Content uploaded by Christoph Langenberg
Author content
All content in this area was uploaded by Christoph Langenberg
Content may be subject to copyright.
Acute Kidney Injury
This book has been made possible by the generous support of
Contributions to Nephrology
Vol. 156
Series Editor
Claudio Ronco,
Vicenza
Acute Kidney Injury
Basel · Freiburg · Paris · London · New York ·
Bangalore · Bangkok · Singapore · Tokyo · Sydney
Volume Editors
Claudio Ronco, Vicenza
Rinaldo Bellomo,
Melbourne, Vic.
John A. Kellum, Pittsburgh, Pa.
63 figures, 2 in color, and 40 tables, 2007
Claudio Ronco Rinaldo Bellomo
Department of Nephrology Department of Intensive Care
St. Bortolo Hospital Austin Hospital
I-36100 Vicenza (Italy) Melbourne, Vic. 3084 (Australia)
John A. Kellum
The CRISMA Laboratory
Department of Critical Care Medicine
University of Pittsburgh, 3550, Terrace Street
Pittsburgh, PA 15261 (USA)
Bibliographic Indices. This publication is listed in bibliographic services, including Current Contents
®
and
Index Medicus.
Disclaimer. The statements, options and data contained in this publication are solely those of the individ-
ual authors and contributors and not of the publisher and the editor(s). The appearance of advertisements in the
book is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness,
quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property
resulting from any ideas, methods, instructions or products referred to in the content or advertisements.
Drug Dosage. The authors and the publisher have exerted every effort to ensure that drug selection and
dosage set forth in this text are in accord with current recommendations and practice at the time of publication.
However, in view of ongoing research, changes in government regulations, and the constant flow of information
relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for
any change in indications and dosage and for added warnings and precautions. This is particularly important when
the recommended agent is a new and/or infrequently employed drug.
All rights reserved. No part of this publication may be translated into other languages, reproduced or
utilized in any form or by any means electronic or mechanical, including photocopying, recording, microcopying,
or by any information storage and retrieval system, without permission in writing from the publisher.
© Copyright 2007 by S. Karger AG, P.O. Box, CH–4009 Basel (Switzerland)
www.karger.com
Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel
ISSN 0302–5144
ISBN 978–3–8055–8271–1
Library of Congress Cataloging-in-Publication Data
Contributions to Nephrology
(Founded 1975 by Geoffrey M. Berlyne)
Acute kidney injury / volume editor[s], Claudio Ronco, Rinaldo Bellomo, John
A. Kellum.
p. ; cm. – (Contributions to nephrology, ISSN 0302-5144 ; v. 156)
Includes bibliographical references and index.
ISBN-13: 978-3-8055-8271-1 (hard cover : alk. paper)
1. Acute renal failure. 2. Kidney–Wounds and injuries. 3. Intensive
care nephrology. I. Ronco, C. (Claudio), 1951- II. Bellomo, R. (Rinaldo),
1956- III. Kellum, John A. IV. Series.
[DNLM: 1. Kidney Failure, Acute. 2. Critical Care. 3. Kidney–injuries.
4. Kidney Failure, Acute–therapy. 5. Renal Replacement Therapy. 6.
Sepsis–complications. W1 CO778UN v.156 2007 / WJ 342 A1885 2007]
RC918.R4A327 2007
616.6⬘14–dc22
2007012324
V
XI Preface
Ronco, C. (Vicenza); Bellomo, R. (Melbourne, Vic.); Kellum, J.A. (Pittsburgh, Pa.)
Critical Care Nephrology Issues
1 Pre-Renal Azotemia: A Flawed Paradigm in
Critically Ill Septic Patients?
Bellomo, R.; Bagshaw, S.; Langenberg, C. (Melbourne, Vic.); Ronco, C. (Vicenza)
10 The Concept of Acute Kidney Injury and the RIFLE Criteria
Kellum, J.A. (Pittsburgh, Pa.); Bellomo, R. (Melbourne, Vic.); Ronco, C. (Vicenza)
17 The Liver and the Kidney: Mutual Clearance or Mixed Intoxication
Ar royo, V. (Barcelona)
24 Critical Care Nephrology: A Multidisciplinary Approach
Vincent, J.-L. (Brussels)
Epidemiology and Pathogenesis of AKI, Sepsis and MOF
32 Incidence, Classification, and Outcomes of Acute Kidney Injury
Hoste, E.A.J. (Ghent/Pittsburgh, Pa.); Kellum, J.A. (Pittsburgh, Pa.)
39 Pathophysiology of Acute Kidney Injury: Roles of Potential
Inhibitors of Inflammation
Bonventre, J.V. (Boston, Mass.)
Contents
47 Sepsis and Multiple Organ Failure
Pinsky, M.R. (Pittsburgh, Pa.)
64 Classification, Incidence, and Outcomes of Sepsis and
Multiple Organ Failure
Vincent, J.-L.; Taccone, F.; Schmit, X. (Brussels)
Evaluation of Illness Severity
75 Genetic Polymorphisms in Sepsis- and Cardiopulmonary Bypass-
Associated Acute Kidney Injury
Haase-Fielitz, A.; Haase, M. (Melbourne, Vic./Berlin);
Bellomo, R. (Melbourne, Vic.); Dragun, D. (Berlin)
92 Predictive Capacity of Severity Scoring Systems in the ICU
Schusterschitz, N.; Joannidis, M. (Innsbruck)
101 Determining the Degree of Immunodysregulation in Sepsis
Cavaillon, J.-M.; Adib-Conquy, M. (Paris)
Metabolism, Electrolytes and Acid-Base Disorders
112 Nutritional Management in Acute Illness and
Acute Kidney Insufficiency
Leverve, X.M. (Grenoble); Cano, N.J.M. (Marseille)
119 Fundamentals of Oxygen Delivery
Yassin, J.; Singer, M. (London)
133 Principals of Hemodynamic Monitoring
Polanco, P.M.; Pinsky, M.R. (Pittsburgh, Pa.)
158 Acid-Base Disorders and Strong Ion Gap
Kellum, J.A. (Pittsburgh, Pa.)
167 Fluid Resuscitation and the Septic Kidney: The Evidence
Licari, E.; Calzavacca, P. (Melbourne, Vic.); Ronco, C. (Vicenza);
Bellomo, R. (Melbourne, Vic.)
Nursing Issues in Critical Care Nephrology
178 Factors Affecting Circuit Patency and Filter ‘Life’
Baldwin, I. (Melbourne, Vic.)
185 Starting up a Continuous Renal Replacement Therapy
Program on ICU
De Becker, W. (Leuven)
Contents VI
191 Is There a Need for a Nurse Emergency Team for
Continuous Renal Replacement Therapy?
Baldwin, I. (Melbourne, Vic.)
197 Information Technology for CRRT and Dose Delivery Calculator
Ricci, Z. (Rome/Vicenza); Ronco, C. (Vicenza)
Early Diagnosis and Prevention of AKI
203 Emerging Biomarkers of Acute Kidney Injury
Devarajan, P. (Cincinnati, Ohio)
213 Diagnosis of Acute Kidney Injury: From Classic Parameters
to New Biomarkers
Bonventre, J.V. (Boston, Mass.)
220 Endotoxin and Cytokine Detection Systems as
Biomarkers for Sepsis-Induced Renal Injury
Opal, S.M. (Providence, R.I.)
227 Quantifying Dynamic Kidney Processes Utilizing
Multi-Photon Microscopy
Molitoris, B.A.; Sandoval, R.M. (Indianapolis, Ind.)
236 Diuretics in the Management of Acute Kidney Injury:
A Multinational Survey
Bagshaw, S.M. (Melbourne, Vic./Edmonton, Alta.); Delaney, A. (Sydney/St. Leonards);
Jones, D. (Melbourne, Vic.); Ronco, C. (Vicenza); Bellomo, R. (Melbourne, Vic.)
250 Stem Cells in Acute Kidney Injury
Bussolati, B.; Camussi, G. (Turin)
Practice Patterns for RRT in the ICU
259 Anticoagulation Options for Patients with Heparin-Induced
Thrombocytopenia Requiring Renal Support in the
Intensive Care Unit
Davenport, A. (London)
267 Nutritional Support during Renal Replacement Therapy
Chioléro, R.; Berger, M.M. (Lausanne)
275 Vascular Access for HD and CRRT
Schetz, M. (Leuven)
287 Dialysate and Replacement Fluid Composition for CRRT
Aucella, F.; Di Paolo, S.; Gesualdo, L. (Foggia)
Contents VII
297 Results from International Questionnaires
Ricci, Z.; Picardo, S. (Rome); Ronco, C. (Vicenza)
Which Treatment for AKI in the ICU
304 Intermittent Hemodialysis for Renal Replacement Therapy
in Intensive Care: New Evidence for Old Truths
Van Biesen, W.; Veys, N.; Vanholder, R. (Ghent)
309 Continuous Renal Replacement in Critical Illness
Ronco, C.; Cruz, D. (Vicenza); Bellomo, R. (Melbourne, Vic.)
320 Sustained Low-Efficiency Dialysis
Tolwani, A.J.; Wheeler, T.S.; Wille, K.M. (Birmingham, Ala.)
325 The Role of the International Society of Nephrology/Renal Disaster
Relief Task Force in the Rescue of Renal Disaster Victims
Vanholder, R.; Van Biesen, W.; Lameire, N. (Ghent); Sever, M.S. (Istanbul)
Extracorporeal Treatment for Specific Indications
333 Renal Replacement Therapy for the Patient with Acute
Traumatic Brain injury and Severe Acute Kidney Injury
Davenport, A. (London)
340 Cardiopulmonary Bypass-Associated Acute Kidney Injury:
A Pigment Nephropathy?
Haase, M.; Haase-Fielitz, A. (Melbourne, Vic./Berlin); Bagshaw, S.M. (Melbourne, Vic.);
Ronco, C. (Vicenza); Bellomo, R. (Melbourne, Vic.)
354 CRRT Technology and Logistics: Is There a Role for a Medical
Emergency Team in CRRT?
Honoré, P.M. (Ottignies-Louvain-la-Neuve); Joannes-Boyau, O. (Pessac);
Gressens, B. (Ottignies-Louvain-la-Neuve)
365 Continuous Hemodiafiltration with Cytokine-Adsorbing
Hemofilter in the Treatment of Severe Sepsis and Septic Shock
Hirasawa, H.; Oda, S. (Chiba); Matsuda, K. (Yamanashi)
371 Blood and Plasma Treatments: High-Volume
Hemofiltration – A Global View
Honoré, P.M. (Ottignies-Louvain-la-Neuve); Joannes-Boyau, O. (Pessac);
Gressens, B. (Ottignies-Louvain-la-Neuve)
387 Blood and Plasma Treatments: The Rationale of
High-Volume Hemofiltration
Honoré, P.M. (Ottignies-Louvain-la-Neuve); Joannes-Boyau, O. (Pessac);
Gressens, B. (Ottignies-Louvain-la-Neuve)
Contents VIII
396 Liver Support Systems
Santoro, A.; Mancini, E.; Ferramosca, E.; Faenza, S. (Bologna)
405 Coupled Plasma Filtration Adsorption
Formica, M.; Inguaggiato, P.; Bainotti, S. (Cuneo); Wratten, M.L. (Mirandola)
411 Albumin Dialysis and Plasma Filtration Adsorption Dialysis System
Nalesso, F.; Brendolan, A.; Crepaldi, C.; Cruz, D.; de Cal, M. (Vicenza);
Bellomo, R. (Melbourne, Vic.); Ronco, C. (Vicenza)
419 Renal Assist Device and Treatment of Sepsis-Induced
Acute Kidney Injury in Intensive Care Units
Issa, N.; Messer, J.; Paganini, E.P. (Cleveland, Ohio)
428 Renal Replacement Therapy in Neonates with Congenital
Heart Disease
Morelli, S.; Ricci, Z.; Di Chiara, L.; Stazi, G.V.; Polito, A.; Vitale, V.; Gior ni, C.;
Iacoella, C.; Picardo, S. (Rome)
New Trials and Meta-Analyses
434 The DOse REsponse Multicentre International
Collaborative Initiative (DO-RE-MI)
Monti, G. (Milan); Herrera, M. (Malaga); Kindgen-Milles, D. (Düsseldorf);
Marinho, A. (Porto); Cruz, D. (Vicenza); Mariano, F. (Turin); Gigliola, G. (Cuneo);
Moretti, E. (Bergamo); Alessandri, E. (Rome); Robert, R. (Poitier); Ronco, C. (Vicenza)
444 Clinical Effects of Polymyxin B-Immobilized Fiber Column in
Septic Patients
Cruz, D.N. (Vicenza/Quezon City); Bellomo, R. (Melbourne, Vic.);
Ronco, C. (Vicenza)
452 Author Index
454 Subject Index
Contents IX
Preface
Four million people will die this year of a condition whose pathophysiology we
do not understand and for which no effective treatment exists. Millions more
will sustain complications and prolonged hospitalizations. Acute kidney injury
(AKI) is complex syndrome for which treatment is lacking and understanding is
limited. Defined as an abrupt change in serum creatinine and/or urine output
and classified according the RIFLE (Risk, Injury, Failure, Loss and End-stage
kidney disease) criteria, AKI is associated with a more than twofold increase in
the risk of death in hospital – even after controlling for other conditions. When
severe enough to require renal replacement therapy, AKI results in hospital
mortality rates of approximately 60%. Yet, even this severe form of AKI is sur-
prisingly common – nearly 6% of all patients admitted to the ICU.
Moreover, as many as two thirds of patients admitted to the ICU have some
evidence of AKI and virtually all may be at risk of this condition. Sepsis, shock,
advanced age and exposure to nephrotoxins lead the list of risk factors and
many patients have more than one. As such, it is absolutely clear that in order
for patients to receive optimal care, the treating physician needs a detailed
working knowledge of multiple aspects of care so that appropriate multidisci-
plinary assistance is sought at the right time and new techniques of organ sup-
port are applied in a safe, timely and effective way. In this volume we have
combined the contributions of experts in various fields to tackle some of the
fundamental and complex aspects of AKI from pathophysiology to epidemiol-
ogy to diagnosis and treatment; from emerging biomarkers to genetic polymor-
phisms. We have also included contributions which focus on the many
complications of AKI and comorbid conditions encountered in patients with
XI
AKI. From abnormalities in oxygen delivery, hemodynamics and acid-base bal-
ance to multi-organ failure, leading experts cover the fundamentals as well as
the latest developments.
Because the immune response to infection is central in determining organ
injury, this volume also focuses on the role of immune dysregulation in deter-
mining renal and lung injury, on the role of immune mediators in inducing dys-
regulation of the immune response, and on the role of genetics in determining
such a response. The roles of nutritional support, metabolic management and
fluid resuscitation in modulating the immune response and influencing patient
outcomes are also considered. As extracorporeal therapies are being increasingly
used in the care of these complex patients, we focus on important technical
aspects of such therapies, including vascular access, anticoagulation, and fluid
composition as well as the logistics of starting continuous renal replacement
therapy programs and keeping them running. As the choice of treatment modal-
ity remains controversial, we also discuss different approaches to renal support
from intermittent dialysis to continuous therapies and hybrid techniques. Finally,
we conclude with a description of advanced extracorporeal techniques of organ
support and discuss their role in the management of sepsis and AKI in the con-
text of an overall strategy of multi-organ failure management.
The overall aim of this volume of the Contributions to Nephrology series is
to provide the medical community involved in the care of critically ill patients
with AKI with a practical and up-to-date summary of current knowledge and
technology as well as a fundamental understanding of pathogenesis and likely
future developments in this field. Just as importantly, this volume serves to
challenge and reexamine the fundamental underlying assumptions we hold with
regard to critical illness in general and AKI in particular. By reexamining age-
old paradigms such as ‘pre-renal azotemia’ and looking to redefine the concepts
of acute renal disease, we can expect to stumble and fall but in the end find our-
selves in a new and better place. Critical care nephrology is an interdisciplinary
field and it is through the continued work of basic, clinical, and translational
researchers from numerous disciplines, and clinicians who will always drive the
field forward, that together we will realize our ambition to improve the standard
of care for patients with AKI, worldwide.
Claudio Ronco, Vicenza
Rinaldo Bellomo, Melbourne, Vic.
John A. Kellum, Pittsburgh, Pa.
Preface XII
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 1–9
Pre-Renal Azotemia: A Flawed Paradigm
in Critically Ill Septic Patients?
Rinaldo Bellomo
a
, Sean Bagshaw
a
, Christoph Langenberg
a
, Claudio Ronco
b
a
Department of Intensive Care, Austin Hospital, Melbourne, Vic., Australia;
b
Department of Nephrology, St. Bortolo Hospital, Vicenza, Italy
Abstract
The term pre-renal azotemia (or on occasion ‘pre-renal renal failure’) is frequently
used in textbooks and in the literature to indicate an acute syndrome characterized by the
presence of an increase in the blood concentration of nitrogen waste products (urea and cre-
atinine). This syndrome is assumed to be due to loss of glomerular filtration rate but is not
considered to be associated with histopathological renal injury. Thus, the term is used to dif-
ferentiate ‘functional’ from ‘structural’ acute kidney injury (AKI) where structural renal
injury is taken to indicate the presence of so-called acute tubular necrosis (ATN). This para-
digm is well entrenched in nephrology and medicine. However, growing evidence from
experimental animal models, systematic analysis of the human and experimental literature
shows that this paradigm is not sustained by sufficient evidence when applied to the syn-
drome of septic AKI, especially in critically ill patients. In such patients, several assumptions
associated with the ‘pre-renal azotemia paradigm’ are violated. In particular, there is no evi-
dence that ATN is the histopathological substrate of septic AKI, there is no evidence that
urine tests can discriminate ‘functional’ from ‘structural’ AKI, there is no evidence that any
proposed differentiation leads or should lead to different treatments, and there is no evidence
that relevant experimentation can resolve these uncertainties. Given that septic AKI of criti-
cal illness now accounts for close to 50% of cases of severe AKI in developed countries,
these observations call into question the validity and usefulness of the ‘pre-renal azotemia
paradigm’ in AKI in general.
Copyright © 2007 S. Karger AG, Basel
Introduction
Acute renal failure or, to use the more recent term, acute kidney injury
(AKI) is a common syndrome in hospitals [1] and intensive care units [2]. It is
Critical Care Nephrology Issues
Bellomo/Bagshaw/Langenberg/Ronco 2
associated with a mortality rate which varies according to definition. However,
if the RIFLE consensus definition is applied [3, 4], the hospital mortality of
severe injury (RIFLE category F) approximates 50% [1]. Although, it has been
argued that patients die with AKI rather than of AKI and that AKI is simply
an expression of illness severity, consistent and strong evidence supports the
notion that AKI has an independent impact on outcome, even after all other
possible variables determining outcome have been corrected for [5]. Finally,
because of the frequent need for extracorporeal supportive therapies, the treat-
ment of established AKI is complex, labor intensive and costly [6]. Clearly,
therefore, AKI is a major issue in acute medicine and increasing our under-
standing of its pathogenesis is vital. As the assessment of renal function in man
is only indirect and limited by the inability to obtain tissue, to reliably and con-
tinuously measure renal blood flow and the need to derive indirect information
on glomerular filtration rate and/or tubular cell status through urine analysis,
animal experimentation has been used to advance our understanding of AKI.
Unfortunately, the animal models that have been used to study AKI have been
predominantly based on ischemia (occlusion of the renal artery for a given
period of time) or the administration of a selected group of toxins instead of
sepsis where renal blood flow may be increased [3, 7, 8]. These ischemic/toxic
models have limited relevance to clinical reality where renal artery occlusion is
an uncommon cause of AKI and the administration of toxins is a similarly less
common trigger of renal injury, especially if severe [2]. In the clinical environ-
ment, sepsis is now the number one trigger of AKI [2] especially in critically ill
patients who now make up the vast majority of cases of severe AKI [9]. Yet real-
istic models of septic AKI, which fully simulate the clinical situation, have not
yet been developed. Most studies have been short-term in design and have
not reported any information on systemic hemodynamics [7]. Importantly,
many such studies have shown clear evidence of a hypodynamic state thus con-
founding septic AKI with AKI secondary to cardiogenic shock and renal
ischemia [7].
Despite all the serious limitations in our understanding of AKI in general
and septic AKI in particular, a variety of paradigms have emerged over the last
30 years to explain AKI, its pathogenesis and its histopathology. They have also
been progressively incorporated into textbooks of medicine, critical care and
nephrology [10–13]. These paradigms provide explanations and descriptions of
the underlying functional and structural changes associated with AKI and offer
a diagnostic and prognostic map to guide clinicians in the interpretation of their
findings. They have progressively become ‘standard teaching’ and, finally, have
ossified into dogma. In this article, we will review some of the problems asso-
ciated with one such paradigm (the ‘pre-renal azotemia’ paradigm) and argue
that such a paradigm suffers from major flaws when applied to critically ill
Pre-Renal Azotemia 3
septic patients. We will then further argue that, given that critically ill septic
patients make up close to 50% of all patients with severe AKI, a paradigm that
cannot apply to such patients cannot be said to reasonably and adequately apply
to AKI in general. Finally, we will argue that a paradigm with such serious flaws
is in need of much rethinking and reviewing.
The Meaning of ‘Pre-Renal’
Any physician in the world is or should be familiar with the term ‘pre-
renal’. The term is widely used in textbooks and the daily practice of clinical
medicine to indicate that the events that might be affecting kidney function are
occurring ‘outside’ the kidney itself and do not include obstruction of the uri-
nary excretory pathways (so-called ‘post-renal’ conditions) [11–13]. However,
the use of the term ‘pre-renal’ often becomes less rigorous as it is also applied
to describe two different but overlapping processes. The first process is the
presence of a trigger outside the kidney which causes AKI. Thus, for example,
a patient with myocardial infarction and low cardiac output syndrome and
rising serum creatinine with oliguria is said to have ‘pre-renal’ renal failure.
Similarly, a patient with severe infectious gastroenteritis and major fluid losses
through diarrhea who has an acute increase in serum creatinine is also said to
have ‘pre-renal’ renal failure. The term ‘pre-renal’ is used here to emphasize
the fact that no ‘parenchymal’ disease (e.g. glomerulonephritis or interstitial
nephropathy) is responsible for the changes in renal function. The second
process describes a condition where the azotemia (increased urea and creati-
nine concentrations) seen in a given patient in the clinical scenario given above
is also taken not only to involve no intrinsic parenchymal disease (see above)
but also to be ‘functional’ and thus relatively rapidly reversible. Here the func-
tional aspect is emphasized to differentiate this pathophysiological state from
one where the pre-renal injury (low cardiac output or severe volume depletion
as described above) has caused ‘structural’ injury or so-called acute tubular
necrosis (ATN; fig. 1). This conceptual framework assumes that, of course,
ATN cannot arise de novo but must by necessity be secondary to sustained or
uncorrected functional injury which has progressed to the point of cell necro-
sis (hence the term ATN). We believe that, this paradigm, although simple and
attractive, should still trigger several logical and pertinent questions before
acceptance:
•
How does one know when ‘functional’ AKI (pre-renal azotemia) becomes
‘structural’ AKI (ATN)?
•
How much structural injury does one need to be able to say that the kidney
has transitioned from ‘pre-renal azotemia’ to ATN?
Bellomo/Bagshaw/Langenberg/Ronco 4
•
What are the treatment implications of being able to accurately classify
AKI into these two types?
•
What experimental data do we have that in the most common condition
leading to AKI (severe sepsis) these pathophysiological states exist?
We will try and use the available evidence and reasoning to address these
questions.
How Do We Know when Pre-Renal Azotemia Becomes ATN?
Logically, in order to know when a clinical entity or syndrome (pre-renal
azotemia) becomes a histopathological entity (ATN), one first needs to have
widely accepted definitions of each. Unfortunately this is currently not the case.
Pre-renal azotemia does not come with ‘consensus criteria’ (whether suffi-
ciently based on evidence or not). Thus it is impossible to know/test/prove that
a given clinician’s pre-renal azotemia is not another clinician’s ATN. This lack
of data and consensus on diagnostic criteria is a huge problem as it makes it
Fig. 1. Convention view of pre-renal azotemia and acute tubular necrosis (ATN).
UNa Urinary sodium; FENa fractional excretion of sodium; UNa/K urinary
sodium/potassium ratio; U/PCr urinary to plasma creatinine ratio; FE
UN
fractional
excretion of urea.
ATN
Pre-renal Azotemia
UNa 15mEq/L
FENa 1%
UNa/K 1/4
U/PCr 20
FE
UN
35%
UNa 20mEq/L
FENa 1%
UNa/K 1/4
U/PCr 15
FE
UN
35%
Pre-Renal Azotemia 5
impossible to know the epidemiology, natural history, inter-rater reliability and
robustness of this paradigm. Further, if we cannot say with some degree of
reproducibility whether a patient does or does not have pre-renal azotemia, how
can we say the he or she does or does not have ATN? Even more problemati-
cally, if we say (as the textbook and literature consistently do) that pre-renal
azotemia can in several cases progress to ATN, how can we decide when the
line between these two conditions that we cannot define has actually been
crossed?
The protagonists of the pre-renal azotemia/ATN paradigm will say that the
answer is simple: either a biopsy will provide the diagnosis or the clinical
course will. Yet both responses are flawed. First, we have absolutely no renal
biopsy series from critically ill patients with clinically suspected ATN let alone
‘pre-renal azotemia’. Thus, we do not actually know that ATN is the histopatho-
logical substrate of non-resolving AKI, which, presumably, most clinicians
would classify as ‘ATN’ and/or, conversely, that it is not the histological sub-
strate of rapidly resolving AKI, which presumably, most clinicians would clas-
sify as ‘pre-renal azotemia’. Second, early postmortem studies of patients in the
intensive care unit (ICU) with sepsis and AKI sufficiently severe to cause death
show that, in 90% cases, the histopathology of the kidney is normal
(Hotchkiss CCM). If people who die of sepsis with AKI have normal renal
histopathology, how can people who are alive with it be expected to have ATN?
Given that sepsis accounts for close to 50% of cases of severe AKI in the ICU,
this is a huge problem. Pre-renal azotemia would then be separated from a diag-
nosis (ATN) that does not exist in the largest group of patients supposed to have
it. If we have no evidence of ATN in sepsis, how can we know that septic
patients with sustained or non-resolving AKI have ATN as the histopathological
substrate while patients with less sustained more rapidly resolving AKI do not?
Again, it is worthy of note that no specific meaning is allocated to words like
‘sustained’ or ‘non-resolving’with values between 48 and 72 h being most com-
monly used but not universally so [10, 14–18]. Finally, if there is an assumed
link between the duration of progressive AKI/azotemia and ATN, how long
should this be? Why 72 h? Why 48? Why 24? Why not 56? The problems with
these nosological entities are obvious.
How Much Structural Injury Is Enough to Change
from ‘Pre-Renal Azotemia’into ‘ATN’?
The protagonists of the pre-renal azotemia/ATN paradigm will see the pre-
vious arguments as insufficient to affect their view of AKI, its course and his-
tology. They will say that ATN is a patchy disease that will be easily missed by
Bellomo/Bagshaw/Langenberg/Ronco 6
a histopathological analysis not specifically searching for its presence and that
it can be easily missed by using renal biopsy methods to diagnose it. In other
words, one can never ‘refute’ its presence to believers. Even accepting this sci-
entifically untenable position, one is still left asking the question, how much
necrosis of tubular cells is needed to diagnose ATN? Is one necrotic cell per
low-powered microscopy field enough? Or does one need 10? How does the
‘necrotic cell count’ correlate with the duration of AKI and its ability to resolve
in less than 48 h in response to appropriate therapy? Does a patient with 100
necrotic cells per low-powered field who improves with appropriate treatment
have pre-renal azotemia or ATN? Could he/she have both simultaneously? What
does ‘improvement in response to therapy’ mean? A return to pre-illness creati-
nine levels? A halt in the progression of azotemia? All of these issues remain
unresolved but cannot be ignored.
What Are the Treatment Implications of Pre-Renal
Azotemia Compared with ATN?
Many of the issues raised above may be easily seen as ‘academic’ rather
than clinical. They appear to deal with scientific rigor rather than the messy but
tangible world of clinical medicine. As such they miss the point. In the clinical
world boundaries are imperfect and recognized as such by physicians. In the
clinical world, paradigms, no matter how flawed, still work reasonably well in
guiding clinicians to the right therapeutic measure. In the clinical world, flawed
paradigms still work a bit like an old map that does not accurately direct the dri-
ver to the right place the way a sophisticated electronic navigation system
might, but is always better than no map at all. In other words, in clinical medi-
cine a ‘flawed paradigm’ is better than no paradigm and always much better
than academic skepticism. The problem with such views is that they are not sus-
tained by sufficient evidence of therapeutic implications. Would a clinician stop
resuscitating a volume-depleted patient with gastroenteritis if he/she thought
that the patient had established ATN instead of pre-renal azotemia? We hope
not. Would a cardiologist not seek immediate revascularization, adequate
inotropic support, pacing if necessary and/or intra-aortic balloon counterpulsa-
tion to treat a patients with myocardial infarction and a rapidly rising creatinine
if he or she thought the patient had ATN instead of pre-renal azotemia? We hope
not. Would a critical care physician and/or nephrologist alone or together not
optimize intravascular volume, cardiac output, oxygenation and blood pressure
while administering appropriate antibiotics and seeking to identify the focus of
infection in a patient with septic AKI if they thought it was due to ATN instead
of pre-renal azotemia? We hope not.
Pre-Renal Azotemia 7
We cannot see any evidence or reason why the paradigm which separates
so-called pre-renal azotemia from so-called ATN can, at this stage, have any
useful therapeutic implications.
What Experimental Data Do We Have that ATN
Occurs in Severe Sepsis?
The protagonists of the pre-renal azotemia/ATN paradigm might be forced
to acknowledge that the evidence for the presence of ATN in human septic AKI
simply does not exist and ATN or its absence can only be inferred. This is
because the factual data are clear. However, they can still argue that, even in the
absence of a tissue diagnosis, the presence or absence of ATN can be confirmed
or refuted by means of urinalysis. This approach introduces a new paradigm,
which is also widely published in textbooks and the literature [11–18], namely
that the measurement of analytes in urine (urea, creatinine, sodium, potassium)
and/or the calculations of derived variables (fractional excretion of sodium,
fractional excretion of urea, various analyte ratios and osmolarity) can be used
to accurately infer preserved tubular function (which equate to pre-renal
azotemia) or lost tubular function (which equates to ATN) [14–18].
This argument is time-honored, widely accepted and is supported by sev-
eral studies [14–18]. However, the problem with this paradigm is that it would
ideally require histopathological confirmation that there is or there is not ATN
for a given set of urinary findings. Unfortunately no such data exist in humans
to confirm the validity of the urinalysis-based approach. Surely the presence of
tubular cell casts proves that there is ATN. Unfortunately, it does not. The cells
in these casts are often viable (accordingly they do not necessarily indicate
‘necrosis’) [19]. Surely, evidence of clearly abnormal urinary biochemistry
proves that tubular function is lost and that tubular damage (ATN) has taken
place. Unfortunately, it does not. Animal experiments show that loss of cell
polarity and relocation of ATPase activity are likely responsible for such
changes, not tubular cell necrosis [20]. Surely, abnormal urinary biochemistry
and microscopy can at least be used to distinguish between sustained (see above
for problems with this concept) from rapidly resolving AKI. This may be true in
ward patients (even though the data are of limited strength) but is certainly not
true of septic critically ill patients. For such patients, often on vasopressor
drugs, receiving large amounts of fluid and often diuretics as well, the data have
been systematically reviewed and found lacking [10].
Even in experimental septic models where the timing of injury is known,
the evidence is simply absent that urinary biochemistry or microscopy can iden-
tify the type, course and outcome of AKI [14].
Bellomo/Bagshaw/Langenberg/Ronco 8
If we cannot use data derived from urine analysis to identify the ‘type’ of
AKI (pre-renal azotemia vs. ATN or sustained vs. rapidly resolving) in septic
patients who make up close to 50% of all patients, how can the accuracy of such
tests be trusted overall?
Conclusions
In this article, we have argued that the paradigm of pre-renal azotemia is
seriously flawed when applied to septic AKI. Given that close to 50% of cases
of severe AKI are associated with severe sepsis, it appears likely that this para-
digm may be flawed in a more general sense. Until our understanding of the
histopathology, clinical course, response to therapy and pathogenesis of septic
AKI has increased sufficiently, we believe it more intellectually honest to talk
about the clinical syndrome of AKI without making any assumptions about the
structural substrate or our ability to predict or reverse its course. If we believe
that, within the syndrome of septic AKI, specific subgroups exist, which can be
usefully identified by clinical tests, epidemiologically separated, shown to carry
a different prognosis, require different treatments and have a different histopathol-
ogy, then we should develop prospective working definitions, diagnostic criteria
and outcome measures and test them in appropriately designed epidemiological
and interventional studies. Until such time, we consider that the flaws of the
current paradigm are too great to go unchallenged.
References
1 Uchino S, Bellomo R, Goldsmith D, Bates S, Ronco C: An assessment of the RIFLE criteria for
acute renal failure in hospitalized patients. Crit Care Med 2006;34:1913–1917.
2 Uchino S, Kellum J, Bellomo R, et al: Acute renal failure in critically ill patients – a multinational,
multicenter study. JAMA 2005;294:813–818.
3 Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P; Acute Dialysis Quality Initiative work-
group: Acute renal failure – definition, outcome measures, animal models, fluid therapy and infor-
mation technology needs: the Second International Consensus Conference of the Acute Dialysis
Quality Initiative (ADQI) Group. Crit Care 2004;8:R204–R212.
4 Bellomo R: Defining, quantifying and classifying acute renal failure. Crit Care Clin 2005;21:
223–237.
5 Metnitz PG, Krenn CG, Steltzer H, et al: Effect of acute renal failure requiring renal replacement
therapy on outcome in critically ill patients. Crit Care Med 2002;30:2051–2058.
6 Bellomo R, Ronco C: Continuous hemofiltration in the intensive care unit. Crit Care 2000;4:
339–345.
7 Langenberg C, Bellomo R, May C, Wan L, Egi M, Morgera S: Renal blood flow in sepsis. Crit
Care 2005;9:R363–R374.
8 Langenberg C, Wan L, Egi M, May CN, Bellomo R: Renal blood flow in experimental septic acute
renal failure. Kidney Int 2006;69:1996–2002.
Pre-Renal Azotemia 9
9 Cole L, Bellomo R, Silvester W, Reeves J: A prospective study of the epidemiology and outcome
of severe acute renal failure patients treated in a ‘closed’ ICU model. Am J Resp Crit Care Med
2000;162:191–196.
10 Bagshaw SM, Langenberg C, Bellomo R: Urinary biochemistry and microscopy in septic acute
renal failure: a systematic review. Am J Kidney Dis 2006;48:695–705.
11 Kasper DL, Braunwald E, Fauci AS, et al. (eds): Harrison’s Principles of Internal Medicine, ed 15.
New York, McGraw-Hill, 2001.
12 Davison A, Cameron JS, Grunfeld JP, Ponticelli C (eds): Oxford Textbook of Clinical Nephrology,
ed 3. Oxford, Oxford University Press, 2005.
13 Fink MP, Abraham E, Vincent J-L, Kochanek PM (eds): Textbook of Critical Care, ed 5.
Philadelphia, Elsevier, 2005.
14 Langenberg C, Wan L, Bagshaw SM, Moritoki E, May CN, Bellomo R: Urinary biochemistry in
experimental septic acute renal failure. Nephrol Dial Transplant 2006;21:3389–3397.
15 Carvounis CP, Nisar S, Guro-Razuman S: Significance of the fractional excretion of urea in the
differential diagnosis of acute renal failure. Kidney Int 2002;62:2223–2229.
16 Espinel CH: The FENa test. Use in the differential diagnosis of acute renal failure. JAMA
1976;236:579–581.
17 Perlmutter M, Grossman SL, Rothenberg S, Dobkin G: Urine/serum urea nitrogen ratio: simple
test of renal function in acute azotemia and oliguria. JAMA 1959;170:1533–1537.
18 Miller TR, Anderson RJ, Linas SL, Henrich WL, Berns AS, Gabow PA, Schrier RW: Urinary diag-
nostic indices in acute renal failure: a prospective study. Ann Intern Med 1978;89:47–50.
19 Graber M, Lane B, Lamina R, Pastoriza-Munoz E: Bubble cells: renal tubular cells in the urinary
sediment with characteristics of viability. J Am Soc Nephrol 1991;1:999–1004.
20 Kwon O, Nelson WJ, Sibley R, et al: Backleak, tight junctions and cell-cell adhesion in post-
ischemic injury in the renal allograft. J Clin Invest 1998;101:2054–2064.
Prof. Rinaldo Bellomo
Department of Intensive Care, Austin Hospital
Melbourne, Vic. 3084 (Australia)
Tel. 61 3 9496 5992, Fax 61 3 9496 3932, E-Mail rinaldo.bellomo@austin.org.au
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 10–16
The Concept of Acute Kidney Injury
and the RIFLE Criteria
John A. Kellum
a
, Rinaldo Bellomo
b
, Claudio Ronco
c
a
Department of Critical Care Medicine, University of Pittsburgh Medical Center,
Pittsburgh, Pa., USA;
b
Department of Intensive Care and Department of Medicine,
Austin Hospital and University of Melbourne, Melbourne, Vic., Australia;
c
Department of Nephrology, Ospedale San Bortolo, Vicenza, Italy
Abstract
Over last half century, the concept of acute renal failure has evolved and with it our
estimates of the incidence, prevalence and mortality. Indeed, until very recently no standard
definition of acute renal failure was available, and this lack of a common language created
confusion and made comparisons all but impossible. In response to the need for a common
definition and classification of acute renal failure, the Acute Dialysis Quality Initiative group
of experts developed and published a set of consensus criteria for defining and classifying
acute renal failure. These criteria which make up acronym ‘RIFLE’ classify renal dysfunc-
tion according to the degree of impairment present: risk (R), injury (I), and failure (F), sus-
tained loss (L) and end-stage kidney disease (E). However, as these criteria were developed,
a new concept immerged. Renal dysfunction was no longer only considered significant when
it reached the stage of failure, but a spectrum from early risk to long-term failure was recog-
nized and codified. Subsequent studies have validated these criteria in various populations
and have shown that relatively mild dysfunction is associated with adverse outcomes. The
term acute kidney injury has subsequently been proposed to distinguish this new concept
from the older terminology of failure.
Copyright © 2007 S. Karger AG, Basel
Acute Kidney Injury: What’s in a Name?
Would a syndrome of any other name still be as deadly? In 1993 the late
Roger Bone penned the following words:
‘Too often, the way we describe a disorder influences, and often limits, the
way we think about that disorder’ [1].
Acute Kidney Injury 11
Although he was talking about sepsis, not renal failure, the similarities are
striking. By focusing on ‘failure’ we might conclude that earlier phases are less
important. By using the term ‘acute tubular necrosis’ we might assume that
ischemia is a predominant pathophysiological mechanism. The entire concept
of ‘pre-renal azotemia’ may have, in its description, influenced our thinking
about a disorder which comprises not discrete entities but a dynamic spectrum
from an early reversible condition to an established disease. Furthermore, given
that acute renal failure (ARF) has been reported to affect from 1 to 25% of
intensive care unit (ICU) patients and has lead to mortality rates from 15 to
60% [2–5], the existing terminology has not influenced our thinking uniformly.
Over the last few years the case for a consensus definition and a classifica-
tion system has repeatedly been made [6, 7]. In these articles we have argued
that the major aim of such a system would be to bring one of the major intensive
care syndromes (acute kidney injury, AKI) to a standard of definition and a
level of classification similar to that achieved by two other common ICU syn-
dromes (sepsis and acute respiratory distress syndrome, ARDS). Following
such advocacy and through the persistent work of the Acute Dialysis Quality
Initiative (ADQI) group, such a system was developed through a broad consen-
sus of experts [8]. The characteristics of this system are summarized in figure 1.
Fig. 1. The RIFLE criteria for AKI (used with permission). ARF Acute renal failure;
GFR glomerular filtration rate; UO urine output.
Risk
Injury
Failure
Loss
ESRD
Increased creatinine 1.5
or GFR decrease 25%
End-stage renal disease
GFR Criteria Urine Output Criteria
UO 0.3 ml/kg/h 24h
or
Anuria 12h
UO 0.5ml/kg/h 12 h
UO 0.5ml/kg/h 6h
Increased creatinine 2
or GFR decrease 50%
Increase creatinine 3
or GFR decrease 75%
or creatinine 4mg/dl
(Acute rise of 0.5mg/dl)
High
sensitivity
High
specificity
Persistent ARFcomplete loss
of renal function 4 weeks
Oliguria
Kellum/Bellomo/Ronco 12
Since its publication, the RIFLE classification system has received much
attention with more than 70,000 electronic hits for its publication site and more
than 70 citations in 2 years. It has also spawned several investigations of its pre-
dictive ability, internal validity, robustness and clinical relevance in a variety of
settings. This review will focus on some of the key findings from these investi-
gations and their broader implications for the concept of AKI.
Validation Studies Using RIFLE
In one of the earliest studies to evaluate RIFLE, Abosaif et al. [9] sought to
evaluate its sensitivity and specificity in patients in the ICU. These investigators
studied 247 patients admitted to the ICU with a serum creatinine of
150 mol/l. This approach identified patients with renal dysfunction on the
first day of admission but not those who developed ARF while in the ICU. Thus,
its findings are limited in scope. The investigators found that the ICU mortality
was greatest among patients classified as RIFLE class ‘F’ (failure) with a
74.5% mortality compared to 50% among those classified as class ‘I’ (injury),
and 38.3% in those classified as RIFLE class ‘R’ (risk). In a significantly larger
single-center multi-ICU study, Hoste et al. [10] evaluated RIFLE as an epi-
demiological and predictive tool in 5,383 critically ill patients. They found that
AKI occurred in a staggering 67% of the patients with 12% achieving a maxi-
mum class of ‘R’, 27% a maximum ‘I’ class, and 28% a maximum ‘F’ class. Of
the 1,510 patients who reached ‘R’, 56% progressed to either ‘I’ or ‘F’. Patients
with a maximum score of ‘R’ had a mortality rate of 8.8%, compared to 11.4%
for ‘I’ and 26.3% for ‘F’. On the other hand, patients who had no evidence of
renal dysfunction had a mortality rate of 5.5%. Furthermore, maximum RIFLE
class ‘I’ (hazard ratio of 1.4) and maximum RIFLE class ‘F’ (hazard ratio of
2.7) were independent predictors of hospital mortality after controlling for
other variables known to predict outcome in critically ill patients. These find-
ings are important as they involve a large cohort of heterogeneous critically ill
patients. They suggest that the incidence of AKI is much greater than previously
appreciated or reported. Even when only patients in RIFLE class ‘F’ are consid-
ered, the incidence of 28% is striking and makes this syndrome, at least in aca-
demic ICUs, more common than ALI or ARDS. The RIFLE classification also
made it possible for these investigators to describe the progression of renal dys-
function over time. Of note, more than 50% of the patients progressed to a more
severe form of renal impairment (RIFLE class ‘I’ or ‘F’) each carrying an inde-
pendent increase in the risk of death.
A further assessment of the validity of the RIFLE classification has been per-
formed in a heterogeneous population of hospitalized patients. Uchino et al. [11]
Acute Kidney Injury 13
focused on the predictive ability of the RIFLE classification in a cohort of
20,126 patients admitted to a teaching hospital for 24 h over a 3-year period.
The authors used the electronic laboratory database and the MDRD equation to
classify patients into the 3 main RIFLE classes and followed them to hospital
discharge or death to assess outcome. They separately analyzed patients who
were admitted once and patients who were readmitted. For both groups the
findings were similar. Close to 10% of patients achieved a maximum RIFLE
class of ‘R’, close to 5% achieved a maximum of ‘I’ and close to 3.5% achieved
a maximum of ‘F’. There was an almost linear increase in hospital mortality
with increasing RIFLE class with patients at ‘R’ having more than 3 times the
mortality rate of patients without AKI. Patients with ‘I’ had close to twice the
mortality of ‘R’ patients and patients with ‘F’ had 10 times the mortality rate of
hospitalized patients without AKI. The investigators performed multivariate
logistic regression analysis to test whether the RIFLE classification was an
independent predictor of hospital mortality. They found that class ‘R’ carried an
odds ratio of hospital mortality of 2.5, class ‘I’ of 5.4, and class ‘F’ of 10.1.
These observations are particularly striking when compared to other important
predictors of outcome such as admission to the ICU (odds ratio of 2.9),
mechanical ventilation (odds ratio of 4.8), or admission to the hematology unit
(odds ratio of 3.1). They suggest that the development of renal injury has
greater prognostic implications that the need for ICU or mechanical ventilation
and that the development of RIFLE class ‘F’ (renal failure) is one of the most
powerful identifiable outcome predictors for hospitalized patients. The findings
of this study are important in the assessment of the RIFLE system, not only
because they link it to outcome but also because they validate its predictive
ability in yet another large population of patients. It is important to note,
however, that hospital mortality may not be the ideal outcome measure for
both syndrome classification systems such as RIFLE or future interventional
trials in such patients. A consensus view is that 60-day mortality may be more
appropriate [8].
RIFLE in Specific Diagnostic Groups
Recently, the RIFLE criteria were evaluated in a cohort of 813 consecutive
patients undergoing cardiac surgery in a university hospital in Finland [12].
These investigators found that 10.9% of patients were classified as ‘R’, and
3.5% as having ‘I’, while 5% developed ‘F’criteria. Compared to a control pop-
ulation of patients who had no evidence of AKI (mortality 0.9%), mortality for
the three groups R, I and F were 8, 21.4 and 32.5%, respectively. Furthermore,
the AUC for the ROC for mortality at 90 days showed good discrimination for
Kellum/Bellomo/Ronco 14
the RIFLE classification with a value of 0.824. Finally, multivariate logistic
regression analysis identified the RIFLE classification as an independent risk
factor for 90-day mortality. These observations are important because they sug-
gest clinical usefulness and good discrimination in a large cohort of patients
and because they are prospective in design. The RIFLE system has also very
recently been tested in a unique population of patients requiring extracorporeal
membrane oxygenation for post-cardiotomy cardiogenic shock [13]. These
patients are uncommon but almost always develop evidence of renal impair-
ment. In a study of 46 such patients, Lin et al. [13] found that there was a pro-
gressive increase in mortality with an increase in RIFLE level, and found that
such a classification calibrated well when tested using the Hosmer-Lemeshow
goodness-of-fit test and that it discriminated well with AUC for the ROC curve
of 0.86. However, this population was small and the confidence intervals wide.
In another restricted population of patients admitted to the ICU with major
burns [14], the RIFLE system also performed well as it did in a recent cohort of
patients with bone marrow transplantation [15]. The RIFLE system has also
been adapted by the Acute Kidney Injury Network, a group involving all major
international societies of critical care medicine and nephrology in an effort to
improve the care of patients with AKI.
Conceptual Developments and Implications
The concept of AKI, as defined by RIFLE, creates a new paradigm. Rather
than focusing exclusively on patients with ARF or on those who receive renal
replacement therapy, the strong association of AKI with hospital mortality
demands that we change the way we think about this disorder. In the study by
Hoste et al. [10], only 14% of patients reaching RIFLE ‘F’ received renal
replacement therapy, yet these patients experienced a hospital mortality more
than 5 times that of the same ICU population without AKI. Is renal support
underutilized or delayed? Are there other supportive measures that should be
employed for these patients? Sustained AKI leads to profound alterations in
fluid, electrolyte, acid-base and hormonal regulation. AKI results in abnormali-
ties in the CNS, immune system and coagulation system. Many patients with
AKI already have multisystem organ failure. What is the incremental influence
of AKI on remote organ function and how does it affect outcome? A recent
study by Levy et al. [16] examined outcomes of over 1,000 patients enrolled in
the control arms of two large sepsis trials. Early improvement (within 24 h) in
cardiovascular (p 0.0010), renal (p 0.0001), or respiratory (p 0.0469)
function was significantly related to survival. This study suggests that outcomes
for patients with severe sepsis in the ICU are closely related to early resolution
Acute Kidney Injury 15
of AKI. While rapid resolution of AKI may simply be a marker of a good prog-
nosis, it may also indicate a window of therapeutic opportunity to improve out-
come in such patients.
Conclusion
AKI, as defined by the RIFLE criteria, is now recognized as a important
ICU syndrome along side other syndromes used in ICU patients for the purpose
of epidemiology and trial execution such as the ALI/ARDS consensus criteria
[17] and the consensus definitions for SIRS/sepsis/severe sepsis and septic
shock [18]. The introduction of the RIFLE system into the clinical arena repre-
sents a useful step in the field of critical care nephrology.
References
1 Bone RC: Why new definitions of sepsis and organ failure are needed. Am J Med 1993;95:
348–350.
2 Silvester W, Bellomo R, Cole L: Epidemiology, management, and outcome of severe acute renal
failure of critical illness in Australia. Crit Care Med 2001;29:1910–1915.
3 Schaefer JH, Jochimsen F, Keller F, et al: Outcome prediction of acute renal failure in medical
intensive care. Intensive Care Med 1991;17:19–24.
4 Liano F, Pascual J; Madrid Acute Renal Failure Study Cluster: Epidemiology of acute renal fail-
ure: a prospective, multicenter, community-based study. Kidney Int 1996;50:811–818.
5 Brivet FG, Kleinknecht DJ, Philippe L, et al: Acute renal failure in intensive care units – causes,
outcome, and prognostic factors of hospital mortality: a prospective, multicenter study. Crit Care
Med 1996;24:192–198.
6 Bellomo R, Kellum J, Ronco C: Acute renal failure: time for consensus. Intensive Care Med
2001;27:1685–1688.
7 Kellum JA, Mehta RL, Ronco C: Acute Dialysis Quality Initiative (ADQI); in Ronco C, Bellomo R,
La Greca G (eds): Blood Purification in Intensive Care. Contrib Nephrol. Basel, Karger, 2001, vol
132, pp 258–265.
8 Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P; Acute Dialysis Quality Initiative work-
group: Acute renal failure – definition, outcome measures, animal models, fluid therapy and infor-
mation technology needs: the Second International Consensus Conference of the Acute Dialysis
Quality Initiative (ADQI) Group. Crit Care 2004;8:R204–R212.
9 Abosaif NY, Tolba YA, Heap M, et al: The outcome of acute renal failure in the intensive care unit
according to RIFLE: model applicability, sensitivity, and predictability. Am J Kidney Dis 2005;46:
1038–1048.
10 Hoste EAJ, Clermont G, Kersten A, et al: RIFLE criteria for acute kidney injury are associated
with hospital mortality in critically ill patients: a cohort analysis. Crit Care 2006;10:R73–R83.
11 Uchino S, Bellomo R, Goldsmith D, et al: An assessment of the RIFLE criteria for acute renal fail-
ure in hospitalized patients. Crit Care Med 2006;34:1913–1917.
12 Kuitunen A, Vento A, Suojaranta-Ylinen R, et al: Acute renal failure after cardiac surgery: evalua-
tion of the RIFLE classification. Ann Thorac Surg 2006;81:542–546.
13 Lin C-Y, Chen Y-C, Tsai F-C, et al: RIFLE classification is predictive of short-term prognosis in
critically ill patients with acute renal failure supported by extra-corporeal membrane oxygenation.
Nephrol Dial Transplant 2006;21:2867–2873.
Kellum/Bellomo/Ronco 16
14 Lopes JA, Jorge S, Neves FC, et al: An assessment of the RIFLE criteria for acute renal failure in
severely burned patients. Nephrol Dial Transplant 2007;22:285.
15 Lopes JA, Jorge S, Silva S, et al: An assessment of the RIFLE criteria for acute renal failure fol-
lowing myeloablative autologous and allogenic haematopoietic cell transplantation. Bone Marrow
Transplant 2006;38:395.
16 Levy MM, Macias WL, Vincent JL, Russell JA, Silva E, Trzaskoma B, Williams MD: Early changes
in organ function predict eventual survival in severe sepsis. Crit Care Med 2005;33: 2194–2201.
17 Bernard G, Artigas A, Briggham KL, et al: The American-European Consensus Conference on
ARDS. Definitions mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir
Crit Care Med 1994;149:818–824.
18 American College of Chest Physicians/Society of Critical Care Medicine Consensus Committee:
Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis.
Chest 1992;101:1658–1662.
John A. Kellum, MD
The CRISMA Laboratory
Department of Critical Care Medicine
University of Pittsburgh, 3550 Terrace Street
Pittsburgh, PA 15261 (USA)
Tel. 1 412 647 6966, Fax 1 412 647 3791, E-Mail kellumja@ccm.upmc.edu
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 17–23
The Liver and the Kidney: Mutual
Clearance or Mixed Intoxication
Vicente Arroyo
Liver Unit, Institute of Digestive and Metabolic Diseases, Hospital Clinic,
University of Barcelona, Barcelona, Spain
Abstract
Hepatorenal syndrome is a frequent complication in patients with cirrhosis, ascites and
advanced liver failure. Its annual incidence in patients with ascites has been estimated at 8%.
Hepatorenal syndrome is a functional renal failure due to low renal perfusion. Renal histol-
ogy is normal or shows lesions that do not justify the decrease in glomerular filtration rate.
The traditional concept is that hepatorenal syndrome is due to deterioration in circulatory
function secondary to an intense vasodilation in the splanchnic circulation (peripheral arterial
vasodilation hypothesis). Over the last decade, however, several features have suggested a
much more complex pathogenesis. In this article new concepts on the pathogenesis of hepa-
torenal syndrome are reported, the current options for prophylaxis are shown, and the most
applicable treatments are described.
Copyright © 2007 S. Karger AG, Basel
Hepatorenal Syndrome: Concept and Clinical Types
Hepatorenal syndrome (HRS) is a frequent complication in patients with
cirrhosis, ascites and advanced liver failure. Its annual incidence in patients
with ascites has been estimated at 8% [1]. HRS is a functional renal failure due
to low renal perfusion. Renal histology is normal or shows lesions that do not
justify the decrease in glomerular filtration rate (GFR). The traditional concept
is that HRS is due to deterioration in circulatory function secondary to an
intense vasodilation in the splanchnic circulation (peripheral arterial vasodila-
tion hypothesis). Over the last decade, however, several features suggest a much
more complex pathogenesis.
Type-1 HRS is characterized by a severe and rapidly progressive renal failure,
which has been defined as a doubling of serum creatinine reaching a level of
Arroyo 18
⬎2.5mg/dl in less than 2 weeks. Although type-1 HRS may develop sponta-
neously, it frequently occurs in close relationship with a precipitating factor, such
as severe bacterial infection, mainly spontaneous bacterial peritonitis (SBP), gas-
trointestinal hemorrhage, major surgical procedure or acute hepatitis super-
imposed to cirrhosis. The association of HRS and SBP has been carefully
investigated [2–4]. Type-1 HRS develops in approximately 25% of patients with
SBP despite a rapid resolution of the infection with non-nephrotoxic antibiotics.
Besides renal failure, patients with type-1 HRS associated with SBP show signs
and symptoms of a rapid and severe deterioration in liver function (jaundice, coag-
ulopathy and hepatic encephalopathy) and circulatory function (arterial hypoten-
sion, very high plasma levels of renin and norepinephrine) [2–5]. It is interesting to
note that in contrast to SBP, sepsis related to other types of infection in patients
with cirrhosis induces type-1 HRS only when there is a lack of response to antibi-
otics [6]. In most patients with sepsis unrelated to SBP responding to antibiotics,
renal impairment, which is also a frequent event, is reversible. Without treatment,
type-1 HRS is the complication of cirrhosis with the poorest prognosis with a
median survival time after the onset of renal failure of only 2 weeks [1].
Type-2 HRS is characterized by a moderate (serum creatinine of
⬍2.5 mg/dl) and slowly progressive renal failure. Patients with type-2 HRS
show signs of liver failure and arterial hypotension but to a lesser degree than
patients with type-1 HRS. The dominant clinical feature is severe ascites with
poor or no response to diuretics (a condition known as refractory ascites).
Patients with type-2 HRS are predisposed to develop type-1 HRS following
SBP or other precipitating events [2–4]. The median survival of patients with
type-2 HRS (6 months) is worse than that of patients with non-azotemic cirrho-
sis with ascites.
Pathogenesis of Type-2 HRS
Portal hypertension in cirrhosis is associated with arterial vasodilation in
the splanchnic circulation due to the local release of nitric oxide and other
vasodilatory substances. The peripheral arterial vasodilation hypothesis pro-
poses that renal dysfunction and type-2 HRS in cirrhosis is related to this fea-
ture (fig. 1). Type-2 HRS represents the extreme expression in the progression
of splanchnic arterial vasodilation. Homeostatic stimulation of the renin-
angiotensin system, the sympathetic nervous system and antidiuretic hormone
is very strong leading to intense renal vasoconstriction and a marked decrease
in renal perfusion and GFR.
Most hemodynamic investigations in cirrhosis have been performed in
non-azotemic patients with and without ascites, and the peripheral arterial
The Liver and the Kidney 19
vasodilation hypothesis was based on these studies. It assumed that type-2 HRS
develops when there is a hyperdynamic circulation and increased cardiac out-
put. However, recent studies assessing cardiovascular function in patients with
type-2 HRS show that cardiac output is significantly reduced compared to
patients without HRS, suggesting that circulatory dysfunction associated with
HRS is due not only to arterial vasodilation but also to a decrease in cardiac
function (fig. 1) [7].
Type-1 HRS Associated with SBP: A Special Form
of Multi-Organ Failure
Type-1 and type-2 HRS have two features in common. They occur in
patients with cirrhosis and ascites, and renal failure is an important component
of both syndromes. However, they show important differences. Type-2 HRS
develops imperceptibly in patients with cirrhosis and ascites who are otherwise
in a stable clinical condition. Circulatory function, although severely deterio-
rated, remains steady or progresses slowly during months as it occurs with the
renal failure. Patients have advanced cirrhosis but the degree of liver failure is
also stable. Hepatic encephalopathy is infrequent. The main clinical problem of
patients with type-2 HRS is refractory ascites. In contrast, type-1 HRS is an
extremely unstable condition. It frequently develops in the setting of an impor-
tant clinical event that acts as a precipitating factor. On the other hand, there is
Fig. 1. Peripheral vasodilation hypothesis (a) and modified peripheral vasodilation
hypothesis (b). According to this latter hypothesis, impairment in arterial blood volume in
cirrhosis could be the consequence of a progression of splanchnic arterial vasodilation and of
a decrease in cardiac output.
Effective arterial
blood volume
Compensated
cirrhosis
Ascites and HRS
ab
Cardiac output
Splanchnic
arterial
vasodilation
Time
Changes
Effective
arterial blood volume
Compensated
cirrhosis
Ascites and HRS
Cardiac output
Splanchnic
arterial
vasodilation
Time
Changes
Arroyo 20
a rapid deterioration in circulatory and renal function within days after the onset
of the syndrome leading to severe arterial hypotension and acute renal failure
with intense oliguria. Finally, there is also rapid deterioration in hepatic func-
tion, with an increase in jaundice and encephalopathy.
Recent studies in patients with SBP have presented data indicating that
type-1 HRS represents a special form of acute multi-organ failure related to the
rapid deterioration in circulatory function (fig. 2). The syndrome develops
when there is a significant decrease in arterial pressure and a marked stimula-
tion of the renin-angiotensin and sympathetic nervous systems in the absence of
changes in systemic vascular resistance, which is consistent with an increase in
arterial vasodilation obscured by the vascular effect of these vasoconstrictor
systems. There is also an acute decrease in the cardiac output that contributes to
effective arterial hypovolemia [5]. The mechanism of this impairment in cardiac
function is complex. There is cirrhotic cardiomyopathy that decreases the car-
diac response to stress conditions. On the other hand, in patients developing
type-1 HRS associated with SBP there is a decrease in cardiopulmonary pres-
sures suggesting a decrease in cardiac preload. Finally, despite the stimulation
of the sympathetic nervous system there is no increase in heart rate indicating
an impaired cardiac chronotropic function.
Fig. 2. Hepatorenal syndrome as part of multi-organ failure. A-II ⫽ Angiotensin II;
NE ⫽ norepinephrine; ADH ⫽ antidiuretic hormone; HRS ⫽ hepatorenal syndrome.
Spontaneous bacterial peritonitis
or other precipitating event
Acute increase in arterial vasodilation
and decrease in cardiac output
↑ A-II, NE, ADH
↑ Resistance to
portal venous flow
Aggravation of
portal hypertension
Kidneys HRS
Encephalopathy
Adrenal insufficiency
Liver failureLiver
Adrenal glands
Brain
Regional arterial
vasoconstriction
The Liver and the Kidney 21
In addition to renal vasoconstriction, patients with type-1 HRS associated
with SBP develop vasoconstriction in the intrahepatic circulation, with a
marked reduction in hepatic blood flow and an increase in portal pressure [5].
The acute deterioration of hepatic function and hepatic encephalopathy may be
related to this feature. Cerebral vascular resistance is increased in patients with
decompensated cirrhosis. A reduction in cerebral blood flow could, therefore,
play a contributory role in hepatic encephalopathy.
Prevention of Type-1 HRS Associated with SBP
Two randomized controlled studies in large series of patients have shown
that type-1 HRS associated with SBP can be prevented either by selective
intestinal decontamination in patients at high risk of developing SBP and HRS
[8] or by circulatory support with intravenous albumin at SBP diagnosis [9].
The first study was performed in cirrhotic patients with a high risk of
developing SBP and type-1 HRS [8]. Primary prophylaxis of SBP, using
long-term oral norfloxacin, was given to patients with low protein ascites of
⬍15 g/l and serum bilirubin of ⱖ3 mg/dl or serum creatinine of ⱖ1.2 mg/dl.
Norfloxacin administration was associated with a significant decrease in the
1-year probability of developing SBP (7 vs. 61%) and type-1 HRS (28 vs. 41%)
and with a significant increase in the 3-month and 1-year probabilities of sur-
vival (94 vs. 62 and 60 vs. 48%, respectively).
In the second study [9], the administration of albumin (1.5 g/kg i.v. at
infection diagnosis and 1g/kg i.v. 48 h later) to patients with cirrhosis and SBP
markedly reduced the incidence of circulatory dysfunction and type-1 HRS
(10% in patients receiving albumin vs. 33% in the control group). The hospital
mortality rate (10 vs. 29%) and 3-month mortality rate (22 vs. 41%) were lower
in patients receiving albumin. Albumin administration to cirrhotic patients with
SBP induces not only an expansion of the plasma volume but also an increase in
systemic vascular resistance. The efficacy of albumin in the prevention of type-1
HRS could, therefore, be related to both an increase in cardiac preload and
cardiac output and a vasoconstrictor effect of albumin in the arterial circulation
related to an attenuation of endothelial dysfunction [10].
Treatment of Type-1 HRS Associated with SBP
Several therapeutic measures can be used in patients developing type-1
HRS associated with SBP. The most effective is liver transplantation. However,
the applicability of this procedure is low. The most applicable treatment
Arroyo 22
consists of the administration of plasma volume expansion with albumin and
vasoconstrictors. The insertion of a transjugular intrahepatic portosystemic
shunt is another possibility that can be used either alone or after reversal of
HRS with vasoconstrictors plus albumin. Finally, extracorporeal albumin dialy-
sis can be used in these patients. Each of these treatments should be considered
after the resolution of infection, since HRS may reverse following effective
antibiotic treatment in a significant number of patients.
Liver Transplantation
Liver transplantation is the treatment of choice for patients with type-1
HRS [11]. Immediately after transplantation a further impairment in GFR may
be observed and many patients require hemodialysis (35% of patients with HRS
as compared with 5% of patients without HRS). After this initial impairment
in renal function, GFR starts to improve. This moderate renal failure persists
during follow-up.
The main problem of liver transplantation in type-1 HRS is its appli-
cability. Due to their extremely short survival, most patients die before trans-
plantation. The introduction of the MELD score, which includes serum
creatinine, bilirubin and the international normalized ratio for listing, has
partially solved the problem as patients with HRS are generally allocated the
first places on the waiting list. Treatment of HRS with vasoconstrictors and
albumin (see below) increases survival in a significant proportion of patients
and therefore the number of patients reaching living transplantation, and
decreases early morbidity and mortality after transplantation and prolongs
long-term survival.
Vasoconstrictors and Albumin
The intravenous administration of vasoconstrictor agents (terlipressin or
noradrenaline) and intravenous albumin over 1–2 weeks is an effective treat-
ment for type-1 HRS [12–14]. The rate of positive response, as defined by a
decrease in serum creatinine to ⬍1.5 mg/dl, is reported in several pilot studies
to be approximately 60%. Remarkably, type-1 HRS does not recur after discon-
tinuation of treatment in most patients.
Reversal of type-1 HRS when terlipressin is given alone (25%) [13] is
lower than that observed in studies in which vasoconstrictors are associated
with intravenous albumin, suggesting that albumin administration is an impor-
tant component in the pharmacological treatment of type-1 HRS. Two random-
ized controlled trials comparing terlipressin plus albumin versus terlipressin
alone have recently been reported in abstract form. Their results confirm that
terlipressin plus albumin is an effective therapy in patients with type-1 HRS
and that reversal of HRS improves survival. However, the effectiveness of the
The Liver and the Kidney 23
treatment reported by these trials (40% rate of reversal of HRS) is lower than
those reported in the pilot studies.
References
1 Gines A, Escorsell A, Gines P, et al: Incidence, predictive factors, and prognosis of the hepatorenal-
syndrome in cirrhosis with ascites. Gastroenterology 1993;105:229–236.
2 Navasa M, Follo A, Filella X, et al: Tumor necrosis factor and interleukin-6 in spontaneous bacte-
rial peritonitis in cirrhosis: relationship with the development of renal impairment and mortality.
Hepatology 1998;27:1227–1232.
3 Follo A, Llovet JM, Navasa M, et al: Renal impairment after spontaneous bacterial peritonitis in cir-
rhosis – incidence, clinical course, predictive factors and prognosis. Hepatology 1994;20:1495–1501.
4 Toledo C, Salmeron JM, Rimola A, et al: Spontaneous bacterial peritonitis in cirrhosis – predictive
factors of infection resolution and survival in patients treated with cefotaxime. Hepatology
1993;17:251–257.
5 Ruiz-del-Arbol L, Urman J, Fernandez J, et al: Systemic, renal, and hepatic hemodynamic derange-
ment in cirrhotic patients with spontaneous bacterial peritonitis. Hepatology 2003;38:1210–1218.
6 Terra C, Guevara M, Torre A, et al: Renal failure in patients with cirrhosis and sepsis unrelated to
spontaneous bacterial peritonitis: value of MELD score. Gastroenterology 2005;129:1944–1953.
7 Ruiz-del-Arbol L, Monescillo A, Arocena C, et al: Circulatory function and hepatorenal syndrome
in cirrhosis. Hepatology 2005;42:439–447.
8 Navasa M, Fernandez J, Montoliu S, et al: Randomized, double-blind, placebo, controlled trial
evaluating norfloxacin in the primary prophylaxis of spontaneous bacterial peritonitis in cirrhotics
with renal impairment, hyponatremia or severe liver failure (abstract). J Hepatol 2006;44:S51.
9 Sort P, Navasa M, Arroyo V, et al: Effect of intravenous albumin on renal impairment and mortality
in patients with cirrhosis and spontaneous bacterial peritonitis. N Engl J Med 1999;341:403–409.
10 Fernandez J, Monteagudo J, Bargallo X, et al: A randomized unblinded pilot study comparing albu-
min versus hydroxyethyl starch in spontaneous bacterial peritonitis. Hepatology 2005;42:627–634.
11 Gonwa TA, Morris CA, Goldstein RM, Husberg BS, Klintmalm GB: Long-term survival and renal
function following liver transplantation in patients with and without hepatorenal syndrome – expe-
rience in 300 patients. Transplantation 1991;51:428–430.
12 Uriz J, Gines P, Cardenas A, et al: Terlipressin plus albumin infusion: an effective and safe therapy
of hepatorenal syndrome. J Hepatol 2000;33:43–48.
13 Ortega R, Gines P, Uriz J, et al: Terlipressin therapy with and without albumin for patients with hepa-
torenal syndrome: results of a prospective, nonrandomized study. Hepatology 2002;36:941–948.
14 Duvoux C, Zanditenas D, Hezode C, et al: Effects of noradrenalin and albumin in patients with
type I hepatorenal syndrome: a pilot study. Hepatology 2002;36:374–380.
V. Arroyo, MD
Liver Unit, Hospital Clínic
Villarroel 170
ES–08036 Barcelona (Spain)
Tel. ⫹34 93 2275400, Fax ⫹34 93 4515522, E-Mail varroyo@clinic.ub.es
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 24–31
Critical Care Nephrology:
A Multidisciplinary Approach
Jean-Louis Vincent
Department of Intensive Care, Erasme Hospital, Free University of Brussels,
Brussels, Belgium
Abstract
Background/Aims: Acute renal failure is a common complication in critically ill
patients, affecting some 25% of intensive care unit (ICU) admissions, and is associated with
high mortality rates of around 40–50%. Acute renal failure in the ICU frequently occurs as
part of multiple organ failure (MOF). Methods: We reviewed the pertinent medical literature
related to the occurrence of acute renal failure in the ICU and its association with other organ
failures. We also reviewed the literature related to different patient management strategies,
notably the differences between ‘closed’ and ‘open’ ICU formats. Results: The increasingly
common association of acute renal failure with other organ failures, in the context of a more
generalized MOF, has important implications on patient care, moving management away from
the realm of nephrologists and towards a more multidisciplinary approach. Closed ICU for-
mats with intensivist-led care, supported by specialist consultation, have been shown to be
associated with improved ICU outcomes. Conclusion: ICU patients with acute renal failure
should be managed using a multidisciplinary team approach led by an intensivist. Good col-
laboration and communication between intensivists and renal and other specialists is essential
to insure the best possible care for ICU patients with renal disease.
Copyright © 2007 S. Karger AG, Basel
Introduction: Acute Renal Failure and
Multiple Organ Failure
Acute renal failure is a common complication in critically ill patients,
affecting some 25% of intensive care unit (ICU) admissions, and is associated
with high mortality rates of around 40–50% [1]. While isolated renal failure
may occur, acute renal failure in the ICU frequently occurs as part of multiple
organ failure (MOF), and is commonly associated with a septic etiology [1–3].
Indeed, patients with severe sepsis and septic shock are at an increased risk of
Critical Care Nephrology 25
developing acute renal failure and have higher mortality rates than patients
without infection [1, 2, 4].
The degree of organ dysfunction in patients with MOF can be assessed
using various organ dysfunction scoring systems, one of the most widely used
being the sequential organ failure assessment (SOFA) score [5]. The SOFA
score (table 1) considers the function of 6 organ systems – cardiovascular, res-
piratory, neurological, hepatic, renal, and coagulation. Data from the SOFA
database showed that 69% of patients with acute renal failure developed MOF
[1]. In addition, while 23% of patients with organ failure had isolated renal fail-
ure on admission, renal failure was often associated with other organ failures –
cardiovascular and hepatic failure (each in 74% of the patients), coagulation (in
72% of the patients), neurologic (67%), and respiratory (57%) failures [6].
It is interesting to note that the mortality from renal failure has not changed
markedly over the years [7] (fig. 1) despite the development and introduction of
new techniques. There are several reasons for this including the lack of effective-
ness of new therapies; for example, even though we have good reason to believe
hemofiltration is superior to intermittent dialysis, there are no real data to support
this [8]. However, perhaps more importantly, the demographics of critically ill
patients with renal failure have changed so that, although therapy and support
may have improved, we are increasingly treating older, sicker patients, with mul-
tiple comorbidities, who survive longer and develop renal failure as part of a more
complex MOF picture with its associated higher mortality rates [1, 9, 10].
Critical Care Nephrology – The Bigger Picture?
Several studies have investigated the importance of acute renal failure in
the ICU in terms of its association with other organ failures. In patients under-
going cardiac surgery, a moderate (20%) increase in plasma creatinine shortly
after surgery occurred in 15.6% of the patients and was associated with other
organ failures in 79% of the patients [11]. The mortality rate in patients with a
postoperative 20% increase in creatinine was 12% in patients with other organ
dysfunctions, but 0% if no other organ dysfunctions were present, demonstrat-
ing the important role of other organ failures on outcomes in this group of
patients. In patients who developed contrast-mediated renal failure, Levy et al.
[12] reported that the mortality rate was 34% compared to 7% in patients who
did not develop renal failure. After adjusting for differences in comorbidity,
renal failure was associated with an odds ratio of dying of 5.5. Patients who
died after developing renal failure had complicated clinical courses with sepsis,
bleeding, delirium, and respiratory failure. Van Biesen et al. [13], showed that
patients with sepsis who later developed acute renal failure needed higher
Vincent 26
Table 1. The sequential organ failure score (SOFA) [5]
Organ system SOFA score
012 3 4
Respiration
PaO
2
/FiO
2
, mm Hg 400 400 300 200 100
with respiratory support
Coagulation
Platelets 10
3
/mm
3
150 150 100 50 20
Liver
Bilirubin, mg/dl 1.2 1.2–1.9 2.0–5.9 6.0–11.9 12.0
(mol/l) (20) (20–32) (33–101) (102–204) (204)
Cardiovascular
Hypotension No MAP dopamine 5 dopamine 5 dopamine 15
hypotension 70 mm Hg or dobutamine or epinephrine 0.1 or epinephrine 0.1
(any dose)
a
or norepinephrine 0.1
a
or norepinephrine 0.1
a
Central nervous system
Glasgow coma score 15 13–14 10–12 6–9 6
Renal
Creatinine, mg/dl 1.2 1.2–1.9 2.0–3.4 3.5–4.9 5.0
(mol/l) (110) (110–170) (171–299) (300–440) (440)
or urine output or 500 ml/day or 200ml/day
a
Adrenergic agents administered for at least 1h (doses given are in g/kg/min).
Critical Care Nephrology 27
inspired oxygen fractions (FiO
2
), even before serum creatinine levels increased,
than patients who did not develop acute renal failure. Using data from the
SOFA database, de Mendonça et al. [1] noted that the number of associated
organ failures on admission in patients with acute renal failure was an indepen-
dent predictor of mortality from acute renal failure (OR 1.24, 95% CI 1.03–1.5,
p 0.02). Acute renal failure with cardiovascular failure was associated with
the highest mortality rate.
Studies have also suggested that renal failure may play a proactive role in
the development and maintenance of MOF [14]. Experimental data indicate
systemic effects of acute renal failure via various mediators [15, 16], and clini-
cal studies have reported that critically ill patients with acute renal failure have
increased oxidative stress compared to other critically ill patients and patients
with end-stage renal disease [17].
In the context of a more generalized MOF, the increasingly common associ-
ation of acute renal failure with other organ failures has important implications
on patient care, moving management away from the realm of the nephrologists
and towards a more multidisciplinary approach.
Fig. 1. Changes in mortality in patients with acute renal failure over the past 48 years.
Line represents numbers of patients included in the studies during each time period. From
Ympa et al. [7] with permission.
42
47
49
63
61
51
52
55
57
0
10
20
30
40
50
60
70
1956–
1960
1961–
1965
1966–
1970
1971–
1975
1976–
1980
1981–
1985
1986–
1990
1991–
1995
1996–
2003
Mortality (%)
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
Number of patients
Vincent 28
The Role of the General Intensivist in Critical Care Nephrology
Intensivists must be trained and ready to take care of a vast range of acute
problems of various etiologies and involving any, or many, organ system(s). As
part of their training, therefore, today’s intensivist must learn how to diagnose,
monitor and manage acute renal failure, acute respiratory failure, circulatory
shock, and any number of other acute and life-threatening conditions. Intensivists
can perhaps be considered as ‘multidisciplinary specialists’.
Importantly, studies have shown that intensivists can really improve patient
outcomes. One way in which the importance of the intensivist has been demon-
strated is in comparing so-called ‘open’ ICUs (those in which patients are cared
for by their admitting physician with limited intensivist input) with ‘closed’
ICUs (in which patient care is transferred to an intensivist); often these studies
are cohort studies assessing outcomes before and after conversion of an open
system to the now more common closed ICU. Several studies have reported
improved resource utilization [18–20] and improved outcomes [19, 21, 22] for
patients managed in a closed compared to an open system. High intensity and
24-hour intensivist coverage have also been associated with improved patient
outcomes [23–25]. In the ‘closed’ ICU system of Victoria, Australia, Cole et al.
[26] reported that patients with acute renal failure were managed by intensivists,
with outcomes comparing favorably to those predicted by illness severity scores,
supporting the ‘closed’ model of care for patients with acute renal failure.
The general intensivist is, therefore, a key element to providing effective
ICU care for all critically ill patients, including those with renal failure, and
other systems of care should be avoided. The nephrologist, therefore, should not
be responsible for ICU patients with acute renal failure (any more than the
pneumologist should care for patients with acute respiratory failure or the car-
diologist for patients with cardiac failure), unless they have received specialist
training in intensive care medicine. This does not, of course, exclude the possi-
bility that many intensivists will have a favorite aspect of intensive care medi-
cine, some preferring respiratory problems, others neurological cases, and
others renal diseases, and their chosen field of expertise will likely result in
research in this area.
In addition, there is no advantage to having multiple small ‘specialist’
ICUs; for example, a renal ICU plus a respiratory ICU plus a trauma ICU, etc.
The problems faced by critically ill patients are basically similar, regardless of
the underlying etiology, and no benefit can be served by spreading resources
and staff among multiple units. Even the separation between surgical and med-
ical ICUs is historical and should be avoided. We need to promote the system of
multidisciplinary departments of intensive care. In our Department of Intensive
Care at Erasme Hospital, Brussels, Belgium, we have 36 beds, with close to
Critical Care Nephrology 29
20 clinical doctors trained in intensive care medicine (and a number of research
fellows), supported by a team of dedicated nurses, physiotherapists, and other
health care personnel.
Collaboration with Nephrologists and Other Specialists
Having promoted above the importance of the closed ICU and the role of
the intensivist in the management of critically ill patients with acute renal fail-
ure, this does not mean there is no place for the nephrologist in the management
of these patients. Indeed, collaboration may be needed with other specialists
also depending on the severity of other associated organ dysfunctions. For
many patients, optimum care will involve collaboration with several specialties;
nevertheless, overall responsibility must lie with the intensivist.
Collaboration with a nephrologist, or other specialist, can be important in
several situations:
Advice and Opinion in Complex Cases
Despite the broad training of modern intensivists, there are cases for which
a specialist consultation is always welcome – however diligent and enthusiastic,
one cannot cover all aspects of every disease state! But such requests must be
reserved for complex cases: a nephrologist should not be called for every olig-
uric patient, just as a neurologist will not be called for every comatose patient or
an endocrinologist for every raised blood glucose level! Excessive consultation
may be a useless consumption of resources and may even have untoward effects
if it causes delay in treatment while waiting for the specialist opinion (although,
in patients with acute renal failure there is usually sufficient time to discuss the
various options). In addition, if specialist consultation is requested for many
patients, the role of the intensivist will be undermined, and the system becomes
increasingly similar to the aforementioned open format again. Importantly very
complex cases with several associated organ failures may require consultation
from several different specialists and good communication is essential to decide
on optimal approaches to treatment; here, the coordinating role of the inten-
sivist can be paramount in integrating and combining specialist opinion.
Need for New Techniques
We are in an era of rapid developments in medical therapies and technol-
ogy, and it is not physically or mentally possible for every physician to keep
abreast of all the latest techniques in all the specialties. Hence, input from a spe-
cialist may be welcome in a patient needing treatment with a new and unfamil-
iar agent or technique.
Vincent 30
Scientific Discussions, Research, and Training
Importantly, specialists should be invited to seminars and other scientific
discussions within the intensive care department whenever the topic includes
aspects relevant to their specific field. Similarly, members of other specialties
should be included in clinical research when it involves specialized aspects of
their particular field. In addition, specialists from all departments, who are most
likely to be able to provide the latest specialist data and results in their field, must
be involved in the training and supervision of junior intensivists and students.
Conclusion
Critically ill patients with renal failure often have multiple ongoing acute
processes and require a multidisciplinary approach to management. This is ide-
ally performed in a multidisciplinary ICU with patients cared for by trained
intensivists available 24h/day, 7 days/week. This does not mean there is no
place for the nephrologist in the care of such patients, but overall responsibility
must lie with the intensivist. Any consultant from any specialty should be wel-
come on the ICU at any time; the patient must be at the center of attention, and
visits from other specialists must be welcomed if they can improve patient care.
In collaborating with other specialties it is important that each party acknowl-
edges the respective competences of the other, and that there be open and hon-
est discussion of all issues associated with patient management. Critical care
nephrology is a multidisciplinary field, and only a multidisciplinary team
approach to patient management, under the ultimate coordination and control of
a trained intensivist, will provide the best possible care for ICU patients with
renal disease.
References
1 de Mendonca A, Vincent JL, Suter PM, et al: Acute renal failure in the ICU: risk factors and out-
come evaluated by the SOFA score. Intensive Care Med 2000;26:915–921.
2 Brivet FG, Kleinknecht DJ, Loirat P, et al: Acute renal failure in intensive care units – causes, out-
come, and prognostic factors of hospital mortality: a prospective, multicenter study. French Study
Group on Acute Renal Failure. Crit Care Med 1996;24:192–198.
3 Uchino S, Kellum JA, Bellomo R, et al: Acute renal failure in critically ill patients: a multina-
tional, multicenter study. JAMA 2005;294:813–818.
4 Neveu H, Kleinknecht D, Brivet F, et al: Prognostic factors in acute renal failure due to sepsis:
results of a prospective multicenter study. Nephrol Dial Transplant 1996;11:293–299.
5 Vincent JL, Moreno R, Takala J, et al: The SOFA (Sepsis-related Organ Failure Assessment) score
to describe organ dysfunction/failure. Intensive Care Med 1996;22:707–710.
6 Vincent JL, de Mendonça A, Cantraine F, et al: Use of the SOFA score to assess the incidence of
organ dysfunction/failure in intensive care units: results of a multicentric, prospective study. Crit
Care Med 1998;26:1793–1800.
Critical Care Nephrology 31
7 Ympa YP, Sakr Y, Reinhart K, et al: Has mortality from acute renal failure decreased? A system-
atic review of the literature. Am J Med 2005;118:827–832.
8 Vinsonneau C, Camus C, Combes A, et al: Continuous venovenous haemodiafiltration versus
intermittent haemodialysis for acute renal failure in patients with multiple-organ dysfunction syn-
drome: a multicentre randomised trial. Lancet 2006;368:379–385.
9 Cappi SB, Sakr Y, Vincent JL: Daily evaluation of organ function during renal replacement therapy
in intensive care unit patients with acute renal failure. J Crit Care 2006;21:179–183.
10 Mehta RL, Pascual MT, Soroko S, et al: Spectrum of acute renal failure in the intensive care unit:
the PICARD experience. Kidney Int 2004;66:1613–1621.
11 Ryckwaert F, Boccara G, Frappier JM, et al: Incidence, risk factors, and prognosis of a moderate
increase in plasma creatinine early after cardiac surgery. Crit Care Med 2002;30:1495–1498.
12 Levy EM, Viscoli CM, Horwitz RI: The effect of acute renal failure on mortality. A cohort analysis.
JAMA 1996;275:1489–1494.
13 Van Biesen W, Yegenaga I, Vanholder R, et al: Relationship between fluid status and its manage-
ment on acute renal failure (ARF) in intensive care unit (ICU) patients with sepsis: a prospective
analysis. J Nephrol 2005;18:54–60.
14 Van Biesen W, Lameire N, Vanholder R, et al: Relation between acute kidney injury and multiple-
organ failure: the chicken and the egg question. Crit Care Med 2007;35:316–317.
15 Rabb H, Wang Z, Nemoto T, et al: Acute renal failure leads to dysregulation of lung salt and water
channels. Kidney Int 2003;63:600–606.
16 Kelly KJ: Distant effects of experimental renal ischemia/reperfusion injury. J Am Soc Nephrol 2003;14:
1549–1558.
17 Himmelfarb J, McMonagle E, Freedman S, et al: Oxidative stress is increased in critically ill
patients with acute renal failure. J Am Soc Nephrol 2004;15:2449–2456.
18 Multz AS, Chalfin DB, Samson IM, et al: A ‘closed’ medical intensive care unit (MICU) improves
resource utilization when compared with an ‘open’ MICU. Am J Respir Crit Care Med 1998;157:
1468–1473.
19 Ghorra S, Reinert SE, Cioffi W, et al: Analysis of the effect of conversion from open to closed sur-
gical intensive care unit. Ann Surg 1999;229:163–171.
20 Hanson CW, Deutschman CS, Anderson HL, et al: Effects of an organized critical care service on
outcomes and resource utilization: a cohort study. Crit Care Med 1999;27:270–274.
21 Baldock G, Foley P, Brett S: The impact of organisational change on outcome in an intensive care
unit in the United Kingdom. Intensive Care Med 2001;27:865–872.
22 Topeli A, Laghi F, Tobin MJ: Effect of closed unit policy and appointing an intensivist in a devel-
oping country. Crit Care Med 2005;33:299–306.
23 Pronovost PJ, Angus DC, Dorman T, et al: Physician staffing patterns and clinical outcomes in
critically ill patients: a systematic review. JAMA 2002;288:2151–2162.
24 Blunt MC, Burchett KR: Out of hours consultant cover and case-mix-adjusted mortality in inten-
sive care. Lancet 2000;356:735–736.
25 Higgins TL, McGee WT, Steingrub JS, et al: Early indicators of prolonged intensive care unit stay:
impact of illness severity, physician staffing, and pre-intensive care unit length of stay. Crit Care
Med 2003;31:45–51.
26 Cole L, Bellomo R, Silvester W, et al: A prospective, multicenter study of the epidemiology, man-
agement, and outcome of severe acute renal failure in a ‘closed’ ICU system. Am J Respir Crit
Care Med 2000;162:191–196.
Dr. J.-L. Vincent
Erasme University Hospital
Route de Lennik 808
BE–1070 Brussels (Belgium)
Tel. 32 2 555 3380, Fax 32 2 555 4555, E-Mail jlvincen@ulb.ac.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 32–38
Incidence, Classification, and Outcomes
of Acute Kidney Injury
Eric A.J. Hoste
a,b
, John A. Kellum
b
a
Intensive Care Unit, Ghent University Hospital, Ghent, Belgium;
b
The Clinical
Research, Investigation, and Systems Modeling of Acute Illness (CRISMA)
Laboratory, Department of Critical Care Medicine, University of Pittsburgh,
School of Medicine, Pittsburgh, Pa., USA
Abstract
Background: Traditionally the epidemiology of acute renal failure was assessed in
patients requiring renal replacement therapy. Recent data emphasized the importance of less
severe impairment of kidney function, hence the terminology acute kidney injury (AKI) was
introduced. Methods: In this paper we present a review of current published data on the epi-
demiology of AKI. Results: The RIFLE classification categorizes the whole severity range
of AKI into 3 severity categories and 2 outcome classes. AKI is associated with increased
costs and worse outcomes. Increasing severity classes are associated with increasing morbid-
ity and mortality. There is an increasing incidence of AKI, while mortality seems to decrease.
Conclusion: Small changes in kidney function have an impact on outcomes and this knowl-
edge has led to the introduction of the terminology AKI, encompassing both discrete and
severe impairment of kidney function. The RIFLE classification describes the whole range
of AKI and has been validated in multiple cohorts. As a consequence of increasing comor-
bidity, the incidence of AKI is increasing. The incidence of acute renal failure requiring renal
replacement therapy even compares to that of acute lung injury, and up to two thirds of gen-
eral ICU patients meet RIFLE criteria for AKI.
Copyright © 2007 S. Karger AG, Basel
The epidemiology of acute renal failure (ARF) has undergone considerable
change over the years. This is not only caused by a change in patient character-
istics, but probably even more importantly, by a change in definition of the dis-
ease. It was Homer W. Smith [1] who introduced the term ‘acute renal failure’
in a chapter on Acute Renal Failure Related to Traumatic Injuries, in his text-
book The Kidney: Structure and Function in Health and Disease (1951). Since
then ARF has become an established term in medical parlance. Confusingly, the
Epidemiology and Pathogenesis of AKI, Sepsis and MOF
Epidemiology of AKI 33
term ARF covers many different meanings. Workgroup 1 of the second confer-
ence of the Acute Dialysis Quality Initiative (ADQI) found in 2002 that over
30 definitions of ARF were used in medical literature, ranging from a 25%
increase of serum creatinine to the need for renal replacement therapy (RRT)
(http://www.ccm.upmc.edu/adqi/ADQI2/ADQI2g1.pdf) [2]. Different defini-
tions of acute kidney injury (AKI) describe different cohorts. This is nicely
illustrated in a study on the occurrence of AKI in 9,210 hospitalized patients in
whom 2 or more serum creatinine levels were assessed [3]. The incidence of
AKI ranged from 1 to 44%.
Initially most emphasis in ARF research was on patients with severe
impairment of kidney function, e.g. defined by the need for RRT. More
recently, several authors have demonstrated that small changes in serum creati-
nine are independently associated with increased morbidity and mortality. Levy
et al. [4] demonstrated that a 25% increase of serum creatinine following radio-
contrast administration was associated with a 5-fold increased risk for in-hospital
mortality. Lassnigg et al. [5] demonstrated similar findings in cardiac surgery
patients. Finally, Chertow et al. [3] demonstrated that a ⱖ0.3 mg/dl increase of
serum creatinine was associated with greater cost, morbidity and mortality in
hospitalized patients. The emphasis of interest is therefore more and more shift-
ing to less severe derangements of kidney function consistent with several calls
toward definitions which identify this patient population [6, 7]. Hence the ter-
minology acute kidney injury (AKI) was introduced. In order to meet the need
for a uniform definition which included different severity grades of AKI the
RIFLE classification was developed by ADQI [7]. This classification system
(table 1) has now been validated in numerous settings [8].
Epidemiology of AKI and ARF
ARF
ARF severe enough to require RRT occurs in approximately 5% of general
ICU patients [9]. This proportion may vary according to the ICU cohort
described, e.g. cardiac surgery versus medical ICU but appears to be surpris-
ingly uniform around the world. Over a period of almost 20 years the incidence
of ARF treated with RRT has more than doubled. Feest et al. [10] and Waikar
et al. [11] found in a cohort from the late 1980s that the incidence of ARF patients
treated with RRT was less than 50 patients per million population. Recent data
from the beginning of this century report an incidence rate of even 270 patients
per million population [11]. In comparison, the incidence of acute lung injury
was estimated at 112–320 patients per million population [12]. ARF requiring
RRT has therefore a comparable incidence to that of acute lung injury. The
Hoste/Kellum 34
reason for the increasing incidence of ARF can probably be explained by the
change in baseline characteristics of patients. Patients nowadays are older, have
more comorbid disease, and are more severely ill at the start of RRT [13, 14].
From ARF to AKI
The incidence of less severe AKI has also increased over time. Perhaps, the
best evidence of this comes from the studies of Hou et al. [15] and Nash et al. [16].
These authors evaluated the occurrence of AKI in a single hospital in 1979 and
1996, and demonstrated that the proportion of patients with AKI had increased
from 4.9 to 7.2% of all hospitalized patients. Two large multicenter, longitudi-
nal databases on AKI in the United States also showed that the incidence has
increased. Both studies used administrative databases, and scored occurrence of
AKI on the basis of reporting of International Classification of Diseases, Ninth
Revision (ICD-9) codes for ARF which include cases we would now consider to
have AKI [11, 17]. Waikar et al. [11] found that the incidence quadrupled from
610 to 2,880 patients per million population during the 15-year study period.
Xue et al. [17] found that during the period from 1992 to 2001, there was a 11%
Table 1. RIFLE classification
Glomerular filtration rate criteria Urine output criteria
Risk serum creatinine ⫻ 1.5 UO ⬍ 0.5 ml/kg/h ⫻ 6h
Injury serum creatinine ⫻ 2 UO ⬍ 0.5 ml/kg/h ⫻ 12 h
Failure serum creatinine ⫻ 3 UO ⬍ 0.3 ml/kg/h ⫻ 24 h
or or
serum creatinine ⱖ4 mg/dl with an anuria ⫻ 12 h
acute rise ⬎0.5 mg/dl
Loss persistent ARF ⫽ complete loss of kidney function ⬎4 weeks
ESKD ESKD ⬎3 months
For conversion of creatinine expressed in conventional units to SI units, multiply by 88.4.
RIFLE class is determined based on the worst of either glomerular filtration criteria or urine
output (UO) criteria. Glomerular filtration criteria are calculated as an increase of serum cre-
atinine above the baseline serum creatinine level. Acute renal dysfunction should be both
abrupt (within 1–7 days) and sustained (more than 24h). When the baseline serum creatinine
is not known and patients are without a history of chronic renal insufficiency, it is recom-
mended to calculate a baseline serum creatinine using the modification of diet in renal dis-
ease (MDRD) equation for assessment of renal function, assuming a glomerular filtration
rate of 75 ml/min/1.73 m
2
. When baseline serum creatinine is elevated, an abrupt rise of at
least 0.5 mg/dl to more than 4 mg/dl is all that is required to achieve ‘failure’.
Epidemiology of AKI 35
yearly increase of the diagnosis of AKI. A limitation to these studies may be
that ICD coding does not currently define strict cutoffs for AKI. Consequently,
sensitivity to detect AKI is low, or in other words, many patients with AKI are
missed when epidemiology is based on administrative databases. Also, report-
ing of AKI may vary across hospitals and during different time periods.
AKI Defined by RIFLE Criteria
The RIFLE classification was used to evaluate the in-hospital epidemiol-
ogy of AKI in a single center setting in Melbourne, Australia [18]. In a cohort of
over 20,000 hospitalized patients, 18% developed AKI according to RIFLE cri-
teria. The maximum severity of AKI was risk in 9.1% of patients, injury in
5.2%, and failure in 3.7% of patients. Our group evaluated the incidence of AKI
in a cohort of 5,383 ICU patients admitted during a 1-period to a tertiary care
hospital. In this cohort two thirds developed AKI according to the RIFLE clas-
sification, 12.4% of patients had a maximum RIFLE Risk, 26.7% had maxi-
mum RIFLE Injury and 28.1% had maximum RIFLE Failure [19].
Outcome of ARF and AKI
Length of Stay
Patients with AKI or ARF are amongst the most severely ill in the ICU.
Therefore, it is not surprising that ICU patients with AKI or ARF have a longer
length of stay in the ICU and in the hospital compared to ICU patients without
these conditions. Patients with less severe forms of AKI have an increased
length of in-hospital stay compared to patients without AKI, and there is a step-
wise increase of length of stay depending on the severity of AKI according to
RIFLE criteria (length of stay for patients without AKI 6 days, RIFLE Risk 8
days, RIFLE Injury 10 days, RIFLE Failure 16 days; p ⬍ 0.01) [19].
End-Stage Kidney Disease
Although the majority of survivors regain kidney function, some patients
do not recover from ARF and evolve to end-stage kidney disease (ESKD) with
permanent need for RRT. In the large multicenter BEST Kidney study 13.8%
(95% confidence interval 11.2–16.3%) of survivors developed ESKD [9]
(RIFLE ‘E’).
Mortality
The in-hospital mortality rate for general ICU patients with ARF who are
treated with RRT is approximately 60% [9]. Mortality rate for ICU patients
with ARF will depend upon the severity of AKI, and the case mix of the
Hoste/Kellum 36
observed cohort (fig. 1) [3, 5, 9, 18, 19]. Despite advances in treatment, the
published mortality rates for ARF patients in individual studies remain more or
less constant at 50% from 1956 on [20]. However, longitudinal collected data
demonstrated an improvement of outcome [11, 17, 21].
Long-Term Outcome
Increasing severity of AKI is associated with increasing 1-mortality rate,
and patients with AKI or ARF have a worse 1-year survival compared to non-
ARF patients [22]. However, after hospital discharge survival curves for
patients with different severity of AKI parallel each other. One-year mortality
rate is comparable for patient groups with different severity of AKI [22].
Patients Die of AKI
Optimism from the past that dictated that renal function could be substituted
by RRT, and therefore, patients died with ARF but not because of ARF, has been
replaced by realism. Observed mortality in patients with ARF or AKI is signifi-
cantly higher than predicted from underlying disease and AKI has been shown to
be an independent predictor of mortality. This appears to be true for the entire
spectrum of AKI, from patients with only minor severity of AKI to patients with
ARF, and for ICU and non-ICU patients. Patients with 25% increase of serum
creatinine after radiocontrast procedures had an in-hospital mortality of 34%
compared to 7% for carefully matched patients without increase of serum creati-
nine (odds ratio 5.5) [4]. After this publication there were several others which
Fig. 1. In-hospital mortality for ICU patients without AKI and ICU patients with
increasing RIFLE class [after 18].
5.50%
8.80%
11.40%
26.30%
0
5
10
15
20
25
30
No AKI RIFLE Risk RIFLE Injury RIFLE Failure
In-hospital mortality (%)
Epidemiology of AKI 37
confirmed the association of small absolute or relative increases of serum creati-
nine in hospitalized, non-ICU patients and in-hospital mortality [3]. In ICU
patients minor severity of AKI is also associated with worse outcome, even after
correction for covariates. This has been demonstrated in several studies that used
the RIFLE classification for AKI, especially classes ‘I’ and ‘F’, and also in stud-
ies that evaluated small increases of serum creatinine [3, 5, 18, 19].
Conclusions
Small changes in kidney function have an impact on outcomes and this
knowledge has led to the introduction of the terminology AKI, encompassing
both discrete and severe impairment of kidney function. The RIFLE classifica-
tion describes the whole range of AKI and has been validated in multiple
cohorts. As a consequence of increasing comorbidity, the incidence of AKI is
increasing. The incidence of ARF requiring RRT even compares to that of acute
lung injury, and up to two thirds of general ICU patients meet RIFLE criteria
for AKI.
References
1 Smith HW: Acute renal failure related to traumatic injuries; in The Kidney: Structure and Function
in Health and Disease. Cary, Oxford University Press, 1951.
2 Kellum JA, Levin N, Bouman C, Lameire N: Developing a consensus classification system for
acute renal failure. Curr Opin Crit Care 2002;8:509–514.
3 Chertow GM, Burdick E, Honour M, Bonventre JV, Bates DW: Acute kidney injury, mortality,
length of stay, and costs in hospitalized patients. J Am Soc Nephrol 2005;16:3365–3370.
4 Levy EM, Viscoli CM, Horwitz RI: The effect of acute renal failure on mortality. A cohort analy-
sis. JAMA 1996;275:1489–1494.
5 Lassnigg A, Schmidlin D, Mouhieddine M, Bachmann LM, Druml W, Bauer P, Hiesmayr M:
Minimal changes of serum creatinine predict prognosis in patients after cardiothoracic surgery: a
prospective cohort study. J Am Soc Nephrol 2004;15:1597–1605.
6 Bellomo R, Kellum J, Ronco C: Acute renal failure: time for consensus. Intensive Care Med
2001;27:1685–1688.
7 Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P; ADQI workgroup: Acute renal failure –
definition, outcome measures, animal models, fluid therapy and information technology needs:
the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI)
Group. Crit Care 2004;8:R204–R212.
8 Hoste EA, Kellum JA: Acute kidney injury: epidemiology and diagnostic criteria. Curr Opin Crit
Care 2006;12:531–537.
9 Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C,
Macedo E, et al: Acute renal failure in critically ill patients: a multinational, multicenter study.
JAMA 2005;294:813–818.
10 Feest TG, Round A, Hamad S: Incidence of severe acute renal failure in adults: results of a com-
munity based study. BMJ 1993;306:481–483.
11 Waikar SS, Curhan GC, Wald R, McCarthy EP, Chertow GM: Declining mortality in patients with
acute renal failure, 1988 to 2002. J Am Soc Nephrol 2006;17:1143–1150.
Hoste/Kellum 38
12 Goss CH, Brower RG, Hudson LD, Rubenfeld GD: Incidence of acute lung injury in the United
States. Crit Care Med 2003;31:1607–1611.
13 Ostermann ME, Taube D, Morgan CJ, Evans TW: Acute renal failure following cardiopulmonary
bypass: a changing picture. Intensive Care Med 2000;26:565–571.
14 McCarthy JT: Prognosis of patients with acute renal failure in the intensive care unit: a tale of two
eras. Mayo Clin Proc 1996;71:117–126.
15 Hou S, Bushinsky D, Wish J, Cohen J, Harrington J: Hospital-acquired renal insufficiency: a
prospective study. Am J Med 1983;74:243–248.
16 Nash K, Hafeez A, Hou S: Hospital-acquired renal insufficiency. Am J Kidney Dis 2002;39:
930–936.
17 Xue JL, Daniels F, Star RA, Kimmel PL, Eggers PW, Molitoris BA, Himmelfarb J, Collins AJ:
Incidence and mortality of acute renal failure in Medicare beneficiaries, 1992 to 2001. J Am Soc
Nephrol 2006;17:1135–1142.
18 Uchino S, Bellomo R, Goldsmith D, Bates S, Ronco C: An assessment of the RIFLE criteria for
acute renal failure in hospitalized patients. Crit Care Med 2006;34:1913–1917.
19 Hoste EA, Clermont G, Kersten A, Venkataraman R, Angus DC, De Bacquer D, Kellum JA:
RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill
patients: a cohort analysis. Crit Care 2006;10:R73.
20 Ympa YP, Sakr Y, Reinhart K, Vincent JL: Has mortality from acute renal failure decreased?
A systematic review of the literature. Am J Med 2005;118:827–832.
21 Desegher A, Reynvoet E, Blot S, De Waele J, Claus S, Hoste E: Outcome of patients treated with
renal replacement therapy for acute kidney injury. Crit Care 2006;10:P296.
22 Bagshaw SM, Mortis G, Doig CJ, Godinez-Luna T, Fick GH, Laupland KB: One-year mortality in
critically ill patients by severity of kidney dysfunction: a population-based assessment. Am J
Kidney Dis 2006;48:402–409.
Eric A.J. Hoste
ICU, 2K12-C, Ghent University Hospital
De Pintelaan 185
BE–9000 Gent (Belgium)
Tel. ⫹32 9 240 27 75, Fax ⫹32 9 240 49 95, E-Mail Eric.Hoste@UGent.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 39–46
Pathophysiology of Acute Kidney Injury:
Roles of Potential Inhibitors of
Inflammation
Joseph V. Bonventre
Renal Division, Brigham and Women’s Hospital and Department of Medicine, Harvard
Stem Cell Institute, Harvard Medical School and Harvard-Massachusetts Institute of
Technology, Division of Health Sciences and Technology, Boston, Mass., USA
Abstract
The pathogenesis of acute kidney injury (AKI) is complex and varies to some extent
based on the particular cause. Inflammation contributes to this pathophysiology in a variety of
contexts. Inflammation can result in reduction in local blood flow to the outer medulla with
adverse consequences on tubule function and viability. Both the innate and adaptive immune
responses are important contributors. With ischemia/reperfusion endothelial cells upregulate a
number of adhesion molecules which have counterreceptors on leukocytes. A number of
vasoactive mediators that are released with injury, such as nitric oxide, may also affect leuko-
cyte-endothelial interactions. Tubule epithelial cells generate proinflammatory and chemotac-
tic cytokines. We and others have found that injection of mesenchymal stem (stromal) cells is
protective against renal injury as assessed by serum creatinine measured 24h after ischemia.
The mechanism of such protection may be through intrarenal paracrine effects to decrease
inflammation or by systemic immune modulation. Resolvins (Rv) and protectins (PD) have
been identified as two newly identified families of naturally occurring n–3 fatty acid docosa-
hexaenoic acid metabolites. In collaboration with Serhan et al. we recently reported that, in
response to bilateral ischemia/reperfusion injury, mouse kidneys produce D series resolvins
(RvDs) and PD1 [J Immunol 2006;177:5902–5911]. Administration of RvDs or PD1 to mice
prior to, or subsequent to, ischemia resulted in a reduction in functional and morphological
kidney injury. Understanding how these anti-inflammatory processes are regulated may pro-
vide insight into how we might intervene to facilitate and enhance them so that we might pre-
vent or mitigate the devastating consequences of AKI.
Copyright © 2007 S. Karger AG, Basel
Body homeostasis depends critically on the ability of the kidney to func-
tion normally. There are many potential causes of acute kidney injury (AKI),
Bonventre 40
some of which are related to a mismatch between oxygen and nutrient delivery
to the nephrons and energy demand of the nephrons. Other causes relate to
direct toxic effects of substances on the epithelium. The kidney is particularly
susceptible to toxic effects of many environmental substances or therapeutics
since many of these compounds are increased in concentration as glomerular
filtrate is reabsorbed from the tubule as the filtrate moves down the nephron. In
many situations in humans acute injury is superimposed on chronic renal dis-
ease and AKI is increasingly being recognized as an important precipitant in
progression to end-stage renal disease.
Whether the injury is related to oxygen deprivation or toxins there are
many common features of the epithelial cell response. The processes of injury
and repair to the kidney epithelium is depicted schematically in figure 1. Injury
results in rapid loss of cytoskeletal integrity and cell polarity. There is shedding
of the proximal tubule brush border, loss of polarity with mislocalization of
adhesion molecules and other membrane proteins such as the Na
⫹
K
⫹
ATPase
and -integrins [1], as well as apoptosis and necrosis [2]. With severe injury,
viable and nonviable cells are desquamated leaving regions where the basement
membrane remains as the only barrier between the filtrate and the peritubular
interstitium. This allows for backleak of the filtrate, especially under circum-
stances where the pressure in the tubule is increased due to intratubular obstruc-
tion resulting from cellular debris in the lumen interacting with proteins such as
fibronectin which enter the lumen [3]. This injury to the epithelium results in
the generation of inflammatory and vasoactive mediators, which can act on the
vasculature to worsen the vasoconstriction and inflammation. Thus inflamma-
tion contributes in a critical way to the pathophysiology of AKI [4]. In contrast
to the heart or brain, the kidney can recover from an ischemic or toxic insult that
results in cell death, although it is becoming increasingly recognized that there
are longer-term detrimental effects of even brief periods of ischemia [5].
Surviving cells that remain adherent undergo repair with the potential to
recover normal renal function. Whether there is a subpopulation of stem or
progenitor cells is a matter of active study at this point in time [6]. When the
kidney recovers from acute injury it relies on a sequence of events that include
epithelial cell spreading and migration to cover the exposed areas of the base-
ment membrane, cell dedifferentiation and proliferation to restore cell number,
followed by differentiation which results in restoration of the functional
integrity of the nephron [7]. The potential role of stem cells derived from the
bone marrow to directly replace cells lost by the injury has been addressed in a
number of publications. We and others have concluded that the bone marrow
does not contribute directly to the replacement of cells but bone marrow-
derived cells may have paracrine effects that may facilitate repair, potentially by
reducing inflammation [6]. In this brief review I will focus on the role of
Pathophysiology of Acute Kidney Injury 41
inflammation and the potential contribution of factors that can mitigate the
inflammation such as bone marrow-derived stem cells via paracrine mecha-
nisms, and naturally occurring anti-inflammatory compounds: resolvins and
protectins.
Fig. 1. Injury and repair to the epithelial cells of the kidney with ischemia/reperfusion.
With injury to the kidney an early response is loss of the polarity of the epithelial cells with
mislocation of adhesion molecules and Na
⫹
/K
⫹
-ATPase and other proteins. In addition there
is a loss of the brush border. With increasing injury, there is cell death by either necrosis or
apoptosis. Some of the necrotic debris is then released into the lumen, where it interacts with
luminal proteins and can ultimately result in obstruction. In addition, because of the misloca-
tion of adhesion molecules viable epithelial cells lift off the basement membrane and are
found in the urine. The kidney can respond to the injury by initiating a repair process if there
are sufficient nutrients and sufficient oxygen delivery and the basement membrane integrity
has not been altered irreparably. Viable epithelial cells migrate and cover denuded areas of the
basement membrane. The source of these cells appears to be from the kidney itself and not
from the bone marrow. Bone marrow cells may contribute to the interstitial cellular infiltrate
and may produce factors that may modulate inflammation and facilitate repair. Cells replacing
the epithelium may derive from differentiated epithelial cells or from a subpopulation of prog-
enitor cells in the tubule or in the interstitium. The cells which populate the recovering epithe-
lium express proteins that are not normally expressed in an adult mature epithelial cell. The
cells then undergo division and replace lost cells. Ultimately, the cells go on to differentiate
and reestablish the normal polarity of the epithelium. In this figure the cells depicted as dark
cells are those that might represent a progenitor pool. The existence and potential role of a
subpopulation of stem/progenitor cells in this process of repair is controversial.
Loss of polarity
Normal Epithelium
Stem cell migration to injured area
Differentiation and
reestablishment
of polarity
Sloughing of viable and dead cells
with luminal obstruction
Ischemia/
reperfusion
Apoptosis
Necrosis
Cell death
Proliferation
Adhesion molecules
Na
⫹
/K
⫹
-ATPase
Toxins
?
?
Bonventre 42
Inflammation
The pathogenesis of ischemic acute renal failure has been attributed to
abnormal regulation of local blood flow following the initial ischemic episode.
Persistent preglomerular vasoconstriction may be a contributing factor; how-
ever, inflammation contributes in an important way to the reduction in local
blood flow to regions of the cortex and the outer medulla with adverse conse-
quences on tubule function and viability.
The Innate Immune Response
Both the innate and adaptive immune responses are important contributors
to the pathobiology of ischemic injury. The innate component is responsible for
the early response to infection or injury and is foreign antigen independent.
Toll-like receptors (TLR) which are important for the detection of exogenous
microbial products [8] and development of antigen-dependent adaptive immu-
nity [9] also recognize host material released during injury [10]. The role of
TLRs was evaluated using an ischemia/reperfusion (I/R) model in TLR2⫺/⫺
and ⫹/⫹ mice [11]. Significantly fewer granulocytes were present in the inter-
stitium of the kidney 1 day post-I/R in the TLR2⫺/⫺ mice and fewer
macrophages were present 1–5 days after I/R. Kidney homogenate cytokines
KC, MCP-1, interleukin-1 (IL-1), and IL-6 were also significantly lower in
the TLR2⫺/⫺ animals as compared to the TLR⫹/⫹ mice. Hence the absence
of TLR2 clearly had an anti-inflammatory effect on the response to I/R. This
anti-inflammatory effect was associated with a functional protection as mea-
sured by serum creatinine at 1 day post-I/R and blood urea nitrogen and tubular
injury score 1 and 5 days post-I/R.
Leukocyte-Endothelial Interactions
With I/R endothelial cells upregulate integrins, selectins, and members of
the immunoglobulin superfamily, including intercellular adhesion molecule-1
(ICAM-1) and vascular cell adhesion molecule. A number of vasoactive com-
pounds may also affect leukocyte-endothelial interactions. Vasodilators, such
as nitric oxide, also can have effects to decrease inflammation. NO inhibits
adhesion of neutrophils to endothelial cells stimulated by TNF-␣, which
would also be protective [12]. It has been known for quite some time now that
there is less flow to the outer medulla in the postischemic kidney [13]. In
addition, as endothelial cells are injured with resulting cell swelling and
increased expression of cell adhesion molecules, leukocytes are also activated.
Enhanced leukocyte-endothelial interactions can result in cell-cell adhesion,
Pathophysiology of Acute Kidney Injury 43
which can physically impede blood flow [14]. Furthermore, these interactions
will additionally activate both leukocytes and endothelial cells and contribute
to the generation of local factors that promote vasoconstriction especially in
the presence of other vasoactive mediators, resulting in compromised local
blood flow and impaired tubule cell metabolism [15]. Due to the anatomical
relationships of vessels and tubules in the outer medulla these leukocyte-
endothelial interactions likely impact the outer medulla to a greater extent than
the cortex.
In an early study to evaluate the significance of endothelial-leukocyte
interactions and inflammation to the pathobiology of ischemic injury we
administered anti-ICAM-1 antibodies and found that when administered prior
to or 2 h following renal I/R they protected the kidney from injury [16]. We fur-
ther confirmed these results in finding that kidneys of ICAM-1 knockout mice
also are protected [17]. We proposed that this upregulation of ICAM-1 was
related to the upregulation of the proinflammatory cytokines TNF-␣ and IL-1
which we measured to be increased by I/R.
Later phases of AKI are characterized by infiltration of macrophages and T
lymphocytes which predominate over neutrophils. ROS are generated during
reperfusion and as a result of the inflammatory response then play a major role
in cell injury. ROS are generated by activated infiltrating leukocytes and by
epithelial cells. ROS are directly toxic to tubular epithelial cells, with ROS gen-
erating systems mimicking the effects of ischemic injury [18].
Tubule Contribution to Inflammatory Injury
Both the S3 segment of the proximal tubule and the medullary thick
ascending limb are located in the outer stripe of the outer medulla. This region
of the kidney is marginally oxygenated under normal conditions and after an
ischemic insult, oxygenation is further compromised because the return in
blood flow is delayed. Both segments of the nephron contribute to the inflam-
matory response in AKI [19]. The tubule epithelial cells are known to generate
proinflammatory and chemotactic cytokines such as TNF-␣, MCP-1, IL-8,
IL-6, IL-1, and TGF-, MCP-1, IL-8, RANTES and ENA-78 [20]. Proximal
tubular epithelia may respond to T lymphocyte activity through activation of
receptors for T cell ligands that are expressed on the proximal tubule cell [21].
When CD40 is ligated in response to interaction with CD154, CD40 ligation
stimulates MCP-1 and IL-8 production, TRAF6 recruitment, and MAPK acti-
vation [21]. CD40 also induces RANTES production by human renal tubular
epithelia, an effect which is amplified by production of IL-4 and IL-13 by Th2
cells, a subpopulation of T cells [22]. B7-1 and B7-2 can be induced on proxi-
mal tubule epithelial cells in vivo and in vitro. After B7-1 and B7-2 induction,
Bonventre 44
proximal tubule epithelial cells costimulate CD28 on T lymphocytes resulting
in cytokine production [23].
Paracrine Effects of Bone Marrow-Derived Stem Cells
There is a potential role for interstitial bone marrow-derived cells in the
production of protective paracrine factors that may facilitate repair of the
epithelium. We and others have found that injection of mesenchymal stem
(stromal) cells (MSCs) is protective against renal injury as assessed by serum
creatinine measured 24 h after ischemia [24]. Other groups have also found
that MSCs protect against ischemic renal injury by a differentiation-indepen-
dent mechanism [25]. In the Adriamycin nephropathy model, injection of the
side population is also protective in the absence of tubular integration [26].
The mechanism of such protection may be through intrarenal paracrine
effects to decrease inflammation or by systemic immune modulation, since
injected cells may be rapidly ingested by immune cells in the spleen, liver and
lungs. It has been increasingly recognized that MSCs can modulate innate
immunity by generating a large number of agents that modify this response
[27].
Role of Resolvins and Protectins as Endogenous Modulators of the
Inflammatory Response in Kidney Injury
Resolvins (Rv) and protectins (PD) are two newly identified families of
naturally occurring n–3 fatty acid docosahexaenoic acid metabolites. These
compounds were identified by Serhan et al. [28], and have been proposed to be
important for the resolution of inflammation. In collaboration with the Serhan
laboratory we recently reported that, in response to bilateral I/R injury, mouse
kidneys produce D series resolvins (RvDs) and PD1 [29]. Administration of
RvDs or PD1 to mice prior to and subsequent to the ischemia resulted in a
reduction in functional and morphological kidney injury. In addition, initiation
of RvD1 administration 10 min after reperfusion also resulted in protection of
the kidney. Both RvDs and PD1 reduced the number of infiltrating leukocytes
and inhibited TLR-mediated activation of macrophages. Interstitial fibrosis
after I/R was reduced in mice treated with RvDs. Thus, production of resolvins
and protectins may represent an endogenous mechanism of the kidney to con-
trol inflammation and may play an important role in resolution of AKI. Whether
tissue injury is due, in part, to failure or inadequacy of this anti-inflammation
response is not known at the present time.
Pathophysiology of Acute Kidney Injury 45
Conclusions
Renal injury is a dynamic process that often exists in the context of multi-
ple organ failure and involves hemodynamic alterations, inflammation and
direct injury to the tubular epithelium followed by a repair process that restores
epithelial differentiation and function. Inflammation plays a considerable role
in the pathophysiology of AKI. Much emphasis has been placed on understand-
ing mechanisms of inflammation that contribute to the pathophysiology. It is
becoming increasingly recognized that there are endogenous mechanisms that
the organism brings to bear to control the inflammation. Understanding how
these anti-inflammatory processes are regulated may provide insight into how
we might intervene to facilitate and enhance them so that we might prevent or
mitigate the devastating consequences of AKI.
Acknowledgments
This work was supported by the National Institutes of Health (grants DK 39773,
DK54741, DK72381).
References
1 Zuk A, Bonventre JV, Brown D, Matlin KS: Polarity, integrin and extracellular matrix dynamics in
the post-ischemic rat kidney. Am J Physiol Cell Physiol 1998;275:C711–C731.
2 Thadhani R, Pascual M, Bonventre JV: Acute renal failure. N Engl J Med 1996;334:1448–1460.
3 Zuk A, Bonventre JV, Matlin KS: Expression of fibronectin splice variants in the postischemic rat
kidney. Am J Physiol Renal Physiol 2001;280:F1037–F1053.
4 Bonventre JV, Zuk A: Ischemic acute renal failure: an inflammatory disease? Kidney Int
2004;66:480–485.
5 Park KM, Byun JY, Kramers C, Kim JI, Huang PL, Bonventre JV: Inducible nitric oxide synthase
is an important contributor to prolonged protective effects of ischemic preconditioning in the
mouse kidney. J Biol Chem 2003;278:27256–27266.
6 Humphreys BD, Duffield JD, Bonventre JV: Renal stem cells in recovery from acute kidney
injury. Minerva Urol Nefrol 2006;58:13–21.
7 Bonventre JV: Dedifferentiation and proliferation of surviving epithelial cells in acute renal fail-
ure. J Am Soc Nephrol 2003;14(suppl 1):S55–61.
8 Aderem A, Ulevitch RJ: Toll-like receptors in the induction of the innate immune response. Nature
2000;406:782–787.
9 Kaisho T, Akira S: Toll-like receptor function and signaling. J Allergy Clin Immunol
2006;117:979–988.
10 Johnson GB, Brunn GJ, Platt JL: Activation of mammalian Toll-like receptors by endogenous ago-
nists. Crit Rev Immunol 2003;23:15–44.
11 Leemans JC, Stokman G, Claessen N, Rouschop KM, Teske GJ, Kirschning CJ, Akira S, van der
Poll T, Weening JJ, Florquin S: Renal-associated TLR2 mediates ischemia/reperfusion injury in
the kidney. J Clin Invest 2005;115:2894–2903.
12 Linas S, Whittenburg D, Repine JE: Nitric oxide prevents neutrophil-mediated acute renal failure.
Am J Physiol 1997;272:F48–F54.
Bonventre 46
13 Vetterlein F, Pethö A, Schmidt G: Distribution of capillary blood flow in rat kidney during postis-
chemic renal failure. Am J Physiol 1986;251:H510–H519.
14 Bonventre JV, Weinberg JM: Recent advances in the pathophysiology of ischemic acute renal fail-
ure. J Am Soc Nephrol 2003;14:2199–2210.
15 Sheridan AM, Bonventre JV: Cell biology and molecular mechanisms of injury in ischemic acute
renal failure. Curr Opin Nephrol Hypertens 2000;9:427–434.
16 Kelly KJ, Williams WW, Colvin RB, Bonventre JV: Antibody to intercellular adhesion molecule-1
protects the kidney against ischemic injury. Proc Natl Acad Sci USA 1994;91:812–816.
17 Kelly KJ, Williams WW, Colvin RB, Meehan SM, Springer TA, Gutierrez-Ramos JC, Bonventre
JV: Intercellular adhesion molecule-1 deficient mice are protected against renal ischemia. J Clin
Invest 1996;97:1056–1063.
18 Malis CD, Weber PC, Leaf A, Bonventre JV: Incorporation of marine lipids into mitochondrial
membranes increases susceptibility to damage by calcium and reactive oxygen species: evidence
for enhanced activation of phospholipase A
2
in mitochondria enriched with n–3 fatty acids. Proc
Natl Acad Sci USA 1990;87:8845–8849.
19 Bonventre JV, Brezis M, Siegel N, Rosen S, Portilla D, Venkatachalam M: Acute renal failure. I.
Relative importance of proximal vs. distal tubular injury. Am J Physiol 1998;275:F623–F632.
20 Bonventre JV, Zuk A: Ischemic acute renal failure: an inflammatory disease? Kidney Int
2004;66:480–485.
21 Li H, Nord EP: CD40 ligation stimulates MCP-1 and IL-8 production, TRAF6 recruitment, and
MAPK activation in proximal tubule cells. Am J Physiol Renal Physiol 2002;282:F1020–F1033.
22 Deckers JG, De Haij S, van der Woude FJ, van der Kooij SW, Daha MR, van Kooten C: IL-4 and
IL-13 augment cytokine- and CD40-induced RANTES production by human renal tubular epithe-
lial cells in vitro. J Am Soc Nephrol 1998;9:1187–1193.
23 Niemann-Masanek U, Mueller A, Yard BA, Waldherr R, van der Woude FJ: B7-1 (CD80) and
B7-2 (CD86) expression in human tubular epithelial cells in vivo and in vitro. Nephron 2002;92:
542–556.
24 Duffield JS, Park KM, Hsiao LL, Kelley VR, Scadden DT, Ichimura T, Bonventre JV: Restoration
of tubular epithelial cells during repair of the postischemic kidney occurs independently of bone
marrow-derived stem cells. J Clin Invest 2005;115:1743–1755.
25 Togel F, Hu Z, Weiss K, Isaac J, Lange C, Westenfelder C: Administered mesenchymal stem cells
protect against ischemic acute renal failure through differentiation-independent mechanisms. Am
J Physiol Renal Physiol 2005;289:F31–F42.
26 Challen GA, Bertoncello I, Deane JA, Ricardo SD, Little MH: Kidney side population reveals
multilineage potential and renal functional capacity but also cellular heterogeneity. J Am Soc
Nephrol 2006;17:1896–1912.
27 Stagg J: Immune regulation by mesenchymal stem cells: two sides to the coin. Tissue Antigens
2007;69:1–9.
28 Serhan C, Hong S, Gronert K, Colgan SP, Devchand PR, Mirick G, Moussignac R: Resolvins: a
family of bioactive products of omega-3 fatty acid transformation circuits initiated by aspirin
treatment that counter proinflammation signals. J Exp Med 2002;196:1025–1037.
29 Duffield JS, Hong S, Vaidya VS, Lu Y, Fredman G, Serhan CN, Bonventre JV: Resolvin D series
and protectin D1 mitigate acute kidney injury. J Immunol 2006;177:5902–5911.
Joseph V. Bonventre, MD, PhD
Renal Division, Brigham and Women’s Hospital, Harvard Institutes of Medicine
4 Blackfan Circle
Boston, MA 02115 (USA)
Tel. ⫹1 617 525 5960, Fax ⫹1 617 525 5965, E-Mail joseph_bonventre@hms.harvard.edu
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 47–63
Sepsis and Multiple Organ Failure
Michael R. Pinsky
Department of Critical Care Medicine, Bioengineering and Anesthesiology,
University of Pittsburgh, Pittsburgh, Pa., USA
Abstract
Background/Aims: Sepsis and multiple organ failure are complex processes that result
from dysregulation of the immune response and its associated hematological, hemodynamic
and metabolic disturbances. Methods: Review of the pathophysiological basis for sepsis and
a review of the literature on its mechanisms of expression. Results: Sepsis is the host
response to an injury, often infectious in origin, that creates both pro- and anti-inflammatory
immune responses. The level and duration of this response roughly correlates with outcome.
Subcellular injury characterized by increased oxidative stress defines the central mitochondr-
ial component of this process. Treatments which minimize the amplification of this response
are usually more effective at reducing tissue injury than are measures aimed at suppressing the
inflammatory response. Conclusions: Sepsis is a complex process whose expression and
treatment are just now being defined. Treatments that minimize the overall host response still
represent the most effective strategies.
Copyright © 2007 S. Karger AG, Basel
Severe sepsis is characterized by an intravascular activation of the host’s
inflammatory pathways releasing potent inflammatory mediators into the circula-
tion [1]. Often referred to as the systemic inflammatory response syndrome [2] it
is a nonspecific inflammatory process to a wide variety of insults that is manifest
by increased circulating levels of numerous active small molecules, each of which
is often capable of inducing a generalized inflammatory response. Furthermore, if
sustained, this system-wide inflammatory process may result in loss of the normal
cardiovascular regulatory adaptations to stress and metabolic demand and nonis-
chemic organ system dysfunction. Importantly, the distribution of organ system
dysfunction and its severity appears to be related to the severity of the circulatory
shock initially induced and its associated inflammatory insult, suggesting that the
tissue hypoperfusion though important in determining tissue function is not the
only process determining organ function in septic shock. Historical perspectives,
Pinsky 48
recent clinical trials of immunomodulating agents and cellular and molecular data
have created an exciting and productive story that only recently led to the first
positive clinical outcome trial of pharmacotherapy for severe sepsis.
Systemic and Cellular Events of Inflammation
Inflammatory processes require immune cell response, immune effector
cell recruitment into the local area and further activation and transcapillary
migration if the inciting infectious stimulus is to be contained and the systemic
inflammatory response is to be sustained. This innate immunity response
involves initial cellular recognition of the stimulant. This initial recognition
process involves a complex interaction of host-derived cofactors, such as
lipopolysaccharide-binding protein, compliment activation and coating of for-
eign biological material (bacteria) and release of procoagulant materials.
Inflammatory mediators bind to cell surface receptors inducing a transmem-
brane signal transduction and intracellular response via activation of several
specific gene promoter proteins. Numerous unrelated exogenous stimulants,
such as endotoxins (lipopolysaccharide or LPS) and exotoxins, and endogenous
stimulants, such as activated compliment (C5a), Hageman’s factor (XIIa), and
products of generalized cell injury can induce immune competent cells (usually
modeled as monocytes) to synthesize TNF-␣ and IL-1. The closest thing we
have to a ‘universal receptor’ is the cell surface molecule CD14. Many foreign
substances, such as endotoxin, require initial binding with lipopolysaccharide-
binding protein before they can bind to CD14. Thus, host-derived factors (usu-
ally proteins) are probably necessary for the host recognition of foreign
materials as foreign and initiate the inflammatory response. However, no such
intermediate binding appears necessary for stimulation of the inflammatory
response via the intrinsic mediators, such as XIIa and C5a.
Several of these activation complexes nonspecifically bind to the host uni-
versal inflammation receptor, CD14, whose transmembrane signal transducing
partners, toll-like receptors, activate an intracellular tyrosine kinase system that
eventually activates the oxidant sensitive proinflammatory promoter, nuclear
factor-kappa B (NF-B) and other proinflammatory promoters. Simultaneously,
expression of novel cell surface receptors, cell adhesion molecules and gene
induction associated with new protein synthesis occur. Following this wave of
intracellular metabolic activity, immune-competent cells and responsive
parenchymal cells make novel protein species necessary to induce a localized
inflammatory response essential to host survival in a host environment.
The release of immune active mediators stimulates primary target cell
(immune cell) responses via similar mechanisms to those described by the primary
Sepsis and Multiple Organ Failure 49
activation sequence, except now more complex and interacting mediator systems
coexist, including activation of the contact system (compliment), fibrinolytic sys-
tem (thrombin and activate protein C) and paired cytokine stimuli (TNF-␣ and IL-
1, or IL-8 and MIP-1). Secondary parenchymal cell response via these soluble
mediators and formed cell interactions then occur via paracrine activation.
Cytokines are presumably present to modulate cellular response and
metabolism on a local or paracrine level. A listing of some of biological actions
of the early phase cytokines (TNF, IL-1, IL-6, and IL-10) are listed in table 1.
With few exceptions, notably pre-pro-IL-1, cytokines have to be synthesized
de novo in response to a specific external stimulus and do not exist in a dormant
state inside the cell. However, once these cytokines are synthesized and
Table 1. Biological effects of various cytokines during acute infections
Cytokine Biological effects
IL-1 Induces COX-2, iNOS, and PG(E
2
) expression
Stimulates release of TNF, IL-6, chemokines and other adhesion molecules
Stimulates myeloid progenitor cells thereby inducing neutrophilia as well
as thrombocytosis
Decreases response to erythropoietin and anemia
TNF Activates macrophages, lymphocytes, neutrophils, eosinophils, fibroblasts,
osteoclasts, chondrocytes, endothelial cells, and nerve cells
Induces the expression of ICAM-1, ELAM, VACAM-1
Activates COX-2, phospholipase A
2
, NOS, PAF, PGE
2
, PGI
2
, NO
Activates complements
Induces release of IL-1, IL-6, IL-8, MCP-1, IL-4, IL-10, IL-1ra, PDGF, IL-2,
endothelin-1, PAI
Downregulates thrombomodulin
Increases the expression of surface adhesion molecules, C3bi receptors,
L-selectin, superoxide production, and phagocytosis
IL-6 Regulates synthesis of ACTH in pituitary gland and neuronal growth factor
Activates Janus kinase/signal transducer and mitogen-activated protein
kinase cascades through binding of gp130 receptor molecules
Induces thrombosis by enhancing the release of von Willebrand factor
fragments and inhibiting its cleaving protease
Induces gut barrier dysfunction
IL-10 Downregulates TNF and other proinflammatory cytokines
Causes T cell anergy, suppresses T cell and Th1 proliferation
Induced T and B cell apoptosis
Causes defects in antigen presentation and induces macrophage ‘paralysis’
COX-2 ⫽ Cyclooxygenase 2; iNOS ⫽ inducible nitric oxide synthase; IL-1ra ⫽ IL-1 receptor antagonist.
Pinsky 50
secreted they rapidly gain access to the bloodstream. If the initiating stimulus is
great or the host response heightened, the systemic delivery of cytokines can
induce a systemic immune cell response. It is still not clear to what extent this
systemic response is adaptive or maladaptive to the host. On the one hand, fever
and the induction and release of acute phase proteins by the liver improve sur-
vival in bacterial infections, and malaise may be a very adaptive symptom in
limiting the host activity and excess consumption of limited energy resources.
However, the generalized inflammatory response may not confer a survival
advantage for the host. Clearly, on a local level, a violent inflammatory
response is highly effective at containing, killing and removing foreign biologi-
cal material (alive or dead) and the lack of a vigorous local inflammatory
response may impair survival. This was best exemplified by the recent docu-
mentation that patients with a mild TNF-␣ response, as characterized by a spe-
cific genetic genotype of a mild TNF-␣ responder, had a greater likelihood of
dying of meningococcal meningitis than patients who had the vigorous TNF-␣
response genotype.
Still, persistent activation, rather than massive pulse activation, appears to
be detrimental for survival. Pinsky et al. [3] measured the circulating levels
of TNF-␣, IL-1, and IL-2 and the immunomodulating cytokines IL-6 and
interferon-␥. Although levels of IL-1, IL-2 and interferon-␥ did not correlate
with any aspect of the generalized inflammatory response, severity of illness or
mortality, sustained elevations of the proinflammatory cytokine TNF-␣ and the
immunomodulating cytokine IL-6 did. Thus, rather than the maximal serum
levels of any specific cytokines, those patients who subsequently develop mul-
tiple organ dysfunction and die display a persistent elevation of TNF-␣ and
IL-6 in their blood [3–5].
Importantly, both proinflammatory cytokines, such as TNF-␣, IL-1, IL-6
and IL-8, and anti-inflammatory species, such as IL-1 receptor antagonist,
IL-10 and the soluble TNF-␣ receptors I and II (sTNFrI and sTNFrII, respec-
tively), coexist in the circulation in patients with established sepsis and presum-
ably within the tissues [6–8]. Thus, sepsis may be more accurately described as
a dysregulation of the innate immunity rather than merely the overexpression of
either proinflammatory substances. This combined pro- and anti-inflammatory
mediator interaction may explain the observed failures of all the major anti-
inflammatory drug trials in the treatment of septic shock [9]. Accordingly, we
proposed ‘malignant intravascular inflammation’ to describe the systemic
process of severe sepsis [10]. This paradoxical expression in the blood of proin-
flammatory mediators and anti-inflammatory species creates an internal milieu
that in sustained sepsis induces impaired host immunity. Experimentally, this
altered immune response state resembles ‘endotoxin tolerance’. Cavaillon [11]
termed this blunted immune response ‘inflammatory-stimuli-induced anergy’,
Sepsis and Multiple Organ Failure 51
because it is universally seen in severe stress states and can be induced by prior
exposure to low levels of one of many proinflammatory stimuli, not only endo-
toxin. Hence, its original name ‘endotoxin tolerance’ is misleading, but will be
used here to describe this nonspecific response. Inflammatory-stimuli-induced
anergy can be induced by anti-inflammatory cytokines, such as TGF-, IL-10,
IL-4 and somewhat by IL-1, but not TNF-␣, IL-6 or IL-8. Furthermore, it is
associated with altered intracellular metabolism of the important regulatory
protein NF-B. Leukocyte dysfunction also occurs in sepsis and in endotoxin-
tolerant states and may be a central determinant of outcome [12–15].
Presumably, inflammatory stimuli-induced anergy minimizes the inflam-
matory response, preventing a chain reaction system-wide activation of the
inflammatory processes. However, it also limits the subsequent ability to mount
an appropriate inflammatory defense to infection. Since endotoxin tolerance
usually requires a few hours to develop in isolated cells and 8–10 h in intact ani-
mals, its expression parallels that of clinical sepsis. Thus, we have previously
hypothesized that endotoxin tolerance carries most if not all of the intracellular
qualities of fully developed severe sepsis. However, the intracellular mecha-
nisms by which inflammatory stimuli-induced anergy are poorly understood,
but reflect in large part intracellular change regarding the activation of the
inflammatory pathways.
Intracellular Inflammatory Response: Pro- or
Anti-Inflammatory Response
NF-B is an oxidant-inducible promoter protein of the proinflammatory
response of immune effector cells [16]. Although other intracellular proinflam-
matory promoters are also present, none have the breadth of gene activation or
complex feedback control mechanisms described for NF-B. NF-B-inducible
proteins include TNF-␣, IL-1 and IL-8 plus the proinflammatory enzyme-
inducible nitric oxide synthase and cyclooxygenase 2 [11]. When viewed from
a purely regulatory perspective NF-B is an excellent target for modulating the
cellular inflammatory response. It is not surprising, therefore, that several intra-
cellular mechanisms exist that modulate NF-B activity.
What is endotoxin tolerance and why can it be used to study the molecular
mechanisms of sepsis? Exposure to small amounts of endotoxin induces an
endotoxin-tolerant state in both cell culture in 4 h and animal models in about
8 h [17]. In this state, subsequent exposure to a previously lethal dose of endo-
toxin does not induce the fatal proinflammatory state. The endotoxin-tolerant
state lasts in decreasing strength for between 24 and 36 h, depending on the
species and the initial dose of endotoxin. Interestingly, following the induction
Pinsky 52
of an endotoxin-tolerant state the initial steps of proinflammatory signal trans-
duction up to cleavage of IB can still occur. However, the liberated NF-B
appears to be dysfunctional [16]. The reasons for this dysfunction are multiple
and not fully defined. However, some specific interactions have been defined
that speak to the importance of the pro- and anti-inflammatory interactions.
Importantly, there is much similarity between endotoxin tolerance and human
sepsis. First, sepsis rarely starts with a massive exposure to an overwhelming
noxious stimulus. Usually, infection or inflammation build over hours. Thus,
the host is exposed to low levels of proinflammatory species with a time course
similar to classic endotoxin tolerance. Second, just as endotoxin tolerance min-
imizes the subsequent proinflammatory response to an additional proinflamma-
tory stimulation, so too sepsis is characterized by a blunted immune effector
cell response. By what mechanisms do proinflammatory stimuli induce an
apparently anti-inflammatory response? The answer appears to relate to the
complex mechanisms by which NF-B is activated and binds to the proinflam-
matory promoter regions.
Intracellular NF-B Activation
The activation of NF-B is central to immune effector cell activation.
Endotoxin can induce the initial steps of signal transduction up to NF-B, but
NF-B activation is required for much of the subsequent intracellular signaling.
The NF-B family is composed of various members, p50 (NF-B1), p52 (NF-
B2), p65 (RelA), RelB and c-Rel, which can form homo- and heterodimers
[18]. The p65 subunit has the most variability, with a common variation being
the RelA subunit substitution for p65. The phosphorylation of the IB-␣ sub-
unit of the NF-B complex following intracellular oxidative stress frees the
dimer to translocate into the nucleus [19]. LPS induces IB-␣ phosphorylation
through activation of the IK kinase [20]. The phosphorylated IB-␣ is rapidly
degraded by proteosomes. Processes that inhibit IB-␣ phosphorylation, such
as 4-hydroxynoneal, prevent NF-B activation [21]. The p65 moiety has a
DNA-binding domain that allows it to bind to numerous specific DNA sites
throughout the genome, regulating gene transcription for most, if not all, of the
proinflammatory species, including TNF-␣ [22], IL-1, inducible nitric oxide
synthase [23], lipoxygenase, and cyclooxygenase. IB-␣ is a heat shock protein
(HSP) and its increased synthesis also downregulates NF-B activation by dis-
sociating the p65-p50 heterodimer from its responsive elements on the genome
and keeping it in an inactive form in the cytoplasm [24].
NF-B DNA-binding activity can be downregulated by processes indepen-
dent of IB-␣. The NF-B dimer can exist in one of two forms: a p65-p50
dimer and a p50-p50 homodimer. The p65 subunit has DNA-binding activity,
whereas the p50 subunit does not. Thus, activation of p65-p50 dimers results in
Sepsis and Multiple Organ Failure 53
markedly increased transcription of mRNA following binding, whereas p50-
p50 activation only minimally increases transcription rates. NF-B dysfunction
reflects both an excess p50 homodimer production, which has impaired tran-
scription activity [16], and excess synthesis of IB-␣ [24]. The balance of NF-
B species is very sensitive to transcription rates, with ratios of NF-B p50-p65
heterodimer to p50 homodimer of 1.8 ⫾ 0.6 conferring activation of the
inflammatory pathways, and a ratio of 0.8 ⫾ 0.1 conferring lack of stimulation
in response to LPS (i.e. endotoxin tolerance). The p50 subunit may inhibit TNF
mRNA synthesis. Using knockout mice for the p50 subunit of NF-B,
Bohuslav et al. [25] demonstrated that endotoxin tolerance was not achieved
from p50⫺/⫺ mice, long-term TNF mRNA synthesis was not blocked in
p50⫺/⫺ macrophages (in contrast to wild-type macrophages), and ectopic
overexpression of p50 reduced transcriptional activation of the TNF promoter.
Finally, analysis of the four B sites from the murine TNF promoter demon-
strated that binding of p50 homodimers to a positively acting B3 element was
associated with endotoxin tolerance. These different B-binding activities may
result from specific differences in their sequences. B2 and B4 sites do not
meet the requirements for p50 binding due to the lack of a GGG motif at the 5⬘-
end and the B1 site lacks a 3⬘-end CCC (present on the B3 site), favorable for
binding p50 homodimers. However, these p65-p50/p50-p50-related processes
can only explain the downregulation of the overall inflammatory process. They
cannot explain why both pro- and anti-inflammatory activation is sustained in
severe sepsis, or why immune suppression, a common characteristic of sepsis,
coexists with a heightened inflammatory state.
Control of NF-B activation can occur at many levels. Excess intracellular
IB excess can pull active p50-p65 dimers off their promoter sites or prevent
their release altogether. IB synthesis is stimulated by NF-B binding to the
genome. Thus, in a negative feedback loop process, NF-B activation stimu-
lates its own inhibition. NF-B dysfunction due to endotoxin tolerance also
reflects excess synthesis of the inhibitor of NF-B, IB-␣ [24], presumably due
to the absence of induction of IB kinase. When endotoxin-tolerant cells are
challenged with a second dose of LPS, cytosolic levels of IB are not reduced,
as they are with the initial challenge, and IB remains in cytoplasm where it
sequesters free NF-B dimers [26]. Furthermore, the IB-␣ promoter can be
upregulated by NF-B, thus providing a negative feedback loop for further NF-
B activation [27]. As sustained proinflammatory activation would be highly
detrimental to an organism, having intrinsic mechanisms to downregulate this
proinflammatory response seems prudent.
As another form of downregulatory control, NF-B can also alter its own
intrinsic gene induction ability. Specifically, NF-B dimers exist primarily in one
of two forms depending on their subunit composition. The specific molecular
Pinsky 54
species usually referred to by the moniker as NF-B is the p50-p65 het-
erodimer. As the name p50-p65 implies, it is comprised of the p65 subunit with
its active DNA consensus domain-binding site and a smaller p50 subunit devoid
of the active binding site. Importantly, the p65 subunit with its DNA consensus
domain-binding site allows gene activation once bound to such promoter
regions on chromosomes. The p50 monomer has no such activity [16].
Although both p50-p50 and p65-p65 homodimers can theoretically exist, very
little p65-p65 has been detected. The primary homodimer is the p50-p50
species. This p50-p50 homodimer can account for more than half the total
amount of intracellular NF-B with immune suppressed cells. Thus, a second
adaptive mechanism involves the balance of NF-B p50-p65 to p50-p50
species. The ratio of p65-p50 to p50-p50 determines NF-B-induced gene tran-
scription rates. Ratios of NF-B p50-p65 heterodimer to p50-p50 homodimer
of 1.8 ⫾ 0.6 or greater are associated with NF-B activation inducing mRNA
synthesis of these genes, while a ratio of 0.8 ⫾ 0.1 or less confers a lack of
mRNA transcription following cleavage of IB [16].
Downregulation of NF-B-related intercellular processes is an important
aspect of the overall intracellular inflammatory response. Potentially, if anti-
inflammatory pathways were activated by the same stimuli that active proinflam-
matory pathways, then an intrinsic mechanism would exist to autoregulate the
inflammatory response on a cellular level. This system may be induced by much
of the same stimuli and act through parallel intracellular processes. The cell does
have several intrinsic anti-inflammatory processes, including antiproteases,
melanoproteins and free radical scavengers. However, the HSP system is by far
the most prevalent, in terms of the mass of protein and scope of its oversight.
HSP and the Stress Response
The HSP system is the oldest phylogenetic cellular defense mechanism
identified. It has widespread and overarching basic roles in cellular defense
against numerous stresses, such as fever, trauma, and inflammation [28].
Intracellular chaperone proteins of the heat shock family appear to be pivotal in
the regulation of the cellular response to inflammatory signals and external
injury. An initial step in the intracellular activation of the inflammatory path-
way is the production of reactive oxygen species (ROS). Mitochondria are not
only very sensitive to oxidative stress from ROS but are a primary intracellular
site of free radical production. Thus, measuring mitochondrial membrane
potential (⌿⌬) allows one to monitor the degree of intracellular oxidative stress
over time [29]. Detection of the mitochondrial permeability transition event
also provides an early indication of the initiation of cellular apoptosis. This
process is typically defined as a collapse in the mitochondrial membrane elec-
trochemical gradient, as measured by the change in the ⌿⌬. Loss of mitochondrial
Sepsis and Multiple Organ Failure 55
⌿⌬ can be detected by a fluorescent cationic dye, 5,5⬘,6,6⬘-tetrachloro-
1,1⬘,3,3⬘-tetraethyl-benzamidazolocarbocyanin iodide, commonly known as
JC-1. Using this sensitive bioassay of mitochondrial membrane ⌿⌬ [30] Polla
et al. [31] showed that HSP70 prevents this mitochondrial oxidative stress
injury and blunts the inflammatory response. Numerous other studies have
shown that HSP are a basic cellular defense mechanism against numerous
stresses, such as fever, trauma, and inflammation [28, 31]. They also minimize
nitric oxide, ROS and streptozotocin cytotoxicity [32, 33]. HSP70 is also an
inducible protective agent in myocardium against ischemia, reperfusion injury
and nitric oxide toxicity [34], but requires some initial trigger for its activation.
Such induced HSP synthesis blunts the inflammatory response to endotoxin
and TNF-␣ in vitro. Thus, the intracellular pathway mobilized upon heat shock
may be important in modulating the intracellular inflammatory signal acting at
the level of NF-B and appears to have many of the immune modulating char-
acteristics of endotoxin tolerance. Since heat shock is a primordial cellular
defense mechanism, upregulates in a matter of minutes and rapidly confers an
immune-depressed state, we consider its actions to reflect a relatively pure
antioxidant, anti-inflammatory response. Although the time courses for produc-
tion of endotoxin tolerance and heat shock are different, similar gene activation
and inhibition occur with both processes.
The heat shock factors (HSFs), particularly HSF-1, are primarily responsi-
ble for inducing the transcription of HSP genes and can be considered the NF-
B equivalent for HSP synthesis. HSF exist preformed in the cytosol bound to
HSP70. The presence of denatured protein (resulting from heat or ROS-induced
protein damage) strips the HSP70 proteins from HSF as HSP70 binds to the
damaged proteins. Trimerized HSF migrates into the nucleus to bind to numer-
ous regions within the genome. HSF-1 DNA-binding sites possess multiple
repeats of a 5⬘-G-A-A-3⬘ triplet, often in an inverted orientation and separated
by at least two nucleotides (e.g. 5⬘-G-A-A- N-N- T-T-C-3⬘). According to Amin
et al. [35], ‘a functional heat shock regulatory element includes a minimum of
three GAA segments although these segments do not have to be consecutive’.
Importantly in natural heat shock response elements (HSEs) single nucleotide
substitutions of the GAA core triplet are found [35]. Although HSF-1 and NF-
B are not structurally related, the DNA recognition elements for both factors
can be somewhat similar. For example, the core NF-B recognition element is
comprised of the sequence 5⬘-G-G-G-R-N-W-T-T-C-C-3⬘ (where R ⫽ purine,
N ⫽ any nucleotide and W ⫽ A or T) [36]. Interestingly, one of the NF-B-
responsive elements within the TNF-␣ gene contains the sequence 5⬘-G-G-G-A-
A-A-G-C-C-C-3⬘ which contains only one mismatch with a partial HSE core
element [36]. Likewise, the interferon-␥ gene promoter contains an NF-B
response element with the sequence 5⬘-G-A-A-T-T-T-T-C-C-3⬘, which contains a
Pinsky 56
perfectly spaced and matched pair of HSE GAA triplets (one in an inverted ori-
entation) [36]. Even in natural HSEs, a third GAA element does not have to be
perfectly matched in sequence or spacing to constitute a functional element.
Thus, the possibility exists that some NF-B sites may be recognized by HSF-1.
Putative recognition of specific NF-B response elements by HSF-1 does
not necessarily imply that transcription would ensue. Transcriptional activation
by DNA-bound HSF-1 requires site-specific phosphorylation, which appar-
ently is brought about by the precise coordination of multiple stress-activated
kinases. Indeed, HSF-1 has been implicated in transcriptional repression of the
IL-1 gene promoter [29]. The mechanism for this repression appears to
require the binding of HSF-1 to an HSE-like sequence within the IL-1 pro-
moter [29]. One of the phenotypes observed in HSF-1 knockout mice also sup-
ports a role for HSF-1 in transcriptional repression of proinflammatory genes.
Following an endotoxic challenge (i.e. Echerichia coli LPS), HSF-1 knockout
mice exhibited a potentiation of proinflammatory TNF-␣ production [37].
Since HSP induction is severely limited in the HSF-1 knockout mice [37], the
effect of HSF-1 on TNF-␣ induction does not appear to require HSF-1-induced
HSP, such as HSP70. Thus, while HSP70 stabilization of IB-␣ has been postu-
lated to participate in downregulation of NF-B [38], other levels of control
may operate under conditions of endotoxic shock. The IB-␣ promoter region
has an NF-B recognition site, and NF-B activation increases IB-␣ synthesis
in a negative-feedback loop fashion [39, 40]. Potentially, HSF-1 may inhibit
NF-B-induced transcription promotion via tethering of NF-B to its respon-
sive element in a fashion similar to glucocorticoids [41]. The anti-inflammatory
action of glucocorticoid hormones is mediated by its binding its cognate recep-
tor, the glucocorticoid receptor. Glucocorticoid repression of NF-B-directed
transcription is brought about, in part, by the association of the glucocorticoid
receptor with DNA-bound NF-B. Glucocorticoids inhibit the inflammatory
process by binding to a glucocorticoid receptor. The glucocorticoid receptor
then can attach to the DNA-bound NF-B in a site remote from the recognition
site preventing transcription. Potentially, HSF-1 inhibition of NF-B activation
involves a tethering mechanism. If so, then competitive binding studies may not
demonstrate NF-B displacement by HSF-1.
HSP confer a survival advantage to their host. Thermal pretreatment is
associated with attenuated lung damage in a rat model of acute lung injury
induced by intratracheal instillation of phospholipase A
2
[42]. Thermal pretreat-
ment reduces mortality rate and sepsis-induced acute lung injury produced by
cecal ligation and perforation [43]. The subsequent increased expression of a
broad variety of HSP confers a nonspecific protection from not only subsequent
oxidative stress but also minimizes the cellular response to proinflammatory
stimuli. Survival in cold-blooded animals given an infectious inoculum is
Sepsis and Multiple Organ Failure 57
linearly related to body temperature. Finally, if specific HSP are depleted, then
multiple intracellular signaling processes can be affected.
The Role of Mitochondria in Intracellular Inflammation
An initial step in the intracellular activation of the inflammatory pathway
between mediator binding to the cell surface and inflammatory gene activation
is the production of ROS and an associated oxidative stress on the mitochondria
[44]. Mitochondria operate by generating a chemiosmotic gradient via the
Krebs cycle inside their inner membrane necessary to drive ATP formation
across this membrane. Essentially they are intracellular batteries whose charge
or polarization level defines their ability to create ATP. Loss of internal mem-
brane polarization induces cytochrome c release from the mitochondria into the
cytosol. Importantly, cytochrome c activates the intrinsic caspase system to ini-
tiate apoptosis or programmed cell death. Thus, preventing mitochondrial depo-
larization would be an important function for HSP if their role were to aid in
cell survival. In vivo measures of mitochondrial polarization are possible using
redox state-sensitive substances that can be introduced into cells in vitro.
HSP70 prevents this mitochondrial oxidative stress and blunts the inflam-
matory response. HSP70 also minimizes nitric oxide, oxygen-free radical and
streptozotocin cytotoxicity [32]. Consistent with the pluripotential effects of
the heat shock response, HSP70 is also an inducible protective agent in
myocardium against ischemia, reperfusion injury and nitric oxide toxicity [34].
Nitric oxide-induced HSP synthesis blunts the inflammatory response to endo-
toxin and TNF-␣ in vitro. Evidence that HSP70 may be active in human sepsis
comes from the observation that higher HSP70 expression is seen in peripheral
mononuclear cells in septic patients. Although not conclusive, these data
strongly suggest that HSP, and HSP70 in particular, may be important in modu-
lating the intracellular inflammatory signal acting at the level of NK-B.
Importantly, Wong et al. [45] identified a potential heat shock-responsive ele-
ment in the IB-␣ promoter that can be activated by HSF after heat shock.
Moreover, the heat shock response can also modulate NF-B inactivation by
this increased IB-␣ expression. Thus, the primary inhibitor of NF-B activa-
tion, IB-␣, is itself an HSP. The interaction between the HSP system and the
proinflammatory pathway, however, has yet to be defined.
Since mitochondrial dysfunction causes leakage of cytochrome c that acti-
vates the caspase system leading to programmed cell death, one of the out-
comes from such activation is a shut-down of all extrinsic cellular activities,
like active transport of ions or, in the case of muscle, contraction. Thus, an early
expression of mitochondrial dysfunction is a nonspecific organ dysfunction,
which if sustained, leads to organ failure. Finally, mitochondria are essential
free radical generators. Their primary role is to create free radicals that drive the
Pinsky 58
electron transport chain to create the chemiosmotic gradient needs to synthe-
size ATP for ADP and inorganic phosphate. Thus, with disruption of mitochon-
drial membrane integrity, intracellular free radical release often occurs
increasing the intracellular oxidative stress. Although the linkage between mito-
chondrial dysfunction and apoptosis is strong and well described, the linkage
between mitochondrial disruption and intracellular oxidative stress is not. It is
not clear if free radical poisoning does occur in sepsis. It does occur with
ischemic necrosis and induces a profound inflammatory response.
Sepsis Is Characterized by Excessive Pro- and
Anti-Inflammatory Activity
Both pro- and anti-inflammatory processes are ongoing in the cell and both
pro- and anti-inflammatory mediators are present in the bloodstream. Thus, the
circulating and fixed immune effector cells also receive mixed messages.
However, the phenotypic response that they make is more difficult to predict.
The normal cellular inflammatory response is essential to survival. It localizes
and eliminates foreign material, including microorganisms. Similarly, some
degree of systemic inflammatory response is useful. Fever reduces microorgan-
ism growth, malaise causes the host to rest, and acute phase protein secretion
minimizes oxidative injury. However, inflammation is also destructive. Local
abscess formation and multiple system organ failure are its very real byproducts.
However, sepsis also carries a strong anti-inflammatory response. The exact bal-
ance or interaction among these two processes with their multiple layers of feed-
back, activation and control is difficult, if not impossible, to assess clinically.
Prior studies documented that sepsis and all acute severe processes result
in the expression in the systemic circulation of cytokines. However, it is diffi-
cult to assess the degree of inflammatory stimulation by measuring serum
cytokine levels. Serum cytokine levels can change within minutes and may be
very different in adjacent tissue compartments [46]. Although TNF and IL-6
serum levels are excellent markers of disease severity and a good positive pre-
dictor of the subsequent development of remote organ system dysfunction,
measuring blood levels of cytokines does not aid in defining the pro- to anti-
inflammatory balance in predicting response to therapy.
Attention has subsequently shifted to examination of the functional status
of circulating immune effector cells. Since polymorphonuclear leukocyte
(PMN) activation and localization represent the initial cellular host defense
against infection, their tight control is essential to prevent widespread nonspe-
cific injury. Subsequently, monocytes localize at the site of inflammation. Their
activity appears to become the predominant process in both host defense and
Sepsis and Multiple Organ Failure 59
repair, especially during the second and third day onward in the course of acute
illness. Thus, inhibition of monocyte immune responsiveness is a powerful
mechanism to downregulate the inflammatory response. Anergy is a cardinal
characteristic of severe illness and reflects macrophage inhibition of antigen
processing. Importantly, antigen processing reflects a primary aspect of this
cellular response. In this regard, the cell surface receptor family, HLA-DR, is
responsible for antigen presentation to antigen processing cells. Immature
monocytes cannot process antigen and have lower cell surface HLA-DR levels.
Docke et al. [47] demonstrated that monocytes require HLA-DR levels ⬎20%
for normal cell-meditated immunity. Lower levels of HLA-DR expression con-
firm immune suppression. Consistent with the overall theme of increased anti-
inflammatory responses in severe sepsis, these workers and others have found a
profound decrease in HLA-DR expression on circulating monocytes from
patients with sepsis.
Rosenbloom et al. [48] examined the relation between circulating cytokine
levels and the expression of the strong 
2
-integrin surface cell adhesion mole-
cules on circulating immune effector cells. The activation state of circulating
immunocytes can be indirectly assessed by measuring the intensity of display
on their 
2
-integrin surface cell adhesion molecules. They showed that all cir-
culating immune effector cells, including PMNs, lymphocytes and monocytes
are activated in critical illness. Furthermore, the level of circulating immune
effector cell activation is proportional to mean circulating IL-6 levels [48].
Importantly, the degree of organ dysfunction but not the level of shock severity
correlated with CD11b expression. CD11b is part of the cell integrin system
essential for cell adhesion in the inflammatory response. Since the level of acti-
vation of circulating PMN, as measured by total PMN count and its display of
immature forms, is used clinically as an indicator of the host response to sys-
temic inflammation, assessment of PMN responsiveness should also be a good
measure of the pro- and anti-inflammatory balance in severe sepsis. They rea-
soned that immune effector cell responsiveness to a known dose of a pro-
inflammatory stimulant would define the host’s immune responsiveness. In
essence, they used the circulating immune effector cells as a bioassay. Prior
studies have shown that PMNs can be both overactive [49] and dysfunctional
[12], and that their CD11b display can be either decreased [50] or increased
[51] in critically ill patients.
Rosenbloom et al. [52] demonstrated that the de novo display of CD11b on
circulating PMNs and its subsequent change in expression of both total CD11b
and its avid form, CBRM1/5 epitope, in response to in vitro stimulation to TNF-␣
characterized the in vivo state of PMN activation and responsiveness. Circu-
lating PMNs of septic humans have a similar phenotype characterized by high
CD11b and low L-selectin expression. This is the phenotype of acute activation
Pinsky 60
of the inflammatory response. Thus, severe sepsis is associated with an inc-
reased de novo activation of circulating immune effector cells. Paradoxically,
however, those same PMNs with a sustained inflammatory state are also
impaired in their ability to upregulate CD11b further or to change surface
CD11b to the avid state [53] in response to an ex vivo challenge by exposure to
biologically significant levels of TNF-␣. Furthermore, circulating PMN for
subjects with severe sepsis have impaired phagocytosis, reduced oxygen burst
capacity and diminished in vitro adhesiveness [54]. This desensitization to
exogenous TNF-␣ was not due to a loss of TNF receptors because the cell sur-
face TNF-␣ receptor density was not reduced in the cells of these septic
patients. Importantly, the hyporesponsiveness observed was extended to all cir-
culating PMN and monocytes in these critically ill patients.
Finally, when peripheral blood monocytes from septic subjects were ana-
lyzed for the NF-B activity and in vitro responsiveness to LPS, Adib-Conquy
et al. [17] observed an NF-B pattern of response similar to that seen with
endotoxin tolerance. The cause of the reduced nuclear translocation of NF-B
was not due to excess IB but rather to an increase in the proportion of the inac-
tive p50-p50 species relative to the active p65-p50 species. Interestingly, sur-
vivors had higher levels of NF-B than did nonsurvivors, suggesting that
although downregulation of inflammation is a normal aspect of sepsis, exces-
sive inhibition of the process is associated with a poor prognosis. Clearly, this is
an area of active investigation and the complex interactions amongst pro- and
anti-inflammatory processes are just now being teased out.
Resuscitation and Reversal of Organ Injury
Given the above complex processes one may reasonably ask: how does any
individual survive a severe inflammatory insult like massive bacteremia, pancre-
atitis or massive trauma? Clearly many patients do survive. Several aspects of
the above processes lend themselves to survival. First, as described above, the
inflammatory response is immediately downregulated by the anti-inflammatory
heat shock response limiting immune effector cell activation and mitochondrial
injury. Thus, if the inciting stimulus is removed, the system tends to right itself.
Second, inflammation does not exist in isolation, but in concert with a deranged
circulatory state, wherein vasomotor tone is reduced and the responsiveness of
the cardiovascular system to increased sympathetic stimulation is diminished.
Within this context, Rivers et al. [55] aggressively resuscitated patients present-
ing in septic shock to the emergency department of a large inner city hospital.
They compared resuscitation to either a normal blood pressure and sensorium or
further increasing O
2
delivery until indirect measures of tissue ischemia were
Sepsis and Multiple Organ Failure 61
resolved. They found that not only did the more aggressively resuscitated patients
have a better outcome in terms of reduced mortality and shorter length of stay in
the hospital, but they also had longer circulating levels of TNF and IL-6. These
preliminary data suggest that aggressive goal-directed resuscitation of patients
early in their septic event may reduce the overall inflammatory response.
Acknowledgment
This work was supported in part by the NIH grants HL67181, HL07820 and HL073198.
References
1 Schlag G, Redl H: Mediators of injury and inflammation. World J Surg 1996;20:406–410.
2 Bone RC: Toward a theory regarding the pathogenesis of the systemic inflammatory response syn-
drome: what we do and do not know about cytokine regulation. Crit Care Med 1996;24:163–172.
3 Pinsky MR, Vincent JL, Deviere J, Alegre M, Kahn R J, Dupont E: Serum cytokine levels in
human septic shock. Relation to multiple-system organ failure and mortality. Chest 1993;103:
565–575.
4 Thijs LG, Hack CE: Time course of cytokine levels in sepsis. Intensive Care Med 1995;21(suppl
2):S258–S263.
5 Blackwell TS, Christman JW. Sepsis and cytokines: current status. Br J Anaesth 1996;77:110–117.
6 Goldie AS, Fearon KC, Ross JA, Barclay GR, Jackson RE, Grant IS, Ramsay G, Blyth AS, Howie
JC: Natural cytokine antagonists and endogenous anti-endotoxin core antibodies in sepsis syn-
drome. The Sepsis Intervention Group. JAMA 1995;274:172–217.
7 Vanderpoll T, Malefyt RD, Coyle SM, Lowry SF: Anti-inflammatory cytokine responses during
clinical sepsis and experimental endotoxemia – sequential measurements of plasma soluble inter-
leukin (IL)-1 receptor type II, IL-10, and IL-13. J Infect Dis 1997;175:118–122.
8 Ertel W, Scholl FA, Trentz O: The role of anti-inflammatory mediators for the control of systemic
inflammation following severe injury; in Faist E, Baue AE, Schildberg FW (eds): The Immune
Consequences of Trauma, Shock, and Sepsis – Mechanisms and Therapeutic Approaches.
Lengerich, Pabst Science Publishers, 1998, pp 453–470.
9 Abraham E: Why immunomodulatory therapies have not worked in sepsis. Intensive Care Med
1999;25:556–566.
10 Pinsky MR: Clinical studies on cytokines in sepsis: role of serum cytokines in the development of
multiple-systems organ failure. Nephrol Dial Transplant 1994;9(suppl 4):94–98.
11 Cavaillon JM: The nonspecific nature of endotoxin tolerance. Trends Microbiol 1995;3:320–324.
12 McCall CE, Grosso-Wilmoth LM, LaRue K, Guzman RN, Cousart SL: Tolerance to endotoxin-
induced expression of the interleukin-1 beta gene in blood neutrophils of humans with the sepsis
syndrome. J Clin Invest 1993;91:853–861.
13 Wenisch C, Parschalk P, Hasenhundl M, et al: Polymorphonuclear leukocyte dysregulation in patients
with gram-negative septicemia assessed by flow cytometry. Eur J Clin Invest 1995;25:418–424.
14 Vespasiano MC, Lewandoski JR, Zimmerman JJ: Longitudinal analysis of neutrophil superoxide
anion generation in patients with septic shock. Crit Care Med 1993;21:666–672.
15 Sorrell TC, Sztelma K, May GL: Circulating polymorphonuclear leukocytes from patients with
gram-negative bacteremia are not primed for enhanced production of leukotriene B4 or 5-hydrox-
yeicosatetraenoic acid. J Infect Dis 1994;169:1140–1151.
16 Ziegler-Heitbrock HWL, Wedel A, Schraut W, Strobel M, Wendelgass P, Sterndorf T, Bauerle PA,
Haas JG, Riethmuller G: Tolerance to lipopolysaccharide involves mobilization of nuclear factor
B with predominance of p50 homodimers. J Biol Chem 1994;269:17001–17004.
Pinsky 62
17 Adib-Conquy M, Adrie C, Moine P, Ashnoune K, Fitting C, Pinsky MR, Dhainaut J-F, Cavaillon J-
M: NF-B expression in mononuclear cells of septic patients resembles that observed in LPS tol-
erance. Am J Respir Crit Care Med 2000;162:1877–1883.
18 Schleiffenbaum B, Fehr J: The tumor necrosis factor receptor and human neutrophil function.
Deactivation and cross-deactivation of tumor necrosis factor-induced neutrophil responses by
receptor down-regulation. J Clin Invest 1990;86:184–195.
19 Kamata H, Hirata H: Redox regulation of cell signaling. Cell Signal 1999;11:1–14.
20 Sanlioglu S, Williams CM, Samavati L, et al: Lipopolysaccharide induces Rac1-dependent reac-
tive oxygen species formation and coordinates tumor necrosis factor – a secretion through IKK
regulation of NF-B. J Biol Chem 2001;276:30188–30198.
21 Page S, Fischer C, Baumgartner B, et al: 4-Hydroxynoneal prevents NF-B activation and tumor
necrosis factor expression by inhibiting IB phosphorylation and subsequent proteolysis. J Biol
Chem 1999;274:11611–11618.
22 Young S-H, Ye J, Frazer DG, et al: Molecular mechanism of tumor necrosis factor-␣ production in
1?3--glucan (zymosan)-activated macrophages. J Biol Chem 2001;276:20781–20787.
23 Taylor BS, de Vers ME, Ganster RW, et al: Multiple NF-B enhancer elements regulate cytokine
induction of human inducible nitric oxide synthase gene. J Biol Chem 1998;273:15148–15156.
24 Larue KEA, McCall CE: A liable transcriptional repressor modulates endotoxin tolerance. J Exp
Med 1994;180:2269–2275
25 Bohuslav J, Kravchenko VV, Parry GCN, Erlich JH, Gerondakis S, Mackman N, Ulevitch RJ:
Regulation of an essential innate immune response by the p50 subunit of NF-B. J Clin Invest
1998;102:1645–1652.
26 Kohler NG, Joly A: The involvement of an LPS inducible IB kinase in endotoxin tolerance.
Biochem Biophys Res Commun 1997;232:602–607.
27 De Martin R, Vanhove B, Cheng Q, Hofer E, Csizmadia V, Winckler H, Bach FH: Cytokine-
inducible expression in endothelial cells of an IB␣-like gene is regulated by NFB. EMBO J
1993;12:2773–2779.
28 Jäättelä M, Wising D: Heat shock proteins protect cells from monocyte cytotoxicity: possible
mechanism of self-protection. J Exp Med 1993;177:231–236.
29 Cahill CM, Waterman WR, Xie Y, et al: Transcriptional repression of the prointerleukin 1beta gene
by heat shock factor 1. J Biol Chem 1996;271:24874–24879.
30 Cossarizza A, Baccarani-Contri M, Kalashnikova G, et al: A new method for the cytofluorimetric
analysis of mitochondrial membrane potential using the J-aggregate forming lipophilic cation
5,5?,6,6’-tetrachloro-1,1’,3,3?-tetraethylbenzimidazolyl-carbocyanine iodide (JC-1). Biochem
Biophys Res Commun 1993;197:40–45.
31 Polla BS, Kantegwa D, François S, et al: Mitochondria are selective targets for the protective
effects of heat shock against oxidative injury. Proc Natl Acad Sci USA 1996;93:6458–6463.
32 Bellmann K, Wenz A, Radons J, Burkart V, Kleemann R, Kolb H: Heat shock induces resistance in
rat pancreatic islet cells against nitric oxide, oxygen radicals and streptozotocin toxicity in vitro. J
Clin Invest 1995;95:2840–2845.
33 Gabai VL, Kabakov AE: Rise in heat-shock protein level confers tolerance to energy deprivation.
FEBS Lett 1993;327:247–250.
34 Malyshev IY, Malugin AV, Golubeva LY, Zenina TA, Manukhina EB, Mikoyan VD, Vanin AF
Nitric oxide donor induces HSP70 accumulation in the heart and in cultured cells. FEBS Lett
1996;391:21–23.
35 Amin J, Ananthan J, Voellmy R: Key features of heat shock regulatory elements. Mol Cell Biol
1988;8:3761–3769.
36 Chen FE, Ghosh G: Regulation of DNA binding by Rel/NF-B transcription factors: structural
views. Oncogene 1999;18:6845–6852.
37 Xiao ZX, Zuo XX, Davis AA, et al: HSF1 is required for extra-embryonic development, postnatal
growth and protection during inflammatory responses in mice. EMBO J 1999;18:5943–5952.
38 Santoro MG: Heat shock factors and the control of the stress response. Biochem Pharmacol
2000;59:55–63.
39 van der Poll T, Calvano SE, Kumar A, et al: Endotoxin induces downregulation of tumor necrosis
factor receptors on circulating monocytes and granulocytes in humans. Blood 1995;86:2754–2759.
Sepsis and Multiple Organ Failure 63
40 Detmers PA, Powell DE, Walz A, et al: Differential effects of neutrophil-activating peptide 1/IL-8
and its homologues on leukocyte adhesion and phagocytosis. J Immunol 1991;147:4211–4217.
41 Steer JH, Kroeger KM, Abraham LJ, et al: Glucocorticoids suppress tumor necrosis factor-alpha
expression by human monocytic THP-1 cells by suppressing transactivation through adjacent NF-
B and c-Jun-activating transcription factor-2 binding sites in the promoter. J Biol Chem 2000;
275:18432–18440.
42 Villar J, Edelson JD, Post M, Mullen BM, Slutsky AS: Induction of the heat stress proteins is asso-
ciated with decreased mortality in an animal model of acute lung injury. Am Rev Respir Dis
1993;147:177–181
43 Villar J, Ribiero SP, Mullen BM, Kuliszewski M, Post M, Slutsky AS: Induction of the heat shock
response reduces mortality rate and organ damage in a sepsis-induced acute lung injury model.
Crit Care Med 1994;22:914–921.
44 Polla BS, Jacquier-Sarlin MR, Kantengwa S, Mariethoz E, Hennet T, Russo-Marie F, Cossarizza
A: TNF-␣ alters mitochondrial membrane potential in L929 but not in TNF␣-resistance L929.12
cells: relationship with the expression of stress proteins, annexin 1 and superoxide dismutase
activity. Free Radic Res 1996;25:125–131.
45 Wong HR, Ryan M, Wispé JR: Stress response decreases N-FB nuclear translocation and
increases I-B␣ expression in A549 cells. J Clin Invest 1997;99:2423–2428.
46 Boutten A, Dehoux MS, Seta N, Ostinelli J, Venembre P, Crestani B, Dombret MC, Durand G,
Aubier M: Compartmentalized IL-8 and elastase release within the human lung in unilateral pneu-
monia. Am J Respir Crit Care Med 1996;153:336–342.
47 Docke WD, Syrbe U, Meinecke A, Platzer C, Makki A, Asadullah K, Klug C, Zuckermann H,
Reinke P, Brunner H, von Baehr R, Volk HD: Improvement of monocyte function – a new thera-
peutic approach? in Reinhart K, Eyrich K, Sprung C (eds): Sepsis: Current Perspectives in
Pathophysiology and Therapy. Berlin, Springer, 1994, pp 473–500.
48 Rosenbloom AJ, Pinsky MR, Bryant JL, Shin A, Tran T, Whiteside T: Leukocyte activation in the
peripheral blood of patients with cirrhosis of the liver and SIRS. Correlation with serum inter-
leukin-6 levels and organ dysfunction. JAMA 1995;274:58–65.
49 Trautinger F, Hammerle AF, Poschl G, Micksche M: Respiratory burst capability of polymor-
phonuclear neutrophils and TNF-alpha serum levels in relationship to the development of septic
syndrome in critically ill patients. J Leukoc Biol 1991;49:449–454.
50 Nakae H, Endo S, Inada K, Takakuwa T, Kasai T: Changes in adhesion molecule levels in sepsis.
Res Commun Mol Pathol Pharmacol 1996;91:329–338
51 Lin RY, Astiz ME, Saxon JC, Rackow EC: Altered leukocyte immunophenotypes in septic shock.
Studies of HLA-DR, CD11b, CD14, and IL-2R expression. Chest 1993;104:847–853.
52 Rosenbloom AJ, Pinsky MR, Napolitano C, Nguyen T-S, Levann D, Pencosky N, Dorrance A, Ray
BK, Whiteside T: Suppression of cytokine mediated 2-integrin activation on circulating neu-
trophils in critically ill patients. J Leukoc Biol 1999;66:83–89.
53 Diamond MS, Springer TA: A subpopulation of Mac-1 (CD11b/CD18) molecules mediates neu-
trophil adhesion to ICAM-1 and fibrinogen. J Cell Biol 1993;120:545–556.
54 Terregino CA, Lubkin C, Thom SR: Impaired neutrophil adherence as an early marker of systemic
inflammatory response syndrome and severe sepsis. Ann Emerg Med 1997;29:400–403.
55 Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich E:
Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med
2001;345:1368–1377.
Michael R. Pinsky, MD
606 Scaife Hall
3550 Terrace Street
Pittsburgh, PA 15261 (USA)
Tel. ⫹1 412 647 5387, Fax ⫹1 412 647 8060, E-Mail pinskymr@upmc.edu
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 64–74
Classification, Incidence, and Outcomes
of Sepsis and Multiple Organ Failure
Jean-Louis Vincent, Fabio Taccone, Xavier Schmit
Department of Intensive Care, Erasme Hospital, Free University of Brussels,
Brussels, Belgium
Abstract
Background/Aims: Sepsis and multiple organ failure are common complications in
intensive care unit (ICU) patients and are associated with considerable morbidity and mor-
tality. Methods: We reviewed pertinent medical literature related to sepsis and multiple
organ failure to determine strategies of classification, the current incidence, and the out-
comes associated with these disease processes. Results: Sepsis affects some 40% of ICU
admissions, severe sepsis occurs in about 30%, and septic shock in 15%. Recent consensus
has improved the definition of sepsis and proposed a new classification system based on pre-
disposing factors, infection, immune response, and organ dysfunction. We discuss the possi-
ble components of each of these four categories. Conclusion: Although there is some
evidence that mortality rates may have decreased in recent years, the incidence of sepsis is
increasing so that overall deaths from this disease are increasing. Improved diagnostic tech-
niques and classification may help target therapies more rapidly and more appropriately.
Copyright © 2007 S. Karger AG, Basel
Sepsis is an increasingly common problem in the intensive care unit (ICU)
and severe sepsis and septic shock are associated with high morbidity, mortal-
ity, and costs. Attempts have been made recently to improve definitions and
characterization of sepsis and multiple organ failure (MOF). Large epidemio-
logical observational studies conducted in recent years have provided important
information about the scale of this problem and its impact, and suggest that it is
becoming more common, likely due to aging populations, the prolonged sur-
vival of patients who would previously have died earlier, and increased use of
immunosuppressive agents. Such data are valuable in determining appropriate
resource allocation and in providing a baseline for ongoing assessment of the
impact of the introduction of new therapeutic interventions and strategies.
Sepsis and MOF 65
Classification of Sepsis and MOF
One of the difficulties in the field of sepsis over the years has been in deter-
mining how best to define it. Unlike many other disease processes, where there
are relatively clear signs and symptoms, a diagnosis of sepsis in a critically ill
patient is often complicated; signs and symptoms are not specific, there is no
simple marker, no magical imaging technique, and microbiological cultures are
unreliable, being negative in some 40% of patients [1]. A Consensus Conference
organized by the American College of Chest Physicians (ACCP) and the Society
of Critical Care Medicine (SCCM) in 1991 [2] coined the term systemic inflam-
matory response syndrome (SIRS) and defined sepsis as SIRS occurring when
infection was present. To meet the SIRS criteria, patients needed to satisfy at
least two of the following: fever or hypothermia, tachycardia, tachypnea or
hyperventilation, leukocytosis or leukopenia. Severe sepsis was defined as sep-
sis complicated by organ dysfunction, and septic shock as severe sepsis with per-
sistent arterial hypotension. However, the SIRS criteria can be met by many ICU
patients and are too nonspecific to be widely useful in the diagnosis of sepsis or
in classifying patients with sepsis. A more recent sepsis definition conference,
sponsored by the SCCM, the ACCP, the American Thoracic Society (ATS), the
European Society of Intensive Care Medicine (ESICM), and the Surgical
Infection Society (SIS), decided that the SIRS concept should be abandoned and
replaced by an expanded list of signs and symptoms of sepsis that are more rep-
resentative of the clinical response to infection [3].
Sepsis is a complex disease process and the population of patients who
develop sepsis is highly heterogeneous. Sepsis can affect all patients of any age
and sex, who may have multiple comorbidities or none at all. It can be caused by
bacteria, viruses, or fungi, and can arise in any organ of the body. Patients with
sepsis develop different degrees of immune response and their response varies
over time. One way of characterizing patients with sepsis is to use mortality pre-
diction or severity of illness scores, like the acute physiology and chronic health
evaluation (APACHE) and simplified acute physiology score (SAPS) systems.
Rather than an aggregate score, an organ dysfunction score, like the sequential
organ failure assessment (SOFA) score, describes better the degree of dysfunc-
tion of each individual organ and can be calculated repeatedly. However, organ
dysfunction and illness severity are only two small components of the intricate
septic process, and a more complex and inclusive staging system is required. The
Sepsis Consensus Conference suggested a way in which all the heterogeneous
characteristics could be included into a staging system to enable patients with
sepsis to be better grouped, thus facilitating research and perhaps enabling
therapies to be more appropriately targeted. They called this system, PIRO, with
the initials of the acronym standing for four key features of the sepsis response:
Vincent/Taccone/Schmit 66
predisposing factors, infection, response, and organ dysfunction [3]. Broadly
based on the TNM (tumor, nodes, metastases) staging system for cancers, it is
envisaged that points could be allocated to each of these aspects, such that a
patient with sepsis could, for example, be staged as P
1
I
2
R
1
O
0
[4], depending on
the features present for each of the four PIRO components.
Predisposing Factors
There are many factors that predispose a patient to developing sepsis, includ-
ing age, sex, comorbidities, and genetic makeup. Older patients are more likely
to develop severe sepsis than their younger counterparts [5], and more likely to
have a worse outcome. Men seem to be at greater risk of developing sepsis and
MOF [6–8] although the effect of sex on outcome from sepsis is less clear [7,
9], and may be related in part to hormonal status. Various genetic factors also
influence the development, severity and outcome from sepsis [10–14], as do the
presence of certain comorbidities or ongoing or recent immunosuppressive
medication.
Infection
Sepsis is associated with an underlying infection, which needs to be char-
acterized, typically by the source of infection and the incriminated organisms.
These features can also be related to severity and outcome. For example, pneu-
monia is more likely to be associated with severe sepsis than urinary tract infec-
tion, certain microorganisms, e.g., Pseudomonas species, are associated with a
greater risk of death [1], and organisms resistant to multiple antimicrobial
agents will be less likely to respond to therapy and, therefore, may also be asso-
ciated with worse outcomes.
Response
The host response in sepsis varies among patients and in the same patient
over time and the severity or nature of the response may influence how a patient
responds to particular therapies. Various factors will influence a patient’s
response to sepsis, as listed under predisposing factors, and techniques to assess
and monitor the degree of host response are urgently needed. Markers of sepsis,
including C-reactive protein and procalcitonin, are useful [15–17], but due to
the complexity of the sepsis response a combination of markers is likely to be
needed to fully assess a patient’s immune status. Advances in proteomic and
microarray techniques are likely to facilitate classification of the host response.
Organ Dysfunction
The presence of sepsis-associated organ dysfunction is an indication of the
severity of the sepsis response, and is associated with outcome, the greater the
Sepsis and MOF 67
number of failing organs, the higher the mortality (fig. 1) [1, 6, 18]. Organ dys-
function can be measured and monitored using various scoring systems, of
which the most widely used is the SOFA score [18]. Any organ can be involved,
but the cardiovascular, respiratory, and renal organs are most commonly
affected [1, 6, 19]. The pattern of organ dysfunction in sepsis varies among
patients and with time in the same patient and careful monitoring can help
improve our understanding of the pathophysiology of sepsis-related organ dys-
function, as well as guiding therapy and predicting outcome.
Incidence of Sepsis and MOF
Many observational or epidemiological studies have been conducted in
recent years assessing the incidence of sepsis, some in individual centers, others
in multiple centers in one geographical area, and others crossing international
boundaries [1, 6, 19–26]. Two key techniques are used to assess how many peo-
ple are affected by sepsis: prevalence, i.e., the percentage of a population that is
affected with sepsis at a given time, and incidence, i.e., the number of new
Fig. 1. Frequency of organ failure in patients in the SOAP study on admission to the
ICU and corresponding ICU mortality [from 1 with permission].
0 1 2 3 4 or more
0
10
20
30
40
50
Patients (%)
0 1 2 3 4 or more
0
10
20
30
40
50
60
70
Mortality
Number of failing organs
Vincent/Taccone/Schmit 68
episodes of sepsis that commence during a specified period of time in a speci-
fied population. The different techniques, different definitions used for sepsis
and organ dysfunction, and different study populations make it difficult to com-
pare study results (table 1).
Using discharge records from hospital in seven states in the United States,
Angus et al. [20] identified 192,980 cases of severe sepsis in 1995, giving
annual estimates of 751,000 cases (3.0 cases/1,000 population and 2.26 cases/
100 hospital discharges), although these numbers may be an overestimate [27].
In 23 ICUs across New Zealand and Australia, Finfer et al. [24] calculated the
incidence of severe sepsis to be 0.77/1,000 population. In the UK, the incidence
of severe sepsis was estimated to be 0.51/1,000 population [19], and in Finland
0.38/1,000 population [25]. In the EpiSepsis study, which included 3,738
admissions to ICUs in France over a 2-week period, 14.6% of patients had
severe sepsis. In the SOAP study, which studied 3,147 patients in 198 ICUs
across Europe, 37% of the patients had sepsis at some point during their ICU
stay, 30% of the patients had severe sepsis, and 15% had septic shock [1];
the frequency varied considerably in the different countries, from 10% in
Switzerland to 63% in Portugal. This study also assessed the incidence of organ
failure, as defined by a SOFA ⬎2 for the organ in question [18], and reported
that 71% of patients had at least one organ failure at some point during their
ICU stay. Sepsis was present in 41% of episodes of organ failure.
Attempts have also been made to determine whether the numbers of
patients with sepsis is changing over time [6, 28]. Using discharge data from
750 million hospitalizations in the United States, Martin et al. [6] assessed the
occurrence of sepsis in 2000 to be 2.4/100,000 population, an increase of some
8.7% compared to the 0.83/100,000 population estimated for 1979. In an analy-
sis of data from 100,554 ICU admissions across France over an 8-year period,
Annane et al. [26] reported that the frequency of septic shock had increased
from 7% in 1993 to 9.7% in 2000. In England, Wales and Northern Ireland
data from 343,860 admissions to 172 adult, general critical care units bet-
ween December 1995 and January 2005 showed an increase in the numbers
of ICU patients admitted with severe sepsis, from 23.5% in 1996 to 28.7% in
2004 [28].
Outcome from Sepsis and MOF
Sepsis was associated with an ICU mortality rate of 27% in the SOAP
study [1], ranging from 10% in Switzerland to 35% in Italy [1]. Severe sepsis
was associated with ICU mortality rates of 32% and in patients with septic
shock, the mortality rate rose to 54% [1]. In a multivariate analysis, the
Sepsis and MOF 69
Table 1. Results from recent large, multicenter epidemiological studies assessing the incidence of and outcome from sepsis
Authors Year of Study Size of study Methodology Diagnosis of Incidence/ Mortality
data population population sepsis used prevalence/
frequency
Angus 1995 Nonfederal 6,621,559 Analysis of Severe sepsis defined 3.0 cases of Hospital mortality
et al. [20] acute hospitals hospitaliza- discharge as documented severe 28.6%
in 7 US states tions records infection and acute sepsis/1,000
organ dysfunction population
using criteria from
the International
Classification of
Diseases
Martin 1979– Nonfederal 750 million Analysis of Severe sepsis defined 82.7 cases of severe Hospital mortality
et al. [6] 2000 acute hospitals hospitaliza- discharge as documented sepsis/100,000 27.8% for 1979–
in USA tions records infection and acute population in 1979 1984; 17.9% for
organ dysfunction to 240.4/100,000 1995–2000
using criteria from population in 2000
the International
Classification of
Diseases
Finfer May 1, 23 ICUs in 5,878 ICU Inception Severe sepsis as 0.77 cases of severe ICU: 26.5%; 28-
et al. [24] 1999 to Australia and admissions cohort study defined by presence sepsis/1,000 day: 32.4%;
July 31, New Zealand of SIRS criteria plus population; 11.8/100 hospital: 37.5%
1999 organ dysfunction ICU admissions
Padkin 1995– 172 ICUs in 56,673 ICU Retrospective Severe sepsis as 0.51 cases of severe ICU: 35%; hospital:
et al. [19] 2000 England, admissions data analysis defined in the sepsis/1,000 47%
Wales, PROWESS study [32] population
Ireland occurring in first 24 h
of ICU admission
Vincent/Taccone/Schmit 70
Guidet 1997– 35 ICUs 65,910 ICU Review of Severe sepsis defined 27.7% ICU mortality
et al. [29] 2001 in France admissions prospectively as infection plus at 10.5% in patients
with stay collected data least one organ with one organ
⬎24h dysfunction failure, 42.7% in
patients with at
least two organ
failures
Brun- November 206 ICUs 3,738 ICU Inception Severe sepsis as Severe sepsis in 35% 30-day
Buisson 19, 2001 in France admissions cohort study defined by 14.6% of patients mortality
et al. [22] to ACCP/SCCM
December consensus
2, 2001 conference
definitions [2]
Silva May 2001 to 5 ICUs 1,383 ICU Observational Sepsis, severe sepsis, Sepsis in 61.4%, 33.9% for sepsis,
et al. [23] January in Brazil admissions cohort study and septic shock as severe sepsis in 46.9% for severe
2002 defined by ACCP/ 35.6%, septic sepsis and 52.2%
SCCM consensus shock in 30% for septic shock
conference
definitions [2]
Vincent May 1– 198 ICUs in 3,147 ICU Prospective Sepsis, severe sepsis Sepsis in 37% of ICU mortality in
et al. [1] 15, 2002 24 European admissions cohort and septic shock as patients, severe sepsis 27%, severe
countries observational defined by sepsis in 30%, sepsis 32.2%,
study ACCP/SCCM septic shock in 15% septic shock 52.1%
consensus
conference
definitions [2]
Table 1. (continued)
Authors Year of Study Size of study Methodology Diagnosis of Incidence/ Mortality
data population population sepsis used prevalence/
frequency
Sepsis and MOF 71
Zahorec July to 12 ICUs in 1,533 ICU Observational Severe sepsis as Severe sepsis in Hospital mortality
et al. [33] December Slovak admissions cohort study defined in the 7.9% of ICU 51.2%
2002 Republic PROWESS admissions
study [32]
Karlsson November 24 ICUs in 4,500 ICU Prospective Severe sepsis as 0.38 cases of severe ICU mortality
et al. [25] 1, 2004 Finland admissions observational defined by sepsis/1,000 15.5%, hospital
to study ACCP/SCCM population mortality 28.3%
February consensus
28, 2005 conference
definitions [2]
Vincent/Taccone/Schmit 72
SAPS II score on admission, the cumulative fluid balance within the first 72 h
of the onset of sepsis, age, SOFA score at the onset of sepsis, bloodstream
infection, cirrhosis, Pseudomonas sp. infection, and being a medical admis-
sion were associated with an increased risk of death [1]. Other studies have
reported similar findings, although again definitions differ among studies
making a direct comparison difficult (table 1). In the UK, Padkin et al. [19]
noted an ICU mortality rate of 35% in patients with severe sepsis; in the USA
hospital mortality was reported as 28.6% [20]; in Australia and New Zealand
the ICU mortality rate for patients with severe sepsis was 27.5% [24], and in
Brazil mortality rates of 34.7, 47.3 and 52.2% were reported for sepsis, severe
sepsis, and septic shock, respectively [23]. In patients with an ICU stay ⬎24 h,
Guidet et al. [29] noted that patients with at least two organ dysfunctions had
longer ICU stays (20.4 vs. 11.6 days) and greater ICU mortality (42.7 vs.
5.5%) than patients with just one organ dysfunction. In the EpiSepsis study,
30-day mortality was 35% and chronic liver and heart failure, acute renal fail-
ure and shock, SAPS II at onset of severe sepsis, and 24-hour total SOFA
scores were the independent factors most associated with an increased risk of
death [22].
Interestingly, sepsis is not only associated with high mortality during the
immediate ICU and hospital stay, but may increase the risk of death for up to 5
years after the septic episode even after adjusting for the presence of comor-
bidities [30]. Health-related quality of life may also be reduced in survivors of
sepsis compared to other ICU survivors [31].
Conclusion
Sepsis affects some 40% of ICU patients at some point during their stay,
and when severe is associated with mortality rates of around 30%. Although
studies have suggested that mortality rates from sepsis may be decreasing
[6, 28], as the incidence of sepsis continues to increase, the numbers of patients
dying from sepsis are also still increasing [28]. The heterogeneous nature of the
patients who develop sepsis, the range of infectious organisms and sources of
infection, and the complexities of the immune response to severe sepsis make
the classification of patients with sepsis a challenge, and yet the ability to do so
is seen as increasingly important for clinical trial development and appropriate
therapeutic targeting. With further development and validation, the PIRO sys-
tem may provide a means by which this can be done, aiding the development of
new therapeutic strategies and ultimately improving outcomes for patients with
sepsis.
Sepsis and MOF 73
References
1 Vincent JL, Sakr Y, Sprung CL, et al: Sepsis in European intensive care units: results of the SOAP
study. Crit Care Med 2006;34:344–353.
2 ACCP-SCCM Consensus Conference: Definitions of sepsis and multiple organ failure and guide-
lines for the use of innovative therapies in sepsis. Crit Care Med 1992;20:864–874.
3 Levy MM, Fink MP, Marshall JC, et al: 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis
Definitions Conference. Crit Care Med 2003;31:1250–1256.
4 Marshall JC, Vincent JL, Fink MP, et al: Measures, markers, and mediators: toward a staging sys-
tem for clinical sepsis. A report of the Fifth Toronto Sepsis Roundtable, Toronto, Ontario, Canada,
October 25–26, 2000. Crit Care Med 2003;31:1560–1567.
5 Martin GS, Mannino DM, Moss M: The effect of age on the development and outcome of adult
sepsis. Crit Care Med 2006;34:15–21.
6 Martin GS, Mannino DM, Eaton S, et al: The epidemiology of sepsis in the United States from
1979 through 2000. N Engl J Med 2003;348:1546–1554.
7 Wichmann MW, Inthorn D, Andress HJ, et al: Incidence and mortality of severe sepsis in surgical
intensive care patients: the influence of patient gender on disease process and outcome. Intensive
Care Med 2000;26:167–172.
8 Frink M, Pape HC, van Griensven M, et al: Influence of sex and age on MODS and cytokines after
multiple injuries. Shock 2007;27:151–156.
9 Eachempati SR, Hydo L, Barie PS: Gender-based differences in outcome in patients with sepsis.
Arch Surg 1999;134:1342–1347.
10 Appoloni O, Dupont E, Andrien M, et al: Association of TNF2, a TNF␣ promoter polymorphism,
with plasma TNF␣ levels and mortality in septic shock. Am J Med 2001;110:486–488.
11 Arnalich F, Lopez-Maderuelo D, Codoceo R, et al: Interleukin-1 receptor antagonist gene poly-
morphism and mortality in patients with severe sepsis. Clin Exp Immunol 2002;127:331–336.
12 Arbour NC, Lorenz E, Schutte BC, et al: TLR4 mutations are associated with endotoxin hypore-
sponsiveness in humans. Nat Genet 2000;25:187–191.
13 Lorenz E, Mira JP, Cornish KL, et al: A novel polymorphism in the toll-like receptor 2 gene and its
potential association with staphylococcal infection. Infect Immun 2000;68:6398–6401.
14 Stassen NA, Leslie-Norfleet LA, Robertson AM, et al: Interferon-gamma gene polymorphisms
and the development of sepsis in patients with trauma. Surgery 2002;132:289–292.
15 Claeys R, Vinken S, Spapen H, et al: Plasma procalcitonin and C-reactive protein in acute septic
shock: clinical and biological correlates. Crit Care Med 2002;30:757–762.
16 Luzzani A, Polati E, Dorizzi R, et al: Comparison of procalcitonin and C-reactive protein as mark-
ers of sepsis. Crit Care Med 2003;31:1737–1741.
17 Christ-Crain M, Stolz D, Bingisser R, et al: Procalcitonin-guidance of antibiotic therapy in
community-acquired pneumonia: a randomized trial. Am J Respir Crit Care Med 2006;174:84–93.
18 Vincent JL, de Mendonça A, Cantraine F, et al: Use of the SOFA score to assess the incidence of
organ dysfunction/failure in intensive care units: results of a multicentric, prospective study. Crit
Care Med 1998;26:1793–1800.
19 Padkin A, Goldfrad C, Brady AR, et al: Epidemiology of severe sepsis occurring in the first
24 hrs in intensive care units in England, Wales, and Northern Ireland. Crit Care Med 2003;31:
2332–2338.
20 Angus DC, Linde-Zwirble WT, Lidicker J, et al: Epidemiology of severe sepsis in the United States:
analysis of incidence, outcome, and associated costs of care. Crit Care Med 2001;29:1303–1310.
21 Alberti C, Brun-Buisson C, Burchardi H, et al: Epidemiology of sepsis and infection in ICU
patients from an international multicentre cohort study. Intensive Care Med 2002;28:108–121.
22 Brun-Buisson C, Meshaka P, Pinton P, et al: EPISEPSIS: a reappraisal of the epidemiology and
outcome of severe sepsis in French intensive care units. Intensive Care Med 2004;30:580–588.
23 Silva E, Pedro MA, Sogayar AC, et al: Brazilian Sepsis Epidemiological Study (BASES study).
Crit Care 2004;8:R251–R260.
24 Finfer S, Bellomo R, Lipman J, et al: Adult-population incidence of severe sepsis in Australian and
New Zealand intensive care units. Intensive Care Med 2004;30:589–596.
Vincent/Taccone/Schmit 74
25 Karlsson S, Varpula M, Ruokonen E, et al: Incidence, treatment, and outcome of severe sepsis in
ICU-treated adults in Finland: the Finnsepsis study. Intensive Care Med 2007;33:435–443.
26 Annane D, Aegerter P, Jars-Guincestre MC, et al: Current epidemiology of septic shock: the CUB-
Rea Network. Am J Respir Crit Care Med 2003;168:165–172.
27 Wenzel RP, Edmond MB: Severe sepsis-national estimates. Crit Care Med 2001;29:1472–1474.
28 Harrison DA, Welch CA, Eddleston JM: The epidemiology of severe sepsis in England, Wales and
Northern Ireland, 1996 to 2004: secondary analysis of a high quality clinical database, the
ICNARC Case Mix Programme Database. Crit Care 2006;10:R42.
29 Guidet B, Aegerter P, Gauzit R, et al: Incidence and impact of organ dysfunctions associated with
sepsis. Chest 2005;127:942–951.
30 Quartin AA, Schein RM, Kett DH, et al: Magnitude and duration of the effect of sepsis on sur-
vival. Department of Veterans Affairs Systemic Sepsis Cooperative Studies Group. JAMA
1997;277:1058–1063.
31 Heyland DK, Hopman W, Coo H, et al: Long-term health-related quality of life in survivors of
sepsis. Short form 36: a valid and reliable measure of health-related quality of life. Crit Care Med
2000;28:3599–3605.
32 Bernard GR, Vincent JL, Laterre PF, et al: Efficacy and safety of recombinant human activated
protein C for severe sepsis. N Engl J Med 2001;344:699–709.
33 Zahorec R, Firment J, Strakova J, et al: Epidemiology of severe sepsis in intensive care units in the
Slovak Republic. Infection 2005;33:122–128.
Dr. J.-L. Vincent
Erasme University Hospital
Route de Lennik 808
BE–1070 Brussels (Belgium)
Tel. ⫹32 2 555 3380, Fax ⫹32 2 555 4555, E-Mail jlvincen@ulb.ac.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 75–91
Genetic Polymorphisms in Sepsis- and
Cardiopulmonary Bypass-Associated
Acute Kidney Injury
Anja Haase-Fielitz
a,b
, Michael Haase
a,b
, Rinaldo Bellomo
a
, Duska Dragun
c
a
Department of Intensive Care, Austin Health, Melbourne, Vic., Australia;
b
Department of Nephrology, Charité University Medicine, Berlin, and
c
Department of Nephrology and Intensive Care and Center for Cardiovascular
Research, Charité, University Berlin, Berlin, Germany
Abstract
Acute kidney injury (AKI) is a major medical problem in critical illness, and has a
separate independent effect on the risk of death. Septic shock and cardiac surgery utilizing
cardiopulmonary bypass are the two most common factors contributing to AKI. Clinical pre-
dictors and biochemical markers identified for the development of AKI can only explain a
part of this individual risk. Another tool to predict the risk of AKI and to improve individual-
ized patient care focuses on the identification of genetic risk factors which might be involved
in the development of AKI. However, to date our knowledge on the importance of such
genetic polymorphisms in influencing the susceptibility to and severity of AKI remains lim-
ited. There is evidence that several genetic polymorphisms accounting for sepsis- or car-
diopulmonary bypass-associated AKI involve genes which participate in the control of
inflammatory or vasomotor processes. In this article, we will review current knowledge con-
cerning the role of genetic polymorphism in the pathogenesis of sepsis- and cardiopul-
monary bypass-associated AKI and discuss possible areas for future developments and
research in this field.
Copyright © 2007 S. Karger AG, Basel
Acute kidney injury (AKI), as shown by rapid oliguria or anuria and an
increase in the blood concentration of kidney-dependent waste products, is a
major medical problem occurring in 5% of all patients admitted to hospital and
30% of those admitted to an intensive care unit [1]. A recent observational
study found that 4.2% of intensive care unit patients developed renal replace-
ment therapy-dependent AKI [2] and that AKI, defined according to the RIFLE
criteria [3], occurred in 5.2% of hospital patients.
Evaluation of Illness Severity
Haase-Fielitz/Haase/Bellomo/Dragun 76
Not only is AKI a common complication of critical illness but it also has a
separate independent effect on the risk of death [4, 5].
In a multinational, multicenter study, septic shock was the most common
contributing factor to AKI [2] accounting for around 47.5% of patients. It was
followed in incidence by cardiac surgery utilizing cardiopulmonary bypass
(CPB), which accounted for 23.2% of cases. Other rare causes comprise the use
of nephrotoxic drugs, hepatorenal syndrome, rhabdomyolysis, rapid glomeru-
lonephritis and obstructive uropathy.
Many clinical predictors and biochemical markers for the development of
AKI have been identified. Though risk stratification based on these factors can
only explain a part of the individual risk of developing AKI [6, 7].
Another tool to predict the risk of AKI and to improve individualized
patient care focuses on the identification of genetic risk factors, which might be
involved in the development of AKI. However, to date our knowledge about the
importance of such genetic polymorphisms in influencing the susceptibility to
and severity of AKI remains limited. In this article, we will review current
knowledge concerning the role of genetic polymorphism in the pathogenesis of
AKI and discuss possible areas of future developments and research in this
field.
Single Nucleotide Polymorphisms
In order to understand the role of genetic factors in the pathogenesis of
AKI, one first needs to become familiar with some fundamental aspects of
genetics. One such fundamental aspect relates to the concept of genetic poly-
morphism. Perhaps the first definition of genetic polymorphism originated in
1940 from Ford [8] who described it as ‘the occurrence together in the same
habitat of two or more discontinuous forms, or “phases”, of a species in such
proportions that the rarest of them cannot be maintained merely by recurrent
mutation’. Most frequently, however, genetic polymorphism is defined as a dif-
ference in DNA sequence among individuals, groups or populations, or as the
presence of two genotypes in a population.
Single nucleotide polymorphisms (SNPs) are DNA sequence variations
that occur when a single nucleotide in the genome sequence is altered. For a
variation to be considered a SNP, at least 1% of the population must present
with it. SNPs make up about 90% of all human genetic variation and occur
every 100–300 bases along the 3-billion-base human genome. Many SNPs have
no effect on cell function, whereas others can predispose people to disease
or influence their response to a drug. Although more than 99% of human
DNA sequences are the same across the population, variations in the remaining
Polymorphisms in Acute Kidney Injury 77
DNA sequence can have a major impact on how humans respond to disease,
environmental factors, such as toxins, drugs and other therapies. This makes
SNPs of great value for biomedical research.
However, currently investigated SNPs may not necessarily be a causal or
functional SNP, but rather, they might be in partial linkage disequilibrium with
a causal SNP and thus, can contribute to conflicting reports.
One of the first studies involving genetic polymorphism was published in
1958 [9], and 50 years later almost 100,000 citations can be found using elec-
tronic search engines such as PubMed for ‘genetic polymorphism’; a number
which is growing daily.
During the last decade, tremendous effort has been made to establish an
association or wherever possible a causality between genetic predisposition and
AKI.
However, the number of genetic polymorphism studies essentially focus-
ing on selected patient populations and specific clinical outcomes or organ dys-
function is still limited.
In the following, we will highlight the relationship of genetic polymor-
phism and acute kidney injury in the context of sepsis and cardiac surgery.
Sepsis-Associated Acute Kidney Injury
Epidemiology
AKI secondary to sepsis is a common diagnosis in the intensive care unit
[10] and occurs in approximately 19% of patients with moderate sepsis, 23%
with severe sepsis, and 51% with septic shock when blood cultures are posi-
tive [11]. The combination of AKI and sepsis shows a higher mortality rate
than AKI alone and hence constitutes a serious medical problem [10, 12].
Despite our increasing ability to support vital organs and resuscitate patients,
the incidence and mortality of patients with sepsis-associated AKI remains
high [13].
Genetic Polymorphisms
Substantial progress has been made toward understanding the mechanisms
whereby sepsis is associated with AKI. Although hemodynamic factors might
play a role in the pathogenesis of sepsis-associated AKI, other mechanisms are
likely to be involved, which are immunologic, toxic or inflammatory in nature
[10, 14]. The initial event in the pathogenesis of sepsis-associated AKI is a sys-
temic infection, which triggers a generalized humoral and cellular immunologic
response to it. Much clinical and molecular biology research suggests that
Haase-Fielitz/Haase/Bellomo/Dragun 78
pro- and anti-inflammatory cytokines, the generation of reactive oxygen
species (ROS) and the formation of glomerular microthrombi cause generalized
injury to glomerular endothelial cells [15–19]. In addition, there is evidence
that human renal tubular cells die by apoptosis as well as necrosis in experi-
mental models of sepsis [20].
Once sufficient evidence had accumulated for the possible pathophysio-
logical role of septic mediators in sepsis-associated AKI, the significance of
genetic polymorphism in patients with sepsis became a field of investigation
[21–23]. There is already evidence for a positive association between genetic
polymorphisms of immune response genes and sepsis in general [24]. However,
the number of studies of genetic variability in sepsis focusing on AKI has
remained limited. We included studies which fulfilled the following criteria:
patients with sepsis-associated AKI and for whom the relationship or distribu-
tion of various genotypes to renal outcome was reported.
Below, we will discuss the evidence available at this stage and focus on the
results of clinical studies focusing on genetic variants of tumor necrosis factor
(TNF)-␣, interleukin (IL)-6 and IL-10 genes and their role in sepsis-associated
AKI (table 1).
Tumor Necrosis Factor
The role of TNF-␣ in endotoxin-related AKI has been studied in both ani-
mals and human studies [23, 33, 34]. TNF has also been demonstrated to play a
major role in the pathogenesis of gram-negative septic shock mediating a broad
spectrum of host responses to endotoxin. In the kidney, endotoxin stimulates
the release of TNF-␣ from glomerular mesangial cells [35]. TNF-␣ is a cytokine
that initiates the inflammatory cascade and induces the production of numerous
additional mediators which are associated with the endothelial and tissue injury
seen in septic multiple organ failure [17, 36, 37]. The TNF-␣ gene is located on
the short arm of chromosome 6. Polymorphism within the 5⬘ flanking region of
the TNF-␣ gene – a region which influences transcriptional activity – has
been associated with high promoter activity and may therefore be of functional
relevance [38–40].
In a retrospective study, genotyping blood from 92 newborns, Treszl et al.
[26] found that the constellation of high TNF-␣-producer and low IL-6-
producer genetic variants (i.e., TNF-␣/IL-6 AG/GC or AG/CC haplotype) was
associated with AKI in neonates with very low birth weight and severe infection
(table 1).
In an international multicentric prospective study of 213 patients with
severe sepsis or septic shock, plasma levels of TNF-␣ were significantly higher
in those who died in the intensive care unit compared to those who survived.
Polymorphisms in Acute Kidney Injury 79
Table 1. Previous studies highlighting the association between a genetic polymorphism and sepsis-associated acute kidney injury
Gene Study design Patients Definition of AKI Definition of Primary Outcome Reference
sepsis inclusion
criteria
(all patients)
TNF-␣/ Retrospective n ⫽ 92 ARF as serum Early onset Sepsis Constellation of Treszl
IL-6 cohort study Neonates creatinine ⬎120 sepsis (at (TNF-␣/IL-6) AG/ et al. [26]
born with mol/l after 2nd least 2 of the GC or AG/CC
birth weight postnatal day or categories) haplotypes were
⬍1,500 g serum urea ⬎9 [25] more often present
mmol/l, diuresis in AKI (26 vs. 6%)
⬍1.0 ml urine/kg
body weight/h
TNF-␣/ Prospective, n ⫽ 213 AKI as part of the Severe sepsis Sepsis No association Gordon
TNF-␣R observational, SOFA score [27] or septic between candidate et al. [29]
multicenter shock [28] TNF or TNF receptor
study polymorphism (SF1A,
B) and renal function
or mortality
IL-10 Genetic n ⫽ 550 AKI defined as Sepsis [28] Sepsis IL-10 haplotype Wattanathum
association days alive and free (⫺592C/734G/ et al. [31]
study of organ dysfunction 3367G) CGG
using the Brussels haplotype more
criteria [30]; renal frequently associ-
support defined ated with AKI and
as HD, PD or any with increased need
continuous renal for renal support
replacement mode compared with other
haplotypes
Haase-Fielitz/Haase/Bellomo/Dragun 80
Table 1. (continued)
Gene Study design Patients Definition of AKI Definition of Primary Outcome Reference
sepsis inclusion
criteria
(all patients)
NADPH Prospective n ⫽ 200 Incremental increase Sepsis [28] AKI Incidence of sepsis Perianayagam
oxidase cohort study in serum creatinine not significantly et al. [32]
p22phox/ by 0.5, 1.0 or 1.5 different according
catalase mg/dl (see [1]) to various NADPH
oxidase or catalase
genotype groups
AKI ⫽ Acute kidney injury; HD ⫽ hemodialysis; IL ⫽ interleukin; PD ⫽ peritoneal dialysis; TNF-␣R ⫽ tumor necrosis factor receptor;
TNF-␣⫽tumor necrosis factor.
Polymorphisms in Acute Kidney Injury 81
However, there were no significant associations between the selected candidate
TNF-␣ or TNF-␣ receptor polymorphisms, or their haplotypes, and susceptibil-
ity to sepsis, illness severity or outcome including the degree of AKI [29]. Thus,
the role of TNF-␣-related genetic polymorphism in the pathogenesis of sepsis-
associated AKI remains uncertain.
Interleukin-10
IL-10 is an important component of the anti-inflammatory cytokine net-
work in sepsis [41, 42]. The IL-10 gene is located on the long arm of chromo-
some 1 and its SNP is also located in the promoter region.
In a genetic association study of 550 patients with sepsis, the IL-10 gene
CGG haplotype (⫺592C/734G/3367G) was associated with increased 28-day
mortality in critically ill patients who had sepsis from pneumonia [31]. Acute
kidney injury and the need for renal replacement therapy was evaluated as days
alive and free of organ dysfunction (for a maximum of 28 days). Patients who
carried one or two copies of the CGG haplotype were alive for longer and free
of AKI including the need for renal support compared to patients carrying no
copies of the CGG haplotype (table 1). Thus it appears that the IL-10 CGG hap-
lotype might be protective from sepsis-associated AKI.
Pro- and Antioxidant Enzymes
NADPH oxidase is a membrane-associated enzyme that catalyses the pro-
duction of superoxide and is highly expressed in neutrophils and endothelial
cells. In contrast, catalase is an antioxidant enzyme which metabolizes hydro-
gen peroxide and thereby limits oxidative stress-mediated injury. In a recent
study [32] the relationship of SNP in the coding region (C to T substitution at
position ⫹242) of the NADPH oxidase p22phox subunit gene to adverse clini-
cal outcomes was evaluated in a cohort of 200 patients with established AKI of
mixed cause and severity. Within this cohort the incidence of sepsis (average
45%) was not significantly different according to the various NADPH oxidase
or catalase genotype groups.
Other studies have also investigated further inflammatory SNPs [43] and
heat shock protein genetic polymorphisms in patients with AKI [44]. However,
in those studies patients with sepsis as the underlying reason for AKI were only
a subgroup, for which further data on a specific association between the genetic
polymorphism and renal outcome was not provided. These studies were
excluded from further review.
In summary, although the pathogenesis of sepsis-associated AKI appears
to be multifactorial, most genetic association studies with renal outcome have
focused on cytokine and pro- and antioxidant enzyme SNPs. Further studies
Haase-Fielitz/Haase/Bellomo/Dragun 82
specifically designed to investigate the relationship between sepsis-associated
AKI and genetic polymorphisms seem desirable.
Cardiopulmonary Bypass-Associated Acute Kidney Injury
Epidemiology
AKI remains a serious complication for patients undergoing cardiac
surgery with an incidence ranging between 5 and 50%, depending on the defin-
ition used [45–48]. AKI following cardiac surgery is associated with an
increase in mortality, morbidity, and prolonged stay in hospital and intensive
care, particularly for patients requiring renal replacement therapy [4, 49].
Indeed, the development of AKI requiring renal replacement therapy has been
identified as the strongest independent risk factor for death with an odds ratio
of 7.9 (95% CI 6–10) in a large retrospective study of cardiac surgical patients
[4]. Even minor degrees of postoperative AKI are associated with increased
mortality [47].
Genetic Polymorphisms
Prognostic risk stratification is used to predict AKI and identify patients
who are at a greater risk of developing postoperative AKI. Several independent
risk factors of AKI including age, pre-existing renal disease, diabetes and low
cardiac output have been identified [7, 50, 51]. Preoperative renal risk stratifi-
cation provides an opportunity to develop strategies for early diagnosis and
intervention. There is the potential that existing clinical scores for renal risk
stratification can be strengthened by taking into account variability in genetic
polymorphisms predisposing to postoperative AKI.
Multiple causes of AKI following cardiac surgery have been proposed,
including perioperative hemodynamic instability and impaired renal blood flow,
ischemia-reperfusion injury, CPB-induced activation of inflammatory pathways
and the generation of ROS [48, 52–55]. In addition, CPB causes hemolysis and
free hemoglobin-induced AKI may be another contributor to CPB-associated
AKI.
Ischemia-reperfusion injury and the generation of oxido-inflammatory
stress represent two conventionally accepted major mechanisms in the patho-
genesis of CPB-associated AKI.
Ischemia-reperfusion injury of the kidney frequently occurring during
cardiac surgery may be an important factor contributing to postoperative AKI
[56, 57]. Decreased tissue oxygen tension under such circumstances might pro-
mote mitochondrial generation of ROS [58] which, in turn, might cause AKI.
Polymorphisms in Acute Kidney Injury 83
Such ischemia may be further aggravated by over-activation of the angiotensinogen/
angiotensin II pathways or decreased endothelial nitric oxide synthase (eNOS)
with resulting excessive renal vasoconstriction.
CPB has also been shown to stimulate neutrophils and to induce the gener-
ation of ROS and inflammatory mediators [59–62]. Increased levels of serum
lipid peroxidation products and an intra- and postoperatively decreased total
serum antioxidative capacity have also been found following CPB [52, 53].
Apolipoprotein-E (APO-E) is a lipoprotein involved in lipid metabolism, tissue
repair and immune response, which might also affect endothelial repair after
lipid peroxidation. The gene is located on chromosome 19q13.2 [63]. APO-E
genetic polymorphisms have been linked to atherosclerosis [64–66] and neu-
rocognitive dysfunction after cardiac surgery [67] suggesting a role in injury
and repair for this protein during and after cardiac surgery. There is much evi-
dence indicating that the generation of ROS may contribute to the initiation and
maintenance of acute tubular necrosis [68]. Oxidative stress is considered to be
an important cause of AKI in patients exposed to CPB [52–55]. CPB is proin-
flammatory and activates components of the nonspecific immune system. The
inflammatory response to CPB generates cytokines (e.g. TNF-␣ and IL-6), both
systemically and locally and in the kidney, that have major effects on the renal
microcirculation and may lead to tubular injury [69–71].
ROS from ischemia-reperfusion injury and oxido-inflammatory response
to CPB also contribute significantly to the deactivation of catecholamines via
oxidation into their corresponding vaso-inactive degradation products and
thereby to systemic vasodilatation and AKI [72, 73].
Reduced degradation of catecholamines by the enzyme, catechol-O-
methyltransferase (COMT), may result in alternative metabolic pathways, which
goes along with increased formation of chemically reactive intermediates and
enhanced generation of ROS [74, 75].
As a consequence of the above observations, genetic polymorphism of sev-
eral of the proteins involved in these postulated injurious pathways has been
considered for investigation (table 2). We included studies for further review if
the relation of a SNP to CPB-associated AKI was evaluated.
Interleukin-6
IL-6 is a pleiotropic cytokine with both pro- and anti-inflammatory prop-
erties. It is located on the long arm of chromosome 7. A polymorphism has
been identified within the promoter region of the IL-6 gene at position ⫺174
(G to C) [77] and ⫺572 (G to C) [78].
In a prospective study [76] of 111 patients receiving coronary artery
bypass surgery, the ⫺174 G/C polymorphism of the IL-6 gene was determined
and correlated with the postoperative plasma IL-6 levels and the development
Haase-Fielitz/Haase/Bellomo/Dragun 84
Table 2. Previous studies highlighting the association between a genetic polymorphism and CPB-associated acute kidney injury
Gene Study design Patients Definition of AKI Type of Outcome Reference
cardiac
surgery
IL-6 Prospective, n ⫽ 111 Perioperative change Coronary IL-6 (G Gaudino et al.
(⫺174 G/C) observational, in serum creatinine bypass graft homozygous) was [76]
single center study surgery associated with
greater increase in
perioperative
creatinine
12 Poly- Prospective, n ⫽ 1,671 Perioperative relative Coronary IL-6⫺572C, Stafford-Smith
morph- observational, change in serum bypass graft angiotensinogen et al. [48]
isms in 7 single center study creatinine surgery 842C allele, APO-E
candi- 448C 4, AGTR1
date 1166C, eNOS 894T
genes and ACE
deletion/insertion
associated with
AKI
APO-E Prospective, n ⫽ 564 Postoperative Coronary 4 allele was Chew et al.
alleles observational, changes in serum bypass graft associated with a [79]
single center study creatinine surgery smaller increase in
postoperative
serum creatinine
than 2, 3 allele
Polymorphisms in Acute Kidney Injury 85
APO-E 4 Prospective, n ⫽ 130 Peak in-hospital Coronary Interaction between MacKensen
allele observational, postoperative serum bypass graft 4 status (4/non- et al. [80]
single center study creatinine (also surgery 4; n ⫽ 26/106) and
perioperative change atheroma burden
in serum creatinine, with a greater peak
change in serum in-hospital
creatinine clearance) postoperative
serum creatinine for
increases in
ascending aorta
atheroma load for
non-4 patients
ACE ⫽ Angiotensin-converting enzyme; AGTR ⫽ angiotensin receptor1; AKI ⫽ acute kidney injury; APO-E ⫽ apolipoprotein E;
eNOS ⫽ endothelial nitric oxide synthase; IL ⫽ interleukin.
Haase-Fielitz/Haase/Bellomo/Dragun 86
of postoperative renal complications. The investigators found evidence that the
IL-6 ⫺174 G/C polymorphism modulated postoperative IL-6 levels and was
associated with the degree of postoperative kidney and pulmonary dysfunction
and with the duration of in-hospital stay after coronary surgery [76].
IL-6, Angiotensin and NO Pathway Interaction
Stafford-Smith et al. [48] assessed twelve candidate genes for polymor-
phism to test the hypotheses that selected gene variants might be associated with
AKI. Two alleles (IL-6 ⫺572C and angiotensinogen 842C) showed a strong
association with AKI. The combination of these polymorphisms was associated
with major postoperative AKI with an average peak serum creatinine level
increase of 121%, which was four times greater than average for the overall
study population. Additional vasoconstrictor gene polymorphisms were also
identified to account for AKI after cardiac surgery (angiotensin, eNOS, and
angiotensin receptor 1). Finally, genetic polymorphism for APO-E was also
found to be associated with the development of AKI.
Apolipoprotein E
Chew et al. [79] further studied the impact of genetic polymorphism for
APO-E and showed a reduced postoperative increase in serum creatinine and a
lower peak serum creatinine after cardiac surgery in patients with the APO-E 4
allele compared with the APO-E 2 allele and APO-E 3 allele.
The same working group evaluated the atheromatous burden of ascending,
arch, and descending aorta and APO-E status in 130 coronary bypass patients.
They found that an equivalent ascending aortic atheromatous burden is associ-
ated with a greater susceptibility to postoperative AKI among patients under-
going cardiac operation who lack the APO-E 4 allele [80].
The above observations provide consistent and reliable evidence that
genetic factors and, in particular SNP for IL-6, the angiotensin-generating path-
ways, a nitric oxide regulation enzyme and lipid pathways significantly influ-
ence the likelihood of a patient developing AKI after cardiac surgery, and do so
in a clinically important way. These observations, however, do not tackle
another likely important pathway in the regulation of the response to stress, the
production, release and uptake of catecholamines and the generation of ROS:
the COMT system.
Potential Role of COMT Polymorphism in
Sepsis- and CPB-Associated AKI
The activity of this crucial enzyme involved in the degradation of circulating
catecholamines, COMT, is controlled by a common autosomal co-dominantly
inherited SNP [81].
Polymorphisms in Acute Kidney Injury 87
This SNP, located at chromosome 22 (22q11.21-q11.23; 22q11.21), results
in a trimodal distribution of low, intermediate and high levels of COMT activity
[82]. Reduced degradation of catecholamines by COMT, associated with the
COMT LL genotype, may result in alternative metabolic pathways with increased
formation of chemically reactive intermediates and enhanced generation of ROS
[74, 75]. The prevention of ROS generation by catecholamines is closely related
to the activity of COMT and thus, where a high activity prevents their conversion
to semiquinones and quinones and therefore blocks the generation of ROS [83], a
low activity is associated with increased generation of ROS by oxidative path-
ways. The biological and therapeutic efficacy of catecholamines to regulate vaso-
motor tone may be reduced by inactivation through ROS [72].
The formation of ROS by auto-oxidation of catecholamines on the one
hand and the inactivation of catecholamines by ROS on the other hand may
contribute to endothelial damage, microcirculatory dysfunction and vasodilata-
tion [84, 85]. Thus, such SNPs may lead to systemic vasodilatation, may signifi-
cantly aggravate CPB-induced vasoplegia and thereby participate in inducing or
sustaning AKI. Unfortunately, this SNP has not yet been investigated in these
patients or in patients with sepsis-associated AKI. Hopefully, future investiga-
tions will help elucidate the role of COMT-related SNP in the development of
both sepsis-associated AKI and CPB-associated AKI.
Conclusion
There is limited but growing evidence for an important role of genetic poly-
morphism in the pathogenesis of AKI in sepsis and after CPB. This evidence
shows that most genetic polymorphisms accounting for sepsis- or CPB-associ-
ated AKI appear to involve genes which participate in the control of inflamma-
tory or vasomotor processes. Thus, it provides further indirect evidence of the
importance of these pathways in the pathogenesis of these syndromes. Based on
this observation, we raise the question whether individual variability in genetic
polymorphism of the COMT enzyme (another powerful controller of vasomotor
state) might not also predispose to sepsis- or CPB-associated AKI and suggest
that studies targeting its genetic variability should be the next step in the investi-
gation of how such variability affects the likelihood of developing sepsis- and
CPB-associated AKI.
In addition, studies involving genetic polymorphism may further help us to
understand the pathogenesis of sepsis- or CPB-associated AKI; discover poten-
tial markers of susceptibility, severity and clinical outcomes; identify markers
for responders vs. non-responders in therapeutic trials, and recognize targets for
therapeutic intervention.
Haase-Fielitz/Haase/Bellomo/Dragun 88
The use of genetic epidemiology may stratify those who may benefit
from preventive measures from those who will not, depending on the individual
genotype.
Genotyping may prove one day to be a routine tool in an individualized
patient care model, which is clinically useful and cost-effective.
References
1 Hou SH, Bushinsky DA, Wish JB, et al: Hospital-acquired renal insufficiency: a prospective study.
Am J Med 1983;74:243–248.
2 Uchino S, Kellum JA, Bellomo R, et al; Beginning and Ending Supportive Therapy for the Kidney
(BEST Kidney) Investigators: Acute renal failure in critically ill patients: a multinational, multi-
center study. JAMA 2005;294:813–818.
3 Uchino S, Bellomo R, Goldsmith D, et al: An assessment of the RIFLE criteria for acute renal
failure in hospitalized patients. Crit Care Med 2006;34:1913–1917.
4 Chertow GM, Levy EM, Hammermeister KE, et al: Independent association between acute renal
failure and mortality following cardiac surgery. Am J Med 1998;104:343–348.
5 De Mendonca A, Vincent JL, Suter PM, et al: Acute renal failure in the ICU: risk factors and out-
come evaluated by the SOFA score. Intensive Care Med 2000;26:915–921.
6 Zanardo G, Michielon P, Paccagnella A, et al: Acute renal failure in the patient undergoing cardiac
operation. Prevalence, mortality rate, and main risk factors. J Thorac Cardiovasc Surg 1994;107:
1489–1495.
7 Chertow GM, Lazarus JM, Christiansen CL, et al: Preoperative renal risk stratification. Circulation
1997;95:878–884.
8 Ford EB: Genetic polymorphism. Proc R Soc Lond B Biol Sci 1966;164:350–361.
9 Hsu TC, Klatt O: Mammalian chromosomes in vitro. IX. On genetic polymorphism in cell popu-
lations. J Natl Cancer Inst 1958;21:437–473.
10 Klenzak J, Himmelfarb J: Sepsis and the kidney. Crit Care Clin 2005;21:211–222.
11 Rangel-Frausto MS, Pittet D, Costigan M, et al: The natural history of the systemic inflammatory
response syndrome (SIRS). A prospective study. JAMA 1995;273:117–123.
12 Schrier RW, Wang W: Acute renal failure and sepsis. N Engl J Med 2004;351:159–169.
13 Marshall JC: Inflammation, coagulopathy, and the pathogenesis of multiple organ dysfunction
syndrome. Crit Care Med 2001;29(suppl):S99–S106.
14 Wan L, Bellomo R, Di Giantomasso D, et al: The pathogenesis of septic acute renal failure. Curr
Opin Crit Care 2003;9:496–502.
15 Parrillo JE: Pathogenetic mechanisms of septic shock. N Engl J Med 1993;328:1471–1477.
16 Bone RC, Grodzin CJ, Balk RA: Sepsis: a new hypothesis for pathogenesis of the disease process.
Chest 1997;112:235–243.
17 Pinsky MR, Vincent JL, Deviere J, et al: Serum cytokine levels in human septic shock. Relation to
multiple-system organ failure and mortality. Chest 1993;103:565–575.
18 Hack CE, Aarden LA, Thijs LG: Role of cytokines in sepsis. Adv Immunol 1997;66:101–195.
19 Messmer UK, Briner VA, Pfeilschifter J: Basic fibroblast growth factor selectively enhances TNF-
alpha-induced apoptotic cell death in glomerular endothelial cells: effects on apoptotic signaling
pathways. J Am Soc Nephrol 2000;11:2199–2211.
20 Lieberthal W, Triaca V, Levine J: Mechanisms of death induced by cisplatin in proximal tubular
epithelial cells: apoptosis vs. necrosis. Am J Physiol 1996;270:F700–F708.
21 Kohan DE: Role of endothelin and tumour necrosis factor in the renal response to sepsis. Nephrol
Dial Transplant 1994;9(suppl 4):73–77.
22 Karnik AM, Bashir R, Khan FA, et al: Renal involvement in the systemic inflammatory reaction
syndrome. Ren Fail 1998;20:103–116.
Polymorphisms in Acute Kidney Injury 89
23 Knotek M, Rogachev B, Wang W, et al: Endotoxemic renal failure in mice: Role of tumor necrosis
factor independent of inducible nitric oxide synthase. Kidney Int 2001;59:2243–2249.
24 Jaber BL, Pereira BJ, Bonventre JV, et al: Polymorphism of host response genes: implications in
the pathogenesis and treatment of acute renal failure. Kidney Int 2005;67:14–33.
25 Dollner H, Vatten L, Linnebo I, et al: Inflammatory mediators in umbilical plasma from neonates
who develop early-onset sepsis. Biol Neonate 2001;80:41–47.
26 Treszl A, Toth-Heyn P, Kocsis I, et al: Interleukin genetic variants and the risk of renal failure in
infants with infection. Pediatr Nephrol 2002;17:713–717.
27 Vincent JL, de Mendonca A, Cantraine F, et al: Use of the SOFA score to assess the incidence of
organ dysfunction/failure in intensive care units: results of a multicenter, prospective study.
Working group on ‘sepsis-related problems’ of the European Society of Intensive Care Medicine.
Crit Care Med 1998;26:1793–1800.
28 American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference:
Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis.
Crit Care Med 1992;20:864–874.
29 Gordon AC, Lagan AL, Aganna E, et al: TNF and TNFR polymorphisms in severe sepsis and sep-
tic shock: a prospective multicentre study. Genes Immun 2004;5:631–640.
30 Sibbald WJ, Vincent JL: Round table conference on clinical trials for the treatment of sepsis.
Brussels, March 12–14, 1994. Intensive Care Med 1995;21:184–189.
31 Wattanathum A, Manocha S, Groshaus H, et al: Interleukin-10 haplotype associated with increased
mortality in critically ill patients with sepsis from pneumonia but not in patients with extrapul-
monary sepsis. Chest 2005;128:1690–1698.
32 Perianayagam MC, Liangos O, Kolyada AY, et al: NADPH oxidase p22phox and catalase gene
variants are associated with biomarkers of oxidative stress and adverse outcomes in acute renal
failure. J Am Soc Nephrol 2007;18:255–263.
33 Reinhart K, Menges T, Gardlund B, et al: Randomized, placebo-controlled trial of the anti-tumor
necrosis factor antibody fragment afelimomab in hyperinflammatory response during severe sep-
sis: The RAMSES Study. Crit Care Med 2001;29:765–769.
34 Gallagher J, Fisher C, Sherman B, et al: A multicenter, open-label, prospective, randomized, dose-
ranging pharmacokinetic study of the anti-TNF-alpha antibody afelimomab in patients with sepsis
syndrome. Intensive Care Med 2001;27:1169–1178.
35 Baud L, Oudinet JP, Bens M, et al: Production of tumor necrosis factor by rat mesangial cells in
response to bacterial lipopolysaccharide. Kidney Int 1989;35:1111–1118.
36 Pinsky MR: The critically ill patient. Kidney Int Suppl 1998;66:S3–S6.
37 Camussi G, Ronco C, Montrucchio G, et al: Role of soluble mediators in sepsis and renal failure.
Kidney Int Suppl 1998;66:S38–S42.
38 Wilson AG, de Vries N, Pociot F, et al: An allelic polymorphism within the human tumor necrosis
factor alpha promoter region is strongly associated with HLA A1, B8, and DR3 alleles. J Exp Med
1993;177:557–560.
39 Morse HR, Olomolaiye OO, Wood NA, et al: Induced heteroduplex genotyping of TNF-alpha, IL-
1beta, IL-6 and IL-10 polymorphisms associated with transcriptional regulation. Cytokine 1999;11:
789–795.
40 Warzocha K, Ribeiro P, Bienvenu J, et al: Genetic polymorphisms in the tumor necrosis factor
locus influence non-Hodgkin’s lymphoma outcome. Blood 1998;91:3574–3581.
41 Bone RC: Toward a theory regarding the pathogenesis of the systemic inflammatory response
syndrome: what we do and do not know about cytokine regulation. Crit Care Med 1996;24:
163–172.
42 Jaber BL, Pereira BJ: Extracorporeal adsorbent-based strategies in sepsis. Am J Kidney Dis
1997;30(suppl 4):S44–S56.
43 Jaber BL, Rao M, Guo D, et al: Cytokine gene promoter polymorphisms and mortality in acute
renal failure. Cytokine 2004;25:212–219.
44 Fekete A, Treszl A, Toth-Heyn P, et al: Association between heat shock protein 72 gene polymor-
phism and acute renal failure in premature neonates. Pediatr Res 2003;54:452–455.
45 Stallwood MI, Grayson AD, Mills K, et al: Acute renal failure in coronary artery bypass surgery:
independent effect of cardiopulmonary bypass. Ann Thorac Surg 2004;77:968–972.
Haase-Fielitz/Haase/Bellomo/Dragun 90
46 Wijeysundera DN, Rao V, Beattie WS, et al: Evaluating surrogate measures of renal dysfunction
after cardiac surgery. Anesth Analg 2003;96:1265–1273.
47 Lassnigg A, Schmidlin D, Mouhieddine M, et al: Minimal changes of serum creatinine predict
prognosis in patients after cardiothoracic surgery: a prospective cohort study. J Am Soc Nephrol
2004;15:1597–1605.
48 Stafford-Smith M, Podgoreanu M, Swaminathan M, et al: Association of genetic polymorphisms
with risk of renal injury after coronary bypass graft surgery. Am J Kidney Dis 2005;45:519–530.
49 Mangano CM, Diamondstone LS, Ramsay JG, et al: Renal dysfunction after myocardial revascu-
larization: risk factors, adverse outcomes, and hospital resource utilization. The Multicenter Study
of Perioperative Ischemia Research Group. Ann Intern Med 1998;128:194–203.
50 Conlon PJ, Stafford-Smith M, White WD, et al: Acute renal failure following cardiac surgery.
Nephrol Dial Transplant 1999;14:1158–1162.
51 Thakar CV, Arrigain S, Worley S, et al: A clinical score to predict acute renal failure after cardiac
surgery. J Am Soc Nephrol 2005;16:162–168.
52 Starkopf J, Zilmer K, Vihalemm T, et al: Time course of oxidative stress during open-heart surgery.
Scand J Thorac Cardiovasc Surg 1995;29:181–186.
53 McColl AJ, Keeble T, Hadjinikolaou L, et al: Plasma antioxidants: evidence for a protective role
against reactive oxygen species following cardiac surgery. Ann Clin Biochem 1998;35:616–623.
54 Doi K, Suzuki Y, Nakao A, et al: Radical scavenger edaravone developed for clinical use amelio-
rates ischemia/reperfusion injury in rat kidney. Kidney Int 2004;65:1714–1723.
55 McCord JM: Oxygen-derived free radicals in postischemic tissue injury. N Engl J Med 1985;312:
159–163.
56 Andersson LG, Bratteby LE, Ekroth R, et al: Renal function during cardiopulmonary bypass:
influence of pump flow and systemic blood pressure. Eur J Cardiothorac Surg 1994;8:597–602.
57 Abbott WM, Austen WG: The reversal of renal cortical ischemia during aortic occlusion by man-
nitol. J Surg Res 1974;16:482–489.
58 Dada LA, Chandel NS, Ridge KM, et al: Hypoxia-induced endocytosis of Na/K-ATPase in alveo-
lar epithelial cells is mediated by mitochondrial reactive oxygen species and PKC-zeta. J Clin
Invest 2003;111:1057–1064.
59 Chello M, Mastroroberto P, Patti G, et al: Simvastatin attenuates leucocyte-endothelial interac-
tions after coronary revascularisation with cardiopulmonary bypass. Heart 2003;89:538–543.
60 Boyle EM Jr, Lille ST, Allaire E, et al: Endothelial cell injury in cardiovascular surgery: athero-
sclerosis. Ann Thorac Surg 1997;63:885–894.
61 Paparella D, Yau TM, Young E. Cardiopulmonary bypass induced inflammation: pathophysiology
and treatment. An update. Eur J Cardiothorac Surg 2002;21:232–244.
62 Partrick DA, Moore EE, Fullerton DA, et al: Cardiopulmonary bypass renders patients at risk for
multiple organ failure via early neutrophil priming and late neutrophil disability. J Surg Res
1999;86:42–49.
63 Strittmatter WJ, Saunders AM, Schmechel D, et al: Apolipoprotein E: high-avidity binding to beta-
amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc
Natl Acad Sci USA 1993;90:1977–1981.
64 Eichner JE, Kuller LH, Orchard TJ, et al: Relation of apolipoprotein E phenotype to myocardial
infarction and mortality from coronary artery disease. Am J Cardiol 1993;71:160–165.
65 Corder EH, Saunders AM, Strittmatter WJ, et al: Gene dose of apolipoprotein E type 4 allele and
the risk of Alzheimer’s disease in late onset families. Science 1993;261:921–923.
66 Corder EH, Saunders AM, Risch NJ, et al: Protective effect of apolipoprotein E type 2 allele for
late onset Alzheimer disease. Nat Genet 1994;7:180–184.
67 Tardiff BE, Newman MF, Saunders AM, et al: Preliminary report of a genetic basis for cognitive
decline after cardiac operations. The Neurologic Outcome Research Group of the Duke Heart
Center. Ann Thorac Surg 1997;64:715–720.
68 Nath KA, Norby SM: Reactive oxygen species and acute renal failure. Am J Med 2000;109:
665–678.
69 Cunningham PN, Dyanov HM, Park P, et al: Acute renal failure in endotoxemia is caused by TNF
acting directly on TNF receptor-1 in kidney. J Immunol 2002;168:5817–5823.
Polymorphisms in Acute Kidney Injury 91
70 Segerer S, Nelson PJ, Schlondorff D: Chemokines, chemokine receptors, and renal disease: from
basic science to pathophysiologic and therapeutic studies. J Am Soc Nephrol 2000;11:152–176.
71 Heyman SN, Rosen S, Darmon D, et al: Endotoxin-induced renal failure. II. A role for tubular
hypoxic damage. Exp Nephrol 2000;8:275–282.
72 Macarthur H, Westfall TC, Riley DP, et al: Inactivation of catecholamines by superoxide gives new
insights on the pathogenesis of septic shock. Proc Natl Acad Sci USA 2000;97:9753–9758.
73 Bindoli A, Rigobello MP, Deeble DJ: Biochemical and toxicological properties of the oxidation
products of catecholamines. Free Radic Biol Med 1992;13:391–405.
74 Stokes AH, Hastings TG, Vrana KE: Cytotoxic and genotoxic potential of dopamine. J Neurosci
Res 1999;55:659–665.
75 Bolton JL, Trush MA, Penning TM, et al: Role of quinones in toxicology. Chem Res Toxicol 2000;13:
135–160.
76 Gaudino M, Di Castelnuovo A, Zamparelli R, et al: Genetic control of postoperative systemic
inflammatory reaction and pulmonary and renal complications after coronary artery surgery.
J Thorac Cardiovasc Surg 2003;126:1107–1112.
77 Fishman D, Faulds G, Jeffery R, et al: The effect of novel polymorphisms in the interleukin-6 (IL-6)
gene on IL-6 transcription and plasma IL-6 levels, and an association with systemic-onset juvenile
chronic arthritis. J Clin Invest 1998;102:1369–1376.
78 Ferrari SL, Ahn-Luong L, Garnero P, et al: Two promoter polymorphisms regulating interleukin-6
gene expression are associated with circulating levels of C-reactive protein and markers of bone
resorption in postmenopausal women. J Clin Endocrinol Metab 2003;88:255–259.
79 Chew ST, Newman MF, White WD, et al: Preliminary report on the association of apolipoprotein
E polymorphisms, with postoperative peak serum creatinine concentrations in cardiac surgical
patients. Anesthesiology 2000;93:325–331.
80 MacKensen GB, Swaminathan M, Ti LK, et al: Preliminary report on the interaction of apolipopro-
tein E polymorphism with aortic atherosclerosis and acute nephropathy after CABG. Ann Thorac
Surg 2004;78:520–526.
81 Weinshilboum RM, Raymond FA: Inheritance of low erythrocyte catechol-o-methyltransferase
activity in man. Am J Hum Genet 1977;29:125–135.
82 Floderus Y, Wetterberg L: The inheritance of human erythrocyte catechol-O-methyltransferase
activity. Clin Genet 1981;19:392–395.
83 Axelrod J, Tomchick R: Enzymatic O-methylation of epinephrine and other catechols. J Biol
Chem 1958;233:702–705.
84 Kalfin RE, Engelman RM, Rousou JA, et al: Induction of interleukin-8 expression during car-
diopulmonary bypass. Circulation 1993;88:II401–II406.
85 Schmid FX, Vudattu N, Floerchinger B, et al: Endothelial apoptosis and circulating endothelial
cells after bypass grafting with and without cardiopulmonary bypass. Eur J Cardiothorac Surg
2006;29:496–500.
Prof. Rinaldo Bellomo
Department of Intensive Care, Austin Hospital
Melbourne, Vic. 3084 (Australia)
Tel. ⫹61 3 9496 5992, Fax ⫹61 3 9496 3932, E-Mail Rinaldo.Bellomo@austin.org.au
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 92–100
Predictive Capacity of Severity
Scoring Systems in the ICU
N. Schusterschitz, M. Joannidis
Medical Intensive Care Unit, Department of Internal Medicine, Division of General
Internal Medicine, Medical University Innsbruck, Innsbruck, Austria
Abstract
Severity scoring systems were first introduced to intensive care units (ICUs) in 1980.
The basis for their development was the intention to provide information on the prognosis of
patients, the efficacy of therapeutic interventions, stratification for clinical studies, workload
and benchmarking of ICUs. Despite the appearance of several specialized scoring systems,
the general mortality prediction systems such as APACHE, SAPS and MPM scores and their
constantly improved successors have become the most popular and widely tested models.
The newest development in this field is SAPS III which is the first ‘global’ model using a
data set acquired from 307 ICUs from all over the world.
Copyright © 2007 S. Karger AG, Basel
Several factors influence the prognosis of critically ill patients. The main
determinants are the severity of damage to organs and systems and the physio-
logical reserve capacity, mainly determined by age and comorbidities. The use
of scoring systems in the intensive care unit (ICU) aims to standardized classi-
fications of illness severity, survival prediction, judgment of the efficacy of
therapeutic modalities in certain diseases, quantification of nursing require-
ments, cost efficiency calculations, stratification of patients when using certain
therapeutic interventions or scientific studies, and quality control. Standardized
mortality ratios comparing observed hospital mortality with the mortality pre-
dicted by statistical models may be used for benchmarking ICUs.
Generally scoring systems can be divided into disease-oriented scores (e.g.
sepsis, trauma, burns), patient-related scores (e.g. children, surgical, medical
ICU patients), and universally adopted mortality prediction scores (APACHE,
SAPS, MPM).
Severity Scoring in the ICU 93
Two major criteria for the validity of a scoring system are discrimination
and calibration. Discrimination is the ability of the scoring system to differenti-
ate between survivors and non-survivors, which can be judged by a receiver
operating curve (ROC) with an area under the curve (AUC) above 0.9 consid-
ered excellent. Calibration reflects the agreement between individual probabili-
ties and actual outcomes and is usually described by the goodness-of-fit
statistic (i.e. Hosmer-Lemeshow
2
statistic).
Mortality Prediction Scores
Applied Physiology and Chronic Health Evaluation
(APACHE) Score
Prognostic scores were introduced in 1981 with the APACHE score [1]
which was the first physiology-based mortality model developed for the ICU.
The APACHE model focused on seven major physiologic systems: cardiovascu-
lar, renal, gastrointestinal, respiratory, hematologic, metabolic and neurologic.
Depending on the degree of derangement from normal a severity score from
0 to 4 was applied for each variable. APACHE scores were well correlated with
mortality and also performed well within a number of specific cardiovascular,
neurological, respiratory and gastrointestinal diagnoses. Importantly, APACHE
scores showed no major difference between European and US centers.
The APACHE score was followed by APACHE II in 1985 (fig. 1a) [2] and
included 12 physiologic variables (heart rate, mean arterial pressure, tempera-
ture, respiratory rate, PaO
2
/A-a gradient, hematocrit, white blood cell count,
creatinine, sodium, potassium, pH/bicarbonate), the Glasgow Coma Scale, age
and previous health status. All 12 variables were considered mandatory and
had to be collected during the first 24 h of admission. External validity of
APACHE II was investigated by a large number of studies proving good over-
all discrimination of the system across a variety of comorbidities as well as
geography.
In 1991 APACHE III was developed [3] to improve calibration and gener-
alization in specialized populations. APACHE III basically used the variables of
APACHE II, the modified Glasgow Coma Scale and added 5 new physiological
variables (blood urea nitrogen, urine output, serum albumin, bilirubin and glu-
cose). Furthermore items were weighted by narrowing the range of normal or
zero points and increasing the points of extremes of a variable. Although
APACHE III showed good discrimination (AUC between 0.8 and 0.9), this sys-
tem never became as popular as the APACHE II system owing to the fact that
some logistic regression coefficients and equations consisted of proprietary
information not made publicly available.
Schusterschitz/Joannidis 94
Simplified Acute Physiology Score (SAPS)
SAPS was developed by Le Gall et al. [4] in 1984 with the intention to
obtain a simpler model but providing the same level of prediction as APACHE.
They included 13 physiological variables, the most severe value within the first
day of admission was recorded. Discrimination reflected by the obtained AUC
of 0.85 was similar to the APACHE score.
SAPS II was developed in 1992 by the same group and was built from a
much larger data set of 13,152 patients from 137 ICUs in North America and
Europe (fig. 1b) [5]. This model utilized 12 physiologic variables. Furthermore
Fig. 1. Expected mortalities as related to mortality prediction scores. a APACHE II,
modified from Knaus et al. [2]. b SAPS II score, modified from Le Gall et al. [5].
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
0–4 5–10 10–14 15–19 20–24 25–29 30–34 ⱖ35
APACHE II scores
Expected mortality (%)
a
b
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
0–9 10–19 20–29 30–39 40–49 50–59 60–69 70–79 80–89 90–99 ⱖ100
SAPS II scores
Expected mortality (%)
Severity Scoring in the ICU 95
it included age and both type of admission (unscheduled surgical, scheduled
surgical and medical) and underlying disease variables (i.e. AIDS, metastatic
cancer and hematological malignancy). Although the area under the ROC was
0.88 in the training set and 0.86 in the test set, studies in some European coun-
tries showed poor calibration until customization for the local case mix was per-
formed [6].
In 2005 SAPS III, a project of the European Society of Intensive Care
Medicine, was introduced [7, 8]. This score was developed on a database
including 16,784 patients admitted to 303 ICUs from all over the world. All
together 20 variables were selected for SAPS III: ten physiological parameters
(Glasgow Coma Scale, total bilirubin, body temperature, creatinine, heart rate,
leukocyte count, pH, platelets, systolic blood pressure and PaO
2
/FiO
2
); five
comorbidities (cancer therapy, heart failure (NYHA IV), hematological disor-
ders and cancer, cirrhosis and AIDS), and several admission parameters
(length of stay before admission, type of admission [emergency room, other
ICU], use of vasoactive drugs, anatomical site of surgery and type of infection
on admission).
The reason for the introduction of SAPS III was the development of new
major diagnostic and therapeutic methods over the last 10 years which led to
shifts towards poor calibration for older models. Secondly SAPS II was devel-
oped on a database built exclusively from patients in Europe and North
America. SAPS III offers the possibility of referring data to different locations
of the world including Australia, Asia as well as Central and South America.
Data are acquired within the 1st hour of ICU admission. A recent study includ-
ing 952 patients with cancer admitted to a specialized ICU in Brazil proved the
good performance of SAPS III in these patients [9].
Mortality Prediction Model (MPN)
Eleven noninvasive and easily collectable parameters were the basis of the
first version of this model published by Teres et al. [10] in 1987. The model was
developed on a dataset including 755 cases collected in 1983. 107 variables
were collected on the training set and, after stepwise univariate analysis, the set
was reduced to 26 admission variables and 44 variables to be taken at 24 h for
patients staying longer in the IUC. After logistic regression the final admission
model (MPN
0
) contained 7 variables (coma/stupor, emergency admission, can-
cer, infection, number of organ system failures, age and systolic pressure). Also
the final 24-hour model (MPN
24
) had 7 variables (coma/stupor, infection,
mechanical ventilation, shock, emergency admission, age and number of organ
system failures).
This model was followed by the MPM II version [11] which again included
two versions: MPM
0
II was designed to use parameters taken at the time of
Schusterschitz/Joannidis 96
admission to the ICU. The admission model contained 15 variables (coma/
stupor, non-elective surgery, cancer, age, systolic blood pressure, heart rate,
chronic renal insufficiency, cirrhosis, acute renal failure, cardiac dysrhythmia,
cerebrovascular accident, gastrointestinal bleeding, intracranial mass effect,
cardiopulmonary resuscitation before admission, and mechanical ventilation).
The MPN
24
II was developed for patients staying longer than 24 h included 5
variables from MPM
0
II and eight additional parameters which have to be taken
after the first 24 h after admission to the ICU (coma/stupor, infection, mechanical,
ventilation, vasoactive drugs, emergency admission, age, cirrhosis, intracranial
mass effect, creatinine, PaO
2
, prothrombin time and urine output). MPM II
appears to be very robust and can be used reliably in medical patients. Although
the MPN II system has the advantage of very few quickly determined parame-
ters, it is mainly validated for the US.
Therapeutic Intervention Scoring System (TISS)
This system is one of the oldest scoring systems developed for ICUs [12].
Originally it was intended to correlate the number of required therapeutic inter-
ventions with changes in physiological parameters. However, TISS turned out
to be insufficient for mortality prediction but has proven to be a reliable mea-
sure for nursing workload and for calculations of requirements in nursing staff.
Containing 76 items (TISS-76), daily scoring turned out to be time-consuming
and many new therapeutic interventions such as positional therapy in acute res-
piratory distress syndrome were not considered. TISS-28 was developed to pro-
vide an updated and simplified version but is still lacking validation by large
international databases [13].
Special Severity Scoring Systems
Numerous specialized scoring systems have appeared over the last years, a
detailed discussion of those, however, lies outside the scope of this article. Just
to mention a few, scores have been developed for pediatric ICUs like the
Pediatric Risk of Mortality (PRISM) and the Pediatric Index of Mortality
(PIM), scores for trauma patients, e.g. Injury Severity Score (ISS) or for
patients with special disease entities such as liver cirrhosis, e.g. Model for End-
Stage Liver Disease (MELD).
Sequential organ dysfunction scores were originally introduced into the
ICU to describe the individual risk of patients on a day-to-day basis. Because
the severity of organ dysfunction has a significant influence on outcome, it was
not surprising that organ severity scores predict mortality in individual patients
as well. This was shown for Sequential Organ Failure Assessment (SOFA) [14]
as well as Multi-Organ Dysfunction Score (MODS) [15].
Severity Scoring in the ICU 97
Performance of Mortality Prediction Models
Several studies have compared the performance of these models in a vari-
ety of ICU populations. Generally spoken the models show good discrimination
but a lack of calibration across different countries and populations of ICU
patients with special diseases such as malignancies, acute renal failure or liver
cirrhosis. Comparison of second-generation models (APACHE II, SAPS,
MPN) found APACHE II to have equivalent or superior discrimination. When
comparing third-generation models (APACHE III, MPM-II and SAPS II), they
all performed equal to or better than the second-generation models (table 1)
[16–30]. In terms of calibration the trend was poor across all studies with the
only exception of APACHE II.
Limitations of Prediction Models
The development of prediction model is subject to several biases.
First of all, it depends on the parameters collected and then selected for the
model. Whereas in the first-generation models variables were originally chosen
by subjective judgment, later models used multiple logistic regression for selec-
tion of the variable which is more accurate. Secondly, calibration depends on
the kinds of complications or syndromes which are known for ICU populations
and their representation by a variable at the time of development of a scoring
system. For example, when investigating the APACHE score in AIDS patients
an under-prediction of mortality in patients with pneumocystis carinii pneumo-
nia was observed. In all other AIDS patients the prediction was good [31]. It is
of importance to note that the model was developed before this clinical syn-
drome had become common, and inclusion of this diagnostic criterion could
improve the model’s prediction.
Table 1. Comparison of popular mortality prediction scores (second and third generation) in terms of
discrimination (AUC) [16–30]
APACHE MPM
0
MPM
24
SAPS APACHE SAPS MPM MPM SAPS
II III II II
0
II
24
III
AUC (mean) 0.82 0.79 0.82 0.77 0.83 0.81 0.8 0.8 0.85
95% CI (0.79– (0.70– (0.76– (0.64– (0.75– (0.77– (0.72– (0.67–
0.85) 0.87) 0.88) 0.90) 0.90) 0.86) 0.87) 0.93)
Number of 15 4 2 3 6 11 4 4 1
studies
Schusterschitz/Joannidis 98
Furthermore, in some surgical ICU environments with high trauma admis-
sions, the APACHE II model performed badly due to the few trauma patients
included in the data development set.
Special patient groups were not seen in the original test sample (e.g. coro-
nary cardiac surgery in APCHE III).
When comparing standardized mortality ratios in different countries by
APACHE II scores, significant differences were noted. This is due to different
case mixes in these situations.
An additional unresolved problem is the question of appropriate classifi-
cation of sedated patients by the Glasgow Coma Scale. SAPS III tries to solve
this problem by using an estimated Glasgow Coma Scale before the use of
sedation [7, 8].
The question of how to deal with missing values is another problem. Most
scoring systems attribute a zero to a missing value which reduces the actual
score.
It has repeatedly been shown that ICU scores underestimate mortality in
patients with acute renal failure requiring renal replacement therapy [32, 33].
APACHE II suggests a doubling of the measured serum creatinine in case of
acute renal failure. Unfortunately this entity is still ill defined leading to inco-
herent treatment of this variable.
Scoring Systems versus Physician’s Judgment
Making a prognosis about patients’ outcomes is an essential part of medical
decision making. Furthermore, patients and relatives rely on the doctor’s prog-
nostic information. When systemically investigating the prognostic capabilities
of critical care and primary care physicians, significant differences depending
on the level of ICU training were found [34]. However, comparison of physi-
cians’ judgment to objective ICU models showed no significant differences in
the AUCs between the physicians and APACHE II [35]. One interesting study
dividing observed to expected mortality ratios into six severity segments found
that the objective model performed better in all segments but the segment with
the highest mortality [36]. In an investigation by Christensen et al. [37] attend-
ing physicians had the best calibration across all training levels being slightly
superior to the objective model prediction.
Finally, the combination of subjective physicians’ mortality estimations
with a physiologic objective model was found to perform superior to either the
subjective or objective model alone [38].
Severity Scoring in the ICU 99
References
1 Knaus WA, Zimmerman JE, Wagner DP, Draper EA, Lawrence DE: APACHE-acute physiology
and chronic health evaluation: a physiologically based classification system. Crit Care Med
1981;9:591–597.
2 Knaus WA, Draper EA, Wagner DP, Zimmerman JE: APACHE II: a severity of disease classifica-
tion system. Crit Care Med 1985;13:818–829.
3 Knaus WA, Wagner DP, Draper EA, Zimmerman JE, Bergner M, Bastos PG, Sirio CA, Murphy
DJ, Lotring T, Damiano A: The APACHE III prognostic system. Risk prediction of hospital mor-
tality for critically ill hospitalized adults. Chest 1991;100:1619–1636.
4 Le Gall JR, Loirat P, Alperovitch A, Glaser P, Granthil C, Mathieu D, Mercier P, Thomas R, Villers
D: A simplified acute physiology score for ICU patients. Crit Care Med 1984;12:975–977.
5 Le Gall JR, Lemeshow S, Saulnier F: A new Simplified Acute Physiology Score (SAPS II) based
on a European/North American multicenter study. JAMA 1993;270:2957–2963.
6 Le Gall JR, Neumann A, Hemery F, Bleriot JP, Fulgencio JP, Garrigues B, Gouzes C, Lepage E,
Moine P, Villers D: Mortality prediction using SAPS II: an update for French intensive. Crit Care
2005;9:R645–R652.
7 Metnitz PG, Moreno RP, Almeida E, Jordan B, Bauer P, Campos RA, Iapichino G, Edbrooke D,
Capuzzo M, Le Gall JR: SAPS 3 – from evaluation of the patient to evaluation of the intensive.
Intensive Care Med 2005;31:1336–1344.
8 Moreno RP, Metnitz PG, Almeida E, Jordan B, Bauer P, Campos RA, Iapichino G, Edbrooke D,
Capuzzo M, Le Gall JR: SAPS 3 – From evaluation of the patient to evaluation of the intensive.
Intensive Care Med 2005;31:1345–1355.
9 Soares M, Salluh JI: Validation of the SAPS 3 admission prognostic model in patients with cancer
in need of intensive care. Intensive Care Med 2006;32:1839–1844.
10 Teres D, Lemeshow S, Avrunin JS, Pastides H: Validation of the mortality prediction model for
ICU patients. Crit Care Med 1987;15:208–213.
11 Lemeshow S, Teres D, Klar J, Avrunin JS, Gehlbach SH, Rapoport J: Mortality Probability Models
(MPM II) based on an international cohort of intensive care unit patients. JAMA 1993;270:2478–2486.
12 Cullen DJ, Civetta JM, Briggs BA, Ferrara LC: Therapeutic intervention scoring system: a method
for quantitative comparison of patient care. Crit Care Med 1974;2:57–60.
13 Miranda DR, de RA, Schaufeli W: Simplified Therapeutic Intervention Scoring System: the TISS-
28 items – results from a multicenter study. Crit Care Med 1996;24:64–73.
14 Ferreira FL, Bota DP, Bross A, Melot C, Vincent JL: Serial evaluation of the SOFA score to predict
outcome in critically ill patients. JAMA 2001;286:1754–1758.
15 Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ: Multiple organ dysfunc-
tion score: a reliable descriptor of a complex clinical outcome. Crit Care Med 1995;23:1638–1652.
16 Beck DH, Smith GB, Pappachan JV, Millar B: External validation of the SAPS II, APACHE II and
APACHE III prognostic models in South England: a multicentre study. Intensive Care Med
2003;29:249–256.
17 Capuzzo M, Valpondi V, Sgarbi A, Bortolazzi S, Pavoni V, Gilli G, Candini G, Gritti G, Alvisi R:
Validation of severity scoring systems SAPS II and APACHE II in a single-center population.
Intensive Care Med 2000;26:1779–1785.
18 Castella X, Gilabert J, Torner F, Torres C: Mortality prediction models in intensive care: acute
physiology and chronic health evaluation II and mortality prediction model compared. Crit Care
Med 1991;19:191–197.
19 Castella X, Artigas A, Bion J, Kari A: A comparison of severity of illness scoring systems for
intensive care unit patients: results of a multicenter, multinational study. The European/North
American Severity Study Group. Crit Care Med 1995;23:1327–1335.
20 Del BC, Morelli A, Bassein L, Fasano L, Quarta CC, Pacilli AM, Gunella G: Severity scores in
respiratory intensive care: APACHE II predicted mortality better than SAPS II. Respir Care
1995;40:1042–1047.
21 Katsaragakis S, Papadimitropoulos K, Antonakis P, Strergiopoulos S, Konstadoulakis MM,
Androulakis G: Comparison of Acute Physiology and Chronic Health Evaluation II (APACHE II)
Schusterschitz/Joannidis 100
and Simplified Acute Physiology Score II (SAPS II) scoring systems in a single Greek intensive
care unit. Crit Care Med 2000;28:426–432.
22 Livingston BM, MacKirdy FN, Howie JC, Jones R, Norrie JD: Assessment of the performance of five
intensive care scoring models within a large Scottish database. Crit Care Med 2000;28:1820–1827.
23 Markgraf R, Deutschinoff G, Pientka L, Scholten T: Comparison of acute physiology and chronic
health evaluations II and III and simplified acute physiology score II: a prospective cohort study
evaluating these methods to predict outcome in a German interdisciplinary intensive care unit. Crit
Care Med 2000;28:26–33.
24 Moreno R, Apolone G, Miranda DR: Evaluation of the uniformity of fit of general outcome pre-
diction models. Intensive Care Med 1998;24:40–47.
25 Nouira S, Belghith M, Elatrous S, Jaafoura M, Ellouzi M, Boujdaria R, Gahbiche M, Bouchoucha
S, Abroug F: Predictive value of severity scoring systems: comparison of four models in Tunisian
adult intensive care units. Crit Care Med 1998;26:852–859.
26 Patel PA, Grant BJ: Application of mortality prediction systems to individual intensive care units.
Intensive Care Med 1999;25:977–982.
27 Rowan KM, Kerr JH, Major E, McPherson K, Short A, Vessey MP: Intensive Care Society’s Acute
Physiology and Chronic Health Evaluation (APACHE II) study in Britain and Ireland: a prospec-
tive, multicenter, cohort study comparing two methods for predicting outcome for adult intensive
care patients. Crit Care Med 1994;22:1392–1401.
28 Tan IK: APACHE II and SAPS II are poorly calibrated in a Hong Kong intensive care unit. Ann
Acad Med Singapore 1998;27:318–322.
29 Vassar MJ, Lewis FR Jr, Chambers JA, Mullins RJ, O’Brien PE, Weigelt JA, Hoang MT, Holcroft
JW: Prediction of outcome in intensive care unit trauma patients: a multicenter study of Acute
Physiology and Chronic Health Evaluation (APACHE), Trauma and Injury Severity Score
(TRISS), and a 24-hour intensive care unit (ICU) point system. J Trauma 1999;47:324–329.
30 Wilairatana P, Noan NS, Chinprasatsak S, Prodeengam K, Kityaporn D, Looareesuwan S: Scoring
systems for predicting outcomes of critically ill patients in northeastern Thailand. Southeast Asian
J Trop Med Public Health 1995;26:66–72.
31 Smith RL, Levine SM, Lewis ML: Prognosis of patients with AIDS requiring intensive care. Chest
1989;96:857–861.
32 Levy EM, Viscoli CM, Horwitz RI: The effect of acute renal failure on mortality. A cohort analy-
sis. JAMA 1996;275:1489–1494.
33 Metnitz PG, Krenn CG, Steltzer H, Lang T, Ploder J, Lenz K, Le Gall JR, Druml W: Effect of
acute renal failure requiring renal replacement therapy on outcome in critically ill patients. Crit
Care Med 2002;30:2051–2058.
34 Poses RM, Bekes C, Copare FJ, Scott WE: The answer to ‘What are my chances, doctor?’ depends
on whom is asked: prognostic disagreement and inaccuracy for critically ill patients. Crit Care
Med 1989;17:827–833.
35 Kruse JA, Thill-Baharozian MC, Carlson RW: Comparison of clinical assessment with APACHE
II for predicting mortality risk in patients admitted to a medical intensive care unit. JAMA
1988;260:1739–1742.
36 McClish DK, Powell SH: How well can physicians estimate mortality in a medical intensive care
unit? Med Decis Making 1989;9:125–132.
37 Christensen C, Cottrell JJ, Murakami J, Mackesy ME, Fetzer AS, Elstein AS: Forecasting survival
in the medical intensive care unit: a comparison of clinical prognoses with formal estimates.
Methods Inf Med 1993;32:302–308.
38 Marcin JP, Pollack MM, Patel KM, Ruttimann UE: Combining physician’s subjective and physiol-
ogy-based objective mortality risk predictions. Crit Care Med 2000;28:2984–2990.
Prof. Dr. Michael Joannidis
Department of Internal Medicine, Medical Intensive Care Unit
Medical University of Innsbruck, Anichstrasse 35
AT–6020 Innsbruck (Austria)
Tel. ⫹43 512 504 81404, Fax ⫹43 512 504 24199, E-Mail michael.joannidis@i-med.ac.at
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 101–111
Determining the Degree of
Immunodysregulation in Sepsis
Jean-Marc Cavaillon, Minou Adib-Conquy
Unit Cytokines & Inflammation, Institut Pasteur, Paris, France
Abstract
During sepsis, the anti-infectious response is closely linked to an overwhelming inflamma-
tory process. The latter is illustrated by the presence in plasma of numerous inflammatory
cytokines, markers of cellular stress (e.g. high mobility group box-1 protein), complement-
derived compounds (e.g. anaphylatoxin C5a), lipid mediators, and activated coagulation factors.
All mediators contribute in synergy to tissue injury, organ dysfunction, and possibly to lethality.
To dampen this overzealous process, a counter-regulatory loop is initiated. The anti-inflamma-
tory counterpart involves few anti-inflammatory cytokines (e.g. interleukin-10, transforming
growth factor-), numerous neuromediators (e.g. adrenalin, acetylcholine), and some other
factors (e.g. heat shock proteins, ligand of TREM-2, adenosine). These mediators modify the
immune status of circulating leukocytes as illustrated by their decreased cell-surface expression
of HLA-DR or their reduced ex vivo pro-inflammatory cytokine production in response to Toll-
like receptor agonists (e.g. endotoxin, lipoproteins). However, circulating leukocytes remain
responsive to whole bacteria and produce normal or even enhanced levels of anti-inflammatory
cytokines. Thus, the immune dysregulation observed in sepsis corresponds to a reprogramming
of circulating leukocytes.
Copyright © 2007 S. Karger AG, Basel
How to Define Immunodysregulation?
Sepsis is associated with an overwhelming proinflammatory response that
is initially aimed at addressing the infectious process. Concomitantly, an anti-
inflammatory response is initiated to counteract the inflammatory process. As a
consequence, the latter regulatory mechanism has an impact on the immune status
of the patients. A long-term alteration in immune status leads to an increased sen-
sitivity to infection and to tumor development. This is illustrated in patients with
chronic inflammatory disorders treated with drugs to neutralize tumor necrosis
factor (TNF), a key cytokine of innate immunity. These patients have an increased
Cavaillon/Adib-Conquy 102
frequency of certain infections (e.g. tuberculosis) or tumors (e.g. lymphoma). It is
believed that a short-term alteration in the immune status in septic and critically ill
patients further renders them more sensitive to nosocomial infection.
A depressed immune status is illustrated by anergy to skin test antigens,
decreased blood cell counts, and reduced expression of surface markers (e.g.
major histocompatibility complex class II antigen, CXCR2, etc.). In vitro
analysis of circulating leukocytes reveals altered natural killer cell activity,
diminished cellular cytotoxicity, reduced antigen presentation, poor prolifera-
tion in response to mitogens, enhanced lymphocyte apoptosis, and depressed
cytokine production by lymphocytes, neutrophils and monocytes. Accordingly,
the terms anergy, immunodepression, or immunoparalysis have been employed
to qualify the immune status of septic patients. We have recently demonstrated
that these terms are far too excessive, and we proposed that the words ‘leuko-
cyte reprogramming’ better defines the phenomenon [1].
Biomarkers in Plasma
Markers of Immune Dysregulation
Proinflammatory cytokines play a key role in innate immunity against
infection. In contrast, either alone in excess or together in synergy, they con-
tribute to lethality in sepsis (table 1). Most tissues participate in their synthesis
during sepsis. In patients, cytokines are produced in excess and are therefore
detectable in blood where they are normally absent. High levels of plasma
cytokines often correlate with organ dysfunction, length of intensive care stay
and poor outcome. However, the circulating cytokines are merely the tip of the
iceberg, and cell-associated cytokines can be shown even when plasma levels
are undetectable [2]. In addition to cytokines, there are other mediators that
have been shown to contribute to sepsis-related lethality. This is the case of
some lipid mediators (prostaglandins, leukotrienes, platelet-activating factor), a
growth factor (i.e. vascular endothelial growth factor), the unidentified ligand
of the triggering receptor expressed on myeloid cells-1 (TREM-1) and anaphy-
latoxin C5a. High-mobility group box-1 protein is a nuclear protein that is
released in sepsis. It is a late mediator and behaves like a cytokine. Like endo-
toxin, it activates leukocytes through Toll-like receptor 4 (TLR4) and seems to
be the mediator that links the occurrence of apoptosis during sepsis and lethal-
ity. Other factors are indirectly involved in sepsis-related mortality, and favor
the release of proinflammatory cytokines. This is the case with factors gener-
ated during coagulation activation (e.g. thrombin, factor Xa) or necrotic cells.
Finally, some markers found in large amounts in plasma are only a reflection of
the overwhelming process. Many studies have established that procalcitonin is a
Immunodysregulation in Sepsis 103
Table 1. Immune dysregulation during sepsis is characterized by an exacerbated production of proinflam-
matory mediators that lead to deleterious effects and lethality, and an exacerbated production of anti-inflammatory
mediators that contribute to induce an immune suppressive status
Mediators Mediators
that contribute to that favor immune
lethality
a
suppression
Cytokines TNF IL-10
IL-1 IL-13
IL-12 TGF
IL-15
IL-18
IL-27
IFN␥
IFN
Granulocyte-macrophage colony-stimulating factor
Leukemia inhibitory factor
Macrophage migration inhibitory factor
Some chemokines: CXCL8 (IL-8)
CCL5
CXCR1 and 2 ligands
CCR1 ligands
CCR4 ligands
b
Growth factors Vascular endothelial growth factor ⫺
Cell markers of stress High mobility group box-1 protein Heat shock proteins
Plasma factors Ligand of TREM-1
c
Ligand of TREM-2
Anaphylatoxin C5a
Lipid mediators Prostaglandins Prostaglandins
Leukotrienes
Platelet-activating factor
Oxidized phospholipids
Hormones ⫺ Glucocorticoids
Neuromediators ⫺ Adrenalin
Acetylcholine
␣-Melanocyte-stimulating hormone
Vasoactive intestinal peptide (VIP)
Urocortin
Adrenomodulin
Cortistatin
Cavaillon/Adib-Conquy 104
good marker for the occurrence of an infection in critically ill patients.
However, other data have challenged the specificity of plasma procalcitonin
elevation as a marker for infection. Soluble TREM-1 was also described as a
specific marker for infection [3]. More recent studies suggest that in certain
noninfectious clinical settings, levels of soluble TREM-1 are also significantly
enhanced [4]. Thus, before using this new marker to diagnose bacterial infec-
tions, a large cohort of patients with various noninfectious causes of systemic
inflammatory response syndrome (SIRS) should be included to attest to the
specific upregulation of TREM-1 during infection.
Molecules Interacting with Lipopolysaccharide
Host serum contains several proteins that interact with lipopolysaccharide
(LPS). Some of these proteins are present at homeostasis, such as soluble CD14
and LPS-binding protein (LBP), but their levels are considerably increased dur-
ing sepsis. Depending on their concentration, these LPS-binding molecules
may facilitate LPS interaction with TLR4-bearing cells or, on the contrary,
decrease cellular response by transferring cell-bound LPS to plasma lipopro-
teins as shown with soluble CD14 [5] and LBP [6]. High-density lipoprotein
(HDL) and other plasma lipoproteins can bind and neutralize the bioactivity of
gram-negative bacterial LPS [7] and gram-positive bacterial lipoteichoic acid
[8]. During sepsis, circulating levels of HDL decline dramatically. However,
when HDL levels decline in critically ill patients, LPS binds preferentially to
low-density and very low-density lipoproteins that maintain their ability to neu-
tralize endotoxin. However, native HDL may enhance the monocyte response to
Table 1. (continued)
Mediators Mediators
that contribute to that favor immune
lethality
a
suppression
Enzymes Cyclooxygenase-2
5-Lipoxygenase
Phospholipase A2
Mast cell dipeptidyl peptidase I ⫺
Coagulation factor Tissue factor ⫺
Thrombin
Purine nucleoside Adenosine (via A
2A
receptor) Adenosine (via A
2A
receptor)
a
As demonstrated in animal models with the help of specific antibodies, inhibitors or antagonists, or with KO mice.
b
Either CCL17 or CCL22.
c
Triggering receptor expressed on myeloid cells.
Immunodysregulation in Sepsis 105
LPS [9]. This enhancing effect was found in the presence of inhibitory concen-
trations of LBP. Bactericidal/permeability-increasing (BPI) protein is a cationic
protein released by activated or killed neutrophils. This protein shows a high
affinity for the lipid A moiety of LPS, and exerts a neutralizing activity and
causes bacterial lysis. BPI levels are increased in critically ill patients with bac-
teremia, and increased circulating BPI is also associated with mortality in
patients with ventilator-associated pneumonia [10]. MD-2 is a soluble protein
that is associated with TLR4 to form the receptor for LPS. Soluble MD-2 has
been detected in the plasma of patients with severe sepsis or septic shock, and
in lung edema fluids from patients with acute respiratory distress syndrome
[11]. Similarly to soluble CD14, soluble MD-2 may enhance the reactivity of
TLR4-positive epithelial cells towards LPS, whereas it would downregulate the
reactivity of cells positive for both TLR4 and MD-2, such as monocytes.
Immunosuppressive Markers
An anti-inflammatory process occurs concomitantly to dampen the
overzealous anti-infectious response. Anti-inflammatory cytokines and soluble
receptors are produced in large amounts during sepsis. They downregulate the
production of proinflammatory cytokines and protect animals from sepsis and
endotoxin shock. This was shown for interleukin (IL)-10, transforming growth
factor- (TGF), IL-4, IL-13, interferon (IFN)␣, and IL-6. IL-6 induces a
broad array of acute-phase proteins that limit inflammation, such as ␣
1
-acid
glycoprotein or C-reactive protein.
The presence of deactivating or immunosuppressive agents within the
blood stream may contribute to the hyporeactivity of circulating leukocytes.
The fact that ‘septic plasma’ behaves as an immunosuppressive milieu [12] is
illustrated in human volunteers by the capacity of endotoxin to induce plasma
inhibitors [13]. Most interestingly, this suppressive effect in septic patients was
significantly reduced after passage of plasma through a resin and after incuba-
tion with anti-IL-10 antibodies [14]. IL-10 was identified as a major functional
deactivator of monocytes in human septic shock plasma [15]. TGF was also
shown in animal models of hemorrhagic shock or sepsis to be the causative
agent of the depressed splenocyte responsiveness [16]. Monocytes from
immunocompromised trauma patients seem to be a source of TGF [17], and
TGF released by apoptotic T cells contributes to this immunosuppressive
milieu [18]. In addition, there is accumulating evidence for a strong interaction
between components of the nervous and the immune systems. Numerous neu-
romediators behave as immunosuppressors. Interaction of adrenaline with its

2
-adrenergic receptor enhances IL-10 production and decreases TNF produc-
tion in vitro and in vivo in LPS-challenged healthy volunteers. ␣-Melanocyte-
stimulating hormone also contributes to immunosuppression by inducing IL-10
Cavaillon/Adib-Conquy 106
production by human monocytes [19]. In addition, vasoactive intestinal peptide
and pituitary adenylate cyclase-activating polypeptide directly inhibit endotoxin-
induced proinflammatory cytokine secretion [20]. Vagal nerve stimulation
attenuates hypotension and reduces plasma and liver TNF levels through an
interaction between acetylcholine and the ␣
7
subunit of the nicotinic receptor at
the macrophage surface [21]. Sepsis is also associated with an activation of the
hypothalamus-pituitary-adrenal axis which leads to the release of glucocorticoids,
well known for their potent ability to limit cytokine production. Prostaglandins
produced during sepsis also contribute to the downregulation of cytokine
production [22]. Finally, adenosine contributes to alter immune status via occu-
pancy of A
2A
receptor expressed on immune cells [23].
Cellular Markers of Immunodysregulation
Cell Surface Markers on Monocytes
In addition to the quantification of circulating cytokines or other soluble
molecules, immunodysregulation in sepsis may be monitored by analyzing the
expression of some cell surface molecules. A profound decrease in HLA-DR
surface expression on monocytes has been regularly reported in sepsis. Low
HLA-DR expression was associated with an increased risk of secondary bacte-
rial infections [24], probably due to a less potent antigen presentation that would
not allow an efficient adaptive immunity. The downregulation of HLA-DR is at
least partially mediated by the immunosuppressive cytokine IL-10 which was
shown to favor the intracellular sequestration of this major histocompatibility
complex type II molecule in human monocytes [25]. HLA-DR downregulation
was also observed on monocytes of patients with a noninfectious systemic
inflammation, such as after pancreatitis, major surgery or trauma.
TREM-1 is a receptor selectively expressed on monocytes/macrophages
and neutrophils that mediates cell activation, but its ligand is still unknown.
TREM-1 expression is increased on human monocytes upon LPS stimulation
in vitro and blockade of TREM-1 protects mice against endotoxic shock [26].
This molecule may also be present in soluble form. Upon LPS injection into
healthy volunteers or during sepsis, TREM-1 surface expression was increased
on circulating monocytes [27, 28]. An upregulation of TREM-1 has also been
found on patients’ monocytes after major abdominal surgery [29].
Ex Vivo Leukocyte Reactivity
One of the hallmarks of sepsis is the decreased ex vivo production of
proinflammatory cytokines by patients’ monocytes in response to LPS stimula-
tion (table 2). This decreased reactivity was also found in many groups of
Immunodysregulation in Sepsis 107
Table 2. Dysregulation of ex vivo cytokine production by patients’ circulating leukocytes during sepsis and systemic inflammatory
response syndrome
Activating Patient Cell type Cytokine production Reference
agent group studied
TLR4 agonist LPS Trauma PBMC IL-1 ↓, IFN␥ ↓ Faist et al: Arch Surg 1988;123:287
LPS Sepsis Monocyte IL-1 ↓, IL-6 ↓, TNF ↓ Munoz et al: J Clin Invest 1991;88:1747
LPS Sepsis Neutrophil IL-1 ↓ McCall et al: J Clin Invest 1993;91:853
LPS Sepsis Whole blood IL-6 ↓, IL-10 ↓, TNF ↓ Marchant et al: J Clin Immunol 1995;15:266
LPS Sepsis Whole blood IL-1Ra ↑ Van Deuren et al: J infect Dis 1994;169:157
LPS Sepsis Whole blood IL-1 ↓, IL-6 ↓, TNF ↓ Ertel et al: Blood 1995;85:1341
LPS Sepsis Whole blood TNF ↓, IL-6 ↓, IL-10 ↓ Marchant et al: J Clin Immunol 1995;15:266
LPS Sepsis Whole blood IFN-␥ ↓, TNF ↓ Mitov et al: Infection 1997;25:206
LPS Sepsis Neutrophil IL-8 ↓ Marie et al: Blood 1998;91:3439
LPS CPB Neutrophil IL-8 ↓ Marie et al: Blood 1998;91:3439
LPS Trauma Neutrophil TNF ↑, IL-8 ↑, IL-1Ra ↔ Zallen et al: J Trauma 1999;46:42
LPS Sepsis Whole blood G-CSF ↑, TNF ↓ Weiss et al: Cytokine 2001;13:51
LPS CPB Whole blood TNF ↓ Wilhelm et al: Shock 2002;17:354
LPS RCA Whole blood IL-1Ra ↑, IL-6 ↓, TNF ↓ Adrie et al: Circulation 2002;106:562
LPS Trauma Whole blood IL-6 ↓, IL-10 ↓, TNF ↓ Adib-Conquy et al: AJRCCM 2003;168:158
LPS Sepsis PBMC MIF ↑, TNF ↓ Maxime et al: J Infect Dis 2005;191:138
LPS Sepsis Monocyte IL-10 ↑, TNF ↓ Adib-Conquy et al: Crit Care Med 2006;34:2377
TLR2 agonist Pam3CysSK4 Sepsis Monocyte IL-10 ↑, TNF ↓ Adib-Conquy et al: Crit Care Med 2006;34:2377
Pam3CysSK4 RCA Monocyte IL-10 ↔, TNF ↔ Adib-Conquy et al: Crit Care Med 2006;34:2377
LPS from
L. interrogans Trauma Whole blood IL-10 ↑, TNF ↔ Adib-Conquy et al: AJRCCM 2003;168:158
TLR9 agonist CpG Trauma Whole blood IL-6 ↓, IL-10 ↔, TNF ↓ Adib-Conquy et al: AJRCCM 2003;168:158
NOD2 agonist MDP Trauma Monocyte IL-6 ↔ Szabo et al: J Clin Immunol 1991;11:326
Trauma PBMC IL-10 ↓ Miller-Graziano et al: J Clin Immunol
1995;15:93
Cavaillon/Adib-Conquy 108
Sepsis Monocyte TNF ↔ Adib-Conquy et al: Crit Care Med 2006;34:2377
RCA Monocyte TNF ↔ Adib-Conquy et al: Crit Care Med 2006;34:2377
G⫺ bacteria S. typhimurium Sepsis Whole blood IL-6 ↑,TNF ↑ Mitov et al: Infection 1997;25:206
E. coli Sepsis Whole blood IL-6 ↓, IL-10 ↓, TNF ↓ Haupt et al: J Invest Surg 1997;10:349
E. coli Trauma Whole blood IL-10 ↑, TNF ↓ Adib-Conquy et al: AJRCCM 2003;168:158
P. aeruginosa Sepsis Whole blood IL-10 ↔,TNF ↓ Rigato et al: Shock 2003;19:113
G⫹ bacteria S. aureus Sepsis Whole blood IL-6 ↑,TNF ↑ Mitov et al: Infection 1997;25:206
S. aureus Sepsis PBMC IL-10 ↔ Muret et al: Shock 2000;13:169
S. aureus Sepsis Whole blood TNF ↓ Wihlem et al: Shock 2002;17:354
S. aureus RCA Whole blood IL-1Ra ↑, IL-6 ↔, TNF ↔ Adrie et al: Circulation 2002;106:562
S. aureus CPB Whole blood TNF ↓ Wilhelm et al: Shock 2002;17:354
S. aureus Trauma Whole blood IL-6 ↑, IL-10 ↑, TNF ↔ Adib-Conquy et al: AJRCCM 2003;168:158
S. aureus Sepsis Monocyte IL-10 ↔, TNF ↔ Adib-Conquy et al: Crit Care Med 2006;34:2377
S. aureus RCA Monocyte IL-10 ↔, TNF ↔ Adib-Conquy et al: Crit Care Med 2006;34:2377
S. pyogenes CPB Neutrophil IL-8 ↓ Marie et al: Blood 1998;91:3439
S. pyogenes Trauma Whole blood IL-6 ↔, IL-10 ↑, TNF ↔ Adib-Conquy et al: AJRCCM 2003;168:158
Cytokine production is indicated as compared to healthy controls: ↑ ⫽ increased production; ↓ ⫽ decreased production; ↔ ⫽ similar
production.
CPB ⫽ Cardiopulmonary bypass; G⫺⫽gram-negative; G⫹⫽gram-positive; LPS ⫽ lipopolysaccharide; PBMC ⫽ peripheral blood
mononuclear cells; RCA ⫽ resuscitated cardiac arrest.
Table 2. (continued)
Activating Patient Cell type Cytokine production Reference
agent group studied
Immunodysregulation in Sepsis 109
patients having noninfectious SIRS (e.g. trauma, major surgery, resuscitation
after cardiac arrest; table 2). As observed with LPS, monocytes from septic
patients produced significantly less TNF ex vivo in response to a TLR2 agonist.
In contrast, we were able to show that this defect did not occur in monocytes
from patients with noninfectious SIRS: the production of TNF was comparable
to that found in healthy controls. The dysregulation in ex vivo cytokine produc-
tion was very different when anti-inflammatory cytokines were analyzed.
Indeed, in contrast to proinflammatory cytokines, increased production of IL-10
and/or IL-1Ra was found in monocytes of septic and SIRS patients stimulated
by TLR4 or TLR2 agonists. Finally, patients’ leukocytes were responsive to
muramyl dipeptide, the minimal motif of bacterial peptidoglycan, and produced
cytokine levels comparable to those obtained in healthy volunteers.
The response to whole bacteria may represent a more relevant pathophysi-
ological approach to monitor the immune status of septic patients. Surprisingly,
in contrast to highly specific TLR2 or TLR4 agonists, the production of
cytokines by patients’leukocytes gives a more contrasted profile. Cytokine pro-
duction by septic and SIRS patients in response to heat-killed gram-positive or
gram-negative bacteria was usually undiminished or even increased when com-
pared to that obtained with healthy donors (table 2). Thus, the immunodysregula-
tion found during sepsis or SIRS is not a generalized hyporeactivity. Depending
upon the cytokine and the stimulus, there may be a decreased, a maintained or
an increased response from monocytes.
In conclusion, immune dysregulation during sepsis may be monitored by
assessing: (i) circulating cytokines, inflammatory mediators and soluble mem-
brane compounds; (ii) expression of cell surface markers, and (iii) ex vivo mono-
cyte reactivity to LPS. However, one should keep in mind that these modifications
are most often similar in patients having sepsis or a non-infectious severe
systemic inflammation.
References
1 Cavaillon JM, Adrie C, Fitting C, Adib-Conquy M: Reprogramming of circulatory cells in sepsis
and SIRS. J Endotoxin Res 2005;11:311–320.
2 Cavaillon JM, Muñoz C, Fitting C, Misset B, Carlet J: Circulating cytokines: the tip of the iceberg?
Circ Shock 1992;38:145–152.
3 Gibot S, Kolopp-Sarda MN, Bene MC, Cravoisy A, Levy B, Faure GC, Bollaert PE: Plasma level
of a triggering receptor expressed on myeloid cells-1: its diagnostic accuracy in patients with sus-
pected sepsis. Ann Intern Med 2004;141:9–15.
4 Adib-Conquy M, Monchi M, Goulenok C, Laurent I, Thuong M, Cavaillon JM, Adrie C:
Increased plasma levels of soluble triggering receptor expressed on myeloid cells-1 and procalci-
tonin after cardiac surgery and cardiac arrest without infection, Shock 2007; in press.
5 Kitchens RL, Thompson PA, Viriyakosol S, O’Keefe GE, Munford RS: Plasma CD14 decreases
monocyte responses to LPS by transferring cell-bound LPS to plasma lipoproteins. J Clin Invest
2001;108:485–493.
Cavaillon/Adib-Conquy 110
6 Wurfel MM, Kunitake ST, Lichenstein H, Kane JP, Wright SD: Lipopolysaccharide (LPS)-binding
protein is carried on lipoproteins and acts as a cofactor in the neutralization of LPS. J Exp Med
1994;180:1025–1035.
7 Cavaillon J, Annane D: Compartmentalization of the inflammatory response in sepsis and SIRS.
J Endotoxin Res 2006;12:151–170.
8 Grunfeld C, Marshall M, Shigenaga JK, Moser AH, Tobias P, Feingold KR: Lipoproteins inhibit
macrophage activation by lipoteichoic acid. J Lipid Res 1999;40:245–252.
9 Thompson PA, Kitchens RL: Native high-density lipoprotein augments monocyte responses to
lipopolysaccharide (LPS) by suppressing the inhibitory activity of LPS-binding protein.
J Immunol 2006;177:4880–4887.
10 Froon AH, Bonten MJ, Gaillard CA, Greve JW, Dentener MA, de Leeuw PW, Drent M,
Stobberingh EE, Buurman WA: Prediction of clinical severity and outcome of ventilator-associ-
ated pneumonia. Comparison of simplified acute physiology score with systemic inflammatory
mediators. Am J Respir Crit Care Med 1998;158:1026–1031.
11 Pugin J, Stern-Voeffray S, Daubeuf B, Matthay MA, Elson G, Dunn-Siegrist I: Soluble MD-2
activity in plasma from patients with severe sepsis and septic shock. Blood 2004;104:4071–4079.
12 Cavaillon J-M: ‘Septic plasma’: an immunosuppressive milieu. Am J Respir Crit Care Med 2002;
166:1417–1418.
13 Spinas G, Bloesch D, Kaufmann M, Keller U, Dayer JM: Induction of plasma inhibitors of inter-
leukin 1 and TNF-alpha activity by endotoxin administration to normal humans. Am J Physiol
1990;259:R993–R997.
14 Ronco C, Brendolan A, Lonnemann G, Bellomo R, Piccinni P, Digito A, Dan M, Irone M, La
Greca G, Inguaggiato P, Maggiore U, De Nitti C, Wratten ML, Ricci Z, Tetta C: A pilot study of
coupled plasma filtration with adsorption in septic shock. Crit Care Med 2002;30:1250–1255.
15 Brandtzaeg P, Osnes L, Øvstebø R, Joø GB, Westwik AB, Kierulf P: Net inflammatory capacity of
human septic shock plasma evaluated by a monocyte-based target cell assay: identification of
interleukin-10 as a major functional deactivator of human monocytes. J Exp Med 1996;184:
51–60.
16 Ayala A, Knotts JB, Ertel W, Perrin MM, Morrison MH, Chaudry IH: Role of interleukin 6 and
transforming growth factor-beta in the induction of depressed splenocyte responses following sep-
sis. Arch Surg 1993;128:89–94.
17 Miller-Graziano CL, Szabo G, Griffey K, Mehta B, Kodys K, Catalano D: Role of elevated mono-
cyte transforming growth factor  production in post-trauma immunosuppression. J Clin Immunol
1991;11:95–102.
18 Chen W, Frank M, Jin W, Wahl S: TGF-beta released by apoptotic T cells contributes to an
immunosuppressive milieu. Immunity 2001;14:715–725.
19 Luger TA, Kalden DH, Scholzen TE, Brzoska T: Alpha-melanocyte-stimulating hormone as a
mediator of tolerance induction. Pathobiology 1999;67:318–321.
20 Delgado M, Pozo D, Martinez C, Leceta J, Calvo JR, Ganea D, Gomariz RP: Vasoactive intestinal
peptide and pituitary adenylate cyclase-activating polypeptide inhibit endotoxin-induced TNF-
alpha production by macrophages : in vitro and in vivo studies. J Immunol 1999;162:2358–2367.
21 Wang H, Yu M, Ochani M, Amella CA, Tanovic M, Susarla S, Li JH, Wang H, Yang H, Ulloa L,
Al-Abed Y, Czura CJ, Tracey KJ: Nicotinic acetylcholine receptor alpha7 subunit is an essential
regulator of inflammation. Nature 2003;421:384–388.
22 Choudhry MA, Ahmad S, Ahmed Z, Sayeed MM: Prostaglandin E2 down-regulation of T cell IL-2
production is independent of IL-10 during Gram negative sepsis. Immunol Lett 1999;67:125–130.
23 Hasko G, Cronstein BN: Adenosine: an endogenous regulator of innate immunity. Trends
Immunol 2004;25:33–39.
24 van den Berk JMM, Oldenburger RHJ, van den Berg AP, Klompmaker IJ, Mesander G, van Son
WJ, van der Bij W, Slooff MJH, The TH: Low HLA DR expression on monocytes as a prognostic
marker for bacterial sepsis after liver transplantation. Transplantation 1997;63:1846–1848.
25 Fumeaux T, Pugin J: Role of interleukin-10 in the intracellular sequestration of human leukocyte
antigen-DR in monocytes during septic shock. Am J Respir Crit Care Med 2002;166:1475–1482.
26 Bouchon A, Facchetti F, Weigand MA, Colonna M: TREM-1 amplifies inflammation and is a cru-
cial mediator of septic shock. Nature 2001;410:1103–1107.
Immunodysregulation in Sepsis 111
27 Knapp S, Gibot S, de Vos A, Versteeg HH, Colonna M, van der Poll T: Cutting edge: expression
patterns of surface and soluble triggering receptor expressed on myeloid cells-1 in human endo-
toxemia. J Immunol 2004;173:7131–7134.
28 Gibot S, Le Renard PE, Bollaert PE, Kolopp-Sarda MN, Bene MC, Faure GC, Levy B: Surface
triggering receptor expressed on myeloid cells 1 expression patterns in septic shock. Intensive
Care Med 2005;31:594–597.
29 Gonzalez-Roldan N, Ferat-Osorio E, Aduna-Vicente R, Wong-Baeza I, Esquivel-Callejas N,
Astudillo-de la Vega H, Sanchez-Fernandez P, Arriaga-Pizano L, Villasis-Keever MA, Lopez-Macias
C, Isibasi A: Expression of triggering receptor on myeloid cell 1 and histocompatibility complex
molecules in sepsis and major abdominal surgery. World J Gastroenterol 2005;11:7473–7479.
Jean-Marc Cavaillon
Unit Cytokines & Inflammation, Institut Pasteur
28 rue Dr. Roux
FR–75015 Paris (France)
Tel. ⫹33 1 45 68 82 38, Fax ⫹33 1 40 61 35 92, E-Mail jmcavail@pasteur.fr
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 112–118
Nutritional Management in Acute Illness
and Acute Kidney Insufficiency
Xavier M. Leverve
a
, Noël J.M. Cano
b
a
INSERM U884 ‘Bioénergétique Fondamentale et Appliquée’,
Université Joseph Fourier, Grenoble, and
b
Centre Hospitalier Privé Résidence
du Parc, Marseille, France
Abstract
There are now powerful compensatory therapies to counteract kidney deficiency and
the prognosis of patients with acute renal failure is mainly related to the severity of the initial
disease. Renal failure is accompanied by an increase in both severity and duration of the
catabolic phase leading to stronger catabolic consequences. The specificity of the metabolic
and nutritional disorders in the most severely ill patients is the consequence of three additive
phenomena: (1) the metabolic response to stress and to organ dysfunction, (2) the lack of
normal kidney function and (3) the interference with the renal treatment (hemodialysis,
hemofiltration or both, continuous or intermittent, lactate or bicarbonate buffer, etc.). As in
many other diseases of similar severity, adequate nutritional support in acutely ill patients
with ARF is of great interest in clinical practice, although the real improvement as a result of
this support is still difficult to assess in terms of morbidity or mortality.
Copyright © 2007 S. Karger AG, Basel
Introduction: Kidney and Whole Body Metabolism
The two kidneys exhibit a high metabolic activity, accounting for 10% of
resting energy expenditure while representing only 0.5% of body mass.
Kidneys are involved in glucose metabolism by their implication in gluconeo-
genesis and their role in hormone degradation, mainly insulin and glucagon [1].
Renal glucose release is of the same order of magnitude as liver gluconeogene-
sis: during the postabsorptive phase, liver glycogenolysis, hepatosplanchnic
gluconeogenesis and renal gluconeogenesis, respectively, account for 50, 30
and 20% of whole body glucose release (10–11 mol/kg body mass/min). In
humans, lactate is the main substrate for renal gluconeogenesis, followed by
Metabolism, Electrolytes and Acid-Base Disorders
Nutrition and Acute Kidney Insufficiency I 113
glutamine, and glycerol, while alanine is only a poor substrate for renal glucose
synthesis [2]. The kidney is responsible for 50% of whole body lactate gluco-
neogenesis, therefore it plays a major role in the Cori cycle. Renal glucose syn-
thesis seems more sensitive to hormone action than hepatic gluconeogenesis
and has a prominent role during adaptation to various physiological and patho-
logical conditions such as hypoglycemia [3].
The kidney plays a key role in amino acid metabolism and in low molecu-
lar weight protein breakdown. It takes up glutamine, proline, citrulline and
phenylalanine and significantly releases serine, arginine, tyrosine, taurine, thre-
onine and lysine [4]. Moreover, it plays a crucial role in the control of acid-base
regulation by finely tuning the elimination of anions versus cations, thanks to
the generation and the elimination of ammonium ions. Glutamine metabolism
in proximal tubules produces ammonium and ␣-ketoglutarate. Acidosis
increases ammonium production and excretion, and gluconeogenesis from
␣-ketoglutarate.
Pathophysiology
Carbohydrate metabolism during renal failure as well as in stress condi-
tions is characterized by insulin resistance, which predominates in peripheral
tissues and mainly concerns nonoxidative glucose metabolism [5]. Insulin
resistance is attributed to the decrease in glycoregulatory-peptide degradation
by the kidney, uremic toxins and acidosis [3]. In addition to insulin resistance,
renal failure is also responsible for hypoglycemic manifestations, which may
reflect the relative inaptitude of the liver to ensure euglycemia in various severe
clinical situations [3]. Hyperinsulinemia together with unadapted glucagon and
adrenaline responses to hypoglycemia may explain these manifestations [6].
Indirect calorimetry measurements showed similar resting energy expenditure
in control subjects and during real failure and hemodialysis [7]. In spite of
abnormalities of plasma lipid transport and clearance, fat is preferentially oxi-
dized after an overnight fast [7], reflecting an accelerated starvation metabo-
lism. Hence, lipids remain a predominant substrate for oxidation in acute renal
failure (ARF) as is shown by the low respiratory quotient observed in these
patients. Acute patients are also characterized by a low T
3
syndrome and a
decrease in testosterone, HGH and IGF levels. In summary, ARF does not much
affect specifically the hormonal pattern in comparison with the effect of stress,
except for insulin level and insulin resistance since this hormone is degraded
mainly in the kidney.
Kidney amino acid exchanges are markedly altered in renal failure [4].
Studies of forearm metabolism during renal failure have shown increased
Leverve/Cano 114
protein synthesis, proteolysis and protein turnover without change in net prote-
olysis [8]. The effect of acidosis on muscle has been extensively studied [9]:
acidosis induces branched-chain AA catabolism and activates the ATP-ubiqui-
tin-dependent cytosolic proteolytic system. Cortisol secretion is activated by
acidosis and further stimulates branched-chain AA oxidation and ATP-ubiqui-
tin-dependent proteolysis [10]. Thus, acidosis induces muscle proteolysis and
increases renal ammonium excretion, muscle proteolysis being controlled by
acid-base balance.
Effect of Renal Replacement Therapy
The metabolic response to acute illness and renal failure can also be
affected by some specific alterations related to the replacement therapy. A cas-
cade of events following the contact between the patient’s blood and dialysis
membrane could be responsible for cytokine and protease activations increas-
ing energy expenditure and protein catabolism [11]. However more biocompat-
ible membranes are often used at present.
Glucose loss during hemodialysis or hemofiltration is similar (between 25
and 50 g) but is higher with hemodiafiltration. The use of 5% dextrose may
deliver very large amounts of glucose to the patient when the hemofiltration
rate is high (0.5 l/h results in 4,800 cal/day) [12]. Since lipids are circulating
only as lipoprotein or are bound to albumin (fatty acids) losses are negligible.
Amino acids are lost with hemodialysis and hemofiltration proportionally
to their plasma concentration [13]. When amino acids are infused for nutritional
supply, hemofiltration results in about a 10% loss of the total amount infused.
Dialysis promotes a net protein catabolism and induces a reduction of protein
synthesis. Four to 9 g of free amino acids and 2–3 g of peptide-bound amino
acids are removed during each session. The frequent reuse of filters exacerbates
the amino acid and albumin loss.
Nutritional Support
Acutely ill patients are highly catabolic and the benefit of adequate nutri-
tional therapy seems important even if it has so far not been possible to show a
clear relationship. Limitations of the use of large volumes may be achieved with
intermittent dialysis, and continuous hemofiltration or hemodiafiltration facili-
tates fluid removal and nutritional intakes [14].
The use of the enteral route is recommended whenever possible. Hyper-
metabolic patients should be fed early and a risk of renal failure should not
Nutrition and Acute Kidney Insufficiency I 115
delay the initiation of nutritional therapy. Conversely, as for the parenteral
route, nutrition is indicated only after the acute phase of shock and severe meta-
bolic disorders has been managed. A polymeric diet (1kcal/ml) is recom-
mended in the majority of clinical situations. In some cases, according to the
severity of the disease (acute pancreatitis, gut dysfunction, very high nutritional
requirements) more specific diets could be indicated. The addition of glutamine
has been proposed since this amino acid plays an important role in the meta-
bolic and immune responses to stress and infection [15].
In hypercatabolic patients with ARF, relatively low levels of energy sup-
ply (26 kcal/kg/day) have been associated with a better nitrogen balance
compared to higher supplies (35kcal/kg/day) [16, 17]. High-energy intake
(40 kcal/kg/day) has been proposed in patients with high urea nitrogen appear-
ance and very negative nitrogen balance. WHO equations for assessing energy
requirement give better estimates than Harris-Benedict equations, and it is
important to use the patients’ estimated dry weight as ARF patients are often
hyperhydrated or have overt edemas. When elevated, heat loss during dialysis
should also be taken into consideration in calculating energy requirement.
Insulin resistance, reduced glucose tolerance and increased gluconeogenesis,
caused by acute uremia itself or acidosis, require a careful monitoring of blood
glucose levels by insulin administration in order to avoid hyperglycemic
episodes.
In acute and chronic kidney failure, there is a decreased ability to utilize
exogenous lipids, and the use of medium-chain triglycerides has not been
shown to offer any advantages over long-chain triglycerides [18]. This experi-
mental finding, together with the frequent occurrence of hypertriglyceridemia,
makes it advisable to limit the percentage of lipids to 20–25% of the total
energy and to monitor triglyceridemia during treatment. However, lipid intake
is important since lipids, besides being a concentrated, low osmolarity source of
energy, are carriers of essential fatty acids and the use of fish oil could be pro-
posed [19].
Essential as well as nonessential (histidine, arginine, tyrosine, serine, and
cysteine) amino acids become indispensable in ARF, while others, such as
phenylalanine and methionine, may accumulate [20]. The use of a mixture of
essential amino acids alone must be avoided, because imbalances and severe
clinical consequences have been described [21]. Protein requirements in these
patients range from 1.0 to 1.5 g/kg/day depending on the severity of catabolism.
There is no evidence that increasing the protein intake further results in a better
nitrogen balance. However, higher protein/amino acid intake (up to
1.5–2.0 g/kg/day) in more severe ARF patients treated with CVVH, CVVHD,
CVVHDF has been advocated. Replacement therapy also produces a consider-
able loss of amino acids and/or protein with the dialysate, especially with high
Leverve/Cano 116
flux dialyzers [16, 17]. This loss should be integrated by artificial nutrition, so
an additional amount of protein or amino acids (0.2 g/kg/day) is recommended.
The nutritional requirement in patients with kidney insufficiency is summa-
rized in table 1 [16, 17].
Trace elements are excreted mainly by the kidney and parenteral adminis-
tration to ARF patients requires great care. However, zinc, manganese, copper,
selenium and chromium can also be effectively eliminated in the gastroenteric
tract. Vitamin A should probably be avoided because of the possibility of accu-
mulation, whereas vitamin C should not exceed 30–50 mg/day, since inappro-
priate supplementation may result in secondary oxalosis. Even though vitamin
D from body storage may provide lasting protection against deficiency, its
Table 1. Expert group recommendations for nutritional treatment of adult patients with
renal insufficiency [16, 17]
Clinical condition Nonprotein energy Protein
kcal/kg/day
a
g/kg/day
Hemodialysis ⱖ35 1.2–1.4
Peritoneal dialysis ⱖ35 (glucose absorption 1.2–1.5 (⬎50% HBV)
from dialysate can
account for 25–30% of
energy needs)
Nephrotic syndrome ⱖ35 0.8–1.0
Acute renal failure
Nonoliguric, in most patients: ⫽1.3 0.55–1.0
nonhypercatabolic BEE
patients
Hypercatabolic, dialyzed ⱖ1.3 BEE 1.0–1.5 (or more, see
patients text); NEAA ⫹ EAA
(ratio ⬎1:1)
Renal transplantation
Preoperative period correction of correction of
malnutrition malnutrition
Early postoperative 30–35 1.3–1.5
period
Late postoperative period adapted to maintain 1.0
an ideal body weight
BEE ⫽ Basal energy expenditure; EAA ⫽ essential amino acids; HBV ⫽ high biological
value; NEAA ⫽ nonessential amino acids.
a
Adapted to individual needs in case of underweight or obesity.
Nutrition and Acute Kidney Insufficiency I 117
active renal metabolite (1,25-OH-cholecalciferol) can be rapidly depleted, mak-
ing repletion necessary. Vitamin K, E, B
6
and folate requirements are also
increased in ARF [22].
References
1 Cano N: Bench-to-bedside review: glucose production from the kidney. Crit Care 2002;6: 317–321.
2 Gerich JE, Meyer C, Woerle HJ, Stumvoll M: Renal gluconeogenesis: its importance in human
glucose homeostasis. Diabetes Care 2001;24:382–391.
3 Adrogue HJ: Glucose homeostasis and the kidney. Kidney Int 1992;42:1266–1282.
4 Tizianello A, De Ferrari G, Garibotto G, Gurreri G, Robaudo C: Renal metabolism of amino acids
and ammonia in subjects with normal renal function and in patients with chronic renal insuffi-
ciency. J Clin Invest 1980;65:1162–1173.
5 Castellino P, Solini A, Luzi L, Barr JG, Smith DJ, Petrides A, Giordano M, Carroll C, DeFronzo
RA: Glucose and amino acid metabolism in chronic renal failure: effect of insulin and amino
acids. Am J Physiol 1992;262:F168–176.
6 Castellino P, Bia M, DeFronzo RA: Metabolic response to exercise in dialysis patients. Kidney Int
1987;32:877–883.
7 Schneeweiss B, Graninger W, Stockenhuber F, Druml W, Ferenci P, Eichinger S, Grimm G,
Laggner AN, Lenz K: Energy metabolism in acute and chronic renal failure. Am J Clin Nutr
1990;52:596–601.
8 Garibotto G, Russo R, Sofia A, Sala MR, Robaudo C, Moscatelli P, Deferrari G, and Tizianello A:
Skeletal muscle protein synthesis and degradation in patients with chronic renal failure. Kidney
Int 1994;45:1432–1439.
9 Mitch WE, Price SR, May RC, Jurkovitz C, England BK: Metabolic consequences of uremia:
extending the concept of adaptive responses to protein metabolism. Am J Kidney Dis 1994;23:
224–228.
10 Price SR, England BK, Bailey JL, Van Vreede K, Mitch WE: Acidosis and glucocorticoids con-
comitantly increase ubiquitin and proteasome subunit mRNAs in rat muscle. Am J Physiol
1994;267:C955–C960.
11 Gutierrez A, Alvestrand A, Wahren J, Bergstrom J: Effect of in vivo contact between blood and
dialysis membranes on protein catabolism in humans. Kidney Int 1990;38:487–494.
12 Frankenfield DC, Reynolds HN, Badellino MM, Wiles CE 3rd: Glucose dynamics during contin-
uous hemodiafiltration and total parenteral nutrition. Intensive Care Med 1995;21:1016–1022.
13 Davenport A, Roberts NB: Amino acid losses during haemofiltration. Blood Purif 1989;7:
192–196.
14 Bellomo R, Parkin G, Love J, Boyce N: Use of continuous haemodiafiltration: an approach to the
management of acute renal failure in the critically ill. Am J Nephrol 1992;12:240–245.
15 Novak I, Sramek V, Pittrova H, Rusavy P, Lacigova S, Eiselt M, Kohoutkova L, Vesela E, Opatrny
K Jr: Glutamine and other amino acid losses during continuous venovenous hemodiafiltration.
Artif Organs 1997;21:359–363.
16 Toigo G, Aparicio M, Attman PO, Cano N, Cianciaruso B, Engel B, Fouque D, Heidland A, Teplan
V, Wanner C: Expert working group report on nutrition in adult patients with renal insufficiency
(Part 2 of 2). Clin Nutr 2000;19:281–291.
17 Toigo G, Aparicio M, Attman PO, Cano N, Cianciaruso B, Engel B, Fouque D, Heidland A, Teplan
V, Wanner C: Expert working group report on nutrition in adult patients with renal insufficiency
(Part 1 of 2). Clin Nutr 2000;19:197–207.
18 Druml W, Fischer M, Sertl S, Schneeweiss B, Lenz K, Widhalm K: Fat elimination in acute renal
failure: long-chain vs medium-chain triglycerides. Am J Clin Nutr 1992;55:468–472.
19 Neumayer HH, Heinrich M, Schmissas M, Haller H, Wagner K, Luft FC: Amelioration of
ischemic acute renal failure by dietary fish oil administration in conscious dogs. J Am Soc
Nephrol 1992;3:1312–1320.
Leverve/Cano 118
20 Kopple JD: The nutrition management of the patient with acute renal failure. JPEN J Parenter
Enteral Nutr 1996;20:3–12.
21 Nakasaki H, Katayama T, Yokoyama S, Tajima T, Mitomi T, Tsuda M, Suga T, Fujii K:
Complication of parenteral nutrition composed of essential amino acids and histidine in adults
with renal failure. JPEN J Parenter Enteral Nutr 1993;17:86–90.
22 Gilmour ER, Hartley GH, Goodship TH: Trace elements and vitamins in renal disease; in Mitch
WE, Klahr S (eds): Nutrition and the Kidney. Boston, Little, Brown, 1993, pp 114–131.
Pr. Xavier Leverve
Bioénergétique Fondamentale et Appliquée INSERM U884
Université Joseph Fourier, BP 53X
FR–38041 Grenoble Cedex (France)
Tel. ⫹33 476 51 43 86, Fax ⫹33 476 51 42 18, E-Mail Xavier.Leverve@ujf-grenoble.fr
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 119–132
Fundamentals of Oxygen Delivery
James Yassin, Mervyn Singer
Bloomsbury Institute of Intensive Care Medicine, University College London,
London, UK
Abstract
Oxygen is vital to life. A series of steps are needed to transport oxygen from the lungs
to the mitochondrion where the bulk of it is used for generation of energy. Understanding this
pathway, which still remains to be properly characterized, will greatly aid both diagnosis and
management of the hypoxic patient.
Copyright © 2007 S. Karger AG, Basel
What Is Oxygen and Why Do We Need It?
In 1774 Joseph Priestley discovered oxygen whilst acting as librarian for
the Earl of Sherbourne. He performed a series of experiments with mercury and
noticed that mercuric oxide, when heated, produced a gas that made a candle
burn much faster, and that supported respiration. He called this gas ‘dephlogis-
ticated air’ [1]. The name oxygen was coined by the French chemist Antoine
Lavoisier, from the Greek ‘acid producer’ as it was mistakenly felt at the time
that all acids contained oxygen [2].
Oxygen (O
2
) is utilized by most eukaryotic cells to produce energy, in the
form of adenosine triphosphate (ATP), by the process of aerobic metabolism
that occurs in mitochondria. There is some debate as to why this happens. In
1883 Andreas Schimper [3] put forward the theory of endosymbiosis, later
refined by Margulis [4], whereby some 3–4 billion years ago, blue-green algae
respired anaerobically and produced oxygen as a by-product of photosynthesis.
To deal with this toxic molecule, new bacteria (probably Rickettsia spp.) devel-
oped aerobic metabolism. Over time, these bacteria were engulfed by anaerobic
cells, the two becoming symbiotic. The host cell produced nutrients for the
ingested aerobic cell that, in turn, produced energy for the host as a mitochondrion.
Yassin/Singer 120
Conclusive proof is obviously lacking, though this theory is strongly supported
by the composition of mitochondrial membranes closely resembling that of
bacteria [4], by both mitochondrial and bacterial DNA being circular, and con-
taining a higher-than-expected percentage of guanine-cytosine [4].
Cellular utilization of O
2
is the last step in a long chain of reactions that
occur in the cell cytoplasm and within the mitochondrion. The aim of aerobic
metabolism is to enhance energy generation from available substrate. This is
then utilized by all cellular functions, from housekeeping duties of membrane
stabilization and maintenance of the resting membrane potential, to the highly
specialized functions that different organs provide, such as contractile function
of cardiac myocytes or the many complex functions of the renal tubules. Failure
of aerobic metabolism, either due to insufficient delivery of O
2
or reduced cel-
lular utilization, results in a greater dependence on anaerobic metabolism (gly-
colysis) in the cellular cytoplasm. While glycolytic activity can upregulate to a
certain degree, this has the disadvantage of a greatly reduced energy generation
per unit of substrate. Enhanced glycolysis will also increase production of
by-products, such as hydrogen ions (H
), carbon dioxide (CO
2
) and lactate.
Though traditionally perceived as negative, there are adaptive roles served by
these by-products that may assist cell function and integrity in critical illness,
for example, a right shift of the oxyhaemoglobin dissociation curve, alternative
substrate provision through lactate, and vasodilatation through activation of tar-
geted ion channels.
Aerobic and Anaerobic Metabolism
Figure 1 describes energy generation from carbohydrate metabolism under
aerobic conditions. Per molecule of glucose, a net two moles of ATP are pro-
duced by glycolysis, a further two are formed in the Krebs cycle and, at
cytochrome oxidase, the last step of the electron transport chain where O
2
is
first introduced, a further 28 or so molecules of ATP are produced. This contri-
bution emphasizes the importance of oxidative phosphorylation toward energy
generation in most of the body’s cells, being responsible for 95% of ATP pro-
duction under healthy, resting conditions. Notable exceptions include the ery-
throcyte, which contains no mitochondria, and neutrophils which have a much
larger glycolytic component.
Figure 2 shows the production of by-products that occur when the cell
respires anaerobically. The lack of oxygen stimulates production of molecules
that reduce the conversion of pyruvate to acetyl-CoA by pyruvate dehydroge-
nase [5]; under these conditions there is a relative increase in production of lac-
tate, CO
2
and H
ions for less ATP.
Fundamentals of Oxygen Delivery 121
Fig. 1. Energy generation from carbohydrate metabolism.
Acetyl-CoA
Electron transport chain
28 ATP
NADH
FADH
2
Cytosol
Mitochondrion
PDH
O
2
Krebs
cycle
2 ATP
Glucose
Pyruvate
Lactate
2 ATP
Fig. 2. Byproducts of glucose metabolism.
NADH
FADH
2
Cytosol
Mitochondrion
PDH
Krebs
cycle
2 ATP
Glucose
Pyruvate
Lactate
2 ATP
CO
2
H
CO
2
HIF-1
Acetyl-CoA
Yassin/Singer 122
Lactate
A reduced O
2
delivery often but not always results in lactic acidosis.
Conversely, lactic acidosis, a common phenomenon in the critically ill, may
occur in the absence of tissue hypoxia. At neutral pH, lactic acid is almost
totally dissociated into lactate and H
. Under normal conditions, the hydrogen
ion is used in oxidative phosphorylation to produce ATP. Thus, when stressed,
the body does not become acidotic unless oxidative phosphorylation is
impaired, for example during sepsis [6]. As can be seen in figure 2, more lactate
is produced under anaerobic compared to aerobic conditions.
Traditionally, hyperlactataemia has been divided into type A and type B.
Tissue hypoxia causes type A, while the causes of type B are non-hypoxic and
include biguanide therapy and also renal replacement therapy, which may use
lactate as a buffer for the dialysate. In this situation, the serum lactate concentra-
tion may be expected to increase, though it is not accompanied by an acidosis,
and then stabilizes at a new and higher level provided liver function is normal.
The ratio of lactate to pyruvate can be measured but is not performed
clinically as pyruvate is very unstable, and quickly breaks down. The lactate:
pyruvate ratio in health is approximately 10:1. Under hypoxic conditions, as
pyruvate dehydrogenase is inhibited, pyruvate is converted into lactate and so
the ratio may be expected to rise to around 40:1 [7].
Oxygen Delivery
Oxygen delivery (DO
2
) can be considered as occurring in a number of
stages: (1) passage of O
2
to the alveolus, (2) transfer of O
2
from the alveolus to
the red cell, (3) movement from alveolar capillary to the pulmonary vein, (4) trans-
port of O
2
to the tissues and (5) cellular uptake of O
2
.
Figure 3 shows the ‘oxygen cascade’. O
2
tension falls progressively from
inspired air to the level of the mitochondria (approximately 0.1–1 kPa), with
shunt and diffusion being represented within the lungs.
Alveolar Oxygen
The partial pressure of O
2
in the alveolus (P
A
O
2
) is influenced by a number
of factors. At sea level, PO
2
is approximately 21 kPa, that is to say 21% of
atmospheric pressure (P
atm
). Under normal physiological conditions, a tidal
breath passes to the carina and the air is warmed and humidified. As a result of
the addition of saturated vapour pressure (6 kPa), the PO
2
drops to [21
(100–6)] 19.7 kPa. This is hardly significant at sea level, but becomes very
important if breathing rarefied air at the summit of Mount Everest. Up to the
Fundamentals of Oxygen Delivery 123
8th generation of conducting airways, O
2
travels by convection, and by the 12th
generation this bulk flow decreases with diffusion becoming increasingly
important [8].
Under conditions of constant O
2
utilization and CO
2
production, P
A
O
2
is
influenced primarily by P
A
CO
2
, as the partial pressure of the other constituents
of air (mostly N
2
) is relatively stable. An increase in P
A
CO
2
as a result of
reduced alveolar ventilation (V
A
) (fig. 4), a relative increase in physiological
dead space, or increased CO
2
production, results in a corresponding decrease in
P
A
O
2
(fig. 5). The relationship between O
2
and CO
2
is thus linear at physiological
gas tensions. The simplified alveolar gas equation (equation 1) allows predic-
tion of P
A
O
2
if PaCO
2
is known and its near match for P
A
CO
2
assumed.
Fig. 3. The oxygen cascade from air to mitochondrion.
PO
2
(kPa)
Atmosphere
Alveolus
Arterial
Shunt
Diffusion
Tissue
21
13
7
5
1
Pulmonary
capillary
Capillary
Mitochondria
Fig. 4. The relationship between alveolar
ventilation and alveolar PCO
2
.
V
A
P
A
CO
2
Yassin/Singer 124
(1)
where P
A
O
2
is the alveolar PO
2
, P
i
O
2
is the partial pressure of inspired O
2
(allowing for vapour pressure), PaCO
2
is the arterial PCO
2
and R is the respira-
tory quotient (usually held to be 0.8). Substituting in some normal values
(equation 2) we can see that the P
A
O
2
is usually 13 kPa.
(2)
From the alveolar gas equation the importance of PaCO
2
becomes clear. The
biggest influence on P
A
CO
2
is alveolar minute volume, i.e. the volume of gas
reaching the alveolus in 1 min. Thus, respiratory depression, usually either from
drugs or a neurological cause, leads to an increase in CO
2
and thus a decrease in
P
A
O
2
.
From Alveolus to Red Cell
The passage of O
2
from alveolus to erythrocyte occurs in two stages:
firstly, diffusion of the gas across the basement membrane into the capillary
and, secondly, binding of O
2
to haemoglobin.
Oxygen is not a particularly soluble gas compared to CO
2
so its transfer to
capillary blood is aided by the small distance it has to travel. It only has to tra-
verse the alveolar basement membrane and capillary endothelium (around
0.6 m) [9]. The speed of diffusion of oxygen across the lung is dependent on
oxygen’s molecular weight and solubility, as well as the area available for diffusion,
P O [(100 6) 0.21] (5.3/0.8) 13.1 kPa
A2
PO PO
PaCO
R
A2 i2
2
Fig. 5. The relationship between alveolar
PO
2
and PCO
2
.
P
i
CO
2
P
A
CO
2
P
A
O
2
P
A
CO
2
Fundamentals of Oxygen Delivery 125
the distance it must travel and the partial pressure difference across its path.
This is described by Fick’s law of diffusion (equation 3):
(3)
where V
oxygen
is the rate of diffusion, A is the area of the lung, T is the thickness
of the alveolar-capillary boundary, D is the diffusion constant (dependent on
molecular weight and solubility), and P
1
– P
2
is the partial pressure difference
across the membrane.
Under normal resting physiological circumstances, haemoglobin is fully
bound by the time it is a third of the way down the alveolar capillary [10]. Thus,
haemoglobin binding is perfusion limited. If the cardiac output is increased, for
example during exercise, full binding will be delayed due to the increased blood
velocity, but will still occur before the end of the capillary. In disease, anything
that increases the difficulty of diffusion, such as pulmonary oedema, lung fibro-
sis or basement membrane disease, will result in an inability to load haemoglo-
bin by the time it leaves the capillary. This is known as diffusion limitation.
Oxygen binds to the four haem moieties within the haemoglobin molecule
in a homotropic fashion. As each molecule binds, the affinity of haemoglobin
for oxygen increases [11]. This gives the characteristic shape of the oxygen dis-
sociation curve (ODC) (fig. 6). The binding affinity of haemoglobin for O
2
is
also heterotropic, and is influenced by H
concentration, temperature, 2,3-
diphosphoglycerate concentration, and CO
2
[12]. The presence of H
and CO
2
V
A
T
D(P P)
oxygen 1 2
∝ ii
Fig. 6. The oxyhaemoglobin dissociation curve.
PO
2
(kPa)
75
50
100
0
25
14204610812
Haemoglobin saturation (%)
P
50
3.5 kPa
Yassin/Singer 126
shifts the ODC to the right and so facilitates O
2
release (Bohr effect) [13],
whilst binding of O
2
leads to the release of CO
2
(Haldane effect) [14].
As well as binding to haemoglobin, a small amount of O
2
also dissolves in
blood. At 1 atm, even when given supplementary O
2
, this dissolved gas is
insignificant when compared with the amount carried on haemoglobin.
From Capillary to Pulmonary Vein
With blood leaving the alveolar capillary is fully saturated with oxygen, it
passes to the pulmonary vein and thence to the left atrium. Along this path it is
mixed with blood that has not been exposed to an alveolus and thus the overall
saturation of blood in the left atrium is lower than would be expected. This
desaturated blood is the result of physiological and pathological shunt (true
shunt) and ventilation/perfusion mismatch (effective shunt). The desaturated
blood is called venous admixture, and is the theoretical volume of desaturated
blood that would need to be added to fully saturated arterial blood in order to
produce the observed arterial SaO
2
.
Physiological shunt allows desaturated venous blood into the systemic cir-
culation via normal vascular anatomy. In the coronary circulation, venous blood
drains directly into the left ventricle via the thebesian veins. Also, blood that
perfuses the bronchial tree drains directly into the pulmonary veins. This shunt
is usually a small fraction of cardiac output, and thus reduces the arterial PO
2
still further.
Pathological shunt may be caused by pulmonary arteriovenous malforma-
tions, or right to left cardiac septal defects. Again, this will depress the arterial
PO
2
, but importantly for both physiological and pathological shunts, neither can
be improved by the administration of 100% O
2
, as nowhere in its path will the
blood be exposed to the oxygen, and the increased P
i
O
2
will have little effect on
the already fully saturated blood passing through the alveoli.
Under normal physiological conditions in the standing subject, both the
ventilation and perfusion of the lung are approximately matched. The result is
that alveoli with a good blood supply are well ventilated, and vice versa. Both
ventilation and perfusion are greater in the dependent parts of the lung.
This effect is due, in part, to the action of gravity on the weight of the lung tis-
sue itself and the column of blood [15]. Anything that alters the ratio of ventila-
tion and perfusion may result in unsaturated blood reaching the systemic
circulation.
Transfer of O
2
to the Tissues
The amount of oxygen delivered to the tissues is a product of the oxygen
content per unit volume of blood and the cardiac output. The O
2
content (CaO
2
)
depends on the amount of haemoglobin, and how well saturated that haemoglobin
Fundamentals of Oxygen Delivery 127
is. There is also a small addition for the dissolved O
2
as discussed earlier. This
is described in equation 4:
(4)
where CaO
2
is the arterial O
2
content in ml/100 ml blood, Hb is the haemoglo-
bin concentration in g/dl, SaO
2
are the arterial haemoglobin saturation, 1.39 is
Huffner’s constant, and refers to the amount of O
2
in milliliters 1 g of haemo-
globin can hold, and P
a
O
2
is the arterial PO
2
in kPa.
For example, a healthy subject breathing room air at 1 atm, saturating at
97% with a haemoglobin concentration of 15 g/dl, we can see that:
Therefore every litre of blood carries 200 ml of O
2
. If this is then
multiplied by cardiac output, the global DO
2
can be calculated (equation 5).
(5)
1,000 ml O
2
is ejected from the left ventricle every minute. This represents the
global DO
2
, but does not indicate the delivery to individual vascular beds, i.e.
splanchnic, renal, or cerebral. Each of these beds can regulate their own blood
flow [16–18], and, with the exception of the coronary system, can alter the
amount of O
2
removed from that regional circulation [19].
Cellular Uptake of O
2
The PO
2
within mitochondria is extremely low, being in the order of
0.5–2.7 kPa [20]. This varies between cell types and is intimately related to
cytosolic PO
2
. In order to enter the cell, the O
2
molecules must first dissociate
from the erythrocyte haemoglobin. This process is aided by the presence of CO
2
and H
ions in the capillary, as described above [21]. Once released, O
2
can
then diffuse down its ‘tension gradient’ into the cell, and thus be available to the
mitochondria.
With respect to the capillary, these gradients are both longitudinal and
radial, and were first described by Krogh [22] and Erlangen with the ‘Krogh
cylinder model’ in 1919 (fig. 7).
The exact site of O
2
transfer into the tissue is the subject of current debate
and research. Investigators have measured a large drop in PO
2
across the termi-
nal arteriole, with less change down the capillary than might be expected and,
finally, an increase in PO
2
in the postcapillary venule [23]. The reasons for this
DO CaO cardiac output 200 5 1,000 ml/min
22
CaO (15 0.97 1.39) (13 0.02) 20.2 0.26 20.5 ml/100 ml
2
CaO (Hb SaO 1.39) (PaO 0.02)
22 2
Yassin/Singer 128
are unclear, but suggestions include protection of the capillary bed from hyper-
oxia (and thus increased free radical production) by a ‘countercurrent’ transfer
of O
2
from the arteriole and venule [24]. The countercurrent transport of CO
2
also allows the capillary pH to remain low, and so increase the efficiency of O
2
offloading from haemoglobin for a given PO
2
[25]. This process may be
affected by anything that prevents O
2
unloading from the red cell, increased
distance from the capillary to the mitochondria, or any inability of the mito-
chondria to use the O
2
for aerobic respiration.
Hypoxia
Hypoxia is defined as the point at which aerobic respiration ceases to con-
tinue, and further metabolism continues anaerobically. This may occur with a
‘normal’ PaO
2
. Hypoxaemia is defined as a low arterial oxygen tension, using
an arbitrary cut-off, of 8 kPa in room air. The two terms are not interchangeable,
and here we shall be discussing hypoxia and its causes.
Using the system outlined above, the causes of hypoxia can be categorized
into: hypoxic hypoxia, anaemic hypoxia, stagnant hypoxia, and cytopathic
dysoxia.
Hypoxic hypoxia is the result of either a failure of gas exchange, and/or an
increase in venous admixture. The causes of these are many, and include
chronic obstructive pulmonary disease, collapse/consolidation, lung fibrosis
and intracardiac shunts. This may respond to administration of oxygen provided
that shunt is not a predominant feature.
Fig. 7. The Krogh cylinder model of oxygen diffusion.
Arteriole
(PO
2
13 kPa)
Venule
(PO
2
⬇ 5 kPa)
Capillary
Tissue
cylinder
Fundamentals of Oxygen Delivery 129
Anaemic hypoxia is self-explanatory and may be treated by an infusion of
red cells. The haemoglobin concentration at which oxygen delivery is insuffi-
cient depends on oxygen requirements. In healthy volunteers a concentration
of 3 g/dl [26] may be tolerated. However, in the critically unwell population, a
target of 7 g/dl has been shown to be sufficient for most patients [27].
However, uncertainty remains about the optimal value in patients with car-
diorespiratory failure. An outcome study investigating early, goal-directed
resuscitation of patients with early sepsis targeted a haematocrit value of 30%
[28].
Stagnant hypoxia refers to the patient with an inadequate cardiac index.
This may be ventricular filling-related, in which case either blood or another
intravenous fluid should be administered. Low cardiac output due to poor ven-
tricular function (either left or right) may respond to inotropes, but a cause
should be identified and, if possible, treated, for example myocardial infarction
or cardiac tamponade.
Cytopathic dysoxia is the last and most recently discovered cause of
hypoxia. This is the result of an inability of the electron transport chain within
the mitochondria to utilize available O
2
from within the cytoplasm of the cell.
A common cause of this is sepsis [29, 30], where the electron transport chain
is inhibited by nitric oxide [6]. Other causes include cyanide or carbon monox-
ide poisoning, where the cyanide or CO molecule binds to the ferric ion of
mitochondrial cytochrome oxidase and stops its ability to respire aerobically
[31].
Table 1 gives a summary of the common causes of hypoxia.
Table 1. Common causes of hypoxia
Causes of hypoxia
↓
O
2
content
↓ P
A
O
2
↓ P
i
O
2
, ↓ MV
↓ diffusion pulmonary oedema, fibrosis
↓ binding to Hb abnormal Hb, alkalosis
↑ admixture ↑ venous admixture
Anaemia
↓ cardiac output hypovolaemia, heart failure
↓ cellular utilization sepsis, drugs, cyanide poisoning,
carbon monoxide poisoning
P
A
O
2
Alveolar PO
2
; P
i
O
2
inspired PO
2
; MV minute
volume; Hb haemoglobin.
Yassin/Singer 130
Oxygen Supply Dependence
Oxygen consumption (VO
2
) is determined by mitochondrial O
2
require-
ments. Under normal conditions there is a relationship between DO
2
and VO
2
(fig. 8, solid line), such that an increase in O
2
requirement results in an increase
in oxygen extraction, thus maintaining aerobic metabolism independent of DO
2
[32]. If DO
2
decreases, a level will be reached where VO
2
cannot be maintained
by increasing extraction alone. This is known as the critical DO
2
(DO
2 CRIT
)
[33]. Different organs under varying conditions of stress will have a differing
ability to extract oxygen, and so figure 8 represents a global picture. Critical ill-
ness changes the relationship between DO
2
and VO
2
(fig. 10, dashed line) such
that the tissue may continue to utilize O
2
as its delivery increases, so called sup-
ply dependency [34].
Conclusion
A series of steps are needed to transport oxygen from the atmosphere to
the mitochondrion where the bulk of it is used for generation of energy.
Understanding this pathway greatly aids diagnosis and management of the
hypoxic patient. Cellular uptake of O
2
is as yet incompletely characterized and
further research will elucidate the physiological control mechanisms and
processes that occur during pathological states.
Fig. 8. The stylized relationship between
oxygen delivery and consumption.
Relationship in health; relation-
ship in sepsis.
DO
2
CRIT
VO
2
DO
2
Fundamentals of Oxygen Delivery 131
References
1 Priestley J: Experiments and Observations on Different Airs. London, Royal Society, 1775.
2 Lavoisier A: Considérations Générales sur la Nature des Acides. Acides, 1778.
3 Schimper AFW: Über die Entwicklung der Chlorophyllkörner und Farbkörper. Bot Ztg
1883;41:105–162.
4 Margulis L: Symbiosis in cell evolution. Freeman, San Francisco, 1981, pp 206–227.
5 Kim JW, Tchernyshyov I, Semenza GL, Dang CV: HIF-1-mediated expression of pyruvate dehy-
drogenase kinase: a metabolic switch required for cellular adaptation to hypoxia. Cell Metab
2006;3:150–151.
6 Brealey D, Brand M, Hargreaves I, Heales S, Land J, Smolenski R, Davies NA, Cooper CE, Singer
M: Association between mitochondrial dysfunction and severity and outcome of septic shock.
Lancet 2002;360:219–223.
7 Levy B, Sadoune LO, Gelot AM, et al. Evolution of lactate/pyruvate and arterial ketone body
ratios in the early course of catecholamine-treated septic shock. Crit Care Med 2000;28:114–119.
8 Wilson TA, Lin K: Convection and diffusion in the airways and the design of the bronchial tree; in
Bouhuys A (ed): Airway Dynamics, Physiology and Pharmacology. Springfield, Thomas, 1970,
pp 5–20.
9 Hogan J, Smith P, Heath D, Harris P: The thickness of the alveolar capillary wall in the human lung
at high and low altitude. Br J Dis Chest 1986;80:13–18.
10 West JB: Respiratory Physiology – the Essentials. Baltimore, Williams & Wilkins, 2000, p 23.
11 Monod J, Wyman J, Changeux J-P: On the nature of allosteric transitions: a plausible model. J Mol
Biol 1965;12:88–118.
12 Imai K: Allosteric Effects in Haemoglobin. Cambridge, Cambridge University Press, 1982.
13 Bohr C, Hasselbalch KA, Krogh A: Über einen in biologischer Beziehung wichtigen Einfluss, den
die Kohlensäurespannung des Blutes auf dessen Sauerstoffbindung übt. Skand Arch Physiol
1904;16:402–412.
14 Christiansen J, Douglas CG, Haldane JS: The absorption and dissociation of carbon dioxide by
human blood. J Physiol (Lond) 1914;48:244–271.
15 West JB, Dollery CT: Distribution of blood flow and ventilation-perfusion ratio in the lung, mea-
sured with radioactive CO
2
. J Appl Physiol 1960;15:405–410.
16 Miller DJ, Stanek A, Langfitt TW: Concepts of cerebral perfusion pressure and vascular compres-
sion during intracranial hypertension. Prog Brain Res 1972;31:411–432.
17 Navar LG: Renal autoregulation: perspectives from whole kidney and single nephron studies. Am
J Physiol Renal Physiol 1978;234:F357–F370.
18 Granger DN, Richardson PDI, Kvietys PR, Mortillaro NA: Intestinal blood flow. Gastroenterology
1980;78:837–863.
19 Allela A, Williams F, Bolene-Williams C, Katz L: Interrelation between cardiac oxygen consump-
tion and coronary flow. Am J Physiol 1955;183:570–582.
20 Nunn JF: Oxygen; in Nunn JF (ed): Nunn’s Applied Respiratory Physiology. Oxford, Butterworth-
Heinmann, 1993, pp 247–305.
21 Pittman RN, Dulling BR: Effects of altered carbon dioxide tension on haemoglobin oxygenation
in the hamster cheek pouch microvessels. Microvasc Res 1977;13:211–224.
22 Krogh A: The number and the distribution of capillaries in muscle with the calculation of the
oxygen pressure necessary for supplying tissue. J Physiol (Lond) 1919;52:409–515.
23 Kerger H, Torres Filho IP, Rivas M, Winslow RM, Intaglietta M: Systemic and subcutaneous
oxygen tension in conscious Syrian golden hamsters. Am J Physiol Heart Circ Physiol 1995;267:
H802–H810.
24 Kobayashi H, Takizawa N: Oxygen saturation and pH changes in cremaster microvessels of the
rat. Am J Physiol Heart Circ Physiol 1996;270:H1453–H1461.
25 Wiebel ER: The Pathway for Oxygen. Cambridge, Harvard University Press, 1984.
26 Fontana JL, Welborn L, Mongan PD, Sturm P, Martin G, Bünger R: Oxygen consumption and car-
diovascular function in children during profound intraoperative normovolemic hemodilution.
Anesth Analg 1995;80:219–225.
Yassin/Singer 132
27 Hébert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, Tweeddale M,
Schweitzer I, Yetisir E: A multimember, randomized, controlled clinical trial of transfusion
requirements in critical care. N Engl J Med 1999;340:409–417.
28 Rivers E, Bryant MPH, Havstad S: Early goal directed therapy in the treatment of severe sepsis
and septic shock. N Engl J Med 2001;345:1368–1377.
29 Astiz M, Rackow EC, Weil MH, Schumer W: Early impairment of oxidative metabolism and
energy production in severe sepsis. Circ Shock 1988;26:311–320.
30 Boekstegers P, Weidenhofer S, Kapsner T, Werden K: Skeletal muscle partial pressure of oxygen in
patients with sepsis. Crit Care Med 1994;22:640–650.
31 Kerns W, Isom G, Kirk MA: Cyanide and hydrogen sulfide; in Goldfrank LR, Flomenbaum NE,
Lewin NA, et al (eds): Toxicologic Emergencies, ed 7. New York, McGraw-Hill, 2002, pp 1493–1513.
32 Leach RM, Treacher DF: The pulmonary physician in critical care 2: oxygen delivery and con-
sumption in the critically ill. Thorax 2002;57:170–177.
33 Schumacher PT, Cain SM: The concept of a critical DO
2
. Intensive Care Med 1987;13:223.
34 Bihari D, Smithies M, Gimson A, et al: The effect of vasodilatation with prostacyclin on oxygen
delivery and uptake in critically ill patients. N Engl J Med 1987;317:397–403.
Prof. Mervyn Singer
Bloomsbury Institute of Intensive Care Medicine, University College London
Cruciform Building
Gower St., London WC1E 6BT (UK)
Tel. 44 207 679 6714, Fax 44 207 679 6952, E-Mail m.singer@ucl.ac.uk
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 133–157
Principals of Hemodynamic Monitoring
Patricio M. Polanco
a
, Michael R. Pinsky
b
a
Division of Trauma, Department of Surgery and
b
Department of Critical Care
Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pa., USA
Abstract
Background/Aims: Hemodynamic monitoring is the cornerstone of patient manage-
ment in the intensive care unit. However, to be used effectively its applications and limita-
tions need to be defined and its values applied within the context of proven therapeutic
approaches. Methods: Review of the physiological basis for monitoring and a review of the
literature on its utility in altering patient outcomes. Results: Most forms of monitoring are
used to prevent cardiovascular deterioration or restore cardiovascular wellness. However, lit-
tle data support the generalized use of aggressive resuscitation protocols in all but the most
acutely ill prior to the onset of organ injury. Outcomes improve with aggressive resuscitation
in some patients presenting with early severe sepsis and in postoperative high-risk surgical
patients. Conclusions: Monitoring should be targeted to meet the specific needs of the
patient and should not be applied in a broad fashion and whenever possible it should be used
as part of a treatment protocol of proven efficacy.
Copyright © 2007 S. Karger AG, Basel
Hemodynamic monitoring is a cornerstone in the care of the critically ill;
high acuity units, such as the emergency department, intensive care unit (ICU)
and operating room and acute treatment centers, such as dialysis units, monitor
the cardiovascular status of their patients to both identify new cardiovascular
insufficiency, diagnose its etiology and monitor response to resuscitation ther-
apy. Modern medicine has witnessed an impressive degree of recent medical
technological advances, allowing monitoring, display, and assessment of an
almost unimaginable number of physiological variables. Yet the utility of most
hemodynamic monitoring is unproven, whereas it is the commonly available
technologies where clinical studies have demonstrated relevance. Furthermore,
despite the many options available, most acute care centers monitor and display
only blood pressure, heart rate and pulse oximetry (SpO
2
), as they have done for
the last 20 years. Furthermore, with few exceptions, such monitoring does not
Polanco/Pinsky 134
drive treatment protocols but rather serves as automated vital signs recorded to
trigger further attention. It is hard to validate the utility of monitoring when it is
used in this fashion because no hemodynamic monitoring device will improve
outcome unless coupled to a treatment which, itself, improves outcome. Thus,
the effectiveness of hemodynamic monitoring to improve outcome is limited to
specific patient groups and disease processes where proven effective treatments
exist. Although, like most of medicine, the utility of hemodynamic monitoring
is not well documented, a primary rationale for the use of hemodynamic moni-
toring is to identify cardiovascular instability and guide therapy.
Rationale for Hemodynamic Monitoring
The arguments to justify the use of specific types of monitoring tech-
niques can be roughly grouped into three levels of defense based on their level
of validation [1]. At the basic level, the specific monitoring technique can be
defended based on historical controls. At this level, prior experience using
similar monitoring was traditionally used and presumed to be beneficial. The
mechanism by which the benefit is achieved need not be understood. The sec-
ond level of defense comes through an understanding of the pathophysiology
of the process being treated. This physiological argument can be stated as
‘knowledge of how a disease process creates its effect and thus preventing the
process from altering measured bodily functions should prevent the disease
process from progressing or injuring remote physiological functions’. Most of
the rationale for hemodynamic monitoring resides at this level. It is not clear
from recent clinical studies in critically ill patients that this argument is valid,
primarily because knowledge of the actual processes involved in the expres-
sion of disease and tissue injury is often inadequate. The third level of defense
comes from documentation that the monitoring device, by altering therapy in
otherwise unexpected ways, improves outcome in terms of survival and quality
of life. In reality, few therapies done in medicine can claim benefit at this
level. Thus, we are left with the physiological rationale as the primary defense
of monitoring critically ill patients.
The Physiological Basis for Hemodynamic Monitoring
Monitoring of critically ill patients usual serves a dual function. First, it is
used to document hemodynamic stability and the lack of a need for acute inter-
ventions and second, it is used to monitor when measured variables vary from
their defined baseline values thus defining disease. Accordingly, knowing the
Principals of Hemodynamic Monitoring 135
limits to which such monitoring reflects actual physiological values is an essential
aspect of their utility.
One the physical side, hemodynamic monitoring can be invasive or non-
invasive, and continuous or intermittent. Monitoring devices can measure phys-
iological variables directly or derive these variables through signal processing.
Signal processing does not minimize the usefulness of physiological variable
analysis; it just separates the output data from the patient by the use of the data
processor. The most common signal processing physiological variables mea-
sured clinically is the electrocardiogram. Although it is a highly processed sig-
nal bearing no simple relationship to the surface electrical potential on the
heart, the electrocardiogram is a highly sensitive and specific measure of
arrhythmias and ischemic injury.
Variables that can be measured noninvasively include body temperature,
heart rate, systolic and diastolic arterial blood pressure, and respiratory fre-
quency. Although central venous pressure (CVP) can be estimated as jugular
venous distention and hepatojugular reflux, these signs can be ambiguous and
are not usually applied in routine clinical practice today. Processed noninvasive
variables include the electrocardiogram, transcutaneous SpO
2
, expired CO
2
,
transthoracic echocardiography, tissue O
2
saturation (StO
2
) and noninvasive
respiratory plethysmography, although the latter two devices are not readily
available. Invasive monitoring reflects intravascular catheter insertion, trans-
esophageal echocardiographic probe insertion and blood component analysis.
Invasive hemodynamic monitoring of vascular pressures is usually performed
by the percutaneous insertion of a catheter into a vascular space and transduc-
ing the pressure sensed at the distal end. This allows for the continual display
and monitoring of these complex pressure waveforms. Similar intrapulmonary
vascular catheters can be used to derive thermal signals and mixed venous O
2
saturation (SvO
2
), central venous O
2
saturation (ScvO
2
), and cardiac output.
How useful this hemodynamic information is to diagnosis, treatment and prog-
nosis is a function of its reliability, established treatment protocols and guide-
lines and the expertise of the operator. What follows is an anatomical survey of
the various monitoring devices and a discussion of their utility.
Arterial Pressure Monitoring
After pulse rate, arterial pressure is the most common hemodynamic vari-
able monitored and recorded. Blood pressure is usually measured noninvasively
using a sphygmomanometer and the auscultation technique. Importantly, very
large and obese subjects in whom the upper arm circumference exceeds the
width limitations of a normal blood pressure cuff will record pressures that are
Polanco/Pinsky 136
higher than they actually are. In such patients using the large thigh blood pres-
sure cuff usually resolves this problem. Blood pressure can be measured auto-
matically using computer-driven devices (e.g. Dynamat
®
) that greatly reduce
nursing time. Sphygmomanometer-derived blood pressure measures display
slightly higher systolic and lower diastolic pressures than simultaneously mea-
sured indwelling arterial catheters, but the mean arterial pressure is usually sim-
ilar and the actual systolic and diastolic pressure differences are often small
except in the setting of increased peripheral vasomotor tone. If perfusion pres-
sure of the finger is similar to arterial pressure, then both blood pressure and the
pressure profile may be recorded noninvasively and continuously using the
optical finger probe (Fenapres). However, finger perfusion is often compro-
mised during hypovolemic shock and hypothermia, limiting this monitoring
technique to relatively well-perfused patients.
Accurate and continuous measures of arterial pressure can be done through
arterial catheterization of easily accessible arterial sites in the arm (axillary,
brachial or radial arterial) or groin (femoral arterial). Rarely are upper extrem-
ity sites other than the radial artery used because of fear of vascular compro-
mised distal to the catheter insertion site, although data supporting these fears
are nonexistent. Arterial catheterization displaying continuous arterial pressure
waveforms lends itself to arterial waveform analysis, essential in calculation
pulse pressure and pulse pressure variations and cardiac output.
The Physiological Significance of Arterial Pressure
Arterial pressure is the input pressure for organ perfusion. Organ perfusion
is usually dependent on organ metabolic demand and perfusion pressure. With
increasing tissue metabolism, organ blood flow proportionally increases by selec-
tive local vasodilation of the small resistance arterioles. If cardiac output cannot
increase as well, as is the case with heart failure, then blood pressure decreases
limiting the ability of local vasomotor control to regulate organ blood flow. If local
metabolic demand remains constant, however, changes in arterial pressure are
usually matched by changes in arterial tone so as to maintain organ blood flow
relatively constant. This local vasomotor control mechanism is referred to as
autoregulation. Cerebral blood flow over the normal autoregulatory range of 65 to
120 mm Hg is remarkably constant. Although autoregulation occurs in many
organs, like the brain, liver, skeletal muscle and skin, it is not a universal phe-
nomenon. For example, coronary flow increases with increasing arterial pressure
because the myocardial O
2
demand increases as the heart ejects into a higher arte-
rial pressure circuit. Furthermore, renal blood flow increases in a pressure-dependent
fashion over its entire pressure for similar reasons. As renal flow increases, so
does renal filtrate flow into the tubules, increasing renal metabolic demand. Thus,
a normal blood pressure does not mean that all organs have an adequate amount
Principals of Hemodynamic Monitoring 137
of perfusion, because increases in local vasomotor tone and mechanical vascular
obstruction can still induce asymmetrical vascular ischemia.
Determinants of Arterial Pressure
Arterial pressure is a function of both vasomotor tone and cardiac output.
The local vasomotor tone also determines blood flow distribution, which itself is
usually determined by local metabolic demands. For a constant vasomotor tone,
vascular resistance can be described by the relation between changes in both arte-
rial pressure and cardiac output. The body defends organ perfusion pressure
above all else in its autonomic hierarchy through alterations in ␣-adrenergic tone,
mediated though baroreceptors located in the carotid sinus and aortic arch. This
supremacy of arterial pressure in the adaptive response to circulatory shock exists
because both coronary and cerebral blood flows are dependent only on perfusion
pressure. The cerebral vasculature has no ␣-adrenergic receptors; the coronary
circulation has only a few. Accordingly, hypotension always reflects cardiovascu-
lar embarrassment, but normotension does not exclude it. Hypotension decreases
organ blood flow and stimulates a strong sympathetic response that induces a
combined ␣-adrenergic (increased vasomotor tone) and -adrenergic (increased
heart rate and cardiac contractility) effect and causes a massive ACTH-induced
cortisol release from the adrenal glands. Thus, to understand the determinants of
arterial pressure one must also know the level of vasomotor tone.
In the ICU setting, arterial tone can be estimated at the bedside by calcu-
lating systemic vascular resistance. Using Ohm’s law, resistance equals the ratio
of the pressure to flow, usually calculated as the ratio of the pressure gradient
between aorta and CVP to cardiac output. Arterial tone can also be calculated as
total peripheral resistance, which is the ratio of mean arterial pressure to cardiac
output. Regrettably, neither systemic vascular resistance nor total peripheral
resistance faithfully describes arterial resistance. Arterial resistance is the slope
of the arterial pressure-flow relation. The calculation of systemic vascular resis-
tance using CVP as the backpressure to flow has no physiological rationale and
the use of systemic vascular resistance for clinical decision making should be
abolished. Regrettably, both systemic vascular resistance and total peripheral
resistance are still commonly used in hemodynamic monitoring because they
allow for the simultaneous assessment of both pressure and flow, while the
actual measure of arterial tone is more difficult to estimate.
The determinants of arterial pressure can simplistically be defined as systemic
arterial tone and blood flow. Since blood flow distribution will vary amongst
organs relative to their local vasomotor tone and arterial pressure is similar for
most organs, measures of peripheral resistance, by any means or formula, reflect
the lump parameter of all the vascular beds, and thus, describe no specific vascular
bed completely. If no hemodynamic instability alters normal regulatory mechanisms,
Polanco/Pinsky 138
then local blood flow will also be proportional to local metabolic demand. Within
this construct, the only reason cardiac output becomes important is to sustain an
adequate and changing blood flow to match changes in vasomotor tone such that
arterial input pressure remains constant. Since cardiac output is proportional to
metabolic demand there is no level of cardiac output that reflects normal values in
the unstable and metabolically active patient. However, as blood pressure
decreases below a mean of 60 mm Hg and/or cardiac indices decrease below
2.0 liters/min/m
2
, organ perfusion usually becomes compromised, and if sustained
it will lead to organ failure and death. Presently, only one clinical trial examined
the effect of increasing mean arterial pressure on tissue blood flow. When patients
with circulatory shock were resuscitated with volume and vasopressors to a mean
arterial pressure range of 60–70, 70–80 or 80–90 mm Hg, no increased organ
blood flow could be identified above a mean arterial pressure of 65 mm Hg.
Clearly, subjects with prior hypertension will have their optimal perfusion pres-
sure range increased over normotensive patients. Thus, there is no firm data sup-
porting any one limit of arterial pressure or cardiac output values or therapeutic
approaches based on these values that have proven more beneficial than any other
has. Accordingly, empiricism is the rule regarding target values of both mean arte-
rial pressure and cardiac output. At present, the literature suggests that maintain-
ing a nonpreviously hypertensive patient’s mean arterial pressure ⬎65mm Hg by
the use of fluid resuscitation and subsequent vasopressor therapy, as needed, is an
acceptable target. Previously hypertensive subjects will need a higher mean arter-
ial pressure to insure the same degree of blood flow [2]. There is no proven value
in forcing either arterial tone or cardiac output to higher levels to achieve a mean
arterial pressure above this threshold. In fact, data suggest that further resuscita-
tive efforts using vasoactive agents markedly increase mortality [3], and the rela-
tively new concept of ‘delayed’ and ‘hypotensive resuscitation’ for traumatic
hemorrhagic shock on the other hand had shown improved outcome in some clin-
ical and experimental studies [4–6]. However, these studies were done in trauma
patients with penetrating wounds and no immediate access to surgical repair.
Once a patient is in the hospital and the sites of active bleeding addressed, then
aggressive fluid and pressor resuscitation is indicated.
Arterial Pressure Variations during Ventilation
The majority of the critically ill surgical patients treated in the ICU are
usually on mechanical ventilation. Ventilation-induced arterial pressure vari-
ations have been described since antiquity as pulsus paradoxus. Inspiratory
decreases in arterial pressure were used to monitor both the severity of bron-
chospasm in asthmatics and their inspiratory efforts [7].
Recently renewed interest in the hemodynamic significance of heart-lung
interactions has emerged. The commonly observed variations in arterial pressure
Principals of Hemodynamic Monitoring 139
and aortic flow seen during positive pressure ventilation have been analyzed
as a measure of preload responsiveness [8]. The rationale for this approach is
that positive-pressure ventilation-induced changes in either systolic arterial
pressure (used to describe pulsus paradoxus), arterial pulse pressure or stroke
volume can predict in which subjects cardiac output will increase in response to
fluid resuscitation. Ventilation-induced changes in systolic arterial pressure
(pulsus paradoxus) and arterial pulse pressure are easy to measure from arter-
ial pressure recordings. The greater the degree of systolic arterial pressure or
pulse pressure variation over the respiratory cycle, the greater the increase in
cardiac output in response to a defined fluid challenge. Recently, measuring
the mean change in aortic blood flow during passive leg raising in sponta-
neous breathing patients has also proven accurate in predicting preload respon-
siveness [9].
Although arterial pressure variations are a measure of preload responsive-
ness [10] the ‘traditional’ preload measures, such as right atrial pressure (Pra),
pulmonary artery occlusion pressure (Ppao), right ventricular (RV) end-
diastolic volume and intrathoracic blood volume, poorly reflect preload respon-
siveness [11]. In essence, preload is not preload responsiveness.
Indications for Arterial Catheterization
The arterial catheter is frequently inserted as a ‘routine’ at the admission of
patients to the ICU for continuous monitoring blood pressure and repetitive
measurements of blood gases. There is no evidence to support this exaggerated
clinical practice. Although probably the only proven indication for arterial
catheterization is to synchronize the intra-aortic balloon of counterpulsation,
there are some others indications where the information obtained is valuable in
the assessment and treatment of the patient, such as cardiovascular instability or
the use of vasopressors or vasodilators during resuscitation. The probable indi-
cations for arterial catheterization are summarized in table 1. Although arterial
catheterization is an invasive procedure that is not free of complications,
a recent systematic review of a large number of cases showed that most of the
complications were minor, including temporary vascular occlusion (19.7%) and
hematoma (14.4%). Permanent ischemic damage, sepsis and pseudoaneurysm
formation occurred in less than 1% of cases [12].
CVP Monitoring
Methods of Measuring CVP
CVP is the pressure in the large central veins proximal to the right atrium
relative to atmosphere. In the ICUs the CVP is usually measured using a fluid-filled
Polanco/Pinsky 140
catheter (central venous line or Swan-Ganz catheter) with the distal tip located
in the superior vena cava connected to a manometer or more often to a pressure
transducer of a monitor, displaying the waveform in a continuous fashion. CVP
can also be measured noninvasively as jugular venous pressure by the height of
the column of blood distending the internal and external jugular veins, when the
subjects are sitting in a semireclined position, such that the small elevations in
CVP will be reflected in a persistent jugular venous distention.
Determinants of CVP
Starling demonstrated the relationship between cardiac output, venous
return and CVP showing that increasing the venous return (and preload)
increases the stroke volume (and cardiac output) until a plateau is reached.
Although the CVP is clearly influenced by the volume of blood in the central
compartment and its venous compliance, there are several physiological
and anatomical factors that can influence its measurement and waveform such as
the vascular tone, RV function, intrathoracic pressure changes, tricuspid valve
disease, arrhythmias, and both myocardial and pericardial disease (table 2).
Monitoring CVP
CVP has been used as a monitor of central venous blood volume and an
estimate of the right atrial pressure for many years, being wrongly used as a
parameter and sometimes goal for replacement of intravascular volume in
Table 1. Arterial catheterization
Indications for arterial catheterization
•
As a guide to synchronization of intra-aortic balloon counterpulsation
Probable indications for arterial catheterization
•
Guide to management of potent vasodilator drug infusions to prevent systemic hypotension
•
Guide to management of potent vasopressor drug infusions to maintain a target mean arterial pressure
•
As a port for the rapid and repetitive sampling of arterial blood in patients in whom multiple arterial
blood samples are indicated
•
As a monitor of cardiovascular deterioration in patients at risk of cardiovascular instability
Useful applications of arterial pressure monitoring in the diagnosis of cardiovascular insufficiency
•
Differentiating cardiac tamponade (pulsus paradoxus) from respiration-induced swings in systolic
arterial pressure
Tamponade reduces the pulse pressure but keeps diastolic pressure constant; respiration reduces
systolic and diastolic pressure equally, such that pulse pressure is constant
•
Differentiating hypovolemia from cardiac dysfunction as the cause of hemodynamic instability
Systolic arterial pressure decreases more following a positive pressure breath as compared to an
apneic baseline during hypovolemia; systolic arterial pressure increases more during positive
pressure inspiration when LV contractility is reduced
Principals of Hemodynamic Monitoring 141
shock patients. The validity of this measure as an index of RV preload is nonex-
istent across numerous studies. It has been shown that CVP has a poor correlation
with cardiac index, stroke volume, left ventricular (LV) end-diastolic volume
and RV end-diastolic volume [13–15].
Although a very high CVP demands a certain level of total circulating
blood volume, one may have a CVP of 20 mm Hg and still have an underfilled
LV that is fluid responsive. For example, in the setting acute RV infarction CVP
can be markedly elevated, whereas cardiac output often increases further with
volume loading. In reported series, some patients with low CVP failed to
respond to fluids and some patients with high CVP responded to challenge of
fluids [16]. Based on this and the poor correlations described above it is impos-
sible to define ideal values of CVP. However there is some evidence that vol-
ume loading in patients with CVP ⬎ 12 mm Hg is very unlikely to increase
cardiac output [17]. Thus, the only usefulness of CVP is to define relative hyper-
volemia, since an elevated CVP only occurs in disease. Two clinical studies
Table 2. Factors affecting the measured CVP
Central venous blood volume
Venous return/cardiac output
Total blood volume
Regional vascular tone
Compliance of central compartment
Vascular tone
RV compliance
Myocardial disease
Pericardial disease
Tamponade
Tricuspid valve disease
Stenosis
Regurgitation
Cardiac rhythm
Junctional rhythm
Atrial fibrillation
A-V dissociation
Reference level of transducer
Positioning of patient
Intrathoracic pressure
Respiration
Intermittent positive pressure ventilation
PEEP
Tension pneumothorax
Polanco/Pinsky 142
showed a potential benefit in specific groups of surgical patients (hip replace-
ment and renal transplant patients) [18, 19] in whom CVP was used to guide
therapy; however, there is no clinical evidence that CVP monitoring improves
outcome in critically ill patients and attempts to normalize CVP in early goal-
directed therapy during resuscitation do not display any benefit [20].
Pulmonary Artery Catheterization and Its Associated
Monitored Variables
Pulmonary arterial catheterization allows the measurement of many clinically
relevant hemodynamic variables (table 3) and, in combination with measures of
Table 3. Physiological measures derived from invasive monitoring and their physiolog-
ical relevance
Arterial pressure
Mean arterial pressure
Organ perfusion inflow pressure
Arterial pulse pressure and its variation during ventilation
LV stroke volume changes and pulsus paradoxus
Preload responsiveness (if assessed during intermittent positive pressure ventilation)
Arterial pressure waveform
Aortic valvulopathy, input impedance and arterial resistance
Used to calculate stroke volume by pulse contour technique
Central venous pressure (CVP)
Mean CVP
If elevated, the effective circulating blood volume is not reduced
CVP variations during ventilation
Tricuspid insufficiency, tamponade physiology
Preload responsiveness (if assessed during spontaneous breathing)
Pulmonary arterial pressure (Ppa)
Mean Ppa
Pulmonary inflow pressure
Systolic Ppa
RV pressure load
Diastolic Ppa and pulse pressure and their variations during ventilation
RV stroke volume, PVR
Diastolic pressure tract changes in intrathoracic pressure during ventilation
Pulmonary artery occlusion pressure (Ppao)
Mean Ppao
Left atrial and LV intraluminal pressure and by inference, LV preload
Backpressure to pulmonary blood flow
Ppao waveform and its variation during occlusion and ventilation
Mitral valvulopathy, atrial or ventricular etiology of arrhythmia, accuracy of mean
Ppao to measure intraluminal LV pressure, and pulmonary capillary pressure (Ppc)
Principals of Hemodynamic Monitoring 143
arterial and mixed venous blood gases, many relevant calculated parameters
(table 4). One can measure the intrapulmonary vascular pressures including
CVP, pulmonary arterial pressure (Ppa), and by intermittent balloon occlusion
of the pulmonary artery, Ppao and pulmonary capillary pressure (Ppc).
Furthermore, by using the thermodilution technique and the Stewart-Hamilton
equation one can estimate cardiac output and RV ejection fraction, global car-
diac volume and intrathoracic blood volume. Finally, one can measure SvO
2
either intermittently by direct sampling of blood from the distal pulmonary arte-
rial port or continuously via fiberoptic reflectometry. Assuming one knows the
hemoglobin concentration and can tract arterial O
2
saturation (SaO
2
), easily
estimated noninvasively by SpO
2
, one can calculate numerous derived variables
that describe well the global cardiovascular state of the patient. These derived
variables include total O
2
delivery (DO
2
), whole body O
2
consumption (VO
2
),
venous admixture (as an estimate of intrapulmonary shunt), pulmonary and
systemic vascular resistance, RV end-diastolic and end-systolic volumes, and
both RV and LV stroke work index.
Table 4. Physiological variables derived from invasive monitoring
Calculated using multiple measured variables including cardiac output by thermodilution
(COtd), arterial and mixed venous blood gases and end-tidal CO
2
(PetCO
2
)
Vascular resistances
Total peripheral resistance ⫽ MAP/COtd
Systemic vascular resistance ⫽ (MAP – CVP)/COtd
Pulmonary arterial resistance ⫽ (mean Ppa – Ppc)/COtd
Pulmonary venous resistance ⫽ (Ppc – Ppao)/COtd
Pulmonary vascular resistance ⫽ (mean Ppa – Ppao)/COtd
Vascular pump function
Left ventricular stroke volume (SVIv) ⫽ COtd/HR
Left ventricular stroke work (SWIv) ⫽ (MAP – Ppao)/SVIv
Preload-recruitable stroke work ⫽ SWIv/Ppao
Oxygen transport and metabolism
Global oxygen transport or delivery (DO
2
) ⫽ CaO
2
/COtd
Global oxygen uptake (VO
2
) ⫽ (CaO
2
– CvO
2
)/COtd
Venous admixture
Ratio of dead space to total tidal volume (Vd/Vt) ⫽ PaCO
2
/(PaCO
2
– PetCO
2
)
RV function using RV ejection fraction (EFrv) catheter-derived data
RV end-diastolic volume (EDVrv) ⫽ SV/EFrv
RV end-systolic volume (ESVrv) ⫽ EDVrv – SV
HR ⫽ Heart rate; MAP ⫽ mean arterial pressure.
Polanco/Pinsky 144
Pulmonary Artery Pressure
The determinants of Ppa are the volume of blood ejected to the pulmonary
artery during systole, the resistance of the pulmonary vascular bed and the
downstream left atrial pressure. The pulmonary vascular bed is a low resistance
circuit with a large reserve that allows increases of cardiac output with minor
changes in the Ppa. On the other hand, increases in the downstream venous
pressure (e.g. LV failure) or in the flow resistance (e.g. lung diseases) rises
the Ppa. Although increases in cardiac output alone do not cause pulmonary
hypertension, having an increased vascular resistance the Ppa can be increased
due to changes in cardiac output. Based on these considerations the Ppa should
not be used as a reliable parameter of ventricular filling in several lung diseases
that conditioned changes in the vascular tone and cardiac output. The normal
range of values for Ppa are systolic 15–30 mm Hg, diastolic 4–12mm Hg, and
mean 9–18 mm Hg [21].
Ppao Monitoring
Methods of Measuring Ppao
Numerous studies of physicians have demonstrated that the ability to accu-
rately measure Ppao from a strip chart recording or a freeze frame snapshot of
the monitor screen is poor. Many initiatives have been put into place to educate
physicians and nurses, but the reality is that since the pressure measured also
reports changes in intrathoracic pressure, a value which is always changing, the
accuracy of Ppao measures is likely to remain poor in graduates of all educa-
tions programs.
The Ppao value is thought to reflect the LV filling because of the unique
characteristic of the pulmonary circulation. Balloon inflation of the pulmonary
artery catheter forces the tip to migrate distally into smaller vessels until the tip
occludes a medium-sized (1.2-cm-diameter) pulmonary artery. This occlusion
stops all blood flow in that vascular tree distal to the occlusion site until such
time as other venous branches reconnect downstream to this venous draining
bed. The point where such parallel pulmonary vascular beds anastomose is at a
point about 1.5 cm from the left atrium. Thus, if a continuous column of blood
is present from the catheter tip to the left heart, then Ppao measures pulmonary
venous pressure at this first junction, or J-1 point, of the pulmonary veins [22].
As downstream pulmonary blood flow ceases, distal Ppa falls in a double expo-
nential fashion to a minimal value, reflecting the pressure downstream in the
pulmonary vasculature from the point of occlusion. The Ppa value where the
first exponential pressure decay is overtaken by the second longer exponential
pressure decay reflects Ppc measures, useful in calculating pulmonary arterial
Principals of Hemodynamic Monitoring 145
and venous resistances. Importantly, the column of water at the end of the
catheter is now extended to include the pulmonary vascular circuit up to this
J-1 point of blood flow. Since the vasculature is compliant relative to the
catheter, vascular pressure signals dampen relative to the nonoccluded Ppa sig-
nal. Thus, the two primary aspects of Ppao measures that are used to identify an
occluded pressure are the decrease in diastolic pressure values to less than dias-
tolic Ppa and the dampening of the pressure signal (fig. 1). If one needed fur-
ther validation that the catheter is actually in an occluded vascular bed, then one
could measure the pH, pCO
2
and pO
2
of blood sampled from the occluded dis-
tal tip of the catheter. Since the sampled blood will be from the stagnant pool of
blood, its removal will make it be pulled back into the pulmonary artery
catheter (PAC) from the pulmonary veins. Since the blood sampled will have
crossed the alveolar capillaries twice, its pCO
2
will be lower than arterial pCO
2
and its pO
2
higher, due to the law of mass action.
Pleural Pressure and Ppao
Although one may measure Ppao accurately relative to atmosphere, the
heart and large vessel pulmonary vasculature live in an intrathoracic compart-
ment and sense pleural pressure (Ppl) as their surrounding pressure. Ventilation
causes significant swings in Ppl. Pulmonary vascular pressures, when measured
relative to atmospheric pressure, will reflect these respiratory changes in Ppl. To
minimize this ‘respiratory artifact’ on intrathoracic vascular pressure recordings,
measures are usually made at end-expiration. During quiet spontaneous breath-
ing, end-expiration occurs at the highest vascular pressure values, whereas
during passive positive-pressure breathing, end-expiration occurs at the lowest
vascular pressure values. With assisted ventilation or with forced spontaneous
ventilation, it is often difficult to define end-expiration [23]. These limitations
are the primary reasons for inaccuracies in estimating Ppao at the bedside.
Even if measures of Ppao are made at end-expiration and Ppao values
reflecting a continuous column of fluid from the catheter tip to the J-1 point,
Fig. 1. Two examples of a Ppa waveform before and then during balloon occlusion to
measure pulmonary artery occlusion pressure (Ppao). Note the left-sided occlusion tracing
has a higher pressure than diastolic Ppa indicating an inaccurate estimate of Ppao, whereas
the right-sided tracing indicates a more accurate estimate of Ppao.
Damped Ppa True Ppao
Mean
Ppa
Polanco/Pinsky 146
these Ppao measures may still overestimate Ppao if Ppl is elevated at end-
expiration. Hyperinflation, due to air trapping, dynamic hyperinflation or the
use of extrinsic positive end-expiratory pressure (PEEP) will all increase end-
expiratory Ppl to a varying degree as a function of airway resistance and lung
and chest wall compliance. It is not possible to predict with accuracy the degree
to which increases in PEEP will increase Ppl. Since differences in lung and
chest wall compliance exist among patients and in the same patient over time,
one cannot assume a fixed relation between increases in Paw and Ppl [24].
Why Measure Ppao?
Ppao is used most often in the bedside assessment of: (1) pulmonary
edema, (2) pulmonary vasomotor tone, (3) intravascular volume status and LV
preload, and (4) LV performance. These points were summarized recently and
will be restated below [25].
Pulmonary Edema
Pulmonary edema can be caused by either elevations of Ppc, referred to as
hydrostatic or secondary pulmonary edema, or increased alveolar capillary
or epithelial permeability, referred to as primary pulmonary edema. Usually
hydrostatic pulmonary edema requires a pulmonary capillary increase to
⬎18 mm Hg. However, if capillary or alveolar cell injury is present, alveolar
flooding can occur at much lower Ppc. Furthermore, in the setting of chronic
pulmonary vascular congestion, increased pulmonary lymphatic flow and
increased respiratory excursions promote a rapid clearance of lung interstitial
fluid minimizing edema formation. Still, measures of Ppao are commonly used
to determine the cause of pulmonary edema. Ppao values ⬍18 mm Hg suggest
a nonhydrostatic cause, whereas values ⬎20 mm Hg suggest a hydrostatic cause
of pulmonary edema [22]. However, many exceptions to this rule exist. As men-
tioned above if increased lung permeability is present then fluid-resuscitation-
induced pulmonary edema may occur at Ppao values much below 18 mm Hg,
and treatment strategies aimed at reducing Ppao will further reduce pulmonary
edema formation. Similarly, if pulmonary venous resistance is increased, then
Ppc may be much higher than the measured Ppao inducing hydrostatic pul-
monary edema despite no increased lung permeability and a low Ppao. Similarly,
Ppao may be ⬎20 mm Hg without any evidence of hydrostatic pulmonary
edema, either because Ppl is also elevated or because of increased pulmonary
lymphatic flow.
Pulmonary Vasomotor Tone
The pulmonary circulation normally has a low resistance, with pul-
monary arterial diastolic pressure only slightly higher than Ppao and mean
Principals of Hemodynamic Monitoring 147
Ppa a few millimeters Hg higher than Ppao. Pulmonary vascular resistance
(PVR) can be estimated using Ohm’s law as the ratio of the pulmonary vascu-
lar pressure gradient (mean pulmonary artery pressure minus Ppao) and car-
diac output [i.e. PVR ⫽ (mean Ppa – Ppao)/cardiac output]. Normal PVR is
between 2 and 4 mm Hg ⫻ l/min/m
2
. Usually these values are multiplied by
80 to give a normal PVR range of 150–250 dyn s/cm
5
. Either an increased
PVR or a passive pressure build-up from the pulmonary veins can induce pul-
monary hypertension. If pulmonary hypertension is associated with an
increased PVR then the causes are primarily within the lung. Diagnoses such
as pulmonary embolism, pulmonary fibrosis, essential pulmonary hyperten-
sion and pulmonary venoocclusive disease need to be excluded. Whereas if
PVR is normal then LV dysfunction is the more likely cause of pulmonary
hypertension [26]. Since the treatments for these two groups of diseases is
quite different despite similar increases in Ppa, the determination of PVR in
the setting of pulmonary hypertension is very important. Regrettably, PVR
poorly reflects true pulmonary vasomotor tone in lung disease states and dur-
ing mechanical ventilation, especially with the application of PEEP. Alveolar
pressure (Palv) can be the backpressure to pulmonary blood flow in certain
lung regions during positive-pressure ventilation and in the presence of
hyperinflation because Palv exceeds left atrial pressure. Furthermore, since
lung disease is usually nonhomogeneous, pulmonary blood flow is preferen-
tially shifted from compressed vessels in West Zone 1 and 2 conditions
(i.e. Ppao ⬍ Palv and Ppa ⫽ Palv, respectively) to those circuits with the low-
est resistance (West Zone 3, i.e. Ppao ⬎ Palv), thus making the lung vascular
pathology appear less than it actually is.
LV Preload
Ppao is often taken to reflect LV filling pressure, and by inference, LV end-
diastolic volume. Patients with cardiovascular insufficiency and a low Ppao are
presumed to be hypovolemic and initially treated with fluid resuscitation,
whereas patients with similar presentations but an elevated Ppao are presumed
to have an impaired contractile function. Although there are no accepted high
and low Ppao values for which LV underfilling is presumed to occur or not,
Ppao values ⬍10 mm Hg are usually used as presumed evidence of a low LV
end-diastolic volume, whereas values ⬎18 mm Hg suggest a distended LV [27].
Unfortunately, there is very little data to support this approach and almost no
data to defend this logic. There are multiple documented reasons for this
observed inaccuracy that relate to individual differences in LV diastolic compli-
ance and contractile function [28]. First, the relation between Ppao and LV end-
diastolic volume is curvilinear and is often very different among subjects and
within subjects over time. Thus, neither absolute values of Ppao or changes in
Polanco/Pinsky 148
Ppao will define a specific LV end-diastolic volume or its change [29]. Second,
Ppao is not the distending pressure for LV filling. It is only the internal pressure
of the pulmonary veins relative to atmospheric pressure. Assuming Ppao
approximated left atrial pressure, it will poorly reflect LV end-diastolic pressure
because it poorly follows the late diastolic pressure rise induced by atrial con-
traction and does not measure pericardial pressure, which is the outside pres-
sure for LV distention. With lung distention Ppl increases increasing pericardial
pressure. Although we can estimate Ppl using esophageal balloon catheters,
pericardial pressure is often different. Changes in pericardial pressure will alter
LV end-diastolic volume independent of Ppao. Finally, even if one knew peri-
cardial pressure and Ppao did accurately reflect LV end-diastolic pressure, LV
diastolic compliance can vary rapidly changing the relation between LV filling
pressure and LV end-diastolic volume. Myocardial ischemia, arrhythmias, and
acute RV dilation can all occur over a few heartbeats. Thus, it is not surprising
that Ppao is a very poor predictor of preload responsiveness. Thus, it is not rec-
ommended to use Ppao to predict response to fluid resuscitation in critically ill
patients.
LV Performance
The four primary determinants of LV performance are preload (LV end-
diastolic volume), afterload (maximal LV wall stress), heart rate and contractil-
ity. Ppao is often used as a substitute for LV end-diastolic volume when constructing
Starling curves (i.e. relationship between changing LV preload and ejection
phase indices). Usually one plots Ppao versus LV stroke work (LV stroke
volume ⫻ developed pressure). Using this construct, patients with heart failure
can be divided into four groups depending on their Ppao (⬎ or ⬍18mm Hg) and
cardiac index values (⬎ or ⬍2.2 liters/min/m
2
) [27]. Those patients with low car-
diac indices and high Ppao are presumed to have primary heart failure, and a low
cardiac output and low Ppao, on the other hand, reflect hypovolemia. Those with
high cardiac indices and high Ppao are presumed to be volume overloaded, and
having high cardiac output and low Ppao reflect increased sympathetic tone.
Although this maybe a useful construct for determining diagnosis, treatment and
prognosis of patients with acute coronary syndrome, it poorly predicts cardio-
vascular status in other patient groups. However, as described above, if LV end-
diastolic volume and Ppao do not trend together in response to fluid loading or
inotropic drug infusion, then inferences about LV contractility based on this
Ppao/LV stroke work relation may be incorrect. This is not a minor point.
Volume loading may induce acute RV dilation markedly reducing LV diastolic
compliance, such that Ppao will increase as LV stroke work decreases. However,
the relationship between LV end-diastolic volume and stroke work need not
have changed at all. Similarly, inotropic drugs, like dobutamine, may reduce
Principals of Hemodynamic Monitoring 149
biventricular volumes by decreasing venous return, decreasing LV diastolic
compliance, even if the heart is not responsive to inotropic therapy. Thus, the
same limitations on the use of Ppao in assessing LV preload must be considered
when using it to assess LV performance.
Cardiac Output Monitoring
Measuring Cardiac Output
Cardiac output can be estimated by many techniques, including invasive
hemodynamic monitoring. Pulmonary blood flow, using a balloon floatation
PAC equipped with a distal thermistor, and transpulmonary blood flow, using
an arterial thermistor, both with a central venous cold volume injection, can be
used. Similarly, minimally invasive echo Doppler techniques can measure
blood flow at the aortic value and descending aortic flow using esophageal
Doppler monitoring. Cardiac output can be measured intermittently by bolus
cold injection or continuously by cold infusion. The advantage of the continu-
ous cardiac output technique and the transpulmonary technique is that neither
is influenced greatly by the ventilation-induced swings in pulmonary blood
flow. Measurement of cardiac output by intermittent pulmonary artery flow
measures using bolus cold indicator and monitoring the thermal decay curve is
the most common method to measure cardiac output at the bedside. However,
such intermittent measures will show profound ventilatory cycle-specific pat-
terns [30]. By making numerous measures at random with the ventilatory
cycle and then averaging all measures with proper thermal decay profiles,
regardless of their values, one can derive an accurate measure of pulmonary
blood flow [31].
Recently, a renewed interest in pulse contour analysis to estimate LV stroke
volume, and therefore cardiac output, from the arterial pressure profile over
ejection has acquired its own set of supporters [32]. Arterial pressure and arte-
rial pulse pressure are a function of rate of LV ejection, LV stroke volume and
the resistance, compliance and inertance characteristics of the arterial tree and
blood. If the arterial components of tone remain constant, then changes in pulse
pressure most proportionally reflect changes in LV stroke volume. Thus, it is
not surprising that the aortic flow variation parallels arterial pulse pressure
variation [33], and pulse contour-derived estimates of stroke volume variation
can be used to determine preload responsiveness [34, 35]. Caution must be
applied to using the pulse contour method because it has not been validated in
subjects with rapidly changing arterial tone, as often occurs in subjects with
hemodynamic instability. Furthermore, it requires the application of abnormally
large tidal volumes [34–36]. Thus, at the present time, the pulse contour-derived
Polanco/Pinsky 150
stroke volume variation technique represents a potentially great but still
unproven clinical decision tool [37].
Currently three commercial devices that use pulse contour analysis of an
arterial line waveform to obtain continuous cardiac output are approved for
clinical use (PiCCO
TM
, LIDCO
TM
and VigileoEdwards
TM
systems). The benefit
of being minimally invasive and the correlation shown with ‘standard’ methods
of measuring cardiac output in some clinical and experimental studies make
them a promising tool for hemodynamic monitoring [38, 39].
Mixed Venous SO
2
Monitoring
Measuring SvO
2
SvO2 reflects the pooled SvO
2
and is an important parameter in the assess-
ment of the adequacy of DO
2
and its relation with VO
2
. A decrease in SvO
2
could
be explained by a decrease in DO
2
or any of the parameters that determine this
like saturation (SaO
2
), cardiac output and hemoglobin, and also by an increase in
VO
2
. A decrease of DO
2
will be followed by stable VO
2
with a consequent
decrease of the SvO
2
until a critical value of DO
2
is reached where the tissues are
not longer able to compensate having a constant VO
2
, and VO
2
becomes depen-
dent on DO
2
in an almost linear relation. At this level SvO
2
, though continuing to
decrease, becomes less sensitive to changes of tissue perfusion.
SvO2 measured from blood drawn from the distal tip of a PAC represents
the true mixed venous value of the blood blended in the right ventricle. Care
must be taken to withdraw blood slowly so that it does not get aspirated from
the downstream pulmonary capillaries. Validation of true mixed venous blood
requires documentation that the measured PvCO
2
is greater than PaCO
2
, because
blood drawn over the capillaries sees alveolar gas twice and will have a lower
PCO
2
than arterial blood. Continuous measures of SvO
2
can be made using
fiberoptic reflectance spectroscopy. Two techniques are commercially available.
Both use one fiberoptic line to send a light signal and another to receive the
reflected light at a different wavelength. However, only one catheter (Abbott
TM
)
uses the Shaw technique of also measuring hemoglobin reflectance and thus
remains accurate of wide changes in hemoglobin concentration. The other
catheter (Edward
TM
) requires recalibration if hemoglobin levels vary by more
than 1 g/dl. Both techniques are valuable to monitoring SvO
2
trends as either
cardiac output, arterial O
2
content or metabolic demand varies.
ScvO
2
Recent interest in central venous O
2
saturation (ScvO
2
) has evolved
over the past years with the positive results of the study of Rivers et al. [40].
Principals of Hemodynamic Monitoring 151
Rivers et al. demonstrated that in patients with septic shock or severe sepsis
admitted to the Emergency Department an early and aggressive resuscitation
guided by ScvO
2
, CVP and mean arterial pressure reduced 28-day mortality
from 46.5 to 30.5%. However, measures of SvO
2
remain the gold standard to
reflect minimal DO
2
. This is because although ScvO
2
and SvO
2
covary and
seem to follow a parallel tracking, their differences can exceed 5%. Furthermore,
during dynamic changes in cardiac output as occur in shock states, ScvO
2
may
exceed SvO
2
by 5% or more or be less than SvO
2
by 5% or more [41]. Thus,
using a defined threshold value for ScvO
2
to identify when to start or stop
resuscitation in a critically ill patient is fundamentally flawed. Still, a low
ScvO
2
(⬍65%) is invariably associated with a low SvO
2
(⬍72%), making it less
sensitive but still clinically useful at lower threshold values.
The Meaning of Cardiac Output and SvO
2
as End Points
of Resuscitation
Although one may potentially measure cardiac output accurately at the
bedside, there is no such thing as a normal cardiac output. Cardiac output is
either adequate for the needs of the body or it is not. For example, the same car-
diac output and DO
2
that is adequate at rest may be grossly inadequate and not
associated with life during periods of increased metabolic demand. Since the
primary goal of the cardiorespiratory system is to continuously maintain ade-
quate amounts of O
2
(DO
2
) to meet the metabolic demands of the tissues (VO
2
),
neither cardiac output nor mean arterial pressure are sensitive or specific mea-
sures of the adequacy of cardiovascular function. Clearly, the best measures of
the adequacy of blood flow are the continued maintenance of normal end-organ
function without evidence of excessive anaerobic metabolism. Normal urine
output, gut activity, mentation, normal blood lactate levels and spontaneous
voluntary muscular activities reflect the most easily validated measures of body
health [42]. Regrettably, many patients present with coexistent organ system
dysfunction, either preexistent or due to the insult. Furthermore, organ function
cannot be monitored quickly enough to allow for titration of care. Thus, one
cannot rely on these absolute markers to direct therapy [43]. Perhaps a more
functional marker of adequacy of DO
2
to the tissues is SvO
2
[44]. Although val-
ues of SvO
2
⬎70% do not insure that all vascular beds are adequately perfused,
SvO
2
values ⬍60% are associated with oxidative impairment of tissues with a
high metabolic rate and values ⬍50% are uniformly associated with evidence
of anaerobic metabolism in some vascular beds [45]. Thus, as a negative pre-
dictive marker, preventing SvO
2
from decreasing below 50% and hopefully
keeping it above 70% by fluid resuscitation, sedation and ancillary support (e.g.
mechanical ventilation to reduce the work cost of breathing), all may improve
DO
2
to metabolically active tissues.
Polanco/Pinsky 152
If the metabolic demand changes, cardiac output should covary with it
[46]. Since this puts an added variable on the analysis of hemodynamic stabil-
ity, a common approach in the cardiovascular management of the critically ill
patient is to minimize the extraneous metabolic demands of the patients during
intervals in which therapeutic interventions and diagnostic processes are being
performed so as to maintain stable baseline O
2
consumption for comparison.
Thus, minimizing the work cost of breathing by using mechanical ventilation,
and reducing sympathetic responses by infusion of sedative agents, reflect sta-
bilizing processes that allow for accurate hemodynamic assessment. This is
often more difficult to achieve than imagined [31]. Even a sedated and mechan-
ically ventilated subject can be expending much effort assisting or resisting the
ventilator-derived breaths. Muscular activities, such as moving in bed or
being turned, ‘fighting the ventilator’, and breathing spontaneously can eas-
ily double resting VO
2
[47]. O
2
supply and demand must covary as a normal and
expected aspect of homeostasis under almost all conditions. In cardiovascular
insufficiency states, such as cardiogenic shock or hypovolemic shock, total
cardiac output is often limited and cannot increase enough in response to
increasing metabolic demand to match the demand. Under these severe condi-
tions VO
2
tends to remain constant by varying the extraction of O
2
in the tissues
rather than by varying total blood flow. Thus, measures of SvO
2
can be used to
identify patients in circulatory shock. Furthermore, resuscitation efforts that
increase SvO
2
to ⬎70% should be associated with improved end-organ
function.
The Controversy of the PAC
One would think that the clinical use of the pulmonary catheter in the
management of the hemodynamically unstable patient would be invaluable.
However, this utility has not been documented. Although there are no proven
indications for the insertion of PAC, there are potential indications (yet
not proven) for its use based on the need to assess cardiac function, global
DO
2
, intravascular volume status and pulmonary pressures as summarized in
table 4.
The controversy over the use of the pulmonary arterial catheter in the man-
agement of critically ill patients continues to rage. Proponents of its use cite
physiological rationale to diagnosis and titration of complex treatments that
may otherwise be detrimental. Opponents of its use cite the almost total lack of
data showing that its use in the management of critically ill patients improves
outcome. Still, one truth remains: no catheter will improve outcome unless cou-
pled to a treatment that itself improves outcome.
Despite some exciting initial uncontrolled reports of markedly improved
outcome in high-risk surgery patients [48, 49], further well-controlled studies
Principals of Hemodynamic Monitoring 153
in both high-risk surgical patients [50] and trauma patients [51, 52] failed to
document that any improved survival when patients were treated based
on pulmonary arterial catheter-derived data. In fact, the patients resusci-
tated aggressively to force DO
2
into these survivor levels suffered a much
higher mortality rate that did the control group treated conservatively [3].
Interestingly, as mentioned above, using only arterial pressure and superior
vena caval O
2
saturation, but with a defined physiology-based treatment algo-
rithm Rivers et al. [40] demonstrated a markedly improved survival in septic
shock patients without the need of PAC. On the other hand, a recent statistical
analysis that includes over 50,000 patients of the National Trauma Data Bank
showed for the first time a decrease in mortality in a selective group of trauma
patients (severely injured, elderly, who arrived in shock) with the use of
PAC [53].
Because of this nonclear benefit of the use of PAC, the fact of it being an
invasive monitoring procedure with potential serious complications acquires a
major relevance when deciding on the risk-benefit indicating its use. Two recent
large prospective multicenter studies showed an incidence of 5 and 10% of
complications [54, 55]. The most frequent complications described in this
series were hematomas, arterial puncture, arrhythmias, and PAC-related infec-
tions, although a long list of complications has been described. No deaths
attributable to PAC were found in this series but other authors had previously
reported mortality generally due to right heart and pulmonary artery perfora-
tion [56, 57].
Beyond the controversial use of the PAC, two recent randomized clinical
trials using active protocols of hemodynamic monitoring and algorithms of
goal-directed therapy guided by esophageal Doppler flowmetry [58] and pulse
contour analysis for cardiac output [59] in postoperative surgical patients had
shown a decreased duration of hospital stay and morbidity. Thus, the literature
suggests that the generalized use of hemodynamic monitoring and aggressive
goal-directed therapy could improve outcome but that one does not need to use
a PAC to achieve these goals. However, the fact that these entire arguments miss
the point of the utility of hemodynamic monitoring is relative, namely, that no
monitoring device, no matter how accurate, safe and simple to use, will improve
outcome unless coupled to a treatment, which itself, improves outcome. Thus,
the question should not be, ‘Does the PAC improve outcome?’ but rather, ‘Do
treatment protocols that require information only attainable from pulmonary
arterial catheterization improve outcome?’ Furthermore, the treatment protocol
itself should also be shown to improve outcome prior to the study, because oth-
erwise if the trail shows no difference in outcome with or without a PAC, the
results may well reflect the fact that there was no benefit for the protocol in
either arm of the study.
Polanco/Pinsky 154
Conclusion
All surgical patients require monitoring to assess cardiovascular stability
and sometimes may benefit from optimization of their hemodynamic status.
Therefore, all surgeons require a basic understanding of physiological under-
pinnings of hemodynamic monitoring. The physiological rationale is still the
primary level of defense for monitoring critically ill patients.
Arterial catheterization to monitor arterial pressure is a safe procedure with
a low complication rate. However, it should be used only when clear indications
exist. There is no evidence that achieving pressures over 65 mm Hg increases
organ perfusion or favors outcome. The analysis of pulse pressure variation is a
useful method to assess preload responsiveness and a potential tool for resuscita-
tion. CVP has being wrongly used as a parameter of goal for replacement of
intravascular volume in shock patients. Volume loading in patients with CVP
⬎12 mmHg is unlikely to increase cardiac output and attempts to normalize
CVP in early goal-directed therapy during resuscitation has no proven benefit.
The use of PAC provides direct access to several physiological parameters, both
as raw data and derived measurements (CO, SvO
2
, DO
2
). At the present targeting
specific levels of DO
2
have proven effective only in high-risk surgery patients in
the perioperative time. Ppao is often used as bedside assessment of pulmonary
edema, pulmonary vasomotor tone, intravascular volume status and LV preload,
and LV performance. Several publications have explored the potential indica-
tions and benefits of the PAC in goal-directed therapies. Beyond this controversy
there is a trend to less invasive methods of hemodynamic monitoring and current
data support protocols of monitoring and goal-directed therapy that could
improve outcome in selected groups of surgical patients.
Acknowledgment
This work was supported by grant federal funding HL67181 and HL0761570.
References
1 Bellomo R, Pinsky MR: Invasive monitoring; in Tinker J, Browne D, Sibbald W (eds): Critical
Care – Standards, Audit and Ethics. Oxford, University Press, 2006, pp 82–104.
2 Ledoux D, Astiz ME, Carpati CM, Rackow EC: Effects of perfusion pressure on tissue perfusion
in septic shock. Crit Care Med 2000;28:2729–2732.
3 Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D: Elevation of systemic oxygen
delivery in the treatment of critically ill patients. N Engl J Med 1994;330:1717–1722.
4 Bickell WH, Wall MJ Jr, Pepe PE, et al: Immediate versus delayed fluid resuscitation for hypoten-
sive patients with penetrating torso injuries. N Engl J Med 1994;331:1105–1109.
Principals of Hemodynamic Monitoring 155
5 Capone AC, Safar P, Stezoski W, Tisherman S, Peitzman AB: Improved outcome with fluid restric-
tion in treatment of uncontrolled hemorrhagic shock. J Am Coll Surg 1995;180:49–56.
6 Kowalenko T, Stern S, Dronen S, Wang X: Improved outcome with hypotensive resuscitation of
uncontrolled hemorrhagic shock in a swine model. J Trauma 1992;33:349–353.
7 Rebuck AS, Read J: Assessment and management of severe asthma. Am J Med 1971;51:788–798.
8 Michard F, Boussat S, Chemla D, et al: Relation between respiratory changes in arterial pulse pres-
sure and fluid responsiveness in septic patients with acute circulatory failure. Am J Respir Crit
Care Med 2000;162:134–138.
9 Monnet X, Rienzo M, Osman D, et al: Passive leg raising predicts fluid responsiveness in the crit-
ically ill. Crit Care Med 2006;34:1402–1407.
10 Gunn SR, Pinsky MR: Implications of arterial pressure variation in patients in the intensive care
unit. Curr Opin Crit Care 2001;7:212–217.
11 Michard F, Teboul JL: Predicting fluid responsiveness in ICU patients: a critical analysis of the
evidence. Chest 2002;121:2000–2008.
12 Scheer B, Perel A, Pfeiffer UJ: Clinical review: complications and risk factors of peripheral arte-
rial catheters used for haemodynamic monitoring in anaesthesia and intensive care medicine. Crit
Care 2002;6:199–204.
13 Michard F, Alaya S, Zarka V, Bahloul M, Richard C, Teboul JL: Global end-diastolic volume as an
indicator of cardiac preload in patients with septic shock. Chest 2003;124:1900–1908.
14 Godje O, Peyerl M, Seebauer T, Lamm P, Mair H, Reichart B: Central venous pressure, pulmonary
capillary wedge pressure and intrathoracic blood volumes as preload indicators in cardiac surgery
patients. Eur J Cardiothorac Surg 1998;13:533–539.
15 Buhre W, Weyland A, Schorn B, et al: Changes in central venous pressure and pulmonary capillary
wedge pressure do not indicate changes in right and left heart volume in patients undergoing coro-
nary artery bypass surgery. Eur J Anaesthesiol 1999;16:11–17.
16 Magder S, Georgiadis G, Tuck C: Respiratory variations in right atrial pressure predict response to
fluid challenge. J Crit Care 2004;7:76–85.
17 Bafaqeeh F, Magder S: CVP and volume responsiveness of cardiac output (abstract). Am J Respir
Crit Care Med 2004;169:A343.
18 Venn R, Steele A, Richardson P, Poloniecki J, Grounds M, Newman P: Randomized controlled
trial to investigate influence of the fluid challenge on duration of hospital stay and perioperative
morbidity in patients with hip fractures. Br J Anaesth 2002;88:65–71.
19 Thomsen HS, Lokkegaard H, Munck O: Influence of normal central venous pressure on onset of
function in renal allografts. Scand J Urol Nephrol 1987;21:143–145.
20 Shoemaker WC, Kram HB, Appel PL, Fleming AW: The efficacy of central venous and pul-
monary artery catheters and therapy based upon them in reducing mortality and morbidity. Arch
Surg 1990;125:1332–1337.
21 Sharkey SW: Beyond the wedge: clinical physiology and the Swan-Ganz catheter. Am J Med
1987;83:111–122.
22 Swan HJ, Ganz W, Forrester J, Marcus H, Diamond G, Chonette D: Catheterization of the heart in
man with use of a flow-directed balloon-tipped catheter. N Engl J Med 1970;283:447–451.
23 Hoyt JD, Leatherman JW: Interpretation of the pulmonary artery occlusion pressure in mechani-
cally ventilated patients with large respiratory excursions in intrathoracic pressure. Intensive Care
Med 1997;23:1125–1131.
24 Pinsky M, Vincent JL, De Smet JM: Estimating left ventricular filling pressure during positive
end-expiratory pressure in humans. Am Rev Respir Dis 1991;143:25–31.
25 Pinsky MR: Clinical significance of pulmonary artery occlusion pressure. Intensive Care Med
2003;29:175–178.
26 Abraham AS, Cole RB, Green ID, Hedworth-Whitty RB, Clarke SW, Bishop JM: Factors con-
tributing to the reversible pulmonary hypertension of patients with acute respiratory failure stud-
ies by serial observations during recovery. Circ Res 1969;24:51–60.
27 Forrester JS, Diamond G, Chatterjee K, Swan HJ: Medical therapy of acute myocardial infarction
by application of hemodynamic subsets (first of two parts). N Engl J Med 1976;295:1356–1362.
28 Raper R, Sibbald WJ: Misled by the wedge? The Swan-Ganz catheter and left ventricular preload.
Chest 1986;89:427–434.
Polanco/Pinsky 156
29 Kumar A, Anel R, Bunnell E, et al: Pulmonary artery occlusion pressure and central venous pres-
sure fail to predict ventricular filling volume, cardiac performance, or the response to volume
infusion in normal subjects. Crit Care Med 2004;32:691–699.
30 Jansen JR, Bogaard JM, Versprille A: Extrapolation of thermodilution curves obtained during a
pause in artificial ventilation. J Appl Physiol 1987;63:1551–1557.
31 Synder JV, Powner DJ: Effects of mechanical ventilation on the measurement of cardiac output by
thermodilution. Crit Care Med 1982;10:677–682.
32 Wesseling K, Wit BD, Weber J, Smith NT: A simple device for the continuous measurement of car-
diac output. Adv Cardiovasc Physiol 1983;5:16–52.
33 Feissel M, Michard F, Mangin I, Ruyer O, Faller JP, Teboul JL: Respiratory changes in aortic blood
velocity as an indicator of fluid responsiveness in ventilated patients with septic shock. Chest
2001;119:867–873.
34 Reuter DA, Felbinger TW, Schmidt C, et al: Stroke volume variations for assessment of cardiac
responsiveness to volume loading in mechanically ventilated patients after cardiac surgery.
Intensive Care Med 2002;28:392–398.
35 Reuter DA, Felbinger TW, Kilger E, Schmidt C, Lamm P, Goetz AE: Optimizing fluid therapy in
mechanically ventilated patients after cardiac surgery by on-line monitoring of left ventricular
stroke volume variations. Comparison with aortic systolic pressure variations. Br J Anaesth 2002;88:
124–126.
36 Berkenstadt H, Margalit N, Hadani M, et al: Stroke volume variation as a predictor of fluid
responsiveness in patients undergoing brain surgery. Anesth Analg 2001;92:984–989.
37 Pinsky MR, Payen D: Functional hemodynamic monitoring. Crit Care 2005;9:566–572.
38 Linton R, Band D, O’Brien T, Jonas M, Leach R: Lithium dilution cardiac output measurement:
a comparison with thermodilution. Crit Care Med 1997;25:1796–1800.
39 Kurita T, Morita K, Kato S, Kikura M, Horie M, Ikeda K: Comparison of the accuracy of the
lithium dilution technique with the thermodilution technique for measurement of cardiac output.
Br J Anaesth 1997;79:770–775.
40 Rivers E, Nguyen B, Havstad S, et al: Early goal-directed therapy in the treatment of severe sepsis
and septic shock. N Engl J Med 2001;345:1368–1377.
41 Reinhart K, Kuhn HJ, Hartog C, Bredle DL: Continuous central venous and pulmonary artery
oxygen saturation monitoring in the critically ill. Intensive Care Med 2004;30:1572–1578.
42 Marik PE: Gastric intramucosal pH. A better predictor of multiorgan dysfunction syndrome and
death than oxygen-derived variables in patients with sepsis. Chest 1993;104:225–229.
43 Pinsky MR: Beyond global oxygen supply-demand relations: in search of measures of dysoxia.
Intensive Care Med 1994;20:1–3.
44 Kandel G, Aberman A: Mixed venous oxygen saturation. Its role in the assessment of the critically
ill patient. Arch Intern Med 1983;143:1400–1402.
45 Miller MJ, Cook W, Mithoefer J: Limitations of the use of mixed venous pO
2
as an indicator of
tissue hypoxia. Clin Res 1979;27:401A.
46 Pinsky MR: The meaning of cardiac output. Intensive Care Med 1990;16:415–417.
47 Weissman C, Kemper M, Damask MC, Askanazi J, Hyman AI, Kinney JM: Effect of routine inten-
sive care interactions on metabolic rate. Chest 1984;86:815–818.
48 Tuchschmidt J, Fried J, Astiz M, Rackow E: Elevation of cardiac output and oxygen delivery
improves outcome in septic shock. Chest 1992;102:216–220.
49 Boyd O, Grounds RM, Bennett ED: A randomized clinical trial of the effect of deliberate periop-
erative increase of oxygen delivery on mortality in high-risk surgical patients. JAMA 1993;270:
2699–2707.
50 Sandham JD, Hull RD, Brant RF, et al: A randomized, controlled trial of the use of pulmonary-
artery catheters in high-risk surgical patients. N Engl J Med 2003;348:5–14.
51 McKinley BA, Kozar RA, Cocanour CS, et al: Normal versus supranormal oxygen delivery goals
in shock resuscitation: the response is the same. J Trauma 2002;53:825–832.
52 Velmahos GC, Demetriades D, Shoemaker WC, et al: Endpoints of resuscitation of critically injured
patients: normal or supranormal? A prospective randomized trial. Ann Surg 2000;232:409–418.
Principals of Hemodynamic Monitoring 157
53 Friese RS, Shafi S, Gentilello LM: Pulmonary artery catheter use is associated with reduced mor-
tality in severely injured patients: a National Trauma Data Bank analysis of 53,312 patients. Crit
Care Med 2006;34:1597–1601.
54 Binanay C, Califf RM, Hasselblad V, et al: Evaluation study of congestive heart failure and pul-
monary artery catheterization effectiveness: the ESCAPE trial. JAMA 2005;294:1625–1633.
55 Harvey S, Harrison DA, Singer M, et al: Assessment of the clinical effectiveness of pulmonary
artery catheters in management of patients in intensive care (PAC-Man): a randomised controlled
trial. Lancet 2005;366:472–477.
56 Ducatman BS, McMichan JC, Edwards WD: Catheter-induced lesions of the right side of the
heart. A one-year prospective study of 141 autopsies. JAMA 1985:253:791–795.
57 Kearney TJ, Shabot MM: Pulmonary artery rupture associated with the Swan-Ganz catheter. Chest
1995;108:1349–1352.
58 McKendry M, McGloin H, Saberi D, Caudwell L, Brady AR, Singer M: Randomised controlled
trial assessing the impact of a nurse delivered, flow monitored protocol for optimisation of circu-
latory status after cardiac surgery. BMJ 2004;329:258.
59 Pearse R, Dawson D, Fawcett J, Rhodes A, Grounds RM, Bennett ED: Early goal-directed therapy
after major surgery reduces complications and duration of hospital stay. A randomised, controlled
trial [ISRCTN38797445]. Crit Care 2005;9:R687–R693.
60 Schlichtig R, Kramer DJ, Boston JR, Pinsky MR: Renal O
2
consumption during progressive hem-
orrhage. J Appl Physiol 1991;70:1957–1962.
Michael R. Pinsky, MD
606 Scaife Hall
3550 Terrace Street
Pittsburgh, PA 15261 (USA)
Tel./Fax ⫹1 412 647 5387, E-Mail pinsky@pitt.edu
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 158–166
Acid-Base Disorders and Strong Ion Gap
John A. Kellum
Department of Critical Care Medicine, University of Pittsburgh Medical Center,
Pittsburgh, Pa., USA
Abstract
The application of modern quantitative physical chemical techniques to clinical acid-base
has yielded important new information about the nature and clinical significance of metabolic
acid-base disorders. Abnormalities identified by the strong ion gap appear to be common in
critically ill patients and are associated with increased mortality especially when identified
early in the course of critical illness. Attempts to identify the exact chemical nature of ions
identified by the strong ion gap have only been of limited success and further study is needed.
Copyright © 2007 S. Karger AG, Basel
Acid-base balance is among the most tightly regulated variables in human
physiology. Acute changes in blood pH induce powerful effects at the level of
the cell, organ, and organism [1]. Yet the mechanisms responsible for local,
regional and systemic acid-base control are incompletely understood and con-
troversy exists in the literature as to what methods should be used to understand
them [2]. Once basic categorization into respiratory and metabolic acid-base
disorders is accomplished, evaluation of ion balance is undertaken in order to
identify complex acid-base disorders and to help narrow the differential diag-
nosis. Evaluation of ion balance can be done using the familiar anion gap (AG)
or the slightly more complex strong ion gap (SIG). Both techniques are based
on the principle of electrical neutrality which dictates that in macroscopic aque-
ous solutions (like blood plasma) the sum of all positive charges (cations) must
equal the sum of all negative charges (anions).
Anion Gap
The AG is calculated, or rather estimated, from the differences between the
routinely measured concentrations of serum cations (Na
⫹
and K
⫹
) and anions
Acid-Base Disorders and Strong Ion Gap 159
(Cl
⫺
and HCO
3
⫺
). Since there can be no actual difference (electrical neutrality
must be preserved), the measured difference reflects missing, or ‘unmeasured’
ions. Normally, this difference or ‘gap’ is filled primarily by the ionized portion
of the weak acids (A
⫺
) principally albumin, and, to a lesser extent, phosphate.
Sulfate and lactate also contribute a small amount to the normal AG, typically
less than 2 mEq/l. However, there are also unmeasured cations such as Ca
2⫹
and
Mg
2⫹
and these tend to offset the effects of sulfate and lactate except when
either is abnormally increased. Plasma proteins other than albumin can be
either positively or negatively charged but in the aggregate tend to be neutral [3]
except in rare cases of abnormal paraproteins such as in multiple myeloma.
When the AG is greater than that produced by albumin and phosphate,
other anions (e.g. lactate, ketones) must be present in higher than normal con-
centrations. In this way, the AG can be used to narrow the differential diagnosis
in the case of a metabolic acidosis. Furthermore, the magnitude of the AG
reflects the concentration of the offending acid and can therefore provide a
means of monitoring when measurement of the acid is difficult (e.g. ketoacido-
sis). Finally the AG may provide a clue to the presence of life-threatening con-
ditions such as poisonings.
In practice the AG is calculated as follows:
AG ⫽ (Na
⫹
⫹ K
⫹
) ⫺ (Cl
⫺
⫹ HCO
3
⫺
)
Because of its low and narrow extracellular concentration, K
⫹
is often
omitted from the calculation. Respective normal values with relatively wide
ranges reported by most laboratories are 12 ⫾ 4 mEq/l (if K
⫹
is considered) and
8 ⫾ 4 mEq/l (if K
⫹
is not considered). The value of a ‘normal AG’ has
decreased in recent years following the introduction of more accurate methods
for measuring Cl
⫺
concentration [4, 5]. However, the various measurement
techniques available mandate that each institution reports its own expected nor-
mal AG.
Importantly, the concept of a normal AG is based on the premise that A
⫺
is
normal, which requires albumin and phosphate, the two major constituents of
the A
⫺
, to be normal, both in concentration and in charge. As it turns out this is
usually the case in healthy subjects but rarely in critically ill patients [6].
Dehydration may induce a parallel increment in the apparent AG by increasing
the concentration of all the ions. Conversely, severe hypoalbuminemia causes a
decrease in the AG and it has been recommended to ‘correct’ the AG for the
prevailing albumin concentration since each gram per deciliter decline in serum
albumin reduces the apparent AG by 2.5–3 mEq/l [7].
Some authors have cast doubt on the diagnostic value of the AG in certain
situations [6, 8]. For example, Salem and Mujais [6] found routine reliance on
the AG to be ‘fraught with numerous pitfalls’. The primary problem with the
Kellum 160
AG is its reliance on the use of a normal range produced by albumin and to a
lesser extent phosphate as discussed above. These constituents may be grossly
abnormal in patients with critical illness leading to a change in the normal
range for these patients. This has prompted some authors to adjust the ‘normal
range’ for the AG by the patient’s albumin [7] or even phosphate [9] concentra-
tion. Because these anions are not strong anions their charge will be altered by
changes in pH. Each gram per deciliter of albumin has a charge of 2.8 mEq/l at
pH 7.4 (2.3 mEq/l at 7.0 and 3.0 mEq/l at 7.6) and each milligram per deciliter
of phosphate has a charge of 0.59 mEq/l at pH 7.4 (0.55 mEq/l at 7.0 and
0.61 mEq/l at 7.6). Thus, under physiological conditions, the variance is reason-
ably small. A convenient way to estimate the normal AG for a given patient is by
use of the following formula [9]:
normal AG ⫽ 2 (albumin g/dl) ⫹ 0.5 (phosphate mg/dl)
or for international units:
normal AG ⫽ 0.2 (albumin g/l) ⫹ 1.5 (phosphate mmol/l)
When this patient-specific normal range was used to examine the presence
of unmeasured anions in the blood of critically ill patients, the accuracy of this
method improved from 33% with the routine AG (normal range ⫽ 12 mEq/l) to
96% [9].
Alternatively, the estimated charge coming from albumin and phosphate
can be added with Cl
⫺
and HCO
3
⫺
as total anions. Lactate can also be consid-
ered and the resultant ‘corrected AG’ (cAG) should be close to zero.
cAG ⫽ (Na
⫹
⫹ K
⫹
) ⫺ [Cl
⫺
⫹ HCO
3
⫺
⫹ 2 (albumin) ⫹ 0.5 (phosphate) ⫹ lactate]
or for international units:
cAG ⫽ (Na
⫹
⫹ K
⫹
) ⫺ (Cl
⫺
⫹ HCO
3
⫺
⫹ 0.2 (albumin) ⫹ 1.5 (phosphate) ⫹ lactate)
Either technique is only accurate within about 5 mEq/l. When more accu-
racy is desired a slightly more complicated method of estimating A
⫺
is required
[10].
Strong Ion Gap
About a decade ago [10], our laboratory applied newly published data by
Figge et al. [3] concerning the net charge on the surface of albumin to ideas
proposed by Stewart [11] a decade earlier. The idea was to compare two meth-
ods of estimating the total charge difference between plasma cations and anions
known as the strong ion difference (SID). The first method, known as the
‘apparent’ SID (SIDa), was to simply measure as many strong (completely or
Acid-Base Disorders and Strong Ion Gap 161
near completely dissociated) cations and anions as possible and sum their
charges. The second was to estimate the SID from the partial pressure of CO
2
(from which HCO
3
⫺
and CO
3
2⫺
can be estimated) and the concentration of the
weak acids (mostly albumin and phosphate as globulins are both cationic and
anionic and in healthy humans their net charge is near zero). This second esti-
mate of SID is termed the ‘effective’ SID (SIDe). Neither estimate is exact.
While considering all of the usual electrolytes and lactate, the SIDa will ‘miss’
strong ions such as ketones and sulfate because they are not measured.
Similarly, the SIDe is only an accurate estimate of SID if there are not signifi-
cant amounts of unmeasured weak acids (e.g. proteins other than albumin
or normal globulins) and if albumin and globulins are themselves normal
in charge, conformation and composition. Neither of these seem very likely in
critically ill patients so neither SIDa nor SIDe should be assumed to be equal to
the SID. However, both SIDa and SIDe should equal SID and hence be equal
to each other in healthy plasma. When SIDa is not equal to SIDe, some unmea-
sured anions or cations must be present. We termed this difference the SIG to
distinguish it from the AG [10]. By convention, SIDa – SIDe ⫽ SIG and hence
SIG is ‘positive’ when unmeasured anions are present in excess of unmeasured
cations, and negative when unmeasured cations exceed unmeasured anions.
Unfortunately, the name SIG might seem to imply that strong ions are involved.
Yet, as detailed above, either strong or weak ions, or both, may produce a ‘gap’
between these two complementary estimates of SID. The SIG cannot tell us
which.
SIDa ⫽ (Na
⫹
⫹ K
⫹
⫹ Ca
2⫹
⫹ Mg
2⫹
) ⫺ (Cl
⫺
⫹ lactate)
SIDe ⫽ 2.46 ⭈ 10
⫺8
⭈ PCO
2
/10
⫺pH
⫹ [albumin] ⭈ (0.123 ⭈ pH ⫺ 0.631) ⫹ [PO
4
2⫺
]
⭈ (0.39 ⭈ pH ⫺0.469)
SIG ⫽ SIDa ⫺ SIDe
Interpreting the Gaps
The utility of the AG/SIG comes primarily from its ability to quickly and
easily limit the differential diagnosis in a patient with metabolic acidosis. If an
increased AG/SIG is present, the explanation will often be found among five
disorders: ketosis, lactic acidosis, poisoning, renal failure and sepsis. Table 1
provides a list, including other disorders associated with an increased AG/SIG.
A number of factors may influence the AG apart from those corrected for
above. Respiratory and metabolic alkalosis are associated with an increase of
up to 3–10 mEq/l in the apparent AG following an enhanced lactate production
Kellum 162
(from stimulated phosphofructokinase enzymatic activity), the reduction in the
ionized weak acids (A
⫺
) and possibly, the additional effect of dehydration (with
its own impact on AG calculation). Low Mg
2⫹
concentration with associated
low K
⫹
and Ca
2⫹
concentrations, as well as the administration of sodium salts
of poorly reabsorbable anions (such as -lactam antibiotics) are known causes
of an increased AG [12]. Certain parenteral nutrition formulations, such as
those containing acetate, may increase both the AG and the SIG and citrate may
rarely have the same effect in the setting of multiple blood transfusions particu-
larly if massive doses of banked blood are used, such as during liver transplan-
tation [13]. None of these rare causes, however, will increase the AG or SIG
significantly, and they are usually easily identified.
Table 1. Causes of an increased AG and SIG
Common causes
Renal failure
Ketoacidosis
Diabetic
Alcoholic
Starvation
Metabolic errors
Lactic acidosis
1
Toxins
Methanol
Ethylene glycol
Salicylates
Paraldehyde
Toluene
Rare causes
Dehydration
Sodium salts
Sodium lactate
1
Sodium citrate
Sodium acetate
Sodium PCN (⬎50 mU/day)
Carbenicilin (⬎30 g/day)
Decreased unmeasured cation
Hypomagnesemia
1
Hypocalcemia
1
Alkalemia
1
Already accounted for by SIG.
Acid-Base Disorders and Strong Ion Gap 163
We expected to find very little, if any, unmeasured ions in the blood of nor-
mal humans and by applying our methodology to a published dataset of healthy
exercising subjects [14] we found, or rather, did not find very much at all. The
total unmeasured anions in the blood of these subjects was a mere 0.3 ⫾
0.6 mEq/l [10]. However, unlike healthy exercising subjects or normal labora-
tory animals [15, 16], critically ill patients seem to have much higher SIG val-
ues [17–22]. Recently, there has been controversy as to what constitutes a
‘normal’ SIG and as to whether an abnormal SIG is associated with adverse
clinical outcomes. Reports from the United States [17, 18, 22] and from
Holland [19] have found that the SIG was close to 5 mEq/l in critically ill
patients while studies from England and Australia [20, 21] have found much
higher values. One might speculate that the difference may lie with the use of
gelatins (an exogenous source of unmeasured ions [23]) for resuscitation in
these countries [24]. In this scenario, the SIG is likely to be a mixture of
endogenous and exogenous anions. Interestingly, these two studies involving
patients receiving gelatins [20, 21] have failed to find a correlation between
SIG and mortality while studies in patients not receiving gelatins [17, 18, 25] a
positive correlation between SIG and hospital mortality has been found. Indeed
Kaplan and Kellum [18] have recently reported that preresuscitation SIG pre-
dicts mortality in injured patients better than blood lactate, pH or injury sever-
ity scores. Dondorp et al. [25] had similar results with preresuscitation SIG as a
strong mortality predictor in patients with severe malaria. More recently,
Durward et al. [26] report yet another instance when SIG and mortality corre-
late in patients not receiving gelatins. These authors also found that SIG, at
admission to the ICU, was superior to lactate and other acid-base variables in
terms of predicting subsequent hospital survival. In this study SIG had equal
predictive accuracy to the pediatric index of mortality score, a risk prediction
tool that comprises eight variables including base excess.
Etiology of SIG
While numerous studies have identified clinical conditions associated with
unmeasured anions, the exact chemical nature of these substances is unknown.
Unmeasured anions have been reported in the blood of patients with sepsis [8,
27] and liver disease [10, 28] and in experimental animals given endotoxin [15].
These anions may be the source of much of the unexplained acidosis seen in
patients with critical illness. However, the very idea that something is happen-
ing during cardiac surgery [26], during the early stages of major vascular injury
[18] and during malarial sepsis [25] as well during other types of critical illness
[17] that results in the release of anions that correlate with subsequent mortality
Kellum 164
is astonishing, especially since we do not know what these anions are. Given
that individual patients may have SIG values of more than 10–15 mEq/l, it
seems unlikely that any strong ion could be present in the plasma at these con-
centrations and be unknown to us. Yet, it seems stranger still, for weak acids
such as proteins to be the cause given that they are, in fact, weak. In healthy
subjects the total charge concentration of plasma albumin is only about
10–12 mEq/l. For a similarly charged protein to affect a SIG of 15 mEq/l, it
would need to be present in very large quantities indeed. Recent attempts to
determine the etiology of SIG in critically ill patients reveal that although cer-
tain low molecular weight anions usually associated with intermediary metabo-
lism are found to be significantly elevated in the plasma obtained from patients
with metabolic acidosis [29] the overall concentration of these molecules
explains less than 50% of the observed SIG.
The answer, probably, is that the identity of the SIG in these patients is
multifactorial. Endogenous strong ions such a ketones and sulfate are added to
exogenous ones such as acetate and citrate. Reduced metabolism of these and
other ions owing to liver [15] and kidney [30] dysfunction likely exacerbates
this situation. The release of a myriad of acute phase proteins, principally from
the liver, in the setting of critical illness and injury likely adds to the SIG.
Furthermore, the systemic inflammatory response is associated with the release
of a substantial quantity of proteins including cytokines and chemokines some
of which, like high-mobility group B1, have been linked to mortality [31]. The
cumulative effect of all of these factors may well be a reflection of both organ
injury and dysfunction. It is perhaps not surprising that there is a correlation
between SIG and mortality.
However, whatever the source of SIG, it appears that its presence in the cir-
culation, especially early in the course of illness or injury, portends a poor prog-
nosis. While the prognostic significance of SIG is reduced (or abolished) when
exogenous unmeasured anions are administered (e.g. gelatins), a SIG acidosis
seems to be far worse than a similar amount of hyperchloremic acidosis and
more like lactic acidosis in terms of significance [26, 32]. Although it is possible
that saline-based resuscitation fluids contaminate the prognostic value of hyper-
chloremia the same way gelatins appear to confound SIG, there remains strong
evidence that not all metabolic acidoses are the same.
References
1 Kellum JA: Diagnosis and treatment of acid-base disorders; in Grenvik A, Ayres SM, Holbrook
PR, Shoemaker WC (eds): Textbook of Critical Care. Philadelphia, Saunders, 2000, pp 839–853.
2 Severinghaus JW: Siggard-Andersen and the ‘great trans-Atlantic acid-base debate’. Scand J Clin
Lab Invest 1993;53(suppl 214):99–104.
Acid-Base Disorders and Strong Ion Gap 165
3 Figge J, Mydosh T, Fencl V: Serum proteins and acid-base equilibria: a follow-up. J Lab Clin Med
1992;120:713–719.
4 Sadjadi SA: A new range for the anion gap. Ann Intern Med 1995;123:807–808.
5 Winter SD, Pearson R, Gabow PG, Schultz A, Lepoff RB: The fall of the serum anion gap. Arch
Intern Med 1990;150:3113–3115.
6 Salem MM, Mujais SK: Gaps in the anion gap. Arch Intern Med 1992;152:1625–1629.
7 Gabow PA: Disorders associated with an altered anion gap. Kidney Int 1985;27:472–483.
8 Mecher C, Rackow EC, Astiz ME, Weil MH: Unaccounted for anion in metabolic acidosis during
severe sepsis in humans. Crit Care Med 1991;19:705–711.
9 Kellum JA: Determinants of blood pH in health and disease. Crit Care 2000;4:6–14.
10 Kellum JA, Kramer DJ, Pinsky MR: Strong ion gap: a methodology for exploring unexplained
anions. J Crit Care 1995;10:51–55.
11 Stewart P: Modern quantitative acid-base chemistry. Can J Physiol Pharmacol 1983;61:
1444–1461.
12 Whelton A, Carter GG, Garth M: Carbenicillin-induced acidosis and seizures. JAMA 1971;218:
1942–1944.
13 Kang Y, Aggarwal S, Virji M: Clinical evaluation of autotransfusion during liver transplantation.
Anesth Analg 1991;72:94–100.
14 Lindinger MI, Heigenhauser GJF, McKelvie RS, Jones NL: Blood ion regulation during repeated
maximal exercise and recovery in humans. Am J Physiol 1992;262:R126–R136.
15 Kellum JA, Bellomo R, Kramer DJ, Pinsky MR: Hepatic anion flux during acute endotoxemia.
J Appl Physiol 1995;78:2212–2217.
16 Kellum JA, Bellomo R, Kramer DJ, Pinsky MR: Etiology of metabolic acidosis during saline
resuscitation in endotoxemia. Shock 1998;9:364–368.
17 Balasubramanyan N, Havens PL, Hoffman GM: Unmeasured anions identified by the Fencl-
Stewart method predict mortality better than base excess, anion gap, and lactate in patients in the
pediatric intensive care unit. Crit Care Med 1999;27:1577–1581.
18 Kaplan L, Kellum JA: Initial pH, base deficit, lactate, anion gap, strong ion difference, and strong
ion gap predict outcome from major vascular injury. Crit Care Med 2004;32:1120–1124.
19 Moviat M, van Haren F, van der Hoeven H: Conventional or physicochemical approach in inten-
sive care unit patients with metabolic acidosis. Crit Care 2003;7:R41–R45.
20 Cusack RJ, Rhodes A, Lochhead P, Jordan B, Perry S, Ball JAS, et al: The strong ion gap does not
have prognostic value in critically ill patients in a mixed medical/surgical adult ICU. Intensive
Care Med 2002;28:864–869.
21 Rocktaschel J, Morimatsu H, Uchino S, Bellomo R: Unmeasured anions in critically ill patients:
can they predict mortality? Crit Care Med 2003;31:2131–2136.
22 Gunnerson KJ, Roberts G, Kellum JA: What is normal strong ion gap (SIG) in healthy subjects
and critically ill patients without acid-base abnormalities (abstract). Crit Care Med 2003;31(12
suppl):A111.
23 Hayhoe M, Bellomo R, Liu G, Kellum JA, McNicol L, Buxton B: Role of the splanchnic circula-
tion in acid-base balance during cardiopulmonary bypass. Crit Care Med 1999;27:2671–2677.
24 Kellum JA: Closing the gap on unmeasured anions. Crit Care 2003;7:219–220.
25 Dondorp AM, Chau TT, Phu NH, Mai NT, Loc PP, Chuong LV, et al: Unidentified acids of strong
prognostic significance in severe malaria. Crit Care Med 2004;32:1683–1688.
26 Durward A, Tibby SM, Skellett S, Austin C, Anderson D, Murdoch IA: The strong ion gap predicts
mortality in children following cardiopulmonary bypass surgery. Pediatr Crit Care Med 2005;6:
281–285.
27 Gilfix BM, Bique M, Magder S: A physical chemical approach to the analysis of acid-base bal-
ance in the clinical setting. J Crit Care 1993;8:187–197.
28 Kirschbaum B: Increased anion gap after liver transplantation. Am J Med Sci 1997;313:107–110.
29 Forni LG, McKinnon W, Lord GA, Treacher DF, Peron JM, Hilton PJ: Circulating anions usually
associated with the Krebs cycle in patients with metabolic acidosis. Crit Care 2005;9:R591–R595.
30 Rocktaschel J, Morimatsu H, Uchino S, Goldsmith D, Poustie S, Story D, et al: Acid-base status of
critically ill patients with acute renal failure: analysis based on Stewart-Figge methodology. Crit
Care 2003;7:R60–R66.
Kellum 166
31 Wang H, Bloom O, Zhang M, Vishnubhakat JM, Ombrellino M, Che J, et al: HMG-1 as a late
mediator of endotoxin lethality in mice. Science 1999;285:248–251.
32 Gunnerson KJ, Saul M, He S, Kellum JA: Lactate versus non-lactate metabolic acidosis: a retro-
spective outcome evaluation of critically ill patients. Crit Care 2006;10:R22–R26.
John A. Kellum, MD
Department of Critical Care Medicine, University of Pittsburgh Medical Center
3550 Terrace Street
Pittsburgh, PA 15216 (USA)
Tel. ⫹1 412 647 6966, Fax ⫹1 412 647 8060, E-Mail kellumja@upmc.edu
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 167–177
Fluid Resuscitation and the Septic Kidney:
The Evidence
Elisa Licari
a
, Paolo Calzavacca
a
, Claudio Ronco
b
, Rinaldo Bellomo
a
a
Department of Intensive Care, Austin Hospital, Melbourne, Vic., Australia;
b
Department of Nephrology, St. Bortolo Hospital, Vicenza, Italy
Abstract
Acute kidney injury (AKI) is a common complication of severe sepsis. Severe sepsis is
the most common cause of AKI in ICU. The widely accepted and practiced initial cornerstone
of treatment for septic AKI is fluid resuscitation. The biological rationale for fluid resuscita-
tion in septic AKI is based on the assumption that septic AKI is an ischemic form of AKI and
that increasing renal perfusion and oxygen delivery by means of fluid resuscitation will pro-
tect the kidney. Whether this is true, however, remains uncertain. In this paper, we discuss
salient pathophysiological aspects of AKI, review the evidence available on the need for fluid
resuscitation, the amount and the type of fluid that might be best suited to AKI and discuss all
major aspects of fluid resuscitation for septic AKI in humans and experimental animals.
Copyright © 2007 S. Karger AG, Basel
Sepsis accounts for a similar number of death per year in the United States
as myocardial infarction [1]. It is the most frequent cause of vasodilatory shock
accounting for more than 200,000 cases per year in the United States [2].
Occurrence of admission in ICU is caused by sepsis in 10% of cases, with mor-
tality ranging from 25 to 80% [3].
Sepsis is the most common factor predisposing to acute kidney injury
(AKI) in critically ill patients [4, 5]. It accounts for around 50% of cases [1, 6].
Patients with AKI due to sepsis have a worse prognosis than those with nonsep-
tic AKI [1]. Unfortunately, until recently, the definition of AKI has been a con-
founding element in its epidemiology. This is because the diagnosis of AKI is
complex and involves data obtained from history, biochemical analysis, body
size, sex, hematological information and imaging. The second International
Consensus Conference of the Acute Dialysis Quality Initiative Group in 2002
proposed a classification scheme for AKI to facilitate communication and
research in this field. It produced the so-called RIFLE criteria for the classification
Licari/Calzavacca/Ronco/Bellomo 168
and definition of AKI [7]. These criteria use creatinine and urine output and
consider changes from baseline creatinine value in reaching a classification [7].
The RIFLE classification has now been applied to understanding the epidemi-
ology of AKI and septic AKI in ICU and the findings confirm the high inci-
dence of AKI in ICU, its high mortality and strong association with sepsis [8].
Thus septic AKI is perhaps the biggest physiological and therapeutic challenge
in critical care nephrology. In order to develop a rational approach of its treat-
ment, one needs to understand some aspects of its pathophysiology.
The Pathophysiology of Septic AKI
A decrease in renal blood flow (RBF) causing renal ischemia has been pro-
posed to be central in the pathogenesis of septic AKI [1, 9, 10]. Sepsis induces
increased nitric oxide synthase activity and generates oxygen radicals. Nitric
oxide is believed, in turn, to cause systemic vasodilatation, peroxynitrite-related
tubular injury and downregulation of renal endothelial nitric oxide synthase [1].
The vasodilation that results from the systemic changes induced by severe sep-
sis shifts blood flow from the renal bed to peripheral vascular beds causing
decreased RBF, subsequent ischemia and acute tubal necrosis.
This paradigm mainly relies on animal data because the measurement of
RBF in man is extremely difficult and requires invasive techniques. However, a
recent systematic review [11] of human and animal studies concluded that the
primary determinant of RBF in sepsis is cardiac output (CO). If the CO is
increased, RBF is typically either increased or preserved. CO in human sepsis is
usually elevated. Only three studies conducted in septic ICU patients were
found in which RBF was measured [12–14]. All showed either preserved or
increased RBF. Recent animal experiments [15–18] using a hyperdynamic
model of sepsis also found that RBF in sepsis was increased due to vasodilation
of renal circulation of both afferent and efferent arteriole. Thus, the pathophys-
iology of septic AKI might not necessarily involve ischemia (acute tubular
necrosis), especially in hyperdynamic sepsis.
These observations have important repercussions on our understanding
and biologic rationale for specific kidney protective interventions, especially in
the field of fluid resuscitation.
The Biologic Rationale for Fluid Resuscitation in Septic AKI
Prompt and aggressive fluid resuscitation is considered a cornerstone for
renal protection and preservation of renal function [19]. The rationale for fluid
Fluid Resuscitation and the Septic Kidney 169
resuscitation is that septic systemic vasodilation and capillary leak cause rela-
tive hypovolemia [2]. Relative hypovolemia would then cause decreased vital
organ perfusion and, therefore, decreased oxygen delivery, which, in turn,
would cause organ dysfunction. This would appear to be particularly relevant to
the kidney, which is considered highly sensitive to hypovolemia.
Correction of septic hypovolemia may require continuous and large vol-
umes of fluid administration to maintain renal oxygen delivery above a critical
threshold and to increase or maintain mean arterial pressure to a level that
allows appropriate distribution of CO and adequate organ perfusion to the kid-
ney [20, 21].
In addition, although the kidney seeks to autoregulate its own perfusion,
this ability may be impaired in severe sepsis [22]. While these considerations
support the use of fluid resuscitation, more recent insights into the pathophysi-
ology of septic AKI, as described above, challenge the notion that ischemia is
responsible for septic AKI and, by implication, also challenge the notion that
fluid resuscitation would be of major benefit.
The Evidence of the Renal Benefits of Fluid
Resuscitation in Septic AKI
Although it is widely recognized as a common practice to start cardiovas-
cular resuscitation of septic patients with a fluid challenge to maintain organ
perfusion and thus renal function, no randomized control trial has ever evalu-
ated the effect of fluid challenge or fluid resuscitation in general on kidney
function in septic AKI [23]. A recent evidenced-based review [24] stated the
need to perform a fluid challenge as soon as a hypovolemic state is supported
by only grade E evidence. This means that no evidence has been found to support
this statement, which, therefore, relies on the expert opinion. On the other hand,
recent evidence has started raising caution about the liberal use of fluid in the
management of critically ill patients [25–29].
For example, the ARDS network enrolled 1,000 patients with acute lung
injury in a randomized trial. The patients were randomly assigned to a strategy
involving either conservative or liberal use of fluids. Over a period of 7 days,
the conservative-strategy group received a fluid balance of –136 ⫾ 491 ml as
compared with 6,992 ⫾ 502 ml in the liberal-strategy group (p ⬍ 0.001). More
than 80% of patients enrolled in this study were septic and, therefore, at high
risk of septic AKI due to relative hypovolemia related not only to sepsis but also
to mechanical ventilation and high levels of PEEP. Within the first 60 days,
there were no significant differences in the percentage of patients receiving
renal replacement therapy (10% in the conservative-strategy group vs. 14% in
Licari/Calzavacca/Ronco/Bellomo 170
the liberal-strategy group, p ⫽ 0.06) (fig. 1) although the trend was clearly in
favor of a conservative approach. In conclusion the conservative strategy
improved lung function and shortened the duration of mechanical ventilation
and intensive care, and, possibly, even offered some protection from severe
AKI.
In the SAFE study, a randomized control trial comparing albumin 4% to
normal saline, 6,997 patients were enrolled. Of these, 3,497 were assigned to
receive albumin and 3,500 assigned to receive saline. The 4 days’ fluid bal-
ance was 3 liters in the albumin group and 4 liters in the saline group.
Unfortunately no data about renal function were provided, making it uncer-
tain whether the different choice of fluids and fluid balance affected the like-
lihood of AKI.
The early goal-directed therapy study [30] is a randomized controlled trial
of resuscitation strategies, which enrolled 263 patients. Of these, 130 were ran-
domly assigned to early goal-directed therapy (EGDT) and 133 to standard
therapy. The initial fluids administered over the first 6 h were greater with
EGDT (5 l in EGDT vs. 3.5 l for standard therapy; p ⬍ 0.001). This difference
might be expected to translate to improved renal function and protection from
AKI. Unfortunately, no data about renal function are provided to tell us whether
the EGDT approach affects the likelihood of AKI.
Thus, in the three major clinical trials of fluid therapy and sepsis, no con-
clusive data are available in relation to the septic kidney. However, it is possible
Fig. 1. Histogram presenting the incidence of the need for renal replacement therapy in
ARDS patients treated with liberal versus conservative fluid strategy in a large multicenter
trial.
Conservative
Number
0
50
100
150
200
250
300
350
400
450
500
550
Liberal
Patients
CRRT
Fluid Resuscitation and the Septic Kidney 171
that the type of fluids used by clinicians might affect the likelihood and course
of septic AKI.
Type of Fluid
Colloids and crystalloids might affect renal function differently. Colloids
have long been used in the care of septic patients to raise oncotic pressure [31].
This approach has the rationale of minimizing edema formation [31]. Unfortunately
little is known [32] about their impact on the septic kidney. Some evidence
exists, however, on three major subtypes of colloid fluids: albumin, starches and
gelatins.
Albumin
The use of albumin in intensive care setting is still debated, many studies
having found very conflicting data. However the SAFE trial (saline vs. albumin
fluid evaluation) [33], a multicentric, randomized, controlled trial that enrolled
nearly 7,000 patients, found no differences in new organ failure, urine output or
in the number of renal replacement therapy days in patients treated with albu-
min or saline for fluid resuscitation. This suggests that there is no intrinsic
advantage or disadvantage in terms of AKI from using albumin or saline.
Starches
There are different types of starches available in different countries.
Pharmacokinetics [34] and pharmacodynamics of starches depend on molec-
ular weight (MW) and rate of substitution. All these macromolecules are
metabolized by serum amylases and excreted by the kidney. Hydroxy-
ethylstarch can be classified according to the MW in high MW (450–
480 kDa), medium MW (200kDa) and low MW (70 kDa), or according to the
rate of substitution (high 0.6–0.7 or low 0.4–0.5), or according to the C2/C6
ratio (high ⬎8, low ⬍8), or according to the concentration (high 10%, low
6%). Although definitive studies addressing safety and efficacy of hydrox-
yethylstarch in preventing alterations of renal function in septic patients are
not conclusive, some evidence exists that at least high MW starch prepara-
tions might contribute to AKI.
In a multicenter randomized study by Schortgen et al. [35] published
in 2001, 129 patients received medium MW hydroxyethylstarch 6%. These
investigators found that starch was associated with an increased risk of devel-
oping AKI or need for renal replacement therapy when compared to a 3%
gelatin (OR 2.57). This observation is consistent with data from autopsy find-
ings reporting osmotic nephrosis-like lesions [36] in 80% of kidneys from
Licari/Calzavacca/Ronco/Bellomo 172
donors treated with starch. In a prospective randomized study in patients under-
going renal transplantation, Cittanova et al. [37] found osmotic nephrosis-like
lesions in the kidneys of donors treated with high MW hydroxyethylstarch and
a statistically significant increase in the need for hemodiafiltration and serum
creatinine level compared to kidneys from donors treated with gelatin only.
These findings are consistent with a recent trial [23] (VISEP trial, Efficacy of
Volume Substitution and Insulin Therapy in Severe Sepsis), a randomized com-
parison of crystalloid (Ringer’s lactate) and colloid (10% HES) fluid therapy in
critically ill patients with severe sepsis. Preliminary data indicate that the use of
starch resulted in a significantly higher incidence of AKI.
Gelatin
Although some case reports [32] exist suggesting gelatin can adversely
affect kidney function, there is no controlled trial comparing gelatin solutions
to crystalloid. Comparisons to starches indicate that gelatin solutions might be
safer (see above).
Hypertonic Saline
A review on the use of hypertonic saline for resuscitation [38] suggested a
favorable effect of this therapy. Unfortunately no randomized controlled trial in
septic humans using hypertonic saline has been done [39].
Saline
No trial has ever been performed comparing saline versus no treatment in
humans; neither has a study comparing saline and other crystalloids (Hartmann’s,
lactated solution) ever been performed in the field of septic AKI. The only ran-
domized control trial in septic critically ill patients addressing saline is the
SAFE trial as discussed above. Although it does not give us any insight into the
effectiveness of saline resuscitation in the treatment of septic AKI, we can state
that there is no evidence in that study that saline is either better or worse than
albumin in resuscitating septic patients.
Lactated Solutions
We could not find any trial of crystalloid solution other than normal saline
in sepsis, so there is no evidence available at the moment on whether lactated
solutions can deliver significant advantages or disadvantages in septic AKI.
In conclusion, insufficient human evidence exists that colloids (except for
high MW starch) are better or worse than crystalloids or that a particular type of
crystalloid is better or worse than another. Thus, we can only rely on data gener-
ated from experimental studies of sepsis and extrapolate from the physiological
Fluid Resuscitation and the Septic Kidney 173
effects of fluid resuscitation on systemic and regional hemodynamics and renal
function.
Animal Studies
Given the limitations of the human data available, a review of the animal
data on the effect of fluids in septic models or septic AKI might be helpful.
Unfortunately many biases are present when dealing with animal studies: ani-
mal size, consciousness of animals, time from septic insult, methods of induc-
ing sepsis, and CO, so that it is very hard to compare different studies and draw
conclusions that apply to humans [11].
Only one randomized, unblinded [40] trial compared isotonic or hyper-
tonic fluid therapy with control in a porcine model of endotoxemic shock. In 24
Landrace pigs a hypodynamic endotoxin shock was induced. The animals were
then randomly divided into three groups: a control group, a group treated with
hypertonic (7.5%) saline-6% dextran 70 (HSD group) at 4 ml/kg and a group
treated with isotonic (0.9%) saline-6% dextran 70 (ISD) at 4 ml/kg. The mortal-
ity rate at 300 min was 67% in the control group, 14% in the HSD group and
75% in ISD group. This gives a relative risk of death in the HSD group when
compared with the control group of 0.17 and a relative risk of death in the ISD
group of 1.2 when compared to control. The findings that isotonic saline was
associated with the highest mortality raise some concerns. More work is
required before any conclusions can be drawn from these unusual observations.
Importantly, for the point of view of fluid therapy in septic AKI, there were no
data on renal function in this study.
Unfortunately there is no other study specifically addressing the need for
fluid resuscitation in order to sustain renal function in septic animals. Some
experimental studies have analyzed the effect of crystalloid resuscitation on
organ blood flow and compared their efficacy versus colloid resuscitation.
Bressack et al. [41] published a study comparing normal saline and 5%
albuminated saline in piglets. They found that RBF was related only to CO and
that organ edema formation occurred only in the saline-treated group.
Wan et al. [42] in 2006 published a randomized controlled crossover ani-
mal study in which they demonstrate that resuscitation with normal saline
increased central venous pressure, CO, urine output, creatinine clearance and
fractional excretion of sodium in sheep but had no effect on RBF. All these
findings were transient (⬍1h).
In a recent study, Garrido et al. [43] compared Ringer’s lactate and hyper-
tonic saline in the resuscitation of a dog model of hypodynamic shock. After
30 min from the start of an infusion of Escherichia coli in mongrel dogs, they
Licari/Calzavacca/Ronco/Bellomo 174
randomized 7 animals to receive lactated Ringer at 32 ml/kg over 30 min or
7.5% hypertonic saline solution 4ml/kg over 5 min. They then observed (120 min)
the effects on hemodynamic parameters and on regional perfusion markers.
Both infusion regimens produced a transient improvement in systemic and
regional blood flow. However, no specific information on the renal effects was
reported. All available animal studies are summarized in table 1.
Unfortunately, no study so far has specifically used animal models of
sepsis or septic AKI to study the optimal amount of fluid resuscitation or the
prevention of AKI. Also no studies have compared resuscitation with fluids
versus resuscitation with vasopressors alone or the combination of the two.
In addition, many different hemodynamic targets are used in titrating fluid
Table 1. Summary of all available animal studies
Authors, year Study performed Main finding
Crystalloid vs. nothing
Oi et al., 2000 [40] No fluid vs. isotonic or OR 1.2 for death in
hypertonic infusion in untreated animals
a porcine model of
endotoxemia
Wan et al., 2006 [42] Normal saline vs. Normal saline
control transiently increases
CVP, CO, mesenteric
flow, UO, CC and FeNa
in sheep, no effect on
RBF
Crystalloid vs. colloid
Bressack et al., Normal saline vs. 5% Organ edema in the
1987 [41] albuminated saline in saline-treated group
piglets
Garrido Adel et al., Ringer lactate and Both transient
2006 [43] hypertonic saline in improvement in
resuscitation of dog in systemic and regional
hypodynamic shock blood flow after infusion;
in the hypertonic group
significant and
sustained reduction
of systemic and
mesenteric oxygen
extraction
CC ⫽ Creatinine clearance; CVP ⫽ central venous pressure; UO ⫽ urine output.
Fluid Resuscitation and the Septic Kidney 175
resuscitation but no controlled trial has been performed to address this specific
issue.
The above observations from animal studies related to septic AKI and fluid
therapy highlight the dearth of information that exists in this field and the need
for further investigations.
Conclusions
Fluid resuscitation is a common empirical practice in the treatment of
patients with septic AKI. The rationale for this is to attempt to restore an ade-
quate RBF in an environment of suspected decreased oxygen delivery due to
ischemia. However, more recent studies challenge this paradigm. In addition,
recent evidence suggests that liberal fluid administration may be dangerous to
both kidney and patient and that high MW starches may be nephrotoxic. Animal
studies remain inadequate in helping us understand what might be the best
approach to fluid therapy in septic AKI. Much more research is needed in this
important field of critical care nephrology.
References
1 Schrier RW, Wang W: Acute renal failure and sepsis. N Engl J Med 2004;351:159–169.
2 Landry DW, Oliver JA: The pathogenesis of vasodilatory shock. N Engl J Med 2001;345:588–595.
3 National Nosocomial Infections Surveillance System (NNIS) report, October 2005. www.cdc.gov/
ncidod/dhqp/nnis_pubs.html.
4 Uchino S, Kellum JA, Bellomo R: Acute renal failure in critically ill patients: a multinational,
multicenter study. JAMA 2005;294:813–818.
5 Cole L, Bellomo R, Silvester W: A prospective, multicenter study of the epidemiology, manage-
ment, and outcome of severe acute renal failure in a ‘closed’ ICU system. Am J Respir Crit Care
Med 2000;162:191–196.
6 De Mondança A, Vincent JL, Suther PM, et al: Acute renal failure in the ICU: risk factors and out-
come evaluated by the SOFA score. Intensive Care Med 2000;26:915–921.
7 Bellomo R, Ronco C, Kellum JA, et al: Acute renal failure-definition, outcome measures, animal
models, fluid therapy and information technology needs: the Second International Consensus
Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004;8:R204–R212.
8 Bellomo R, Kellum JA, Ronco C: Defining and classifying acute renal failure: from advocacy to
consensus and validation of the RIFLE criteria. Intensive Care Med 2007;33:409–413.
9 Badr KF: Sepsis-associated renal vasoconstriction: potential targets for future therapy. Am J
Kidney Dis 1992;20:207–213.
10 De Vriese AS, Bourgeois M: Pharmacologic treatment of acute renal failure in sepsis. Curr Opin
Crit Care 2003;9:474–480.
11 Langenberg C, Bellomo R, May C, Li W, Moritoki E, Morgera S: Renal blood flow in sepsis. Crit
Care 2005;9:R363–R374.
12 Brenner M, Schaer GL, Mallory DL, et al: Detection of renal blood flow abnormalities in septic
and critically ill patients using a newly designed indwelling thermodilution renal vein catheter.
Chest 1990;98:170–179.
Licari/Calzavacca/Ronco/Bellomo 176
13 Lucas CE, Rector FE, Werner M, et al: Altered renal homeostasis with acute sepsis. Clinical sig-
nificance. Arch Surg 1973;106:444–449.
14 Rector F, Goyal S, Rosemberg IK, et al: Sepsis: a mechanism for vasodilatation in the kidney. Ann
Surg 1973;178:222–226.
15 Langenberg C, Wan L, Egi M, May CN, Bellomo R, et al: Renal blood flow in experimental septic
acute renal failure. Kidney Int 2006;69:1996–2002.
16 Heemskerk AE, Huisman E, van Lambalgen AA, et al: Gram-negative shock in rats depends on
the presence of capsulated bacteria and is modified by laparotomy. Shock 1996;6:418–425.
17 Heemskerk AE, Huisman E, van Lambalgen AA, et al: Laparotomy and renal function during
endotoxin shock in rats. Shock 1996;6:410–417.
18 Heemskerk AE, Huisman E, van Lambalgen AA, et al: Renal function and oxygen consumption
during bacteremia and endotoxemia in rats. Nephrol Dial Transplant 1997;12:1586–1594.
19 Dellinger RP, Carlet JM, Masur H: Surviving sepsis campaign guidelines for management of
severe sepsis and septic shock. Crit Care Med 2004;32:858–873.
20 Landry DW, Oliver JA: The pathogenesis of vasodilatory shock. N Engl J Med 2001;23:588–595.
21 Ince C, Sinaasappel M: Microcirculatory oxygenation and shunting in sepsis and shock. Crit Care
Med 1999;27:1369–1377.
22 Bersten AD, Holt AW: Vasoactive drugs and the importance of renal perfusion pressure. New
Horiz 1995;3:650–661.
23 Bagshaw S, Bellomo R: Fluid resuscitation and the septic kidney. Curr Opin Crit Care 2006;12:
527–530.
24 Vincent JL, Gherlac H: Fluid resuscitation in severe sepsis and septic shock: an evidence-based
review. Crit Care Med 2004;32(suppl 11):s451–s454.
25 Van Biesen W, Yegenaga I, Vanholder R: Relationship between fluid status and its management on
acute renal failure AKI in intensive care unit ICU patients with sepsis: a prospective analysis.
J Nephrol 2005;18:54–60.
26 Wiedemann HP, Wheeler AP, Bernard GR: Comparison of two fluid-management strategies in
acute lung injury. N Engl J Med 2006;354:2564–2575.
27 Sakr Y, Vincent JL, Reinhart K: High tidal volume and positive fluid balance are associated with
worse outcome in acute lung injury. Chest 2005;128:3098–3108.
28 Simmons RS, Berdine GG, Seidenfeld JJ: Fluid balance and the adult respiratory distress syn-
drome. Am Rev Respir Dis 1987;135:924–929.
29 Mehta RL, Clark WC, Schetz M: Techniques for assessing and achieving fluid balance in acute
renal failure. Curr Opin Crit Care 2002;8:535–543.
30 Rivers E, Nguyen B, Havstad S: Early goal-directed therapy in the treatment of severe sepsis and
septic shock. N Engl J Med 2001;345:1368–1377.
31 Rackow EC, Falk JL, Fein IA: Fluid resuscitation in circulatory shock: a comparison of the car-
diorespiratory effects of albumin, hetastarch, and saline solutions in patients with hypovolemic
and septic shock. Crit Care Med 1983;11:839–850.
32 Davidson IJ: Renal impact of fluid management with colloids: a comparative review. Eur J
Anaesthesiol 2006;23:721–738.
33 Finfer F, Bellomo R, Boyce N: A comparison of albumin and saline for fluid resuscitation in the
intensive care unit. N Engl J Med 2004;350:2247–2256.
34 Treib J, Baron JF, Grauer MT, Strauss RG: An international view of hydroxyethyl starches.
Intensive Care Med 1999;25:258–268.
35 Schortgen F, Lacherade JC, Bruneel F, et al: Effects of hydroxhyethylstarch and gelatin on renal
function in severe sepsis: a multicentre randomized study. Lancet 2001;357:911–916.
36 Legendre C, Thervet E, Page B, et al: Hydroxyethylstarch and osmotic-nephrosis-like lesions in
kidney transplantation. Lancet 1993;342:248.
37 Cittanova ML, Leblanc I, Legendre C, et al: Effect of hydroxyethylstarch in brain-dead kidney
donors on renal function in kidney transplant recipients. Lancet 1996;348:1620–1622.
38 Wade CE: Hypertonic saline resuscitation in sepsis. Crit Care 2002;6:397–398.
39 Oliveira RP, Velasco I, Soriano FG, Friedman G: Clinical review: hypertonic saline resuscitation in
sepsis. Crit Care 2002;6:418–423.
Fluid Resuscitation and the Septic Kidney 177
40 Oi Y, Aneman A, Svensson M, et al: Hypertonic saline-dextran improves intestinal perfusion and
survival in porcine endotoxin shock. Crit Care Med 2000;28:2843–2850.
41 Bressack MA, Morton NS, Hortop J: Group B streptococcal sepsis in the piglet: effects of fluid
therapy on venous return, organ oedema, and organ blood flow. Circulation 1987;61:659–669.
42 Wan L, Bellomo R, May CN: The effect of normal saline resuscitation on vital organ blood flow
and septic sheep. Intensive Care Med 2006;32:1238–1242.
43 Garrido Adel P, Cruz Junior RJ, Poli de Figueiredo LF, et al: Small volume of hypertonic saline as
the initial fluid replacement in experimental hypodynamic sepsis. Crit Care 2006;10:R62–R71.
Prof. Rinaldo Bellomo
Department of Intensive Care, Austin Health
Melbourne, Vic. 3084 (Australia)
Tel. ⫹61 3 9496 5992, Fax ⫹61 3 9496 3932, E-Mail rinaldo.bellomo@austin.org.au
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 178–184
Factors Affecting Circuit Patency
and Filter ‘Life’
Ian Baldwin
Department of Intensive Care, Austin Hospital, and Department of Nursing
and Health Sciences, RMIT University, Melbourne, Vic., Australia
Abstract
Frequent clotting applying continuous renal replacement therapy means treatment can
be inadequate and with increased costs for circuits and nursing time. Patency of the extracor-
poreal circuit is commonly achieved using anticoagulants such as heparin. When anticoagu-
lants are not used, or clotting occurs within a few hours of use, with anticoagulation, blood
flow failure is a likely cause. The blood pump can fail to deliver without operator awareness.
Clotting within the membrane and/or venous ‘air-trap’ chamber is common where resistance
to blood flow is high with stasis and turbulence. The design of the venous chamber allows the
blood fill level to oscillate and form a clot, with a blood filter at the exit of the chamber also
causing clot development. Several practices attempt to prevent clotting, however most with-
out evidence. Adding heparin to the circuit during the preparation phase, ensuring that the
access catheter is not obstructed, a blood flow setting of ⱖ200 ml/min, and administration of
substitution fluids before the membrane (predilution) can be useful strategies for increasing
circuit patency. An audit of filter life is useful and necessary feedback to nursing staff train-
ing strategies. This promotes safety and, when circuit patency is poor, may reflect poor trou-
bleshooting ability.
Copyright © 2007 S. Karger AG, Basel
In this article we will review issues affecting the patency of the extracor-
poreal circuit (EC) and filter ‘life’ in relation to continuous renal replacement
therapy (CRRT) by first discussing where clotting commonly occurs. This will
be followed by a review of factors affecting circuit patency including consider-
ation of heparin during preparation, the access catheter, blood flow, the mem-
brane, the administration of substitution fluids, the venous ‘air-trap’ chamber,
and the importance of training and education for staff.
Nursing Issues in Critical Care Nephrology
Factors Affecting Circuit Patency and Filter ‘Life’ 179
Circuit Patency, Where Clotting Occurs
Clotting occurs most commonly in the hemofilter (membrane) and or the
‘venous’ air-trap chamber [1, 2]. Figure 1 is a drawing of a hemofilter schemat-
ically highlighting that clot formation can initially occur as blood enters the
membrane, and resistance is high prior to the ‘potting’ material for the filter
fibers. Further clotting and clogging of the fibers with proteins then possibly
occurs as a secondary additional process. Figure 2 is a drawing of a venous
chamber schematically highlighting flow resistance factors at the chamber exit
due to the chamber blood filter. If the venous chamber clots, this can obstruct
blood flow completely with blood not easily returned to the patient, an undesir-
able event. This event is often misdiagnosed as ‘filter clotting’. Therefore the
venous chamber is an important factor in circuit and filter patency.
Anticoagulants
Anticoagulant drugs prevent or delay the formation of clots in the EC, and
many methods are used with different levels of evidence supporting their suc-
cess [3]. Anticoagulants are the focus of another article in this book and will be
not discussed other than to suggest that circuit clotting may not always be due to
insufficient or incorrect anticoagulation. Maintaining blood flow can be as
important in preventing clotting. These flow ‘mechanics’ are particularly
Fig. 1. Schematic drawing of the CRRT membrane indicating sites of clotting: before
the potting medium at blood entry, and within fibers by cellular protein plugs.
Potting medium
Fibers
cellular ‘plugs’
and clotting
Blood tubing
connection
Filter outer casing
Blood space
clot formation
Dialysate or
filtration
Baldwin 180
important when no anticoagulation is used in patients at risk of bleeding, and
useful to understand for all treatments and for circuit patency in every patient.
Access Catheter
The veno-venous access catheter used for CRRT is not a common site for
clot formation, however the access catheter may be associated with formation
of clots in the EC. Increasing negative (‘arterial’) and positive (‘venous’) pres-
sure in the EC reflects access lumen obstruction, causing blood pump failure
and reduction in prescribed output. This can then cause slowing of blood flow in
the membrane, and clotting as ultrafiltration continues regardless of blood flow
indicator speed [4]. This access catheter failure may be unrecognized with
reduced blood flow over long periods causing membrane clotting [4].
The insertion site (e.g. femoral, subclavian, internal jugular) of the catheter
may also be implicated in circuit patency with anecdotal experience suggesting
that poor blood flow can occur when repositioning patients during nursing care.
Nursing care and physical therapy must be managed with caution and a
response to changes in arterial and venous circuit pressures indicating catheter
obstruction is often needed. It is an important aspect of CRRT nursing knowl-
edge to recognize catheter obstruction, and modify the patient position to main-
tain blood flow without excessive ‘arterial’ (ⱖ100 mm Hg) or ‘venous’
(⬎150 mm Hg) pressures. Sometimes subtle patient position changes can be
sufficient to alleviate such pressure changes and facilitate correct blood flow.
Fig. 2. Schematic drawing of the venous air-trap chamber showing where clot forma-
tion commonly occurs: top of the chamber by blood oscillation building a clot (cross-section
indicated) and the bottom around the outlet filter (picture included).
Vertical smearing
of blood at top of
chamber: clotting
Blood clot at filter
Blood entry
Gas
space
Clot: cross-section
Factors Affecting Circuit Patency and Filter ‘Life’ 181
Blood Pump – Flow Speed
Blood flow controlled by the blood pump setting may influence the devel-
opment of EC clot formation [5, 6]. In theory: the faster the blood flow, the less
clot formation. A blood flow rate of 200 ml/min is adequate for all modes of
CRRT, however hemoconcentration of blood can occur in the membrane if the
ultrafiltrate and blood flow are not correctly mixed (filtration fraction error).
This highlights a possibly important relationship between clearance mode or
technique (diffusion or convection) and the potential for clotting in the EC.
Whilst there is minimal evidence, some clinicians do suggest that convective
clearance may have a higher potential for clotting in comparison to diffusive
clearance [7, 8].
Membrane and Circuit Patency
The addition of heparin into a circuit during or after priming in order to
‘coat’ the membrane is suggested by many authors to be a useful strategy to
prevent clotting [9–12]. There is no good evidence to suggest this has any
effect on all membranes, however plastic and membrane surfaces do take up
heparin particularly after treating to neutralize negative charge [13, 14].
Therefore, unless contraindicated, heparin can be added to a circuit after prim-
ing whilst awaiting connection to the patient, with this heparinized saline
pumped around the circuit to promote the coating effect. Different fiber types
may also influence clotting [15–17]. Although synthetic membranes are con-
sidered biocompatible, premature clotting when exposed to an acrylonitrile
membrane may be reduced when changed to polysulfone, with some differ-
ences in racial and genetic disposition to clotting for different membrane expo-
sures a possible association [18]. Finally, larger surface area membranes offer
less resistance to blood flow [19] and may increase circuit patency and func-
tional life. Increasing membrane size from 1.0 to 1.4 m
2
is of minimal clinical
consequence in the adult patient, and often at no increased cost, but may
increase filter life.
Circuit Patency and Administration of Substitution Fluids
There is evidence that use of predilution reduces membrane clotting in
comparison to postdilution in pure convective modes [20]. The amount of predi-
lution volume required to achieve this affect is not clear, however an alteration in
hematocrit is the likely mechanism. Citrate anticoagulation can be performed by
Baldwin 182
addition of the citrate to predilution replacement fluids in a convective therapy.
This combines the effect of anticoagulation and predilution [21].
Venous Chamber
An important safety feature of any EC is an air-trap chamber placed in the
EC prior to the blood returning to the patient, however this EC component is
often a site for clotting [1, 5, 9, 10, 22].
The chamber allows turbulent flow to occur with the blood level in the cham-
ber oscillating or rising and falling. This oscillation is consistent with the pulsatile
flow generated by the blood pump and the varying resistance at the venous lumen
of the access catheter. The chamber cannot be completely full as a pocket of gas
(air ⫾ CO
2
) above the blood level acts as a medium for pressure readings to a trans-
ducer, and prevents blood entering the transducer line. The oscillating blood level
causes a constant smearing and cell deposition on the inside of the chamber even-
tually developing a ring of deposited cells inside the chamber, building a clot, as
indicated in figure 2. Attempts have been made to prevent this clotting by adding
heparin into the chamber before and during use [23], adding fluids to the chamber
(postdilution), and use of a tubing design such that incoming blood enters under the
blood chamber level. This last approach can create a cell–plasma separation with a
small layer of plasma separating to the top of the chamber. This provides a plasma
layer protecting the cells from exposure to the gas and reduces cell smearing.
While evidence is lacking, circuit patency may be enhanced by keeping the
venous chamber level close to full with a minimal gas pocket, adjusting the
level down when a ring of clot begins to form, adding postdilution fluids into
this chamber when used, and adding heparin into the chamber during the prim-
ing procedure – heparin coating.
Staff Training, Education
Safe and skilled use of CRRT machines requires nursing education and
training activities with theoretical and practical components. Inability to manage
and/or correct simple alarm events may be associated with poor circuit patency
and/or serious patient harm [24].
Current day machines with rigid priming and preparation sequences, trou-
bleshooting prompts, and automated alarms are a safety net for use, but are not
absolute in respect to the human–machine interface. An inability to manage an
alarm event where the blood pump stops can be the cause of circuit clotting and
failure. No blood flow for 3–5 min or less can cause cell and plasma separation,
clotting, and no recovery.
Factors Affecting Circuit Patency and Filter ‘Life’ 183
There are many strategies to train nurses for these events using a simulation
set up of the machine and EC, interactive video activities of these alarms, and
simple tutorial activities [10, 25]. All of these are useful towards providing safe
and successful CRRT, and are another strategy for maintaining circuit patency.
With regard to training programs, bedside records of ‘filter life’ are useful
and important data to review. This type of circuit patency or filter life audit pro-
vides useful feedback to teachers, particularly when circuit ‘life’ is poor (e.g.
⬍3–4 h). Repeated events of this can reflect the adequacy of nursing education
and training needs.
Conclusion
Clotting in the circuit during CRRT can be delayed or prevented by both
the administration of anticoagulants and prevention of blood stasis and resis-
tance in the circuit. The access catheter, blood flow, membrane and venous
chamber are in a relationship and this has an association with circuit patency.
The use of heparin during preparation by coating the circuit and administration
of substitution as predilution are also useful strategies to prevent clotting.
Training and education with audit of each circuit or filter ‘life’ are useful data to
collect and provide feedback. These strategies promote safety and successful
use where circuit patency is one measure of success.
References
1 Davenport A: The coagulation system in the critically ill patient with acute renal failure and the
effect of an extracorporeal circuit. Am J Kidney Dis 1997;30(suppl 4):s20–s27.
2 Keller F, Seeman J, Preuschof L, Offermann G: Risk factors of system clotting in heparin free
hemodialysis. Nephrol Dial Transplant 1990;5:802–807.
3 Oudemans-Van Straaten HM, Wester JPJ, de Pont ACJM, Schetz MRC: Anticoagulation strategies
in continuous renal replacement therapy: can the choice be evidence based ? Intensive Care Med
2006;32:188–202.
4 Baldwin I, Bellomo R: The relationship between blood flow, access catheter and circuit failure
during CRRT; a practical review; in Ronco C, Bellomo R, Brendolan A (eds): Sepsis, Kidney and
Multiple Organ Failure. Contrib Nephrol. Basel, Karger, 2004, vol 144, pp 203–213.
5 Webb AR, Mythen, MG, Jacobsen D, Mackie IJ: Maintaining blood flow in the extracorporeal cir-
cuit: hemostasis and anticoagulation. Intensive Care Med 1995;21:84–93.
6 Bellomo R, Ronco C: Circulation of the continuous artificial kidney: blood flow, pressures, clear-
ance and the search for the best; in Artigas A, Bellomo R, Ronco C (eds): Circulation in Native
and Artificial Kidneys. Karger, Basel, 1997, pp 138–149.
7 Mitchell A, Daul AE, Beiderlinden M, Schafers RF, Heemann U, Kribben A, et al: A new system
for regional citrate anticoagulation in continuous veno-venous hemodialysis (CVVHD). Clin
Nephrol 2003;59:106–114.
8 Kutsogiannis DJ, Mayers I, Chin WD, Gibney RT: Regional citrate anticoagulation in continuous
veno-venous hemodiafiltration. Am J Kidney Dis 2000;35:802–811.
Baldwin 184
9 Schetz M: Anticoagulation for continuous renal replacement therapy. Curr Opin Anaesthesiol
2001;14 143–149.
10 Baldwin IC, Elderkin TD: CVVH in intensive care. Nursing perspectives. New Horiz 1995;3:738–747.
11 Baldwin IC, Bridge NP, Elderkin TD: Nursing issues, practices, and perspectives for the manage-
ment of continuous renal replacement therapy in the intensive care unit; in Bellomo R, Ronco C
(eds): Critical Care Nephrology. Dordrecht, Kluwer Academic, 1998, pp 1309–1327.
12 Davenport A: Extracorporeal anticoagulation for intermittent and continuous forms of renal
replacement therapy in the intensive care unit; in Murray PT, Brady HR, Hall JB (eds): Intensive
Care Nephrology. London, Taylor & Francis, 2006, pp 165–180.
13 Lavaud S, Canivet E, Wuillai A, Maheut H, Randoux C, Bonnet J M, Renaux JL, Chanard J:
Optimal anticoagulation strategy in haemodialysis with heparin coated polyacrylonitrile mem-
brane. Nephrol Dial Transplant 2003;18:2097–2094.
14 Lavaud S, Paris B, Maheut H, Randoux C, Renaux JL, Rieu P, Chanard J: Assessment of the
heparin-binding AN69 ST hemodialysis membrane: II. Clinical studies without heparin adminis-
tration. ASAIO J 2005;51:348–351.
15 Jones CH: Continuous renal replacement therapy in acute renal failure: membranes for CRRT.
Artif Organs 1998;22:2–7.
16 Salmon J, Cardigan R, Mackie I, Cohen SL, Machin S, Singer M: Continuous venovenous
haemofiltration using polyacrylonitrile filters does not activate contact system and intrinsic coag-
ulation pathways. Intensive Care Med 1997;23:38–43.
17 Frank RD, Weber J, Dresbach H, Thelan H, Weiss C, Floeg J: Role of contact system activation in
hemodialyzer-induced thrombogenicity. Kidney Int 2001;60:1972–1981.
18 Hadef S, Raoudha G, Saoussen A, Imen K, Henda E, Jalel G, Abdelaziz H: A prospective study of
the prevalence of heparin-induced antibodies and other associated thromboembolic risk factors in
pediatric patients undergoing hemodialysis. Am J Hematol 2006;81:328–334.
19 Jenkins RD: The extra-corporeal circuit: physical principles and monitoring; in Ronco C, Bellomo R
(eds): Critical Care Nephrology, ed 1. Dordrecht, Kluwer Academic, 1998, pp 1189–1197.
20 Uchino S, Fealy N, Baldwin I, Morimatsu H, Bellomo R: Pre-dilution vs. post-dilution during con-
tinuous veno-venous hemofiltration: impact on filter life and azotemic control. Nephron Clin
Pract 2003;94:94–98.
21 Egi M, Naka T, Bellomo R, Cole L, French C, Trethewy C, Wan L, Langenberg CC, Fealy N,
Baldwin I: A comparison of two citrate anticoagulation regimens for continuous veno-venous
hemofiltration. J Artif Organs 2005;28:1211–1218.
22 Keller F, Seeman J, Preuschof L, Offermann G: Risk factors of system clotting in heparin free
hemodialysis. Nephrol Dial Transplant 1990;5:802–807.
23 Baldwin I, Tan HK, Bridge N, Bellomo R: Possible strategies to prolong circuit life during
hemofiltration: three controlled studies. Ren Fail 2002;24:839–848.
24 Ronco C: Fluid balance in CRRT: a call to attention! Int J Artif Organs 2005;28:763–764.
25 Baldwin I: Training management and credentialing for CRRT in critical care. Am J Kidney Dis
1997;30(suppl 4):S112–S116.
Prof. Ian Baldwin
Department of Intensive Care, Austin Hospital
Melbourne, Vic. 3084 (Australia)
Tel. ⫹61 3 9496 5000, Fax ⫹61 3 9496 3932, E-Mail ian.baldwin@austin.org.au
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 185–190
Starting Up a Continuous Renal
Replacement Therapy Program on ICU
Wilfried De Becker
Department of Intensive Care, University Hospital Gasthuisberg, Leuven, Belgium
Abstract
Background/Aim: The questions as to which treatment is the most effective for the
replacement of renal function in critically ill patients with acute renal failure and the qualifica-
tions needed by nurses to manage the continuous renal replacement therapy (CRRT) device are
part of an ongoing debate between nephrologists and intensivists, between nurses of the renal
ward and the ICU. Methods: The keys to a successful CRRT program are a well-balanced and
practical education program, a user-friendly dialysis machine, and technical support 24 h/day.
A computerized data management system will diminish the workload to an acceptable level.
Results: Intensive care nurses on our ICUs are well trained to execute CRRT without the
involvement of nephrology nurses. On the ICU, the 24-hour presence of an intensivist is an
additional advantage to solve medical problems involving CRRT. The daily cost of CRRT is
only dependent on the devices and independent of human resources. Conclusion: Initiating
and maintaining a CRRT program is a great challenge for the ICU nurse. The possible problems
remain within the ICU staff’s ability to solve if they follow an education program. If the work-
load for the nurses is well monitored, extra personnel can be avoided.
Copyright © 2007 S. Karger AG, Basel
Introduction
The questions as to which treatment is the most effective for the replace-
ment of renal function in critically ill patients with acute renal failure and the
qualifications needed by the nurses to manage the continuous renal replace-
ment therapy (CRRT) device is an ongoing debate between nephrologists and
intensivists as well as between nurses in the renal ward and the ICU [2, 3, 8].
Some centers create a specialized CRRT team with nurses and physicians from
both disciplines [3, 6]. Our experience is that all ICU nurses are well educated
and well trained to perform all actions of the CRRT.
De Becker 186
In 1980 continuous arteriovenous hemofiltration was introduced to our
ICU. The intensivist installed the circuit and the nurse managed the fluid
balance. A few years later we used continuous veno-venous hemofiltration
CVVH with a roller pump device (BSM22™ by Gambro™). In the beginning
the intensivists still installed the filter but step-by-step the nurses took over this
procedure. The Prisma™ (Gambro) was introduced in 1994. All nurses were
trained to do the installation and the physician only managed the settings and
electrolyte balance. The responsibility and the workload for the ICU nurse
increased and more education was needed. The physiology of renal failure,
CRRT techniques, care plan and practical training were included in the educa-
tion program.
The introduction of the new Prismaflex™ device (Gambro) was preceded
by an update lesson about renal failure (1 h) by a member of the medical staff
and one about the prevention of filter failure (1 h) by a staff nurse. A 2-hour
practical training was organized for every nurse in small sessions of 6 people by
a representative of Gambro.
Our Intensive Care Department has 56 ICU beds divided into 4 units: 2
general ICUs with 16 beds each; a pediatric ICU with 10 beds, and a burn unit
with 14 beds. Each unit has a head nurse and 2 assistant head nurses. One of
them has a special interest in CRRT and is also an instructor in the education
program.
The nursing staff rate is 2.8 full time equivalents for each ICU bed. In prac-
tice every nurse takes care of 2 patients during every shift (morning, evening
and night).
The medical staff members are intensivists who are well trained in the crit-
ical care of kidney failure.
Nurses and intensivists work complementarily in the handling of all CRRT
tasks. So no dialysis nurse or nephrologist is needed in this setting. Only inter-
mittent hemodialysis (IHD) is performed by staff from the dialysis department
of our hospital.
The incidence of CVVH in 2006 was 4.9/day (0–9) and IHD was 2/day.
Keys to a Successful CRRT Program [5]
Physicians well trained in intensive care medicine and renal failure must be
available on the ward 24 h/day so that nurses can easily ask their advice, even
for minor problems involving the CRRT.
Intensive care nurses must be well trained in care plans for renal failure
and CRRT techniques, and the CRRT devices must be user-friendly.
Starting Up a CRRT Program on ICU 187
A CRRT education program using the ‘Bachelor after bachelor in intensive
care and emergency care’ program [1] gives a theoretical background and large
practical experience is offered to future ICU nurses.
The ICU nurse is supported by experienced staff nurses who are able to
organize additional training on a regular basis.
A computerized monitoring system manages all pressures and fluid values
in the patient’s file. Full technical support must be offered by the manufacturer
(24 h/day).
Education [1]
The education starts at nursing school and contains: anatomy and physiology
of the kidney (4 h); 4 h of theory and 6 h of case study regarding pathology and
nursing of patients with renal failure, especially chronic renal failure with inter-
mittent hemodialysis and peritoneal dialysis, and acid-base balance (2 h).
Nurses who have completed the nursing school have a Bachelor Diploma
in Nursing and they can continue with the Bachelor after Bachelor program.
The nursing high school which is affiliated with our university hospital
organizes a Bachelor after Bachelor education program in 4 modules as shown
in table 1.
This education program takes a whole year; in-practice training included,
and is completed after an oral jury examination on the basis of cases from the
work field. All the lecturers are physicians and nurses from our hospital, and
most of them work on the intensive care or emergency ward of our hospital.
After this course the nurses can easily find work on any ICU, PICU, NICU or
emergency ward.
Table 1. The 4 modules of the education program
Module Lessons In-practice training
I General intensive care basic I 85 h 121 h on general ICU
II General intensive care basic II 80 h 121 h on general ICU
III Emergency care 80 h 167 h on emergency ward
IV Pediatric intensive care 36 h 121 h on PICU or NICU
Module II includes several topics about renal failure: physiology of kidney failure (1 h);
epuration techniques (1 h); nursing care plan for CRRT (1.5 h); practical laboratory installa-
tion on the Prismaflex™ (1.5h), and practice on an ICU in the presence of qualified nurses.
De Becker 188
Equipment
The choice of an appropriate CRRT device is very important for the ICU
nurse. Before changing from Prisma to Prismaflex, other manufacturers were
also invited to demonstrated with their latest devices which we tested at the bed-
side with a small dedicated group of nurses. The test focused on the ergonomic
aspects of the device and how easy it was to learn the filter set installation. The
beta-trial of the Prismaflex device was also performed by this group. They all
choose this device to succeed the Prisma because of the recognizable installa-
tion procedure and the new ergonomic aspects such as higher scales. The
Prismaflex is a good example of a user-friendly machine: easy to learn, and an
easy to install ‘all-in-one’ set with a ‘step-by-step’ explanation on the color dis-
play. Many tasks run automatically. All alarms and possible actions are clearly
explained on the display. No handout is needed. All these advantages save time
in the workload of the ICU nurse. If a dual lumen catheter is already in place,
therapy can be installed and running 30min after the decision to start CRRT is
made.
Our department has 6 devices on standby and has the ability to extend this
with 2 devices from the medical ICU. When the demand is larger we still have
2 Prisma devices and on exception we can call for support from Gambro to rent
more devices.
Computerized Monitoring
Prismaflex is connectable to our patient data management system (PDMS;
MetaVision™ of iMDsoft™). The data from settings, pressure monitoring,
fluids, scales and therapy status plus additional fluid balance calculations and
laboratory results are displayed every minute on the bedside computer screen.
Pop-up warning menus are shown when some actions must be done. The run
time of every filter is monitored and can be used as a quality parameter for
CRRT. This advantage improves the quality of collecting data and significantly
diminishes the workload for the nurse.
Tasks for the Nurses
The tasks of the nurses are the following: priming the circuit using 1 l of
normal saline containing 5,000 IU heparin, an initial bolus of heparin is not
used; installation of the circuit, input of the prescribed settings of CVVH
and starting up; dialysis machine resetting after any trouble; replacement of
Starting Up a CRRT Program on ICU 189
exhausted fluids (substitution and anticoagulant); return blood before filter is
clotted and re-initiate CRRT day and night; circuit removal and catheter refill-
ing with anticoagulant after circuit clotting or major machine troubles day and
night; monitoring pressures and fluid loss with a PDMS; fluid balance control,
automatically calculated by the PDMS every hour; blood sampling for coagula-
tion (aPTT control) and electrolytes every 4 h; catheter care, and assist the
intensivist while placing the catheter.
Standard CVVH Settings
We simultaneously use two 5-liter bags of a commercialized electrolyte
solution hanging on the dialysate and substitution scales. The effluent is cap-
tured in 5-liter bags. If heparin is prescribed we use 20-ml syringes (standard
2,000 IU/20 ml) with the infusion pump of the machine. The filter is a M100
(AN69 membrane), and the blood flow rate is 200–300 ml/min, the substitution
rate is 2,000–3,500 ml/h, predilution is 30%, postdilution is 70%, and a 13-
french dual-lumen venous catheter (Hemo-Access™, Gambro) is used.
Discussion
CVVH does not require extra personnel cost because the cost of an ICU
nurse and intensivists are already calculated in the total budget of an ICU. 16%
of the total cost can be saved [7]. When IHD therapy is needed, a dialysis nurse
from the renal ward will do the whole procedure during 2–4 h. Sometimes they
do two IHD therapies simultaneously.
The affinity of the ICU nurse to their patient is greater and this affects
morbidity in a positive way [4].
Most ICU nurses have great technical skills with all kinds of ICU equip-
ment. The more they use a CVVH device by installing the set, the better they
can anticipate all kinds of alarms. This affects filter survival.
The down-time between two filters can be kept very low [9, 10] by chang-
ing the filter immediately after clotting by the ICU nurses themselves. They do
not have to wait for a renal nurse and they can organize their own work better.
References
1 Deneire M: Bachelor after bachelor in the intensive care and emergency care: education program
of the Catholic High School for Nursing Leuven (Belgium). European Credit Transfer System
form, 2006.
De Becker 190
2 Craig MA, Depner TA, Tweedy RL, Hokana L, Newby-Lintz M: Implementing a continuous renal
replacement therapies program. Adv Ren Replace Ther 1996;3:348–350.
3 Gilbert RW, Caruso DM, Foster KN, Canulla MV, Nelson ML, Gilbert EA: Development of a con-
tinuous renal replacement program in critically ill patients. Am J Surg 2002;184:526–533.
4 Politoski G, Mayer B, Davy T, Swartz MD: Continuous renal replacement therapy. A national per-
spective AACN/NKF. Crit Care Nurs Clin North Am 1998;10:171–177.
5 Willis J, Hodge KS, Dwyer J: Keys for a successful CRRT educational program. Blood Purif
2004;22:229–247.
6 Hodge KS, Willis J, Dwyer J: Development of a consolidated CRRT program. Blood Purif
2004;22:229–247.
7 Vitale C, Bagnis C, Marangella M, Belloni G, Lupo M, Spina G, Bondonio P, Ramello A: Cost
analysis of blood purification in intensive care units: continuous versus intermittent hemodiafil-
tration. J Nephrol 2003;16:572–579.
8 Ronco C, Zanella M, Brendolan A, Milan M, Canato G, Zamperetti N, Bellomo R: Management
of severe acute renal failure in critically ill patients: an international survey in 245 centers.
Nephrol Dial Transplant 2001;16:230–237.
9 Uchino S, Fealy N, Baldwin I, Morimatsu H, Bellomo R: Continuous is not continuous: the inci-
dence and impact of circuit ‘down-time’ on uraemic control during continuous veno-venous
haemofiltration. Intensive Care Med 2003;29:575–578.
10 Baldwin I, Bellomo R, Koch B:Blood flow reductions during continuous renal replacement ther-
apy and circuit life. Intensive Care Med 2004;30:2074–2079.
Wilfried De Becker, RN, CCRN
Department of Intensive Care, University Hospital Gasthuisberg
Herestraat 49
BE–3000 Leuven (Belgium)
Tel. ⫹32 16 344778, Fax ⫹32 16 344015, E-Mail wilfried.debecker@uz.kuleuven.ac.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 191–196
Is There a Need for a Nurse Emergency
Team for Continuous Renal Replacement
Therapy?
Ian Baldwin
Department of Intensive Care, Austin Hospital, and Department of Nursing and
Health Sciences, RMIT University, Melbourne, Vic., Australia
Abstract
The use of an emergency response team for unwell patients has provided an improve-
ment in hospital care standards by reducing medical and postoperative adverse outcomes.
Use of a nurse emergency team for patients treated with continuous renal replacement ther-
apy (CRRT) also has potential to reduce adverse outcomes with CRRT, where staff may lack
experience or find troubleshooting CRRT difficult in an ICU with many critically ill
patients in their care. Differing nursing models are used to provide CRRT in the ICU, and all
of these could benefit from a nursing response team at some time. The response must be
immediate, with suitably available and CRRT-experienced nurses. As with medical emer-
gency team use, the nursing emergency team for CRRT would be called when a deviation
from a standard criterion list occurs. The list could include: prolonged blood pump stoppage
(⬃2 min); air detection alarm; blood leakage; sudden circuit pressure changes – transmem-
brane pressure (⬎200 mm Hg) or venous pressure (⬎200 mm Hg) or arterial pressure nega-
tive (ⱖ 100 mm Hg); the need to override a fluid balance alarm 3 times in 5 min; patient
hypotension; cardiac arrest or similar event, or the nurse is concerned that the machine is
malfunctioning. The ‘human resource’ is the biggest challenge to developing a suitable
response team 24/7, however where ICU and nephrology nurses work in a collaborative
approach for CRRT, a response team would be more easily established and may not be
required continuously.
Copyright © 2007 S. Karger AG, Basel
The use of a rapid response team, with acute care skills, can better manage
the unwell patient in a hospital ward or subacute area by preventing cardiac and
respiratory arrest [1]. The team is called usually via the hospital public
address system or a pager text when patient carers detect changes in vital signs
Baldwin 192
reflecting a deterioration and/or they are simply ‘worried’ about the patient [2].
This approach (Medical Emergency Team, MET) is being adopted in hospitals
to prevent the traditional ‘crisis’intervention associated with cardiac or respira-
tory arrest [3].
CRRT Emergency Team
Some parallels may be made between the unwell patient on a hospital ward
and the continuous renal replacement therapy (CRRT) machine when in use on
a critically ill patient. Both have a blood pump system with measurable pres-
sures reflecting a normal and abnormal state. Changes in ‘vital signs’ can
reflect a deterioration that, if not corrected, may cause the ‘pump’ to stop and
death to occur.
In the context of CRRT, prolonged blood pump stoppage in response to
acute circuit pressure changes may cause a clotting event and an extracorporeal
circuit ‘death’ [4–6]. This will mean a loss of treatment, a need to replace the
circuit and membrane, and commonly the loss of patient blood. All of these are
undesirable, reflect inefficiency, and are potentially preventable.
If subtle and less acute changes in measured circuit pressures are identified
and corrected, this event may not occur; the cardiac arrest is prevented. Is there
a need for a similar response team to better manage CRRT?
Nursing Models and CRRT
Different nursing models are used to apply CRRT to the critically ill
patient in the ICU. Nursing management and care of the CRRT can be provided
by ICU nurses alone, a mixed or collaborative arrangement with the ICU and
nephrology unit nurses, or by nephrology nurses only [7, 8].
The continuous application of the treatment requires a high percentage
of staff to be trained and skilled, and in some centers this has been impo-
ssible [9]. With increasing use of CRRT in the ICU it can be difficult to
provide enough ICU nurses with CRRT training, for a 24-hour, 7-days/week
use of CRRT with multiple patients being treated. This can also be a
challenge for the collaborative models. In contrast, when a small number of
patients are treated, knowledge and skills can be difficult to acquire and
maintain.
Therefore regardless of nursing models, situations can occur where inex-
perienced staff are managing a patient on CRRT. In addition, skilled nurses may
Need for a Nurse Emergency Team for CRRT 193
be caring for an acutely unstable patient or more than one critically ill patient,
and their attention to CRRT may be compromised.
CRRT Machines and Alarm Events
The design of new CRRT machines has made preparation routines more
simple and automated, alarm systems that set and adjust automatically, and user
interfaces with messages and prompts for potential malfunctions [10].
However, when the blood pump stops with an alarm event, rapid and
skilled intervention may be required. This may be due to a circuit pressure
change, an air bubble detected, or fluid flow problem [5]. However if the alarm
is ‘latched’ the blood pump will stop and not restart until the situation is cor-
rected and the alarm reset [5]. Without a rapid response, secondary alarms may
then begin such that multiple alarms are occurring making troubleshooting the
task for an expert.
In addition to blood pump stoppage, other acute alarm events are important
to manage appropriately as these can cause serious harm to the critically ill
patient, e.g. continued override and ignoring a fluid balance alarm may cause
hypovolemia and cardiac arrest [11, 12].
CRRT Nurse Emergency Team
The intentions of the MET are to provide skilled help quickly to an imme-
diate need, where an acute deterioration is thought to be imminent. In addition,
the MET provides an education role, is collaborative and makes useful human
links for a better hospital where knowledge is disseminated outside the closed
walls of a specialized area [2]. It is important to acknowledge that a MET ser-
vice is not to reduce knowledge of carers or take over their skill development.
The MET is not a forum for grandstanding and promoting superiority [13],
creating a culture of hesitation where staff feel they will be criticized after
making a call [14].
The use of a similar service for CRRT could benefit patients and nurses in
many centers. The criteria for calling the team may vary depending on the set-
ting, however table 1 is a suggested list potentially applicable to all. The main
focus of the criteria would be to prevent blood pump stoppage and serious
adverse patient events related to fluid balance and circuit function. Routine
preparation, checking, settings and disconnection routines are not activities
where a CRRT emergency team would be called. These skills may require a
Baldwin 194
support and education service, and constitute areas for basic and ongoing learn-
ing with CRRT in a classroom simulation or similar.
Challenges to a Nurse Emergency Team for CRRT
A MET service relies on suitably qualified people who respond quickly to
help 24 h, 7 days/week. The nursing or medical responders must be readily
available to attend the call, and have a high level of knowledge. This suggests
that they either do not have a patient allocation or can be backed up immedi-
ately during their absence for potentially long periods. Is this possible for a
CRRT emergency team? Which nurses would fit this demand? In addition,
would they have the knowledge necessary? For example, if a nephrology nurs-
ing group becomes the CRRT response team, they would be required to leave a
patient, during dialysis, be backed up by others while they were absent, and
then have the knowledge for a different machine technology used for CRRT in
the ICU.
Yes, this is possible in some settings providing the human resources are
provided, and appropriate training and expertise is obtained by the nurse.
Where, collaborative models are currently used [8], this would be more easily
achieved as the human resources and knowledge required may already exist;
nephrology nurses have a role in the ICU and experience with the CRRT
machine. Where an ICU manages CRRT alone, these factors may be more chal-
lenging to achieve, suggesting that a response team may be very difficult and/or
not worthwhile. This is possibly why many ICU areas have adopted a daily dial-
ysis (SLED) approach during the day as this reduces the training requirement of
CRRT and allows a more achievable collaborative situation with nephrology
Table 1. Nurse emergency team CRRT: criteria for calling
Call for the CRRT nurse emergency team if:
Blood pump stoppage and unable to restart before ⬃2min
Significant air in the circuit
Blood leakage anywhere
Sudden rise in either transmembrane pressure (⬎200 mm Hg) or venous
pressure (⬎200 mm Hg) or arterial pressure negative (ⱖ100 mm Hg)
Override ‘fluid balance’ alarm 3 times in 5 min
Significant patient hypotension
At cardiac arrest or similar
You are concerned or worried the machine is malfunctioning
Need for a Nurse Emergency Team for CRRT 195
nurses [15–17], particularly when machine technology is very similar in both
areas.
Conclusion
The use of a nurse emergency team has potential to apply CRRT in the ICU
with more success, reduce circuit failure and prevent mistakes or serious
adverse events. A similar approach to the MET established for unwell patients
on a general ward would ask suitable nurses to rapidly attend the ICU and assist
others managing CRRT. This could be to a criteria or list including contextual
factors of a nurses ‘concern’ and/or aberrations of measured circuit pressures
and/or alarms suggesting incorrect function. Such a team would need to
respond quickly, being able to leave their area and have the knowledge required
on arrival. This concept would be more easily applied in hospitals where collab-
orative models of nursing for CRRT in the ICU already exist. The availability of
nurses or the human resource is the most important key to success.
References
1 Buist MD, Moore GE, Bernard SA, et al: Effects of a medical emergency team on reduction in
incidence of and mortality from unexpected cardiac arrest in hospital: preliminary study. BMJ
2002;324:387–390.
2 Bellomo R, Goldsmith D, Uchino S, Buckmaster J, et al: Prospective controlled trial of the effect med-
ical emergency team on postoperative morbidity and mortality rates. Crit Care Med 2004;32:916–921.
3 DeVita M: Medical emergency teams: deciphering clues to crises in hospitals. Crit Care
2005;9:325–326.
4 Webb AR, Mythen MG, Jacobsen D, Mackie IJ: Maintaining blood flow in the extracorporeal cir-
cuit: haemostasis and anticoagulation. Intensive Care Med 1995;21:84–93.
5 Baldwin IC, Elderkin TD: CVVH in intensive care. Nursing perspectives. New Horiz 1995;3:
738–747.
6 Baldwin IC, Bridge NP, Elderkin TD: Nursing issues, practices, and perspectives for the manage-
ment of continuous renal replacement therapy in the intensive care unit; in Bellomo R, Ronco C
(eds): Critical Care Nephrology. Dordrecht, Kluwer Academic, 1998, pp 1309–1327.
7 Giuliano K, Pysznik E: Renal replacement therapy in critical care: implementation of a unit-based
continuous venovenous hemodialysis program. Crit Care Nurse 1998;18:40–51.
8 Dirkes S: Continuous renal replacement therapy: dialytic therapy for acute renal failure in inten-
sive care. Nephrol Nurs J 2000;27:6:581–586.
9 Kihara M, Ikeda Y, Shibata K, Masumori S, Fujita H, Ebira H, Toya Y, Takagi N, Shionoiri H,
Umemura S, Ishii M: Slow hemodialysis performed during the day in managing renal failure in
critically ill patient. Nephron 1994;67:36–41.
10 Ronco C, Brendolan A, Dan M, Piccinni P, Bellomo R: Machines for continuous renal replace-
ment therapy; in Ronco C, Bellomo R, La Greca G (eds): Blood Purification in Intensive Care.
Contrib Nephrol. Basel, Karger, 2000, vol 132, pp 323–334.
11 Schultz D: Gambro Prisma Continuous Renal Replacement System. FDA Updated Public Health
Notification. 2005: Retrieved December 19th, 2006, from http:www.fda.gov/cdrh/safety/022706 –
gambro.html
Baldwin 196
12 Ronco C: Fluid balance in CRRT: a call to attention! Int J Artif Organs 2005;28:8:763–764.
13 Foraida MI, DeVita MA, Braithwaite S, et al: Improving the utilization of medical crisis teams
(condition C) at an urban tertiary care hospital. J Crit Care 2003;19:87–94.
14 Jones D, Baldwin I, McIntyre T, Story D, Mercer I, Miglic A, Goldsmith D, Bellomo R: Nurses
attitudes to a medical emergency team service in a teaching hospital. Qual Saf Health Care
2006;15:427–432.
15 Schlaeper C, Amerling R, Manns M, Levin N: High clearance continuous renal replacement ther-
apy with a modified dialysis machine. Kidney Int 1999;56(suppl 72):S20–S23.
16 Kumar VA, Craig M, Depner TA, Yeun JY: Extended daily dialysis: a new approach to renal
replacement for acute renal failure in the intensive care unit. Am J Kidney Dis 2000;36:294–300.
17 Marshall MR, Golper TA, Shaver MJ, Chatoth DK: Hybrid renal replacement modalities for the
critically ill; in Ronco C, Bellomo R, La Greca G (eds): Blood Purification in Intensive Care.
Contrib Nephrol. Basel, Karger, 2001, vol 132, pp 252–257.
Prof. Ian Baldwin
Department of Intensive Care, Austin Hospital
Melbourne, Vic. 3084 (Australia)
Tel. ⫹61 3 9496 5000, Fax ⫹61 3 9496 3932, E-Mail ian.baldwin@austin.org.au
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 197–202
Information Technology for CRRT
and Dose Delivery Calculator
Zaccaria Ricci
a,b
, Claudio Ronco
b
a
Department of Pediatric Cardiology and Cardiac Surgery, Bambino Gesù Hospital,
Rome, and
b
Department of Nephrology, Dialysis and Transplantation, St. Bortolo
Hospital, Vicenza, Italy
Abstract
Background: The application of information technology (IT) to the field of critical
care nephrology is a process that may reduce errors in care delivery, improve monitoring,
decrease unintentional practice variation, increase the quality and accuracy of delivered treat-
ments. Methods: This review presents some examples of potential applications of recent IT
achievements to clinical practice. Results: The adequacy calculator for continuous therapy
dose prescription was recently shown to accurately predict urea clearance. When clearances
above 60 ml/min where prescribed, the calculator tended to overestimate effective clearances;
this overestimation generally remained within an error of 15%. Nevertheless, the delivered
Kt/V in 24 h will always approach the target value of 1.2. The use of the calculator enabled
strict monitoring of treatments. Furthermore, the so-called ‘next generation’ machines have
technical characteristics in common that allow the highest safety and accuracy levels: some of
these aspects are addressed and commented on in the present review. Conclusion: IT is hav-
ing and will likely have a significant impact on patient safety, practice variation, patient
assessment and monitoring, and documentation of the demographics of acute renal failure and
dialysis. One of the most recent and potentially interesting aspects of IT implementation on
acute dialysis might be the renal replacement dose monitoring and calculation: close control
of the therapy delivery and, eventually, prescription adjustments might be optimized.
Copyright © 2007 S. Karger AG, Basel
Introduction
Information technology (IT) in the medical field is a synonym for techni-
cal modernization and improvement in medical equipment. Over the last few
years, the application of IT to the field of critical care nephrology has been an
ongoing process and, when properly utilized, it may result in both a better
Ricci/Ronco 198
understanding of the disease as well as in improvements in patient outcomes.
Some fundamental targets and priorities of IT have recently been shown by the
Acute Dialysis Quality Initiative (ADQI) [1]. These aspects can be summarized
by four points: (i) current and future IT innovations should be used to reduce
errors in care delivery which could potentially lead to patient harm; (ii) IT
should improve monitoring of the current practice of acute dialysis; (iii) IT
should be implemented to reduce unintentional practice variation without limit-
ing practice preferences, and (iv) advances in technology should be applied to
acute dialysis pump systems in order to improve the quality of acute dialysis
care delivery.
Application of IT to Dialysis Dose Calculation
Software called the ‘Adequacy Calculator for ARF’ has recently been
described [2]. This is a simple Microsoft Excel-based program [3] that calcu-
lates urea clearance and estimates fractional clearance and Kt/V
urea
for all con-
tinuous renal replacement therapy (CRRT) modalities. The calculator works on
the assumption that the urea sieving coefficient is 1 for convective therapies and
that complete saturation of spent dialysate occurs during continuous dialysis. In
a pilot evaluation of a small cohort of patients, the value of clearance predicted
by the calculator correlated significantly to the value obtained from direct
blood and dialysate determination during the first 24 h of treatment, regardless
of the renal replacement modality used. Renal replacement therapy (RRT), in
fact, consists of various modalities which differ in many features including the
continuity of treatment (intermittent vs. continuous), vascular access (arteriove-
nous vs. veno-venous), and the mechanism of solute removal (diffusion vs.
convection). Accordingly, it is difficult to find an ideal marker and an universal
method to compare the doses of these different treatments. Using urea as a
marker molecule, as is done in chronic hemodialysis, treatment dose of CRRT
can be defined by various aspects such as efficiency, intensity, frequency, and
clinical efficacy [4]. Of course, urea simply represents a surrogate of the low
molecular weight toxin. The major shortcoming of the traditional solute
marker-based approach to dialysis dose in acute renal failure (ARF) lies well
beyond any methodological critique of single-solute kinetics-based thinking. In
patients with ARF, the majority of whom are in intensive care, a restrictive
(solute-based only) concept of dialysis dose seems grossly inappropriate. In
these patients the therapeutic needs that can/need to be affected by the ‘dose’ of
RRT are much more than the simple control of small solutes as represented by
urea. They include control of acid-base, tonicity, potassium, magnesium, cal-
cium and phosphate, intravascular volume, extravascular volume and temperature;
Information Technology for CRRT and Dose Delivery Calculator 199
furthermore, the avoidance of unwanted side effects associated with the deliv-
ery of solute control must be considered [4].
Nonetheless, the possibility of implementing a simple software in the rou-
tine management of CRRT prescription is an ideal example of IT: the calculator
can be used to reduce errors in care delivery (mostly, undertreatment or fluid
balance errors); it would improve monitoring of the actual delivery of RRT ses-
sions; it could truly reduce unintentional practice variation without limiting
practice preferences, and if directly implemented in dialysis machine monitors,
the calculator would certainly improve the quality of acute dialysis care deliv-
ery. This tool or its future developments would finally help operators in the
field of prescription, delivery and monitoring of RRT, would improve standard-
ization of dose selection, and could potentially facilitate dialysis prescription in
a large scale trial on RRT dose.
In chronic hemodialysis, treatment dose of RRT is defined as a fractional
clearance, Kt/V, where K is the instantaneous clearance, t is treatment time and
V is the volume of distribution of the marker molecule. This is a dimensionless
parameter that represents the efficacy of treatments, and allows comparison
among different therapies and among different patients. In fact, different instan-
taneous clearances, representing treatment efficiency, can yield to comparable
results in terms of efficacy only if correlated to treatment time and patient total
body weight. A Kt/V value of 1.2 is an established marker of adequacy shown
to correlate with morbidity and mortality in patients with end-stage kidney dis-
ease [5, 6]. Kt/V has not yet been validated as a marker of adequacy in ARF
patients but it seems that a good rationale exists for its use in continuous thera-
pies. Theoretically speaking, in its original concept, clearance was signed to
evaluate renal function among disparate individuals where, however, function
was operating 24 h/day and blood levels were at steady state. This is the reason
why, after some days of CRRT, patients’ urea levels approach a real steady state
(never obtained in the intermittent dialysis population) and post-dialysis
rebound is not present. It is finally reasonable to consider urea distribution
equivalent in total body weight, as in the case of a single pool kinetic model
(spKt/V). Ronco et al. [7] demonstrated an improved outcome with post-
dilution hemofiltration delivered at 35 ml h
1
kg
1
in a population of 450
patients. Setting a spKt/V threshold that could guide clinicians towards ade-
quate treatments, we could possibly meet the target of 35 ml h
1
kg
1
which,
delivered as 24-hour treatment, may translate into a spKt/V of 1.4 independent
of the RRT modality.
We tested the adequacy calculator [2] and found that it was able to accu-
rately predict the delivered urea clearance, apart from which the CRRT modal-
ity was selected; the correlation between prediction and effective delivery
remained high in a range of time of 24 h. When clearances above 60 ml/min
Ricci/Ronco 200
where prescribed, the calculator showed a tendency to overestimate effective
clearances: this overestimation remained generally within an error of 15% (fig. 1).
Considering our results and the dissociation between treatment delivery and
calculator estimation when high clearances are involved, as could be the case of
slow efficiency extended dialysis, a slight correction to prevent overestimation
of effective treatment delivery is strongly advised. Nevertheless, even in the
presence of an error of up to 15%, the delivered Kt/V in 24 h will always
approach the target value of 1.2. The use of the calculator allowed us to strictly
monitor our treatments during the study period, and described an average
10.7% (p 0.05) reduction in therapy delivery, when compared to the pre-
scribed dose. In our population, this delivery reduction was mainly due to oper-
ative treatment time, often shorter than the prescribed treatment time (during
the substitution of bags and filter change, treatment is not administered). Our
observation is consistent with a recent large retrospective analysis [8]. In this
setting, when a ‘standardized’ downtime is foreseen, treatment prescription
might be adjusted to correct for the time of zero clearance.
However, all these considerations must be seen in the light of an absolute
lack of any previous attempt to adjust treatment dose to specific target levels.
Furthermore, a clear understanding of the adequate levels of RRT has still to be
achieved. Despite all the uncertainty surrounding the meaning of adequate
Fig. 1. Bland-Altman analysis illustrates correlation between urea clearance obtained
by the two methods. Urea clearance calculated using the described software (K
CALC
) and urea
clearance obtained by direct measurement on pre-filter blood and effluent samples (K
DEL
).
The mean K
CALC
– K
DEL
value for each treatment is presented on the x axis; the difference
between both methods is shown on the y axis. Parallel lines indicate the standard deviation.
The difference between K
CALC
and K
DEL
tends to increase for mean K
CALC
– K
DEL
values
above 60 ml/min (vertical line). It is possible to distinguish the correlations between K
CALC
and K
DEL
at the start of therapy (T0) and after 18–24 h of uninterrupted therapy (T18).
1401201008060
(K
DEL
K
CALC
)/2
K
DEL
K
CALC
difference
4020
40
30
20
10
10
20
30
0
0
T18
T0
Information Technology for CRRT and Dose Delivery Calculator 201
CRRT dose and the gross shortcomings related to its accuracy in patients with
ARF, the idea that there might be an optimal dose of solute removal continues to
have a powerful hold in the literature. According to some authors this concept
seems optimistic [4]: nonetheless, a dose prescription should be made after the
indication for an extracorporeal blood purification technique is given and the
treatment dose delivered should be carefully monitored.
Application of IT to Dialysis Accuracy and Safety
A new generation of CRRT equipment is being developed more than 30
years after the first continuous treatment was instituted. Today technological evo-
lution is supporting new clinical indications, new safety and accuracy standards
while also trying to reduce the workload for the operators. The so-called ‘next
generation’ machines [9] have in common technical characteristics of high-level
safety requirements, pressure measurements of all crucial segments of the circuit
(catheter inlet and outlet, filter inlet and outlet, ultrafiltrate and dialysate ports),
accurate ultrafiltration and therapy delivery control, obtained by four or more
roller volumetric pumps (blood, replacement, dialysate and effluent), and ultra-
precise scales. The accuracies of these systems allow errors in fluid adminis-
tration/ultrafiltration of a few grams per hour. Another important feature of
this integrated systems is a user-friendly operator interface. These monitors per-
form a complete range of therapies including slow continuous ultrafiltration,
continuous veno-venous hemofiltration, continuous veno-venous hemodialysis, con-
tinuous veno-venous hemodiafiltration, and plasma exchange. Complete moni-
toring of fluid balance and all important events is provided by continuous
recording of the history of the last 24 h (or more) of treatment. When an alarm
occurs a message on the screen suggests the most appropriate intervention
required. Blood flow ranges from 10 to 400 ml/min. Hemofiltration and dia-
lysate range from 0 to about 10l/h. The possibility of making fluid balance errors
during CRRT has been recognized from the beginning of this treatment strategy.
The possibility of errors is continuous, and hardware and circuits are highly chal-
lenged by the uninterrupted utilization. The advent of automated machines and
the implementation of IT has partially overcome this problem. Nevertheless, there
are conditions and operation modes in which the potential for errors is still pre-
sent. In particular, a fluid balance error can lead to fatal outcomes [10]. Features
and alarms, even the safest, can be manipulated by operators creating the oppor-
tunity for serious errors. Physicians and nurses involved in the prescription and
delivery of CRRT should have precise protocols and defined procedures in rela-
tion to machine alarms to prevent major clinical problems. There is no solution to
the unwise utilization of a perfect system.
Ricci/Ronco 202
Conclusion
IT is having and will likely continue to have a significant impact on patient
safety, practice variation, patient assessment and monitoring, and documentation
of the demographics of ARF and dialysis. Continuing work in these areas is nec-
essary together with repeated assessment of the targets achieved in order to aug-
ment IT effects on routine practice. While increased system accuracy and strict
safety controls are already a reality, dialysis dose monitoring and calculation
might be an interesting implementation in future developments of CRRT
machines: treatment prescription could theoretically be guided by a ‘calculator’
in successive steps before RRT is started. Such semi-automated treatment should
implement a close control of therapy delivery and eventually, in a closed loop,
suggest prescription adjustments when the target of the dose is not reached.
References
1 Savage B, Marquardt GW, Paolini F, Schlaeper C: Information technology and acute dialysis. Curr
Opin Crit Care 2002;8:544–548.
2 Ricci Z, Salvatori G, Bonello M, Bolgan I, D’Amico G, Dan M, Piccinni P, Ronco C: In vivo val-
idation of the adequacy calculator for continuous renal replacement therapies. Crit Care
2005;9:R266–R273.
3 Pisitkun T, Tiranathanagul K, Poulin S, et al: A practical tool for determining the adequacy of renal
replacement therapy in acute renal failure patients; in Ronco C, Bellomo R, Brendolan A (eds):
Sepsis, Kidney and Multiple Organ Dysfunction. Contrib Nephrol. Basel, Karger, 2004, vol 144,
pp 329–349.
4 Ricci Z, Bellomo R, Ronco C: Dose of dialysis in acute renal failure. Clin J Am Soc Nephrol
2006;1:380–388.
5 NKF/DOQI Clinical practice guidelines for haemodialysis adequacy: updater 2000. Am J Kidney
Dis 2001;37(suppl 1):S7–S64.
6 Owen WF Jr, Chertow GM, Lazarus JM, Lowrie EG: Dose of haemodialysis and survival: differ-
ences by race and sex. JAMA 1998;280:1764–1768.
7 Ronco C, Bellomo R, Homel P, Brendolan A, Dan M, Piccini P, La Greca G: Effects of different
doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective
randomised trial. Lancet 2000;355:26–30.
8 Venkataraman R, Kellum JA, Palevsky P: Dosing patterns for CRRT at a large academic medical
center in the United States. J Crit Care 2002;17:246–250.
9 Ricci Z, Salvatori G, Bonello M, Ratanarat R, Andrikos E, Dan M, Piccinni P, Ronco C: A new
machine for continuous renal replacement therapy: from development to clinical testing. Expert
Rev Med Devices 2005;2:47–55.
10 Ronco C, Ricci Z, Bellomo R, Baldwin I, Kellum J: Management of fluid balance in CRRT: a
technical approach. Int J Artif Organs 2005;28:765–776.
Zaccaria Ricci
Department of Pediatric Cardiology and Cardiac Surgery, Bambino Gesù Hospital
Piazza S. Onofrio
IT–00100 Rome (Italy)
Tel. 39 06 6859 3333, E-Mail z.ricci@libero.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 203–212
Emerging Biomarkers of Acute
Kidney Injury
Prasad Devarajan
Nephrology and Hypertension, Cincinnati Children’s Hospital Medical Center,
University of Cincinnati, Cincinnati, Ohio, USA
Abstract
Background: Acute kidney injury (AKI) is a major clinical problem with a rising inci-
dence and high mortality rate. The lack of early biomarkers has resulted in an unacceptable
delay in initiating therapies. Methods: Here we will update the reader on promising new
blood and urinary biomarkers that have recently emerged through the application of innova-
tive technologies such as functional genomics and proteomics to human and animal models
of AKI. Results: The most promising biomarkers of AKI for clinical use include a plasma
panel (NGAL and cystatin C) and a urine panel (NGAL, Il-18 and KIM-1). Conclusions:
As they represent tandem biomarkers, it is likely that the AKI panels will be useful for tim-
ing the initial insult and assessing the duration and severity of AKI. Based on the differential
expression of the biomarkers, it is also likely that the AKI panels will distinguish between the
various types and etiologies of AKI. It will be important in future studies to validate the sen-
sitivity and specificity of these biomarker panels in clinical samples from large cohorts and
from multiple clinical situations.
Copyright © 2007 S. Karger AG, Basel
Acute kidney injury (AKI), previously referred to as acute renal failure
(ARF), is a significant and devastating problem in clinical medicine [1–4]. The
incidence of AKI varies from 5% in hospitalized patients to 30–50% of patients
in intensive care units, and there is now substantial evidence that the incidence is
rising at an alarming rate. Despite significant improvements in therapeutics, the
mortality and morbidity associated with AKI remain dismally high. Outstanding
advances in basic research have illuminated the pathogenesis of AKI and have
paved the way for successful therapeutic approaches in animal models. However,
translational research efforts in humans have yielded extremely disappointing
results. A major reason for this is the lack of early markers for AKI, akin to tro-
ponins in acute myocardial disease, and hence an unacceptable delay in initiating
Early Diagnosis and Prevention of AKI
Devarajan 204
therapy [5–7]. In current clinical practice, AKI is typically diagnosed by measur-
ing serum creatinine. Unfortunately, creatinine is an unreliable indicator during
acute changes in kidney function [8]. First, serum creatinine concentrations may
not change until about 50% of kidney function has already been lost. Second,
serum creatinine does not accurately depict kidney function until a steady state
has been reached, which may require several days. However, animal studies have
shown that while AKI can be prevented and/or treated by several maneuvers,
these must be instituted very early after the insult, well before the rise in serum
creatinine. The lack of early biomarkers of AKI in humans has hitherto crippled
our ability to launch potentially effective therapies in a timely manner. Indeed,
human investigations have now clearly established that earlier intervention
improves the chance of ameliorating renal dysfunction [5]. The lack of early bio-
markers has negatively impacted on a number of landmark clinical trials investi-
gating highly promising therapies for AKI [9, 10].
In addition to aiding in the early diagnosis and prediction, biomarkers may
serve several additional purposes in AKI. Thus, biomarkers are also needed for:
(a) discerning AKI subtypes (pre-renal, intrinsic renal, or post-renal); (b) iden-
tifying AKI etiologies (ischemia, toxins, sepsis, or a combination); (c) differen-
tiating AKI from other forms of acute kidney disease (urinary tract infection,
glomerulonephritis, interstitial nephritis); (d) predicting the AKI severity (risk
stratification for prognostication as well as to guide therapy); (e) monitoring the
course of AKI, and (f) monitoring the response to AKI interventions. Furthermore,
AKI biomarkers may play a critical role in expediting the drug development
process. The Critical Path Initiative issued by the FDA in 2004 stated that
‘Additional biomarkers (quantitative measures of biologic effects that provide
informative links between mechanism of action and clinical effectiveness) and
additional surrogate markers (quantitative measures that can predict effective-
ness) are needed to guide product development’. Identification of novel AKI
biomarkers has been designated as a top priority by the American Society of
Nephrology [11]. The concept of developing a new toolbox for earlier diagnosis
of disease states is also prominently featured in the NIH Road Map for biomed-
ical research [12].
Desirable characteristics of clinically applicable AKI biomarkers include:
(a) they should be noninvasive and easy to perform at the bedside or in a stan-
dard clinical laboratory, using easily accessible samples such as blood or urine;
(b) they should be rapidly and reliably measurable using a standardized assay
platform; (c) they should be highly sensitive to facilitate early detection, and
with a wide dynamic range and cutoff values that allow risk stratification, and
(d) they should be highly specific for AKI, and allow the identification of AKI
subtypes and etiologies. This will almost certainly involve a combination of a
panel of biomarkers, along with clinical information.
Emerging Biomarkers of AKI 205
The quest for AKI biomarkers is an area of intense contemporary research
[13–15]. Conventional urinary biomarkers such as casts and fractional excre-
tion of sodium have been insensitive and nonspecific in the clinical setting of
AKI. Other traditional urinary biomarkers such as filtered high molecular
weight proteins and tubular proteins or enzymes have also suffered from a lack
of specificity and a dearth of standardized assays [15]. Fortunately, the applica-
tion of innovative technologies such as cDNA microarrays and proteomics to
human and animal models of AKI has uncovered several novel genes and gene
products that are emerging as biomarkers [13, 16]. The most promising of these
are outlined in table 1, and their current status in human AKI is chronicled in
this article.
Novel AKI Biomarkers under Evaluation in Humans
Neutrophil Gelatinase-Associated Lipocalin
Human neutrophil gelatinase-associated lipocalin (NGAL) was originally
identified as a 25-kDa protein covalently bound to gelatinase from neutrophils.
NGAL is normally expressed at very low levels in several human tissues,
including kidneys, lungs, stomach, and colon. NGAL expression is markedly
induced in injured epithelia. For example, NGAL concentrations are elevated in
the serum of patients with acute bacterial infections, the sputum of subjects
with asthma or chronic obstructive pulmonary disease, and the bronchial fluid
from the emphysematous lung [17]. NGAL was recently identified by micro-
array analysis as one of the earliest and most robustly induced genes and proteins
in the kidney after ischemic or nephrotoxic injury in animal models, and NGAL
protein was easily detected in the blood and urine soon after AKI [18–22].
Table 1. Current status of promising AKI biomarkers in various clinical situations
Biomarker Sample Cardiac Contrast Sepsis or Kidney Commercial
name source surgery nephropathy ICU transplant test?
NGAL Plasma Early Early Early Early Biosite
a
Cystatin C Plasma Intermediate Intermediate Intermediate Intermediate Dade-Behring
NGAL Urine Early Early Early Early Abbott
a
IL-18 Urine Intermediate Absent Intermediate Intermediate None
KIM-1 Urine Intermediate Not tested Not tested Not tested None
a
In development.
Devarajan 206
These findings have spawned a number of translational studies to evaluate
NGAL as a novel biomarker of human AKI.
In a cross-sectional study, human adults in the intensive care unit with
established ARF (defined as a doubling of the serum creatinine in less than
5 days) secondary to sepsis, ischemia, or nephrotoxins displayed a greater than
10-fold increase in plasma NGAL and more than a 100-fold increase in urine
NGAL by Western blotting when compared to normal controls [21]. Both
plasma and urine NGAL correlated highly with serum creatinine levels. Kidney
biopsies in these patients showed intense accumulation of immunoreactive
NGAL in 50% of the cortical tubules. These results identified NGAL as a wide-
spread and sensitive response to established AKI in humans.
In a prospective study of children undergoing cardiopulmonary bypass,
AKI (defined as a 50% increase in serum creatinine) occurred in 28% of the
subjects, but the diagnosis using serum creatinine was only possible 1–3 days
after surgery [23]. In marked contrast, NGAL measurements by Western blot-
ting and ELISA revealed a robust 10-fold or more increase in the urine and
plasma within 2–6 h of surgery in patients who subsequently developed AKI.
Both urine and plasma NGAL were powerful independent predictors of AKI,
with an outstanding area under the curve (AUC) of 0.998 for the 2-hour urine
NGAL and 0.91 for the 2-hour plasma NGAL measurement [23]. Thus, plasma
and urine NGAL emerged as sensitive, specific, and highly predictive early bio-
markers of AKI after cardiac surgery in children. It should be emphasized that
the patients in this study were primarily children with congenital heart disease,
who lacked many of the common comorbid conditions (such as diabetes, hyper-
tension, and atherosclerosis) that are frequently encountered in adults. Never-
theless, these findings have now been confirmed in a prospective study in
adults who developed AKI after cardiac surgery, in whom urinary NGAL was
significantly elevated 1–3 h after the operation [24]. AKI, defined as a 50%
increase in serum creatinine, did not occur until the 3rd postoperative day.
However, patients who did not encounter AKI also displayed a significant
increase in urine NGAL in the early postoperative period, although to a much
lesser degree than in those who subsequently developed AKI. The AUC
reported in the adult study was 0.74 for the 3-hour NGAL and 0.80 for the
18-hour NGAL, which is perhaps reflective of the confounding variables one
typically accumulates as we age.
NGAL has also been evaluated as a biomarker of AKI in kidney transplan-
tation. Biopsies of kidneys obtained 1 h after vascular anastomosis revealed a
significant correlation between NGAL staining intensity and the subsequent
development of delayed graft function [25]. In a prospective multicenter study
of children and adults, urine NGAL levels in samples collected on the day of
transplant clearly identified cadaveric kidney recipients who subsequently
Emerging Biomarkers of AKI 207
developed delayed graft function and dialysis requirement (which typically
occurred 2–4 days later). The receiver-operating characteristic curve for predic-
tion of delayed graft function based on urine NGAL at day 0 showed an AUC of
0.9, indicative of an excellent predictive biomarker [26]. Urine NGAL has also
been shown to predict the severity of AKI and dialysis requirement in a multi-
center study of children with diarrhea-associated hemolytic uremic syndrome
[27]. Preliminary results also suggest that plasma and urine NGAL measure-
ments represent predictive biomarkers of AKI following contrast administration
[28–30] and in the intensive care setting [31].
In summary, NGAL is emerging as a center-stage player in the AKI field,
as a novel predictive biomarker. However, it is acknowledged that the studies
published thus far are small, and NGAL appears to be most sensitive and spe-
cific in relatively uncomplicated patient populations with AKI. NGAL meas-
urements may be influenced by a number of coexisting variables such as
preexisting renal disease [32] and systemic or urinary tract infections [15, 17].
Large multicenter studies to further define the predictive role of plasma and
urine NGAL as a member of the putative ‘AKI panel’ have been initiated, robust
assays for commercialization are nearly complete, and the results are awaited
with optimism.
Cystatin C
Cystatin C is a cysteine protease inhibitor that is synthesized and released
into the blood at a relatively constant rate by all nucleated cells. It is freely fil-
tered by the glomerulus, completely reabsorbed by the proximal tubule, and not
secreted. Since blood levels of cystatin C are not significantly affected by age,
gender, race, or muscle mass, it is a better predictor of glomerular function than
serum creatinine in patients with chronic kidney disease [33]. Urinary excretion
of cystatin C has been shown predict the requirement for renal replacement
therapy about 1 day earlier in patients with established AKI, with an AUC of
0.75 [34]. In the intensive care setting, a 50% increase in serum cystatin C pre-
dicted AKI 1–2 days before the rise in serum creatinine, with an AUC of 0.97
and 0.82, respectively [7].
A recent prospective study compared the ability of serum cystatin C and
NGAL in the prediction of AKI following cardiac surgery [35]. Of 129 patients,
41 developed AKI (defined as a 50% increase in serum creatinine) 1–3 days
after cardiopulmonary bypass. In AKI cases, serum NGAL levels were elevated
2 h after surgery, whereas serum cystatin C levels increased only after 12 h.
Both NGAL and cystatin C levels at 12 h were strong independent predictors of
AKI, but NGAL outperformed cystatin C at earlier time points.
Thus, both NGAL and cystatin C represent promising tandem biomarker
candidates for inclusion in the blood ‘AKI panel’. An advantage of cystatin C is
Devarajan 208
the commercial availability of a standardized immunonephelometric assay,which
is automated and provides results in minutes. Additionally, routine clinical storage
conditions, freeze/thaw cycles, the presence of interfering substances, and the eti-
ology of the AKI do not affect serum cystatin C measurements.
Kidney Injury Molecule-1
Kidney injury molecule-1 (KIM-1) is a transmembrane protein that is
highly overexpressed in dedifferentiated proximal tubule cells after ischemic or
nephrotoxic AKI in animal models [36, 37], and a proteolytically processed
domain is easily detected in urine [38]. In a small human cross-sectional study,
KIM-1 was found to be markedly induced in proximal tubules in kidney biop-
sies from patients with established AKI (primarily ischemic), and urinary KIM-1
distinguished ischemic AKI from pre-renal azotemia and chronic renal disease
[36]. Patients with AKI induced by contrast did not have increased urinary
KIM-1.
Recent preliminary studies have expanded the potential clinical utility of
KIM-1 as a predictive AKI biomarker. In a cohort of 103 adults undergoing
cardiopulmonary bypass, AKI (defined as a 0.3-mg/dl increase in serum creati-
nine) developed in 31% in whom the urinary KIM-1 levels increased by about
40% 2 h after surgery and by more than 100% at the 24-hour time point [39]. In
a small case-control study of 40 children undergoing cardiac surgery, 20 with
AKI (defined as a 50% increase in serum creatinine) and 20 without AKI,
urinary KIM-1 levels were markedly enhanced, with an AUC of 0.83 at the
12-hour time point [40].
Thus, KIM-1 represents a promising candidate for inclusion in the urinary
‘AKI panel’. An advantage of KIM-1 over NGAL is that it appears to be more
specific to ischemic or nephrotoxic kidney injury, and not significantly affected
by chronic kidney disease or urinary tract infections. It is likely that NGAL and
KIM-1 will emerge as tandem biomarkers of AKI, with NGAL being most sen-
sitive at the earliest time points and KIM-1 adding significant specificity at
slightly later time points.
Interleukin-18
Interleukin (IL)-18 is a proinflammatory cytokine that is induced and
cleaved in the proximal tubule, and subsequently easily detected in the urine
following ischemic AKI in animal models [41]. In a cross-sectional study, urine
IL-18 levels were markedly increased in patients with established AKI, but not
in subjects with urinary tract infection, chronic kidney disease, nephritic syn-
drome, or pre-renal failure [42]. Urinary IL-18 levels displayed a sensitivity and
specificity of ⬎90% for the diagnosis of established AKI. In addition, IL-18 in
Emerging Biomarkers of AKI 209
urine obtained on the day of kidney transplantation was significantly increased
in patients who subsequently developed delayed graft function, with an AUC of
0.95. Urinary IL-18 was significantly upregulated up to 48 h prior to the
increase in serum creatinine in patients with acute respiratory distress syn-
drome who develop AKI, with an AUC of 0.73, and represented an independent
predictor of mortality in this cohort [43].
Urinary NGAL and IL-18 were recently shown to represent early, predic-
tive, sequential AKI biomarkers in children undergoing cardiac surgery [44]. In
patients who developed AKI 2–3 days after surgery, urinary NGAL was
induced within 2h and peaked at 6 h whereas urine IL-18 levels increased
around 6 h and peaked at over 25-fold 12 h after surgery (AUC 0.75). Both
NGAL and IL-18 were independently associated with the duration of AKI
among cases. Urine NGAL and IL-18 have also emerged as predictive bio-
markers for delayed graft function following kidney transplantation [26]. In a
prospective multicenter study of children and adults, both NGAL and IL-18 in
urine samples collected on the day of transplant predicted delayed graft func-
tion and dialysis requirement with AUC of 0.9.
Thus, IL-18 also represents a promising candidate for inclusion in the uri-
nary ‘AKI panel’. IL-18 is more specific to ischemic AKI, and not affected by
nephrotoxins, chronic kidney disease or urinary tract infections. It is likely that
NGAL, IL-18 and KIM-1 will emerge as sequential urinary biomarkers of
AKI.
Conclusions
The tools of modern science have provided us with promising novel bio-
markers for AKI, with potentially high sensitivity and specificity. These
include a plasma panel (NGAL and cystatin C) and a urine panel (NGAL,
IL-18, and KIM-1). Since they represent tandem biomarkers, it is likely that
the AKI panels will be useful for timing the initial insult and assessing the
duration of AKI (analogous to the cardiac panel for evaluating chest pain).
Based on the differential expression of the biomarkers, it is also likely that the
AKI panels will help distinguish between the various types and etiologies of
AKI. However, they have hitherto been tested only in small studies and in a
limited number of clinical situations. It will be important in future studies
to validate the sensitivity and specificity of these biomarker panels in clini-
cal samples from large cohorts and from multiple clinical situations. Such
studies will be markedly facilitated by the availability of commercial tools for
the reliable and reproducible measurement of biomarkers across different
laboratories.
Devarajan 210
References
1 Lameire N, Van Biesen W, Vanholder R: Acute renal failure. Lancet 2005;365:417–430.
2 Schrier RW, Wang W, Poole B, Mitra A: Acute renal failure: definitions, diagnosis, pathogenesis,
and therapy. J Clin Invest 2004;114:5–14.
3 Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C,
Macedo E, Gibney N, Tolwani A, Ronco C; Beginning and Ending Supportive Therapy for the
Kidney (BEST Kidney) Investigators: Acute renal failure in critically ill patients: a multinational,
multicenter study. JAMA 2005;294:813–818.
4 Devarajan P: Update on mechanisms of ischemic acute kidney injury. J Am Soc Nephrol 2006;17:
1503–1520.
5 Schrier RW: Need to intervene in established acute renal failure. J Am Soc Nephrol 2004;15:
2756–2758.
6 Hewitt SM, Dear J, Star RA: Discovery of protein biomarkers for renal diseases. J Am Soc
Nephrol 2004;15:1677–1689.
7 Herget-Rosenthal S, Marggraf G, Hüsing J, Goring F, Pietruck F, Janssen O, Phillip T, Kribben A:
Early detection of acute renal failure by serum cystatin C. Kidney Int 2004;66:1115–1122.
8 Bellomo R, Kellum JA, Ronco C: Defining acute renal failure: physiological principles. Intensive
Care Med 2004;30:33–37.
9 Allgren RL, Marbury TC, Rahman SN, Weisberg LS, Fenves AZ, Lafayette RA, Sweet RM,
Genter FC, Kurnik BR, Conger JD, Sayegh MH: Anaritide in acute tubule necrosis. Auriculin
Anaritide Acute Renal Failure Study Group. N Engl J Med 1997;336:828–834.
10 Hirschberg R, Kopple J, Lipsett P, Benjamin E, Minei J, Albertson T, Munger M, Metzler M,
Zaloga G, Murray M, Lowry S, Conger J, McKeown W, O’Shea M, Baughman R, Wood K, Haupt M,
Kaiser R, Simms H, Warnock D, Summer W, Hintz R, Myers B, Haenftling K, Capra W, Pike M,
Guler H-P: Multicenter clinical trial of recombinant human insulin-like growth factor 1 in patients
with acute renal failure. Kidney Int 1999;55:2423–2432.
11 American Society of Nephrology: American Society of Nephrology Renal Research Report. J Am
Soc Nephrol 2005;16:1886–1893.
12 Zerhouni E: The NIH roadmap. Science 2003;302:63–65.
13 Devarajan P, Mishra J, Supavekin S, Patterson LT, Potter SS: Gene expression in early ischemic
renal injury: clues towards pathogenesis, biomarker discovery, and novel therapeutics. Mol Genet
Metab 2003;80:365–376.
14 Han WK, Bonventre JV: Biologic markers for the early detection of acute kidney injury. Curr Opin
Crit Care 2004;10:476–482.
15 Zhou H, Hewitt SM, Yuen PST, Star RA: Acute kidney injury biomarkers – needs, present status,
and future promise. Nephrology Self Assessment Program, American Society of Nephrology.
NephSAP 2006;5:63–71.
16 Nguyen M, Ross G, Dent C, Devarajan P: Early prediction of acute renal injury using urinary
proteomics. Am J Nephrol 2005;25:318–326.
17 Xu S, Venge P: Lipocalins as biochemical markers of disease. Biochim Biophys Acta 2000;1482:
298–307.
18 Supavekin S, Zhang W, Kucherlapati R, Kaskel FJ, Moore LC, Devarajan P: Differential gene
expression following early renal ischemia-reperfusion. Kidney Int 2003;63:1714–1724.
19 Mishra J, Ma Q, Prada A, Zahedi K, Yang Y, Barasch J, Devarajan P: Identification of neutrophil
gelatinase-associated lipocalin as a novel urinary biomarker for ischemic injury. J Am Soc
Nephrol 2003;4:2534–2543.
20 Mishra J, Mori K, Ma Q, Kelly C, Barasch J, Devarajan P: Neutrophil Gelatinase-Associated
Lipocalin (NGAL): a novel urinary biomarker for cisplatin nephrotoxicity. Am J Nephrol 2004;24:
307–315.
21 Mori K, Lee HT, Rapoport D, Drexler I, Foster K, Yang J, Schmidt-Ott KM, Chen X, Li JY,
Weiss S, Mishra J, Cheema FH, Markowitz G, Suganami T, Sawai K, Mukoyama M, Kunis C,
D’Agati V, Devarajan P, Barasch J: Endocytic delivery of lipocalin-siderophore-iron complex
rescues the kidney from ischemia-reperfusion injury. J Clin Invest 2005;115:610–621.
Emerging Biomarkers of AKI 211
22 Schmitt-Ott KM, Mori K, Kalandadze A, Li J-Y, Paragas N, Nicholas T, Devarajan P, Barasch J:
Neutrophil gelatinase-associated lipocalin-mediated iron traffic in kidney epithelia. Curr Opin
Nephrol Hypertens 2005;15:442–449.
23 Mishra J, Dent C, Tarabishi R, Mitsnefes MM, Ma Q, Kelly C, Ruff SM, Zahedi K, Shao M, Bean J,
Mori K, Barasch J, Devarajan P: Neutrophil gelatinase-associated lipocalin (NGAL) as a bio-
marker for acute renal injury following cardiac surgery. Lancet 2005;365:1231–1238.
24 Wagener G, Jan M, Kim M, Mori K, Barasch JM, Sladen RN, Lee HT: Association between
increases in urinary neutrophil gelatinase-associated lipocalin and acute renal dysfunction after
adult cardiac surgery. Anesthesiology 2006;105:485–491.
25 Mishra J, Ma Q, Kelly C, Mitsnefes M, Mori K, Barasch J, Devarajan P: Kidney NGAL is a novel
early marker of acute injury following transplantation. Pediatr Nephrol 2006;21:856–863.
26 Parikh CR, Jani A, Mishra J, Ma Q, Kelly C, Barasch J, Edelstein CL, Devarajan P: Urine NGAL
and IL-18 are predictive biomarkers for delayed graft function following kidney transplantation.
Am J Transplant 2006;6:1639–1645.
27 Trachtman H, Christen E, Cnaan A, Patrick J, Mai V, Mishra J, Jain A, Bullington N, Devarajan P:
Urinary neutrophil gelatinase-associated lipocalin in D⫹HUS: A novel marker of renal injury.
Pediatr Nephrol 2006;21:989–994.
28 Bachorzewska-Gajewska H, Malyszko J, Sitniewska E, Malyszko JS, Dobrzycki S: Neutrophil
gelatinase-associated lipocalin and renal function after percutaneous coronary interventions. Am J
Nephrol 2006;26:287–292.
29 Malyszko J, Bachorzewska-Gajewska H, Malyszko JS, Dobrzycki S: Could NGAL predict renal
function after percutaneous coronary interventions? J Am Soc Nephrol 2006;17:406A.
30 Devarajan P, Hirsch R, Dent C, Pfriem H, Allen J, Beekman R, Mishra J, Ma Q, Kelly C, Mitsnefes M:
NGAL is an early predictive biomarker of acute kidney injury following contrast administration.
J Am Soc Nephrol 2006;17:48A.
31 Zappitelli M, Washburn K, Arikan AA, Mishra J, Loftis L, Devarajan P, Goldstein SL: Urine
NGAL is an early predictive biomarker of acute kidney injury in critically ill children. J Am Soc
Nephrol 2006;17:404A.
32 Mitsnefes M, Kathman T, Mishra J, Kartal J, Khoury P, Nickolas T, Barasch J, Devarajan P: Serum
NGAL as a marker of renal function in children with chronic kidney disease. Pediatr Nephrol
2007;22:101–108.
33 Dharnidharka VR, Kwon C, Stevens G: Serum cystatin C is superior to serum creatinine as a
marker of kidney function: a meta-analysis. Am J Kidney Dis 2002;40:221–226.
34 Herget-Rosenthal S, Poppen D, Husing J, Marggraf G, Pietruck F, Jakob HG, Phillip T, Kribben A:
Prognostic value of tubular proteinuria and enzymuria in nonoliguric acute tubular necrosis. Clin
Chem 2004;50:552–558.
35 VandeVoorde RG, Katlman TI, Ma Q, Kelly C, Mishra J, Dent CA, Mitsnefes MM, Devarajan P:
Serum NGAL and cystatin C as predictive biomarkers for acute kidney injury. J Am Soc Nephrol
2006;17:404A.
36 Han WK, Bailly V, Abichandani R, Thadani R, Bonventre JV: Kidney injury molecule-1 (KIM-1):
a novel biomarker for human renal proximal tubule injury. Kidney Int 2002;62:237–244.
37 Ichimura T, Hung CC, Yang SA, Stevens JL, Bonventre JV: Kidney injury molecule-1: a tissue and
urinary biomarker for nephrotoxicant-induced renal injury. Am J Physiol Renal Physiol 2004;286:
F552–F563.
38 Vaidya VS, Ramirez V, Ichimura T, Bobadilla NA, Bonventre JV: Urinary kidney injury molecule-1:
a sensitive quantitative biomarker for early detection of kidney tubular injury. Am J Physiol Renal
Physiol 2006;290:F517–F529.
39 Liangos O, Han WK, Wald R, Perianayagam MC, Mackinnon RW, Dolan N, Warner KG, Symes JF,
Bonventre JV, Jaber BL: Urinary kidney injury molecule-1 level is an early and sensitive marker of
acute kidney injury following cardiopulmonary bypass. J Am Soc Nephrol 2006;17:403A.
40 Han WK, Waikar SS, Johnson A, Curhan GC, Devarajan P, Bonventre JV: Urinary biomarkers for
early detection of acute kidney injury. J Am Soc Nephrol 2006;17:403A.
41 Melnikov VY, Ecder T, Fantuzzi G, Siegmund B, Lucia MS, Dinarello CA, Schrier RW, Edelstein CL:
Impaired IL-18 processing protects caspase-1 deficient mice from ischemic acute renal failure.
J Clin Invest 2001;107:1145–1152.
Devarajan 212
42 Parikh CR, Jani A, Melnikov VY, Faubel S, Edelstein CL: Urinary interleukin-18 is a marker of
human acute tubular necrosis. Am J Kidney Dis 2004;43:405–414.
43 Parikh CR, Abraham E, Ancukiewicz M, Edelstein CL: Urine IL-18 is an early diagnostic marker
for acute kidney injury and predicts mortality in the intensive care unit. J Am Soc Nephrol
2005;16:3046–3052.
44 Parikh CR, Mishra J, Thiessen-Philbrook H, Dursun B, Ma Q, Kelly C, Dent C, Devarajan P,
Edelstein CL: Urinary IL-18 is an early predictive biomarker of acute kidney injury after cardiac
surgery. Kidney Int 2006;70:199–203.
Prasad Devarajan, MD
Nephrology and Hypertension, MLC 7022, Cincinnati Children’s Hospital Medical Center
3333 Burnet Avenue
Cincinnati, OH 45229–3039 (USA)
Tel. ⫹1 513 636 4531, Fax ⫹1 513 636 7407, E-Mail prasad.devarajan@cchmc.org
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 213–219
Diagnosis of Acute Kidney Injury: From
Classic Parameters to New Biomarkers
Joseph V. Bonventre
Renal Division, Brigham and Women’s Hospital and Department of Medicine, Harvard
Stem Cell Institute, Harvard Medical School and Harvard-Massachusetts Institute of
Technology, Division of Health Sciences and Technology, Boston, Mass., USA
Abstract
A change in serum creatinine is the standard metric used to define and monitor the pro-
gression of acute kidney injury (AKI). This marker is inadequate for a number of reasons
including the fact that changes in serum creatinine are delayed in time after kidney injury and
hence creatinine is not a good indicator to use in order to target therapy in a timely fashion.
There is an urgent need for early biomarkers for the diagnosis of AKI. There is also a need for
biomarkers that will be predictive of outcome and which can be used to monitor therapy. There
are a limited number of biomarkers that are being validated by a number of groups and from
this list clinically useful reagents are likely to be derived over the next few years. In this article
the status of 5 potential urinary biomarkers for AKI are discussed: kidney injury molecule-1,
N-acetyl--
D-glucosaminidase, neutrophil gelatinase-associated lipocalin, cystatin C, and
interleukin-18. Considerable progress has been made although much continues to be needed to
validate these markers for routine clinical use. Armed with these new tools the future will look
much brighter for the patient with AKI as it is likely that early diagnosis and better predictors of
outcome will lead to new therapies which can be introduced earlier in the course of disease.
Copyright © 2007 S. Karger AG, Basel
Acute kidney injury (AKI) is a common condition that is associated with a
high mortality rate. It has been recognized that routinely used measures of renal
function, such as blood urea nitrogen and serum creatinine concentrations, sig-
nificantly increase only after substantial kidney injury occurs and then with a
time delay. In addition multiple factors can effect blood urea nitrogen and
serum creatinine. Serum creatinine is known to be secreted by the renal tubule
and this secretion can be modified by pharmacological agents. Serum levels
can be modified by changes in volume status which often occurs especially in
postoperative patients who are at risk of AKI and for whom early diagnosis of
AKI would be desirable. Creatinine production varies greatly among individuals
Bonventre 214
and consequently the changes in serum creatinine will vary greatly from indi-
vidual to individual even in the context of an equivalent change in renal func-
tion. There are also technical problems with the assay as there are urine
components which can interfere.
The insensitivity of tests used to detect injury to the kidney delays the
diagnosis in humans, making it particularly challenging to administer putative
therapeutic agents in a timely fashion. While significant attempts have been
made to increase the utility of serum creatinine in the steady state with the
introduction of equations to calculate an ‘effective GFR’ [1], this does not alle-
viate the problem with this biomarker of renal function when kidney function is
changing rapidly as it does with AKI. Furthermore, the insensitivity of tradi-
tional markers of kidney damage affects the evaluation of toxicity in preclinical
studies by allowing drug candidates, which have low but nevertheless important
nephrotoxic side effects in animals, to pass the preclinical safety criteria only to
be found to be clinically nephrotoxic at great human costs. In this brief review I
will summarize the importance of better biomarkers for kidney injury and dis-
cuss the current status of specific biomarkers to detect preclinical and clinical
renal injury and potentially predict outcome of AKI in humans.
Urgent Need for Early Biomarkers in the Management of AKI
One of the disappointments in the care of patients with AKI is that little
has been shown to be effective in preventing the syndrome or arresting its pro-
gression. We have support therapies that can provide correction of metabolic
consequences of kidney failure but despite this the mortality rates with AKI
remain very high, especially in patients in the ICU. It has recently been recog-
nized that even small changes in kidney function lead to marked increases in
mortality [2, 3]. Thus it is particularly important to diagnose injury early.
Furthermore progress in the field has been significantly impaired by inadequate
clinical studies of therapeutic agents because treatment protocols are adversely
effected by late diagnosis. By the time creatinine increases and the study inves-
tigators confirm this increase, the acute event resulting in the AKI is long past.
This is analogous to initiating treatment in patients with acute myocardial
infarction 48 h after coronary occlusion.
What Are Biomarkers?
Biomarkers can be any parameter in a patient that can be quantitated and
provides useful information about a normal or pathobiological state. We will
Diagnosis of AKI: Biomarkers 215
limit our discussion primarily to proteins or other molecules which are found in
blood or urine. A biomarker may be used to help stratify patients at risk of AKI,
or diagnose or predict the natural history of a disease process. A biomarker
might be used as an indicator to guide the timing and type of therapy, to predict
mortality and monitor the response to therapy. Response to therapy can be
quantitated with a biomarker. The best biomarkers are measured noninvasively,
and are easy to determine in a timely fashion so that clinical decision making
can be facilitated. Ideally the measurement is made at the bedside. The bio-
marker should not be affected by changes in the composition of the fluid in
which it is measured, especially as patients with kidney injury have marked
changes in the composition of blood and urine. It is possible that no single bio-
marker will be ideally suited to satisfy all these needs. It is possible that a set of
biomarkers will ultimately be used.
To validate biomarkers it would be ideal if there were a ‘gold standard’ to
which the marker is being compared. Unfortunately this is not a trivial task as we
have already indicated that creatinine is not such a standard. Pathology would be
very useful but we routinely have few AKI patients in whom renal biopsies have
been obtained. Urinary casts and changes in the fractional excretion of sodium
may be helpful but both of these markers have their own inherent insensitivity
and problems with specificity, especially with quantitation of injury. In this con-
text it is very useful to consider animal data because in the animal pathology is
more easily obtained and a detailed time course can be evaluated relative to a
very precise time definition of the insult. Newly proposed biomarkers should be
simultaneously and quantitatively compared with those used in past.
Potential Biomarkers for AKI
There are a limited number of biomarkers that are being validated by vari-
ous groups and from this list clinically useful reagents are likely to derive over
the next few years. Of these biomarkers, kidney injury molecule-1 (KIM-1),
N-acetyl -
D-glucosaminidase (NAG), neutrophil gelatinase-associated lipocalin
(NGAL), cystatin C, and interleukin-18 (IL-18) will be discussed.
Kidney Injury Molecule-1
KIM-1 (or Kim-1 in rodents) encodes a type I cell membrane glycoprotein
containing, in its extracellular portion, a novel six-cysteine immunoglobulin-
like domain and a threonine/serine and proline-rich domain characteristic of
mucin-like O-glycosylated proteins, suggesting its potential involvement in
cell–cell and/or cell–matrix interactions [4] (fig. 1). We found the mRNA that
encodes this protein to be upregulated in the kidney more than any other mRNA
Bonventre 216
in response to experimental AKI in rodents. This was confirmed using the
nephrotoxin, cisplatin, in an unbiased genomic approach taken by a pharmaceuti-
cal consortium [5]. After proximal tubular kidney injury, the ectodomain of
KIM-1 is shed from proximal tubule cells in vivo into the urine in rodents [6–8]
and humans [9]. In preclinical and clinical studies urinary Kim-1 serves as an
early diagnostic indicator of kidney injury. In rodents Kim-1 is increased earlier
than any of the conventional biomarkers, e.g. plasma creatinine, blood urea nitro-
gen, glycosuria, increased proteinuria or increased urinary NAG levels [8, 9].
N-Acetyl-b-
D-Glucosaminidase
NAG is a proximal tubular brush border-specific lysosomal enzyme which
has been used as an indicator of renal proximal tubule injury. In our hands in
humans and rodents this biomarker performs well [8, 10], although not as well
as Kim-1 in rodents where it is less sensitive. Appropriate comparisons are not
complete in humans. It is know that some metals and other nephrotoxicants can
directly inhibit NAG activity, and therefore in such cases NAG cannot be used
as a biomarker [11, 12].
Neutrophil Gelatinase-Associated Lipocalin
NGAL is a 25-kDa protein bound to gelatinase which is expressed and
secreted by hepatocytes, neutrophils, and kidney epithelial cells during inflam-
mation and AKI [13]. NGAL is involved in iron shuttling from extracellular to
N-glycan
171
Membrane
Cytoplasmic
domain
Tyr-P
Metalloproteinase
CCCC
CC
CCCC CC TSP Rich
117
290 311 359
Mucin domainIg-like domain
Cytoplasmic
domain
Fig. 1. Drawing showing the structure of KIM-1. The protein is a type-1 membrane
protein with most of the protein made up of an extracellular domain that consists of a signal
peptide, and Ig domain and a mucin domain. There is a short cytoplasmic domain with at
least one important tyrosine phosphorylation domain. The protein is cleaved by a metallo-
proteinase and the ectodomain appears in the urine of rodents and humans with AKI.
Diagnosis of AKI: Biomarkers 217
intracellular compartments. NGAL was identified as being one of the seven
genes whose expression was upregulated ⬎10-fold within the first few hours
after ischemic renal injury in a mouse model. It has recently been reported that
on day 0 both urine NGAL and IL-18 (see below) predicted the trend in serum
creatinine in the post-transplant period after adjusting for effects of age, gender,
race, urine output, and cold ischemia time [14].
Cystatin C
Cystatin C is a 13-kDa cysteine protease inhibitor and one of the 12 mem-
bers of human cystatin family [15]. Cystatin C is unique amongst all cystatins
as it seems to be produced by all human nucleated cells. It has been argued by
some that serum cystatin C should be used to replace creatinine clearance in
patients with chronic kidney disease. For example, while a meat meal increases
serum creatinine and reduces the calculated effective GFR significantly, there is
no significant change in the serum cystatin concentration [16]. Herget-
Rosenthal et al. [17] showed that urinary excretion of cystatin C at entry to the
study in patients with non-oliguric acute tubular necrosis had a higher sensitiv-
ity and specificity than 
1
-microglobulin, NAG, and the Liano severity of ill-
ness score [18] in predicting the requirement for renal replacement therapy.
Interleukin-18
IL-18 is a cytokine whose levels have been reported to be elevated in the
urine of patients with AKI and delayed graft function compared with normal
subjects and patients with pre-renal azotemia, urinary tract infection, chronic
renal insufficiency, and nephritic syndrome [19]. Parikh et al. [20] have
reported that IL-18 is an early, predictive biomarker of AKI after cardiopul-
monary bypass, and that NGAL and IL-18 are increased in tandem after car-
diopulmonary bypass.
Conclusions
Acute kidney injury is a complex entity with multiple causes. It is possible
that no single marker will provide the levels of sensitivity and specificity nec-
essary to be clinically useful across the full spectrum of AKI. It is possible that
a panel of markers will be optimal to satisfy all the previously mentioned
requirements. Laboratories carrying out biomarker studies are collaborating
with clinical investigators performing prevention and treatment studies. A net-
work should be developed within the biomarker field. Multiple biomarkers
should be studied simultaneously. For both the discovery of new biomarkers
and validation of the ones at hand, it is necessary to have close interactions
Bonventre 218
between clinician scientists and laboratory scientists. Armed with these new
tools the future will look much brighter for the patient with AKI as it is likely
that clinical trials will lead to new therapies which can be introduced earlier in
the course of disease and be more effective in leading to mitigation of the sever-
ity of disease and potentiation of recovery of this fundamentally reversible
condition.
Acknowledgments
This work was supported by the National Institutes of Health (grants DK 39773,
DK54741, DK72381).
References
1 Levey AS, Coresh J, Balk E, Kausz AT, Levin A, Steffes MW, Hogg RJ, Perrone RD, Lau J,
Eknoyan G: National Kidney Foundation practice guidelines for chronic kidney disease: evalua-
tion, classification, and stratification. Ann Intern Med 2003;139:137–147.
2 Chertow GM, Burdick E, Honour M, Bonventre JV, Bates DW: Acute kidney injury, mortality,
length of stay, and costs in hospitalized patients. J Am Soc Nephrol 2005;16:3365–3370.
3 Lassnigg A, Schmidlin D, Mouhieddine M, Bachmann LM, Druml W, Bauer P, Hiesmayr M:
Minimal changes of serum creatinine predict prognosis in patients after cardiothoracic surgery: a
prospective cohort study. J Am Soc Nephrol 2004;15:1597–1605.
4 Bailly V, Zhang Z, Meier W, Cate R, Sanicola M, Bonventre JV: Shedding of kidney injury mole-
cule-1, a putative adhesion protein involved in renal regeneration. J Biol Chem 2002;277:
39739–39748.
5 Amin RP, Vickers AE, Sistare F, Thompson KL, Roman RJ, Lawton M, Kramer J, Hamadeh HK,
Collins J, Grissom S, et al: Identification of putative gene based markers of renal toxicity. Environ
Health Perspect 2004;112:465–479.
6 de Borst MH, van Timmeren MM, Vaidya VS, de Boer RA, van Dalen MB, Kramer AB, Schuurs
TA, Bonventre JV, Navis G, van Goor H: Induction of kidney injury molecule-1 in homozygous
Ren2 rats is attenuated by blockade of the renin-angiotensin system or p38 MAP kinase. Am J
Physiol Renal Physiol 2007;292:F313–F320.
7 Ichimura T, Hung CC, Yang SA, Stevens JL, Bonventre JV: Kidney injury molecule-1:a tissue and
urinary biomarker for nephrotoxicant-induced renal injury. Am J Physiol Renal Physiol
2004;286:F552–F563.
8 Vaidya VS, Ramirez V, Ichimura T, Bobadilla NA, Bonventre JV: Urinary kidney injury molecule-1:
a sensitive quantitative biomarker for early detection of kidney tubular injury. Am J Physiol Renal
Physiol 2006;290:F517–F529.
9 Han WK, Bailly V, Abichandani R, Thadhani R, Bonventre JV: Kidney injury molecule-1 (KIM-
1): a novel biomarker for human renal proximal tubule injury. Kidney Int 2002;62:237–244.
10 Liangos O, Perianayagam MC, Vaidya VS, Han WK, Wald R, Tighiouart H, Mackinnon RW, Li L,
Balakrishnan VS, Pereira BJ, et al: Urinary N-acetyl-{beta}-(D)-glucosaminidase activity and kid-
ney injury molecule-1 level are associated with adverse outcomes in acute renal failure. J Am Soc
Nephrol 2007;Epub ahead of print.
11 Wiley RA, Choo HY, Traiger GJ: The effect of nephrotoxic furans on urinary N-acetylglu-
cosaminidase levels in mice. Toxicol Lett 1982;14:93–96.
12 Vaidya VS, Shankar K, Lock EA, Bucci TJ, Mehendale HM: Renal injury and repair following
S-1,2 dichlorovinyl-L-cysteine administration to mice. Toxicol Appl Pharmacol 2003;188:110–121.
Diagnosis of AKI: Biomarkers 219
13 Schmidt-Ott KM, Mori K, Li JY, Kalandadze A, Cohen DJ, Devarajan P, Barasch J: Dual action of
neutrophil gelatinase-associated lipocalin. J Am Soc Nephrol 2007;18:407–413.
14 Parikh CR, Jani A, Mishra J, Ma Q, Kelly C, Barasch J, Edelstein CL, Devarajan P: Urine NGAL
and IL-18 are predictive biomarkers for delayed graft function following kidney transplantation.
Am J Transplant 2006;6:1639–1645.
15 Filler G, Bokenkamp A, Hofmann W, Le Bricon T, Martinez-Bru C, Grubb A: Cystatin C as a
marker of GFR–history, indications, and future research. Clin Biochem 2005;38:1–8.
16 Preiss DJ, Godber IM, Lamb EJ, Dalton RN, Gunn IR: The influence of a cooked-meat meal on
estimated glomerular filtration rate. Ann Clin Biochem 2007;44:35–42.
17 Herget-Rosenthal S, Poppen D, Husing J, Marggraf G, Pietruck F, Jakob HG, Philipp T, Kribben
A: Prognostic value of tubular proteinuria and enzymuria in nonoliguric acute tubular necrosis.
Clin Chem 2004;50:552–558.
18 Liano F, Junco E, Pascual J, Madero R, Verde E: The spectrum of acute renal failure in the inten-
sive care unit compared with that seen in other settings. The Madrid Acute Renal Failure Study
Group. Kidney Int Suppl 1998;66:S16–S24.
19 Parikh CR, Jani A, Melnikov VY, Faubel S, Edelstein CL: Urinary interleukin-18 is a marker of
human acute tubular necrosis. Am J Kidney Dis 2004;43:405–414.
20 Parikh CR, Mishra J, Thiessen-Philbrook H, Dursun B, Ma Q, Kelly C, Dent C, Devarajan P,
Edelstein CL: Urinary IL-18 is an early predictive biomarker of acute kidney injury after cardiac
surgery. Kidney Int 2006;70:199–203.
Joseph V. Bonventre, MD, PhD
Brigham and Women’s Hospital, Renal Division, Harvard Institutes of Medicine
4 Blackfan Circle
Boston, MA 02115 (USA)
Tel. ⫹1 617 525 5960, Fax ⫹1 617 525 5965, E-Mail joseph_bonventre@hms.harvard.edu
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 220–226
Endotoxin and Cytokine Detection
Systems as Biomarkers for
Sepsis-Induced Renal Injury
Steven M. Opal
Infectious Disease Division, Brown Medical School, Providence, R.I., USA
Abstract
Background: A reliable biomarker as an indicator of the presence of severe sepsis is
an unmet medical need. Methods: Review of recent literature on this topic focusing upon
endotoxin and cytokine assays. Results: The ideal biomarker for sepsis would be readily
available, technically easy to perform with a quick turn-around time, inexpensive, highly
specific, very sensitive, and preferably highly correlated in quantitative terms with disease
severity. Such a test would provide early diagnostic accuracy, prognostic information, and
indicate responsiveness to treatment interventions. Regrettably no such biomarker exists for
sepsis at present, and it is not likely that such an ideal assay will be developed in the foresee-
able future. Conclusions: Despite their shortcomings, a number of existing and candidate
biomarker assays are available and can provide some useful information to the clinician car-
ing for septic patients. The relative merits of endotoxin measurement, interleukin-6 levels
and a variety of other sepsis markers are reviewed. Full implementation of these biomarkers
may improve diagnostic accuracy over the standard clinical criteria for sepsis.
Copyright © 2007 S. Karger AG, Basel
One of the major unmet medical needs in the management of sepsis is the
development of reliable methods to differentiate patients who are becoming
septic from those who have a physiological systemic response to infection. As
Machiavelli noted over 500 years ago in his famous treatise The Prince, ‘Hectic
fever at its inception is difficult to recognize but easy to treat, left untended it
becomes easy to recognize but too difficult to treat’. Regrettably this same
statement could be made about sepsis even today. Despite intense clinical and
laboratory monitoring, early recognition of the early phases of sepsis remains a
fundamental challenge for clinicians and investigators alike. When does the
systemic host response to infection convert from a controlled and advantageous
LPS and IL-6 as Biomarkers 221
host defense response to a deleterious pathophysiological process capable of
diffuse tissue injury with potentially lethal consequences?
There is no single diagnostic test that defines the clinical syndrome that we
now refer to as sepsis, and clinical criteria alone are notoriously inaccurate in
determining the precise time when an injurious and dysfunctional septic host
response has developed. We can readily measure evidence of end-organ injury
and tissue hypoperfusion (defined as severe sepsis), but these are late findings of
established organ dysfunction. Waiting to recognize that a patient was septic in
retrospect is often too late for successful intervention to prevent further tissue
injury. For these reasons, a biomarker assay has been much sought after to assist
the clinician in the detection of the early signs of sepsis. The most promising bio-
markers have been detection assays for microbial mediators or early host
response mediators indicative of a pathologic state of an injurious host response.
Some of these candidate biomarkers will be briefly described in the following
paragraphs. Their relative merits as each biomarker are summarized in table 1.
Table 1. Summary of some of the currently available biomarkers for the recognition of sepsis
Biomarker Advantages Disadvantages
Endotoxin (LPS) High levels predictive of outcome, Difficult assays and not readily
approved assays available available, not highly specific
Interleukin-6 Easily measured and high levels Not available in clinical
predictive of outcome laboratories, highly variable,
nonspecific
Procalcitonin Well-documented predictive value, Not available in many clinical
reliable, easily measured laboratories
C-Reactive protein Readily available assays, easily Nonspecific assay, limited
measured acute phase protein predictive value in sepsis
aPTT wave form analysis Easy, rapid assay if equipment is Limited clinical experience,
available, highly predictive mechanism of action unclear
Coagulation markers Standard assays are available, Not readily available as rapid
(protein C, platelets, D-dimer, levels correlated with outcome assays, indirect and nonspecific
TAT, F1.2, etc.)
HLA-DR expression Clear pathophysiological link with Not readily available, less
sepsis-induced immune valuable for early sepsis
suppression
Soluble TREM-1 Highly predictive in early clinical Limited experience, not readily
studies available clinically
aPTT ⫽ Activated partial thromboplastin time; F1.2 ⫽ prothrombin fragment 1.2; HLA ⫽ human leukocyte
antigen; LPS ⫽ lipopolysaccharide; TAT ⫽ thrombin antithrombin complex; TREM ⫽ triggering receptor expressed
on myeloid cells.
Opal 222
Endotoxin (LPS) Assays
Bacterial endotoxin, or lipopolysaccharide (LPS), is an intrinsic compo-
nent of the outer membrane of gram-negative bacteria and is essential for the
viability of enteric bacteria [1]. Endotoxin functions as an alarm molecule alert-
ing the host to the presence of microbial invasion by gram-negative bacteria.
The potentially lethal consequences following endotoxin release into the circu-
lation is attributable to the exaggerated host response to the endotoxin, rather
than the endotoxin molecule itself.
LPS is a di-phosphorylated, polar macromolecule that contains hydropho-
bic elements within its lipid A core structure, and hydrophilic elements in its
repeating polysaccharide surface components. LPS forms microaggregates
(micelles). LPS signaling is mediated by interactions with a hepatically derived,
acute-phase plasma protein known as LPS-binding protein (LBP) [2, 3]. LBP
functions as a shuttle carrier protein transferring LPS monomers to CD14. CD14
is a glycosyl phosphatidylinositol-linked protein found primarily on the cell sur-
faces of myeloid cells. After docking to membrane-bound CD14, LPS is trans-
ferred to the essential soluble adaptor protein MD2. This LPS-MD2 complex is
then presented to the extracellular domain of TLR4 where they aggregate on
lipid rafts on the cell surface to activate intracellular signaling. Through a well-
characterized series of activation steps by specific threonine/serine kinases,
intracellular signaling ultimately leads to phosphorylation, ubiquinylation and
degradation of inhibitory B (IB). IB degradation releases nuclear factor B
(NFB) from its cytoplasm stores, NFB then translocates into the nucleus acti-
vating a myriad of transcriptional programs including clotting elements, comple-
ment, acute phase proteins, cytokines, chemokines and nitric oxide synthase
genes. The outpouring of these inflammatory mediators is central to the patho-
genesis of septic shock induced by gram-negative bacteria [1, 2].
High levels of endotoxin in patients with severe sepsis correlated with the
presence of hypotension and is associated with worse prognosis [3, 4]. The
problems with endotoxin assays are that the measurable levels are affected by
endotoxin-binding proteins such as LBP, soluble CD14, and high- and low-
density lipoproteins. The levels of these endotoxin-binding proteins are highly
variable in septic patients. Moreover, the pathophysiological impact of a given
amount of circulating endotoxin is highly dependent upon the responsiveness of
host tissues to this microbial mediator. As an example, older patients tolerate
endotoxemia poorly when compared with younger patients [3].
Measurement of endotoxin in the blood has been performed by a techni-
cally difficult and time-intensive bioassay, the Limulus amebocyte assay.
Recently, a new rapid endotoxin activity assay has been developed that may
prove to be of greater clinical utility. This assay is based upon the degree of
LPS and IL-6 as Biomarkers 223
priming of the circulating neutrophil population by endotoxin exposure. In a
study of patients admitted to a mixed surgical-medical ICU, 58% of the patients
had elevated endotoxin levels by this assay. This percentage increased to 85% in
patients with severe sepsis [4]. This new endotoxin detection method correlated
with excess ICU mortality, septic shock and gram-negative bacteremia.
IL-6 as a Biomarker for Cytokine Networks
Inflammatory cytokines play a pivotal role in the pathogenesis of sepsis.
The major proinflammatory cytokines, tumor necrosis factor (TNF)-␣ and
interleukin (IL)-1, induce their hemodynamic and metabolic effects in concert
with an expanding group of host-derived inflammatory mediators that work in a
coordinated fashion to produce the systemic inflammatory response. The multi-
tude of inflammatory cytokines and chemokines found in excess quantities in
the bloodstream in patients with septic shock is impressive and is only matched
by an equally daunting group of anti-inflammatory mediators. The proinflam-
matory mediators tend to predominate locally and in the early phases of sepsis,
whereas the endogenous anti-inflammatory components often prevail systemi-
cally in the later phases of sepsis. Monocyte-macrophage-generated cytokines
and chemokines primarily drive the early septic process; whereas, the lympho-
cyte-derived cytokines and interferons become important in the later phases of
sepsis [5, 6].
Cytokine levels are notoriously variable in the blood and have proven rather
difficult to assess from routine blood samples. The most reliable and widely uti-
lized cytokine measure continues to be IL-6 [2–6]. This GP130 receptor ligand is
found and easily measured in the majority of septic patients but these levels are
capricious and vary by orders of magnitude in patients with similar clinical pre-
sentations. IL-6 has both anti-inflammatory and proinflammatory actions and is
not lethal by itself, as is TNF or IL-1, when injected into experimental animals.
IL-6 is viewed as an indicator cytokine during periods of excess cytokine syn-
thesis in acute severe illness. Its measurement has proven to be of significant but
limited clinical relevance as a biomarker for severe sepsis [5, 6].
Monocyte Deactivation and Immunodepression as
Marker for Sepsis
The sepsis-induced immune suppressive phenomenon is part of a gene-
ral, compensatory, host defense mechanism designed to limit the potentially inju-
rious impact of ongoing systemic inflammation. This occurs primarily at the
Opal 224
transcriptional level, with downregulation of genes encoding for proinflammatory
cytokines and other acute phase proteins. Stress hormones (i.e. catecholamines,
corticosteroids) and anti-inflammatory cytokines such as IL-10 are upregulated.
Excess CD4⫹ lymphocyte apoptosis and a shift to a TH2-type cytokine response
are common concomitants following an initial septic insult [7]. This relative
immune refractory state places the patient at increased risk of secondary bacterial
or fungal infection. Methods to detect this immunosuppressed state and restore
immune competence are under active clinical investigation. Patients with
depressed expression of MHC class II antigens (e.g., HLA-DR) on the cell sur-
face of macrophages may be in a functionally immunosuppressed state [8, 9]. The
level of expression of monocyte HLA-DR is measurable by a rapid bioassay to
assess immune function and prognosis in sepsis [8–10].
aPTT Biphasic Wave Form Analysis as a Marker of Sepsis
A remarkably specific abnormality in the optical transmission waveform
obtained during measurement of the activated partial thromboplastin time on
some photometric hemostasis autoanalyzers deserves mention as a potential bio-
marker for sepsis [11]. A biphasic waveform abnormality is related to a complex
of C-reactive protein and very low-density lipoprotein associated with the clinical
diagnosis of disseminated intravascular coagulation and severe sepsis (fig. 1).
0
20
40
60
80
100
5 10152025303540455055
Time (s)
Transmittance (%)
Biphasic
slope
Normal
monophasic
response
Normal
Sepsis
Severe sepsis
Fig. 1. Hypothetical appearance of optical transmission of coagulation events in the
activated partial thromboplastin time. The normal response is a monophasic, sudden drop in
percent transmittance when fibrin polymerization occurs (solid line). In septic plasma, com-
plexes of CRP and lipoproteins form immediately and result in a biphasic decrease in trans-
mittance (dotted lines). The slope of the initial decline in percent of transmittance is
correlated with disease severity in sepsis.
LPS and IL-6 as Biomarkers 225
Several studies have confirmed the association of a biphasic waveform and the
diagnosis of severe sepsis [11–13]. The diagnostic accuracy of the abnormal
waveform for severe sepsis appears to be comparable to procalcitonin with the
advantage of a higher negative predictive value. This simple and rapidly available
test may prove to be of added diagnostic value during the workup for common
coagulation abnormalities in sepsis. An early procoagulant state is almost uni-
formly generated in the initial stages of sepsis and simple measurement of coagu-
lation parameters have provided consistent prognostic value in many sepsis
studies [12].
Procalcitonin in Sepsis
Perhaps the most extensively studied biomarker for sepsis is procalcitonin,
the propeptide of calcitonin. In septic patients, procalcitonin is generated by
numerous extrathyroidal tissues and possesses many favorable attributes as a
diagnostic test for sepsis. It has a long half-life (approximately 24 h), and blood
levels increase from undetectable to over 100 ng/ml during the course of septic
shock. Procalcitonin levels do not become elevated as rapidly as the cytokine
IL-6 and more reliably distinguish between non-infectious versus bacterial
causes of inflammation than do cytokine measures in numerous studies [5, 6,
14, 15]. Recent studies indicate that procalcitonin levels could be utilized to
determine the appropriate use of antibacterial agents for bacterial respiratory
infections versus supportive care alone for viral infections in community-
acquired pneumonia [15]. Procalcitonin is approved for risk assessment for
patients with sepsis. Assay time is less than 20 min and results are available
in 1 h.
Soluble TREM-1 as a Potential Biomarker for Sepsis
Triggering receptor expressed on myeloid cells (TREM-1) is a member of
the immunoglobulin superfamily and is expressed on the cell surface of neu-
trophils and monocytes. TREM-1 is upregulated in the setting of bacterial and
fungal infection and, after binding with its as yet unknown ligand, acts syner-
gistically with LPS to induce cytokine production. A soluble form of TREM is
released during infection and at a cutoff of 60ng/ml in the blood, soluble TREM
has an excellent sensitivity and specificity in differentiating systemic inflam-
matory response syndrome from sepsis [16]. The ultimate diagnostic utility of
soluble TREM-1 measurement awaits further testing.
Opal 226
References
1 Calvano SE, Wenzhong X, Rihards DR, Felciano RM, Baker HV, Cho RJ, Chen RO, Brownstein
BH, Perren J, Tschoeke SK, Miller-Graziano C, Moldawer LL, Mindrinos MN, Davis RW,
Tompkins RG, Lowry SF; Inflammatory Host Responses to Injury Large Scale Collaborative
Research Program: A network-based analysis of systemic inflammation in humans. Nature
2005;437:1032–1037.
2 Opal SM: The clinical relevance of endotoxin in human sepsis: a critical analysis. J Endotoxin Res
2002;8:473–476.
3 Opal SM, Scannon PJ, Vincent J-L, White M, