can be mediated by exaggerated production of TNFa, IL-l, and other proinflammatory cytokines, leading to vascular collapse, shock, and death, has emphasized that inappropriate activation of innate immunity leading to a sort of vascular malignant inflammation can be lethal [12].

What is SIRS?

Infection is considered an invasion of normally sterile host tissue by microorganisms; bacteremia represents a condition with viable bacteria in the blood; systemic inflammatory response syndrome (SIRS) was introduced as a generic and permissive condition characterized by temperature;::: 38°C or 36°C, heart

200 A. Gullo

rate ~ 90 beats/min, respiratory rate ~ 20 breaths/min or paC02 ~ 32torr (~ 4.3KPa), WBC ~ 12,000 celllmm3, or ~ 4,000 cells/mm3 or ~ 10% immature forms. However, the use of this definition remains controversial. The central concept to which the consensus definitions relate is the SIRS. The Canadian group abandoned the term sepsis syndrome in favor of SIRS. The evolving concept that the term infection does not mean sepsis, and that sepsis is one form of SIRS, remains under debate. SIRS also includes non-infective causes and this definition remains very controversial, but was developed to simplify and explain matters.

Changing terminology results from the evolution of patient care, and from attempts to clarify or better describe what is happening. This has occurred in the last decade, during which several new data have increased understanding of the biological events that generate the SIRS; this definition represents a common denominator for sepsis, severe sepsis, septic shock, organ dysfunction, and eventually death in critically ill medical, surgical, and severely traumatized patients [13].

Criteria defining SIRS became controversial when Knaus [14] identified 503 of 519 patients (97%) admitted to ICUs with a primary diagnosis of sepsis-SIRS definition represents an arbitrary and generic definition that was reached by consensus without a process of validation.

Vincent [15] criticized the concept of SIRS for several reasons, particularly because this arbitrary definition appears too sensitive .... and, it doesn't help us to understand the pathophysiology ... .it doesn't help in our clinical trials ..... it doesn't help us in our practice ... this definition doesn't help intensivists in decision making ..... this concept doesn't stimulate physicians to explore the source of infection ..... and so for other several reasons .... we think that we don't like it (as Vincent declared).

Generally speaking, the concept of SIRS may be considered a sort of non-specific bio-humoral permissive condition with alarming signals from the body, with a possible progression towards a continuum represented by a sequence of inflammation, infection, sepsis, and MODS. However, the definition of SIRS remains generic and is not useful for guiding therapy. This concept is sometimes useful, but generally represents a confounding situation.

Sepsis and organ dysfunction

Sepsis produces a powerful, series of reactions in the host organism (fever or hypothermia, leukocytosis or leukopenia, tachypnea or hyperventilation) caused by increased secretion of cytokines and humoregulator peptides.


Unfortunately, none of these situations is specific for the diagnosis of sepsis. The problems of sepsis were summarized by an expert committee of the

European Society of Intensive Care Medicine [16]. The aim of this initiative was to improve the stratification of patients in terms of severity and estimation of the risk of mortality from sepsis; the objective was to identify the patients who are most likely to benefit from new forms of therapy. It is crucial to distinguish infections from non-infectious processes; the underlying clinical status is an important prognostic factor in sepsis; it may be hazardous to lump together different kinds of patients with infection from various sources; the adequate control of the infection with appropriate use of anti-microbial drugs and surgical drainage of the source of infection when possible represents a priority. The Gram-negative endotoxin is a potent mediator of the septic response; is well established that Gram-positive bacteria and their products can have a potent effect on the same network of mediators. Infections due to fungi or viruses are more frequent and in some circumstances they present the same level of danger for patients suffering from bacterial sepsis. These infections are more likely to occur in debilitated or immuno-compromised patients, and the prognosis can be worse; the presence of circulatory shock and the importance of the degree of organ dysfunctiOn/failure represent a key element in the evolution of the septic process. The presence of demonstrable microorganisms in the blood is not a requirement for the diagnosis of sepsis. The presence of bacteremia has been associated with a higher incidence of circulatory shock. About 15-40% of patients suffering from sepsis have documented positive blood culture.

The importance of blood cytokine levels and the timing of cytokine extrusion represents a critical point. In several international sepsis trials reversal of the destructive effects of different mediators was considered a priority. The end point of inadequate tissue perfusion and progressive organ dysfunction is multiple organ failure. Many different criteria have been used to define multiple organ failure, or even failure of a single organ or system, because organ function is a continuous variable from normality, through increasing degrees of organ dysfunction, to overt organ failure, i.e., impairment sufficient to lead to patient death without specific support or artificial replacement of that organ's function.

The many published definitions of multiple organ failure use two types of criteria: intervention based (e.g., mechanical ventilation, dialysis) or physiology based (A-a02 gradient, plasma creatinine concentration) etc. [17]. It is time to focus our attention on the importance of categorizing patients, understanding the basic mechanisms underlying the response to injury, and better defining immunological disturbance in patients at risk. We need to reconsider the algorithm of the sepsis trials; advances in immuno-modulatory therapy represent the ongoing challenges of sepsis for the future.

202 A. Gullo

The use of prognostic index

In order to improve the precision of evaluation of new therapies for the treatment of sepsis, to monitor their utilization, and to refine their indications, it has been recommended that mortality risk stratification or a severity of illness scoring system be utilized in clinical trials and in practice.

There are generic factors to assess the performance of an organ or a system; paO:zlFi02 ratio is a good index for oxygenation; chest X-ray is an important and simple technique useful at the bedside for identifying different grading of severity in the lungs, and sometimes considered a "primum movens"; mechanical ventilation describes a situation where an artificial supportive measure becomes necessary to maintain body homeostasis or improve severe hypoxemia without or with positive end-expiratory pressure. Computed tomography or nuclear magnetic resonance imaging may be useful techniques to evaluate the seriousness of illness in the lung parenchyma; these techniques are necessary to maintain an open lung for improving oxygenation.

Lactacidemia, the ketone body ratio, and osmolal gap are useful to monitor cell dysfunction as a consequence of mismatch perfusion and reduced tissue oxygenation in the microcirculation. On the other hand, creatinine blood levels represent a simple, cheap, and good index of impending renal dysfunction/failure; the proteins of the acute-phase response are useful to control metabolic activity or specifically to establish liver function by the control of enzymes and bilirubin blood levels.

Coagulation maintains a central role in patients suffering from critical illness, particularly during sepsis; platelet count, fibrinogen, prothrombin, and activated partial thromboplastin time are considered the screening tests in patients at high risk for developing consumption coagulopathy. Blood level of ATIII and D-dimer value represent a confirmatory test for the diagnosis of disseminated intravascular coagulation.

Glasgow Coma Scale is a useful and very important index to evaluate the general condition of the cerebral nervous system; although, this score is very difficult to use in severe sepsis and during sedative infusion. Mortality prediction has also been carried out by assessment of endotoxin or IL-l, IL-6, and TNF-a. plasma concentrations.

While increased levels of these substances have been correlated with increased mortality, difficulty with bioassays and their sporadic appearance in the bloodstream prevent these measurements from being practically applied [18]. In the last 20 years several sepsis scoring systems have been developed and applied in clinical practice; the scores more frequently used are APACHE I, II, III, SAPS, MPM, MSOF, and SOFA [19, 20] (Table 3).

Several severity scoring methods [e.g., APACHE, Mortality Prediciton


Model (MPM)] were specifically devised to predict global ICU mortality. Most systems, especially the APACHE systems, have been used for purposes of severity of illness stratification, in the context of clinical research evaluating an intervention. The Multiple System Organ Failure, Organ Dysfunction, and Acute Organ System Failure systems are based solely on the number of organ systems failing, and in the case of the Septic Severity Score and Multiple Organ Failure, based on the score of the severity of the organ failure.

