Sepsis and Organ Dysfunction Bad and Good News on Prevention and Management

Springer Milano Berlin Heidelberg New York Barcelona Hong Kong London Paris Singapore Tokyo

A.E.Baue G. Berlot A. Gullo J.-1. Vincent (Eds)

Sepsis and Organ Dysfunction Bad and Good News on Prevention and Management

ORGAN FAILURE ACADEMY

Springer

A.E. BAUE, M.D. Department of Surgery, Saint Louis University, Health Sciences Center, St. Louis - USA

G. BERLOf, 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 BATIISTI, 1 - 34125 TRIESTE (ITALY)

Steering Committee A.E. Baue, M.D., Department of Surgery, Saint Louis University Health Sciences Center, St. Louis - 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, Gorizia - Italy

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

ISBN-13: 978-88-470-0137-4 001: 10.1007/978-88-470-2229-4

e-ISBN-13: 978-88-470-2229-4

© Springer-Verlag Italia, Milano 2001

Springer-Verlag Italia A member of BertelsmannSpringer Science+Business Media GmbH

Library of Congress Cataloging-in-Publication Data: Applied for

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, re-use of illustrations, recitation, broadcast­ing, reproduction on microfilms or in other ways, and storage in data banks. Duplication of this publica­tion or parts thereof is only permitted under the provisions of the Italian Copyright Law in its current version and permission for use must always be obtained from Springer-Verlag. Violations are 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 guarantee 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: 10789915

Table of Contents

Epidemiology of Infections in ICUs: Where Are We? M. VIVIANI, G. BERLOT, AND A. GULLO ...................................................................................... II

Prothrombin Fragment 1+2 Levels Are Associated with Pulmonary and Renal Responses to Cardiopulmonary Bypass B. DIXON, J.D. SANTAMARIA, AND D.l CAMPBELL ..................................................................... 23

Oxidative Stress and Apoptosis in Sepsis and the Adult Respiratory Distress Syndrome T. WHITEHEAD, AND H. ZHANG .................................................................................................... 33

Neutrophil Defensins in Lung Inflammation H. ZHANG, AND T. WHITEHEAD.................................................................................................... 39

Nitric Oxide: Lessons Learned and Areas of Success W.M. ZAPOL, AND R. JENNEY ...................................................................................................... 47

Protecting Renal Blood Flow in the Intensive Care Unit J.A. KELLUM ................................................................................................................................ 53

Bad and Good News in Pathophysiology, Prevention, and Management of Sepsis R.P. DELLINGER............................................................................................................................ 63

Light and Shadow: Perspectives on Host-Microbial Interactions in the Pathogenesis of Intensive Care Unit-Acquired Infection J.C. MARSHALL............................................................................................................................ 75

Identification and Characterization of Protein Tyrosine Phosphatases Expressed in Human Neutrophils J. KRUGER, T. FUKUSHIMA, AND G.P. DOWNEy............................................................................ 85

Treatment of Sepsis and Endotoxemia by Extracorporeal Endotoxin Adsorption with Immobilised Human Serum Albumin K. REINHART, AND M. ZIMMERMANN ........................................................................................... 103

Hemofiltration in Intensive Care G. BERLOT, AND M. VIVIANI ........................................................................................................ III

Sepsis and Organ Dysfunction: An Overview of the New Science and New Biology A.E. BAUE .................................................................................................................................... 123

Index ........................................................................................................................................... 133

Authors Index

BaueA.E. Professor of Surgery Emeritus, Vice President for the Medical Center Emeritus, Department of Surgery, Saint Louis University School of Medicine. Fishers Island, Ney York (U,S,A,)

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

Campbell D.J. Intensive Care Centre. St Vincent's Hospital, Melbourne (Australia)

Dellinger R.P. Department of Internal Medicine. Rush University, Chicago, Illinois (U,S,A,)

Dixon B. Intensive Care Centre. St Vincent's Hospital, Melbourne (Australia)

Downey G.P. Clinical Sciences Division. Medical Sciences Building, University of Toronto, Toronto, Ontario (Canada)

Fukushima T. Department of Medicine, Division of Respirology, The University of Toronto, Toronto, Ontario (Canada), and The Hospital for Sick Children Research Institute, Toronto. Ontario (Canada)

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

Jenney R. Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts (U,S.A)

Kellum J.A. Departments of Anaesthesiology/CCM and Medicine, University of Pittsburgh Medical Centre, Pittsburgh, Pennsylvania (U.S.A)

Kruger J. Department of Medicine, Division of Respirology, The University of Toronto, Toronto, Ontario (Canada)

Marshall J.e. Department of Surgery and Programme in Critical Care Medicine, Toronto General Hospital and the University of Toronto, Toronto, Ontario (Canada)

Reinhart K. Clinic for Anaesthesiology and Intensive Care, Friedrich-Schiller University Jena, Jena (Germany)

Santamaria J.D. Intensive Care Centre, St Vincent's Hospital, Melbourne (Australia)

Viviani M. Department of Clinical Sciences, Section of Anaesthesia, Intensive Care and Pain Clinic, Trieste University Medical School, Trieste (Italy)

Whitehead T. Divisions of Respiratory and Critical Care Medicine, Medical Sciences Building, University of Toronto, Toronto, Ontario (Canada)

VIII

ZapoIW.M. Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts (U.S.A.)

ZbangH. Divisions of Respiratory and Critical Care Medicine, Medical Sciences Building, University of Toronto, Toronto, Ontario (Canada)

Zimmermann M. Fresenius HemoCare Adsorber Technology GmbH, St. Wendel (Germany)

Abbreviations

ALT, alanine aminotransferase

ARD, acute respiratory distress syndrome

ARF, acute renal failure

ATN, acute tubular necrosis

CAVH, continuous arteriovenous haemofil­tration

CDC, Centers for Disease Control

cGMP, cyclic guanosine monophosphate

CHF, continuous haemofiltration

CI, cardiac index

CNS, coagulase-negative Staphylococcus

CPB, cardiopulmonary bypass

CVP, central venous pressure

CVVH, continuous venovenous haemofilta­tion

DFP, diisopropylfluorophosphate

DTT, dithiothreitol

ECL, enhanced chemiluminescence

ECMO, extracorporeal membrane oxygena­tion

EDRF, endothelium-derived relaxing factor

GFR, glomerular filtration rate

Hb, haemoglobin

HNP, human neutrophil peptides

HRP, horse radish peroxidase

iHSA, immobilized human serum albumin

LAL, Iimulus amoebocyte lysate

LOCM, low osmolality radiocontrast media

LOS, length of stay

LV, left ventricle

MAP, mean arterial pressure

MBP, mannose-binding protein

MODS, multiple organ dysfunction syndrome

MPAP, mean pulmonary artery pressure

MRSA, methicillin-resistant Staphylococcus

MuLV, murine leukemia virus

NO, nitric oxide

PAIl, plasminogen activator inhibitor I

PAl, plasminogen activator inhibitor

PaOz/Fi02, arterial partial pressure of oxy-gen to inspired fraction of oxygen ratio

PCD, programmed cell death

PCR, polymerase chain reaction

PCWP, pulmonary capillary wedge pressure

PNPP, p-nitrophenyl phosphate

PPHN, persistent pulmonary hypertension of the newborn

PPM, potentially pathogenic microorganisms

PTF 1+2, prothrombin fragment 1+2

PTK, protein tyrosine kinases

PTP, protein tyrosine phosphatases

PVRI, pulmonary vascular resistance index

RBF, renal blood flow

ROS, reactive oxygen species

RT-PCR, reverse transcription polymerase chain reaction

SDD, selective decontamination of the diges­tive tract

SDS-PAGE, sodium dodecyl sulfate-poly­acrylamide gel electrophoresis

sGC, soluble guanylate cyclase

SIRS, systemic inflammatory response syn-drome

SOD, superoxide dismutase

SVRI, systemic vascular resistance index

YAP, ventilator-associated pneumonia

VRE, vancomycin-resistant Enterococcus

WCC, white cell count

XOD, xanthine oxidase

Epidemiology of Infections in ICUs: Where Are We?

M. VIVIANI, G. BERLOT, A. GULLO

Patients admitted to intensive care units (lCUs) represent 8 to 15% of overall hospital population [1]. A frequent problem related to lCU stay is the relative high incidence of infections, rates of which are higher than 40% as reported in some prevalence studies [2, 3]. This percentage is 5-10 times higher when com­pared with the infections in patients admitted in the normal wards; furthermore, in some hospitals of United States, recurrence of infections in lCUs represents more than 20% of overall nosocomial infections [4].

lCU-acquired infections represent a major concern reflecting 80% of total episodes one week after admission in intensive care [2]. Localisation and epi­demiology of infections are well reported in the current literature. Pulmonary, urinary tract and bloodstream infections are mainly observed [5], but etiology is changing in the last years. After the predominant presence of Gram-negative bacteria, generally sensitive to common antimicrobial agents, a progressive in­crease of Gram-positive infections is documented during the 1980s [6] and 1990s. Particularly, resistant strains (coagulase-negative Staphylococcus, Staphylococcus aureus and Enterococci) are often involved in nosocomial and lCU epidemiology. This change in etiology is mainly associated to aggressive antibiotic therapy, type of admission and environmental factors. Hence preven­tion of infection and a tailored policy of antimicrobial administration are recom­mended [7], but, actually, the proposed strategies have obtained conflicting re­sults. The selective decontamination of the digestive tract (SDD) has been re­cently re-evaluated and considered a good strategy for the prevention of respira­tory tract infection [8]. Unfortunately a consensus is not yet established in the clinical practice because some controversies, such as emergence of resistant mi­croorganisms and the impact on mortality, are attributed to SDD application.

Epidemiology of infection in intensive care units Many publications [1-3] report a lot of data about infections and stress the im­portance of epidemiology monitoring as a fundamental step in diagnosis, treat­ment and prevention strategies. This is a relevant matter because lCUs receive 5-10% of hospital patients accounting for more than 20% of overall nosocomial

12 M. Viviani, G. Beriot, A. Gullo

infections [9]. This high incidence has been correlated with the length of stay (LOS) in intensive care [2, 5] which, in tum, depends on other factors such as the severity of illness and the use of invasive procedures. Furthermore, over­crowding and animate reservoirs (colonized subjects), promoting cross trans­mission, represent other major risk factors. A large European multicenter trial, published in 1995, evaluated the prevalence of infections and related risk factors in intensive care [2]. The authors observed that half of the subjects admitted to ICUs were infected (50% of them had acquired the infection in ICU); further­more increased risk of death was associated to LOS, pneumonia and sepsis. A recent similar multicenter study [3] and some incidence studies reported similar results [10] (Table 1).

Table 1. Infections in ICU pati~nts. The percentage of community, hospital and ICU episodes are referred to overall infected subjects enrolled in the studies

Vincent Leon-Rosales Legras et al. [2] et al. [3] et al. [10]

Total infected patients 44.8% 58.2% 42.4% Community - Acq. Infection 32.5% 41.1% 49.5% Hospital - Acq. Infection 21.6% 19.0% 19.6% ICU - Acq. Infection 45.9% 39.9% 30.9%

Low respiratory tract, urinary tract and bloodstream represent the main local­izations of infections in ICU patients. Anyway the incidence varies among countries, hospitals and even different ICUs in the same hospital. This particular distribution is associated to widespread use of mechanical ventilation, bladder catheterisation and intravascular catheters [11]; moreover considerable varia­tions can be attributed to intrinsic factors (i.e. age, LOS, severity of illness, type of admission) and extrinsic factors (i.e. differences in diagnostic methods and prevention measures). Consequently even large prevalence multicenter trials show some limits in the evaluation of this complex problem.

The most relevant point lies on the fact that cross-sectional studies provide only a snapshot at the time of infective episodes, which implies the overestima­tion of long-duration infections and the underestimation of short-duration episodes. Then, longitudinal trials seem to be more useful to assess the inci­dence rates of infectious diseases because the calculation can be corrected on patients' ICU stay or patients' days devices [11, 12].

Interestingly Weber et al. [13], evaluating NNISS data, reported that the fre­quency of different sites of infections and the attributed risk factors change ac­cording to the different type of ICU. The rate of ventilator-associated pneumo­nia, for instance, is higher in surgical, neurosurgical, bum and trauma units than

Epidemiology of Infections in ICDs: Where Are We? 13

in the medical and coronary ICUs, whilst paediatric and bum patients show a greater incidence of central venous lines associated infections. Noteworthy, Wallace observed that trauma patients admitted in intensive care were signifi­cantly more infected than surgical subjects [14]. In this study the increased risk was attributed to many emergency procedures (intubation and vascular cannula­tion), massive transfusion and head injury found in the trauma group.

The characteristic of patients and the type of admission are not the only fac­tors influencing the distribution of infections. In a comparative analysis involv­ing five French ICUs, Legras et al. [10] showed a significant variation of ICU­acquired infection among different units. Moreover, the incidence rates varied in the same ICU when two distinct periods were considered. The authors conclud­ed that such variations could be related to the difficulty of achieving standard­ized definitions of infections (especially pneumonia). The impossibility of ob­taining homogeneous diagnostic methods was a further confounding factor showing the difficulty to achieve good comparative results even when standard­ized protocols were adopted.

In conclusion, although large trials are essential for the epidemiologic as­sessment, the knowledge of local micro-ecology may play an important role for a correct approach to infection monitoring, therapy and prevention program.

Classification of infections: are we adopting the right approach? The classification of infections is very important, particularly for the prevention and management strategies adopted in the intensive care setting. In 1988 the Centers of Disease Control and Prevention (CDC) [15] proposed the well-known time related criteria to distinguishing infections in ICU. Unfortunately, CDC es­tablished no specific "cut off" time to determine the difference among commu­nity, hospital and ICU acquired infections. Arbitrary periods, varying from 48 to 120 hours were employed in epidemiological trials [16, 17]. Adopting this mod­el and establishing the "cut off" at 48 hours, many studies showed that infections were mainly ICU-acquired (Table 1) whilst lower rates of community-acquired episodes were found. Only de Leon-Rosales [3] observed a high rate of commu­nity infections in patients admitted to ICUs. The explanation consisted in the common finding of untreated chronic infections outside the hospitals.

As stated before, the main criticism to the CDC model is that a change in the "cut off" time implies a modification in the attributed classification confounding the ICU acquired infections with the other two categories. In fact, some authors, attempting to overcome this problem, introduced the concept of "early onset" and "late onset" infections [18, 19]. Furthermore, short knowledge about the ICU ecology is obtained adopting the time criterion.

A novel pathophysiological classification, based on patients' carrier state has been proposed to distinguish the origin of infections and to organize a suitable

14 M. Viviani, G. Berlot, A. Gullo

prevention program in the intensive care setting. Using this approach, infections are split into primary endogenous, secondary endogenous and exogenous [20]. The first is caused by potentially pathogenic microorganisms [21] (PPM) car­ried by the patient in throat and rectum on admission to ICUs. The second type is invariably caused by hospital PPM, which are acquired in the oropharynx and gut during the ICU stay but not present on admission. The causative bacteria of exogenous infections are hospital PPM never carried by the patient during the ICU stay.

A recent survey published by Silvestri and colI. [22] interestingly observed that primary endogenous infections accounted for 60% of total diagnosed episodes. These results were in stark contrast with the CDC time criterion (adopting this method 80% of infections were ICU-acquired) and are in accor­dance with data obtained in our experience (unpublished data) (Fig. 1). More­over the authors found that the best time "cut off" between ICU-acquired infec­tions and communitylhospital acquired infections was 9 days.

Hence the carrier-state criterion gives a new insight into the type of infection showing that most episodes are not really ICU-acquired but imported into the intensive care unit. As a consequence change in prevention and therapy strate­gies is necessary. Finally this method, unlike the CDC classification, suggests to the intensivists a novel pathophysiological approach to the infection problem stressing the pivotal role of causative bacteria and ICU ecology.

Resistant bacteria: a major problem in leu Many epidemiological studies show that Gram-negative bacteria are the main etiologic agents isolated in ICU infections [2, 5]. Anyway increasing evidence of Gram-positive infections was observed in the last years [3, 10]. Possibly, this trend could be explained by the selective prevention of Gram-negative infec­tions (antibiotic use), the widespread usage of indwelling devices and the grow­ing importance of some Gram-positive bacteria (coagulase-negative Staphylo­coccus or CNS) as pathogenic agents [6]. In contrast data about yeasts infec­tions are controversial. Even if large epidemiological trials report a relative high rate of Candida spp. infections accounting for 8-15% [2, 5] of cases, a lower in­cidence « 2%) was found in others studies [23,24]. The substantial discrepan­cy between these data could be explained by the high rate of Candida coloniza­tion among critically ill patients [25] (especially in the lower airways and uri­nary tract) and the difficult diagnostic approach to fungal infections [26] (i.e. isolation of yeasts from specimens and the need of pathological findings). Simi­larly to other bacteria, many mycetes isolated in intensive care units are resistant to antimicrobials, especially to azoles [27]. This is an emerging crucial problem because previous prolonged course of azoles can induce resistance among Can­dida species [28].

Epidemiology of Infections in ICUs: Where Are We?

80 0 Silvestri L. (74 infectious episodes)

70

60

50

% 40·

30 i

20

10

o

P. E.

III Viviani M. (206 infectious episodes)

u.i 0 & & x u u cti UJ <: <:

cj :I:

15

& ~

;:J S2

Fig. 1. Classification of infections in intensive care using pathophysiological (P.E. = primary en­dogenous; S.E. = secondary endogenous; Exo. = exogenous) and time (c. Acq. = community-ac­quired; H. Acq. = hospital-acquired; ICU Acq. = ICU-acquired) criteria (see full text). From Sil­vestri [22] and Viviani (unpublished data)

Another important issue consists in the distribution of causative bacteria among different intensive care units. It varies in accordance to the site of infec­tion, type of ICU, type of admission, length of stay and countries. Actually, many authors [5, 12-14] observed that Gram-negatives were mainly involved in low respiratory and urinary tract infections, whilst bloodstream and cardiovas­cular infections were caused principally by Gram-positive bacilli (Staphylococ­cus aureus, CNS, Enterococci). Hence strong antibiotic therapy is often adopted to control such infections promoting the risk of emerging resistant strains.

Current literature reports the crucial problem regarding the resistant micro­organisms as cause of particularly nosocomial and ICU infections [13]. Al­though Gram-negative resistant bacteria such as Pseudomonas and Acinetobac­ter are well described in intensive care [ 13, 29], data obtained from epidemio­logical trials show that infections due to Gram-positive resistant strains are emerging (Fig. 2) in the last years [6, 13, 30, 31]. Methicillin-resistant Staphylo­coccus (MRSA), Enterococcus resistant to vancomycin (VRE) and CNS are iso­lated frequently in critically ill patients especially in the bloodstream, wounds and low respiratory tract. Interestingly the distribution of MRSA infections varies among different countries. A recent European study showed a wide diffu­sion of MRSA in Italy, France and Spain (> 30%), whilst the presence of MRSA in northern Europe is minimal. The basic mechanism related to resistance relies upon some important considerations [30, 32]. First, over time bacteria accumu­late mutations which inhibit the activity of entire classes of antimicrobials; sec­ond, the severity of illness, the large use of indwelling devices, the length of stay are factors favouring bacteria overgrowth especially in the gastrointestinal

16 M. Viviani, G. Berlot, A. Gullo

70

60 rII c:: .~ 50

rII

N 40

.!!! 30 rII e ..... 0 20

?f- -'-MRSA(2)

lO -+- Enterococcus

Fig. 2. Proportion of resistant strains over time. (eNS = Schaberg et aI. [6]; MRSA (1) = resistant strains isolated in bloodstream, Livermore [30]; MRSA (2) = Huang et al. [31]; Enterococcus = Weber et al. [13]

tract [33]; third, the excessive antibiotic usage is related to the development of resistance; last but not least, the failing development of new antibiotics in the re­cent years promotes the acquisition of resistance [32, 34].

Prevention of infections As described herein, the infections in intensive care represent a very complex is­sue for the intensivist. The different epidemiological observed data, the impor­tance of classification and the evident spread of resistance are stimulating argu­ments for a tailored policy approach to leu infections. Prevention plays a piv­otal role in this context, but conventional measures [35] including isolation and cohorting of colonized and/or infected patients, hand washing, aseptic proce­dures, early removal of devices and antibiotic restriction have led to conflicting results. Weinstein [1] argued that these manoeuvres are directed against exoge­nous pathogens, whilst leu infections are due principally to endogenous flora. Moreover the wide variation in the adoption of prevention practices may explain the limited benefits of infection control protocols [36]. Some authors [37-39] ar­gue that high standards of hygiene, isolation of carriers and halting transmission of bacteria between patients and healthcare workers may reduce the morbidity associated with multi-resistant pathogens in leus. Anyway, up to now, there are no properly validating surveys showing a relationship between hand washing and reduction of leu infections [40]. The use of gloves and gowns is not stan­dardised in intensive care and it is unhelpful when adopted as single practice

Epidemiology of Infections in ICDs: Where Are We? 17

[41,42]. The environmental problems including inadequate location of intensive care beds, ventilation systems, structures for storage favour the acquisition and transmission of exogenous pathogens [43]. Furthermore, leu overcrowding and shortage of nursing staff hamper the isolation of carriers and infected patients with resistant organisms. Lastly, the implementation of antibiotic policies in in­tensive care settings is very difficult because many factors are involved in the decision-making process [44]. First, the bad clinical conditions of patients and the high probability of resistant pathogen infections contribute to the prescrip­tion of broad-spectrum antimicrobials. Second, microbiologic surveillance, an­tibiotic therapy monitoring and strict cooperation with microbiologists and pharmacologists are necessary to improve antibiotic control policy.

Many investigators have evaluated the relationship between leu infections, intestinal flora and gut physiology [45-48]. Bacterial translocation is well docu­mented in animals but it is debated in humans. Possibly the main reason consists in experimental technical difficulties [49]. Anyway O'Boyle et al. [50] argues that low levels of bacteria translocation associated to microorganism over­growth, immunosuppression and increased intestinal permeability represent im­portant factors promoting infections. This is an intriguing hypothesis knowing that the widespread systemic antimicrobial usage adopted in leu settings pro­mote multi-resistant PPM overgrowth in the gut overwhelming normal intestinal flora [32, 51].

Selective decontamination of digestive tract has shown to reduce low respira­tory tract infections and mortality in critically ill patients [8, 52]. The philoso­phy of this novel approach consists in the eradication of PPM from the gut with minimal impact on normal intestinal flora [53]. The full SDD protocol includes the following: parenteral infusion of cephalosporins (i.e. cefotaxime) on pa­tients' admission for the prevention of primary endogenous infections; topical application of non-absorbable antibiotics for the prevention of secondary en­dogenous infections; high standards of hygiene and surveillance cultures from throat and rectum for the prevention of exogenous infections and for the evalua­tion of resistant strains and efficacy of SDD manoeuvres.

The main criticisms to SDD application are [54]: the threat of emerging re­sistant pathogens, high costs, the promotion of Gram-positive colonization and infection. Interestingly a recent meta-analysis [8], evaluating more than 30 tri­als, didn't report the emergence of resistant microorganisms during SDD treat­ment. Differently from systemic administration of antimicrobials, topical appli­cation of non-absorbable antibiotics eradicates PPM (Pseudomonas, Enterobac­ter etc.) because high concentrations in saliva and faeces are achieved. Some in­vestigators [55, 56] undertook studies evaluating total costs of stay in leu set­tings. They observed that expenditure was lower in SDD treated patients com­pared to control groups. Gram-positives, in particular MRSA, are intrinsically resistant to SDD. Hence SDD application could be questioned because of pro­moting MRSA gut overgrowth [56] and infections. Furthermore other factors, such as parenteral administration of broad-spectrum antibiotics, underlying dis-

18 M. Viviani, G. Berlot, A. Gullo

eases, large use of invasive devices, breaches of hygiene, person-to-person transmission contribute to the spread of MRSA infections in intensive care set­tings [57]. Some investigators evaluated the effects of MRSA carriage eradica­tion from nasal and oropharyngeal cavity on MRSA infection in ICUs [58,59]. Mupirocin nasal ointment application reduced significantly MRSA carriage and infections. Unfortunately rapid emergence of resistant strains to mupirocin [60, 61] discouraged the use of this strategy. Topical bacitracin [59], chlorhexidine, vancomycin [62] and povidone-iodine cream [63] has been evaluated as alterna­tive to mupirocin. In our two-year experience topical application of vancomycin in the oropharynx and through nasogastric tubes abolished MRSA gastrointesti­nal overgrowth reducing secondary endogenous infections caused by MRSA in ICU intubated patients [64].

Actually selective decontamination of the digestive tract has to be consid­ered an evidence-based-medicine intervention; probably other difficulties must be considered for its limited diffusion [65]. Mainly, proper application of SDD strategy requires a strict cooperation between ICU staff and microbiologists and pharmacologists; special efforts are requested for monitoring efficacy of SDD policies, diagnosis of infections and emergence of resistance. Then a highly professional ICU team is necessary considering that pharmaceutical companies are not interested in supporting a much cheaper method for the prevention of infections.

References 1. Weinstein RA (1991) Epidemiology and control of nosocomial infections in adult intensive

care units. AmJ Med 91[SuppI3B]:179S-184S 2. Vincent JL, Bihari DJ, Suter PM et al (1995) The prevalence of nosocomial infections in in­

tensive care units in Europe. JAMA 274:639-644 3. de Leon-Rosales SP, Molinar-Ramos F, Dominguez-Cherit G et al (2000) Prevalence of infec­

tions in intensive care units in Mexico: a multicenter study. Crit Care Med 28:1316-1321 4. Craven DE, Kunches LM, Lichtemberg DA et al (1988) Nosocomial infection and fatality in

medical and surgical intensive care unit patients. Arch Intern Med 148:1161-1168 5. NNlS System Report (1997) National Nosocomial Infections Surveillance (NNIS) report,

data summary from October 1986-April 1997, issued May 1997. Am J Infect Control 25: 477-487

6. Schaberg DR, Culver DH, Gaynes RP (1991) Major trends in the microbial etiology of noso­comial infection. Am J Med 91[SuppI3B]:72S-75S

7. Bergogne-Berezin E (1999) Current guidelines for the treatment and prevention of nosocomi­al infections. Drugs 58:51-67

8. D'Amico R, Pifferi S, Legnetti C et al (1998) Effectiveness of antibiotic prophylaxis in criti­cally ill adult patients: systematic review of randomised controlled trials. Br Med Journal 316:1275-1285

9. Trilla A (1994) Epidemiology of nosocomial infections in adult intensive care units. Int Care Med 20:S1-S4

10. Legras A, Malvy D, Quinioux AI et al (1998) Nosocomial infections: prospective survey of incidence in five French intensive care units. Int Care Med 24:1040-1046

Epidemiology of Infections in ICUs: Where Are We? 19

11. Gaynes RP, Culver DH, Emori TG et al (1991) The National Nosocomial Infections Surveil­lance System: plans for the 1990s and beyond. Am J Med 91[SuppI3B]: l16S-120S

12. Richards MJ, Edwards JR, Culver DH et al (1999) Nosocomial infections in medical inten­sive care units in the United States. Crit Care Med 27:887-892

13. Weber DJ, Raash R, Rutala WA (1999) Nosocomial infections in the ICU: the growing impor­tance of antibiotic resistant pathogens. Chest 115:34S-41S

14. Wallace WC, Cinat M, Gornik WB et al (1999) Nosocomial infections in the surgical inten­sive care unit: a difference between trauma and surgical patients. Am Surg 65:987-990

15. Gamer JS, Jarvis WR, Emori TG et al (1988) CDC definitions for nosocomial infections. Am lInfect Control 16: 128-140

16. Potgieter PD, Hammond JM (1992) Etiology and diagnosis of pneumonia requiring ICU ad­mission. Chest 101:199-203

17. Mosconi P, Langer M, Cigada M et al (1991) Epidemiology and risk factors of pneumonia in critically ill patients. Eur J Epidemiol 7:320-327

18. Antonelli M, Moro ML. D'Errico RR et al (1996) Early and late onset bacteremia have a dif­ferent risk factors in trauma patients. Int Care Med 22:735-741

19. Troullet JL, Chastre J, Vuagnat A et al (1998) Ventilator associated pneumonia caused by po­tentially drug resistant bacteria. Am J Respir Crit Care Med 157:531-539

20. Van Saene HKF, Damjanovic V, Murray AE, de la Cal MA (1996) How to classify infections in intensive care units - the carrier state, a criterion whose time has come? J Hosp Infect 33: 1-12

21. De la Cal MA, Cerda E, Garcia-Hierro P (1998) Classification of micro-organisms according to their pathogenicity. In: van Saene HKF, Silvestri L. de la Cal MA (eds) Infection control in the Intensive Care Unit. Springer Verlag, pp 9-16

22. Silvestri L, Monti Bragadin C, Milanese M et al (1999) Are most ICU infections really noso­comial? A prospective observational cohort study in mechanically ventilated patients. J Hosp Infect 42: 125-133

23. Nolla-Salas J, Sitges-Serra A, Leon-Gil C et al (1997) Candidaemia in non-neutropenic criti­cally ill patients: analysis of prognosis factors and assessment of systemic antifungal therapy. Int Care Med 23:23-30

24. Petri MG, Konig J, Moeke HP et al (1997) Epidemiology of invasive mycosis in ICU pa­tients: a prospective multicenter study in 435 non-neutropenic patients. Int Care Med 23: 317-325

25. Pittet D, Monod M, Suter PM et al (1994) Candida colonization and subsequent infections in critically ill surgical patients. Ann Surg 220:751-758

26. Vincent JL, Anaissie E, Bruining H et al (1998) Epidemiology, diagnosis and treatment of systemic Candida infection in surgical patients under intensive care. Int Care Med 24:206-216

27. Flanagan PG, Barnes RA (1998) Fungal infection in the intensive care unit. J Hosp Infect 38: 163-177

28. Johnson EW, Warnok DW (1995) Azole drug resistance in yeasts. J Antimicrob Chemother 36:751-755

29. Archibald L, Phillips L, Monnet D et al (1997) Antimicrobial resistance in isolates from inpa­tients and outpatients in the United States: increasing importance of the intensive care unit. Clin Infect Dis 24:211-215

30. Livermore DM (2000) Epidemiology of antibiotic resistance. Int Care Med 26:S14-S21 31. Huang AH, Yan JJ, Wu JJ (2000) Rapid dissemination of Staphylococcus aureus with classic

oxacillin resistance phenotype at a new university hospital. J Hosp Infect 44:309-315 32. van Saene, de la Cal MA, Petros AJ et al (1998) Management of antimicrobial resistance in

the intensive care unit. In: van Saene HKF, Silvestri L, de la Cal MA (eds) Infection control in the Intensive Care Unit. Springer Verlag, pp 323-340

33. Brun-Buisson C, Legrand P, Rauss A et al (1989) Intestinal decontamination for control of multiresistant Gram-negative bacilli. Ann Intern Med 110:873-881

20 M. Viviani, G. Berlot, A. Gullo

34. Moellering RC (1998) The specter of glycopeptide resistance: current trends and future con-siderations. Am J Med 104:3S-6S .

35. Goldman 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

36. Moro ML, Jepsen OB (1996) Infection control practices in intensive care units of 14 Euro­pean countries. Int Care Med 22:872-879

37. Shlaes DM, Gerding DN, John IF et al (1997) Society for Healthcare Epidemiology of Amer­ica and Infectious Diseases Society of America Joint Committee on the Prevention of Antimi­crobial Resistance: guidelines for the prevention of antimicrobial resistance in hospitals. Clin Infect Dis 25:584-599

38. Souweine B, Traore 0, Aublet-Cuvelier B et al (2000) Role of infection control measures in limiting morbidity associated with multi-resistant organisms in critically ill patients. J Hosp Infect 45: 107-116

39. Kollef MH (1999) The prevention of ventilator associated pneumonia. N Engl J Med 340: 627-634

40. Rahman M, Chattopadhyay B (2000) Handwashing: unanswered questions and compliance. J Hosp Infect 45:249-250

41. Sax H, Pittet D (2000) Disinfectants that do. Curr Opin Infect Dis 13:395-399 42. Flaherty JP, Weinstein RA (1996) Nosocomial infections caused by antibiotic-resistant organ­

isms in the intensive care unit. Infect Control Hosp EpidemioI17:236-248 43. O'Connel NH, Humphreys H (2000) Intensive care unit design and environmental factors in

the acquisition of infection. J Hosp Infect 45:255-262 44. Jarvis WR. Preventing the emergence of multidrug-resistant microorganisms through antimi­

crobial use controls: the complexity of the problem 45. Van Der Waaij D (1989) The ecology of human intestine and its consequences for overgrowth

by pathogens such as Clostridium difficile. Annu Rev MicrobioI43:69-87 46. Marshall JC, Christou NY, Meakins JL (1993) The gastrointestinal tract. The "undrained ab­

scess" of multiple organ failure. Ann Surg 218:111-119 47. Marshall JC (1999) Gastrointestinal flora and its alterations in critical illness. Curr Opin Crit

Care 51:119-125 48. Sganga G, Gangheri G, Castagneto M (1998) The gut: a central organ in the development of

multiple organ system failure. In: van Saene HKF, Silvestri L, de la Cal MA (eds) Infection control in the Intensive Care Unit. Springer Verlag, pp 269-277

49. Feltis BA, Wells CL (2000) Does microbial translocation playa role in critical illness? Curr Opin Crit Care 6: 117-122

50. O'Boyle CJ, MacFie J, Mitchell CJ et al (1998) Microbiology of bacterial translocation in hu­mans. Gut 42:29-35

51. Edlund C, Hedberg M, Nord CE (2000) Ecological impact of beta-Iactam treatment on nor­mal human intestinal flora. In: Vmcent JL (ed) 2000 Yearbook of Intensive Care and Emer­gency Medicine. Springer Verlag, pp 111-122

52. Nathens AB, Marshall JC (1999) Selective decontamination of the digestive tract in surgical patients: a systematic review of evidence. Arch Surg 134: 170-176

53. 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. Int Care Med 22:699-706

54. Fowler RA, Cheung AM, Marshall JC (1999) Selective decontamination of the digestive tract in the critically ill patients. Int Care Med 25: 1323-1326

55. Rocha RA, Martin MJ, Pita S et al (1992) Prevention of nosocomial infections in critically ill patients by selective decontamination of the digestive tract. A randomised double blind place­bo-controlled study. Int Care Med 18:398-404

56. Garcia MS, Galoche JAC, Diaz JL et al (1998) Effectiveness and cost of selective decontami­nation of the digestive tract in critically ill intubated patients: a randomized, double-blind, placebo-controlled, multicenter trial. Am J Respir Crit Care Med 158:908-916

Epidemiology of Infections in ICUs: Where Are We? 21

57. Coello R, Glynn JR, Gaspar C et al (1997) Risk factors for developing clinical infection with methicillin-resistant Staphylococcus aureus (MRSA) amongst hospital patients initially only colonized with MRSA. J Hosp Infect 37:39-46

58. Dacre J, Emmerson AM, Jenner EA (1986) Gentamicin-methicillin-resistant Staphylococcus aureus: epidemiology and containment of an outbreak. J Hosp Infect 7: 130-136

59. Wenzel RP, Nettleman MD, Jones RN, Pfaller MA (1991) Methicillin-resistant Staphylococ­cus aureus: implications for the 1990s and effective control measures. Am J Med 91 [Suppl 3B]:22IS-227S

60. Cookson BD (1990) Mupirocin resistance in Staphylococci. J Antimicrob Chemother 25: 497-503

61. Miller MA, Dascal A, Portnoy J, Mendelson J (1996) Development of mupirocin resistance among methicillin-resistant Staphylococcus aureus after widespread use of nasal mupirocin ointment. Infect Control Hosp Epidemiol 17:811-813

62. Pugin J, Auckenthaler R, Lew DP, Suter PM (1991) Oropharyngeal decontamination decreas­es incidence of ventilator-associated pneumonia. JAMA 265:2704-2710

63. Hill RLR, Casewell MW (2000) The in-vitro activity of povidone-iodine cream against Staphylococcus aureus and its bioavailability in' nasal secretions. J Hosp Infect 45: 198-205

64. Viviani M, Berlot G, Dezzoni R et al (2000) The impact of topical vancomycin on MRS A gut carriage and infections in ICU patients. Int Care Med 26:S292

65. Silvestri L, Mannucci F, van Saene HKF (2000) Selective decontamination of the digestive tract: a life saver. J Hosp Infect 45: 185-190

Prothrombin Fragment 1+2 Levels Are Associated with Pulmonary and Renal Responses to Cardiopulmonary Bypass

B. DIXON, J.D. SANTAMARIA, D.J. CAMPBELL

Acute inflammation promotes the elimination of invading microorganisms and the repair of damaged tissues. While acute inflammation is clearly beneficial, there is evidence that inflammation may also cause tissue injury [1-4]. How in­flammation causes tissue injury is poorly understood. A potential mechanism is cellular hypoxia following microvascular thrombosis. Endothelial cells activated by inflammatory mediators develop increpsed permeability and flood the inter­stitial spaces with fluid. With sustained activation, tissue factor and plasminogen activator inhibitor 1 (PAll) are expressed on the luminal surface of the endothe­lial cells [5, 6]. Tissue factor activates the extrinsic clotting cascade resulting in intra-vascular fibrin deposition [7], whereas PAll neutralizes tissue plasmino­gen activator, thus preventing fibrinolysis of intra-vascular fibrin [8, 9]. Expres­sion of tissue factor and PAil may therefore transform the endothelium from an anti-coagulant to a pro-coagulant surface, thereby inducing the formation of mi­crovascular thrombi. Microvascular thrombosis may result in cellular hypoxia, lactic acidosis and irreversible tissue injury [5, 7, 9-11].

Cardiopulmonary bypass (CPB) triggers a transient inflammatory response. Inflammatory mediators are generated by activation of the complement and con­tact cascades on the surface of the oxygenator of the CPB circuit [12-14]. En­dothelial cells throughout the body are activated when these inflammatory me­diators are pumped back into the circulation. The subsequent inflammatory re­sponse is associated with increased postoperative morbidity [14-17].

While the morbidity associated with the systemic inflammatory response to CPB is mild in comparison to other inflammatory conditions such as septic shock, the study of these patients offers the advantage of an inflammatory insult that is reproducible and has a defined time of onset. Measurements may there­fore be made before, during and after the insult.

Prothrombin fragment 1+2 (PTF 1+2) is released when prothrombin is con­verted to thrombin, following activation of the extrinsic clotting cascade by tis­sue factor. Plasma PTF 1+2 levels therefore reflect the extent of intra-vascular fibrin deposition. CPB is associated with inhibition of intra-vascular fibrinolysis due to increased PAl 1 expression [18]. In this setting PTF 1 + 2 levels may re­flect the magnitude of microvascular thrombosis.

24 B. Dixon, J.D. Santamaria, D.J. Campbell

Therefore to elucidate the role of microvascular thrombosis in tissue injury associated with acute inflammation we examined the relationship between PTF 1 + 2 levels and organ responses in patients undergoing elective CPB.

Methods The Human Research Ethics Committee of St Vincent's Hospital approved this study and all patients gave written informed consent. Thirty patients scheduled for elective cardiac surgery requiring CPB were studied. Exclusion criteria in­cluded preoperative blood creatinine greater than 0.2 mmollL, bilirubin greater than 28 micromollL, severe airways disease, evidence of systemic inflamma­tion (temperature greater than 37.5°C or white cell count greater than 12 x 1091L) , clotting abnormalities, and immunosuppressive therapies, including glucocorticoid therapy. Plasma levels of PTF 1+2 and PAl activity were meas­ured preoperatively and at 0, 1, 2, 3, 4, and 7 hours following CPB. Tempera­ture, haemodynamic variables, and arterial and mixed venous lactate levels were measured preoperatively and at 1, 2, 3, 4, 7, 10, 13, and 16 hours follow­ing CPB. The arterial partial pressure of oxygen to inspired fraction of oxygen ratio (Pa02IFP2) was measured preoperatively and at 1, 2, 3, 4, and 7 hours following CPB. Urine output was measured at 0, 1,2, 3, and 4 hours following CPB. Fluid balance was calculated from the time of admission to the Intensive Care Unit until sixteen hours post-CPB. White cell count, haemoglobin, platelet count, creatine kinase (CK) and CKMB levels were measured preoperatively and on the first postoperative day. Plasma urea, creatinine, bilirubin and alanine aminotransferase levels were measured preoperatively and on the second post­operative day.

Patients were induced with fentanyl (15-50 microg/kg), propofol (0.5-1 mg/kg) and pancuronium (0.1 mg/kg). Anaesthesia was maintained with isoflu­rane. All patients received 5 mg of epsilon-aminocaproic acid IV and 5 mg into the pump prime. No patients received aprotinin. Following induction of anaes­thesia, a 6 lumen Swan Ganz VIP catheter was inserted (Baxter Healthcare, Irvine, CA, USA), the bladder was catheterized, and trans-oesophageal echocar­diography (TOE) was performed to assess left ventricular function.

Five surgeons participated in the study. CPB was performed using a Cobe membrane oxygenator (Cobe Cardiovascular, Inc, Arvado, CO, USA) with a 40 micron filter pre-patient. Pump prime was with plasmalyte 1000 mL, colloid 500 mL, sodium bicarbonate 20 mmol and calcium chloride 5 mmol. Patients were heparinised to maintain an activated clotting time above 400 seconds and reversed with protamine at the conclusion of CPB. For all but one of the sur­geons, arrest of the heart was managed using a combination of antegrade and retrograde blood cardioplegia without active cooling, and central temperatures plateaued at 32-33°C. The remaining surgeon used cold crystalloid antegrade

Prothrombin Fragment 1+2 Levels 25

cardioplegia with active cooling to 28°C. All patients were actively rewanned to 37°C before discontinuation of CPB.

Intra-vascular pressures and cardiac outputs were recorded at end-expiration using standard strain gauge pressure transducers calibrated to zero at the mid­axillary line fourth intercostal space and coupled to a monitor with a calibrated screen (Merlin, Hewlett Packard Ltd., Andover, MA, USA). Cardiac output was measured by the thermodilution technique, with 10 mL of 5% glucose at ambi­ent temperature. Cardiac index, systemic vascular resistance index, pulmonary vascular resistance index, oxygen delivery index and oxygen consumption index were calculated using standard formulae. Arterial and mixed venous bloods were obtained within 5 minutes of the cardiac output determinations and imme­diately analyzed for pOz, oxygen saturation and lactate.

Data analysis Patients were divided into high and low PTF 1+2 groups by the median value of the area under the curve calculated for PTF 1+2 levels measured at 0, 1,2,3,4, and 7 hours post-CPB. Categorical variables were compared using chi-square analysis, parametric variables with Student's t-test, and non-parametric vari­ables with the Mann-Whitney U test. Repeated Measures ANOVA was used to compare variables measured over time. A P value of < 0.05 was considered sta­tistically significant. Statistical analysis was performed with Stata (Release 6, Stata Corporation, TX, USA).

Results Of the 30 patients studied, 11 were women; 24 patients underwent coronary ar­tery bypass grafting, 4 had a valve replacement, and 2 had a combined proce­dure. Preoperatively the high PTF 1+2 group had increased pulmonary capillary wedge pressure (14 v 9 mmHg, P < 0.05). However, this is unlikely to indicate impaired cardiac function because the high PTF 1 + 2 group also had a higher preoperative cardiac index (3.3 v 2.8 Llminlm2, P < 0.05) and the proportion of patients with normal left ventricular function was similar for the two groups (Table 1). Preoperative plasma creatinine (0.09 v 0.08 mmollL, P < 0.05) and PTF 1+2 levels (0.99 v 0.74 nmollL, P = 0.06) were slightly higher in the high PTF 1 + 2 group. Otherwise, the two groups had similar preoperative characteris­tics (Table 1).

26 B. Dixon, J.D. Santamaria, D.J. Campbell

Table 1. Preoperative variables for patients with high and low PTF I +2 levels post-CPB

Variable HighPTF 1+2 LowPTF 1+2 P value

PCWP(mmHg) 14 (10-16) 9(6-10) <0.05 CVP(mmHg) 8 (6-14) 6 (2-7) 0.06 MAP (mmHg) 100 (98-111) 91 (79-102) NS MPAP(mmHg) 24 (18-39) 18 (13-24) 0.08 SVRI (dyne.slem5/m2) 2114 (2022-2346) 2461 (2116-2853) NS PVRI (dyne.slems/m2) 284(185) 283 (177) NS CI (L/minlm2) 3.3 (0.5) 2.8 (2.6-3.0) <0.05 Lactate (mmollL) 1.4 (0.6) 1.3 (0.4) NS

PaOiFj02 (mmHg) 373 (350-419) 367 (329-391) NS Normal LV function (%) 67 40 NS Creatinine (mmollL) 0.09 (0.09-0.13) 0.08 (0.08-0.01) <0.05 Urea (mrnollL) 6.9 (5.5-8.3) 5.4 (5.0-6.4) NS ALT (UIL) 25 (16-36) 18 (14-34) NS Bilirubin (micromollL) 8 (5-12) 8 (8-13) NS WCC (1091L) 7.3 (5.3-8.5.) 6.6 (6.4-7.6) NS Platelets (1 091L) 246 (80) 220 (74) NS Hb (gIL) 138 (17) 134(15) NS PTF 1 +2 (nmollL) 0.99 (0.65-1.7) 0.74 (0.65-0.85) 0.06 PAl activity (U/mL) 3.2 (2.6) 2.1 (1.7) NS

Data shown as mean ± SD or median (interquartile range). PCWP. pulmonary capillary wedge pressure; CVP, central venous pressure, MAP. mean systemic blood pressure; MPAP. mean pulmonary artery pressure; Cl, cardiac index; SVRl, systemic vas­cular resistance index; PVRl, pulmonary vascular resistance index; PaO,lFiO" arterial partial pressure of oxygen to inspired fraction of oxygen ratio; ALT, alanine aminotransferase; WCC, white cell count; Hb, haemoglobin; LV, left ventricle; PTF 1+2, prothrombin fragment I +2; PAl, plasminogen activator inhibitor

PTF 1+2 levels and PAl activity

Both PTF 1+2 levels and PAl activity post-CPB were significantly elevated above preoperative levels, PTF 1+2 levels peaked at one-hour post-CPB and re­mained elevated at seven hours post-CPB (Fig, 1), PAl activity was also markedly elevated following CPB and levels peaked at seven hours post-CPB (Fig. 2). The high PTF 1 +2 group was associated with a trend towards increased PAl activity, P = 0.07 (Fig. 2).

Pulmonary function

Both groups showed deterioration in P a02IFP2 ratio following CPB, with a greater deterioration in the high PTF I +2 group (Fig. 3). Duration of mechani­cal ventilation was similar for the two groups (Table 2).

Prothrombin Fragment 1 + 2 Levels

N + 12 ~

.$ c 9 Q)

E~ 0>::::: co "0 6 -= E . ~ c ..c ~ E 3 e .c (5

0 a:

It't,. -t

0,0-0'0'0 ~. ~ ---0

I I I I I I I I

pre 0 2 3 4 5 6 7 Time post-bypass (hours)

27

• High PTF1+2 0 Low PTF1+2

Fig. 1. Change in PTF 1 +2 levels (mean ± SE) with time for patients with high and low PTF 1 +2 levels post-CPB

12

o I I I

pre 0 2 3 4 5 6 7 Time post-bypass (hours)

• High PTF1 +2 o Low PTF1+2

Fig. 2. Change in PAl activity (mean ± SE) with time for patients with high and low PTF 1+2 lev­els post-CPB. Patients with high PTF 1+2 levels had a trend towards higher PAl activity, P = 0.07

Ci 500 J: E 400 I S :8 300 ~+-9-0-0-0 ~ -t-t-t-t

N 200 • High PTF1+2 0

0 Low PTF1+2 LL--- 100 N

0 '" a.. 0 I I I I I I I I

pre 0 2 3 4 5 6 7 Time post-bypass (hours)

Fig. 3. Change in PaO/FPl ratio (mean ± SE) with time for patients with high and low PTF 1+2 levels post-CPB. Patients with high PTF I +2 levels had lower P aO/FPl ratios, P < 0.05

28 B. Dixon, J.D. Santamaria, D.J. Campbell

Renal function and fluid balance

Despite a more positive fluid balance (Table 2), the high PTFl+2 group had a lower urine output (Fig. 4). In addition, plasma urea levels did not decrease from preoperative levels on the second postoperative day in the high PTF 1 +2 group (Table 2).

Lactate and oxygen consumption

There was a trend to increased arterial lactate levels measured from 1 to 16 hours post-CPB in the high PTF 1+2 group (P = 0.09), with a maximal differ­ence between the two groups at 7 hours post-CPB (Fig. 5). Oxygen consump­tion post-CPB (\102) was similar for the two groups (data not shown).

CPB time and postoperative variables including haemodynamics, blood haematology, liver function tests, CK, CKMB and inotrope use were similar for the two groups (Table 2).

Table 2. Operative and postoperative variables for patients with high and low PTF 1 +2 levels post­CPB

Variable HighPTF 1+2 LowPTF1+2 Pvalue

CPB time (min) 108 (27) 97 (24) NS

lla Creatinine (mmollL) 0.0 (-0.02-0.01) 0.0 (-0.01-0.0) NS

lla Urea (mmollL) 0.2 (-1.2-2.4) - 1.4 (- 2.2-0.Q7) <0.05 lla ALT (UIL) -2.0 (-8-3.0) -3.5 (-9.5-0.5) NS

lla Bilirubin (micromollL) 3 (1-4) 1 (0.5-4) NS

CK(UIL) 857 (571-1000) 674 (448-803) NS

CKMB (UIL) 33 (24-60) 41 (33-58) NS

WCC (1091L) 10 (9-15) 11 (9-14) NS

Platelets( 1 091L) 133 (114-186) 139 (125-164) NS

Hb (gIL) 97.3 (8.7) 95.9 (10.4) NS

Hours of ventilation 10 (8-14) 12 (10-14) NS

Total fluid balanceb (mL) 2220 (1431) 1269 (1097) 0.05

Urine volumeb (mL) 1480 (973-2110) 1989 (1409-2574) 0.10

Post-op noradrenaline (%) 20 13 NS

Post-op adrenaline (%) 27 13 NS

Hours in ICU 22.3 (20.5-45.5) 21.5 (20.5-23.0) NS

Data shown as mean ± SD or median (interquartile range). CPB, cardiopulmonary bypass; ALT, alanine aminotransferase; CK, creatine kinase; WCC, white cell count; Hb, haemoglobin; ICU, Intensive Care Unit. 'Value on second postoperative day mi­nus preoperative value. bFluid balance calculated from the time of admission to the Intensive Care Unit until sixteen hours post-CPB

Prothrombin Fragment 1+2 Levels

-c­::J o

400

oE 300 I "5200 .e-::J o 2! 100

.;::

::::> o

o 2 3 4 Time post-bypass (hours)

• High PTF1 +2 o Low PTF1+2

29

Fig. 4. Change in urine output (mean ± SE) with time for patients with high and low PTF I +2 lev­els post-CPB. Patients with high PTF I +2 levels had lower urine output, P < 0.05

~ 5 :::::: (5

A-f~, E 4 E

2 3 113

Y 9-9-¢~~::~ t3 2 • High PTF1+2 ~

~ 0 Low PTF1+2 ... 1 ~ Q) t:: <{

0 I I I [ I I I [ I i I I r I I I I i

pre 0 4 8 12 16 Time post-bypass (hours)

Fig. 5. Change in arterial lactate (mean ± SE) with time for patients with high and low PTF 1+2 levels post-CPB. Patients with high PTF 1+2 levels had a trend towards higher arterial lactate lev­els, P = 0.09

Discussion To investigate the role of microvascular thrombosis in mediating tissue injury associated with acute inflammation we studied patients undergoing CPB. A ho­mogenous group of patients was selected with a low risk of postoperative mor­bidity. In general, these patients have an uneventful postoperative course and are discharged from the Intensive Care Unit within 24 hours of CPB. In such a group of patients it is unrealistic to expect evidence of tissue injury based on clinically defined outcomes such as organ failure. Instead, we evaluated the rela­tionship between PTF 1+2 levels and organ responses to CPB.

Our main findings were a marked increase from baseline in PTF 1+2 levels and PAl activity post-CPB and an association between PTF 1 +2 levels and

30 B. Dixon, 1.0. Santamaria, 0.1. Campbell

pulmonary and renal responses to CPB. Following CPB, PTF 1+2 levels in­creased by approximately 9-fold and PAl activity by 3-fold from preoperative levels. The increased PTF 1+2 levels and PAl activity illustrate the marked in­flammatory response associated with CPB, given that patients undergoing coronary artery bypass grafting without CPB show only a 30% increase in PTF 1+2 levels [19].

The main differences between the high and low PTF 1+2 groups were the pulmonary and renal responses to CPB. Whereas both groups showed deteriora­tion in pulmonary function, as assessed by the P a02IFP2 ratio, deterioration was greater in the high PTF 1+2 group. Tissue injury associated with microvas­cular thrombosis may be responsible for this deterioration in pulmonary func­tion, although increased endothelial permeability and associated pulmonary oedema may also explain these findings [14, 16, 20]. The high PTF 1+2 group also showed evidence of an impaired renal response to CPB in that, despite a more positive fluid balance, these patients had reduced urine output and plasma urea levels did not decrease from preoperative levels on the second postopera­tive day. This impaired renal response was unlikely to be due to increased en­dothelial permeability and renal interstitial oedema, and was more consistent with tissue injury following microvascular thrombosis. Prerenal mechanisms were unlikely to contribute to the renal impairment because there were no sig­nificant differences between the high and low PTF 1 +2 groups in haemodynam­ic variables following CPB.

Cellular hypoxia following microvascular thrombosis may be expected to give rise to increased lactate production and reduced oxygen consumption. A trend to increased lactate production was evident in the high PTF 1 +2 group, with arterial lactate levels peaking at 7 hours post-CPB. This delayed increase in lactate levels was consistent with microvascular thrombosis given that endothe­lial expression of tissue factor and PAIl requires de novo synthesis and is there­fore delayed by 3 to 4 hours following an inflammatory insult [21]. We found no difference in V02 between the two groups. However V02 is influenced by postoperative variables additional to impaired tissue perfusion due to microvas­cular thrombosis. For example, muscle activity is an important determinant of V02, and may be increased by shivering and agitation, or reduced by sedation. Therefore, small changes in V02 due to cellular hypoxia associated with mi­crovascular thrombosis are unlikely to be detected in this setting.

Conclusion PTF 1 +2 levels were associated with pulmonary and renal impairment post-CPB and a trend to increased arterial lactate levels. These findings support a role for microvascular thrombosis in mediating tissue damage associated with acute in­flammation.

Prothrombin Fragment 1+2 Levels 31

References I. Tracey KJ, Beutler B, Lowry SF et al (1986) Shock and tissue injury induced by recombinant

human cachectin. Science 234:470-474 2. Jansen PM, Boermeester MA, Fischer E et al (1995) Contribution of interleukin-I to activa­

tion of coagulation and fibrinolysis, neutrophil degranulation, and the release of secretory­type phospholipase A2 in sepsis: studies in nonhuman primates after interleukin-I alpha ad­ministration and during lethal bacteremia. Blood 86: 1027 -1034

3. Ohlsson K, Bjork P, Bergenfeldt M et al (1990) Interleukin-I receptor antagonist reduces mortality from endotoxin shock. Nature 348:550-552

4. Hinshaw LB, Emerson TE Jr, Taylor FB Jr et al (1992) Lethal Staphylococcus aureus-induced shock in primates: prevention of death with anti-TNF antibody. J Trauma 33:568-573

5. Yamamoto K, Loskutoff OJ (1996) Fibrin deposition in tissues from endotoxin-treated mice correlates with decreases in the expression of urokinase-type but not tissue-type plasminogen activator. J Clin Invest 97:2440-2451

6. Schleef RR, Loskutoff OJ, Podor TJ (1991) Immunoelectron microscopic localization of type I plasminogen activator inhibitor on the surface of activated endothelial cells. J Cell Bioi 113:1413-1423

7. Lawson CA, Yan SO, Yan SF et al (1997) Monocytes and tissue factor promote thrombosis in a murine model of oxygen deprivation. J Clin Invest 99: 1729-1738

8. Seki T, Healy AM, Fletcher OS et al (1999) IL-I beta mediates induction of hepatic type I plasminogen activator inhibitor in response to local tissue injury. Am J Physiol 277 :G80 1-809

9. Pinsky OJ, Liao H, Lawson CA et al (1998) Coordinated induction of plasminogen activator inhibitor-I (PAl-I) and inhibition of plasminogen activator gene expression by hypoxia pro­motes pulmonary vascular fibrin deposition. J Clin Invest 102:919-928

10. Ivanyi B, Thoenes W (1987) Microvascular injury and repair in acute human bacterial pyelonephritis. Virchows Arch A Pathol Anat Histopathol 41 1:257-265

II. Asaka S, Shibayama Y, Nakata K (1996) Pathogenesis of focal and random hepatocellular necrosis in endotoxemia: microscopic observation in vivo. Liver 16: 183-187

12. Wachtfogel YT, OeLa Cadena RA, Colman RW (1993) Structural biology, cellular interac­tions and pathophysiology of the contact system. Thromb Res 72: 1-21

13. te Velthuis H, Jansen PG, Hack CE et al (1996) Specific complement inhibition with heparin­coated extracorporeal circuits. Ann Thorac Surg 61: 1153-1157

14. Shimamoto A, Kanemitsu S, Fujinaga K et al (2000) Biocompatibility of silicone-coated oxy­genator in cardiopulmonary bypass. Ann Thorac Surg 69: 115-120

15. Fitch Je, Rollins S, Matis L et al (1999) Pharmacology and biological efficacy of a recombi­nant, humanized, single-chain antibody C5 complement inhibitor in patients undergoing coro­nary artery bypass graft surgery with cardiopulmonary bypass. Circulation 100:2499-2506

16. Chai PJ, Nassar R, Oakeley AE et al (2000) Soluble complement receptor-I protects heart, lung, and cardiac myotilament function from cardiopulmonary bypass damage. Circulation 101:541-546

17. Gu YJ, Mariani MA, van Oeveren W et al (1998) Reduction of the int1ammatory response in patients undergoing minimally invasive coronary artery bypass grafting. Ann Thorac Surg 65:420-424

18. Ray MJ, Marsh NA (1997) Aprotinin reduces blood loss after cardiopulmonary bypass by di­rect inhibition of plasmin. Thromb Haemost 78: 1021-1026

19. Mariani MA, Gu Y J, Boonstra PW et al (1999) Procoagulant activity after off-pump coronary operation: is the current anticoagulation adequate? Ann Thorac Surg 67: 1370-1375

20. Imamura S, Matsukawa A, Ohkawara S et al (1997) Involvement of tumor necrosis factor-al­pha, interleukin-I beta, interleukin-8, and interleukin-I receptor antagonist in acute lung in­jury caused by local Shwartzman reaction. Pathol Int 47: 16-24

21. ten Cate JW, van der Poll T, Levi M et al (1997) Cytokines: triggers of clinical thrombotic disease. Thromb Haemost 78:415-419

Oxidative Stress and Apoptosis in Sepsis and the Adult Respiratory Distress Syndrome

T. WHITEHEAD, H. ZHANG

There has been considerable research into the role of reactive oxygen species (ROS) and more recently reactive nitrogen species in critical illness during the past decade. These highly reactive, short-lived compounds are generated contin­uously in normal cellular metabolism. Initially thought of as merely by-products to be detoxified as quickly as possible by the body's numerous antioxidants, it is now recognized that they have important regulatory functions. Nevertheless, it is a widely accepted theory that during critical illness, particularly the acute respi­ratory distress syndrome (ARDS) and sepsis, and during conditions of ischemia­reperfusion, the increased production of reactive species overwhelms the de­fense systems resulting in oxidant stress. This may contribute significantly to the morbidity and mortality of these conditions. Although a compelling theory, attempts to improve outcome in the critically sick with antioxidants have gener­ally failed. In this article we shall briefly review the evidence for oxidant stress in sepsis and ARDS. We will then focus on one aspect of ROS in critical illness - their possible role in apoptosis, which is increasingly recognized as a key process in the critically ill patient.

Redox imbalance in ARDS and sepsis Three main lines of evidence indicate that oxidative forces are important in the pathophysiology of ARDS and sepsis. Firstly, several studies have shown in­creased oxidants or oxidant damage in these conditions. Ventilated patients with ARDS have increased levels of breath condensate H202 compared with controls [1]. In sepsis and ARDS increased lipid peroxidation has been demonstrated [2, 3]. Plasma levels of the enzyme xanthine oxidase (XOD) are increased in ARDS and sepsis [4,5]. XOD oxidizes xanthine and hypoxanthine with the formation of superoxide and H202. XOD is itself formed by the oxidation of its thiol groups or limited poteolysis. Plasma protein thiol levels are lower in non-sur­vivors of ARDS compared with survivors, suggesting a worse outcome is asso­ciated with greater oxidant stress [4]. Secondly, antioxidant defenses are re­duced, with decreased plasma levels of iron-binding capacity, and antioxidant compounds including a-tocopherol, ascorbic acid, selenium, and retinol [6-8].

34 T. Whitehead, H. Zhang

Thirdly, in animal models of sepsis [9, to] and ARDS [11-13] it has been shown that a variety of antioxidant compounds can reduce the mortality or favorably affect the disease process.

Apoptosis and ROS in critical illness Apoptosis, or programmed cell death (PCD) , is the process by which cells 'commit suicide' in response to a variety of signals. In a physiological sense, it serves to eliminate cells that are no longer needed, are surplus, or are damaged. The process occurs in an ordered stereotyped way. In contrast to necrotic cell death, membrane integrity is maintained, such that potential inflammatory intra­cellular material does not spill into the extracellular space prior to ingestion by macrophages. This is believed to be particularly important for neutrophils that have the potential to cause significant tissue damage [14]. Many cell types un­dergo apoptosis when exposed to mild oxidative stress, with necrosis taking place under greater oxidative stress [15]. There is considerable evidence that apoptosis occurs in the critically sick, and it is possible that ROS have an impor­tant role in the PCD of some cell types.

Neutrophil apoptosis in sepsis and ARDS Neutrophils have a limited life span of a few hours and undergo apoptosis in vi­vo or in vitro. Several studies have shown that apoptosis of neutrophils from septic patients is delayed, and serum from patients with systemic inflammatory response syndrome (SIRS) and sepsis can delay apoptosis in neutrophils taken from healthy controls [16, 17]. Cytokines including interleukin (IL )-1[3, IL-6, timor necrosis factor (TNF)-a., interferon-y, granulocyte/macrophage colony­stimulating factor (GM-CSF), and G-CSF have been implicated as mediating this effect. The evidence that ROS are involved comes from studies demonstrat­ing that neutrophil apoptosis can be accelerated by H20 2 in vitro, whereas an­tioxidants and catalase inhibit PCD [18, 19]. Recently Fanning et al. [17] stud­ied the effect of serum taken from patients with SIRS on apoptosis in neu­trophils taken from healthy controls. They found that the ability of SIRS plasma to delay neutrophil apoptosis could be attenuated by the presence of anti-GM­CSF antibodies, but not antibody to IL-6, IL-l[3, and TNF-a.. Supplementation of the plasma with the anti-inflammatory cytokine IL-to promoted apoptosis. Under conditions in which apoptosis was inhibited, generation of ROS was re­duced, suggesting these may be involved in the intracellular control of neu­trophil apoptosis.

Unstimulated and activated neutrophils may undergo apoptosis via different final pathways. Caspases are probably involved in PCD of unstimulated neu-

Oxidative Stress and Apoptosis in Sepsis 35

trophils. However, these enzymes contain numerous cysteine residues and are redox sensitive, and in the oxidizing environment of the activated neutrophil there is evidence that apoptosis occurs by a caspase-independent route [20].

In ARDS delayed neutrophil apoptosis has also been observed, probably me­diated by factor(s) in the lavage fluid [21, 22]. It is tempting to speculate that this prolongation of the neutrophil life span is an adaptive measure to fight in­fection, and in that sense is beneficial, inappropriate persistence of neutrophils could contribute to the tissue damage and systemic inflammatory changes char­acteristic of ARDS and SIRS.

Lymphocyte apoptosis in sepsis Whereas neutrophil apoptosis is delayed in sepsis, studies in both animal mod­els and humans indicate that lymphocyte populations undergo accelerated PCD [16,23]. ROS have been implicated in lymphocyte apoptosis within the physio­logical setting of elimination of activated T cells [24], and in the pathological context of death of T cells in human immunodeficiency virus disease [25]. There is some evidence from animal models that ROS may be involved in caus­ing lymphocyte apoptosis. Freeman et al. [26] recently studied a cecal ligation and puncture (CLP) model of sepsis in mice that either constitutively overex­pressed the CuZn superoxide dismutase (SOD), or were partially deficient in this intracellular enzyme. CuZn SOD-deficient mice showed greater lymphocyte depletion than controls, although it was not possible to quantify the difference. Interestingly, overexpression of CuZn SOD did not prevent apoptosis, despite enzyme activity 2-3 times that of the control group.

Whether lymphocyte apoptosis in sepsis is detrimental - possibly by induc­ing a state of immune suppression, beneficial - by reducing inflammation, or merely an association of critical illness is not known. However, one study [27] has suggested that apoptosis in sepsis is harmful. Again using a CLP mouse model, it was found that lymphocyte apoptosis and mortality is reduced in trans­genic mice that overexpress the anti-apoptosis Bcl-2 gene in T-Iymphocytes compared with control mice. This may also be interpreted as evidence in favor of ROS-stimulated lymphocyte apoptosis, since Bcl-2 may act as an antioxidant [28], although this is controversial [29].

Apoptosis in specific organs PCD may contribute to multiple organ dysfunction syndrome (MODS) in the critically ill [30]. In a CLP model in mice, apoptosis is detectable in parenchy­mal cells of the ileum, colon, lung, kidney, and skeletal muscle, but not heart, brain or liver [31]. Injection of a lethal dose of lipopoly saccharide (LPS) into

36 T. Whitehead, H. Zhang

mice gave similar findings [32]. There is little evidence for extensive end-organ apoptosis in human sepsis, however. Hotchkiss et al. [31] recently conducted a postmortem study of 20 patients who had died of sepsis and MODS. Although epithelial cell apoptosis was evident in the colon and ileum, the dominant find­ing was extensive lymphocyte apoptosis.

Endothelial cell apoptosis may also contribute to organ dysfunction. TNF-a and LPS can induce widespread endothelial cell apoptosis in mice [33], preced­ing non-endothelial cell damage. Since antioxidants can protect against LPS-in­duced apoptosis of endothelial cells in vitro [34] and in vivo [35], it is possible that this process is mediated by ROS. Nitric oxide appears to have a protective role [36].

Signalling of apoptosis by ROS Thus there is evidence that ROS are important in signalling apoptosis in sepsis and ARDS. It is likely that a variety of pro-apoptotic signals, such as LPS and TNF-a, can trigger intracellular signalling involving ROS, and that it may be considered as either pathological or physiological, depending on the circum­stances. The intracellular pathway by which ROS mediate apoptosis is not known. However the second messenger role for ROS is not confined to the regu­lating PCD; there is good evidence that ROS act in signal transduction for the cytokines platelet-derived growth factor [37] and opidermal growth factor [38], where catalase is found to block ligand-induced phosphorylation of several tar­get proteins.

As well as an action in intracellular signal transduction, it is possible that de­rivatives of oxidant stress within the systemic circulation can have effects on apoptosis in remote cells or organs. Compton et al. [39] recently showed that perfusion of isolated rat lungs with a solution containing 50 Ilmolll of 4-hydrox­ynonenal (HNE) led to edema and DNA laddering characteristic of apoptosis in the perfused lung. HNE is a lipid peroxide present at a concentration of < 1 Ilmolll in the plasma of unstressed subjects. However, levels of up to 50 Ilmolll have been reported in a dog model of hypovolemic shock [40], although such high concentrations have not been recorded in humans. HNE is much more stable than superoxide or hydroxyl radicals, peroxynitrite, or H20 2, and its half­life in the circulation is around 2 min. Thus it is conceivable that HNE or other lipid peroxides act as 'toxic messengers' in critical illness.

Conclusion There is increased oxidant stress in the critically ill, and it is likely that this has a deleterious effect. It is not clear whether damage is entirely due to the indis-

Oxidative Stress and Apoptosis in Sepsis 37

criminate attack by ROS on lipids, DNA, and proteins, or whether there is more­subtle effect, with alterations in cell behavior mediated by ROS acting as sig­nalling molecules.

During critical illness the rate of apoptosis decreases in neutrophils, but markedly increases in lymphocytes and possibly in certain end-organ parenchy­mal cells. In vitro and animal studies have indicated that ROS can influence apoptosis, usually accelerating the process. The importance of PCD in critical illness remains to be clarified, as does the role of ROS in apoptosis, and whether antioxidant drugs can influence this process. It is to be hoped that a better un­derstanding of these issues will lead to therapeutic advances in critical care.

References 1. Kietzmann D, Kahl R, Muller M et al (1993) Hydrogen peroxide in expired breath condensate

of patients with acute respiratory failure and with ARDS. Intensive Care Med 19:78-81 2. Ogilvie AC, Groeneveld AB, Straub JP et al (1991) Plasma lipid peroxides and antioxidants

in human septic shock. Intensive Care Med 17:40-44 3. Quinlan GJ, Lamb NJ, Evans TW et al (1996) Plasma fatty acid changes and increased lipid

peroxidation in patients with adult respiratory distress syndrome. Crit Care Med 24:241-246 4. Quinlan GJ, Lamb NJ, Tilley Ret al (1997) Plasma hypoxanthine levels in ARDS: implica­

tions for oxidative stress, morbidity and mortality. Am J Respir Crit Care Med 155:479-484 5. Galley HF, Davies MJ, Webster NR (1996) Xanthine oxidase activity and free radical genera­

tion in patients with sepsis syndrome. Crit Care Med 24: 1649-1653 6. Metnitz PGH, Bartens C, Fischer Met al (1999) Antioxidant status in patients with the respi­

ratory distress syndrome. Intensive Care Med 25: 180-185 7. Gutteridge JMC, Quinlan GJ, Mumby S et al (1994) Primary plasma anti-oxidants in adult

respiratory distress syndrome patients: changes in iron oxidising, iron-binding and free radi­cal-scavenging proteins. J Lab Clin Med 124:263-267

8. Goode HF, Cowley HC, Walker BE et al (1995) Decreased antioxidant status and increased lipid peroxidation in patients with septic shock and secondary organ dysfunction. Crit Care Med 23:646-651

9. Kong CW, Tsai K, Chin JH et al (2000) Magnolol attenuates peroxidative damage and im­proves survival ofrats with sepsis. Shock 13:24-28

10. Bakker J, Zhang H, Depierreux M et al (1994) Effects of n-acetylcysteine in endotoxic shock. J Crit Care 9:236-243

11. Gonzalez PK, Zhuang J, Doctrow SR et al (1996) Role of oxidant stress in the adult respirato­ry distress syndrome: evaluation of a novel antioxidant strategy in a porcine model of endo­toxin-induced lung injury. Shock 6[SupplJl :S23-26

12. Davreux CJ, Soric I, Nathens AB et al (1997) n-acetyl cysteine attenuates lung injury in the rat. Shock 8:432-438

13. Robbins CG, Horowitz S, Merritt TA et al (1997) Recombinant human superoxide dismutase reduces lung injury caused by inhaled nitric oxide and hyperoxia. Am J Physiol 272:L903-L907

14. Haslett C (1997) Granulocyte apoptosis and inflammatory disease. Br Med Bull 53:669-683 15. Clutton S (1997) The importance of oxidative stress in apoptosis. Br Med Bull 53:662-668 16. Mahidhara R, Billiar TR (2000) Apoptosis in sepsis. Crit Care Med 28[SuppIJ4: 104-113 17. Fanning NF, Kell MR, Shorten GD et al (1999) Circulating granulocyte macrophage colony­

stimulating factor in plasma of patients with the systemic inflammatory response syndrome

38 T. Whitehead, H. Zhang

delays neutrophil apoptosis through inhibition of spontaneous reactive oxygen species gener­ation. Shock 11:167-174

18. Rollet-Labelle E, Grange MJ, Elbim C et al (1998) Hydroxyl radical as a potential intracellu­lar mediator of polymorphonuclear neutrophil apoptosis. Free Radic BioI Med 24:563-572

19. Aoshiba K, Yasui S, Hayashi M et al (1999) Role of mitogen-activated protein kinase in spon­taneous apoptosis of human neutrophils. J ImmunolI62:1692-l700

20. Fadeel B, Ahlin A, Henter Jet al (1998) Involvement of caspases in neutrophil apoptosis: reg­ulation by reactive oxygen species. Blood 92:4808-4818

21. 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

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

23. 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

24. Hildeman DA, Mitchell T, Teague TK et al (1999) Reactive oxygen species regulate activa­tion-induced T cell apoptosis. Immunity 10:735-744

25. Romero-Alvira D, Roche E (1998) The keys of oxidative stress in acquired immune deficien­cy syndrome apoptosis. Med Hypotheses 51:169-173

26. Freeman BD, Reaume AG, Swanson PE et al (2000) Role of CuZn superoxide dismutase in regulating lymphocyte apoptosis during sepsis. Crit Care Med 28:1701-1708

27. Hotchkiss RS, Swanson PE, Cobb JP et al (1999) Overexpression of Bcl-2 in transgenic mice decreases apoptosis and improves survival in sepsis. J ImmunolI62:4148-4156

28. Hockenbury DM, Oltavai ZN, Yin XM et al (1993) Bcl-2 functions as an antioxidant pathway to prevent apoptosis. Cell 75:241-251

29. Jacobson MD (1996) Reactive oxygen species and programmed cell death. Trends Biochem Sci 21:83-86

30. Papathanassoglou ED, Moynihan JA, Ackerman MH (2000) Does programmed cell death (apoptosis) playa role in the development of multiple organ dysfunction in critically ill pa­tients? A review and a theoretical framework. Crit Care Med 28:537-549

31. Hotchkiss RS, Swanson PE, Cobb JP et al (1997) Apoptosis in lymphoid and parenchymal cells during sepsis: findings in normal and T- and B-cell-deficient mice. Crit Care Med 25: 1298-1307

32. Bohlinger I, Leist M, Gantner F et al (1996) DNA fragmentation in mouse organs during en­dotoxic shock. Am J Patho149:l381-1393

33. Haimovitz-Friedman A, Cordon-Cardo C, Bayoumy S et al (1997) Lipopolysaccharide in­duces disseminated endothelial apoptosis requiring cerarnide generation. J Exp Med 186: 1831-1841

34. Abello PA, Fidler SA, Buchman TG (1994) Thiol reducing agents modulate induced apopto­sis in porcine endothelial cells. Shock 2:79-83

35. Drab-Weiss EA, Hansra IK, Blazek ER et al (1998) Aminothiols protect endothelial cell pro­liferation against inhibition by lipopolysaccharide. Shock 10:423-429

36. Dimmeler S, Zeiher AM (1996) Nitric oxide - an endothelial cell survival factor. Cell Death Differ 6:964-968

37. Sundaresan M, Yu S-Z, Ferrans V et al (1995) Requirement for generation of H20 2 for platelet-derived growth factor signal transduction. Science 270:296-299

38. Bae YS, Kang SW, Seo MS et al (1997) Epidermal growth factor (EGF) - induced generation of hydrogen peroxide. J BioI Chern 272:217-221

39. Compton CN, Franko AP, Murray MT et al (1998) Signaling of apoptotic lung injury by lipid hydroperoxides. J Trauma 44:783-788

40. Esterbauer H, Schaur RJ, Zollner H et al (1999) Chemistry and biochemistry of 4-hydrox­ynonenal, malonaldehyde and related aldehydes. Free Radic BioI Med 11:81-128

Neutrophil Defensins in Lung Inflammation

H. ZHANG, T. WHITEHEAD

Defence against infection involves a multitude of factors and cells of the innate and acquired immune systems. Antimicrobial peptides, which include the de­fensins, are one component of innate immunity. Defensins have well document­ed bactericidal properties. In addition, recent studies have suggested that de­fensins have other, pro-inflammatory, activities. Furthermore there is evidence that defensins can damage host tissue, and may contribute to the pathogenesis of some conditions. This review summarises the recent developments in de­fensin biology that are important in understanding their role in lung infection and inflammation.

Classification of defensins

Antimicrobial peptides that contain six cysteine residues have been classified as defensins [1, 2]. The cysteine residues form 3 characteristic intramolecular disulfide bridges that are believed to be essential for their antimicrobial activity. Defensins were discovered in 1968 when Zeya and Spitznagel reported the pres­ence of "lysosomal cationic proteins" with antimicrobial properties in rabbit and guinea pig neutrophils [3, 4]. The term "defensins" was first used 15 years ago to describe human neutrophil peptides (HNP)-I, -2, -3, the human homologues of these rabbit peptides, and later adopted to encompass equivalent molecules found in other species [1, 5].

Defensins can be classified into at least three structural groups: the "classi­cal" a-defensins, the 13-defensins and insect defensins. Of the six a-defensins, four are the human neutrophil peptides - (HNP)-I, -2, -3 and -4, which are found in the azurophilic granules of the neutrophils [2, 6, 7]. The other two a-defensins, human defensin (HD)-5 and -6, are present in the secretory gran­ules of intestinal Paneth cells and in epithelial cells of the female reproduc­tive tract [8-10]. Two human 13-defensins (hBD) are described as hBD-l, lo­cated in epithelial cells of various organs [11-15], and hBD-2 in psoriatic scales [16].

40 H. Zhang, T. Whitehead

a-defensins Neutrophil defensins, or a-defensins, are encoded by a cluster of genes on chro­mosome 8, locus p23 [7, 17, 18]. HNP-2lacks the N-terminal amino acid pres­ent in the HNP-l and HNP-3 and therefore is believed to be a proteolytic prod­uct of either one or both of the HNP-l or HNP-3 peptides. HNP-l, 2 and 3 ac­count for almost 99% of the total defensin content of neutrophils [2, 19]. In hu­mans, these defensins constitute up to 5% of the total protein content of mature neutrophils and more than 50% of the total protein within the azurophilic gran­ules. The composition of azurophilic granule proteins is approximately (in ng/106 neutrophils): Elastase - 1500; Cathepsin G - 2500; Proteinase 3 - 1000; and defensins - 6000 [20].

HNP-4 is a small, cationic, arginine-rich peptide that shares only 32% amino acid homology with HNP-I-3 and lacks antimicrobial activity [21, 32]. Com­pared to a mixture of HNP-I-3, HNP-4 was found to be approximately 100 times more potent against E. coli and four times more potent against both S.fae­calis and C. albicans [21]. However, it accounts for only around 1 % of the total defensin content of neutrophils [2, 19]. During neutrophil development, the HNP-4 peptides are produced as a result of processing of a propeptide, which it­self lacks antimicrobial activity [35].

This review concentrates on the dominant neutrophil defensins - HNP 1-3, which for the sake of simplicity will be referred to as defensins. HNP-4 and the [3-defensins will not be discussed further.

Antimicrobial activity

Defensins were initially identified by their antimicrobial activity, which is di­rected against gram-negative and gram-positive bacteria. At concentrations be­tween 10 and 100 Jlg/mL purified defensins kill a wide variety of bacteria in mi­crobiological media, which have low ionic strength and low calcium and mag­nesium concentrations [5, 22, 23]. We have demonstrated that defensins also have significant antimicrobial effect against E. coli and P. aeruginosa in tissue culture media, with physiological pH, temperature and salt conditions. 48 h after incubation with defensins, the number of E. coli had decreased by about 65-fold, and P. aeruginosa by 20-fold compared to cultures in the absence of de­fensins. Thus the antimicrobial activity of defensins may be greater towards E. coli than P. aeruginosa.

Defensins can also kill fungi such as C. albicans [24] and herpes simplex virus [25]. However, they lack significant activity against the non-enveloped viruses reovirus and echovirus, suggesting that interaction with the membrane is essential for their antimicrobial function.

Neutrophil Defensins in Lung Inflammation 41

The mechanism by which defensins kill bacteria is thought to occur in 3 phases [24]. Firstly, the cationic defensin molecules bind to the bacterial outer membrane, which carries a negative charge. Individual defensin molecules or dimers then penetrate the membrane by electromotive force. This disrupts the membrane integrity and causes permeabilization. Finally irreversible target cell injury occurs, with entry into the cell of defensins and other normally excluded molecules, and by the outward leakage of essential minerals and metabolities from the target cell. Defensins can damage DNA directly and contribute to the cessation of DNA, RNA, and protein synthesis [26].

We have found that the anti-bacterial activity of defensins is enhanced by the presence of lung tissue. Defensin killing of E. coli or P. aeruginosa was 100 to 1000 fold greater when incubated in the presence of cultured primary lung ex­plants. Lung tissue alone did not show any significant microbicidal effect. This suggests that defensins indirectly cause bacterial death by stimulation of lung tissue. One possible mechanism for this is by inducing the release of reactive oxygen species toxic to micro-organisms. We found increased levels of hydro­gen peroxide in lung tissue cultures following defensin stimulation. The produc­tion of hydrogen peroxide may be initiated either by the release of cytokines, since the time course of hydrogen peroxide release was much delayed, or by a direct effect of defensins on lung tissue. We suggest that defensins kill bacteria directly by their microbicidal activity, and indirectly by inducing bacterial killing products that may be dependent on early release of cytokines from lung tissues.

Defensins as mediators of inflammation Initial studies on defensins focused on their antimicrobial activity, but subse­quent research revealed that defensins may also play a role in inflammation and regulation of specific immune response [27-29].

Defensins have chemotactic properties and can stimulate release of chemo­tactic factors from other cells. Territo et al. investigated the monocyte-chemo­tactic activity of fractionated extracts of human neutrophil granules [28]. Mono­cyte-chemotactic activity was found predominantly in the defensin containing fraction of the neutrophil granules. Purified preparations of human defensins were then tested. HNP-l and HNP-2 are chemotactic for monocytes, but HNP-3 failed to demonstrate chemotactic activity. HNP-l and HNP-2 are also potent chemoattractants for T-cells [27]. Van Wetering et al. have recently shown that defensins induce IL-8 synthesis in the airway epithelial A549 cells [29, 30], and defensins have been shown to stimulate the production of the neutrophil chemoattractant leukotriene B4 and IL-8 by alveolar macrophages [31]. We have demonstrated that defensins induce release of the c-c chemokine MCP-l from cultured lung explant tissue [32]. MCP-l is a potent chemotactic sub­stance for monocytes [42]. In addition, we also found that defensins stimulate

42 H. Zhang, T. Whitehead

the production of the cytokine tumour necrosis factor ~TNF-~ from cultured lung tissue [32].

Taken together, these studies suggest that defensins have the potential to promote the inflammatory process in the lung. They may be involved in the recruitment of inflammatory cells, including other neutrophils, monocytes and T-cells, both by their own inherent chemotactic properties and by increasing re­lease of other chemotactic agents from other cells. Furthermore they can stimu­late TNF-a release. The importance of these activities in clinical situations has not yet been addressed.

Defensins in lung injury Extremely high concentrations of defensins have been reported in various in­flammatory diseases. Panyutich et al. found that patients with bacterial meningi­tis and sepsis had plasma defensin levels ranging from 120 ng/mL to 170,000 ng/mL, compared to a mean concentration of 42 ng/mL in healthy blood donors [33]. Soong et al. reported high sputum defensin levels, ranging from 300 to > 1600 Ilg/mL (the upper detection limit in the study), in patients with cystic fi­brosis [34]. Ihi et al. have recently found high defensin concentrations in pleural fluid (13.3 ± 1.9 mg/mL [mean ± SED of patients with empyema, in BAL fluid (2.0 ± 0.9 mg/mL) of patients with bacterial pneumonia, and in cerebrospinal fluid (3.4 ± 1.2 mg/mL) of patients with bacterial meningitis [35]. However, the exact role of defensins in these conditions is not known.

We have administered defensins intratracheally at doses ranging from 1 to 10 mg/mL into the lungs of healthy mice. These doses were chosen to reflect the intrapulmonary levels that may occur clinically, according to the studies de­scribed above. We demonstrated that at low concentrations defensins do not cause significant lung dysfunction at 5 hours. However, at higher concentra­tions, defensins induce hypoxia, increased lung permeability, and an elevation of mitochondrial cytochrome c, which is usually released from mitochondria following propagation of a "death signal" [36]. This suggests that high concen­trations of defensins are detrimental to the lung.

Our investigations into the toxic activity of defensins were stimulated by in vitro studies of defensin cytotoxicity. Purified human defensins kill various hu­man and murime tumour targets in a concentration- and time-dependent fashion [37]. Soong et al. reported that 72 h after exposure to defensins, the total num­ber of tracheal epithelial cells was reduced and the permeability to trypan blue was increased [34]. Okrent et al. showed that defensins (100-200 Ilg/mL) in­creased chromium release from the lung-derived cell lines MRC-5, A549, and human umbilical vein endothelial cells at 10-20 h [38]. Cytotoxicity was also observed when 51-Cr-labeled A549 cells were stimulated for 3-20 h with 100 Ilg/mL of defensins [30, 37]. These data suggest that high doses of defensins

Neutrophil Defensins in Lung Inflammation 43

may damage the two separate barriers in the lung - the alveolar epithelium and the microvascular endothelium.

Defensins probably kill mammalian cells by disruption of the cell mem­brane in a manner similar to that described above for bacterial killing. De­fensins bind to mammalian target cells with biphasic kinetics. The initial bind­ing phase occurs within 2 minutes, with the second phase beginning by 10 min and gradually increasing to reach a plateau after 60 min. Although the impaired membrane integrity occurring during the first 30 min is reversible, it is suc­ceeded by a third phase of injury that leads to cell death [39]. This final, lethal phase begins after 30 to 60 min of incubation and requires the continued pres­ence of defensins.

Defensins may also be toxic to cells by reducing cellular antioxidant defens­es. Van Wetering et al. recently observed that defensins decrease glutathione levels in airway epithelial cells [40]. Glutathione is a potent antioxidant present in the lung and confers protection against endogenous and exogenous oxidants. The defensin-induced decrease in glutathione may lead to an increased suscep­tibility to oxidant-mediated damage. Support for this comes from the observa­tion that defensin-mediated cell lysis is increased in the presence of hydrogen peroxide [41].

Various studies have shown that in the lungs of patients with neutrophil­dominated diseases, such as ARDS. there are signs of increased oxidative stress and a decreased antioxidant capacity, which leads to an impaired oxidant-an­tioxidant balance [42]. Defensins may contribute to this imbalance by reducing glutathione levels in airway epithelial cells.

Conclusion A large body of evidence has proven that defensins exert a significant antimicro­bial activity. A number of in vitro studies and our in vivo study have also shown that defensins are cytotoxic. Potentially, therefore, defensins have both benefi­cial and harmful actions in lung disease. Further studies will hopefully clarify the relevance of defensins in clinical disease, and identify possible therapeutic interventions.

44 H. Zhang, T. Whitehead

References 1. 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 2. Selsted ME, Harwig SS, Ganz Tet al (1985) Primary structures of three human neutrophil de­

fensins. J Clin Invest 76: 1436-1439 3. Zeya HI, Spitznagel JK (1966) Antimicrobial specificity of leukocyte lysosomal cationic pro­

teins. Science 154:1049-1051 4. Zeya HI, Spitznagel JK (1968) Arginine-rich proteins of polymorphonuclear leukocyte lyso­

somes. Antimicrobial specificity and biochemical heterogeneity. J Exp Med 127:927-941 5. Ganz T, Selsted ME, Szklarek D et al (1985) Defensins. Natural peptide antibiotics of human

neutrophils. J Clin Invest 76:1427-1435 6. Lehrer RI, Ganz T, Szklarek D, Selsted ME (1988) Modulation of the in vitro candidacidal

activity of human neutrophil defensins by target cell metabolism and divalent cations. J Clin Invest 81:1829-1835

7. Linzmeier R, Michaelson D, Liu L, Ganz T (1993) The structure of neutrophil defensin genes. FEBS Lett 326:299-300

8. 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 Pathol 152:1247-1258

9. Jones DE, Bevins CL (1992) Paneth cells of the human small intestine express an antimicro­bial peptide gene. J BioI Chern 267:23216-23225

10. Cunliffe RN, Mahida YR (2000) Antimicrobial peptides in innate intestinal host defence. Gut 47:16-17

11. Becker MN, Diamond G, Verghese MW, Randell SH (2000) CDl4-dependent LPS-induced ~-defensin-2 expression in human tracheobronchial epithelium. J BioI Chern 275:29731-29736

12. Bensch KW, Raida M, Magert HJ et al (1995) hBD-l: a novel beta-defensin from human plasma. FEBS Lett 368:331-335

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

14. McCray PB Jr, Bentley L (1997) Human airway epithelia express a beta-defensin. Am J Respir Cell Mol Bioi 16:343-349

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

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

17. Palfree RG, Sadro LC, Solomon S (1993) The gene encoding the human corticostatin HP-4 precursor contains a recent 86-base duplication and is located on chromosome 8. Mol En­docrinoI7:199-205

18. Sparkes RS, Kronenberg M, Heinzmann C et al (1989) Assignment of defensin gene(s) to hu­man chromosome 8p23. Genomics 5:240-244

19. Ganz T and Lehrer RI (1994) Defensins. Curr Opin Immun 6:584-589 20. Gabay JE, Almeida RP (1993) Antibiotic peptides and serine protease homologs in human

polymorphonuclear leukocytes: defensins and azurocidin. Curr Opin ImmunoI5:97-102 21. Wilde CG, Griffith JE, Marra MN et al (1989) Purification and characterization of human

neutrophil peptide 4, a novel member of the defensin family. J BioI Chern 264: 11200-11203 22. Miyasaki KT, Bodeau AL, Ganz T et al (1990) In vitro sensitivity of oral, gram-negative, fac­

ultative bacteria to the bactericidal activity of human neutrophil defensins. Infect Immun 58:3934-3940

23. Patterson-Delafield J, Martinez RJ, Lehrer RI (1980) Microbicidal cationic proteins in rabbit alveolar macrophages: a potential host defense mechanism. Infect Immun 30: 180-192

24. Lehrer RI, Lichtenstein AK, Ganz T (1993) Defensins: antimicrobial and cytotoxic peptides of mammalian cells. Annu Rev Immunol 11: 105-128

Neutrophil Defensins in Lung Inflammation 45

25. Daher KA, Selsted ME, Lehrer RI (1986) Direct inactivation of viruses by human granulocyte defensins. J Virol 60: 1068-1074

26. Lehrer RI, Barton A, Daher KA et al (1989) Interaction of human defensins with Escherichia coli. Mechanism of bactericidal activity. J Clin Invest 84:553-561

27. Chertov 0, Michiel DF, Xu L et al (1996) Identification of defensin-I, defensin-2, and CAP37/azurocidin as T-cell chemoattractant proteins released from interleukin-8-stimulated neutrophils. J Bioi Chern 271 :2935-2940

28. Territo MC, Ganz T, Selsted ME, Lehrer R (1989) Monocyte-chemotactic activity of de­fensins from human neutrophils. J Clin Invest 84:2017-2020

29. Van Wetering S, Mannesse-Lazeroms SP, Dijkman JH, Hiemstra PS (1997) Effect of neu­trophil serine proteinases and defensins on lung epithelial cells: modulation of cytotoxicity and IL-8 production. 1 Leukoc Bioi 62:217-226

30. Van Wetering S, Mannesse-Lazeroms SP, Van Sterkenburg MA et al (1997) Effect of de­fensins on interleukin-8 synthesis in airway epithelial cells. Am J Physiol 272:L888-L896

31. Paone G, Krein P, Rouhani F, Brantly ML (1999) Human neutrophil peptides stimulate alveo­lar macrophage production of LTB4 and IL-8 [abstract]. Am J Respir Crit Care Med 159:511

32. Zhang H, Hsia C, Porro GA et al (2000) Human neutrophil defensins regulate chemokine pro­duction in lung tissues. Am J Respir Crit Care Med 161 :A518

33. Panyutich AV, Panyutich EA, Krapivin VA et al (1993) Plasma defensin concentrations are el­evated in patients with septicemia or bacterial meningitis. 1 Lab Clin Med 122:202-207

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

35. 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 Infect Dis 25: 1134-1140

36. Thress K, Kornbluth S, Smith 11 (1999) Mitochondria at the crossroad of apoptotic cell death. 1 Bioenerg Biomembr 31 :321-326

37. Lichtenstein A, Ganz T. Selsted ME, Lehrer RI (1986) In vitro tumor cell cytolysis mediated by peptide defensins of human and rabbit granulocytes. Blood 68: 1407-1410

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

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

40. Van Wetering S. Rahman I, Hiemstra PS, MacNee W (1998) Role of intracellular glutathione in neutrophil defensin-indllced IL-8 synthesis and cytotoxicity in airway epithelial cells. ElIr RespirJ 12:420s

41. Lichtenstein AK. Ganz T, Selsted ME, Lehrer RI (1988) Synergistic cytolysis mediated by hydrogen peroxide combined with peptide defensins. Cell Immunol 114: 104-116

42. Gutteridge 1M. Mitchell J (1999) Redox imbalance in the critically ill. Br Med Bull 55:49-75

Nitric Oxide: Lessons Learned and Areas of Success

W.M. ZAPOL, R. JENNEY

Pulmonary hypertension with severe hypoxemia complicates the care of patients with diseases such as acute respiratory distress syndrome (ARDS), pulmonary embolism, chronic respiratory failure, chronic pulmonary hypertension, and af­ter cardiopulmonary bypass. Pulmonary hypertension is also common in new­born patients with prematurity, meconium aspiration, pneumonia, sepsis, and congenital diaphragmatic hernias and in infants and children with many forms of congenital heart disease. Currently available intravenous vasodilator therapies often cause untoward effects when used to reduce pulmonary hypertension. For example, systemic vasodilatation and hypotension occur with all vasodilators in­fused at dosages sufficient to decrease pulmonary hypertension. In addition, in­travenous infusions of systemic vasodilators often increase the venous admix­ture due to right to left shunting of venous blood. It is extremely desirable to identify a vasodilator that selectively relaxes constricted pulmonary arteries, and does not cause systemic dilatation or increased pulmonary venous admixture.

Nitric oxide In 1987, the gaseous molecule nitric oxide (NO) was identified as an endotheli­um-derived relaxing factor (EDRF) [1]. NO is an ideal local transcellular mes­senger because of its diffusibility, lipophilic nature, and short duration of action. The numerous functions of NO in various tissues have been reviewed [2]. In vascular endothelial cells, NO is synthesized by nitric oxide synthase from the terminal guanidine nitrogen of L-arginine and molecular oxygen. NO rapidly diffuses into subjacent smooth muscle cells. There, NO binds to the heme iron complex of soluble guanylate cyclase (sGC) and stimulates the production of cyclic guanosine 3', 5'-monophosphate (cGMP). Through several incompletely understood mechanisms, cGMP causes vascular smooth muscle relaxation. The biological activity of NO is evanescent because of several mechanisms. In the presence of oxygen, NO forms higher oxides (NOJ that do not stimulate sGc. Additionally, NO binds to other heme-containing proteins and thereby is un­available to activate sGc. When NO diffuses into the intravascular space, its bi­ologic activity is limited by avid binding to hemoglobin. Interestingly, the ni-

48 WM. ZapoJ, R. Jenney

troso-vasodilators that have been used for decades to treat hypertensive patients, such as nitroglycerin and nitroprusside, act by releasing NO.

Endothelium-dependent relaxation in pulmonary arteries occurs in response to a variety of physical and pharmacological stimuli. Endogenous NO can be measured in the exhaled gas of rabbits, guinea pigs, and human beings [3]. In normal lungs, the baseline pulmonary vascular tone is low and pulmonary vas­cular resistance is unchanged by increased levels of NO. In some acute and chronic pulmonary hypertensive states, the production of endogenous NO is im­paired [4]. In these conditions, introducing NO into the vasculature often causes vasodilatation. Increasing NO by infusing acetylcholine or NO donor agents or by inhalation [5] can reduce pulmonary vascular resistance.

NO inhalation in ARDS We hypothesized that inhaled NO would diffuse into the pulmonary vasculature of ventilated lung regions and cause relaxation of pulmonary vascular smooth muscle [6, 7]. Since the NO is inhaled, the gas should be distributed predomi­nantly to well-ventilated alveoli and not to collapsed or fluid-filled regions of the lung. In the presence of increased vasomotor tone, the selective vasodilata­tion of well-ventilated lung regions by inhaled NO should cause a "steal" or di­version of pulmonary artery blood flow towards well-ventilated alveoli, and thereby improve the matching of ventilation to perfusion and thus enhancing ar­terial oxygenation (Pa02)' Such an effect would be in marked contrast to the in­creased mismatch of ventilation and perfusion caused by intravenously adminis­tered vasodilators such as nitroprusside, nitroglycerin, and prostacyclin. Al­though these intravenous agents decrease pulmonary artery pressure, they in­crease intrapulmonary shunting of deoxygenated blood, by non-selectively dilat­ing hypoventilated lung segments, and reduce the systemic Pa02' Also, inhaled NO should not produce systemic vasodilatation because, unlike available intra­venous vasodilators, it is avidly bound to haemoglobin and rapidly inactivated.

Rossaint and coworkers compared the effects of inhaling 18 and 36 ppm (parts per million by volume) NO to intravenously infused prostacyclin in nine patients with ARDS [8]. Inhaled NO selectively reduced mean pulmonary artery pressure from 37 ± 3 to 30 ± 2 mmHg (mean ± SE) and improved oxygenation by decreasing venous admixture (QvA/Qt). The improved efficiency in oxygen exchange during NO inhalation was reflected in an increase of the Pa02IFi02 ratio from 152 ± 15 mmHg to 199 ± 23 mmHg. While the intravenous infusion of prostacyclin also reduced pulmonary artery pressure, mean arterial pressure and Pa02 decreased as QvA/Qt increased. Subsequent reports have documented that inhalation of lower concentrations of NO (less than 20 ppm) also decreases pulmonary artery pressure and improves Pa02 levels [9]. Even very small in­haled concentrations of NO (as low as 250 parts per billion) may be effective in some patients [9]. Right ventricular ejection fraction increases in some patients

Nitric Oxide: Lessons Learned and Areas of Success 49

breathing inhaled NO, suggesting that decreasing pulmonary artery pressure may unload the right heart and be haemodynamically beneficial [9, 10].

A marked variation has been reported for the haemodynamic and respiratory effects of clinical NO inhalation, both among patients and within the same pa­tient at different times in their illness. It is possible that preexisting pulmonary disease as well as the concomitant administration of other vasoactive drugs and the effects of septic mediators may contribute to the observed variability. In gen­eral, the baseline level of pulmonary vascular resistance predicts the degree of pulmonary vasoconstriction that is reversible by NO inhalation. Those with the greatest degree of pulmonary hypertension appear to respond best to NO inhala­tion. A dose-response analysis of a randomized trial of NO in 177 ARDS pa­tients was recently reported by Dellinger and co-workers [11].

Tachyphylaxis has not been observed even when NO inhalation was contin­ued for up to 53 days [8]. Pulmonary artery pressure and Pa02 quickly return to baseline values, however, after discontinuation of the gas. Occasionally, sudden discontinuation of inhaled NO can produce problematic pulmonary vasocon­striction and possibly bronchoconstriction. The reason for this is unclear.

Pharmacological means to augment the improved matching of ventilation and perfusion associated with NO inhalation have been investigated. For exam­ple, the vasoconstrictor almitrine bemesylate has been given intravenously to enhance pulmonary vasoconstriction during NO breathing. This agent was ob­served to further reduce Qs/Qt in ARDS in combination with NO inhalation [12]. Further studies of the safety of such combination therapies are warranted before widespread clinical use can be advocated.

NO inhalation in neonatal respiratory failure In the foetus, intense pulmonary vasoconstriction causes oxygenated blood re­turning from the placenta to shunt right-to-Ieft across the patent foramen ovale and ductus and bypass the collapsed lungs. At birth, the lungs are distended with air and there is a sustained decrease in pulmonary vascular resistance and an increase in pulmonary blood flow. In some babies, pulmonary blood flow does not increase after birth. Persistent pulmonary hypertension of the newborn (PPHN) is characterized by an increased pulmonary vascular resistance, right­to-left shunting of deoxygenated blood across the ductus arteriosus and fora­men ovale, and severe systemic hypoxemia. Although breathing high levels of oxygen and induced alkalosis decrease pulmonary hypertension in some pa­tients with PPHN, these therapies are often unsuccessful. The use of intra­venous vasodilator therapy is limited by severe systemic hypotension, which may further reduce the Pa02 by increasing right-to-Ieft shunting in patients with PPHN. Extracorporeal membrane oxygenation (EeMO) is often used to support babies who remain hypoxemic despite maximal ventilator and medical

50 W.M. Zapol, R. Jenney

therapies. Endogenous production of NO by the pulmonary vasculature is likely to be decreased in PPHN. Therefore, a therapeutic strategy that selectively in­creases NO activity in the lung may be beneficial to many infants with pul­monary hypertension.

Clinical studies of NO inhalation have been performed in neonates [13, 14], and infants and children with pulmonary hypertension [15]. In general, they demonstrate that inhalation of NO selectively decreases pulmonary hyperten­sion and increases Pa02' In newborns with PPHN breathing NO acutely increas­es systemic oxygen levels and significantly decreases the need for ECMO [16, 17, 21]. Additionally, in children with many forms of congenital heart disease inhaled NO decreases pulmonary hypertension.

In paediatric as well as adult patients the pulmonary vasodilator response to NO inhalation is variable. In the neonatal lung, the degree of improvement with NO depends on the initial degree of pulmonary vasoconstriction and hypoxemia [16] and the recruitment of an adequate lung volume.

Inhaled NO and the prevention of pulmonary vascular disease Laboratory studies suggest that inhaled NO protects the injured lung from vas­cular disease. Lung injury is associated with endothelial cell dysfunction, leak­age of serum into the vessel wall, and proliferation of smooth muscle cells or their precursors. The resulting hypertrophic and hyperplastic response of pul­monary cells is associated with increased pulmonary artery tone, restriction of lung blood flow, and pulmonary hypertension.

NO may decrease the hyperplastic response to lung injury. NO decreases cell proliferation in culture [18] in part by inhibiting transduction of growth factor signalling, expression of regulators of cell cycle progression, and in­hibiting the activities of several critical enzymes required for DNA synthesis. It is also possible that decreased endogenous NO production in injured en­dothelial cells leads to unbridled proliferation of pulmonary artery cells and vascular remodelling.

Several studies suggest that inhaling NO attenuates the remodelling of in­jured pulmonary arteries. In hypoxic infant [19] and adult rats [20] chronically inhaled NO attenuates pulmonary artery muscularization, and right ventricular hypertrophy. Although studies in adult animals suggest that pulmonary vasocon­striction is required for the protective effect of inhaled NO, a recent report sug­gests that NI prevents neomuscularization of pulmonary arteries and prolifera­tion of smooth muscle precursor cells in injured infant rat lungs without pul­monary hypertension. These studies suggest that inhaled NO might have an im­portant role in preventing many forms of pulmonary vascular disease. They also provide the basis for clinical studies currently underway to determine whether inhaled NO prevents pulmonary artery disease in the injured lungs of patients.

Nitric Oxide: Lessons Learned and Areas of Success 51

For a more complete and thorough review of inhaled nitric oxide please refer to a recent review [22].

References 1. Ignarro LJ, Buga GM, Wood KS, Byrns RE (1987) Endothelium-derived relaxing factor pro­

duced and released from artery and vein is NO, Proc Natl Acad Sci USA 84:9265-9269 2. Ignarro LJ (1990) Signal transduction mechanisms involving nitric oxide. Biochem Pharma­

col 41 :485-490 3. Gerlach H, Rossaint R. Pappert 0 (1994) Autoinhalation of nitric oxide after endogenous

synthesis in nasopharynx. Lancet 343:518-519 4. Dinh Xuan AT, Higenbottam TW, Clelland C et al (1991) Impairment of endothelium-de­

pendent pulmonary artery relaxation in chronic obstructive lung disease. N Engl J Med 324: 1539-1547

5. Pepke-Zaba 1, Higenbottam TW, Dinh-Xuan AT et al (1991) Inhaled nitric oxide as a cause of selective pulmonary vasodilation in pulmonary hypertension. Lancet 338: 1173-1174

6. Frostell C, Fratacci M-D, Wain JC et al (1991) Inhaled nitric oxide: A selective pulmonary vasodilator reversing hypoxic pulmonary vasoconstriction. Circulation 83:2038-2047

7. Roberts JD Jr, Chen T-y, Kawai N et al (1993) Inhaled nitric oxide reverses pulmonary vaso­constriction in the hypoxic and acidotic newborn lamb. Circulation Research 72:246-254

8. Rossaint R. Falke KJ, Lopez F et al (1993) Inhaled nitric oxide for the adult respiratory dis­tress syndrome. N Engl J Med 328:399-405

9. Gerlach H, Pappert 0, Lewandowski K et al (1993) Long term inhalation with evaluated low doses of nitric oxide for selective improvement of oxygenation in patients with adult respira­tory distress syndrome. Intensive Care Medicine 19:443-449

10. Wysocki M, Vignon P, Roupie E et al (1993) Improvement in right ventricular function with inhaled nitric oxide in patients with the adult respiratory distress syndrome (ARDS) and per­missive hypercapnia. Am Rev Respir Dis 147:A350

11. Dellinger RP, Zimmerman JL, Taylor RW et al (1998) Effects of inhaled nitric oxide in pa­tients with acute respiratory distress syndrome: Results of a randomized phase II trial: Crit Care Med 26: 15-23

12. Wysocki M, Delclaux C, Roupie E et al (1994) Additive effect on gas exchange of inhaled ni­tric oxide and intravenous almitrine bismesylate in the adult respiratory distress syndrome. Inten Care Med 20:254-259

13. Roberts JD, Polaner OM, Lang P, Zapol WM (1992) Inhaled nitric oxide in persistent pul­monary hypertension of the newborn. Lancet 340:818-819

14. Kinsella JP, Shaffer E, Neish SR, Abman SH (1992) Low-dose inhalational nitric oxide in persistent pulmonary hypertension of the newborn. Lancet 340:8819-8820

15. Roberts JD Jr, Lang P, Bigatello L et al (1993) Inhaled nitric oxide in congenital heart dis­ease. Circulation 87:447-453

16. Roberts JD Jr, Fineman JR, Morin FC et al (1997) Inhaled nitric oxide and persistent pul­monary hypertension ofthe newborn. N Engl J Med 336(9):605-610

17. The Neonatal Inhaled Nitric Oxide Study Group (1997) Inhaled nitric oxide in full-term and nearly full-term infants with hypoxic respiratory failure. N Engl J Med 336(9):597-604

18. Garg UC, Hassid A (1989) Nitric oxide-generating vasodilators and 8-bromo-cyclic guano­sine monophosphate inhibit mitogenesis and proliferation of cultured rat vascular smooth muscle cells. Journal of Clinical Investigation 83: 1774-1777

19. Roberts JD, Roberts CT, Jones RC, Zapol WM (1995) Continuous nitric oxide inhalation re­duces pulmonary arterial structural changes, right ventricular hypertrophy, and growth retar­dation in the hypoxic newborn rat. Circ Res 76:215-222

52 W.M. Zapol, R. Jenney

20. Kouyoumdjian C, Adnot S, Levame M et al (1994) Continuous inhalation of nitric oxide pro­tects against development of pulmonary hypertension in chronically hypoxic rats. Journal of Clinical Investigation 94:578-584

21. Clark RH, Kueser TJ, Walker MW et al for the Clinical Inhaled Nitric Oxide Research Group (2000) Low-dose nitric oxide therapy for persistent pulmonary hypertension of the newborn. N Engl J Med 342:469-474

22. Steudel W, Hurford WE, Zapol WM (1999) Inhaled Nitric Oxide: Basic Biology and Clinical Applications. Anesthesiology 91: 1090-1121

Protecting Renal Blood Flow in the Intensive Care Unit

l.A. KELLUM

Acute renal failure in the intensive care unit Acute renal failure (ARF) in the intensive care unit (ICU) represents a different spectrum of disease compared to ARF occurring outside the ICU. As much as 95% of ARF in the ICU is secondary to acute tubular necrosis (ATN). Incidence and mortality rates for ARF in and outside the ICU are quite different [1-3]. For example, the incidence of ARF is about 5% ou~side the ICU and mortality rates are usually < 30%. However, in the ICU, the incidence can be as high as 15% with a mortality rate between 50% and 90%. Severe ARF (defined as requiring dialysis) rarely occurs in isolation, and most often occurs in association with multiple organ failure (MOF) [4]. Ischemia, principally of the renal medulla, is estimated to contribute to 85% of cases of ARF [5], and multiple causes of medullary ischemia have been identified [6]. Most ARF occurs with multiple in­sults. Common conditions causing or exacerbating medullary ischemia are shown in Table 1. Thus, it seems reasonable that preserving renal blood flow (RBF) should be seen as an imperative for the intensivist. Unfortunately, this goal is easier to espouse than to achieve, and increasing RBF may not always be beneficial.

Table 1. Common conditions causing or exacerbating medullary ischaemia

Inadequate renal perfusion pressure Hypovolemia Hypoxia Radiocontrast dyes Medications NSAIDs

Amphotericin Cyclosporin

Myoglobulin Sepsis

Systemic hemodynamic effects Direct renal vascular effects PMN-endothelium interactions and medullar congestion

NSAIDS non-steroidal anti-inflammatory drugs, PMN polymorphonuclear leukocytes

54 J.A. Kellum

Renal blood flow There are several unique aspects of the renal vasculature. First, unlike most oth­er organs, RBF serves not to satisfy the metabolic demands of the tissue but rather the needs of filtration. As illustrated in Figure 1, total RBF well exceeds the metabolic needs of the kidney. Roughly 20% of plasma flow is filtered into Bowman's capsule yielding a glomerular filtration rate (GFR) of approximately 120 mIJmin. Changes in renal hemodynamics are mediated both at the afferent and efferent arterioles. Renal vascular resistance is 40-60% afferent and 30% ef­ferent in origin. Changes in systemic hemodynamics influence renal vascular re­sistance in predictable ways. Systemic arterial hypotension leads to decreased afferent arteriolar tone and increased efferent arteriolar tone. These effects are mediated by prostaglandins E2 and 12 and angiotensin II, as well as by the myo­genic response. The balance of afferent and efferent tone determines the hy­draulic pressure in the glomerulus and this, in tum, largely determines GFR. Obviously then, an agent that decreases both afferent and efferent arteriolar tone in parallel would have the effect of increasing RBF without increased (or per­haps even decreasing) GFR. Conversely, an agent that increases both afferent and efferent arteriolar tone would decrease RBF while maintaining GFR. For example vasopressin causes both afferent and efferent vasoconstriction and thus does not effect GFR. Thus, GFR cannot be used as a marker of adequacy of REF, except to the extent that REF is adequate for filtration. As Figure 1 sug­gests, oxygen delivery to the renal medulla may be compromised at levels of RBF sufficient to maintain GFR.

The second important feature of the renal vasculature is the relatively high autoregulatory threshold and the extent to which autoregulation is "fragile". All vascular beds in the body have some degree of autoregulation, such that changes in perfusion pressure are not associated with changes in blood flow. The pres­sure at the lower limit of the autoregulatory zone (or plateau) is called the au­toregulatory threshold. For most mammalian kidneys, the autoregulatory thresh­old is set near a mean arterial pressure of 80 mmHg. When pressure drops be­low this level there will be an associated decrease in RBF. In addition, there is evidence that renal vascular autoregulation is easily lost. In ischemic ARF, renal autoregulation vanishes and remains gone for months following recovery [7, 8]. Furthermore, even small doses of endotoxin are capable of disrupting renal au­toregulation, even at doses that do not produce renal injury or shock [9]. In this way, blood pressure changes always influence RBF in the injured and, even in some cases, the non-injured kidney.

Finally, the renal vasculature is unique in the way it responds to adenosine. For most vascular beds, adenosine is a safety net. When blood flow is compro­mised to the extent that substrate and/or oxygen delivery falls below critical lev­els, ATP is hydralyzed completely to adenosine which causes local vasodilation and increased blood flow. In the kidney, adenosine works differently. Local gen­eration of adenosine occurs from the macula densa in response to increased Na-

Protecting Renal Blood Flow in the Intensive Care Unit

Cortex

Outer medulla

Inner medulla

proximal tubule pars recta

medullary th ick ascending limb

55

p02 (mm Hg)

70

50

30

10

Fig. 1. Renal oxygen delivery. Schematic representation of a nephron. Note the site of lowest in­trarenal oxygen P02 and the important structures located there. Adapted from Heyman et al. III]

CI concentration in the distal tubular fluid . Adenosine is then released into the juxtaglomerular apparatus causing preglomerular vasoconstriction and de­creased RBF [10]. Renin release by the juxtaglomerular apparatus is also de­creased by adenosine. These effects are collectively known as the tubu­loglomerular feedback mechanism. The primary purpose of tubuloglomerular feedback is thought to be the regulation of electrolyte reabsorption. The effect keeps fluid and electrolyte delivery to the distal nephron within certain limits. Since roughly 10% of the fluid and electrolytes filtered enter the distal tubule and since hormonal regulation primarily effects fractional reabsorption, a rela­tively constant load to the distal tubule is necessary in order to assure that hor­monal regulation can be effective in controlling electrolyte reabsorption. How­ever, another result of tubuloglomerular feedback is that medullary oxygen con­sumption is regulated.

Increased delivery of solute to the distal tubule increases medullary oxygen consumption. As seen in Figure I, the medulla is always on the brink of dysox­ia. Thus, by regulating solute delivery to the distal tubule, tubuloglomerular feedback works as a form of autoprotection.

56

Clinical conditions

Radiocontrast-induced ARF

J.A. Kellum

Administration of radiocontrast dye is one of the most-common anticipated eti­ologies of ATN. This complication is rare in patients without underlying renal, cardiac, or hepatic dysfunction, and is most common in patients with diabetic nephropathy. The incidence of radiocontrast-induced ARF approaches 50% in this group, depending on how it is defined, the degree of underlying renal dys­function, and the use of ionic versus non-ionic contrast. The mechanism of tubu­lar injury secondary to radio contrast is not entirely understood, but appears to be due to medullary ischemia. However, while initial descriptions emphasized that RBF decreases with exposure to radiocontrast, more-recent work suggests that the vasoconstriction phase occurs after tubular injury and RBF actually in­creases early in the course of radiocontrast-induced ARF [11]. As outer medullary blood flow increases, medullary oxygen requirements increase and hypoxia is worsened due to increased solute delivery to the distal tubule. Thus, insuring or even increasing RBF does not insure medullary oxygen delivery and may not be protective. In fact, to the extent that increasing Na delivery increases medullary cell work, increasing RBF may even be deleterious if it does not in­crease oxygen delivery in a proportional manner.

Sepsis Although sepsis affects the kidney in a variety of ways (Table 1), the effects on systemic hemodynamics are among the most important. Vasodilatation and arte­rial hypotension jeopardize renal perfusion. Again the region of the kidney most at risk is the outer medulla. In this setting, the clinician is faced with the diffi­cult position of maintaining mean arterial pressure and at the same time avoid­ing vasoconstriction of the renal vasculature. Most of the vasopressor agents at our disposal have the potential to decrease RBF and might therefore aggravate medullary ischemia. One agent, norepinephrine, has been evaluated extensively for its effect on renal perfusion. In animals, high-dose norepinephrine (1 Ilg/kg per min) infused directly into the renal artery induces a marked decrease in RBF and subsequent ARF [12]. In contrast, when the drug is given systemically, a decrease in RBF is seen only when mean arterial pressure is well above the au­toregulatory threshold. Furthermore, there has been noted subsequent "auto-reg­ulatory escape" in this setting [13]. Interestingly, medullary blood flow redistri­bution toward the outer medulla has also been shown with norepinephrine infu­sion in animals [14]. In humans, less data are available. Normotensive volun­teers exhibited a decrease in RBF (although not GFR) with moderate doses of norepinephrine, and this effect was reversed by dopamine [15]. However, as the authors point out, these effects may be different in hypotensive patients with systemic vasodilatation. Indeed, patients with hyperdynamic septic shock had increased creatinine clearance and urine output when blood pressure was re­stored with norepinephrine [16, 17].

Protecting Renal Blood Flow in the Intensive Care Unit 57

However, creatinine clearance and urine output during treatment do not in­sure RBF. As long as afferent and efferent arteriolar tone are regulated in con­cert, renal hemodynamics, and thus GFR, may be preserved despite a fall in RBF. However, following ischemic ARF in the rat, RBF was unchanged by in­fusion of norepinephrine into the renal artery [18]. Similarly, in hypodynamic/ hypotensive animals, no adverse renal effects were seen when norepinephrine was used to restore arterial blood pressure [19, 20]. Furthermore, endotoxemic dogs given norepinephrine not only exhibited improved renal perfusion second­ary to increased systemic arterial pressure, but actually had increased RBF that could not be attributed to arterial pressure alone [21]. Presumably, this increase was due to redistribution of blood flow away from other vascular beds, such as skin and adipose tissue. Whatever the mechanism, norepinephrine appears to be associated with improved renal hemodynamics when given to treat shock in pa­tients and animals with sepsis and sepsis-like conditions.

Dopamine and RBF

Dopamine may increase RBF through a variety of mechanisms. First, dopamine may improve renal perfusion pressure by its actions on a- and ~-adrenergic re­ceptors. Second, particularly at "low doses", dopamine directly vasodilates af­ferent arterioles through stimulation of DA-1 receptors. However, "high-dose" dopamine is less effective than norepineprhine in increasing arterial pressure (and hence renal perfusion pressure) in patients with shock [16, 17]. The use of so-called low-dose or renal-dose dopamine has become routine in many IeUs, despite sound rationale to avoid it. First, there is no such thing as "renal-dose". The action of a drug is dependent on the concentration of drug at the site of ac­tion and for dopamine, across different critically ill patients, there is no relation­ship between plasma dopamine levels and the dosage infused [22]. Individual patients may experience pressor effects (a- and/or ~-adrenergic effects) of dopamine at doses thought to stimulate only dopamine receptors. There is no evidence that low-dose dopamine is beneficial in patients with, or at risk for, re­nal dysfunction. In a systematic review of 19 studies, benefit was defined as a positive effect on renal function persisting after the intervention, a decrease in mortality, or a decrease in requirement for hemodialysis. Among these trials on­ly 2 were positive, both were methodologically inferior, and the remaining 17 studies, collectively enrolling over 700 patients, were all negative [23]. More of­ten, clinicians use "religion-dose" dopamine in that, despite the lack of evi­dence, we somehow have faith that dopamine is doing something good for our patients. This faith is largely driven by the observation that urine output increas­es with dopamine. Indeed urine output increases with dopamine in part because this drug inhibits the Na-KATPase at the tubular epithelial cell level, resulting in natriuresis [24]. Thus, apart from any other effect, dopamine increases urine output because it is a diuretic!

58 J.A. Kellum

There are several potential problems with the use of "diuretic-dose" dopamine. First, diuretics may confound the assessment of renal perfusion and volume status and may lead to volume depletion. Furthermore, although dopamine increases RBF, it may not increase medullary oxygenation [25] and, in fact, results in increased oxygen utilization by increasing solute delivery to the distal tubule [26]. Thus, one would expect that dopamine would worsen re­nal injury during conditions in which medullary oxygenation is already at risk. This is exactly what was seen by Wiesberg et al. [27] when they studied the ef­fect of dopamine at 2 Ilg/kg per min in contrast-induced ARF in patients with diabetes. Renal injury was increased despite increased RBF. In addition, dopamine inhibits tubuloglomerular feedback [28], a valuable renal defense mechanism. In addition to these risks, low-dose dopamine may result in bowel mucosal ischaemia, digital necrosis, immune suppression, atrial and ventricular arrhythmias and hypo-pituitarism.

Alternative methods of renal protection

Pharmacological agents A variety of pharmacologic agents have been studied for their potential to pre­serve renal function in various clinical conditions (Table 2). To date, these agents have been extremely disappointing in clinical trials [29]. The only agents that have been shown to be of potential value are the calcium channel antago­nists in the setting of renal transplantation. These agents have now been shown to improve allograft function by reducing ischemia-reperfusion injury [30]. Fur­thermore, small randomized trials have demonstrated that calcium antagonists can reverse cyclosporin-mediated renal vasoconstriction and confer protection

Table 2. Pharmacological agents evaluated for renal protective effects

Dopaminergic agents Dopamine Dopexamine Fenolapam

Atrial natriuretic peptide analogues Atrial natriuretic peptide Anaritide Urodilatin

Adenosine antagonists Theophylline Pentoxifylline Rolipram

Calcium antagonists Nifedipine Diltiazem

Protecting Renal Blood Flow in the Intensive Care Unit 59

in cadaver kidney transplant recipients treated with cyclosporin [31, 32]. How­ever, calcium channel antagonists also interfere with the metabolism of cy­closporin and tacrolimus and therefore make dosing more difficult. This prob­lem can be easily avoided if calcium channel antagonists are started immediate­ly following transplant while immunosuppression is still being titrated.

With the exception of calcium channel antagonists, the remaining agents list­ed in Table 2 have been unsuccessful. One reason may be that global RBF may not be the problem that leads to ischemic renal injury. Indeed, dopamine antag­onists are not nephrotoxic even in the critically ill. For example, metoclo­pramide (10 mg) produces renal vasoconstriction and decreases RBF in normal humans [33], but is not associated with any renal toxicity. Similarly, low doses of metoclopramide and haloperidol completely abolish dopamine-mediated in­creases in RBF [34].

Non-pharmacological approaches to renal protection

Fluid loading is perhaps the most-effective means at our disposal to limit renal injury. Although the evidence is limited (no randomized trials), fluid loading has been associated with dramatic benefits in certain groups compared with histori­cal controls (e.g., traumatic rhabdomyolysis). However, given the strong physio­logical justification and little potential for harm (recommended volumes are 1-2 1 at rates of < 500 mllh) , fluid loading to ensure adequate circulating volume should be undertaken in all high-risk patients.

The second approach to reducing renal injury is to reduce the risk from nephrotoxins. Several approaches have proven valuable in this regard. First, sin­gle rather than multiple daily doses of aminoglycosides should be used. Al­though this approach has been evaluated by one meta-analysis (not limited to leu patients) and found to produce no difference in the risk of ARF (relative risk 0.78, 95% confidence interval 0.31-1.94), the review also concluded that single daily dosing was as effective in treating infection and not associated with any other risks [35]. One high-quality randomized trial (n = 95), in which renal toxicity was defined as an increase in serum creatinine> 0.5 mg/dl, showed that renal toxicity occurred in 24% of patients on multiple daily dosing versus only 5% in patients on single daily dosing (relative risk 4.89, 95% confidence inter­val 1.15-20.74) [36]. In a similar vein, lipid complex amphotericin holds prom­ise as a much-safer, although considerably more-expensive, approach to treating fungal infections in patients with or at risk for renal insufficiency. Although there have yet to be any direct comparative randomized trials, one phase II trial of 556 patients, in which renal toxicity was defined as any increase in serum creatinine, showed a 24% incidence of renal toxicity with lipid complex ampho­tericin [37] compared with 60-80% incidence with standard preparations. In ad­dition, patients with pre-existing renal dysfunction on standard amphotericin had a reduction in serum creatinine with lipid complex [37].

60 J.A. Kellum

Finally, low osmolality radiocontrast media (LOCM) should be used in high­risk patients. A systematic review of 31 randomized trials (n = 5,146) [38] con­cluded that LOCM is associated with less increase in serum creatinine overall, but no change in the occurrence of ARF or need for dialysis (rare events). How­ever, in patients with underlying renal insufficiency, LOCM significantly re­duced the risk of ARF (odds ratio 0.5, 95% confidence interval 0.36-0.70). Al­though LOCM is expensive, it is still likely to be cost effective in high-risk groups (diabetic nephropathy), but this has not yet been formally established.

Conclusions Although ARF in the ICU is usually a manifestation of medullary ischemia, it is important to understand that there are multiple mechanisms responsible for this injury. First, there are "supply side" reasons for medullary ischemia, including systemic arterial hypotension and, less commonly, selective renal vasoconstric­tion. However, just as important are the "demand side" mechanisms, which in­clude increased solute delivery to the distal tubule. ARF in the critically ill car­ries a very high mortality and current treatment is disappointing. Prevention of ARF in the ICU involves avoidance of nephrotoxins, maintenance of adequate circulating volume and arterial pressure. Selective manipulation of RBF by va­soactive agents has not been shown to be of benefit and may be harmful.

References 1. Hou SH, Bushinsky DA, Wish JB et al (1983) Hospital-acquired renal insufficiency: a

prospective study. Am J Med 74:243-248 2. Brivet FG, Kleinknecht DJ, Loirat P et al (1996) Acute renal failure in intensive care units­

causes, outcome, and prognostic factors of hospital mortality; a prospective, multicenter study. French Study Group on Acute Renal Failure. Crit Care Med 24: 192-198

3. Liano F, Junco E, Pascual J et al (1998) The spectrum of acute renal failure in the intensive care unit compared with that seen in other settings. The Madrid Acute Renal Failure Study Group. Kidney Int [Suppl]66:S16-S24

4. Tran DD, Oe PL, de Fijter CW et al (1993) Acute renal failure in patients with acute pancre­atitis: prevalence, risk factors, and outcome. Nephrol Dial Transplant 8: 1079-1084

5. Thadhani R, Pascual M, Bonventre N (1996) Acute renal failure. N Engl J Med 334:1448-1460

6. Brezis M, Rosen S (1995) Hypoxia of the renal medulla - its implications for disease. N Engl J Med 332:647-655

7. Adams PL, Adams FF, Bell PD et al (1980) Impaired renal blood flow autoregulation in is­chemic acute renal failure. Kidney Int 18:68-76

8. Kelleher SP, Robinette JB, Conger JD (1984) Sympathetic nervous system in the loss of au­toregulation in acute renal failure. Am J Physiol 246:F379-F386

9. Bersten AD, Holt AW (1995) Vasoactive drugs and the importance of renal perfusion pres­sure. New Horiz 3:650-661

Protecting Renal Blood Flow in the Intensive Care Unit 61

10. Osswald H, Muhlbauer B, Schenk F (1991) Adenosine mediates tubuloglomerular feedback response: an element of metabolic control of kidney function. Kidney Int [Suppl]32:S128-S131

11. Heyman SN, Fuchs S, Brezis M (1995) The role of medullary ischemia in acute renal failure. New Horiz 3:597-607

12. Patak RV, Fadem SZ, Lifschitz MD et al (1979) Study of factors which modify the develop­ment of norepinephrine-induced acute renal failure in the dog. Kidney Int 15:227-237

13. Banks RO (1988) Vasoconstrictor-induced changes in renal blood flow: role of prostaglandins and histamine. Am J Physiol 254:F470-F476

14. Yang S, Silldorff EP, Pallone TL (1995) Effect of norepinephrine and acetylcholine on outer medullary descending vasa recta. Am J PhysioI269:H71O-H716

15. Hoogenberg K. Smit AJ, Girbes AR (1998) Effects of low-dose dopamine on renal and sys­temic hemodynamics during incremental norepinephrine infusion in healthy volunteers Crit Care Med 26:260-265

16. Martin C, Papazian L, Perrin G et al (1993) Norepinephrine or dopamine for the treatment of hyperdynamic septic shock? Chest 103: 1826-1831

17. Redl-Wenzl EM. Armbruster C, Edelmann G et al (1993) The effects of norepinephrine on hemodynamics and renal function in severe septic shock states. Intensive Care Medicine 19:151-154

18. Conger JD. Robinette JB, Hammond WS (1991) Differences in vascular reactivity in models of ischemic acute renal failure. Kidney Int 39: 1087-1097

19. Breslow MJ, Miller CF, Parker SD et al (1987) Effect of vasopressors on organ blood flow during endotoxin shock in pigs. Am J PhysioI252:H291-H300

20. Hussain SNA. Rutledge F, Roussos C et al (1998) Effects of norepinephrine and fluid admin­istration on the selective blood flow distribution in endotoxic shock. J Crit Care 3:32-42

21. Bellomo R, Kellum JA, Wisniewski SR et al (1999) Effects of norepinephrine on the renal vasculature in normal and endotoxemic dogs. Am J Respir Crit Care Med 159: 1186-1192

22. Zaritsky A, Lotze A, Stull R et al (1988) Steady-state dopamine clearance in critically ill in­fants and children. Crit Care Med 16:217-220

23. Kellum JA (1997) The use of diuretics and dopamine in acute renal failure: a systematic re­view of the evidence. Crit Care Med 1:53-59

24. Seri I, Kone BC, Gullans SR et al (1988) Locally formed dopamine inhibits Na+-K+-ATPase activity in rat renal cortical tubule cells. Am J Physiol 255:F666-F673

25. Heyman SN, Kaminski N, Brezis M (1995) Dopamine increases renal medullary blood flow without improving regional hypoxia. Exp NephroI3:331-337

26. Olsen NV, Hansen JM, Ladefoged SD et al (1990) Renal tubular reabsorption of sodium and water during infusion of low-dose dopamine in normal man. Clin Sci (Colch) 78:503-507

27. Weisberg LS, Kumik PB, Kumik BR (1994) Risk of radiocontrast nephropathy in patients with and without diabetes mellitus. Kidney Int 45:259-265

28. Schnermann J, Todd KM, Briggs JP (1990) Effect of dopamine on the tubuloglomerular feed­back mechanism. Am J Physiol 258:F790-F798

29. Dishart MK, Kellum JA (2000) An evaluation of pharmacological strategies for the preven­tion and treatment of acute renal failure. Drugs 59:79-91

30. Neumayer HH, Wagner K (1987) Prevention of delayed graft function in cadaver kidney transplants by diltiazem: outcome of two prospective, randomized clinical trials. J Cardiovasc PharmacollO[Suppl]IO:S170-S177

31. Dawidson I, Rooth P, Fry WR et al (1989) Prevention of acute cyclosporine-induced renal blood flow inhibition and improved immunosuppression with verapamil. Transplantation 48:575-580

32. Neumayer HH, Kunzendorf U, Schreiber M (1992) Protective effects of diltiazem and the prostazycline analogue iloprost in human renal transplantation. Ren Fail 14:289-296

33. Manara AR, Bolsin S, Monk CR et al (1991) Metoclopramide and renal vascular resistance. Br J Anaesth 66: 129-130

62 J.A. Kellum

34. Felder RA, Blecher M, Calcagno PL et al (1984) Dopamine receptors in the proximal tubule of the rabbit. Am J PhysioI247:F499-F505

35. Hatala R, Dinh IT, Cook OJ (1997) Single daily dosing of aminoglycosides in immunocom­promised adults: a systematic review. Clin Infect Dis 24:810-815

36. Prins JM, Buller HR, Kuijper EJ et al (1993) Once versus thrice daily gentamicin in patients with serious infections. Lancet 341:335-339

37. Walsh TJ, Hiemenz JW, Seibel NL et al (1998) Amphotericin B lipid complex for invasive fungal infections: analysis of safety and efficacy in 556 cases. Clin Infect Dis 26: 1383-1396

38. Barrett BJ, Carlisle EJ (1993) Metaanalysis of the relative nephrotoxicity of high- and low-os­molality iodinated contrast media. Radiology 188: 171-178

Bad and Good News in Pathophysiology, Prevention, and Management of Sepsis

R. P. DELLINGER

Pathophysiology

Bad news

Complexity and duplication of cytokine cascade

To date, models that describe the interactions of inflammatory mediators during severe sepsis are dominated by a single mediator (e.g. IL-I or TNF) and often fail to account for the redundancy and interdependence of inflammatory re­sponses. The net response elicited by a cytokine on a target cell is not an isolat­ed event, but depends on the state of cell-associated cytokine receptors (which may be up or down-regulated) and the presence of other pro-inflammatory and counter-regulatory molecules. Ascribing qualifiers of benefit or harm to any in­flammatory mediator (e.g. IL-I or TNR) depends on the clinical context and the net effects of change over time during an infection.

The complexity of the host defence system suggests that it is unlikely that anyone agent will be effective in the majority of patients with severe sepsis. The development of a multiple modality approach directed at various steps, in­cluding components of immune modulation and immune enhancement, has the­oretical appeal to correct abnormalities or imbalances in the host inflammatory response. For example, combination immunotherapy has been shown to be more efficacious than single anti-mediator therapy in a rat model of Pseudomonas sepsis [1]. The limitations in applying this approach to clinical sepsis are identi­cal to those of single agent anti-mediator therapy. The clinician has little infor­mation available to assess what aspect of the host response is abnormal and im­balanced in order to direct the timing, dose, duration, and the specific compo­nents of these single or combination therapies.

Lack of total understanding of the interaction of pro- and anti-inflammatory cytokines

The body has great capability to counter a pro-inflammatory response with a counter-regulating anti-inflammatory response. It is possible that a significant number of patients initially diagnosed with severe sepsis and pro-inflammatory response may transition into a period of predominate anti-inflammatory state [2,

64 R. P. Dellinger

3]. While current therapies for severe sepsis are focused on selective down-regu­lation of immune response, emerging data suggest that many patients are al­ready immunosuppressed or may become functionally immunosuppressed due to compensatory down-regulation of the inflammatory response. Thus, an alter­native strategy in such patients would be to use agents that enhance or restore immune responses of severely infected patients.

Good news

Potential use of genetic profiling

An increasing understanding of the genetic variability from patient to patient as it relates to cytokine response to initiations of the inflammatory response may prove useful in patient selection. HLA subtype has been demonstrated to corre­late with intensity of in vitro macrophage release of inflammatory cytokines in response to endotoxin. More recently, the TNF2 polymorphism on the TNF-a promoter region (upstream of the start codon of the TNF-a gene) was present in 39% of 89 septic shock patients but in only 18% of healthy controls [4]. Among patients with septic shock, non-survivors were significantly more likely to have at least one polymorphism (61 % versus 34%). In addition, several studies indi­cate significant genetic link to cytokine production and outcome in life-threaten­ing meningococcal disease [5,6].

Working toward better definitions

Why has the standardization of definitions not improved our capability to carry out successful trials of innovative immunologic therapy in severe sepsis? Two reasons are clearly identified [7].

First is our realization that success of immunotherapy is likely equally de­pendent on the timing of therapy, the activity of the drug and the use of a single agent in such a complex biologic environment.

The second, and perhaps the one that will be ultimately easier to solve is that any clinical classification of patients with sepsis does not reliably predict media­tor milieu which would seem to be most important for any therapeutic effect of immunologic therapy. Therefore, the search for mediator profiles that predict re­sponse to immunologic therapy is paramount.

Clinical profiles are unlikely to correlate well with underlying biologic pro­files. The identification of patients who are likely to benefit from specific im­muno-modulation based purely on clinical information is unlikely. What is like­ly needed is a biologic profile that matches with the therapy to be administered. The need to ascertain the overall effect of the biologic profile by measuring sta­tus of the target cell (turned on-turned off) is also likely needed.

There is intense interest in establishing relationships among biologic markers of sepsis and stages of sepsis. Elevations of IL-l, IL-6, IL-8, IL-lO, TNF-a and

Bad and Good News in Pathophysiology, Prevention, and Management of Sepsis 65

TNF soluble receptor have been demonstrated in sepsis [8-10]. Profiles of these pro-inflammatory and anti-inflammatory mediators may prove to have treatment pertinence [11].

Prevention

Bad news

Physician compliance/cooperation in basic proven techniques to prevent infec­tion and decrease incidence of resistant organisms remains poor. The classic ex­ample of this is physician non-compliance with hand washing before and after patient contact. The prevention of infection with resistant organisms is also handicapped by physician practices which include prescribing antibiotics for longer than standard/recommended duration, prescribing antibiotics empirically without a definite diagnosis or bacterial identification, and prescribing board spectrum antibiotics when equally effective narrow spectrum antibiotics are available.

Good news

Prevention of pneumonia in mechanically ventilated patients

It is becoming clearer that simple inexpensive techniques in the ICU wi11likely decrease the incidence of pneumonia in the mechanically ventilated patient [12]. Semirecumbent position of patients has been demonstrated to decrease as­piration risk. Reducing gastric volumes by limiting narcotic/analgesic agents, monitoring gastric volumes, use of pro-motility agents, and post-pyloric place­ment of feeding tubes are expected to reduce the incidence of ventilator-ac­quired pneumonia. Oral intubation is preferred over nasotracheal intubation. Accumulated condensate in ventilation tubing should be regularly removed. The use of granulocyte colony stimulating factor for the febrile, neutropenic, me­chanically ventilated patient is now well accepted. Avoidance of intubation with proper use of noninvasive positive pressure ventilation in some patient popula­tions, and early identification of patients ready for extubation may also decrease the incidence of hospital-acquired pneumonia. Other techniques such as supra­glottic suctioning and chlorhexidine oral rinse may also have a role in the pre­vention of pneumonia.

Prevention of line sepsis

Infection associated with central lines can result in serious medical conse­quences and expense. Prospective, randomized trials of CVP catheters coated with minocycline and rifampin or chlorhexidine and silver sulfadiazine have been associated with reduced rates of catheter colonization and blood stream in-

66 R. P. Dellinger

fections compared to unimpregnated catheters. A study by Darouiche and col­leagues published in January 1999 compared two antimicrobial-impregnated central venous catheters. A study of 865 inserted catheters was performed. Catheters impregnated with minocycline and rifampin were one-third as likely to be colonized as catheters impregnated with chlorhexidine and silver sulfadi­azine. More importantly, the comparison of blood stream infections was 7.9% versus 22.8% [13].

Early detection of sepsis Although endotoxemia is a conceptually attractive marker of invasive Gram­negative infection or altered gastrointestinal permeability in the critically ill pa­tient, the assay of endotoxin in biological fluids has been notoriously difficult and unreliable. The most widely used assay is the limulus assay that has proven to be unreliable in biologic fluids. A recently devised alternate strategy for de­tecting endotoxin in biological fluids is the technique of neutrophil chemilumi­nescence with specific monoclonal antibody against endotoxin. The presence of a common polysaccharide core has made it possible to generate antibodies with specificity for endotoxin, and a broad range of cross reactivities for multiple en­dotoxin strains. A complex of endotoxin and anti-endotoxin antibody primes the patient's cells for an enhanced respiratory burst in response to zymosan. Respi­ratory burst activity is readily detected using the lumiphor, luminol. The magni­tude of the priming influence is directly proportional to the concentration of antigen-antibody complex: in the presence of antibody excess, this is dependent upon the concentration of endotoxin. This technique appears to be both sensitive and specific, and to reliably detect endotoxin from a wide strain of Gram-nega­tive bacteria in spiked samples of whole blood. The assay can be performed us­ing 10-20 microlitres of whole blood and results are obtained within 30 minutes. Preliminary studies with the chemiluminescent assay have shown that endotoxin at a concentration of more than 50 pglml (0.04 EU/ml) can be detected in more than half of patients admitted to a mixed medical-surgical ICU [14]. It is invari­ably present in patients with Gram-negative infections, and elevated levels are associated with a prolonged ICU stay.

In recent trials, C-reactive protein (CRP) and procalcitonin (PCT) have been used as markers of sepsis [15-17]. High levels of CRP have been shown to dif­ferentiate bacterial infections from viral infections [18]. Decreasing CRP levels predict the recovery phase of sepsis [19, 20]. Although reported more sensitive than fever or leukocytosis in identifying sepsis, high levels of CRP have also been documented in trauma and surgery, especially cardiac surgery. Ugarte et al. [21] found that the combination of CRP and PCT was very specific for infec­tion. However in that study CRP demonstrated greater sensitivity and specificity than PCT for infection. Both PCT and CRP are not 100% sensitive or specific. Normal values are probably most useful in defining that infection is unlikely. The ability of these markers to facilitate identification and classification of sep­sis is not clear.

Bad and Good News in Pathophysiology, Prevention, and Management of Sepsis

Management

Bad news

Resistant bacteria

67

Critical care areas are a prominent breeding ground for the multiple antimicro­bial-resistant bacterial strains that have markedly reduced the antimicrobial op­tions available to the intensivist [22]. An increasing prevalence of methicillin­resistant staphylococcus, vancomycin-resistant enterococcus, and ceftazidime­or imipenem-resistant Pseudomonas or Enterobacter are now found in the in­tensive care units. An increasing number of centres are reporting S. aureus strains with reduced vancomycin sensitivity. The seemingly inevitable appear­ance of high level glycopeptide resistant S. aureus will have enormous impact on ICU care.

Innovative therapy trial results published thus far

Over the last 12 years there has been great interest in the potential of innovative therapy for severe sepsis. Many products targeting bacterial toxins and media­tors generated in response to these toxins have reached phase II clinical trials. All clinical trials to date have failed to demonstrate clinical efficacy (Table O. The most likely reasons for trial failure are [23]:

1. Animal study problems. Failure of clinical trials may be related to the poor predictability of severe sepsis animal models. Studies with small animals (mice, rats, rabbits) are most economical but may not adequately reproduce human sepsis. Studies with primates are very expensive. No animal model, rodent or primate, reproduces the complex interactions of human sepsis. Ani­mals used are previously healthy while most humans with sepsis are not. Agents under investigation are often administered to animals before or im­mediately after the septic challenge, conditions that can rarely be achieved in clinical trials. Non-blinded studies are common in animal research and may introduce bias. Ideally, animal studies should be blinded and randomized.

2. Entering the right patient in the study based on biologic profile. The ideal patient population for a certain therapy should have a biologic profile that matches the need for the therapy being tested. Most human studies have not targeted a specific biologic profile known to be present in the enrolled pa­tients. This problem is further handicapped by the lack of rapid turnaround assays to identify patients' biologic status.

3. Dosing and timing of therapy. As alluded to earlier, agents are typically shown to benefit in animals when given immediately before, at the same time, or shortly after toxin challenge. It is rare in humans to be able to in­stitute therapy immediately after what is judged to be the onset of severe sepsis. In addition, the optimum relation of dosing to onset of symptomolo­gy is undefined. There may be a window of opportunity for successful anti­inflammatory therapy that is typically missed or perhaps can never be en-

68 R. P. Dellinger

Table 1. Major phase II/Ill clinical trials of anti-inflammatory therapy

Therapy Sponsor Number of patients

Human anti-lipid A MAB Centocor 534 Human anti-lipid A MAB Centocor 2199 Mouse anti-lipid A MAB Zoma 468 Mouse anti-lipid A MAB Zoma 830 Anti-enterobacterial antigen MAB Chiron 826 Interleukin-l receptor antagonist Synergen 893 Interleukin-l receptor antagonist Synergen 906 Anti-bradykinin Cortech 251 and 504 Anti-PAP Ipsen 262 Anti-PAP Ipsen 608 Anti-TNF MAB Celltech 80 Anti-TNF MAB BayerlMiles-Norasept I 971 Anti-TNF MAB BayerlMiles-Intercept 553 Anti-TNFMAB BayerlMiles-Noracept II 1879 Anti-TNF MAB Knoll 122 Soluble TNF-receptor Immunex 141 Soluble TNF-receptor Hoffman/Roche 444 Ibuprofen N/A 455 Soluble TNF-receptor Hoffman/Roche 1340 NOS inhibitor Glaxo-Wellcome 797

NIA, not available; MAB, monoclonal antibody

tered using clinical findings. Furthermore, since there are no readily avail­able measure of biologic activity for which the therapy is targeted, it is dif­ficult to establish dose response curves as well as judge optimal duration of therapy.

4. Entering the right patient population in the study based on likelihood of sur­vival. The ideal patient populations for a study are patients with intermediate risks of mortality and who are likely to survive longer than 24 hours inde­pendent of effect of therapy. Many patients enrolled in studies, as reflected by the placebo popUlation, survive less than 24 hours or do remarkably well based on factors such as site of infection (line-related bacteremias and urosepsis) .

5. Other considerations. Additional reasons why studies may have failed in­clude a) outcome was determined by co-morbid condition and not severe sepsis, b) the patient population was too heterogeneous due to variability in care of sepsis from physician-to-physician, hospital to hospital, and country to country, c) the presence of physiologic benefit but mortality was too in­sensitive an outcome measure, and d) phase III trials were based on post hoc analysis of phase II data.

Bad and Good News in Pathophysiology, Prevention, and Management of Sepsis 69

Good news

Changes in our approach to clinical trials in sepsis Now that the likelihood of innovative therapy producing a major decrease in mortality is recognized, a more realistic estimation of potential decrease in mor­tality has led to studies being powered in the 2000-3000 patient range as op­posed to earlier studies which were powered to detect large differences in mor­tality with studies of 300-500 patients. In addition, the majority of studies have established stricter exclusion criteria so that patients who have been in a state of severe sepsis or septic shock for extended periods of time are not enrolled in the study. It has long been thought that the earlier the intervention, the more likely the success of innovative therapy of sepsis.

Most important is the move to trial design which targets a specific biologic profile. Enrolling patients in clinical trials solely based on clinical criteria is thought to be the major reason previous trials have failed. Three recent trials that suggest benefit from innovative therapy are the MAK-195 anti-TNF anti­body targeted toward septic patients with increased IL-6, the activated protein e study targeted toward the decrease in activated protein e that is known to occur in the majority of patients with severe sepsis and septic shock, and the French steroid trial targeted toward patients with "relative adrenal insufficiency". These trials are reviewed briefly below. - MAK-195. The US study of the Knoll MAK-195 monoclonal anti-TNF anti­

body in severe sepsis was recently reported as a late breaking communication at the American Thoracic Society meeting in Toronto (April 2000). With 2634 patients enrolled, the good news was that the intent to treat analysis for all patients enrolled revealed a 3.6% reduction in 28-day all-cause mortality with a p = .049. The bad news was that the pre-identified, a priori primary analysis group was not intent to treat but those patients with an IL-6 ~ 1000. There were 998 of these patients with a 4% reduction in mortality; however, with the smaller number of patients, the p-value was .058 and did not reach statistical significance. The study plan had called for a logistic regression analysis for confounding variables. Three potential confounding variables that had been identified prior to study commencement included PSAPS, leU type, and country effect. None of these were found to be confounding (de­fined as effecting odds ratio by ~ 10%). The only confounding variable iden­tified was SOFA score survival prediction which affected the odds ratio by 12.6%. With logistic regression adjustment for SOFA score, the p-value was .041 in favor of the MAK 195, with a projected mortality reduction of 6.9%.

- Activated protein C. A press release in July 2000 by Eli Lily reported that the activated protein e study in patients with severe sepsis and septic shock had been prematurely discontinued by the Data and Safety Monitoring Board. This decision to stop the study was based on the Data and Safety Monitoring Board's assessment that the drug showed efficacy. The study was discontin­ued with 1524 evaluable patients. At the time of the study discontinuation,

70 R. P. Dellinger

300 additional patients had been enrolled and that data will also be available for final analysis. The study was powered to 2300 patients to detect a relative mortality reduction of 18%. An early stopping rule in place would suggest that the p-value was at least .01 or less at the interim analysis.

- Steroids for relative adrenal insufficiency in sepsis. The French study evaluat­ing the effect of steroids in early septic shock was a randomized, prospective, double-blinded, placebo-controlled study performed in 16 centres. Patients were randomized to receive either 50 mg hydrocortisone IV q 6 h + 50 Ilg fludrocortisone q.d. or placebo. Patients received 7 days of therapy. Inclusion criteria were site of infection, SIRS, refractory hypotension, mechanical ven­tilation, and organ dysfunction. Enrolled patients had to have an ACTH stim­ulation test within 8 hours of shock and had to be enrolled in the study within 24 hours of onset of shock. The primary analysis was mortality in non-re­sponders to ACTH stimulation test. Non-responders to ACTH stimulation test were defined as failure to increase cortisol level by 10 Ilg/dl or greater. The study is now completed and final data analysis is in progress. The interim analysis was presented at the 5th World Congress on "Trauma, Shock Inflam­mation, and Sepsis" in Munich, Germany in March 2000. The interim analy­sis was performed on 220 patients and showed encouraging results with a to­tal patient population reduction in mortality from 64% to 54% (p-value, .029). For the 167 non-responders, the reduction in mortality was from 65% to 50% with ap-value of .051.

Preventing and treating sepsis induced organ injury

Lung protection strategy

In the late 1990s, four clinical trials were performed internationally in an at­tempt to demonstrate alteration in clinical outcomes in patients with acute lung injury and ARDS when lung protection strategy was utilized. Lung protection strategy was primarily targeted toward avoiding hyperinflation injury, but some studies combined this with minimal PEEP strategy. Three of these studies were negative, and one single institution study produced rather striking results sug­gesting benefit. This later study was not considered as definitive evidence that lung protection strategy altered clinical outcome in ARDS.

On 4 May 2000, the ARDS network (ARDSnet) sponsored by the National Heart Lung and Blood Institute of the National Institutes of Health in the USA published in the New England Journal of Medicine the results of a large multi­centre trial which was designed to study the difference between ventilating ARDS patients with a tidal volume of 6 ml per kilogram (low stretch) versus 12 ml per kilogram (high stretch) [24]. Enrolment criteria were the consensus defi­nition of acute lung injury and the patients could not have met definition of acute lung injury for longer than 36 hours. All patients were ventilated with vol­ume-assist-control. Minimal tidal volume allowed in the low stretch group to

Bad and Good News in Pathophysiology, Prevention, and Management of Sepsis 71

achieve an inspiratory plateau pressure of 30 or less was 4 ml per kilogram. In the high stretch group tidal volume was increased in increments of 1 ml per kilogram to reach an inspiratory plateau pressure of 45 or a tidal volume of 12 ml/kg. In the low stretch group in the presence of severe dyspnoea, the tidal vol­ume could be increased 7-8 ml per kilogram as long as the inspiratory plateau pressure remained 30 cm or less. Inspiratory plateau pressures were measured with a half-second inspiratory pause. The study was stopped after the fourth in­terim analysis at 861 patients with the demonstration of striking benefit in the low tidal volume group (p-value, .007). Mortality was 31 % in the low stretch group and 39.8% in the high stretch group. A comparison of days alive and off mechanical ventilation revealed 12 ± 11 in the low stretch group versus 10 ± 11 in the high stretch group (p = 0.007). The mean tidal volumes on days 1 to 3 were 6.2 ± 0.8 and 11.8 ± 0.8 in the low and high stretch groups and the inspira­tory plateau pressures were 25 ± 6 and 33 ± 8, respectively.

The relative reduction in mortality was 22%. PEEP requirements were higher during the initial part of the study in the low stretch group. There was no differ­ence in barotrauma. Organ failures were reduced in the low tidal volume group, as were plasma interleukin-6 concentrations. The success of this study despite failure of most of the earlier studies of lung protection may be related to:

- a greater difference in inspiratory plateau pressures between the high and low stretch groups compared to previous studies;

- greater statistical power to detect differences;

- treatment of respiratory acidosis associated with the use of permissive hy-percapnia;

- a greater positive end-expiratory pressure that was applied during the early parts of the study in order to allow oxygenation with lower tidal volumes.

Steroid therapy for the fibroproliferative phase of ARDS

Although clinical trials using steroids in the early phase of ARDS failed to show any benefit, some investigators now advocate the use of steroids in later stages of ARDS, the so-called fibroproliferative phase, to decrease progression to fi­brosis. Towards the end of the first week of ARDS non-survivors, when com­pared to survivors, have histological evidence of a more intense inflammatory and fibrotic activity with maladaptive lung repair. Furthermore, in patients with persistent ARDS compatible with an overly exuberant fibrotic response, persist­ent high levels of cytokines can be demonstrated in blood and BAL fluid.

Meduri and colleagues in 1998 published in JAMA the report of a random­ized, double-placebo-controlled trial carried out at four medical centres compar­ing placebo versus steroid therapy for ARDS [25]. Twenty-four patients with se­vere ARDS who had failed to improve lung injury score by the seventh day of respiratory failure were randomized. Randomization was 2 to 1 in favour of steroids. Sixteen patients received steroids and 8 received placebo. The consen­sus definition of ARDS was used for enrolment. The criteria for persistent

72 R. P. Dellinger

ARDS included 7 days of mechanical ventilation with an LIS of 2.5 or greater and less than a 1 point reduction from day 1 point of ARDS. Bronchoscopy was performed to insure that there was no untreated infection. Ventilator manage­ment targeted limited inspiratory plateau pressure to 35 cm or less. Steroid dose was a loading dose of 2 mg/kg followed by 2 mglkg per day until day 14. Be­tween day 15 and 21 the dose was half and half again between day 22 and 28, and between day 31 and 32. Patients extubated prior to day 14 were advanced to day 15 steroid therapy. Patients who failed to improve their lung injury score by at least 1 point after 10 days of treatment were blindly crossed over to the alter­native treatment. At day 10, in the steroid group there was significant reduction in lung injury score, improvement in oxygenation and decreased multiple organ dysfunction syndrome (MODS) score. There was also increased successful ex­tubation and improved mortality with a p-value of .002. None of 16 patients originally randomized to steroids died versus 5 of 8 of those initially random­ized to placebo. Although this trial was greeted with much optimism about pos­sible benefits of steroids, a larger trial is needed to confirm definitive clinical outcome benefit from steroid therapy in late ARDS. This trial is currently un­derway by the National Heart Lung and Blood Institute ARDSnet group.

References 1. Cross AS, Opal SM, Palardy JE et al (1993) The efficacy of combination immunotherapy in

experimental Pseudomonas sepsis. J Infect Dis 167: 112-118 2. Bone RC (1996) Sir Issac Newton, sepsis, SIRS, and CARS. Crit Care Med 24:1125-1128 3. Bone RC (1996) Why sepsis trials fail. JAMA 276:565-566 4. Mira JP, Cariou A, Grall F et al (1999) Association ofTNF2, a TNF-alpha promoter polymor­

phism, with septic shock susceptibility and mortality. JAMA 282:501-568 5. Westendorp RGJ, Hottenga J-J, Slagboom PE (1999) Variation in plasminogen-activator-in­

hibitor-l gene and risk of meningococcal septic shock. Lancet 354:561-563 6. Westendorp RGJ, Langermans JAM, Huizinga TWJ et al (1997) Genetic influence on cy­

tokine production and fatal meningococcal disease. Lancet 349: 170-173 7. Perez J, Dellinger RP (2000) Sepsis definitions. In: Eichacker PQ, Pugin J (eds) Evolving

concepts in sepsis and septic shock. Kluwer Academic, Norwell (in press) 8. Reinhart K, Wiegand-Lohnert C, Grimminger F et al (1996) Assessment of the safety and ef­

ficacy of the monoclonal anti-tumor necrosis factor antibody-fragment, MAK 195F, in pa­tients with sepsis and septic shock: A multicenter, randomized, placebo-controlled, dose­ranging study. Crit Care Med 24:733-742

9. Hack CE, DeGroot ER, Felt-Bersma RJF et al (1989) In plasma levels of interleukin-6 in sep­sis. Blood 7: 1704-1710

10. Dinarello CA, Gelfland JA, Wolff SM (1993) Anticytokine strategies in the treatment of sys­temic inflammatory response syndrome. JAMA 269:1829-1835

11. Dellinger RP (1997) Tumor necrosis factor in septic shock and multiple system trauma. Crit Care Med 25:1771-1773

12. Kollef MH (1999) The prevention of ventilator-associated pneumonia. N Engl J Med 340: 627-634

13. Darouiche RO, Raad I, Heard SO et al (1999) A comparison of two antimicrobial-impregnat­ed central venous catheters. N Engl J Med 340: 1-8

Bad and Good News in Pathophysiology, Prevention, and Management of Sepsis 73

14. 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

15. Abraham E, Glauser MP, Butler T et al (1997) p55 tumor necrosis factor receptor fusion pro­tein in the treatment of patients with severe sepsis and septic shock: A randomized controlled multicenter trial. JAMA 277: 1531-1538

16. Kushner I (1990) C-reactive protein and the acute phase response. Hosp Pract 25: 13-28 17. Chwals WJ, Fernandez M, Jamie A et al (1994) Detection of postoperative sepsis in infants

with the use of metabolic stress monitoring. Arch Surg 129:437-442 18. Shaw AC (1991) Serum C-reactive protein and neopterin concentrations in patients with viral

or bacterial infection. J Clin PathoI44:596-599 19. Yentis SM, Soni N, Sheldon J (1998) C-reactive protein as a indicator of resolution of sepsis

in the intensive care unit. Inten Care Med 21 :602-605 20. Provoa P, Almeida E, Moreira P (1998) C-reactive protein as an indicator of sepsis. Inten

Care Med 24: 1052-1056 21. Ugarte H, Silva E, Mercan D (1999) Proca\Citonin as a marker of infection in the intensive

care unit. Crit Care Med 27:498-504 22. Linden PK (1999) Management of hospital-acquired infection: Focus on Gram-positive re­

sistance. Medical Education Network Presentations in Focus: Report's and Expert Commen­tary. [From data presented at a symposium at the Society of Critical Care Medicine 28th Edu­cational and Scientific Symposium]

23. Dellinger RP, Opal SM, Rotrosen D et al and the Future of Sepsis Research Planning Com­mittee and Panel Members (1997) From the bench to the bedside. The future of sepsis re­search. Executive Summary of an American College of Chest Physicians, National Institute of Allergy and Infectious Disease, and National Heart, Lung, and Blood Institute Workshop. Chest 111:744-753

24. 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

25. Meduri GU, Headley AS, Golden E et al (1998) Effect of prolonged methylprednisolone ther­apy in unresolving acute respiratory distress syndrome: A randomized controlled trial. JAMA 280: 159-165

Light and Shadow: Perspectives on Host-Microbial Interactions in the Pathogenesis of Intensive Care Unit-Acquired Infection

J.e. MARSHALL

It has been said that in the race for supremacy, microbes are sprinting ahead ... technological developments are offering new insights and opportunities to deal a decisil'e blow to microbes, and improve the health and quality of life of mankind.

Sir Richard Sykes Hamao Umezawa Memorial Award Lecture. 21st ICC, Japan, 1999

Conventional wisdom holds that the relationship between humans and the mi­crobial world is intrinsically inimical. Microorganisms have been responsible for some of the greatest scourges that have faced humanity - the plague, tuber­culosis, smallpox, and the spectrum of diseases that threaten the majority of people in the developing world, including malaria, bacillary dysentery, and par­asitic diseases. Modern infectious epidemics, such as the influenza pandemic of 1918 and AIDS, have had a devastating effect on large numbers of people. In the developed world, an increasing number of illnesses are being found to have an infectious origin, notably peptic ulcers [1] and atherosclerosis [2]. From the perspective of the clinician, the need to administer a definitive blow to the mi­crobial world has never seemed greater, at a time when our arsenals have never been more potent.

But viewed from an evolutionary perspective, the relationship between the mammalian and microbial worlds is profoundly different. Symbiosis, rather than destruction, is the predominant pressure. Evolution is both promiscuous and op­portunistic: interactions between the organism and its environment are exploited to benefit the survival of the organisms, and if this can be accomplished to mu­tual benefit, so much the better. Fruits, for example, survive because they serve as food for animals; ingestion of the fruit and its contained seeds permits the seed to be transported in feces and ensures the propagation and survival of the plant. On the other hand, when the long-standing ecological interactions that have regulated populations over evolutionary time are disrupted, as they were, for example, with the introduction of the cane toad to Australia or the zebra mussel to the North American great lakes, the consequences can be disastrous. The interaction between the human host and its microbial flora represents a

76 J. C. Marshall

complex evolutionary microcosm that reflects these principles. Its disruption in critical illness can have important consequences.

The advent of the intensive care unit (ICU) is forcing a fundamental re-eval­uation of concepts of disease, for never before have we as clinicians been faced with treating illnesses whose very existence is a consequence of our therapeutic successes [3]. Acute respiratory distress syndrome only develops in patients whose lives have been saved through the availability of mechanical ventilation, but its evolution is a direct consequence of that technology [4]. The dysregulat­ed inflammatory response of sepsis only arises when the acute lethality of infec­tion is subverted by resuscitation and organ system-specific support, and a process that evolved to be protective comes to threaten the survival of the pa­tient. Similarly the morbidity of infection in critical illness requires a fundamen­tal re-evaluation of concepts, and a heightened awareness of the dynamic inter­play that exists between the host, the microbial flora, and the homeostatic changes that therapy produces. This brief review considers some of the elements that must be incorporated into that re-evaluation.

Normal host-microbial interactions The average healthy human being consists of roughly 1,013 mammalian cells, differentiated into approximately 250 different cell types. However, each of us from birth harbors a complex indigenous microbial flora that is both more nu­merous and more complex. The mucosal surfaces of the skin and gastrointesti­nal tract contain approximately 1,014 microorganisms, representing between 400 and 600 individual microbial species [5]. Each occupies a distinct ecologi­cal niche, and the composition of this flora remains relatively constant over the life of the individual [6]. The stability of this flora over time is a consequence of many different factors (Table 1); one of the most important is the inhibitory ef­fects of the other organisms that comprise that flora.

The anaerobic flora of the gastrointestinal tract plays an important role in in­hibiting gut colonization by pathogens, an influence that has been termed colo­nization resistance [7]. Anaerobic organisms carpet the mucosa of the distal small intestine and colon, preventing the adhesion of other potential pathogens from the gut lumen. Moreover, they produce volatile fatty acids that directly in­hibit the growth of such important ICU pathogens as Pseudomonas and other enterobacteriaceae [8]. Elimination of the anaerobic flora jn rodents through the administration of oral antibiotics results in cecal overgrowth with Gram-nega­tive organisms, and the translocation of these organisms from the gut lumen into mesenteric lymph nodes [9].

Members of the indigenous gastrointestinal flora also synthesize and release factors with potent activity against other bacteria. These bacterial peptides, col­lectively known as bacteriocins [10], have diverse spectra of activity, inhibiting

Light and Shadow: Perspectives on Host-Microbial Interactions

Table 1. Factors regulating the composition of the indigenous flora of the gastrointestinal tract *

Physiological Gastric acid Bile salts Gastrointestinal motility Luminal oxygen tension Nutrient availability

Immunological IgA from bile and gut mucosa Gut-associated lymphoid tissues Paneth cells- BPI (bactericidal permeability increasing protein), defensins, lactoferrin, phospholipase A2 Epithelial cell-derived cytokine

Microbiological Competition for nutrients Competition for binding sites Microbial-derived antimicrobial agents (bacteriocins) Anaerobic flora (volatile fatty acids)

• Adapted from [47] BPI stands for bactericidal permeability increasing protein

77

the growth of such potential pathogens as group A streptococci and Candida, Lactobacilli, in particular, produce antimicrobial peptides directed against Staphylococus aureus, Salmonella, and Clostridium [11], and even support the synthesis by the host of specific IgA antibody against gut pathogens such as Salmonella typhi [12],

The complex ecological interactions between the many species that comprise the endogenous gut flora, and between organisms of these species and the host, are poorly understood, Nonetheless it is clear that, in health, a stable balance ex­ists between the host and microbial worlds. Disease can disrupt this balance. Conversely its disruption can produce disease.

The indigenous flora also exerts a significant impact on normal physiology, and on the development and maturation of the immune system, Animals raised in a germ-free environment fail to develop a luminal flora and manifest multiple abnormalities. The lymphoid tissues of the gut wall are underdeveloped, and the intestinal villi blunted. The cardiac output is, for reasons that are not apparent, elevated. The ability to mount a delayed hypersensitivity response to a new anti­gen is impaired, as is the capacity to become tolerant to enterally administered antigens. Susceptibility to lethal infection with organisms such as S. aureus and Klebsiella is increased [13], It is more difficult to document a beneficial role for the normal gut flora during human development, although recent studies into the epidemiology of asthma suggest that reduced exposure to orofecal and food­borne microorganisms may predispose to allergic asthma [14].

78 J.e. Marshall

In summary, then, a consideration of pathological interactions between the microbial world and the human host must start from the recognition that our destinies are intertwined, and that a normal state of health is both associated with, and dependent upon, an intact indigenous flora.

What is infection in critical illness? Infection is conventionally defined as the invasion of normally sterile tissues by viable microorganisms [15]. In the patient presenting with perforated sigmoid diverticulitis or meningococcal meningitis, this definition is both sensible and readily satisfied. In the ICU patient with evidence of a florid septic response, diffuse pulmonary infiltrates, and negative or inconclusive culture data, the defi­nition may be harder to translate into sensible clinical decisions. The ambigui­ties reflect changes in normal host-microbial homeostasis, and create diagnostic dilemmas that currently cannot be resolved.

If infection is defined as the presence of microorganisms in normally sterile tissues, and if the characteristic clinical syndrome (and its associated morbidity) associated with infection is triggered by the host response to microbial products such as endotoxin, then is the patient with negative cultures who is endotoxemic infected or not? If by infection we are describing the clinical state that evolves from the interaction of host and pathogen, then clearly he or she is. If on the other hand, we are attempting to make a decision to institute antibiotics, or to undertake a surgical procedure, then he or she is not. Endotoxemia is common in critically ill patients [16, 17], yet its presence, as measured by the convention­al Limu1us amebocyte lysate assay, correlates poorly with the presence of Gram-negative infection [18]. Currently there are no assays for inflammatory stimuli from Gram-positive organisms (e.g., peptidoglycan or lipotechoic acid) or fungi (e.g., mannan).

Infection in the critically ill patient - particularly infection that develops fol­lowing admission to the ICU - may be less a state that is either present or ab­sent, than a continuum of abnormality that reflects changes in the indigenous flora. Proximal gastrointestinal flora is altered early following ICU admission, and the nature of these changes mirrors the microbial pattern of nosocomial ICU-acquired infection. The stomach and proximal small bowel, normally ster­ile or sparsely populated with Gram-positive organisms such as streptococci and lactobacilli, become densely overgrown with Candida, coagulase-negative staphylococci, Pseudomonas, and enterococci - the same organisms that pre­dominate in ICU-acquired infections [19], and gut colonization is significantly associated with invasive infection with the same organism [20]. In both animal models [9] and human studies [21], changes in gut flora alone are sufficient to promote microbial translocation from the gut lumen into adjacent mesenteric lymphatic tissue. Thus an alternate model of infection in critical illness suggests that the inflammatory response is driven by the constant entry into the body of

Light and Shadow: Perspectives on Host-Microbial Interactions 79

bacterial products such as endotoxin, or of small inocula of microorganisms across epithelial surfaces whose barrier function and patterns of normal colo­nization have been altered by the acute disease process.

Nosocomial infection in critical illness: an alternate model Nosocomial infections arising in the ICU patient differ strikingly from commu­nity-acquired infections in their microbiology, risk factors, and response to ther­apy. It seems appropriate to re-evaluate our concepts of infection in the critically ill, based on the fundamental epidemiological differences between community­and nosocomial acquired infection. An alternate model is proposed, having at its base the changes in host-microbial homeostasis that arise in critical illness (Table 2).

Table 2. A comparison of community-acquired and nosocomial intensive care unit (ICU) infections

Community-acquired

Caused by live organisms Infecting species are exogenous organisms Arise in setting of normal colonization Occur through bolus exposure to large

inoculum of organisms Respond to conventional anti-infective

therapy Prevented by specific immunotherapy

Treated by eliminating pathogen A cause of illness

ICU-acquired

Caused by live organisms or microbial products Infecting species are endogenous organisms Arise in setting of pathological colonization Occur through continuous low-level exposure

to spectrum of organisms Disappointing response to conventional

anti-infective therapy Prevented by measures directed against

pathological colonization Treated by restoring optimal homeostasis A consequence of illness

Conventional infections are caused by live microorganisms, invading the host in a single large inoculum. ICU-acquired infection, on the other hand, may result from exposure to live bacteria or bacterial products, and occurs through continuous low-grade exposure resulting from microbial overgrowth, or from a breach of the protective epithelial barriers of the skin, and respirato­ry, urinary, and gastrointestinal mucosa. In keeping with this difference, the on­set of community-acquired infection can often be determined with considerable precision, and elimination of the microbial inoculum results in resolution of the infection, while the onset and resolution of nosocomial infections are more dif­ficult to determine.

Community-acquired infections typically result from exposure to exogenous organisms, not normally resident in the host, or from an anatomical abnormali-

80 J.C. Marshall

ty of the gastrointestinal tract, for example a colonic perforation, or biliary tract obstruction. ICU-acquired infections are caused by endogenous species, whose proliferation within the host is favored by the circumstances of critical illness. These organisms are typically resistant to first-line antimicrobial agents, suggesting that their proliferation in the host is favored by depletion of the normal indigenous flora. Alternatively they possess characteristics that per­mit them to thrive under the unique circumstances of critical illness. Coagu­lase-negative staphylococci elaborate slime that facilitates their adhesion to prosthetic devices [22], while adhesins from both Pseudomonas and Candida have a particular affinity for glycosphingolipid receptors [23]. Candida are also able to degrade IgA [24], an ability that may facilitate adhesion to gut mucosal surfaces.

Community-acquired infections typically arise de novo in the patient with essentially normal patterns of colonization, whereas ICU-acquired infections develop in patients with abnormal colonization of the gastrointestinal tract [20, 25] or oropharynx [26]. As a consequence, the former reflect a bolus exposure to a sizeable inoculum of organisms, whereas the latter reflect continuous low­level exposure, often to a number of organisms.

Resolution of the infectious process usually follows adequate anti-infective therapy in patients with community-acquired infections [27,28], whereas the re­sponse to anti-infective therapy for patients with nosocomial infection is less clear [29]. While community-acquired infections are clearly a cause of serious illness, nosocomial infections can equally be considered to be a consequence of that illness [30]. Conventional infection is prevented by specific immunothera­py, and treated by eliminating the pathogen. ICU-acquired infections are pre­vented by measures that prevent pathological colonization [31], and are treated by restoring normal physiology.

The distinction between these two models of infection is exemplified by the problem of recurrent or tertiary peritonitis in the critically ill patient. Secondary bacterial peritonitis usually results from an anatomical breach of the gastroin­testinal tract, and is caused by gut organisms - Escherichia coli, Bacteroides jragilis, and the enterococcus; the infecting species are sensitive to a variety of first-line agents, and empirical therapy is generally effective [27, 32]. The re­sponse to conventional source control measures is resolution of the infection, and of the clinical manifestations of a septic response [33]. Recurrent or tertiary peritonitis, on the other hand, arises following apparently adequate therapy of primary or secondary peritonitis, and is dominated bacteriologically by organ­isms such as Candida, Pseudomonas, and coagulase-negative staphylococci [34, 35], the same species that overgrow the proximal gastrointestinal tract of the critically ill [20]. The response to either appropriate antimicrobial agents or ag­gressive surgical management is disappointing [34, 36, 37].

Light and Shadow: Perspectives on Host-Microbial Interactions 81

Conclusions The relationship between the human host and the microbial world is a dynamic and usually symbiotic one; derangements in one have consequences in the other. Anti-infective therapy is one of the more-important and deliberate interventions that alter host-microbial interactions. For the patient with pneumococcal pneu­monia or Gram-negative cholangitis from obstruction of the common bile duct, this intervention brings significant benefit. For the ICU patient with tertiary peritonitis or ventilator-associated pneumonia, the benefits are much less clear.

What is apparent, however, is that intervention in a dynamic system such as this alters that system. The widespread use of antimicrobial agents has resulted in unprecedented levels of antibiotic resistance in common infecting organisms such as S. aureus, Enterococcus, and a number of Gram-negative species [38-40]. In the individual patient, antibiotic use favors the emergence of superinfec­tion with resistant organisms such as C. difficile or Candida [41,42]. Thus it is important for clinicians to reconsider the rationale for antibiotic use in the criti­cally ill patient, and to develop strategies for minimizing the adverse ecologic consequences of such use [43].

The distinction made here between community-acquired and nosocomial in­fections is not absolute, and many patients with infection will have features of both. Nonetheless the concept provides an alternate approach to the vexing problem of infection in the critically ill patient that may merit consideration in the development of future strategies to prevent and treat infection in the critical­ly ill. Both epidemiological [44,45] and interventional [46] studies support the concept that ICU-acquired infections result in significant morbidity and mortali­ty, and new approaches to their prevention and management are needed [47].

References 1. Veldhuyzen van Zanten SJD, Sherman PM (1994) Helicobacter pylori infection as a cause of

gastritis, duodenal ulcer, gastric cancer and nonulcer dyspepsia: a systematic overview. Can MedAssocJ 150:177-185

2. Gupta S, Camm AJ (1997) Chronic infection in the etiology of atherosclerosis - the case for Chlamydia pneumoniae. Clin Cardiol 20:829-836

3. Marshall JC (1999) Rethinking sepsis: from concepts to syndromes to diseases. Sepsis 3:5-10 4. Slutsky AS, Tremblay LN (1998) Multiple system organ failure. Is mechanical ventilation a

contributing factor? Am J Resp Crit Care Med 157: 1721-1725 5. Savage DC (1977) Microbial ecology of the gastrointestinal tract. Annu Rev Med 31:107-133 6. Lee A (1985) Neglected niches. The microbial ecology of the gastrointestinal tract. Adv Mi­

crobial Ecol 8: 115-162 7. Van DerWaaij D, Berghuis De Vries JM et al (1971) Colonization resistance of the digestive

tract in conventional and antibiotic treated mice. J Hyg Camb 69:405-411 8. Levison ME (1973) Effect of colon flora and short chain fatty acids on growth in vitro of

Pseudomonas aeruginosa and enterobacteriaceae. Infect Immun 8:30-35 9. 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

82 J.C. Marshall

10. Govan JRW (1986) In vivo significance of bacteriocins and bacteriocin receptors. Scand J In­fect Dis [Suppl]49:31-37

11. Shahani KM, Ayebo AD (1980) Role of dietary lactobacilli in gastrointestinal microecology. Am J Clin Nutr 33:2448-2457

12. Link-Amster H, Rochat F, Saudan KY et al (1994) Modulation of a specific humoral immune response and changes in intestinal flora mediated through fermented milk intake. FEMS Im­munol Med Microbio1 10:55-63

13. Marshall JC (1991) The ecology and immunology of the GI tract in health and critical illness. J Hosp Infect 19:7-17

14. Matricardi PM, Rosmini F, Riondino S et al (2000) Exposure to foodborne and orofecal mi­crobes versus airborne viruses in relation to atopy and allergic asthma: epidemiological study. BMJ 320:412-417

15. Bone RC, Balk RA, Cerra FB et al (1992) ACCP/SCCM Consensus Conference. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest 101:1644-1655

16. Danner RL, Elin RJ, Hosseini JM et al (1991) Endotoxemia in human septic shock. Chest 99:169-175

17. van Deventer SJ, Buller HR, Ten Cate JW et al (1988) Endotoxemia: an early predictor of septicaemia in febrile patients. Lancet 1 :605-609

18. Bates DW, Parsonnet J, Ketchum PA et al (1998) Limulus amebocyte lysate assay for detec­tion of endotoxin in paients with sepsis sydrome. Clin Infect Dis 27:582-591

19. Marshall JC, Christou NV, Hom R et al (1988) The microbiology of multiple organ failure. The proximal GI tract as an occult reservoir of pathogens. Arch Surg 123:309-315

20. Marshall JC, Christou NV, Meakins JL (1993) The gastrointestinal tract. The "undrained ab­scess" of multiple organ failure. Ann Surg 218: 111-119

21. Krause W, Matheis H, Wu1f K (1969) Fungaemia and funguria after oral administration of Candida albicans. Lancet 1 :598-599

22. Patrick CC, Plaunt MR, Hetherington SV et al (1992) Role of the Staphylococcus epidermidis slime layer in experimental tunnel tract infections. Infect Immun 60: 1363-1367

23. Yu L, Lee KK, Hodges RS et al (1994) Adherence of Pseudomonas aeruginosa and Candida albicans to glycosphingolipid (asialo-GM1) receptors is achieved by a conserved receptor­binding domain present on their adhesins. Infect Immun 62:5213-5219

24. Reinholdt J, Krogh P, Holmstrup P (1987) Degradation of IgA1, IgA2, and S-IgA by Candida and Torulopsis species. Acta Path Microbiol Immunol Scand 95:265-274

25. Garvey BM, McCambley JA, Tuxen DV (1989) Effects of gastric alkalization on bacterial colonization in critically ill patients. Crit Care Med 17:211-216

26. Johanson WG, Pierce AK, Sanford JP (1969) Changing pharyngeal bacterial flora of hospital­ized patients. N Engl J Med 281:1137-1140

27. Bohnen JMA, Solomkin JS, Dellinger EP et al (1992) Guidelines for clinical care: anti-infec­tive agents for intra-abdominal infection. A Surgical Infection Society policy statement. Arch Surg 127:83-89

28. Solomkin JS, Reinhart HH, Dellinger EP et al (1996) Results of a randomized trial comparing sequential intravenous oral treatment with ciprofloxacin plus metronidazole to imipenem cilastatin for intra-abdominal infections. Ann Surg 223:303-315

29. Wunderink RG (1998) Attributable mortality of ventilator-associated pneumonia. Sepsis 1: 211-221

30. Nathens AB, Chu PTY, Marshall JC (1992) Nosocomial infection in the surgical intensive care unit. Infect Dis Clin North Am 6:657

31. D'Amico, Pifferi S, Leonetti C et al (1998) Effectiveness of antibiotic prophylaxis in critical­ly ill adult patients: systematic review of randomized controlled trials. BMJ 316:1275-1285

32. Lorber B, Swenson RM (1975) The bacteriology of intra-abdominal infections. Surg Clin North Am 55:1349-1354

33. Rotstein 00, Meakins JL (1990) Diagnostic and therapeutic challenges of intraabdominal in­fections. World J Surg 14:159-166

Light and Shadow: Perspectives on Host-Microbial Interactions 83

34. Nathens AB, Rotstein 00, Marshall IC (1998) Tertiary peritonitis: clinical features of a com­plex nosocomial infection. World I Surg 22: 158-163

35. Rotstein 00, Pruett TL, Simmons RL (1986) Microbiologic features and treatment of persist­ent peritonitis in patients in the intensive care unit. Can I Surg 29:247-250

36. Bunt TJ (1986) Non-directed relaparotomy for intraabdominal sepsis: a futile procedure. Am Surg 52:294

37. Norton LW (1985) Does drainage of intraabdominal pus reverse multiple organ failure? Am 1 Surg 149:347-350

38. Archibald L, Phillips L, Monnet 0 et al ( 1997) Antimicrobial resistance in isolates from inpa­tients and outpatients in the United States: increasing importance of the intensive care unit. Clin Infect Dis 24:211-215

39. Centers for Disease Control (1993) Nosocomial enterococci resistant to vancomycin - United States - 1989-1993. MMWR 42:597-599

40. Berkelman RL, Hughs 1M (1993) The conquest of infectious diseases: Who are we kidding? Ann Intern Med 119:426-428

41. Gorbach SL (1999) Antibiotics and Clostridium difficile. N Engl 1 Med 341: 1690-1691 42. Bross 1, Talbot GH, Maislin G et al (1989) Risk factors for nosocomial candidemia: a case

control study in adults without leukemia. Am I Med 87:614-620 43. Marshall IC, Evans DC (1998) Antimicrobial therapy for ICU-acquired infection: time for a

reappraisal. In: Vincent I-L (ed) Yearbook of Intensive Care and Emergency Medicine. Springer-Verlag, Berlin Heidelberg New York, pp 283-291

44. Emmerson AM (1990) The epidemiology of infections in intensive care units. Intensive Care Med 16fSuppI3]:SI97-S200

45. Weinstein RA (1991) Epidemiology and control of nosocomial infections in adult intensive care units. Am I Med 91 [Suppl3B [:3B-179S-3B-184S

46. Nathens AB, Marshall IC (1999) Selective decontamination of the digestive tract (SOD) in surgical patients. Arch Surg 134: 170-176

47. Marshall JC (1999) The gastrointestinal flora and its alterations in critical illness. Curr Opin Crit Care 5: 119-125

Identification and Characterization of Protein Tyrosine Phosphatases Expressed in Human Neutrophils

J. KRUGER, T. FUKUSHIMA, G.P. DOWNEY

The neutrophil is an important component of the innate immune system respon­sible for host defense. In this regard, it has a dual function: to destroy pathogen­ic microorganisms and to remove inflammatory debris. This functional role is achieved through a series of rapid and coordinated responses that include chemotaxis, phagocytosis, and intracellular killing of invading microorganisms. The latter is accomplished by release of a variety of microbicidal enzymes and cationic proteins contained in granules (exocytosis) and by production of reac­tive oxygen intermediates by the NADPH oxidase [1, 2]. In certain situations however, these toxic compounds can injure host tissues as is believed to occur in disorders characterized by inflammatory damage such as rheumatoid arthritis [3], inflammatory bowel disease [4, 5], and acute lung injury [6-8]. Thus to maintain homeostasis and minimize tissue damage, leukocyte microbicidal re­sponses must be precisely regulated by processes including selective triggering and rapid termination of activation cascades once the initial stimulus has been removed. Currently the mechanisms which regulate neutrophil activation are in­completely understood.

One of the earliest biochemical events evoked by myeloid cell receptor en­gagement is the phosphorylation of cellular proteins on serine, threonine and ty­rosine residues [9-11]. Although stoichiometrically tyrosine phosphorylation ac­counts for less than one percent of total cellular phosphorylation, it plays a criti­cal role in the regulation of important cellular functions of the neutrophil [12, 13]. Increases in tyrosine phosphorylation can be elicited by a variety of soluble and particulate stimuli and correlate temporally with the occurrence of cellular responses [12, 14, 15]. The importance of tyrosine phosphorylation to leukocyte function is further underscored by the observation that inhibitors of protein tyro­sine kinases block many microbicidal responses, including adherence [16], chemotaxis [17], phagocytosis [18] and production of reactive oxygen interme­diates [12, 13, 19].

The net level of cellular tyrosine phosphorylation is determined by the op­posing activities of protein tyrosine kinases (PTK) and protein tyrosine phos­phatases (PTP). These enzymes are responsible for phosphorylation and dephos­phorylation of tyrosine residues respectively. While PTK have been extensively studied, current knowledge of the identity and functional importance of PTP ex-

86 J. Kruger, T. Fukushima, G.P. Downey

pressed in neutrophils is comparatively less. To date, only CD45 and SHP-l have been identified in neutrophils but, by inference from studies in other cell types, additional PTP are likely to be expressed.

The purpose of the current study was to identify PTP expressed by human neutrophils. To accomplish this, we took advantage of the highly conserved cat­alytic domain of all known PTP ('signature' domain) to design degenerate primers for use in reverse transcription polymerase chain reaction (RT-PCR). This technique has been used successfully by others to identify mRNAs which bear such highly conserved sequences of interest including phosphatases [20-22]. In the current manuscript we demonstrate that in addition to CD45 and SHP-l neutrophils express mRNA and functional protein for the protein tyro­sine phosphatases PTPIB, and MEG2 which may play an important role in reg­ulation of these phagocytes.

Materials and methods PTPMEG2 and degenerate primers were synthesized by General Synthesis and Diagnosis (Toronto, Ont). The TA Cloning kit and One-Shot kit were purchased from Invitrogen (San Diego, CA). MuLV reverse transcriptase and Ampli Taq DNA polymerase were purchased from Perkin Elmer-Cetus (Applied Biosys­terns, Mississauga, Ont). The T7 Sequencing kit was purchased from Pharmacia LKB Biotechnology Inc (Bai D'Urfey, Quebec). Pfu polymerase was purchased from Stratagene (La Jolla,CA).

Cell isolation Human neutrophils were isolated from citrated whole blood obtained by venipuncture. Dextran sedimentation and discontinuous plasma-Percoll gradi­ents were used as previously described [23]. The purity of the neutrophils isolat­ed exceeded 98%. After isolation, cells were resuspended in KRPD buffer at a concentration of 8 x 106 cells/ml and gently rotated at room temperature until used (generally within 1 hr).

Cell culture HL-60, U937, MEG-Ol, and COS 7 cells were obtained from the American Type Culture Collection (ATCC). PLB cells were a generous gift from M. Dineaur (Riley Hospital for Children, Indiannapolis, IN). HL-60, MEG-Ol, U937, and PLB cells were grown in RPMI 1640 media containing 10% heat-in­activated fetal bovine serum, penicillin (100 units/ml), streptomycin (100 Ilg/ml) , and 0.5 mM L-glutamine. COS 7 cells were grown in high glucose DMEM containing 10% heat-inactivated fetal bovine serum, penicillin (100 g/ml), streptomycin (100 g/ml), and 0.5 mM L-glutamine.

Identification and Characterization of Protein Tyrosine Phosphatases Expressed 87

RNA isolation Total RNA was isolated from human peripheral blood neutrophils, MEG-O 1, HL-60, PLB, and U937 cell lines by the guanidinium isothiocyanate-cesium chloride protocol [24, 25].

Primers used for PCR

Degenerate primers were designed targeting highly conserved sequences in the catalytic domain of PTPases. The primer sequences correspond to the amino acid sequences DYINAS and VHCSAG and were as follows: forward primer (5' -GCGGGATCCGA(T IC)TA(T IC)AT(T IC)AA(T/C)GCIAG(T/C)TT-3'); re­verse primer (5' -GCGGAATTCAT/CICCIGCA/GCTA/GCAA/GTGIAC-3'). Two sets of primers were also designed for PTP MEG2 (accession no. M83738). Primers for set #1 corresponded to base pairs 943-1124 of the cDNA sequence and were as follows: forward primer (5'-GAGGAATGGACTGGT­GTTTATC-3'); reverse primer (5' -GGAGGAGACTGATGATGGAATA-3'). Primers for set #4 corresponded to base pairs 1402-1589 of the cDNA sequence and were as follows: forward primer (5' -GCAAAAGCAAGGAATCTATGAG-3'); reverse primer (5'-TGGCATTGATGTAATCTGTCTG-3').

Reverse transcriptase (RT) PCR

cDNA was reverse transcribed from total RNA isolated from human peripheral blood neutrophils and HL-60 cells. Murine leukemia virus (MuLV) reverse tran­scriptase and random hexamers from the Gene Amp RNA PCR kit were used for the reaction which was carried out in a Perkin-Elmer Cetus DNA Thermal Cycler 480 (Perkin-Elmer Cetus, Emeryville, CA) according to the manufactur­er's instructions. Transcribed cDNA was then amplified using the degenerate primers for an initial 5 cycles with the following parameters: denaturation at 94°C for 30 seconds, annealing at 42°C for 30 seconds, and extension at noc for 2 min. Another 30 cycles followed with denaturation at 94°C for 30 seconds, annealing at 55°C for 30 seconds, and extension at noc for 2 min. Transcribed cDNA was amplified using MEG2 primer sets #1 and #4 for 35 cycles with de­naturation at 94°C for 30 seconds, annealing at 50°C for 30 seconds, and exten­sion at noc for 2 min. All components of the amplification mixtures were test­ed for contamination by running 35 PCR cycles in absence of the template RNA. Genomic DNA contamination was tested for by performing the RT-PCR in absence of the MuLV reverse transcriptase. All reactions were analyzed by agarose gel electrophoresis using ethidium bromide staining. Amplified cDNA sequences from the RT-PCR were cloned into the Invitrogen TA vector. Se­quencing was performed using the Pharmacia T7 sequencing kit. Automated fluorescent sequencing (Pharmacia ALF) was additionally performed on select­ed sequences at the Biotechnology Centre of the University of Toronto.

88 J. Kruger, T. Fukushima, G.P. Downey

Antibodies Anti-PTPIB murine monoclonal antibody was purchased from Oncogene Sci­ence Inc. (Cambridge, MA) and used at a 1: 1000 dilution for immunoblotting, and a 1: 20 dilution for immunofluorescence. Mouse IgG2a Negative Control antibody was purchased from Serotec (Raleigh, NC). Anti-hemagglutinin (HA) mouse monoclonal antibody was purchased from the Berkeley Antibody Com­pany (Berkeley, CA) and used at a 1:5000 dilution. Rabbit immune and preim­mune serum were used in a 1: 100 dilution. Anti-mouse and anti-rabbit horse radish peroxidase (HRP) conjugated secondary antibodies (Amersham) were used at a 1:5000 dilution. Goat anti-mouse FITC conjugated secondary (Jackson Immunoresearch laboratories, West Grove, Penn) was used at a 1:50 dilution.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting SDS-PAGE and immunoblotting were carried out essentially as described previ­ously [26]. Briefly, cell lysates were fractionated by SDS-PAGE gel elec­trophoresis in a 10% or 4-20% gradient gel. Samples were transferred to nitro­cellulose membranes at 100 V for 2 hr using a Bio-Rad transfer apparatus. Membranes were incubated overnight in a blocking buffer (5% skim milk, 0.05% Tween-20 in PBS) shaking at 4°C. Membranes were incubated shaking for 2 hr at room temperature in the primary antibody buffer (primary antibody diluted in blocking buffer), and then washed three times with blocking buffer for 15, 10, and 10 min respectively. Membranes were then incubated shaking for 1 hr at room temperature in the secondary antibody buffer (secondary antibody di­luted in blocking buffer), washed once with a 3% Tween-20 PBS buffer for 2 min, and three times in PBS for 5 min each wash. The Western blots were then developed using the enhanced chemiluminescence (ECL) detection system fol­lowing the manufacturer's instructions (Amersham).

Immunofluorescence Neutrophils (50,000) were cytospun for 5 min at 800 RPM, fixed in 1.6% paraformaldehyde (PFA) for 20 min and washed in PBS three times, 5 min per wash. Samples were then incubated for 20 min in a permeablizationlblocking buffer containing 0.2% Triton X-l00 and 10% normal goat serum in PBS. After another 3 washes in PBS, samples were incubated in a primary antibody buffer for a minimum of 4 hr at 4°C. After washing, samples were incubated in a sec­ondary antibody buffer (secondary antibody diluted in PBS) for 30 min at room temperature, washed and mounted (Dako mounting medium). Cells were ana­lyzed with confocal imaging (Laser Scanning Microscope, Zeiss).

COS 7 cells were transfected with the MEG2-HA construct and grown on glass coverslips. Cells were fixed in 1.6% paraformaldehyde (PFA) for 20 min-

Identification and Characterization of Protein Tyrosine Phosphatases Expressed 89

utes and washed in PBS three times, 5 minutes per wash. Cells were incubated in a 100 mM glycine-PBS buffer (pH 7.4) for 20 minutes, washed, and then in­cubated for 20 minutes in a permeablizationlblocking buffer containing 0.1 % Triton X-lOO, 500 mM PIPES, 5 mM EGTA, 100 mM KOH, pH 6.8. After an­other 3 washes in PBS, samples were incubated in a primary antibody buffer (anti-HA mouse monoclonal antibody diluted in PBS containing 0.5% BSA) for a minimum of 4 hours at 4°C. After washing, samples were incubated in a sec­ondary antibody buffer (secondary antibody diluted in PBS containing 0.5% BSA) for 30 minutes at room temperature, washed 3 additional times, mounted, and analyzed with confocal imaging.

Immunoprecipitations and phosphatase assays

Lysates were prepared from neutrophils or transiently transfected COS 7 cells by lysing cells in 1 ml of a buffer containing 0.2% NP40 and the protease in­hibitors PMSF, leupeptin, and aprotinin. Prior to cell lysis, neutrophils were pre­treated in 2.5 mM diisopropylfluorophosphate (DFP) for 30 min at room tem­perature. Samples were immediately incubated on ice for 10 min, centrifuged at 15,000 RPM at 4°C for 15 min, and supernatants were then recovered. Protein A/G beads preblocked with BSA (Santa Cruz) were used for immunoprecipita­tion. Lysates were precleared by incubating with 50 III of the protein A/G beads for 2 hr, rotating at 4°C. Protein A/G beads were pre armed with anti-PTPIB, anti-MEG2, or anti-HA antibody for 2 hr, rotating at 4°C. Pre armed beads were washed 3 times in the lysis buffer (without protease inhibitors) and added to the precleared lysate. Samples were incubated overnight, rotating at 4°C.

MEG2-HA phosphatase activity was assessed by incubation of the immuno­precipitate in 200 III of a buffer containing: 10 mM p-nitrophenyl phosphate (PNPP) substrate (Sigma), 10 mM dithiothreitol (DTT) , 25 mM HEPES, and 0.5 mM EDTA. Samples were incubated for 2 hr shaking at 37°C. Samples were diluted with 800 III of 0.2 M NaOH and absorbance was measured at 420 nm using a UV spectrophotometer (Milton Roy).

When endogenous phosphatase activity for MEG2 and PTPIB was examined in neutrophils, the specificity of the PNPP assay was found to be suboptimal (see discussion). A phosphopeptide substrate (Malachite Green kit; UBI) was then successfully used to demonstrate specific phosphatase activity in PTP im­munoprecipitates. The immunoprecipitates were washed several times with the assay buffer (25 mM HEPES, 0.5 mM EDTA, pH 7.0). Immunoprecipitates were resuspended in 50 III of assay buffer and the phosphopeptide RRLIEDAEp YAARG (100 IlM) was added. Samples were shaken for 4 hr at 37°C and centrifuged briefly to recover 40 III of the supernatant. 100 III of Malachite Green solution was added for 5 min, and absorbance was measured at 650 with an ELISA reader (Molecular Devices).

90 J. Kruger, T. Fukushima, G.P. Downey

MEG2 Northern blot analysis Total RNA was isolated from human peripheral blood neutrophils, MEG-Ol, HL-60, PLB, and U937 cell lines. Equal amounts of RNA (4 ~g) were elec­trophoresed in a 1 % agarose/formaldehyde gel and transferred to a nylon mem­brane. A 1.4 kb radiolabeled cDNA probe was prepared corresponding to base pairs 359 to 1777 of the MEG2 cDNA sequence. Hybridization was performed at 49°C in a buffer composed of: 50% formamide, 5X SSC, 0.1 % Ficoll, 100 mg/ml salmon sperm DNA, 0.5% SDS, 10% dextran sulfate, 50 mM sodium phosphate, pH 6.5. The membrane was prewashed once at 49°C in a wash buffer of 2X SSC/0.3% SDS, and then washed twice more at 49°C in the wash buffer for 30-60 min each time.

Peptide antibody The Kyte-Doolittle, Hopp-Woods, and Surface algorithms were used to select a 20 amino acid region of the MEG2 sequence for immunization. The sequence consisting of amino acids 440-459 (NHYKKTTLEIHNTEERQKRQ) was iden­tified. This peptide was synthesized and used for immunization of 2 rabbits. Keyhole limpet hemacyanin (KLH) was coupled as a carrier protein. Three bleeds (including a preimmune bleed) were obtained from the rabbits, from which the serum was prepared. Immunoreactivity against the immunizing pep­tide was confirmed by ELISA by the manufacturer (Synpep Corp., Dublin, CA).

Affinity-purification of peptide antibody A SulfoLink Coupling Gel (Pierce, Rockford, IL) binding the peptide was con­structed for the purpose of affinity purifying the anti-peptide immune serum. The serum was diluted 10 times in a 10 mM Tris-HCI binding buffer (pH 7.4), incubated with the resin overnight rotating at 4°C, and was poured into a col­umn allowing the unbound components to pass through. The resin was washed with 20 bed volumes of binding buffer, following which the bound antibody was eluted (ImmunoPure Gentle Ag/Ab Elution Buffer, Pierce).

Construction of MEG2-hemagglutinin fusion protein The cloned MEG2 cDNA in pBluescript II SK(+) (Stratagene) was received as a generous gift from P. Majerus (Washington University School of Medicine). The MEG2 insert was then transferred to the pCDNA3 vector (Invitrogen). A 3' primer (CCCCCTCGAGTTACGCATAGTCAGGAACATCGTATGGGTACT­GACTCTCCACGGCCAGCAGG) was constructed containing the final 24 bp of the coding MEG2 sequence, the hemagglutinin (HA) sequence, a stop codon, and a XhoI restriction site. A 5' primer (GGCAGAAACGCCAGGTGACCC) was synthesized which targeted an upstream region in the MEG2 cDNA con-

Identification and Characterization of Protein Tyrosine Phosphatases Expressed 91

taining a BstEll restriction site. Twenty-five cycles of the polymerase chain re­action (PCR) were performed with denaturation for I min at 94 DC, annealing at 55°C for I min, and extension at noc for I min using Pfu polymerase. The PCR product was double digested (BstEII, XhoI) and ligated into the MEG2-pcDNA3 construct in place of the existing BstEII-XhoI sequence in the native MEG2 sequence.

Expression of MEG2-hemagglutinin fusion protein in COS 7 cells COS 7 cells were transiently transfected with the MEG2-HA construct using ei­ther Lipofectamine (Gibco-BRL) or Superfect (Qiagen) according to the manu­facturers' instructions. Cells were typically harvested on day 3 post-transfection.

Results

Identification of transcriptionally active PTP in neutrophils

To identify PTP genes that were transcriptionally active in neutrophils, an RT­PCR strategy was employed using degenerate primers targeted at highly con­served regions within the catalytic domain of PTP. Selected studies were also done in HL-60 cells, a myeloid cell line that is known to be more transcription­ally active than neutrophils. Total RNA was isolated from neutrophils and HL-60 cells, and used as a template for RT-PCR. The resultant PCR products were approximately 450 bp in length (Fig. I) and were cloned into the TA cloning vector for sequencing. A total of ISO sequences was analyzed from neutrophils. Thirty four of the sequences were identified as ribosomal RNA, and 18 se­quences had no homology « 1%) with known proteins. The remaining 98 se­quences shared complete sequence identity with 4 known PTP as follows: CD45 (n = 30), SHP-I (11 = 26), PTPIB (n = 22), and MEG2 (n = 20).

The expression of CD45 and SHP-I in neutrophils has previously been re­ported by our laboratory and others [26-31]. The detection of PTP previously documented to be expressed in neutrophils served as a confirmation of the valid­ity of this technique for identification of PTP in these cells. Expression of the tyrosine phosphatases PTPlB and MEG2 has not been previously reported and, accordingly, we focused our attention on these phosphatases.

Functional PTPIB protein is expressed by neutrophils

To determine if protein for PTP-I B was present in human neutrophils, Western blots of whole cell extracts were conducted. Figure 2a illustrates that a 50 kDa band was present in whole cell extracts of neutrophils. This molecular weight corresponds to observations in other cell types reported previously [32]. A 64 kDa band of much lower immunoreactivity that was observed was likely the re-

92

bp

700 ~

600 -500 -

J. Kruger, T. Fukushima, G.P. Downey

DNA RT-P R

400 J 1--_____ -....1

Fig. 1. Degenerate PTP RT-PCR product. Degenerate primers targeting the conserved amino acid sequences DYINAS and VHCSAG were used in RT-PCR to generate product of approximately 450 bp. Total RNA was isolated from neutrophils, reverse transcribed into cDNA and amplified using PCR as described in Materials and methods

suIt of non-specific binding of the secondary antibody because it was apparent in the absence of the primary antibody (not shown) or when non-immune mouse serum was used as the primary antibody.

To determine the enzymatic activity of immunoreactive PTPIB in neu­trophils [33, 34], a phosphatase assay on anti-PTPlB immunoprecipitates was conducted. Phosphatase assays were performed on neutrophils from 4 donors and confirmed that the enzyme was catalytically active (Fig. 2b).

To identify the subcellular localization of PTPIB in neutrophils, immunoflu­orescence studies were performed. Figure 2c demonstrates that the protein lo­calizes to the cytoplasm which is in agreement with the predicted amino acid se­quence and the results of previous studies in other cell types [32].

Functional MEG2 protein is expressed by neutrophils

To confirm that mRNA for MEG2 was expressed in neutrophils an additional RT-PCR analysis was conducted with 2 additional sets of specific (i.e. nonde­generate) primers. This additional analysis targeted two distinct areas of the cloned MEG2 sequence [35]: primer set #1 targeted base pairs 943-1124 of the MEG2 cDNA sequence and primer set #4 targeted a region in the C terminal di­rection corresponding to base pairs 1402-1589. Bands of the predicted size (181-187 bp) resulted from RT-PCR amplification of total RNA from neu­trophils (Fig. 3). For confirmation, this analysis was repeated using HL-60 cells, because expression of MEG2 mRNA has previously been reported in these cells [35]. Figure 3 illustrates that bands of the appropriate size also resulted from amplification of RNA from HL-60 cells. DNA sequence analysis confirmed that these PCR products corresponded to MEG2.

Identification and Characterization of Protein Tyrosine Phosphatases Expressed

."1 11 -

1'"1 Pill

(a)

\ 111 1- 1' II'I B

111 (l II ~ ('

s.: r II 111

--'

§ = on ' 0 .., u iii .0 .... 0 <, {i!

(c)

04

03

02

01

1l01l illllllUlI1!

(b)

93

d ll ll-!' 11'113

Fig. 2. PTPI B is expressed in the cytoplasm of neutrophils and contains phosphatase activity. a) Neutrophil whole cell Iysates were dissolved in Laemmli sample buffer, separated by SDS-PAGE, and blotted to nitrocellulose as described in Materials and methods. Blots were probed using either an anti-PTP I B murine monoclonal antibody (left lane). or non-immune mouse serum (right lane) . A 50 kDa band is recognized corresponding to the predicted molecular weight of PTP I B. Addition­ally, a nonspecific 64 kDa band is present in the lane blotted with non-immune serum and faintly in the lane blotted with immune serum. b) Neutrophils were isolated, lysed and used for immunopre­cipitation with either the affinity-purified anti-PTP I B antibody or non-immune mouse serum and Protein A/G beads (described in Materials and methods). Immunoprecipitates were incubated with a phosphopeptide substrate and phosphatase activity was quantified using malachite green. Ab­sorbance was measured at 650 nm. c) Fifty thousand neutrophils were cytospun, fixed, perme­ablized, and incubated with primary and F1TC-conjugated secondary antibodies. Samples were analyzed using confocal microscopy as described in Materials and methods. Size bar equals 5 J.1m

94 J. Kruger, T. Fukushima, G.P' Downey

As PCR can detect minute quantities of DNA, we endeavored to quantify the amount of mRNA expression using Northern analysis. A lA-kb probe corre­sponding to base pairs 359-1777 of the MEG2 cDNA sequence was constructed and labeled. Total RNA was isolated from neutrophils and the cell lines HL-60, PLB, U937, and MEGOI. Northern blotting confirmed mRNA expression in HL-60 and MEG01, as found by Gu et al. [35] (Fig. 4). Expression of MEG2 mRNA was also detected in U937 and PLB cells. However, MEG2 mRNA could not be detected in preparations of neutrophil total RNA by Northern blot­ting (not shown) consistent with the low sensitivity of this technique compared with RT-PCR and with the low levels of mRNA known to be present in these terminally differentiated cells.

In order to determine if MEG2 protein was expressed in neutrophils, a poly­clonal anti-peptide antibody was raised in rabbits as described in Materials and methods. To test the specificity of this antibody, Western analysis was first per­formed on cell Iysates of COS 7 cells transiently transfected with an epitope­tagged MEG2. Figure 5 (left) illustrates that the immune but not pre-immune serum recognized a 70 kDa protein corresponding to the predicted molecular weight of the recombinant fusion protein. When the blots were stripped and reprobed with a monoclonal anti-HA antibody, a single band corresponding to that detected by the anti-MEG2 peptide serum was apparent indicating that the epitope tag did not interfere with the ability of the anti-peptide serum to recog­nize recombinant MEG2. This anti-peptide serum recognized a single 68 kDa band in neutrophils corresponding to the predicted molecular weight of MEG2 (Fig. 5), indicating that MEG2 protein was expressed in these cells.

Pnmer#l Pnmer#4

pAW109 PMN HL60 PMN HL60

kb M + + + +

Fig. 3. Detection of MEG2 mRNA by RT-PCR in PMN and HL-60 cells using two distinct sets of primers. Nondegenerate primer sets #1 and #4 targeted two distinct regions of the MEG2 se­quence. RT-PCR was performed on RNA isolated from neutrophils and HL-60 cells. Predicted PCR products for primer sets #1 and #4 were 181 and 187 bp respectively. No product was detect­ed in negative controls for RT-PCR excluding template RNA. As a positive control, the plasmid pAW I 09 was used with specific primers which amplify a region of 308 bp

Identitication and Charactcrization of Protein Tyrosine Phosphatases Expressed 95

1.4 k b

H L-60 M -01 U937 PLB Fig. 4. MEG2 mRNA detccted in HL-60, MEG-O I, U937, and PLB cell lines. Total RNA was iso­lated from HL-60, MEG-OI, U937, and PLB cell lines, fractionated on a 1% agarose formalde­hyde gel, and transferred to a nylon membrane. Blot, were probed using a 1.4 kb probe correspon­ding to base pairs 359-1777 of the MEG2 cDNA. Equal amounts of total RNA were loaded as de­termined by intensity of the ISS and 2SS RNA on the ethidium bromide-stained gel

To study the enzymatic activity of MEG2, phosphatase assays were initially performed on anti-HA immunoprecipitates of recombinant HA-MEG2 ex­pressed in COS 7 cells. Figure 6 (left) illustrates that the recombinant MEG2-HA fusion protein possessed phosphatase activity as determined by the ability to dephosphorylate pNPP. To determine if MEG2 endogenously expressed in neutrophils possessed enzymatic activity, affinity-purified anti-MEG2 peptide serum was used for immunoprecipitation from neutrophil lysates. Initial at­tempts to measure phosphatase activity using pNPP as a substrate proved unsuc­cessful due to substantial nonspecific activity. However through use of a phos­phopeptide substrate with detection using Malachite Green (Upstate Biologi­cals, Lake Placid, NY), specific phosphatase activity of endogenous MEG2 was readily detectable in neutrophil celllysates (Fig. 6, right). To determine the sub­cellular localization of endogenous MEG2, COS 7 cells were transiently trans­fected with the MEG2-HA construct and studied using immunofluorescence. Figure 7 illustrates that MEG2 was cytoplasmic in location in agreement with the predicted amino acid sequence. The polyclonal anti-MEG2 peptide anti­serum was not useful for immunofluorescence.

96 1. Kruger, T. Fukushima, G.P. Downey

70 kDa .. 71 kOa .... .

nli-II 1EG2 pn:immune illllll une !>eruill

i In 111 1111 C pre i In III Lin e

Fig. 5. Anti-MEG2 peptide immune serum recognizes epitope tagged and endogenous MEG2 pro­tein. Left part: COS 7 cells were transiently transfected with either the MEG2-HA construct or the pcDNA3 vector alone as described in Materials and methods. Three days post transfection, cells were lysed, separated by SDS-PAGE and transferred to nitrocellulose. Western blotting was per­formed using anti-HA mouse monoclonal antibody (left lane), anti-MEG2 peptide immune serum (center lane), and preimmune serum (right lane). Right part: MEG2 protein is expressed in neu­trophils. Neutrophils were isolated, dissolved in Laemmli sample buffer, proteins separated by SDS-PAGE and transferred to nitrocellulose as described in Materials and methods. Blots were probed using either immune or preimmune serum from a rabbit immunized with MEG2 peptide. The immune serum recognizes a 68 kDa band corresponding to the predicted molecular weight of MEG2

Discussion

In the present study we utilized degenerate RT-PCR to determine the identity of PTP expressed by neutrophils. This strategy targeted highly conserved regions within the catalytic domain of PTP to identify those which were expressed. Of the 150 sequences that were analyzed, 4 known PTP were identified including: CD45, SHP-l, PTPIB, and MEG2, The expression of PTPlB and MEG2 mRNA in neutrophils has not been previously described. The number of se­quences identified as PTPIB (n = 22) or MEG2 (n = 20) is comparable to the number of sequences identified as CD45 (n = 30) or SHP-I (n = 26). To the ex­tent that these values reflect the relative abundance of the respective transcripts in neutrophils, the level of expression of PTPIB and MEG2 is comparable to that of CD45 and SHP-l. The identity of the eighteen sequences which had no significant homology to known tyrosine phosphatases remains to be determined.

Identification and Characterization of Protein Tyrosine Phosphatases Expressed 97

().~ 0.4

S PNPP S

Phospilopcptidc

= Malachite Green 0 0.3 = 0.3 ~l 0 ..,. ",

CI> \.0 ... CI>

= 0.2 ... ~ = 0.2 ~ ~ ... ~ 0 ... III 0 ~ III I': 0.1 ~ 0.1 ~

() 0 vcctor MEG2-IIA non im m une anti-M EG2

Fig. 6. Epitope tagged and endogenous MEG2 have phosphatase activity. Left part: COS 7 cells were transiently transfected with either the MEG2-HA construct, or the pcDNA vector alone. Three days post-transfection, cells were lysed and an immunoprecipitation was performed using a monoclonal anti-HA antibody. The immunoprecipitate was incubated with the substrate p-nitro­phenyl phosphate (pNPP) as described in Materials and methods, and absorbance of the sample was measured at 420 nm. Right part: Neutrophils were isolated, lysed and used for immunoprecip­itation with either the affinity-purified anti-MEG2 antibody or control (non-immune) rabbit IgG as described in Materials and methods. Immunoprecipitates were incubated with a phosphopeptide substrate and quantified using malachite green. Absorbance was measured at 650 nm

While these sequences might represent novel PTP, this possibility is unlikely be­cause of the very low levels of homology « 1%) with the highly conserved phosphatase domain.

While RT-PCR was felt to be an appropriate technique to identify PTP ex­pressed by neutrophils, the procedure has several shortcomings which merit dis­cussion. First, although the RT-PCR primers used were degenerate and the initial cycles of PCR were conducted using a permissive annealing temperature, the affinity of the primers for other PTP sequences may have been insufficient to al­low annealing and amplification thus excluding them from detection. Second, as we limited our sequencing to 150 clones, PTP transcripts of low abundance might not have been identified. On the other hand, since PCR is such a sensitive technique, DNA contamination could lead to detection of PTP that were not ex­pressed. However, we utilized immunoblotting to confirm protein expression of each of the PTP identified by RT-PCR. Lastly, an alternative strategy to identify PTP would be to utilize Western blotting to screen neutrophil lysates. Such a strategy would limit detection of PTP to those for which antibodies were avail­able and would exclude the possibility of identifying novel PTP.

The expression of MEG2 mRNA in neutrophils and in cultured cell lines of myeloid origin (HL-60) was confirmed using RT-PCR with specific primers.

98 1. Kruger, T. Fukushima, G.P. Downey

M EG2-H A transfection vector alone

Fig. 7. MEG2-HA fusion protein has cytoplasmic localization in transiently transfected COS 7 cells. COS 7 cells were transfected with the MEG2-HA construct and grown on glass coverslips. Cells were fixed, permeablized, and incubated with a monoclonal anti-HA antibody followed by a FlTC-conjugated secondary antibody. Samples were analyzed using confocal microscopy as de­scribed in Materials and methods. Size bar equals 10 Ilm

Additionally, Northern blotting was used to quantify the amount of MEG2 mRNA in these cell types, as well as several other cell lines. MEG2 mRNA was detected in HL-60 and MEG-Ol cell lines, in agreement with previous findings [35], as well as in the PLB and U937 lines, cells of myeloid and myelomono­cytic origin respectively. Our inability to detect MEG2 mRNA in neutrophils using Northern blotting likely reflects the low levels of mRNA in these termi­nally differentiated cells in combination with the relative insensitivity of North­ern blotting.

Two artificial substrates were used in this study to examine the phosphatase activity of MEG2: pNPP and a phosphopeptide. Using the pNPP substrate, we were able to detect phosphatase activity for the epitope-tagged MEG2 transient­ly expressed in COS cells. When the endogenous phosphatase activity of MEG2 in neutrophils was examined, the pNPP assay was found to be suboptimal. This likely reflects in part the higher levels of expression of recombinant protein in COS cells compared with endogenous levels in neutrophils. A more sensitive and specific assay using a phosphopeptide substrate was then successfully used to demonstrate specific phosphatase activity in MEG2 immunoprecipitates. An­other factor to be considered is the nature of these artificial substrates. pNPP is a small molecule that can be hydrolyzed by enzymes other than PTP as well as undergo spontaneous hydrolysis (albeit slowly). Additionally, PTP have evolved

Identification and Characterization of Protein Tyrosine Phosphatases Expressed 99

the capacity to recognize specific phosphotyrosine residues in the context of the structure of a protein. The phosphopeptide substrate presumably binds more tightly within the binding cleft of the phosphatase than a small tyrosine ana­logue (pNPP) that does not provide the flanking amino acids that facilitate binding.

Immunofluorescence confocal microscopy confirmed that PTPIB is predom­inantly cytoplasmic in neutrophils. The cellular location of PTPIB has previous­ly been analyzed by Frangioni et al. [32] in HeLa cells. They reported that PTPIB was a cytoplasmic protein, specifically localized in the endoplasmic reticulum (ER) via a highly hydrophobic 35 amino acid C-terminal sequence. In transiently transfected COS 7 cells, the MEG2-HA fusion protein was shown to have a cytoplasmic localization as well. Sequence analysis of the MEG2 cDNA predicts the protein to be cytosolic based on the lack of a leader or highly hy­drophobic (transmembrane domain) sequence. Although our findings confirmed these proteins were localized diffusely within the cytoplasm in neutrophils, cel­lular fractionation studies would be required to resolve the specific subcellular localization of the proteins within the cytoplasm.

The role of specific protein tyrosine phosphatases in leukocyte function is beginning to be defined. CD45 for example has been demonstrated to be critical for the maintenance of integrin-mediated adhesion in bone marrow-derived macrophages [36]. Evidence exists that the activation of the Src family members p56lck and p59fyn is dependent on a CD45-mediated dephosphorylation of these kinases [37]. Overexpression of SHP-I in established myeloid cell lines has been shown to suppress cell growth [38], and SHP-l has been shown to associ­ate with and regulate the function of a number of growth factor receptors includ­ing the EGF, PDGF, Kit, and IL-3 receptors [39]. Recent work from our lab has shown that SHP-l negatively regulates adhesion and the oxidative burst in myeloid cells in response to PMA stimulation [40]. By analogy to CD45 and SHP-l, it is likely that PTP such as MEG2 and PTPIB contribute significantly to regulation of leukocyte function. Studies to determine their role in leukocyte regulation are currently underway in our laboratory.

Conclusion Neutrophils are important in host defense but, in pathologic circumstances, may mediate inflammatory tissue injury. Tyrosine phosphorylation, which represents the balance of activity of protein tyrosine kinases and protein tyrosine phos­phatases, regulates signaling pathways involved in neutrophil activation. Rela­tive to tyrosine kinases, little is known about the identity and functional impor­tance of phosphatases in neutrophils. Our goal was to identify tyrosine phos­phatases expressed in human neutrophils as a prelude to determination of their functional importance. Reverse transcriptase PCR was employed using degener­ate primers based on highly conserved motifs in the catalytic domain of PTP to

100 J. Kruger, T. Fukushima, G.P. Downey

identify tyrosine phosphatases expressed by these cells. Using this technique, 4 sequences were identified from total RNA that represented tyrosine phos­phatases including SHP-l, CD45, PTPlB, and MEG2. Expression of SHP-l and CD45 by neutrophils has previously been documented. Western blotting with commercially available antibodies confirmed the expression of protein for PTPIB. A polyclonal antibody against a MEG2 peptide was raised in rabbits and Western blotting of neutrophil whole cell lysates demonstrated an im­munoreactive band of the predicted molecular weight. Phosphatase activity was confirmed for each of the phosphatases using an in vitro assay. The subcellular distribution of PTPIB and MEG2 was found to be cytoplasmic, in agreement with their predicted structure. In summary, our results indicate that in addition to CD45 and SHP-l, human neutrophils express supplementary tyrosine phos­phatases including PTPIB and MEG2 that may participate in their regulation.

References 1. Babior BM (1978) Oxygen-dependent microbial killing by phagocytosis. N Engl J Med

298:659-668 2. Badwey JA, Curnutte JT, Karnovsky ML (1979) The enzyme of granulocytes that produces

superoxide and peroxide. An elusive pimpernel. N Engl J Med 300: 1157-1160 3. Weissmann G, Korchak H (1984) Rheumatoid arthritis. The role of neutrophil activation. In­

flammation 8[SuppIJ:S3-14 4. Chester JF, Ross JS, Malt RA, Weitzman SA (1985) Acute colitis produced by chemotactic

peptides in rats and mice. Am J Pathology 121:284-290 5. Wandall JH (1985) Function of exudative neutrophilic granulocytes in patients with Crohn's

disease of ulceritive colitis. Scandinavian Journal of Gastroenterology 20: 1151-1156 6. Stevens JH, Raffin TA (1984) Adult respiratory distress syndrome - I. Aetiology and mecha­

nisms. Postgrad Med J 60:505-513 7. Weiland JE, Davis WB, Holter JF et al (1986) Lung neutrophils in the adult respiratory dis­

tress syndrome: Clinical and pathophysiologic significance. Am Rev Respir Dis 133:218-225 8. Repine JE, Beehler CJ (1991) Neutrophils and adult respiratory distress syndrome: Two inter­

locking perspectives in 1991. Amer Rev Resp Dis 144:251-252 9. Babior BM (1988) Protein phosphorylation and the respiratory burst. Arch Biochem Biophys

264:361-367 10. Dusi S, Donini M, Della Bianca V, Rossi F (1994) Tyrosine phosphorylation of phospholipase

C-gamma 2 is involved in the activation of phosphoinositide hydrolysis by Fc receptors in hu­man neutrophils. Biochem Biophys Res Commun 201: 1100-1108

11. Greenberg S, Chang P, Silverstein SC (1993) Tyrosine phosphorylation is required for Fc re­ceptor-mediated phagocytosis in mouse macrophages. J Exp Med 177:529-534

12. Berkow RL, Dodson RW, Kraft AS (1989) Human neutrophils contain distinct cytosolic and particulate tyrosine kinase activities: Possible role in neutrophil activation. Biochim Biophys Acta 997:292-301

13. Kusunoki T, Higashi H, Hosoi S et al (1992) Tyrosine phosphorylation and its possible role in superoxide production by human neutrophils stimulated with FMLP and IgG. Biochem Bio­phys Res Commun 183:789-796

14. Huang CK, Laramee GR, Casnellie JE (1988) Chemotactic factor induced tyrosine phospho­rylation of membrane associated proteins in rabbit peritoneal neutrophils. Biochem Biophys Res Commun 151:794-801

Identification and Characterization of Protein Tyrosine Phosphatases Expressed 101

15. Gomez-Cambronero J, Huang CK, Bonak VA et al (1989) Tyrosine phosphorylation in hu­man neutrophil. Biochem Biophys Res Commun 162:1478-1485

16. McGregor PE, Agrawal DK, Edwards JD (1994) Attenuation of human leukocyte adherence to endothelial cell monolayers by tyrosine kinase inhibitors. Biochem Biophys Res Commun 198:359-365

17. Gaudry M, Caon AC, Gilbert C et al (1992) Evidence for the involvement of tyrosine kinases in the locomotory responses of human neutrophils. J Leukoc Bioi 51: 103 -I 08

18. Kobayashi K, Takahashi K, Nagasawa S (1995) The role of tyrosine phosphorylation and Ca2+ accumulation in Fc gamma- receptor-mediated phagocytosis of human neutrophils. J Biochem(Tokyo) 117:1156-1161

19. Grinstein S, Furuya W (1991) Tyrosine phosphorylation and oxygen consumption induced by G proteins in neutrophils. Am J PhysioI260:C1019-1027

20. LeBoeuf RD, Galin FS, Hollinger SK et al (1989) Cloning and sequencing of immunoglobu­lin variable-region genes using degenerate oligodeoxyribonucleotides and polymerase chain reaction. Gene 82:371-377

21. Yang Q, Tonks NK (1991) Isolation of a cDNA clone encoding a human protein-tyrosine phosphatase with homology to the cytoskeletal-associated proteins band 4.1, ezrin, and talin. PNAS 88:5949-5953

22. Rotin D, Goldstein BJ, Fladd CA (1994) Expression of the tyrosine phosphatase LAR-PTP2 is developmentally regulated in lung epithelia. Am J PhysioI267(3.1):L263-270

23. Haslett C, Guthne LA, Kopamak MM et al (1985) Modulation of multiple neutrophil func­tions by preparative methods or trace concentrations of bacterial lipopolysaccharide. Am J PathoII19:101-110

24. Chirgwin JM, Przybyla AE, MacDonald RJ, Rutter WJ (1979) Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochem 18:5294-5299

25. Groppe JC, Morse DE (1993) Isolation of full-length RNA templates for reverse transcription from tissues rich in RNase and proteoglycans. Analyt Biochem 210:337-343

26. Fialkow L, Chan CK, Rotin D et al (1994) Activation of the mitogen-activated protein kinase signaling pathway in neutrophils. Role of oxidants. J Bioi Chern 269:31234-31242

27. Lacal P, Pulido R, Sanchez-Madrid F, Mollinedo F (1988) Intracellular location of T200 and Mol glycoproteins in human neutrophils. The Journal of Biological Chemistry 263:9946-9951

28. Pulido R, Lacal P, Mollinedo F, Sanchez-Madrid F (1989) Biochemical and antigenic charac­terization of CD45 polypeptides expressed on plasma membrane and integral granules of hu­man neutrophils. FEBS Letters 249:337-342

29. Harvath L, Balke JA, Christiansen NP et al (1991) Selected antibodies to leukocyte common antigen (CD45) inhibit human neutrophil chemotaxis. J ImmunoI146:949-957

30. Fialkow L, Chan CK, Downey GP (1997) Regulation of CD45 in neutrophils: Modulation by oxidants. J Immunol 158:5409-5417

31. Brumell JH, Chan CK, Butler J et al (1997) Regulation of Src homology 2-containing tyro­sine phosphatase I during activation of human neutrophils. Role of protein kinase C J Bioi Chern 272:875-882

32. Frangioni JV, Beahm PH, Shifrin V et al (1992) The nontransmembrane tyrosine phosphatase PTP-IB localizes to the endoplasmic reticulum via its 35 amino acid C-terminal sequence. Cell 68:545-560

33. Hannig G, Ottilie S, Schievella AR, Erikson RL (1993) Comparison of the biochemical and biological functions of tyrosine phosphatases from fission yeast, budding yeast and animal cells. Yeast 9: 1039-1052

34. Mishra S, Hamburger AW (1993) A microtiter enzyme-linked immunosorbent assay for pro­tein tyrsoine phosphatase. Biochimica et Biophysica Acta 1157:93-101

35. Gu M, Warshawsky I, Majerus PW (1992) Cloning of a cytosolic megakaryocyte protein-ty­rosine-phosphatase with sequence homology to retinaldehyde-binding protein and yeast SECI4p. Proc Natl Acad Sci USA 89:2980-2984

102 1. Kruger, T. Fukushima, G.P. Downey

36. Roach T, Slater S, Koval M et al (1997) CD45 regulates Src family member kinase activity associated with macrophage integrin-mediated adhesion. Current Biology 7:408-417

37. Burns CM, Sakaguchi K, Appella E, Ashwell JD (1994) CD45 regulation of tyrosine phos­phorylation and enzyme activity of src family kinases. 1 Bioi Chern 269: 13594-135600

38. Yi T, Mui AL, Krystal G, Ihle IN (1993) Hematopoietic cell phosphatase associates with the Interleukin-3 (IL-3) receptor beta chain and down-regulates IL-3-induced tyrosine phospho­rylation and mitogenesis. Mol Cell Bioi 13:7577-7586

39. Bignon IS, Siminovitch KA (1994) Identification of PTPIC mutation as the genetic defect in motheaten and viable motheaten mice: A step toward defining the roles of protein tyrosine phosphatases in the regulation of hemopoietic cell differentiation and function. Clin Immunol Immunopath 73: 168-179

40. Dong Q, Siminovitch KA, Fialkow L et al (1999) Negative regulation of myeloid cell prolif­eration and function by the SH2 domain-containing tyrsoine phosphatase-I. 1 Immunol 162:3220-3230

Treatment of Sepsis and Endotoxemia by Extracorporeal Endotoxin Adsorption with Immobilised Human Serum Albumin

K. REINHART, M. ZIMMERMANN

Endotoxin and sepsis syndrome Endotoxin (lipopolysaccharides = LPS) are heat-stable amphiphilic macromole­cules located on the outer cell wall of Gram-negative bacteria. The LPS mole­cule consists of three structure elements [1]: the lipid A, a glucosamine unit that contains fatty acids with a 10- to 20-carbon-atom chain length, a carbohydrate core, and the polysaccharide 0 antigen with repeating sequences of either linear or branched oligosaccharides which vary in chain length among the various strains of bacteria [2]. Studies using X-ray crystallography suggest that the lipid A component of the molecule is in a highly ordered conformation within the outer membrane of Gram-negative bacteria and it is relatively concealed within this membrane, where it presumably has an important role in maintaining struc­tural integrity.

LPS released from bacteria in vivo or administered in an isolated form exert powerful pathophysiological effects in higher organisms and thus represent im­portant virulence factors in Gram-negative sepsis. The lipid A substructure has been shown to constitute the endotoxic and biologically active principle of the LPS molecule [3,4]. The endotoxicity of LPSllipid A is thought to be based on the excessive cascade-like induction of highly active endogenous mediators [5] as a result of the activation of various cell systems [6] such as monocytes/ macrophages, endothelial cells, lymphocytes, granulocytes, fibroblasts and smooth muscle cells and the activation of humoral cascade systems such as the coagulation and the complement system [7, 8]. The initial binding to the CD14 receptor of the macrophages/monocytes and following signal transduction me­diated by different receptor systems primarily leads to release of tumour necro­sis factor alpha (TNF-a) and interleukins such as IL-l, IL-6 and IL-8 [9-11]. These in tum induce the release of a large number of secondary mediators such as histamine, prostaglandins, leucotrienes, platelet activating factor, reactive oxygen intermediates, nitric oxide, and probably others as yet undetermined [12-14]. This inflammatory cascade combines to induce septic shock and multi­organ failure in the presence of systemic Gram-negative bacterial infection and endotoxemia.

From the clinical aspect the precise role of LPS in the generation of multior­gan failure and septic shock remains a subject of some controversy. Elevated

104 K. Reinhart, M. Zimmermann

plasma LPS levels are more frequent in patients with sepsis syndrome, septic shock, and organ failure [15,16] but do not always clearly predict outcome [16,17]. The latter can be explained by genomic polymorphism in host response [18] and the different morbidity depending on age and the underlying disease. Another still existing problem is the accurate measurement of endotoxin [19]. Nevertheless LPS administered to human volunteers in experimental studies at a dose range of 0.5-4 ng/kg resulted in rapidly elevated levels of stress hormones like epinephrine, cortisol and ACTH accompanied by profound monocytopenia [20] and followed by hypothermia, chills, increased cardiac index and heart rate [21]. Within the first few hours after administration changes in the coagulation and fibrinolytic pathway [22] as well as disturbed splanchnic and peripheral tis­sue metabolism were also observed [23]. These effects are similar to the clinical pathophysiological situation seen in septic patients and also underline the con­cept of LPS as one important trigger of the sepsis syndrome.

Extracorporeal blood purification in critically ill patients Continuous haemofiltration (CHF), introduced by Kramer et al. 1977 [24], has been applied safely and successfully to manage renal failure in severe sepsis and MOP. This technology has also been proposed for removal of sepsis associ­ated inflammation mediators. Improvements in cardiovascular and pulmonary function have been observed in patients undergoing arteriovenous haemofiltra­tion (CAVH), independent of fluid removal [25] but CAVH did not improve sur­vival in a canine model of septic shock [26]. Heering and co-workers [27] demonstrated in 33 ventilated patients with acute renal failure of septic and car­diovascular origin that continuous venovenous haemofiltation (CVVH) does re­move TNF-a and cytokines from the blood. Improvements of cardiovascular and haemodynamic parameters were also observed. Nevertheless in this study there was no evidence that CVVH decreases cytokine blood levels. Up to now no clinical study clearly demonstrates beneficial effects of CHF with regard to outcome.

More promising clinical results were reported from selective extracorporeal adsorption according to the principle of affinity chromatography. For e.g. endo­toxin adsorption by direct haemoperfusion over Polymyxin B immobilised on the surface of synthetic material led to a decrease of plasma endotoxin levels in septic patients with simultaneous improvements mainly of cardiovascular and haemodynamic parameters as well as an improved outcome in clinical studies with low patient numbers [28, 29].

An endotoxin absorber based on immobilised human serum albumin (iHSA) has recently been described by our group [30]. The absorber uses the LPS bind­ing properties of human serum albumin to eliminate endotoxin also by direct haemoperfusion. The absorber material consists of a synthetic carrier (poly­methacrylate) on which human serum albumin is immobilised. During the pro-

Treatment of Sepsis and Endotoxemia by Extracorporeal Endotoxin Adsorption 105

duction of the absorber material the ligand is covalently coupled to the surface of the polyacrylic beads. The immobilisation process leads to a significantly in­creased affinity for endotoxin compared to native albumin. The absorber shows effective removal of LPS and lipid A from blood and plasma. In the following we would like to describe first clinical experience with iHSA in patients with severe sepsis and septic shock and confirmed endotoxemia.

Endotoxin adsorption over immobilized human serum albumin (iHSA)

Endotoxin adsorption properties and feasibility of iHSA treatment under clini­cal conditions were investigated in six patients (Table 1) with Gram-negative sepsis and confirmed endotoxemia. Endotoxin measurement was performed us­ing a kinetic limulus amoebocyte lysate (LAL) assay as described elsewhere [30]. A LPS concentration above 20 pg/ml within a period of 24 hours before first treatment (baseline) was regarded as confirmed endotoxemia. Three to four hour adsorption treatments (1 per day) with 500 ml of iHSA were performed 2 to 6 times per patient (23 treatments in total) with a blood flow of 50 mllmin. Treated blood volume per session was 1.5 times total individual blood volume (6-10 L). The extracorporeal circuit (Fig. 1) was maintained by continuous cit­rate anticoagulation.

Table 1. iHSA treated patients with confirmed endotoxemia *

Patient Age Sepsis APACHE Treat-No (years) Source II ments

46 Pneumonia 23 2 2 52 Pneumonia 24 3 3 35 Peritonitis 27 6 4 66 Pneumonia 29 5 5 42 Peritonitis 20 4 6 79 Pneumonia 18 3

* Plasma endotoxin concentration> 20 pg/ml

Extracorporeal endotoxin adsorption over immobilized human serum albu­min was well tolerated. There were no coagulation problems and no other tech­nical problems in maintaining the extracorporeal circuit. Under the study condi­tions blood endotoxin concentrations were reduced by about 37% from 280 pg/ml (mean) to 177 pg/ml directly before/after the first three to four treatments (Table 2) and further decreased by 37% at mean after the following treatments.

106 K. Reinhart, M. Zimmermann

Hemoadsorptlon unit 4008 ADS

~

1 I

'----- ~ ... L ;, -.'.~ __ -ry-:.. ..... .1

Fig. 1. Operating flow chart of the extracorporeal circuit used with an absorber based on immo­bilised human serum albumin (iHSA)

Table 2. Changes of pre/post adsorption plasma endotoxin values

Endotoxina * [pg/ml]

Treatment Patients pre/post Mean No [n] adsorption change [0/0]

6 Baseline 73 pg/ml 6 2801177 - 37

2 6 36/37 + 3 3 5 33/21 - 36 4 3 74/19 - 74 5 2 54/29 - 46 6 (26/ND) ** (-100)

Mean: - 37

* Mean of three repeated measurements * * ND= not detected, detection limit of the LAL assay around 8 pglml

All 6 critically ill patients included survived at day 28 and were discharged ei­ther from the leu or the hospital.

Figure 2 depicts the impact of endotoxin adsorption in a patient after heart failure and severe cardiogenic shock after heart transplantation. The patient ex­hibited major elevations of procalcitonin (100 nglrnl) which is known to in-

Treatment of Sepsis and Endotoxemia by Extracorporeal Endotoxin Adsorption

Endotoxin EU/ml

2.0

1.5

1.0

0.5

G. , \ \. ,

peT ...... peT 100 ng/ml

-e- Endotoxin 100 90 80 70 60 50 40 30 20 10

O.O...L-~-::-----'9'--~---€...--~---e!T--~O before before '" apheresis I

NE 0,65 0,33 Epi 0,55 0,05

0,33 0,05

019 t[1-I9/kg/ml1 0'05 after apheresis

107

Fig. 2. Apheresis by endotoxin absorber in patient with cardiogenic shock after heart transplanta­tion. NE = norepinephrine; Epi = epinephrine; peT = procaicitonin. Treatments by apheresis were performed on three consecutive days over a 4-hour period

crease in patients with endotoxemia [31]. It is well known that both cardiopul­monary bypass and cardiogenic shock may result in endotoxemia most likely due to intestinal translocation of endotoxin and/or bacteria [32, 33 J.

Despite high doses of vasosopressors (norepinephrine 0.65 ng/kg/min and epinephrine 0.55 ng/kg/min) the patient was haemodynamically unstable and re­mained in a state with low cardiac output. That is why he was scheduled for the implantation of a cardiac assist device. Several hours before the patient was tak­en to the operating theatre for this procedure epheresis by the endotoxin ab­sorber was performed for a 4-hour period which resulted in a marked haemody­namic stabilization of the patient that also allowed the reduction of both norepi­nephrine and epinephrine from 0.65 to 0.33 and 0.55 to 0.05 ng/kg/min respec­tively. The scheduled surgery for the implantation of the assist device was can­celled although the patient had already been referred to the operating room for this procedure.

As depicted in Figure 2 retrospective analysis of endotoxin plasma levels af­ter the first as well as after the following aphereses that were performed during the following days resulted in complete elimination of plasma endotoxin. Due to progressing further haemodynamic stabilization the vasopressor and inotropic support could be further reduced.

Currently the hypothesis is tested within the framework of a European multi­center study that apheresis by this new endotoxin adsorber results in an earlier reduction of the APACHE II score and improvements of organ dysfunctions in patients with severe sepsis and septic shock, in comparison to a control group of patients receiving standard treatment only.

108 K. Reinhart, M. Zimmermann

References 1. Rietschel ET, Kirikae T, Schade FU et al (1994) Bacterial endotoxin: Molecular relationships

of structure to activity and function. FASEB J 218:217-225 2. Luderitz 0, Freudenberg MA, Galanos C et al (1982) Lipopolysaccharides of gram-negative

bacteria. In: Razin S, Rottem S (eds) Current topics in membranes and transport. Academic, New York, vol 17, pp 79-151

3. Ziihringer U, Lindner B, Rietschel ET (1994) Molecular structure of lipid A, the toxic center of bacteriallipopolysaccharides. Adv Carbohydr Chern Biochem 50:211-276

4. Galanos C, Luderitz 0, Rietschel ET et al (1985) Synthetic and natural Escherichia coli free lipid A express identical endotoxic activities. Eur J Biochem 148: 1-5

5. Morrison DC, Ulevitch RJ (1978) The effects of bacterial endotoxins on host mediation sys­tems. Am J PathoI93:526-617

6. Galanos C, Freudenberg MA, Katschinski T et al (1992) Tumor necrosis factor and host re­sponse to endotoxin. In: Ryan JL, Morrison DC (eds) Immunopharmacology and pathophysi­ology. CRC Press, Boca Raton, pp 75-102

7. Morrison DC, Cochrane CG (1974) Direct evidence for Hageman factor (factor XII) activa­tion by bacteriallipopolysaccharides (endotoxins). J Exp Med 140:797-811

8. Van Deventer SJH, Buller HR, ten Cate JW et al (1990) Experimental endotoxemia in hu­mans: Analysis of cytokine release and coagulation, fibrinolytic, and complement pathways. Blood 76:2520-2526

9. Wright SD, Ramos RA, Tobias PS et al (1990) CD14, a receptor for complexes of lipopolysaccharide (LPS) and LPS binding protein. Science 249:1431-1433

10. Nathan CF (1987) Secretory products of macrophages. J Clin Invest 79:319-323 11. Vick JA, Hehlman B, Heiffer MH (1971) Early histamine release and death due to endotoxin.

Proc Soc Exp BioI Med 137:902-906 12. Luderitz T, Brandenburg K, Seydel U et al (1989) Structural and physicochemical require­

ments of endotoxins for the activation of arachidonic acid metabolism in the mouse peritoneal macrophages in vitro. Eur J Biochem 179:11-16

13. Braquet P, Touqui L, Shen TY, Vargaftig BB (1987) Perspectives in platelet-activating factor research. Pharmacol Rev 39:97-112

14. Warren JS, Kunkel SL, Cunningham TW (1988) Macrophage-derived cytokines amplify im­mune complex-triggered oxygen responses by rat alveolar macrophages. Amer J Pathol 130:498

15. Casey LC, Balk RA, Bone RC (1993) Plasma cytokine and endotoxin levels correlate with survival in patients with sepsis syndrome. Ann Int Med 119:771-778

16. Guidet B, Barakett V, Vassal T et al (1994) Endotoxemia and bacteremia in patients with sep­sis syndrome in the intensive care unit. Chest 106: 1194-1201

17. Fugger R, Hamilton G, Rogy M et al (1990) Prognostic significance of endotoxin determina­tion in patients with severe intraabdominal infection. J Infect Dis 161: 1314-1315

18. StUber F, Petersen M, Bokelmann F, Schade U (1996) A genomic polymorphism within the TNF locus influences plasma tumor necrosis factor - concentrates and outcome of patients in severe sepsis. Crit Care Med 24:381-384

19. Hurley JC (1995) Endotoxemia: Methods of detection and clinical correlates. Clin Microbiol Rev 8:268-292

20. Richardson RP, Rhyne CD, Fong Y et al (1989) Peripheral blood leukocyte kinetics following in vivo lipopolysaccharide (LPS) administration to normal human subjects. Influence of elicited hormones and cytokines. Ann Surg 210:239-245

21. Suffredini AF, Fromm RE, Parker MM et al (1989) The cardiovascular response of normal humans to the administration of endotoxin. N Engl J Med 321 :280-287

22. Suffredini AF, Harpel PC, Parrillo JE (1989) Promotion and subsequent inhibition of plas­minogen activation after administration of intravenous endotoxin to normal subjects. N Engl J Med 320:1165-1172

Treatment of Sepsis and Endotoxemia by Extracorporeal Endotoxin Adsorption 109

23. Fong Y, Marano MA, Moldawer LL et al (1990) The acute splanchnic and peripheral tissue metabolic response to endotoxin in humans. J Clin Invest 85: 1896-1904

24. Kramer A, Wigger W, Rieger J (1977) Arteriovenous haemofiltration: A new and simple method for treatment of over hydrated patients resistant to diuretics. Klin Wochenschr 55: 1121-1122

25. Groeneveld ABJ (1990) Septic shock and multiple organ failure: Treatment with hemofiltra­tion. Int Care Med 16:489-490

26. Freeman BD, Yatsiv I, Natanson C et al (1995) Continuous arteriovenous hemofiltration does not improve survival in a canine model of septic shock. J Am Coll Surg 180:286-292

27. Heering P, Morgera S, Schmitz FI et al (1997) Cytokine removal and cardiovascular hemody­namics in septic patients with continuous venovenous hemofiltration. Intensive Care Med 23: 288-296

28. Aoki H, Kodama M, Tani T, Hanasawa K (1994) Treatment of sepsis by extracorporeal elimi­nation of endotoxin using polymyxin B-immobilized fiber. Am J Surg 167:412-17

29. Tani T, Hanasawa K, Endo Yet al (1998) Therapeutic apheresis for septic patients with organ dysfunction: Hemoperfusion using a polymyxin B immobilized column. Artif Organs 22 (12): 1038-1044

30. Zimmermann M, Busch K, Kuhn S, Zeppezauer M (1999) Endotoxin adsorbent based on im­mobilized human serum albumin. Clin Chern Lab Med 37:373-379

31. Dandona P, Nix D, Wilson MF (1994) Procalcitonin increase after endotoxin injection in nor­mal subjects. ICE & M 79: 1605-1608

32. Niebauer J, Volk HD, Kemp M et al (1999) Endotoxin and immune activation in chronic heart failure: A prospective cohort study. The Lancet 353: 1838-1842

33. Martinez-Pellus AE, Merino P, Bru M et al (1997) Endogenous endotoxemia of intestinal ori­gin during cardiopulmonary bypass. Intens Care Med 23: 1251-1257

Hemofiltration in Intensive Care

G. BERLOT, M. VIVIANI

Since the very beginning of the medical era, the occurrence of several diseases has been attributed to the action of some endogenous or exogenous toxins, whose removal from the organism was warranted in order to re-establish a nor­mal condition. Although in more-recent times this approach has been aban­doned or largely modified, there is an ever-increasing number of clinical condi­tions associated with the presence of biologically active substances, whose ac­cumulation can be considered either a marker of a determined disease (e.g., creatinine or urea during acute or chronic renal failure) or a causative agent (e.g. autoantibodies, immunocomplexes) [1]. The discovery of "toxic" sub­tances has prompted the development of purifying techniques such as he­modialysis (HD), peritoneal dialyis (PD), plasmapheresis (PP), and plasma ex­change (PE), which have been and/or are currently used to treat different clini­cal conditions. In the critical care setting, in which both sepsis and acute renal failure (ARF) are particularly common and often associated with a poor prog­nosis, the blood purifying approach is particularly appealing. In the last 2 decades, several techniques have been aimed at removing fluids and waste products from the organism without creating cardiovascular problems. These techniques are based on convection, which implies the removal of electrolytes, urea, creatinine, and other waste products with microvolumes of plasma water [2, 3]. In the current scientific literature, convective-based techniques are de­scribed as ultrafiltration or hemofiltration. Although these terms are often used as synonyms, there is a relevant conceptual difference, as the former indicates the removal of fluid that is not fully replaced (thus obtaining a negative fluid balance), whereas with the latter procedure the removed fluid is fully replaced with dedicated solutions, whose composition can vary according to the clinical indications. Independent of these basic characteristics, these techniques do not allow the removal of substances with elevated molecular weights (MW) and protein-bond substances.

The aim of this review is to illustrate the principles of functioning, the cur­rent indications, and the clinical experience with convection-based techniques in the treatment of intensive care unit (leU) patients.

112

Principles of functioning

Hemofiltration and ultrafiltration

G. Berlot, M. Viviani

Intermittent HD can exert deleterious cardiorespiratory effects on critically ill patients, due to 1) the interaction between the blood and the extracopropreal cir­cuit, leading to the activation of leukocytes with the subsequent release of in­flammatory mediators with cardiodepressant and vasodilating properties and 2) the rapid removal of large amounts of fluids and solutes from the intravascular compartment, which is too fast to allow a compensatory refilling from the inter­stitial space. The introduction of more-biocompatible membranes and the use of bicarbonate instead of acetate as buffer in the dialysis bath allowed an increased tolerability, but this is insufficient in hemodynamically unstable patients [4, 5]. For these reasons, since the early 1980s, a number of extracorporeal renal puri­fying techniques have been developed, which, despite some external similari­ties, are very different from HD [2, 3]. In the latter, a semipermeable membrane separates the patient's blood from a fluid whose composition basically deter­mines both quantitatively and qualitatively the purifying process and its velocity (Fig. 1). The escape of solutes occurs along a concentration gradient, as the dial­ysis fluid contains physiological or paraphysiological amounts of sodium, and is poor in potassium and other substances retained in excessive amounts in during ARF or chronic renal failure (CRF). The membranes commonly used in HD are not particularly permeable to water, whose removal occurs mainly by applying a pressure gradient between the two sides of the membrane and/or varying the concentration of sodium or other osmolatically active substance in the dialysis fluid. In the diffusion-based transport, the passage of a determined substance is basically related to its MW: low MW molecules (e.g., urea, MW = 60) are re­moved more efficiently than those with a higher MW (creatinine, MW = 113) [6]. These features can explain, among others, the pathogenesis of some abnor­malities commonly observed in CRF patients undergoing HD, such as the accu­mulation of phosphate and of other toxins with a MW exceeding the cut-off val­ue of the membrane (so-called middle MW molecules) [4, 5]. The more recently introduced convective techniques, which have been made possible by the devel­opment of more-biocompatible and more-water permeable membranes, caused a revolution in the treatment of ARF in ICU patients. With these techniques, water is removed more gradually from the bloodstream, and the gradual depletion of fluid and solutes from the intravascular space is compensated by the contempo­rary, gradual refilling from the interstitial space (Fig. 2). The cut-off MW of the membranes commonly used in the convective treatments is 50-60 kilodaltons, allowing an increased clearance of substances with a middle MW compared with conventional HD.

Initially, the process was based on the arterovenous gradient of pressure by interposing a filter between the arterial and the venous side (continuous arterovenous hemofiltration, CAVH). As a consequence, the overall efficacy of the treatment was based 1) on the presence of an adequate pressure gradient be-

Hemofiltration in Intensive Care 113

interstitial space intravascular space dialysate

m HC03-

Na+ CI- Na+ CI-K+ K+ K+

Creat Creat N

m ~~ N N Creat

P04-- P04-- N

Fig. 1. Diffusive transport: the passage of solute from the blood to the dialysis fluid occurs along a concentration gradient

interstitial space intravascular space ultrafiltrate

Na+ CI-K+

Creat, Na+ CI- Na+ CI-N K+

~~ K+

Creat Creat Na+ CI- N N

K+ Creat

N

Fig. 2. Convective transport: each box represents a volume of water

114 G. Berlot, M. Viviani

tween the two sides and 2) the characteristics of the blood flowing inside the fil­ter (Qb), which, in tum, were influenced by its viscosity and the hydrodynamic properties of the filter. Actually, according to Starling's mechanism, the increase of protein concentration related to the progressive loss of water determines the increase of the oncotic pressure to the point beyond which the filtration ceases. To overcome these limitations and to avoid iatrogenic complications deriving from arterial cannulation, this technique has been modified with the interposi­tion of a roller pump into the extracorporeal circuit, thus making the use of arte­rial cannulae needless. This modification, the continuous venovenous hemofil­tration (CVVH), has made CAVH obsolete. Further developments basically con­sisted in the combination of the convection with the diffusion, the continuous venovenous hemodiafiltration (CVVHD), which is accomplished by counter­current flow of a dialysis fluid in the external part of the filter chamber. In prac­tical terms, ultrafiltration is indicated when the simple removal of fluids is nec­essary, as in clinical conditions characterized by the incapacity to eliminate fluid in excess. Conversely, hemofiltration is usually a more-complicated and pro­longed procedure, basically consisting in the removal and replacement of fluid at an overall zero balance. With this latter approach, the amount of fluid re­moved and replaced, ultimately determines the purification capabilities. The in­troduction of these techniques constituted the basics of continuous renal re­placement therapy (CRRT) [2]. Basically, these techniques differ in terms of both purifying technique(s) and volume exchanged (Table 1).

The main advantages associated with the convective techniques are [2]: - a 24-h treatment, which can be continued indefinitely; although CVVH is less

effective per unit of time in terms of electrolytes and catabolic wastes re­moval compared with HD, this latter is associated with marked water and ion­ic disturbances that could be harmful in critically ill patients; conversely, dur­ing CVVH, there is no decrease in the intravascular water content as it is con­tinuously replaced by the fluid from the interstitial space;

- the continuous removal of water and electrolytes allows the liberal adminis­tration of fluids, including enteral and parenteral nutrition, drugs, blood, and derivates;

- the removed fluid can be totally or partially replaced with solutions whose composition varies according to the clinical needs; this is a particularly rele­vant feature of the convective techniques, as it allows of the change the "in­ternal milieau". However, the description of these techniques would be largely incomplete

without the description of their inherent limitations, which are: - the relatively reduced purifying abilities per unit of time compared with con­

ventional HD; this contraindicates their use to treat most exogenous poison­ing, some reports suggest a possible role in selected patients [7, 8];

- the need of large-bore intravenous cannulas, with the consequent risk of hem­orrhage, thrombosis, infections, and air embolism;

Hemofiltration in Intensive Care 115

Table 1. Technical details and main indications of some techniques used in continuous renal replacement therapy

Denomination Qb Qf Qd K Clinical indications

Slow continuous ultrafiltration (SCUF) 20-100 2-5 Fluid overload, CHE AHF Continuous hemofiltration (CAVH, CVVH) 20-200 8-25 12-36 ARF (sepsis?) Continuous hemodialysis (CAVHD-CAVHD) 50-200 2-4 10-20 14-36 ARF (sepsis'l) Continuous hemodialfiltration (CAVHDF-CAVVHDF) 50-200 10-20 20-40 ARF (sepsis'l) Continuous high-flux dialysis (CAVHDF-CVVHFD) 50-200 2-8 50-200 40-60 ARF (sepsis?)

Qh blood 110v. (ml IminJ. Qf"ultraliltration rate (ml Iminl. Qd dialysate 110\\ (ml Imin). K clearance (1124 day). CHF chronic heart failure. AHF acute heart failure. ARF acute renal failure

similar to other treatments performed using extracorporeal devices, it is nec­essary to administer heparin to prevent blood coagulation inside the lines and/or the filter itself; this obviously limits their use in patients at hemorrhag­ic risk; however, there are some way to limit the risk of bleeding; first, the an­ticoagulation can be restricted to the extracoroporeal circuit r91; second, he­parin itself can be avoided by the use of other anticoagulants, including prostacyclin [10], prostaglandin E 1, and low MW heparins [11, 12]; third, re­cently heparin has been bound to the connecting lines and the filter by the manufacturer.

Clinical applications Three main fields of applications have been developed for convective tech­niques:

the support of renal function [2, 3];

the treatment of congestive heart failure and/or overhydration states not re­sponsive to other measures, including fluid restriction and diuretics [13, 14];

the treatment of sepsis; this latter indication is based on the assumption that many mediators of sepsis can be neutralized from the bloodstream through transfilter removal or absorption on the filter membrane [15].

The support of renal function The occurrence of ARF in critically ill patients represents a particularly harm­ful complication for several reasons [16]. First, in this setting ARF is often as-

116 G. BerIot, M. Viviani

sociated with the failure of other organs and systems; under these circum­stances, the mortality rate can peak: at 80-100%, in sharp contrast to that recorded when ARF occurs in isolation, ranging from 15 to 20%. Second, ICU patients are exposed to a host of risk factors, including the use of nephrotoxic agents, low cardiac output states, and the action of endogenous nephrotoxins (e.g., myoglobin). Third, HD in and by itself can cause or aggravate cardiores­piratory disturbances.

Then, it appears that CVVH and derived techniques represent the treatment of choice for ICU patients with ARF [2]. Actually, some investigators [13] have demonstrated a better hemodynamic tolerance compared with HD, which could be still necessary in cases of severe hyperkalemia or when the treatment is initiated in such an advanced phase of ARF that the convective mechanism alone is no longer able to sufficiently treat the azotemia and the related symp­toms. The maintenance of cardiovascular stability appears to be more marked in hemodynamically unstable patients, as in a cross-over study the authors did not find any change in cardiovascular variables when normotensive patients were shifted from HD to CAVH and viceversa [17]. Furthermore, there is a direct relationship between the intensity of the treatment and the outcome. Stock et al. [18] demonstrated that critically ill patients with ARF treated with CVVH had a better outcome than those treated with CAVH, and attributed this effect to the reduced efficacy of this latter technique, both in terms of re­nal support and clearance of sepsis mediators. More recently, Ronco et al. [19] were able to demonstrate a "dose-effect relationship" in three groups of ARF patients treated with CVVH at different Qf (20, 35 or 45 ml/kg per hour-I, respectively), since patients treated at the lowest level had a signifi­cantly worse outcome than those in two other groups. Thus, due to the very principle of functioning, the effective treatment of the ARF with these tech­niques requires the exchange of substantial amounts of fluids, reaching > 40 IJday in critically ill patients. This large shift of fluids implies some practical difficulties that cannot be overlooked when initiating a treatment. First, the amount of fluids exchanged requires a careful evaluation of the in- and out­balance, in order to avoid the risk of over-or underplacement of the removed fluid: this goal can be easily accomplished with a computerized weight con­trol of both the collecting and infusion bags. Second, drugs, nutrients, vita­mins, and hormones can be lost during the procedure [20]; to overcome these shortcomings, the dosage of some relevant drugs should be adjusted by re­peatedly measuring their blood levels or by using the available tables [21]. Lastly, the passage of the blood along the membrane of the filter is associated with the deposit of a proteinaceous film, which, after several hours, reduces its permeability and thus the efficacy of the purification. Although this process is influenced by individual factors, including the velocity of the blood flow and its viscosity, change of the filter every 24-36 h is indicated to main­tain the filter capabilities.

Hemofiltration in Intensive Care 117

Treatment of heart failure and fluid overload The mainstay of the treatment of acute and chronic heart failure (AHF and CHF, respectively) is the administration of drugs aimed at improving the cardiac out­put and reducing the excess fluid accumulated in the interstitial space. This lat­ter point is particularly relevant since: - the appearance of acute pulmonary edema represents a potentially lethal

complication; - the accumulation of tissue edema causes the increase of the distance separat­

ing capillaries from the cells, thus exposing them to a reduced local P02 and the consequent risk of dysoxia. The main goal of the treatment of AHF and CHF is increase of the cardiac

output, with the subsequent blunting of the neuroendocrine response involving the sympathetic nervous system and the renin-angiotensin-aldosterone system, whose combined action ultimately leads to the increased renal reabsorption of sodium and water [22]. In this setting, the restriction of water and sodium in­take, possibly associated with the use of diuretics, is indicated to counteract the interstitial fluid accumulation, whose severity can range from a modest pretibial edema to acute pulmonary edema, ascites, and anasarca. Although this approach is usually effective, there are some clinical settings in which the elimination of the excess fluid cannot be achieved pharmacologically (e.g., oligoanuric ARF and CFR) or the fluid limitation is hard to achieve (e.g., burn and trauma pa­tients). In these circumstances, convective techniques facilitate the management of fluids, as part of the excess fluid can be eliminated without hemodynamic de­rangements, which are prevented by the continuous refilling from the interstitial space [14]. Since these conditions are short lasting, the use of CVVH and de­rived techniques can be limited to the time during which large amounts of fluids are likely to be infused. As an example, CVVH and related techniques can be employed in patients undergoing cardiac surgery during the last phase of the ex­tracorpoeral circulation to eliminate the excess fluid, thus preventing the occur­rence of postoperative pulmonary edema [23].

Treatment of sepsis

The wide array of metabolic and cardiorespiratory disturbances commonly seen in septic patients is determined by the production and release of inflammatory mediators triggered by the interaction between the host and the infecting organ­isms [24]. A number of cells are involved in this process, including neutrophils, lymphocytes, circulating and fixed macrophages, and endothelial cells. Al­though in the early 1990s they were supposed to invariably cause a diffuse in­flammatory response, later it became clear that many substances produced dur­ing sepsis actually exert a powerful anti-inflammatory action, aimed at counter­balancing and thus limiting the effects of the initial systemic inflammatory reac­tion. In the past few years, two strategies have been developed to counteract the

118 G. BerIot, M. Viviani

deleterious effect of the pro-inflammatory septic mediators. The first involves the administration of substances theoretically able to blunt their action [25], but, despite the promising experimental results, results of clinical studies have been substantially disappointing. The second approach is based on the removal of the mediators from the bloodstream by means of CVVH and derived techniques [2, 15]. To be clinically effective in this setting, a blood purifying treatment must fulfil some criteria. First, the cut-off value of the membrane used must allow the passage of the high MW mediators or its chemico-physical properties should al­low their adsorption on the filter surface [26]. Second, sepsis mediators must be present in the bloodstream at the time of the procedure, as both too early or too late treatment can be ineffective. In the former case the triggering of the media­tors has not yet occurred, whereas in the second, if the end-organ damage has already occurred, any attempt to decrease their concentration could be futile [15,27,28]. There are both experimental and clinical results indicating that con­vective techniques can either ameliorate some physiological variables altered by sepsis mediators and reduce the mortality of treated subjects. In experimental sepsis, CVVH at zero balance was associated with a significant improvement of the considered parameters [29]. Moreover, the intensity of treatment can influ­ence the outcome, as in septic patients with ARF the CVVH was associated with a better outcome than CAVH, and this effect has been attributed to a more­effective removal of mediators obtained with the former technique [18]. Howev­er, the role exerted by CVVH and derived techniques on blood purification from sepsis mediators remains largely undetermined, as measurements of their con­centration in the bloodstream of the treated patients are conflicting: a decrease has been demonstrated by some investigators, whereas their levels remained un­changed or even increased in other studies [26]. As an example, some years ago, Bellomo et al. [30] demonstrated that tumor necrosis factor (TNF)a and inter­leukin-I ~ could be effectively removed from the blood of septic patients treated with CVVHD using a AN 69 membrane, and these substances could be recov­ered in substantial amounts in the ultrafiltrate. Conversely, more recently Ro­giers et al. [31] were able to demonstrate that, in septic dogs treated with CVVH using a polysulphone membrane, the improvement of several systemic and re­gional hemodynamic variables was not associated with the removal of TNF, and only minor amounts of this mediator were recovered in the ultrafilrate. Several factors can account for these inconsistencies, including the timeframe and the duration of the treatment, the heterogeneity of the patients enrolled, the different properties of the membrane used, and the volume of fluid exchanged. Each point is worth discussing separately. First, it appears that the secretion of most sepsis mediators is pulsatile, and that their half-life is short. Any treatment will be futile if initiated after the end-organ damage. Second, in sepsis patients a number of factors can influence the outcome independently of the purifying technique used, including the age, the underlying conditions, the number of fail­ing organs, and the overall appropriateness of care. Third, the mechanism of mediator inactivation is probably more complicated than the pure convective

Hemofiltration in Intensive Care 119

clearance previously hypothesized, and the chemico-physical properties of the membrane used probably playa major role in determining the different results currently available in the literature. In a recent study, DeVriese et al. [32] demonstrated in a group of patients with sepsis and ARF treated with CVVH using an AN69 membrane that 1) the decrease of TNF and other sepsis media­tors in septic patients treated with CVVH was largely attributed to the adsorp­tion of these substances of the membrane surface, 2) the absorption was time dependent, being maximal in the first few hours after the beginning of the treat­ment and decreasing later, and 3) the convective removal of cytokines was more pronounced at elevated Qb. Then, one could argue that any attempt to enhance the removal of sepsis mediators by maximizing the ultrafiltrate production will be futile if the membrane used is poorly permeable to these substances but has elevated absorptive capabilities; in this case, rather, the increase of the Qb could be warranted. Then, should this kind of study be repeated to elucidate the ab­sorptive and/or convective capabilities "f other available membranes, it could provide key information on the adjustment of both the Qb and the Qf during the treatment. At present, however, it is not clear if CRRT alone or in combination with other techniques (e.g., plasmapheresis or PE) is valuable for attenuation of the action of the sepsis mediators, and a large, randomized multicenter trial has been advocated [33].

Conclusions The techniques based on the convective transport of solutes have gained wide­spread popularity among European intensivists. The established indications basically reflect those of HD, but they are associated with better hemodynam­ic tolerance. There are some suggestions from both experimental and clinical investigations that these techniques can neuralize the sepsis mediators, either via their convective removal of absorption on the filter membrane. Whereas the first indications are widely accepted, making CRRT the treatment of choice among critically ill patients, the second still has to be confirmed in large clinical trials.

120 G. BerIot, M. Viviani

References 1. BerIot G, Tomasini A, Silvestri L, Gullo A (1998) Plasmapheresis in the critically ill. Kidney

Int 53[SuppI66]: 178-181 2. Bellomo R, Ronco C (1999) Continuous renal replacement therapy in the intensive care unit.

Intensive Care Med 25:781-789 3. Forni IG, Hilton PJ (1997) Current concepts: continuous hemofiltration in the treatment of

acute renal failure. N Engl J Med 336: 1303-1309 4. Mallick NP, Gokal R (1999) Haemodialyis. Lancet 353:737-742 5. Port FK (1994) Morbidity and mortality in dialysis patients. Kidney Int 46:1728-1737 6. Dhondt A, Vanholder R, van Biesen W, Lameire N (2000) The removal of uremic toxins. Kid­

ney Int 58[SuppI76]:S47-S59 7. Bellomo R, Kearly Y, Parkin G et al (1991) Treatment of life-threatening lithium toxicity with

continuous arterio-venous hemodiafiltration. Crit Care Med 19:836-837 8. Leblanc M, Pichette V, Madore F et al (1995) n-acetylprocainamide intoxication with torsade

de pointes treated by high dialysate flow rate continuous arteriovenous hemodiafiltration. Crit Care Med 23:589-593

9. Mehta RL, McDonald BR, Aguilar MM, Ward DM (1990) Regional citrate anticoagulation for continuous arteriovenous hemodialysis in critically ill patients. Kidney Int 378:976-981

10. Langenecker-Kozek SA, Felfernig M, Werba A et al (1994) Anticoagulation with prostacyclin and heparin during continuous venovenous hemofiltration. Crit Care Med 27: 1774-1781

11. Langenecker-Kozek SA, Kettner SC, Ouismueller C et al (1998) Anticoagulation with prostaglandin El and unfractioned heparin during continuous venovenous hemofiltration. Crit Care Med 26:1208-1212

12. De Pont ANJM, Oudemans-van Straaten Roozendal KJ, Zandtsra DF (2000) Nadroparin ves­rus dalteparin in high-volume, continuous venovenous hemofiltration: a double blind, ran­domized, crossover study. Crit Care Med 28:421-425

13. Davenport A, Will E, Davidson AM (1993) Improved cardiovascular stability during conti­nous modes of renal replacement therapy in critically ill patients with acute hepatic and renal failure. Crit Care Med 21 :328-338

14. Tortorella G, Gonzi G, Zambrelli.P et al (1997) Continuous veno-venous hemofiltration in acute renal failure associated with heart failure. Cardiologia 42:845-848

15. van Bommel EFH (1997) Should continuous renal replacement therapy be used for "non re­nal" indications in critically ill patients with shock? Resuscitation 33:257-270

16. Liano F, Junco E, Pascual J (1998) The spectrum of acute renal failure in the intensive care unit compared with that seen in other settings. Kidney Int 53[SuppI66]:S 16-S24

17. Misset B, Timsit JF, Chevret S et al (1996) A randomized cross-over comparison of the hemo­dynamic response to intermittent hemodialysis and continuos hemofiltration in ICU patients with acute renal failure. Intensive Care Med 22:742-746

18. Storck M, Hartl WH, Zimmerer E, Inthorn D (1991) Comparison of pump-driven and sponta­neous continuous hemofiltration in postoperative acute renal failure. Lancet 337:452-455

19. Ronco C, Bellomo R, Hormel P et al (2000) Effects of different doses in continuous veno-ve­nous hemofiltration on outcomes of acute renal failure: a prospective, randomised trial. Lancet 355:26-30

20. Story DA, Ronco C, Bellomo R (1999) Trace element and vitamin concentrations and losses in critically ill patients treated with continuous venovenous haemofiltration. Crit Care Med 27:220-223

21. Schetz M, Ferdinande P, Van der Berghe G et al (1995) Pharmacokinetics of continuous renal replacement therapy. Intensive Care Med 21 :612-620

22. Schrier RW, Abraham WT (1999) Mechanisms of disease: hormones and hemodynamics in heart failure. N Engl J Med 341 :577-585

23. Baudoin SV, Wiggins J, Keogh BF et al (1993) Continuous veno-venous haemofiltration fol­lowing cardio-plmonary bpass. Intensive Care Med 19:290-293

HemofiItration in Intensive Care 121

24. Adrie C, Pinsky MR (2000) The inflammatory balance in human sepsis. Intensive Care Med 26:364-375

25. Cain BS, Meldrum DR, Harken AH, Mcintyre (1998) The physiologic basis for anticytokine clinical trials in the treatment of sepsis. J Am Coli Surg 186:337-350

26. De Vriese AS, Vanholder RC, Pascual M et al (1999) Can inflammatory cytokine be removed efficiently by continuous renal replacement therapies? Intensive Care Med 25:903-910

27. Got1oib L (1996) Hemofiltration in multiorgan failure syndrome secondary to sepsis: a criti­cal analysis of heterogeneity. Nephron 73: 125-130

28. Schetz M, Ferdinande P, Van Den Berghe Verwaest C, Lauvers P (1995) Removal of proin­f1ammatory cytokines with renal replacement therapy: sense or non-sense? Intensive Care Med 21: 169-176

29. Stein B, Pfenninger E, Grunert A et al (1990) Influence of continuous arteriovenous haemofiItration on haemodynamics and central blood volume in experimental endotoxic shock. Intensive Care Med 6:494-499

30. Bellomo R, Ripping P, Boyce N (1993) Continuous veno-venous hemofiltration with dialysis removes cytokines form the circulation of septic patients. Crit Care Med 21 :522-526

31. Rogiers P, Zhang H, Smail N et al (1999) Continuous venovenous hemofiltration improves cardiac function by mechanisms different other than tumor necrosis factor-a attenuation dur­ing endotoxic shock. Crit Care Med 27: 1848-1855

32. De Vriese AS, Colardyn FA, Philippe 11 et al (1999) Cytokine removal during hemofiltration in septic patients. J Am Soc Nephrol 10:846-853

33. Rogiers P (1999) Hemofiltration treatment for sepsis. Is it time for controlled trials? Kidney Intern 56[Suppl72J:S99-S I 03

Sepsis and Organ Dysfunction: An Overview of the New Science and New Biology

A.E. BAUE

The discrepancy between what we know and what we can do What is new and exciting in the new millennium about sepsis? Do we know more about it? Yes, and that is new and exciting. Do we know more about the contributions of sepsis and mediators to organ dysfunction? Yes, and that is new and exciting. Can we use this exciting information to better care for our pa­tients? Not yet. Thus, there is a large discrepancy between how much we know and what we can do. This is a challenge. There are a number of reasons for the gulf between what we know and what we can do [1]. The excitement of molecu­lar biology, genomic studies, and the human genome project cannot as yet be translated into clinical usefulness [2]. The process of science dissects phenome­na and mediators - studies them in isolation - as in genetically pure animals or knockout mice with well-controlled experiments and only one variable. In pa­tient care things are complex. Patients differ in age, sex, ethnicity, prior illness, chronic illness, genetic background, and life (antigen) experiences. These differ­ences are becoming clearer as we learn of genetic polymorphisms and antigen exposures [3]. Sepsis is not a diagnosis or a disease and you cannot treat it. An infection which has become systemic should be called that - an infection with systemic manifestations.

Years ago this process was called "blood poisoning", a term more descriptive and colorful than "sepsis". When a local infection invaded lymphatics, lym­phangitis with red streaks led to fever and malaise, which are now due to cy­tokines. We have lumped together many different diseases. What actually is sep­sis? Is it systemic inflammatory response syndrome (SIRS) or multiple organ dysfunction syndrome (MODS), or inflammation? However you define it, you can only give palliative treatment - drugs to decrease fever, to help one feel bet­ter. A nonsteroidal, anti-inflammatory drug or a COX-2 inhibitor may make you feel a little better, but will not alter the course of a disease leading to multiple organ failure (MOF). If antibiotics are to be used, for what? I believe we should stop lumping all of these together under the rubric "sepsis" and focus on indi­vidual problems of infection with or without systemic manifestations. Treating systemic manifestations without adequately treating the original infection is fu­tile [4]. Before cytokines were given that name, interleukin-l was called leuko­cyte endogenous mediator and produced fever with infection. Tumor necrosis

124 A.E. Baue

factor was called cachectin and induced weight loss in cancer. There are now many complex, interlocking, overlapping, and redundant mediators that damage organs in infection. Some of these phenomena are still unexplained, such as in­creased oxygen in venous blood with infection - is it due to microcirculatory shunting, peripheral edema or mitochondrial dysfunction? The hyperdynamic circulation with infection and decreased peripheral vascular resistance leading to septic shock may be due to metabolites and nitric oxide (NO) production, but all we can do is support the circulation by conventional means. NO synthase in­hibition has not helped as yet. Better understanding of these complexities will increase our knowledge, but mayor may not improve therapy. Organ dysfunc­tion will vary in individual patients because of prior illness or injury [5]. Thera­py must be specific for the infection, both location and cause. Peritonitis, pan­creatitis, a wound infection, ventilator-associated pneumonia (VAP), other noso­comial infections, sinusitis, endocarditis, and catheter infections are all differ­ent. Injuries must be separated, such as blunt trauma, long bone fractures, pene­trating trauma with visceral injury, operative trauma, and infection in various or­gans, such as the urinary tract, the wound, the peritoneal cavity, and the lung, etc. Hopefully, molecular biology will eventually contribute to better patient care. At present we must emphasize treating specific diseases and abnormalities.

Infection and intensive care uiifts A recent issue of Intensive Care Medicine focused on "Infection and critical ill­ness" [1]. The problems reviewed included: microbial resistance [2], prevention and control of intensive care unit (ICU) infections, antibiotic usage [3], media­tor and receptor polymorphisms [4], leukocytes [5], VAP [6], and intravascular catheter sepsis [7]. Of the 20 articles, only 2 used the expression sepsis [7, 8]. Only 1 article referred to SIRS [9]. Perhaps these expressions are on the way out, as they should be. In all these articles, and elsewhere in the literature, a ma­jor current problem is infection-nosocomial infection acquired or brought into the ICU, transmission from patient to patient, contamination of ICUs, infection from intravascular catheters, endotracheal tubes, nasal tubes, sinusitis, and pul­monary contamination from the upper gastrointestinal tract (GI) tract with gas­tric acid suppression.

Are there solutions to these problems? I believe there are and they do not re­quire molecular biology. Bacterial resistance to antibiotics will always be with us [10]. Bacteria will develop ways to survive by genetic changes, as described by Livermore [11]. He said: "There is no simple cure for resistance but the best opportunities for control lie in lesser and better use of antibiotics backed by swifter and more accurate microbiology; in developing new antibiotics; and in protecting old ones from resistant determinants. All this must be supported by good local knowledge of the epidemiology of infections and resistance and the likelihood of particular antibiotics to produce resistance" [11].

Sepsis and Organ Dysfunction: An Overview of the New Science and New Biology 125

Prevention of transmission from one patient to another should be simple -isolation of contaminated, infected patients and standards of hygiene, hand washing, and bactericidal solutions for hand soaking. These are difficult to maintain because of carelessness, staffing problems, urgent situations in the next cubicle, and other matters. We must do a better job. leUs must be kept cleaner. Transmission of organisms from one patient to another should be preventable. Scott [12] indicates that this problem is less frequent (better infection control) in small leUs, particularly those used mainly for postoperative patients. In larger units with different patients, inadequate staffing, and frequent antibiotic use, chronic colonization and infection are inevitable. The length of time a patient spends in an leu also contributes to colonization.

I found no articles in the literature on how to keep leUs clean. Clean walls, floors, ceilings, equipment, and air handling are important features in an operat­ing room to protect an open wound. This is important also for the susceptible, perhaps immunosuppressed leu patient. I believe we must rethink our Ieus and how they are used.

A modest suggestion I recommend a paradigm shift to a concept of short- and long-term IeUs. The short-term leu would be a conventional unit, with street clothing, visitor traffic, clean but not scrupulously so. This unit would be used for observation and mon­itoring for a few days, as with postoperative patients after cardiac operations or other major operations or illnesses. The maximum stay would be about 4 days.

The long-term leu would be for patients with a predicted need for leU care for days or weeks. This would be for patients who are on a ventilator, with or­gan failure and with other serious conditions, where prolonged intensive care seems necessary. This unit should be like an operating room with scrub clothes, caps and masks, air exchange from the top of the room to the bottom, gowns and gloves, visitors limited, and zip-on clothing covers with caps and masks. The walls and floors are scrubbed daily. Each patient is cultured on admission and isolated as needed. Personnel are cultured periodically. Travel from one pa­tient to another requires a change of gown and gloves and with gowns for each patient.

I recognize problems with this approach. It is expensive in personnel, equip­ment, supplies, and facilities in a cost-conscious time. However, we could have a phase-in period. Ambulatory surgery may allow use of excess operating rooms. The present postanesthesia recovery rooms could be used for short-term IeUs very easily by simply keeping them open all the time. Another option would be to clean up our leUs and make them bacteriologically safe for pa­tients. Jasny and Bloom [13] said about nosocomial infection: "To have invested in science, achieved understanding of the steps that need to be taken, and then failed to act on that knowledge would be folly of the highest order".

126 A.E. Baue

The contest between bacterial resistance to antibiotics, our innate defenses and new antibiotics Resistance of bacteria to antimicrobial agents is a serious problem in ICUs [2, 3, 10, 11]. They frequently cause nosocomial infections and are often lethal. ICUs are a reservoir for them as are nursing homes, as described by Bonten and Wein­stein [14]. They review Levy's five principles of antimicrobial resistance: " ... a) given sufficient time and drug use, antibiotic resistance will emerge; b) resist­ance is progressive, evolving from low levels to high levels; c) organisms resist­ant to one drug are likely to become resistant to others; d) once resistance ap­pears, it is likely to decline slowly, if at all; e) antibiotics in anyone person af­fects others in the extended as well as the immediate environment". Resistant bacteria are unlikely to ever revert back to sensitivity and full usefulness [15]. Major resistant organisms are shown in the Table.

Table 1. Antibiotic resistance

Vancomycin-resistant enterococci (VRE) are endemic [16-20] Methicillin-resistant Staphylococcus aureus (MRS A) are endemic [21] Cephalosporin-resistant Gram-negative organisms [22] Resistant Klebsiella strains [23] Resistant Pseudomonas aeruginosa [23] Extended (broad) spectrum beta-Iactamases [24] on plasmids (ESBLS) - over 43 now described Escherichia coli, Klebsiella, Enterobacter, including Serratia, Pseudomonas, Proteus, Citrobacter Nursing homes are a main reservoir for ESBLS Fluconazole-tolerant fungi [25]

Jasny and Bloom [13] wrote: "The rates of infection and death from hospi­tal-related infection and the fact that antibiotic resistance is spreading in envi­ronments where the public expects vigilance to be at its best are further indica­tors that preventive medicine is underutilized. Roughly 88,000 people in the United States alone die each year as a result of complications from nosocomial infections, a third of which are estimated to be preventable".

The biology of bacterial species to produce disease is better defined now. For example, type IV bundle-forming pili are required for virulence of Escherichia coli [26]. Type III secretion machines in several Gram-negative bacteria are in­volved in their pathogenesis [27]. Invasin, a bacterial integrin-binding protein, promotes bacterial entry into cells by binding to host cell integrins in Yersinia, for example [28]. Bacteria have quorum-sensing systems composed of mole­cules, such as acylated homoserine lactone (acyl-HSL), which allow them to communicate and form large groups [29]. Some bacteria, Mycobacterium tuber­culosis for example, have taken on eight human genes that encode a protein that breaks down hydrogen peroxide from white cells. These bacteria are now safe

Sepsis and Organ Dysfunction: An Overview of the New Science and New Biology 127

from the process of hydrogen peroxide killing bacteria [30]. Costerton et al. [31] described "bacteria that attach to surfaces and aggregate in a hydrated polymeric matrix of their own synthesis to form biofilms". These sessile communities are resistant to antimicrobial agents and contribute to persistent, chronic infections. Biofilms form on any implanted device and cause many nosocomial infections [32]. Many problems are due to biofilms from dental caries to osteomyelitis to valvular endocarditis. Bacteria also have a defense system to kill other closely related strains in a sort of "chemical arms race", as described by Riley. This pro­duces bacterial diversity. The weapons are colicins or bacteriocins which evolve by positive selection [331. Superantigens or bacterial exoproducts are able to by­pass the rate-limiting step of normal antigen processing. Bacteria such as Staphylococcus aureus can produce T cell shock, the toxic shock syndrome, and severe septicemia [34]. The virulence of group A streptococcus causing necro­tizing fasciitis is another example [35].

Innate immunity is a first-line host system serving to limit infection in the early hours after exposure to bacteria, before the adaptive immune system is ac­tivated [36]. Phagocytosis by blood cells begins the process. Then there are pro­teolytic cascades, potent antimicrobial peptides (over 400 in number), and pat­tern recognition molecules, or receptors to recognize bacterial molecular arrays. These include complement, collectins, and a battery of antimicrobial peptides. For lipopolysaccharide there is BPIP and LBP. The second family consists of collectins, which include lung surfactant protein (SP-A) and mannose-binding protein (MBP).

The innate defense system includes a-defensins from granulocytes and Paneth cells [37] in the GI tract. Matrilysin cleaves pro-defensin to form a-de­fensin. p-defensins are found in the genitourinary tract, peripheral blood cells, and skin. They disrupt cytoplasmic membranes of bacteria [38-40] and may also recruit T cells. Cathelicidins are myeloid antimicrobial peptides. The proteolytic cascades begin with the three pathways of complement activation. Study of this innate immune system may lead to new antimicrobial agents and defense mech­anisms for sick and injured patients [41]. An example is the powerful rhesus theta defensin -1 (RTD 1).

New antibiotics continue to be found or developed, but not as rapidly as re­sistant organisms. Linezolid, a new and novel antimicrobial agent, seems effec­tive against resistant Gram-positive bacteria [42, 43]. It is available now for compassionate use and will soon be approved for general use. An antibacterial peptide being studied, nisin Z, kills its target by permeabilizing the plasma membrane [44, 45]. It resembles a peptide antibiotic from frog skin called maigainin [46]. A steroid from sharks, squalamine, was found to kill organisms, particularly Candida albicans, but not much has been heard about it recently [46]. Synercid, a recently approved antibiotic, is effective against methicillin-re­sistant S. aureus.

An agent for cleaning ICUs may be a unique emulsion of oil, water, and two common laboratory detergents (Triton X-IOO and tributyl phosphate), which

128 A.E. Baue

kills Gram-positive and -negative bacteria, fungi, and enveloped viruses [47]. This could be used to keep ICUs cleaner than they previously have been and de­crease cross-contamination.

Prevention of specific abnormalities - support of organ functions before they fail For prevention it is necessary to be specific. What is the problem? What is the disease or process we are dealing with? Many treatments, agents, or drugs to support or prevent may help in some circumstances, but it is 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-drugs to make a difference (multi-agent therapy). There are treatments which 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 hemorrhagic pancreatitis [48], but was no help in general trauma patients [49].

Enteral immunonutrition is a great help in patients having surgery for malig­nancy, whereas it is no help in patients just having major operations [50]. High­dose steroids in patients with end-stage acute respiratory distress syndrome (ARDS), as used by Meduri et al. [51], 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 potential ventilatory failure. The same may be true for patients with severe infection who are having ventilatory difficulty.

Prevention of thrombophlebitis, pulmonary embolism, and GI 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 metoprofol will preserve organ function. Contributors to organ failure must be recognized, such as the ab­dominal compartment syndrome and hypothermia, coagulopathy, and acidosis during abdominal exploration of trauma patients. The importance of a high car­diac output after operation or injury and with infection is critical [52, 53].

Much has been written about support of organ or systems function: lungs, circulation, liver, kidneys, coagulation, the central nervous system, metabolism, musculoskeletal system, etc. There is much we can do. I refer the reader to the many specific recommendations in the literature [52-54].

Thangathurai et al. [55] maintained intraoperative tissue perfusion by nitro­glycerin and fluids in high-risk patients, and in 155 such patients none devel­oped ARDS. Shoemaker et al. [56] used intraoperative evaluation of tissue per-

Sepsis and Organ Dysfunction: An Overview of the New Science and New Biology 129

fusion in high-risk patients by invasive and noninvasive hemodynamic monitor­ing. They suggest that blood flow, oxygen delivery, and tissue oxygenation of patients who do not survive became inadequate during the end of the operation. This suggests potential intraoperative therapy.

Differences in infections Sepsis has been used by many to indicate the systemic effects of an infection somewhere in the body. There are great differences in infections and we must recognize those. For example, mediastinitis is a rapidly lethal infection because of its relationship to the heart and lungs. Necrotizing fasciitis is a very severe problem as is clostridial myositis or gas gangrene. An empyema after a thoracic operation is not a problem, unless it is a postpneumonectomy empyema in which a bronchopleural fistula develops, which can be lethal. Ruptured appen­dicitis is different from perforated diverticulitis. A leaking colonic anastomosis or a duodenal stump blowout are great problems. A ruptured tubal ovarian ab­scess is different from an ectopic pregnancy. A pulmonary infection in a patient on a ventilator is not similar to acute pancreatitis. I plea for differentiation of in­fections into specific components and specific entities that can then be treated appropriately.

We must now integrate all this information into a whole In Schultz' description of integrative biology he said: "In biology the parts are dynamic or plastic. They can change shape when they are brought together -that's what nonlinearity means. The shape of the 'whole' cannot be predicted by knowing only the shapes of the separated parts. Thus, molecular biology pro­vides parts of a puzzle. Putting them together is now our problem" [57]. Walter Cannon, the father of homeostasis, who contributed so much to our knowledge of physiology, would focus on the phases of inflammation and pro- and anti-in­flammatory responses. This is now out of date. We must go beyond that. Buch­man [58] said, "If rested-nonlinear models better represent human physiology than Cannon's negative feedback servomechanisms, therapy should be directed toward transitions to a basal range not manipulating cytokines, nitric oxide or other mediators".

Conclusions I have reviewed the discrepancy between what we know and what we can do, the problems of nosocomial infection in the ICUs, and I have made a modest

130 A.E. Baue

proposal. I reviewed antibiotic resistance, the innate defenses, and bacterial of­fenses that are becoming known, the necessity for prevention and treatment of specific abnormalities rather than general concept such as sepsis, SIRS, MODS, or MOE I described the need to breakdown sepsis into specific infections rather than trying to treat sepsis which is not an entity. Finally, I indicate that we must integrate all of this information into the complexities and interrelationships of mediators, therapy, and the biological response to injury and infection in nonlin­ear biology. 1. There is a great discrepancy between how much we know and how little we

can use that information in caring for our patients. 2. Nosocomial infection is a great ICU problem which requires new approaches

to ICU design and function. I recommend short- (conventional) and long­term ICUs which are ultraclean. At the Society of Critical Care Medicine in January 1999 there were several presentations on stepdown units with exact­ly the concept that I have recommended [59, 60].

3. There is a contest between bacterial resistance to antibiotics, their adapta­tions to cause disease and our innate defense systems, and new antibiotics. The bugs seem to be winning.

4. Prevention of infection [61, 62] and organ failure is the key to better care. The treatment of specific diseases rather than inflammation (sepsis, SIRS, MODS, and MOF) will help.

5. Nonlinear or integrative biology is the next scientific revolution [63].

References 1. Bion JF, Brun-Buisson C (2000) Introduction - infection and critical illness: genetic and envi­

ronmental aspects of susceptibility and resistance. Intensive Care Med 26:S I-S2 2. Vincent I-L (2000) Microbial resistance: lessons from the EPIC study. Intensive Care Med

26:S3-S8 3. Emmerson M (2000) Antibiotic usage and prescribing policies in the intensive care unit. In­

tensive Care Med 26:S26-S30 4. van Deventer SIH (2000) Cytokine and cytokine receptor polymorphisms in infectious dis­

ease. Intensive Care Med 26:S98-S102 5. Bellingan G (2000) Leukocytes: friend or foe. Intensive Care Med 26:S111-S118 6. Cook D (2000) Ventilator associated pneumonia: perspectives on the burden of illness. Inten­

sive Care Med 26:S31-S37 7. Elliott T (2000) Intravascular catheter-related sepsis-novel methods of prevention. Intensive

Care Med 26:S45-S50 8. Marshall IC (2000) Clinical trials of mediator-directed therapy in sepsis: what have we

learned? Intensive Care Med 26:S75-S83 9. Brun-Buisson C (2000) The epidemiology of the systemic inflammatory response. Intensive

Care Med 26:S64-S74 10. Hawkey PM (2000) Mechanisms of resistance to antibiotics. Intensive Care Med 26:S9-S 13 11. Livermore DM (2000) Epidemiology of antibiotic resistance. Intensive Care Med 26:S 14-S21 12. Scott G (2000) Prevention and control of infections in intensive care. Intensive Care Med 26:

S22-S25

Sepsis and Organ Dysfunction: An Overview of the New Science and New Biology 131

13. Jasny BR, Bloom FE (1998) It's not rocket science - but it can save lives. Science 280:1507 14. Bonten MJM, Weinstein RA (1999) Bird's-eye view of nosocomial infections in medical

ICU: blue bugs, fungi, and device-days. Crit Care Med 27:853-854 15. Morell V (1997) Antibiotic resistance: road of no return. Science 278:575-576 16. Fleenor-Ford A, Hayden MK, Weinstein RA (1999) Vancomycin-resistant enterococci: impli­

cations for surgeons. Surgery 125: 121-125 17. Davis JM, Huycke MM, Wells CL et al (1996) Surgical infection society position paper on

vancomycin resistant enterococcus. Arch Surg 131: 1061-1068 18. Dahms RA, Johnson EM. Statz CL et al (1998) Third-generation cephalosporins and van­

comycin as risk factors for postoperative vancomycin-resistant enterococcus infection. Arch Surg 133:1343-1346

19. Bonten MJM, Slaughter S. Hayden MK et al (1998) External sources of vancomycin-resistant enterococci for intensive care units. Crit Care Med 26:2001-2004

20. Farr B (1998) Hospital wards spreading vancomycin-resistant enterococci to intensive care units: returning coals to Newcastle. Crit Care Med 26: 1942-1943

21. Chaix C. Durand-Zaleski L Alberti C. Brun-Buisson C (1999) Control of endemic methi­cillin-resistant staphylococcus aureus. A cost-benefit analysis in an intensive care unit. JAMA 282:1745-1751

22. D'Agata EMC, Venkataraman L DeGirolami P et al (1999) Colonization with broad-spec­trum cephalosporin-resistant Gram-negative bacilli in intensive care units during a nonout­break period: prevalence, risk factors. and rate of infection. 27: 1090-1095

23. Harbarth S, Pittet D (1999) Multiresistance of Gram-negative bacteria in intensive care units: bad news from without. Crit Care Med 27: 1037-1038

24. Fierer J, Guiney D (1999) Extended-spectrum ~-lactamases a plague of plasmids. JAMA 281 :563-564

25. Gleason TG. Addison KM, Caparelli D et al (1997) Emerging evidence of selection of flu­conazole-tolerant fungi in surgical intensive care units. Arch Surg 132: 1197 -1202

26. Bieber D, Ramer SW, Wu C-Y et al (1998) Type IV pili, transient bacterial aggregates, and virulence of enteropathogenic Escherichia coli. Science 280:2114-2118

27. Galan JE, Collmer A (1999) Type III secretion machines: bacterial devices for protein deliv­ery into host cells. Science 284: 1322-1328

28. Hamburger ZA, Brown MS, Isberg RR, Bjorkman PJ (1999) Crystal structure of invasin: a bacterial integrin-binding protein. Science 286:291-295

29. Strauss E (1999) A symphony of bacterial voices. Science 284:1302-1304 30. Pennisi E (1999) Is it time to uproot the tree of life? Science 284: 1305-1307 31. Costerton JW, Stewart PS, Greenberg EP (1999) Bacterial biofilms: a common cause of per­

sistent infections. Science 284: 1318-1322 32. Adair CG. Gorman SP, Feron BM et al (1999) Implications of endotracheal tube biofilm for

ventilator-associated pneumonia. Intensive Care Med 25: 1072-1076 33. Morell V (1997) Antibiotic resistance: road of no return. Science 278:575-576 34. Heeg K, Miethke T, Wagner H (1996) Superantigen-mediated lethal shock: the functional

state of ligand-reactive T cells. In: Heeg K, Miethke T, Wagner H (eds) Current topics of mi­crobiology. Springer-Verlag, Heidelberg Berlin New York, pp 83-100

35. Kaul R, McGeer A, Low DE et al (1997) Population-based surveillance for group a strepto­coccal necrotizing fascitis: clinical features, prognostic indicators, and microbiologic analysis of seventy-seven cases. Am J Med 103: 18-24

36. Hoffmann JA, Kafatos FC, Janeway CA Jr, Ezekowitz RAB (1999) Phylogenetic perspectives in innate immunity. Science 284:1313-1318

37. Tang Y-Q, Yuan J, Osapay K et al (1999) A cyclic antimicrobial peptide produced in primate leukocytes by the ligation of two truncated a-defensins. Science 286:498-502

38. Gans T (1999) Defensins and host defense. Science 286:420-421 39. Fang X-M, Book M, Hoeft A. Stuber F (1999) Impaired inducible expression of human ~-de­

fensin gene in peripheral blood of patients with severe sepsis. Crit Care Med 27[Suppl]:A36

l32 A.E. Baue

40. Yang D, Chertov 0, Bykovskaia SN et al (1999) ~-defensins: linking innate and adaptive im­munity through dendritic and T cell CCR6. Science 286:525-528

41. Stenger S, Rosat JP, Bloom BR et al (1999) Granulysin: a lethal weapon of cytolytic T cells. Immunol Today 20:390-394

42. Birmingham MC, Craig RR, Hafkin B et al (1999) Critical care patients with significant, re­sistant, Gram-positive infections enrolled in the linezolid compassionate use protocol. Crit Care Med 27[Suppl]:A33

43. Moellering RC Jr (1999) A novel antimicrobial agent joins the battle against resistant bacte­ria. Ann Intern Med l30: 155-157

44. Enserink M (1999) Promising antibiotic candidate identified. Science 286:2245-2247 45. Breukink E, Wiedemann I, van Kraaij C et al (1999) Use of the cell wall precursor lipid II by

a pore-forming peptide antibiotic. Science 286:2361-2364 46. Stone R (1993) Deja vu guides the way to new antimicrobial steroid. Science 259:1125 47. Alper J (1999) From the bioweapons trenches new tools for battling microbes. Science 284:

1754-1755 48. Luiten EJT, Wim CJ, Lange JF, Bruining HA (1995) Controlled clinical trial of selective de­

contamination for the treatment of severe acute pancreatitis. Ann Surg 222:57-65 49. 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 50. Braga M, Vignali A, Gianotti L et al (1995) Benefits of early postoperative enteral feeding in

cancer patients. Infusionsther Tranfusionsmed 22:280-284 51. Meduri GU, Cinn AJ, Leeper KV et al (1994) Corticosteroid rescue treatment of progressive

fibroproiferation in late ARDS. Patterns of response and predictors of outcome. Chest 105: 1516-1527

52. Baue AE (1994) Multiple organ failure, multiple organ dysfunction syndrome and the sys­temic inflammatory response syndrome-where do we stand? Shock 2:385-397

53. Baue AE, Durham R, Faist E (1998) Systemic inflammatory response syndrome (SiRS), mul­tiple organ dysfunction syndrome (MODS), multiple organ failure (MOF): are we winning the battle? Shock 10:79-89

54. Baue AE (1997) Multiple organ failure, mUltiple organ dysfunction syndrome, and systemic inflammatory response syndrome: Why no magic bullets? Arch Surg l32:703-707

55. Thangathurai D, Charbonnet C, Wo CCJ et al (1996) Intraoperative maintenance of tissue per­fusion prevents ARDS. New Horiz 4:466-474

56. 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

57. Schultz SG (1996) Homeostasis, humpty-dumpty and integrative biology. News Physiol Sci 11:238-246

58. Buchman TG (1996) Physiologic stability and physiologic state. J Trauma 41:599-605 59. Zimmerman JE, Seneff M, Wood S, Knaus WA (1999) A comparison of outcome and re­

source use for stepdown units and ICU patients. Crit Care Med 27: 154 60. Barie P, Eachempati SR, Hydo U (1999) Impact of a new intermediate care unit on utilization

and outcomes of the surgical intensive care unit. Crit Care Med 27 :28 61. Hopf HW, Hunt TK, West 1M et al (1997) Wound tissue oxygen tension predicts the risk of

wound infection in surgical patients. Arch Surg l32:997 -1004 62. Leaper DJ, Ali B, Melling A (2000) Local or systemic warming reduces the risk of wound in­

fection after clean surgery. Shock l3[Supp12]:6A 63. Baue AE (1999) Molecular biology and minimal surgery. CUIT Opin Crit Care 5 :284-289

INDEX I

acinetobacter 15 acute renal failure II I acute respiratory distress syndrome 33,47 antibiotic therapy monitoring 17 apoptosis 33 aprotinin 24 bacterial translocation 17 Candida 14, 78 cardiac index 25 cardiac output 25 cardiac surgery 24 cardiopulmonary bypass 23,47 carrier state 13 cephalosporins 17 chemotaxis 85 chlorhexidine 18 chronic pulmonary hypertension 47 chronic respiratory failure 47 classification of infections 13 coagulase-negative staphylococci 78 community-acquired infections 79 continuous arterovenous hemofiltration 112 continuous renal replacement therapy 114 continuous venovenous hemodiafiltration 114 cytokines 34 defensins 39 dopamine 57 endotoxemia 78 endotoxin 103 enhanced chemiluminescence 88 enteral immunonutrition 128 enterobacter 17 enterococci 78 enterococcus 15 epidemiology of infections II epsilon-aminocaproic acid [V 24 fibrinolysis 23 Gram-negative bacteria 14 Gram-negative infections 14 Gram-positive bacteria 14 Gram-positive colonization 17 Gram-positive infections 14 haemodialysis III haemotiltration III human host 75 hypoxemia 47

[CU overcrowding 17

ICU-acquired infections 80

immunofluorescence 88

infection 78, 123

inflammation 23

intensive care III length of stay 12

lung inflammation 39

lung injury 50

mechanical ventilation 26

microbiologic surveillance 17

mupirocin 18

neutrophil defensins 39

neutrophil 34,85,99

nitric oxide 36,47

organ dysfunction 124

oxidative stress 33

peritoneal dialyis III phagocytosis 85

plasma exchange III plasmapheresis III pneumonia 47

prothrombin fragment 1+223

prothrombin 23

pseudomonas 15, 78

pulmonary function 26

pulmonary hypertension 47

pulmonary embolism 47

renal function 28

resistance 15

sepsis 3,47,56,123,129

sepsis syndrome 103

serine 85

Staphylococcus Aureus II tachyphylaxis 49

threonine 85

thrombin 23

thrombosis 29

trauma patients I3

tyrosine phosphorylation 99

tyrosine 85

tyrosine phosphatases 99

vancomycin 18

ventilator-associated pneumonia 12

ORGAN FAILURE ACADEMY

Sepsi e insufficienza d' organo sono condizioni che mettono in apprensione chi prende in CUTa iI pazietJ.te critico. Qualunque sia la natura dell'evento scateDante, esso viene ad alterare uno stato biologico frutto di una annonica proporzione di elementi in equilibrio e contrastanti. La biochimica e la fisiologia, la fisiopatologia e la biotecnologia applicata offrono alIa clinica solo spooti di interprelazione degli innumerevoli meccanis~i che stanno alla base della sepsi e dell'insufficienza d'organo; ta1i eventi singoIannente 0 in associazione sono tra Ie cause maggiori di mortalit. nei pazienti degenti in lerapia intensiva. E giooto il momeoto di lanciare uoa idea stimolante; creare un "consensus" pennanenle di stndiosi ove ciascuoo porti il cootributo del proprio sapore e sia disponibile agli scambi inlerdisciplinari. L'Accademia per 10 studio della sepsi e dell'insufficienza d'orgaoo deve propagarsi in ogni direzione, avere carattere itinerante, proporre modelli di studio e di rieerca utili per la prevenziooe e il trattamenlo di uoa patologia e di uoa sindrome che coodizionaoo l'iter clinico del malato critico.

ORGAN FAILURE ACADEMY

Flir den kritische Patienlen behandelnden Arzl sind Sepsis ood Organschwache besorgniserregende Zuslande. Unabhiingig voo ihrem Ausliiser sltiren sie das biologische Gleichgewicht, das sich aus dem hannoniscben Zusammenspiel widerspriichlicber Elemenle ergibt. Biochemie uod Physiologie sowie Pathophysiologie, angewandle Biotechnologie hieten dem Kliniker nur Interpretationsansitze flir die oozihligen Vorginge, die Sepsis- und Organschwicheerscheinungen zugruodeliegen; allein oder in Kombination zihlen sie zu den hiufigsten Sierblichkeilsursachen bei Intensivpatienten. Nun iSI hiichste Zeit, eine slimulierende Vorstellung zu wagen; Die Sch;tffung eines slandigen Gremiums, in dessen Rahmen die verschiedenen Forscher ihren Kenntnisbeitrag leisten und sich am interdiszipliniiren Erfahrungsauslausch beleiligen konnen. Die Akademie zui Forschung von Sepsis- und Organschwiicheerscheinungen ist als facheriibergreifeoder, wandernder Zusammenschlu6 mil dem Ziel zu gestalten, StudieD- uod Forschoogskonrepte vorzuschlagen, die zur Vorbeugung und Behandlung von den kliniscben Verlauf eines kritischen Patienten beeinllu6eoden Pathologien uod Syndromen dienen konnen.

AIIA,tlBIIIIS no II3fIII!HIIIl BOIIPIJCOB CEIICHCA B III\AOCTA'I'(l'IHOCTH OPI'AROB 'II!.IOBI!'IJ!CHDI'O '1'IW.

Ha.nHtlHe ycnoBHfI cenCHca H HBAOCTa'J'Ol.lHOCTH y opraHOB 60JlbHOrO Bbi3b1BaeT HeMaJIYlO 3at'5oTY y BpaqeA, 3aHHMBlOCIHXCJl naq8HHeN Tex naLIHeHTOB, KOTopble H,axO,AllTC.fI B KpHTHqeCKOA CTa,I\HH 3at'5oneBaHM. B03HHKHOBeHHe 3at'5oneBaHMft, He38.BHCHNO 0'1' BH,Aa npHtlMHJlXll::ero ¢ax'l'OPS, H3IIeHJleT t'5HonOrMtieCKOe COCT01lHHe ,llaHHOrO opraHa, onpe,lleJUleMoe rapMOHHtlHhIM COOTHomeHHew B38HMOKOMlleHCHpyxH:UOtCJl eneueHTOB. BHOXHMHft H iIIH3HonorM, K8.K. iIIH3HonaTonorM, npeACT8.BnmoT KnHHHKe nHmb HCXOAHble 3neMeHTli ,IVl.fl HHTepnpeTaL\HH COBOxynHOCTH MeXaHH3NOB, Bbl3b1B8lXIHX B03HHKHOBeHHe cenCHCS H He,[lOCTSTOqHOCTH y OpraHOB qenOBeqeCKOro Tans. 3TH UeX8.HH3Yl11, ,IteACTBYJ:lIItHe OT,IJ;enbHO HJIM B COtleT8.HHH APyr C llPYI'OM, RBJIJDITCR O,ll;HOA a3 -rn8.BHbIX npHtrnH CMepTHTanbHOCTH cpe,ItH naUHeHTOB, HaxOAJ'ililHXC1I no,£t PexHMON HHTeHCHBHoA 'l'epanHH. nepe,[l HaMH CTOHT, n03TOIQ. BUHaR 3a,[IB.'la. T.e. C03,£tSHHe nOCT01lHHOrO KOW1TeTa yqeHW<. Ka.JIt,[U:.IA 1-13 KOTOPbIX ,!:IOMeH CiY,IleT ,IlenSTb caoH BKna,A B ,[lena H3ytleHMft yxa3aHHbIX BonpOCOB H cnocoCicTBoBaTb N6&OTpacneaOMY ot'5NeHY HHtPopuallH.RMH. AKa,qeMHH no H3ytleHMIO BonpOCOB cenCHca H He,llOCTaTOtlHOCTH OpraHOB tlenOBe'leCKOrO Tena ,IlOJIXHa 6Y,£teT HMeTb "nepe,llBH&HoA" xspsKTep. T. e. pscnpOCTpSHJlTbCSi 80 Bcex . HanpaBJIeHH.fIX. H paapa6aTblaa'l'b HCCJIeAOBSTenbcKHe npoeKTbI, He06xo,llHMble AJISi npoillHnaKTHKH naTonorHH H netleHH.fI 3a6oneBaHHH, o6ycnOBJIHBaJOIl{HX KJIHHH'leCKHA no,qxo,ll K 60JIbHOMY a KpHTHtleCKOA CTB.,D;HH aa60neBaHHH.

t..1.,.Jt..,JoI1 ~ -~",t...I .... "'L,.

~I £>!aIt. ... t.l1t......,u ",t-..a:lIl.....:iI1 ...,..JI

....,l~.£ It...11,...,..t1

~I ~t.lP""l &,,1 "'t...I .... ~1

~.,!WII ......... 4.,.011 "WI .,. ..... ....u e'l-ll.,....t.ll J.JI ~I JAilt.l<'"""'l eoI'IL.,..,..-,

'<!o.,.... ... ,.".,.: ... \,flll.,-.,JolI~I.:ut.JIl"Io."""'.,.~1 uPJ, ltI~I,,~11

.ll:l~tVlt..I""rl-~.W.11~U;,r'ovlo"'''II.".,SIl,,!. ~l.,H.""'.I~u.l.'-_u. ~I ~t.lP""l e<l"I.,I1 \,foU-l",,1 ",I,l-II...ItO.~.I~I ..... .loli~ ,..., .... 1 ... 11

. .,..........t&II;:>!aI1",11o..,.'t;,Il",1Io.r-I'--.,I1, I..~ ... ), • ...,., 10......,.,.,...,,,,,1

...w...;... .:,. ":""411.:,. ,..lLl..o J.w. ,...u. JlO. .,....-m , • ....-J....u ' ........ '§..A c"IoJ "t.11 "I .....

'-"I. ~11 "'L-..alII ........ ".., ~ " .... J~ J.,-..,. ... t.....~JS:,....., ",~I .,.~uPJ __ · '4'IJS:.,.~1 ~t.~leoI'II"'t...I .... """":I",~ ..... """,

.It...lI.,J.. '~"".lt..-. .... -'"'.:,. ;:""t.~lVl.,II". .. \,flll"..toJI.....,u;:,,", ~ .40.111 t.lt.s

. ~..-J1"'11o.JJ..;r..-&11

ORGAN FAILURE ACADEMY

La septicemie ell'insuffisance d'urganes represenlenl des conditions pouvanl • bon droil engendrer 00 etat d'apprehen~on de la part du midecin qui doit ""ter ses soins au patienl critique. Quelle que soit la nature du declencbement de la maladie, il en decoule bien evidemment une alteration de l'etat biologique qui eslle resultat d'un rapport hannonieox entre des elements qui peuvent etre, a leur lour, soil en equilibre soil en contrasle. La biochimie, la physiologie, la physiopathologie ella biotechnologie appliquee n'olfrent a l'analyse clinique que des elements ""Iiminaires d'inlerpretation des nombreu, mecanismes qui so trouvenl a la base de la septicemie el de l'insuffisanee d'urganes: les differentes pathologies en jeu, ptises individuellemenl ou associ"s les ones aux autres, constituent les principales causes de mnrtalite chez les patients critiques soumis • traitement intensif. Le moment est venu de proposer one idee stimtilanle, a savoir la creation d'un "consensus" pennanent a ee sujet de la part de tous les experts en la matiere; pour ee faire, il irnpone bien enteodu que chacun soil dispose a fnurnir son propre apport en onnnaissanees el qu'il soit a la fuis ouvert a loul type d'echanges interdisciplinaires. Aussi faul-il que l'Academie pour l'etude de la septicemie el de l'insuffisance d'organes so developpe dans loules les directions, qu'elle ail un caractere itineranl el qu'elle soit en mesure de proposer des modeles d'etude el de recherche utiles pour la prevention elle traitement d'one pathologie et d'oo syndrOme qui inIluencenl en bonne mesure l'evolution clinique du patienl critique.

ORGAN FAILURE ACADEMY

Sepsis e insuficiencia de organa son afecciones que crean cierta inquielud al curar el paciente critico. Cualquiera que sea el gtnero del factor desencadenador, esle termina pur alterar un estado biol6gico fruto de una proporcion annonica de elementos en equilibrio y contrastantes. La bioqufmica y la fi~ologf~ la fisiopalologia y la biotecnologia aplicada olreeen a la clinica solamente unas indicaciones para la interpretacion de los mecanismos innumerables que estan a la bose de la sepsis y de la insuficiencia de organa; estos factores pur separadn 0 conjuntarnente se hallan entre las causas principales de mortalidad­entre los pacientes que reciben terapia inlensiva Ha lIegado el momenta de lanzar uoa idea inspiradora; de crear un "consensus" pennaneDle de cientificos doode cada uno lIeve su propio saber y sea disponibl. a los intercambios interdisciplinarios. La Academia para el "tudio de la sepsis y de la insuficiencia de Organo debe propagarse en todas las direcciones, lener caracrer itinerante, proponer modelos de estudio y de investigacio. utiles para la prevencio. y el tratantiento de ooa patologia y de uoa sindrome que afeelen el curso clinico del enfermo critico.

J!!ultlliE&I1l11 .... 7 :h 7' ~ -!8ultlll&tFlI .... I1, alll.l.tr*lftll! t.n' 05trl::.!:?"t, 40'1::_*£05,", .~~05. ~ft'~~"t~lltt*~8g~~.~.!:"~~~~~!J!:~.~~

1IJ!l ~~,&tFjl/I!!II::J;?"tfl' .ftl::l!gIti~Dftl::J; 05 t> ~~<I; 05. ~.'II:!J!:, ~

1I!J!:, ~1I11111!J!:f,j;.!:'1;t, .'*-IIt!J!:l::tH'"t, !Ut:1fD.1iE&tFlII" •• ~il5*l!!Jt,l: ~CID.~~.~ID.C*4~05. ~1::~ft'ID!Ut:IfD.IiE&tFlII1f •• I1,.­t> t. C I1M~ t., 1ftIl!Ijl.l..~~ C ~'JEt::l!~.!: f,j;? n' 05. ~~af,j;tt~IDf,j;.~,.,D.!:~ID~~*.Z~05~~*"t"~.~~!l.

.~;M;llIIiIf~~lIjfi.!:t."t,~ftl::.4t., • .ll*~:b~.IDIII*1.,D.!:~ ~·~~~~~~·.!:t."t~,!J!:~l.lI*fI'05.~~05. ~~!Ut:lfD.ll&tFlII1f

•• ~7:h7'~-I1,~nWI::.&t.,~I::.W~~*~!l,~~.!:lIIttIDIft .&tF of: IDIiEMIJI:: J; 05allli!!.~III'*-I,'t.1::1III4~ 05i1f~Aif*.Z t."t~, C t>ID~<I;l..

~"~~~tt~:f~ft

~ff~.~"A~.~~~~~tt~:f±~~~~ • t!::f~- ~ltai1~~~rt-J:fl6Jfjt~rt-J~HtM~-t-EB:;P: tJj~.~t§&W.flc~if1ti9fm~rt-J~~~~· ~T~. ~~~~tt~:f±rt-JtT~~Wrt-J~~ufi. ~~~. ~Jf~. ~JfM~ RjMiiJt-~ti~- it~W.7C1tj:.!kI:l:lJ,{ kEH~I:I:IJ,{t~H~~jJlJ~Jf~Art-J~gWZ-. m~J:iI1Jjl~:tiI:l:l~rt-Jiltv<. ~J~~ ~ ttrt-J~#zfij-a-t'f: rt-Jttft. ij!.-t~#~~i$ ~ 2rt-J"ff~. ni;f]M#rt-J~­~.~~~~m~:f±~ft4~~~~~~~~~ *~a&~Jf~~.~~Art-J~*ufi~~~~tt~~ rt-J~Jf~~~-a-~ .