All other systems generate scores that are derived from a variety of physiological, anatomical, and clinical findings. Importantly, only two [21, 22] of the methods have been specifically tested in patients with sepsis, and most were validated in patients with intra-abdominal infection, septic shock, or a plethora of other infectious diagnoses. The SOFA score introduced by Vincent was originally perceived to be useful in sepsis, but is now applied more globally in multiple organ system dysfunction scoring [23].

There is a continuing desire to predict outcome, especially mortality. This desire has perhaps reached its logical extension in the development of APACHE I through ITI scores [24].

Simplified Acute Physiology Scores [25] and other scores include additional clinical information such as age, pre-existing disease, and patient location in the derivation of the score, in an attempt to add precision and accuracy to the mortality prediction.

Organ scores can also be used in clinical trials to compare treatment groups

Table 3. SOFA score I 234 (GCS :: Glasgow Coma Score)

SOFA score 1 2 3 4

Respiration ~OO ~300 ~200 ~IOO

pa02/Fi02 (mmHg)

Coagulation Platelets X 103/mm3

~ISO ~IOO ~O QO

Liver 1.2-1.9 2.0-S.9 6.0-11.9 ~12.0 bilirubin, mg/dl (20-32) (33-101) (102-201) (~204)

(mm01l1)

Cardiovascular MAP F (mmHg) Dopamine ~S or Dopamine ~ S or Dopamine ~ I.S or hypotension" dobutamine epinephrine ~ 0.1 epinephrine ~O.l or

(any dose) norepinhrine ~O.I norepinephrine ~O.I

Central nervous system 13c l4 10-12 6-9 ~6 GCS

Renal 1.2-1.9 2.0-3.4 3.5-4.9 Creatinine, mg/dl (1l0-l70) (171-299) (300-440) or (440) or (mm01l1) or urine output ~SOOmllday ~220mllday

a Adrenergic agents administered for at least I h (doses given in mg/kg x min)

204 A. Gullo

or to evaluate the effects of experimental drugs, devices, and procedures. Recently, four scoring systems have been proposed for clinical use - SOFA, MODS, LOD and Brussels Scores - [19, 20, 26, 27] with the objective to improve standardization in the process of sepsis and organ dysfunction.

Acute-phase responses

Critical care physicians have long recognized that tissue damage can produce a number of different host responses. Some of these responses are adaptive. For example, the demargination of white blood cells in the face of injury contains infection, mediates repair, and leads to recovery. Most responses, however, fall in a nebulous area.

One of the great challenges facing the clinical intensivist and the critical care investigator is differentiating the adaptive from the maladaptive, or defining the point at which a beneficial response becomes detrimental. Part of the adaptive response to injury or inflammation is the elaboration, by the liver, of a series of "acute-phase reactant" proteins, whose production is under transcriptional control and whose function is the modulation of certain aspects of the inflammatory response.

There are proteins that are essential for the restoration of hemostasis, such as fibrinogen, and other proteins such as aI-antitrypsin in humans or ar -macroglobulin in rats, that limit, in rats, the effects of leukocyte-derived proteolytic enzymes. A final group - including heptoglobin, hemopexin, transferrin, and, in rodents, aI-acid glycoprotein - binds cellular and molecular break-down products for transport and disposal. The transcription of these species is controlled, in part, by inflammatory cytokines, including TNF-a, IL-l, and IL-6. These proteins have survival benefit following tissue damage.

Cressman et al. [28] recently found impaired regeneration and increased mortality in genetically engineered IL-6 "knockout" mice subjected to partial hepatectomy. This response could be reversed with a single dose of IL-6 administered before hepatectomy. These data also further demonstrate the dual nature of many cytokine-mediated responses.

High concentrations of IL-6 have been correlated with the development of SIRS/MODS after trauma. An excessive inflammatory response accompanies the initial stages of severe infection and appears to contribute to associated organ system failure and death [29].


Sepsis trials and immunomodulatory therapies

Sepsis trials

The most likely reasons for the failure of trials are reported in this section considering the important suggestions of Bone [30] and Dellinger [31] in 1995 and 2001, respectively. Trials should not start with humans; however failure of clinical trials may be related to the poor predictability of animal models of severe sepsis; no animal model reproduces the complex interactions of human sepsis. An appropriate patient population must be better defined, since defining a homogeneous patient population for the treatment of sepsis is a major problem.

Standardization of trials should be considered carefully (antibiotic policy, fluid resuscitation, vasoactive drugs, nutritional therapy). Entering the right patient in the study based on biological profile represents a preliminary obligation. Full and rapid reporting of all clinical trials of sepsis, regardless of results (withdrawals, dropouts, crossovers, success, and failure), should be mandatory to avoid publication bias and to inform the intensive care community about this rapidly growing literature.

The number of patients enrolled for each center should be of sufficient size and should include a breakdown of the number of patients accrued, mortality rates, and the patient characteristics at each individual center. Pivotal trials should be preceded by sufficient pilot or phase II studies. It is difficult in humans to be able to start treatment immediately after what is judged to be the onset of severe sepsis. Correct drug dosage and appropriate dosage timing should be delineated in pilot studies.

Large, multicenter, double-blind, placebo-controlled, randomized trials of sufficient power should be used to maximize internal validity, minimize random error, avoid type II error, and enhance the applicability of trials results. Analyses should be planned a priori, and should include, as a minimum, an intention-totreat analysis and a power calculation.

Definitions for the target population should be explicit, reproducible, and include illness severity scores. Outcomes should be clinically relevant, reproducible, specified a priori, and should include both measures of benefit and harm. MODS and its reversal should be considered an end point, in addition to mortality. Quality of life should be also considered an end point.

Economic analysis should be included as a part of the clinical design. Formal evaluation of the adequacy of source control should be a critical component of any study to evaluate new strategies for the treatment of sepsis. Standardized clinical mediator assays should be pursued. Monitoring of oxygen metabolism and/or related variables should be considered (blood lactate, ipH). Placebo

206 A. Gullo

patients of clinical trials should be studied for a better understanding of the pathogenesis and epidemiology of the SIRS.

Dellinger [31] concluded that the ideal patient population for a study are patients with intermediate risks of mortality who are likely to survive longer than 24 h independent of effect of therapy. The sepsis trials failed for additional reasons, including co-morbid conditions, and a heterogeneous population due to variability in care of sepsis from physician-to-physician, hospital-to-hospital, and country-to-country. Evidence-based medicine should be used to evaluate the validity of clinical evidence. Bone [30] proposed, in his excellent article, that it is time to regroup and continue our odyssey with sepsis.

Immunomodulatory therapy

It is well known that cytokines playa pivotal role in initiating sepsis and shock. Of particular relevance for the induction of cytokines are three types of bacterial cell wall components:

endotoxin (lipopolysaccharide, LPS), LPS is only present in Gram-negative bacteria lipoteichic acid (LTA) only in Gram-positive bacteria peptidoglycan (PG) in both groups

The best-known cytokine-inducing exotoxins are those derived from Gram-positive bacteria. Bacterial products act as the starting point of sepsis, initiating a pathophysiological chain of events, including cytokine release, that may then continue independently, leading to mUltiorgan failure, shock, and death.

During infection, numerous host and pathogen factors interact to determine whether infection is resolved or progresses to severe sepsis. Cytokines serve as chemical messengers between cells for cell growth and differentiation, tissue repair and remodeling, and regulation of the immune response [32]. Only by fully understanding how these extraordinarily complex proteins work will we be able to develop effective agents that combat their destructive effects while preserving their protective functions.

A number of immunomodulatory therapies aimed at decreasing the dysregulated inflammatory response have been examined in patients with sepsis. Despite the initially encouraging results emerging from several Phase II clinical trials, larger studies have been unable to demonstrate benefit in reducing mortality (Table 4).


Table 4. Result of trials of immunomodulatory therapies (IL-I '" interleukin-l, PAF '" platelet-activating factor, TNF '" tumor necrosis factor, NSAIDs '" non-steroidal anti-inflammatory drugs)

1Ype oftrial Number of trials Total number of patients Mortality Placebo Therapy

Antiendotoxin 4 2,010 35% 35%

Antibody to IL-l receptor 3 1,898 35% 31%

Antibradykinin 3 735 36% 39%

Anti-PAF 2 870 45% 50%

Anti-TNF 8 4,153 41% 40%

Soluble TNF receptor 2 688 35% 40%

NSAIDs 3 514 40% 37%

Steroids 9 1,267 35% 39%

All studies 34 12,135 38% 38%

In contrast, therapies directed against specific pro-inflammatory cytokines, such as TNF-a or IL-l, have produced remarkable clinical response in diseases such as rheumatoid arthritis and Chron's disease. The failure of immunomodulatory therapies to improve outcome in sepsis raises several questions about drug development strategies and the design of clinical trials in this area.

First, there are reasons to believe that some of the agents tested were insufficiently potent to block the mediator of interest or had properties directed at mediators that were not of central importance in determining clinical outcome. It is therefore not surprising that clinical studies with these agents were negative.

Second, most of the immunomodulatory therapies used in the treatment of patients with sepsis showed impressive efficacy in animal models. Their subsequent failure in clinical trials raises concerns about the relevance of pre-clinical experimental models. Specifically, differences between the biochemical and immunological responses of patients with a clinical diagnosis of sepsis and animals with known bacterial infections or endotoxemia may explain the divergence of results between experimental and clinical studies.

Third, the fact that agents such as anti-TNF-a antibodies are clinically effective in the setting of rheumatoid arthritis and Chron's disease raises questions about the heterogeneous nature of patients with sepsis.

The various combinations of treatment, criteria used for patient inclusion, relative efficacy of pharmacological therapy against the mediator or pathophy-

208 A. Gullo

siological process of interest, timing of the administration of the agent, and appropriateness of the therapeutic target all probably contribute to the negative results of clinical trials [33] (Table 4). There is reason to believe that addressing these issues in future studies may demonstrate a beneficial role of immunomodulatory therapies in carefully defined groups of patients with severe infection.

Cohen [34] , recently commented on the failure of clinical trials in sepsis; this condition is related frequently to heterogeneous category of patients, and he proposed that although sepsis is an easily recognizable problem at the bedside, it is not necessarily a single definable, clinico-pathological entity. Sepsis remains a concept that eludes precise definition. Hence, sepsis and the related problems appear an inextricable puzzle for researchers and clinicians.

Ongoing challenges of sepsis

An important challenge is to progress from the clinical syndrome, as presently defined, to more well-defined entities that are delineated by alterations in specific immunological or biochemical pathways. Such mechanistic definitions will provide homogeneous groups of patients who can be identified at early stages of their clinical course. This approach encourages focused investigation of pathways leading to organ system dysfunction and death and, also, provides an efficient framework for the development of new therapies useful in critically ill patients [35].

Chemokines

The events that lead to an inflammatory response are characterized by recognition of the site of injury by inflammatory cells, specific recruitment of sub-populations of leukocytes into tissue, removal of the offending agent and "debridement" of the injured cells/tissue, and repair of the site of injury with attempts to re-establish normal parenchymal, stromal, and extracellular matrix components. The molecular regulation of this complex physiological process involves the interaction between cell surface, extracellular matrix, and soluble mediators, such as chemokines [36].

Chemokine activities are mediated through G-protein-coupled receptors. This is the largest known family of cell-surface receptors, which mediate transmission of stimuli as diverse as hormones, peptides, glycopeptides, and chemokines. One population of cells may respond directly to specific stimuli by the production of a particular cytokine or chemokine to exert distinct effects upon another popUlation of cells. The targets respond by producing chemoki-


nes, which may serve as feedback signals to initiate a cascade of events by activating and recruiting yet another array of target cells. It is clear from in vitro studies that the same signaling pathway can lead to different outcomes, depending on the cell involved [37].

Apoptosis

Apoptosis occurs systematically in many types of cells in patients with sepsis and shock [38]. The lymphoid organs and the columnar cells of the gastrointestinal tract are particularly vulnerable. Investigators have postulated that apopto sis-induced loss of lymphocytes may be responsible for the immune depression that typifies the disorder. Other types of cell, including hepatocytes and vascular endothelial cells, may die by apoptosis in sepsis/endotoxemia. A current theory suggests that apoptosis may contribute to the multiple organ dysfunction of sepsis.

Furthermore, it is difficult to determine whether apoptosis is beneficial or detrimental to the host. If apoptosis is determined to be detrimental to host survival, drugs that inhibit caspase-3 may be useful therapeutically.

Definitions for sepsis, severe sepsis, septic shock, acute lung injury, and acute respiratory distress syndrome were developed by consensus conferences with the goal of achieving standardization of terminology and improved homogeneity of patient populations in clinical studies [39].

Although such definitions have been useful in epidemiological investigations, the criteria specified by the consensus conferences are broad and insufficiently specific to address the problem of heterogeneous mechanisms leading to the clinical syndrome.

Recent news in research and clinical trials

Recently, we have gained some good news. From laboratory studies we have learned about the biology of endotoxin and its interaction with serum proteins such as LPS-binding protein, and the exciting discovery of cell-associated proteins, for example, the Toll-like receptors (TLRs) ... defined the "guardians of the immune system" [40] that recognize the binding of LPS to the cell and transduce the signal across the cell membrane. These developments appear very promising in terms of more knowledge on sepsis mechanisms.

Inside the cell, we have come to understand the role of a myriad of "second messenger", kinases that activate a complicated series of interlocking reactions, resulting in activation of molecule such as NFkB. This activation in tum controls transcription of pro-inflammatory cytokines such as TNF-a., IL-l, and

210 A. Gullo

many other related molecules, which in their turn upregulate enzymes such as nitric oxide synthase, leading to a direct effect on the vasculature.

TLRs are the principle sensors used by the innate immune system in the pathological processes underlying sepsis and septic shock [40]. A new era is dawning in the control of infection and probably the proposal of new definitions. Proposed signaling pathways for TLRs involve the release of LPS from bacterial membranes and recruitment to CD14 by the LPS-binding protein, a plasma protein.

Soluble CD 14 also serves a similar function. Experimental data indicate the utility of anti-CD 14 monoclonal antibody therapy in septic shock and the potential value of targeting intracellular kinases to modulate harmful cellular responses during sepsis. The key element to each of these approaches is to blunt, but not eliminate, the dysregulated inflammatory response that may occur in sepsis [41].

The TLR4IMD-2 complex serves as the LPS signaling receptor. Experimental evidence suggests that the TLR proteins and IL-l receptor (lL-IR) activate similar signaling pathways, as might have been deduced from the known homology between TLRs and IL-IR. Engagement of LPS by TLR4 is thought to induce dimerization of the TLRs, as might be suggested by homophilic binding of the myeloid differentiation factor 88 (MyD88) and recruitment of IL-IR-associated kinase (IRAK). lRAK is subsequently autophosphorylated and can then associate with TNF receptor-associated factor 6 (TRAF6), ultimately resulting in the activation of the transcription factor NF-KB. Note that the great diversity of microbial recognition can theoretically be achieved by the formation of TRL heterodimers (such as TLTX and tLRY).

The principal function of cytokine receptors is to convert an extracellular signal, namely, the specific binding of a cytokine to a target cell, into a intracellular signal, such as activation of kinase or a transcription factor that can trigger a target of cell response. The possibility of blocking TLR signal transduction by means of drugs that directly interact with the TollfIL-l-related domain of different TLRs, or perhaps with the TollfIL-l related domain of the transducer MYD88, also beckons as a possible intervention to mitigate the untoward consequences of global innate immune activation [42].

Endothelium, inflammation and coagulation

There are complex interactions between the endothelium, inflammation, and coagulation system [43]. Endothelial cells are not inert but they exert an active function upon interaction with inflammatory mediators - for example, by promoting fibrin formation; in addition, they may generate inflammatory mediators themselves.


The endothelial system plays a pivotal role in the pathogenesis of the systemic inflammatory response involved in the pathogenesis of sepsis; endothelial dysfunction or injury may be the trigger event for the development of organ failure. The presence of pro-inflammatory cytokines acts on the coagulation cascades via an effect up.)n tissue factor, which is a key player in the coagulation cascade. The relationship with coagulopathy is now recognized as one of the fundamental components of sepsis-related morbidity and mortality [44].

The result of the derangement of the coagulation process during sepsis causes a depletion in protein C, ATIII, and Cl inhibitor, and a decrease in fibrinolysis. The overall effect during sepsis is a marked pro-coagulant balance. This alteration of the coagulation system determines a series of amplification loops. For example, thrombin can induce an upregulation of P and E-selectin, and contact factor activation can induce the production of bradykinin, worsening hypotension and tissue hypoperfusion. In clinical studies, both ATIII and activated protein C levels are sharply decreased, and mortality of septic patients is inversely correlated with the levels of those two products. ATIII was studied in some randomized, small, double-blind studies [45-47]. The duration of disseminated intravascular coagulation was reduced [45], as well as the number of organ failures. The mortality rate was not different; even a meta-analysis [47] performed showing a 22.9% reduction in mortality did not reach statistical significance. A recent large multicenter, prospective, double-blind study showed no significant improvement in survival [48].

Other drugs such as activated protein C and tissue factor inhibitors are not currently available; clinical studies seem promising but multicenter clinical studies are in progress. Protein C is activated by the complex of thrombin and (endothelial cell-bound) thrombomodulin.

Activated protein C and its cofactor, protein S, exert anticoagulant and anti-inflammatory actions. During sepsis, the coagulation cascade is compromised by several factors, such as the progressive downregulation of thrombomodulin and low levels of free protein S due to an increase in C4b-binding protein. The cytokine-mediated pathogenetic pathways of microvascular thrombosis in sepsis are critical.

Recently, Bernard et al. [49] conducted a randomized, double blind, placebo-controlled, multicenter trial in patients with systemic inflammation and organ failure due to acute infection. Drotrecogin alfa (activated), or recombinant human activated protein C, has antithrombotic, anti-inflammatory, and profibrinolytic properties. A total of 1,690 randomized patients were enrolled (840 in the placebo group and 850 in the drotrecogin alfa activated group). The mortality rate was 30.8% in the placebo group and 24.7% in the drotrecogin group. The conclusion of the authors was that drotrecogin alfa (activated)

212 A. Gullo

reduces mortality in patients with severe sepsis and may be associated with an increased risk of bleeding.

Conclusion

Sepsis is a condition with a high risk for patient mortality. The failure of some, many or most clinical trials of so-called or potential "magic bullets" suggested that, if we could document the exact situation of a patient after injury, illness, surgery or with sepsis, we could treat the patient more specifically or exactly by blocking or promoting some of the causes of the problem [50]. There are many complexities concerning sepsis and organ dysfunction and there is also considerable confusion about them [51]. In the last 10 years there has been a consensus about definitions and increasing knowledge of mechanisms and different strategies applied by intensivists for the prevention and management of this critical condition. Nevertheless, continuing advances in the clinical management of sepsis, in terms of survival, are not keeping pace. Can we use this exciting information for an improved care of our patients? Not yet, says Baue [52]. This is a challenge. There are a number of causes of the exsisting gap between what we know and what we can do [53]. Physicians and nurses and allied personnel are obliged to consume higher resources. All clinical trials conducted in the last 10 years were not able to demonstrate the efficacy of new forms of treatment. Revolutionary changes in molecular genetics and more knowledge of cell biology may advance knowledge and treatment in the next decade.

References 1. Salvo I, De Cian W, Musicco M, et al (1995) The Sepsis Study Group; The Italian sepsis study:

preliminary results on the incidence and evolution of SIRS, sepsis, severe sepsis, and septic shock. Intensive Care Med 21:S244-S249

2. Rangel Frausto MS, Pittet D, Costigan M, et al (1995) The natural history of the systemic inflammatory response syndrome (SIRS). JAMA 273: 117 -123

3. Opal SM, Fisher CJ Jr, Dhainaut JF, et al (1997) Conftrmatory interleukin-l receptor antagonist trial in severe sepsis. A phase III, randomised, double-blind, placebo-controlled, multicenter trial. Crit Care Med 25:1115-1124

4. Bernard GR, Wheeler AP, Russell JA, et al (1997) The effects of ibuprofen on the physiology and survival of patients with sepsis. The Ibuprofen in Sepsis Study Group. N Engl J Med 336:912-918

5. Centers for Disease Control (1990) Increase in national hospital discharge survey rates for septicaemia-United States, 1979-1987. JAMA 263:937-938

6. Kieft H, Hoepelman AI, Zhou W, et al (1993) The sepsis syndrome in a Dutch university hospital. Clinical observations. Arch Intern Med 153:2241-2247


7. Bone RC, Balk RA, Cerra FB, et al (1992) Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee, American College of Chest Physicians/Society of Critical Care Medicine. Chest 101:1644--1655

8. Angus DC, Linde-Zwirble, Lidicker J, et al (2001) Epidemiology of severe sepsis in the United States: analysis of incidence, outcome, and associated costs of care. Crit Care Med 29: 1303-1310

9. Sprung CL, Bernard GR, Dellinger RP (2001) Guidelines for the management of severe sepsis and septic shock. Intensive Care Med 27 [Suppl]: SI-S134

10. Sprung CL, Bernard GR, Dellinger RP (2001) Guidelines for the management of severe sepsis and septic shock. Introduction. Intensive Care Med 27 [Suppll]: SI-S2

11. Stedman's Medical Dictionary (1995) 36th edn. Williams and Wilkins, Baltimore 12. Pinsky MR, Vincent JL, Deviere J, et al (1993) Serum cytokine levels in human septic shock.

Relation to multiple-system organ failure and mortality. Chest 103:565-575 13. American College of Chest Physicians/Society of Critical Care Medicine (1997) Consensus

Conference defmitions of the systemic inflammatory response syndrome and allied disorders in relation to critically injured patients. Crit Care Med 25: 1789-1795

14. Knaus WA, Harrell FE, WagnerDP (1994) Defming sepsis. JAMA 271: 1876-1878 15. VincentJL (1997) Dear SIRS, I'm sorry to say that I don't like you ... Crit Care Med 25: 372-374 16. Problem of sepsis. European Society of Intensive Care Medicine (1994) Intensive Care Med

20:300-304 17. Dobb GJ (1991) Multiple organ failure-words mean what I say they mean. Intensive Care World

8:157-160 18. Barriere SL, Lowry SF (1995) An overview of mortality risk prediction in sepsis. Crit Care Med

23:376-393. 19. Vincent JL, Moremuno R, Takala J, et al (1996) The SOFA (Sepsis-related organ failure

assessment) score to describe organ dysfunction/failure. Intensive Care Med 22:707-710 20. Vincent JL, De Mendonca A, Cantraine F, et al (1998) Use of the SOFA score to assess the

incidence of organ dysfunction/failure in intensive care units: results of a multicentric, prospective study. Crit Care Med 26: 1793-1800

21. Baumgartner JD, Bula C, Vaney C, et al (1992) A novel score for predicting the mortality of septic shock patients. Crit Care Med 20:953-960

22. Hebert PC, Drummond AJ, Singer J, et al (1993) A simple multiple system organ failure scoring system predicts mortality of patients who have sepsis syndrome. Chest 19: 137 -144

23. Bernard GR (1998) Quantification of organ dysfunction: seeking standardization. Crit Care Med 26:1767-1768

24. Knaus WA, Wagner DP, Draper EA, et al (1991) The APACHE ill prognostic system. Risk prediction of hospital mortality in critically ill hospitalised adults. Chest 100: 1619-1636

25. Le Gall JR, Loirat P, Alperovitch A, et al (1984) A simplified acute physiology score of ICU patients. Crit Care Med 12:975-977

26. Marshall JC, Cook DJ, Christou NV, et al (1995) Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome. Crit Care Med 23:1638-1652

27. Le Gall JR, Klar J, Lemeshow S, et al for the ICU scoring group (1996). The Logistic Organ Dysfunction System. A new way to assess organ dysfunction in the intensive care unit. JAMA 376:802-810

28. Cressman DE, Greenbaum LE, DeAngelis RA, et al (1996) Liver failure and defective hepatocyte regeneration in interleukin-6-deficient mice. Science 274: 1379-1383

29. Deutschman CS (1998) Acute-phase responses and SIRSIMODS: the good, the bad, and the nebulous. Crit Care Med 26: 1630-1631

30. Bone RC (1995) Sepsis and controlled trials: the odyssey continues. Crit Care Med 25: 1313-1315 31. Dellinger RP (2001) Bad and good news in pathophysiology, prevention, and management of

sepsis. In: Sepsis and organ dysfunction. Baue AE, Berlot G, Gullo A, Vincent JL (eds) Springer, Berlin Heidelberg New York

214 A. Gullo

32. Bone RC (1996) Toward a theory regarding the pathogenesis of the inflammatory response syndrome: what we do and do not know about cytokine regulation. Crit Care Med 24: 163-172

33. Abraham E (1999) Why immunomodulatory therapies have not worked in sepsis. Intensive Care Med 25:556-566

34. Cohen J (1999) The "failure" of clinical trials in sepsis. Curr Opin Crit Care 339-340 35. Marshall JC (ed) (1999) Sepsis, vol. 3. Kluwer 36. Keane MP, Strieter RM (2000) Chemokine signalling in inflammation. Crit Care Med 28 [Suppl]:

N13-N26 37. Yaffe MB, Fink MP (2000) Cellular signalling in critical care. Putting the pieces together. Crit

Care Med 28 [Suppl]: NI-N2 38. Hotchkiss RS, Swanson PE, Freeman BD, et al (1999) Apoptotic cell death in patients with

sepsis, shock, and multiple organ dysfunction. Crit Care Med 27: 1230-1251 39. Abraham E, Matthay MA, Dinarello CA, et al (2000) Consensus conference definitions for

sepsis, septic shock, acute lung injury, and acute respiratory distress syndrome: time for a reevaluation. Crit Care Med 28: 232-235

40. Le-Barillec K, Henneke P, Golenbock DT (2000) Toll receptors: guardians of the immune system and effectors of septic shock. Advances in sepsis. ReMEDICA, pp 23-30

41. Beutler B, Poltorak A (2001) Sepsis and evolution of the innate immune response. Crit Care Med 29: S2-S7

42. Ulevitch RJ (2001) New therapeutic targets revealed through investigations of innate immunity. Crit Care Med 29: S8-S 12

43. Carlet J (2001) Immunological therapy in sepsis: currently available. Intensive Care Med 27:S93-S103

44. Dhainaut JF, Giroir B, Opal S (2001) Introduction to the second Margaux Conference on Critical Illness. Sepsis: interface between inflammation, coagulation, and the endothelium. Crit Care Med 29: SI

45. Fourrier F, Chopin C, Huart JJ, et al (1993) Double-blind, placebo-controlled trial of antithrombin III concentrates in septic shock with disseminated intravascular coagulation. Chest 104: 882-888

46. Baudo F, Caimi TM, Cataldo F de, et al (1998) Antithrombin III (ATIII) replacement therapy in patients with sepsis and/or post-surgical complications: a controlled double-blind, randomized, multicenter study. Intensive Care Med 24:336-342

47. Eisele B, Lamy M, Thijs LG, et al (1998) Anti-thrombin III in patients with severe sepsis: a randomised, placebo-controlled, double blind. Multicenter trial plus a meta-analysis on all randomised placebo-controlled, double-blind trials with antithrombin III in severe sepsis. Intensive Care Med 24:663-672

48. Riess H (2000) Antithrombin in severe sepsis. "New" indication of an "old" drug. Intensive Care Med 26:657-665

49. Bernard GR, Vincent JL, Laterre PF, et al (2001) Efficacy and safety of recombinant human activated protein C for severe sepsis. N Engl J Med 344:699-709

50. Baue AE (1999) Introduction in sepsis and organ dysfunction. In: Baue AE, Berlot G, Gullo A, Vincent JL (eds) Sepsis and organ dysfunction, Springer Verlag Italia, Milan 13-19

51. Baue AE (1998) The complexities of sepsis and organ dysfunction. In: Baue AE, Berlot G, Gullo A, Vincent JL (eds) Sepsis and organ dysfunction, Springer Verlag Italia, Milan 23-31

52. Baue AE (2001) Sepsis and organ dysfunction: an overview of the new science and new biology. In: Baue AE, Berlot G, Gullo A, Vincent Jl (eds) Sepsis and organ dysfunction, Springer Verlag ltalia, Milan 123-132

53. Bion JF, Brun-Buisson C (2000) Introduction - infection and critical illness: genetic and environmental aspects of susceptibility and resistance. Intensive Care Med 26: S I-S2

In memory of Gian Paolo Novelli

February 22, 1932 - August 24, 2001

1iNDExl

A acid sphingomyelinase, 79, 85

activated protein C, 23,32, 112, 118, 161, 167, 173-175,185,189,193,196,211,214

acute lung injury, 38-39, 42, 48, 57, 59, 63, 67, 69-70,72,75,77,79-81, 83, 85, 87-93, 163,209, 214

acute respiratory distress syndrome, 25, 38, 59, 63, 65,75,83,91-93,127,158,163,166,209,214

alveolar epithelium, 37, 39, 41, 43, 45, 47, 49

alveolar macrophages, 37,41,46,48,80-82,87,91

amphotericin B, 127

ampicillin, 134, 138

angiotensin II, 81, 90-91

anoxia, 20

antibiotics, 26-27, 30-31, 63-64,123, 126, 131, 133, 136,139,141-150,160,170-171,175,179,184

apoptosis, 55, 57, 74-75, 77-83, 85-93,196,209

Aspergillus fomigatus, 40-41

B

bacteria, 22-23, 40-42, 46, 60-61, 95-96, 98, 101, 121-124, 126, 128-129, 131, 137, 141, 143-147, 150, 152, 160-161, 177-178, 183, 199, 201,206

bacterial translocation, 20, 32, 123, 125-129, 166

bronchiolitis, 37

Burkholderia cepacia, 40-41, 147

c cadherin, 76

Candida, 64, 124-125, 129, 135, 138

Candida albicans, 64,124, 129, 135, 138

cardiac surgery, 30, 63, 103, 109-110, 117, 144, 156, 166

cardiovascular surgery, 103, 105, 107, 109

caspase inhibitor, 79, 81, 85, 89, 91

caspases, 78, 80, 86, 89

cathepsin G, 60

ceftazidime, 134-135, 138, 143-144, 148-149

ceramide, 79, 85, 89-90, 92

chemokines, 22, 37-42, 45-47, 67, 81, 83, 97, 105, 152,208

children, 38,158,164,175,177,179-188

Clara cells, 80

CO2 production, 103, 109

coagulation, 20, 23, 28-30, 32, 67, 97, 102, 105, 109, 152, 161, 177, 184, 189, 192, 196, 202-203,210-211,214

coagulation cascade, 152, 211

complement 3a, 159, 166

C-reactive protein, 151, 154, 157, 164-165, 186, 192, 195

critical illness, 32, 111, 122, 125, 128,202,214

croup, 37

cytokine production, 37, 39-41, 43-47, 49, 90, 96, 101, 105, 109, 182, 196

cytokines, 21-23, 29, 37-42, 45-48, 54, 59, 62, 67, 77, 79, 81-83, 85, 87, 89-91, 93, 95-98, 101, 105, 109, 129, 152-154, 156, 162-165, 170, 173, 181, 186, 192, 199-200, 204, 206-207, 209,211

cytoskeleton, 40, 44, 46, 49, 52, 57, 76, 122

D defensins, 40, 53, 59-65

dexamethasone, 80, 90, 182, 188

dibenzy1ine, 24

drotrecogin alfa, 193, 211

dysoxia,20

E early diagnosis, 151, 162, 165

elastase, 20-21, 23, 38, 53, 55-56, 59-60, 63, 73-74, 76, 159, 166

endothelial cell, 23, 32, 79, 85, 89, 188

endothelium, 19,23,32,67,192,210,214

endotoxemia, 22,47,55,93,97,99, 101-102, 109, 118,123,126,129,175,199,207

endotoxin, 20-23, 25-26, 28, 41, 50, 83, 88-89, 91, 95-99, 102, 104, 111, 118-119, 121, 124, 126, 129,155-156,159-160,165-167,181-183,188, 190, 195,201-202,206,209

energy expenditure, 103

Enterobacter spp., 134

epidemiology of sepsis, 24, 32, 175, 197

220

epithelial cell, 42, 47-48, 61, 74, 76, 85, 90, 93, 122-123

epithelium, 37, 39, 41, 43, 45, 47, 49, 64, 67-71, 73, 80,105,121-122

Escherichia coli, 60, 64, 98, 100, 102, 118, 121, 132, 138, 188

ethambutol, 27

extraction ratio, 103-104, 109

F Fast-Track management, 107

fentanyl, 30

fungi, 60, 95, 127, 135,201

G glutamine, 28

Golgi complex, 45, 49

Gram-negative bacteria, 60, 96,143-144,160,183

Gram-positive bacteria, 23, 131, 145, 152, 161, 201,206

granulocytes, 40, 64, 82-83, 87, 96, 98, 152, 154, 159

H heparin-coated circuits, 105-107, 110

hepatorenal syndrome, 20

herpes simplex, 60

host defence, 37-39, 41-43, 45-47, 49,51,82

hypotension, 151, 173, 178, 181,203,211

hypoxia, 20, 153

I infection, 19-25,28-31,37-39,41,43,49,55,59,

62, 64, 70, 82, 95, 97-101, 124-128, 132-138, 141, 143, 148, 150-151, 153-159, 162-166, 169-171,175,177-179,182,185-186,192-193, 195,198-201,203-204,206,208,210-211,214

inflammation, 19-25,28-33,38,46,57,59,61,63, 65,67,77, 81-83, 85, 87-93,95-99, 101, 109. 123. 125. 128, 151. 153. 156, 159, 161-162, 164-166. 181, 192, 196, 199-200. 204, 210-211,214

inflammatory cascade, 104-105, 109-110, 181

integrins, 69, 79-80, 89

interferon-y, 80

Index

interleukin, 21, 47, 49, 54, 99, 101-102, 104, 108, 110, 126, 153, 164, 179, 188, 190, 195

interleukin-l-converting enzyme, 78, 85

J Jarisch-Herxheimer reaction. 191

K kallikrein-kinin-system, 152

Klebsiella pneumoniae, 41, 98, 132, 134, 138

L laser Doppler technique, 112-113

Listeria, 122

lung, 20, 37-44,46-49,51,53,55,57,59-65,67-70, 72,75-83,85-93,119,163,166,202,209,214

lung inflammation, 38, 46, 57, 59, 61, 63, 65, 91

lung injury, 38-39, 41-42, 46, 48, 51, 53, 55, 57, 59, 62-63,67-70,72,75,77-81,83,85,87-93,163, 209,214

lymphocytes, 37-38, 80,95-97, 152,209

M macrophages, 37-39, 41, 44, 46-49, 51, 53, 55-57,

80-83,87,91,95-96, 102, 122, 152, 159

markers, 31,103,105,151-152,154,162,165-166, 174,181,192

mechanical ventilation, 62, 85-86, 92, 131, 147, 149, 170, 184,201

mediators, 19,21-25,28-29,31,37-40,59,70,77-78,81-83,85-87,95,97-98,112,152,154,161, 164, 172, 177, 179, 181, 184, 186, 188-192, 198-199,201,207,210

meningococcemia, 185, 188

microcirculation, 19, 111-119, 202

microvascular blood flow, 111-113, 117, 119

milrinone lactate, 180, 187

monocyte chemoattractant proteins, 38

multiple organ dysfunction syndrome, 19, 31-32, 127, 175. 179

multiple organ failure. 19,21-22,31-33,59,92,98, 100, 102, 111, 118, 128-129, 156, 163, 166, 196,201,203

Mycobacterium tuberculosis, 40-41. 57

Index

N nailfold, 112-114, 116

neopterin, 159, 164, 166

neutrophil, 38, 40-42, 48, 53, 55-57, 59-65, 70, 72-76,82-83,91-93, 105, 123, 128, 159, 166, 185

neutrophil a-defensins, 59, 61, 63, 65

neutrophil defensins, 60, 64-65

neutrophil elastase, 38, 63, 73-74, 159

nitric oxide, 21, 33, 81, 97,112,118,161,174,179, 181, 187-188,210

o opsonins, 51, 70

oxygen consumption, 20, 103-104, 109

oxygen delivery, 30, 103-104, 118, 180, 199

p

pancreatitis, 30, 33, 67. 123, 125, 127, 129, 152-153, 158, 164, 167

peritonitis, 26, 89,98-100, 102, 123, 126, 165

permeability, 22, 32, 53, 60, 62, 67, 70, 76, 79, 91, 122

phagocytosis, 44, 51-57, 69, 76, 81, 87,91-92

phospholipases, 160, 167

plasma exchange, 185

platelet activating factor, 79, 179

platelets, 22, 203

Pneumocystis carinii, 40-41

pneumonia, 25, 37, 46-48, 61, 70,80,82,89-90,98-99,101-102,123,127,134,137,144-147,149-150, 154, 165

polymorphonuclear neutrophil granulocytes, 82

polymyxin, 127

post-ischaemic reperfusion, 105

procalcitonin, 151, 154-155, 157, 164-166, 187, 192, 195

protein C, 23, 32,107,112, 118, 161, 167, 173-175, 181,184-185,188-189,192-193,196,211,214

protein C pathway, 185, 192

proteinase 3, 60

Pseudonwnas, 40-41, 43, 49, 125-126, 134, 144, 146-147

221

Pseudomonas aeruginosa, 41, 43, 49,134,146-147

pulmonary embolism, 30, 161

pulmonary fibrosis, 40, 48, 78, 83, 92

R reactive oxygen species, 49, 55, 59, 61

recombinant human activated protein C, 23, 32, 118, 167, 175, 196,214

respiratory system, 37

s sepsis, 19-21,23-25,27,29,31-35,38,47,56,62-

63,67,69-73,75,77,83,95-96,98, 101, 111-112, 117-119, 129, 133, 151-157, 159-167, 169-170, 172-175, 177-214

septic shock, 20, 23-25, 28, 30-31, 63, 75, 78, 85, 95-96, 101, 114, 117-119, 151, 155-157, 162-163, 165, 167, 169-173, 175, 177-184, 186-188,192,194-200,203,209-210,212-214

severe sepsis, 23, 32, 112, 118-119, 151-153, 155-157, 161-163. 165, 167, 169-170, 173-175, 178, 188-189, 191-198, 200, 202, 205-206, 209,212-214

Shigella, 122

soluble TNF receptor, 22, 39, 207

splanchnic hypoperfusion, 123

Staphylococci, 125, 131, 133

Staphylococcus aureus, 133, 137, 142, 147, 149

steroids, 24, 30, 80-81, 92,107, 170, 182,207

Streptococcus pneumoniae, 82, 98

sublingual area, 114, 117

sulbactam, 134

surfactant, 25, 37, 70

systemic inflammatory response syndrome, 19,31-32,103,156,163,166,169,175,178,191,195, 199,212-213

T tazobactam, 148

teicoplanin, 133

thermal injury, 32, 123

thrombocytes, 152

thrombomodulin, 181,211

222

thrombophlebitis, 30

ticarcillin, 148

tobrarnycin, 127, 148

tracheo-bronchitis, 37

tumour necrosis factor, 47, 79, 89, 153, 196

tumour necrosis factor-a, 38, 95, 179

type II pneumocytes, 37, 90

u urinary tract infection, 126

V vancomycin, 131, 133, 137-138, 143

video microscopy, 112

vitamins C, 28

Index

Finito di stampare nel novembre 200 I dalla Stella Arti Grafiche s.r.l. - Trieste


Sepsi e insufflcienza d'organo sono condizioni che mellono In apprensione chi prende in cura il paziente critico. Qualunque sia la natura dell'evento scatenante, esso viene ad a1terare uno stato biologico fruno di una annonica proporzione di elementi in equilibrio e contrastanti. La biochimica e la fisioiogia, la ftsiopalologia e la biolecnologia applicata oflrono alia clinica solo spunti di inlerprelazione degli innumerevoli meccanismi che stan~o alia base della sepsi e dell'insufficienza d'organo; tali eventi singoiannente 0 in associazione SODO tra Ie cause maggiori di mortalita nei pazienti degenti in terapia intensiva. E giunto iI momento di lanciare una idea stimolante; creare un "consensus" pennanente di studiosi ove ciascuno porti il contributo del proprio sapere e sia disporubile agli scambi interdisciplinari. L' Accademia per 10 sludio della sepsi e dell'insufficienza d'organo deve propagarsi in ogni direzione, "ere caraltere itineranle, proporre modelli di studio e di ricerca utili per la prevenzione e il trattarnento di una patologia e di una sindrome c.e con<lizionano I'iter e1inico del malato cotico.


Fur den kritische Palienlen behandelnden Arzl sind Sepsis und Organschwac.e besorgniserregende Zustiinde. UnabbUngig von ihrem Ausliiser sliiren sie das biologische Gleichgewichl, d.s sich aus dem .armonischen Zusarnmenspiel widerspriichlicher E1emente ergibt. Bioc.emie und Physiologie ",wie Pathophysiologie, angewandte Biotecnoologie bieten dem Kliniker nur Inlerprelalionsansalze fiiI die unzithligen Vorgange, die Sepsis- und Organschwac.eersc.einungen zu-grundeliegen; allein oder in Kombination zahlen sie zu den hauftgsten Sterblichkeitsursachen bei Intensi vpatienten. Nun ist hOchste Zeit, eine stimulierende Vorstellung zu wagen: Die Schaffung eines standigen Gremiums, in dessen Ranmen die verschiedenen Forscher ihren Kenntnisbeitrag leisten und sich am inlerdiszipliniiren Erfanrungsaustausch beteiligen konnen, Die Akademie zur Forschung von Sepsis- und Organschwacheerscheinungen iSI als facheriibergreifender, wandemder ZusammenschluB mil dem Ziel zu gestallen, Studien- und Forschungskonzeple vorzuschlagen, die zor Vorbeugung und BebandIung von den klinisc.en Verlauf eines kritischen Pal~nlen beein/loBenden Pal.ologien und Syndromen dienen konnen.

AKA,IlEIIIIIl no II3Y'IEl!lII) BOlIPOCOB CEIICHCA H m;nOCTATo<lHOCTH OprAHOB 'IEllOBt!'!ECKOI'O TEJIA

Ha.nHtiHe YCJlOBHA cenCHca H He~OCTaT01.lHOCTH y OpraH08 C)Oll.:bHOJ'O Ebl3b1BSeT HeMaJIYJO as60TY Y Spaqefl, aaHI1M8!OLlHXC.R JIeqeHH6M T6X naUHeHTOB I KOTopble HSXO,[Vl:TCH B KpHTI11.leCKOfi CT~VI 3s6oJIesatiwi. B0:3HHKHOB6HHe as60neBSHHfi, HeaaBHCHMO OT BH,IJ.S npHtlHH.fIlOl'J.ero ¢aKTopa, H3ueHReT 6HOJIOI'loNeCKOe COCTO.RHHe ,l:I8.HHoro opraHa, onpe,QemleMOe rapMOHmHW COOTHoweHIof6M B31U1l40KOMneHCHpyll!:\HXCH 3JIeMeHTOB. BHOXHMWI H $H3HOJlOrWl, KaIC 11 $113110naTOJlOrHH, npe,ll;CTaBJI.IIIOT KJlHHHKe JlHWb HCXO,ll;Hbl9 :3J19MeHTbl WlR I1HTepnpeTB.UI1I1 C080KYfIHOCTH MexaHHSMOB, BbiSbl.B8.IOlI1HX B03HlIlKHOBeHHe cenCHca H HeAOCTBTOQHOCTH y OpraHOB 1.IeJlOB61.1eCXOro Tens. 3TH wexaHH3Mb1, ,[Iei1lcTBytKD,He OTAeJIbHO j.fJllof B COQeTaHHH APyr C .tIpyrOM, HBJIJOOTCfi O,IlHO~ Ha I'JIRSHbfX npJoNMH CMeprl1TeJIbHOCTJ.1 cpe,r:tH naL\H6HTOB, HaXO,IV'lIilIDCCR no,Q pellHMOM J.1HTeHCHBHOI1 TepanlofH. nepe,lJ. HaMH CTOHT, nOSTOMY, BaIHM 3a.o,aQa, T.e. C03,1l;8HHe nOCTOJ'iHHQrO KOMl1TeTa ytl:eHWC. K~blt1 113 KOTOpW:: ,Il,OJllteH 6YAeT ,neJIaTb CBOI1 BRJIa,o; B ,Il;eJIO lof3yqeHHJi YKasaHHW<: BonpocoB H cnoco6CTBoBaTb Ue&OTpaCJI6BOwy 06MeHY HH$OPM8.ltH.IIMH. AKa,IJ;eMKfI no J.13yq9HHIO BonpOCOB cenCJ.1ca H He,ll;OCTaTOtlHOCTH OpI"aHOB tteJIOBelleCKoro TeJta ,n:OJl)!l;Ha 6y,n:eT IDoIeTb "nepeJl.BJ.1.J.HOt1" xapa:KTep, T. e. pacnpOCTpaHHTbCH BO Bcex HanpaBJleHH.fIX, 11 pa3pa6aTbiSaTb I1ccne,[tOBaTenbCKHe npoeKTbl, Heo6xOAHMble ,lI,JIR np0!IJJ.1nB.XTJ.1HJ.1 naTOJlOI"HJ.1 11 !!eqeHHH 3a6oJIeS6HI1H, 06YCJlOBJIHBa:Kll!lffi( KJIHHloNeCKJ.1il! no~o,ll; K 60JIbHOMY S Kpl1Tl1l.leCHO~ CTa):\l1Yl 3aCloJleBaHID!.

t-1.,..It..,JoI1 ~ - "-'-" "' .... ,,,,. ......

;,t...I1r,:,wt. .;.t.l1t.".,...."'~1,...;I1 ~I

.."J~.r.: lI:l...11,..,..t1

~I J.:.IIt. IP"'" &,,1 "' .... ,,,,. ~ .... I!:I

~~I~ ...... ".JI~t..JI"."'".,..u "Lo...JI.,....,.".;JI~IJ.:.Mt.~I&'I~

......... ...w'"..l"..<l1I1~1 "..,u..t1';';~1 Ut..*".,.~I~. 1tI~1".4-11

'~I""'lVl""I""rke-.l..oa11""U;,,"""""""'.~I.:.!. t,l"L..,IWoj!.l.uWu;."...u.

~I J.:.IIt.IP"'" &'1' .)' <1.0)..1.",' "'~I""""~j~1 ...... J.U~ f"'.Jl .... , .... 1,

. .,...,..tI-=Ir,:""'1",1t..""""'"dt..~.)I, t...<J..,l .... ;"'''''"', t.._"......"" ...w..;.... V-~~I V-I""~ J,...,..u. Jl.i."...,...;...Q , l-.l-v'~ l.,u c"t.J .:.\,11.:.1 .....,

U>1 . ...,.11"'t....-;.;Il..l--'"..,~(l .... Jt... ... ". ......... f'<'-J!:I""oI,"'~1

". Wl.. ~ __ •• 4" J!:'; ~I J,.l.IIt...,..." &'I' ",L..I",."""ts:1 "'~ ....... .,..",

jL...lI.,JaI.J1).. .... jl><" ........ V- r.:"""\,~LJ"I1 .)1 V' ... <l1I1".-loI'",.,.n r.: ..... ~ ',+j~1 <..lIS

.... ".tI"'~t..JJ~~1

ORGAN F AlLURE ACADEMY

La sepllcemie el I'insuffisance d' organes represenlenl des conditions pouvanl a bon droil engendrer WI ,tat d' apprehension de la part du medecin qui doil preter ses soins au patienl crilique. Quel. que "'il I. nature du <!eelenebemenl de Ia maladie, i1 en dieoule bien ividernment une alteration de I'ltal blologique qui esl Ie resultal d'un rapport narmonieux entre des lllments qui peuvent etre, a leur lour, soil en iquilibre soit en conlraste. La bioc.imie, la p.ysiologie, la physiopalhoiogie ella blOlecnnologie appliquee n'offrenl a I'analyse e1inique qne des eiements preliminaires d'inlerpritation des nomhreux mlcanismes qui se lrouvenl a la base de la septicemie el de I'insuffisance d'organes: les differentes patbologies en jeu, prises individuellemenl ou associles les unes aux aulreS, conslituenl les princlpales causes de mortalite chez les patients critiques soumis a traitement intensif. Le moment esl venu de proposer une idee stimulanle, ; savoir la creation d'un "consensus" pernnanenl a ce sujet de la part de tous les experts en la matiere: pour ce faire, illmpone bien entendu que chacun soil dispose a fournir son propre apport en connaissances et qu'il "'il a la fois ouvert a toul type d'/cbanges inlerdisciplinaires. Au"i faUl-il que I'Aca<!emie pout I'Ilude de la septicemie el de I'insuffisance d'organes se developpe dans loules les directions, qu'elle ail un caract,re itineranl el qu'elle ",it en mesure de proposer des modele, d'ilude el de recherche utiles pour I. prevenlion el Ie trailemenl d'une pathologie el d'un syndrome qui influencent en bonne mesure l'evolution clinique du patient critique.


Sepsis e insuficiencia de 6rgano son afecciones que crean cierta inquietud at curar el paciente critico. Cualqwera que sea el genero del factor desencadenador, este termina por alterar un estado biol6gico fruto de una proporcion arm6nica de elementos en equilibrio y contrastantes. La bioquimica y la fisioiogia, la fisiopatologfa y la biotecnologia aplicada ofrecen a la clfnica solamente unas indicaciones para la interpretaci6n de los mecanismos irmumerables que estan a la base de la sepsis y de la insuficiencia de Organo; eslos faclores por separado 0 conjunlamen~ se hallan entte las causas principales de mortaJidad entre los pacienles que reciben terapia intensiva. Ha lIegado el momenlo de lamar una idea inspiradora; de crear un "consensus" pennanente de cienliftcos donde cada uno 11m su propio saber y sea disponible a los inlercambios interdisciplinarios. La Academia para el estudio de la sepsis y de la insuficiencia de organo debe propagarse en mdas las direcciones, lener caracler itinerante, proponer modelos de estudio y de investigacion litiles para la prevenci6n y el tratarniento de una patologia y de una s{ndrome que afecten el curso clinico del enfenno crilico.

J&JlnIiE&US11I'1I.7tJ:1"

J&JIn!l&1:1S11I'IIl1Fl;t, Ii:lIII.f.<W-t,iill! Vnq,<W.: c ? -r, ,*.:""-t,£Q~ lJI!:b<"'.,. ;:t1.;;Q)~-rQ)!l~-t,~iUil;: "t1llll!lcPo; Q)1;t, !t~"ltQ)lnllQ)

i!J!H~\'&1:1~I'lI,:J;?-rfi';;t1.t,,~ltiQ)IIlIt1..:.t7.>{,Q)1!"'7.>. !tit"lt,!t

1I"lt, !tllllllll"ltt"I:'I;t, I!IIiI<l!i:"lt':8P-r, J&JIn!l&1:1511!'lIl1FQ)1lIi;$:lI'Jt .. .(Q)IIIll!lQ)~MQ)*o.4~ ••• ~;:t1.;;Q)J!l(JlnIiE&1:15111'.lIFl;t,.

{, l- ( I;l:fi\'~ l-, illIl!t!'!l!<WQ). < <ll¥Et:IIIll!l c tJ:?"(\' ••

;:Q)afl~&Q)flfi'1!,~, v-c?Q)~~-t,!lU;;"t7.>q:b<*-r\'7.>. ?*I),

.:b<~~~~Q).~cl--r,~t1.':.4~,.~-t,~b"t.Q)W*7.>,v-c?

Q)·~~~~~A·Cl--rQ),"lt~7.> •• fi'''.1!'''7.>. ;:Q)~JIn~&~5~

•• Q)7~T,-I;t,~hB':.&l-,.t"m~~~-t,M~,~~cllll~Q)iiI

.&1:1 ~ Q)!lMlIJ': J; 7.> IIill'i.f.<WQ) lIIIiI<l.'tll': 1lII4"t 7.>~'91aBl-t,lI& l- -r \'

(t><!.l1!"'''.

JIt{.~~~-g-tt~~1:~~

}aff~'1~5~Aat, JIt{"~~-g-tt~IF1:MYt@!I!4: • ~~1t ~lgi!~~~~IFIEJflt~~IN":tM~-tEl3Jf tli-1N.~t§&~*~~iWl:t.~~~~4:~:tK:~- xtT~. ~~~-g-ttt~~1:~1!F~~~~~~ulI, 4:~~. 4:J!~, 4:JfM~;tW.j!i1t-!MI~· i!~IZil*X1B:I:!MkiliW. ~~*£ili~~£~~~~~J!~~A~~4~~zm~1Jt111.@i3t:tlili:Jr~~~, ~J~~ ~ tt~ftzfij-a-tr: ~tt~, @!.1-~~t~£~ ~ c~~t~, tt~~~#~mt· ~.~~.-g-ttt~~1:~~4(l!]*!lktt#1f:&i~iVf:;t;jJ~ *1!jj}ft~HtJf;:;j-~'1~,5~ A~ lai,,*uli1"4:~;E:tt~~ ~M:l*0gj(;~-a-@ .

sepsis and organ dysfunction: ...from chaos to rationale

Documents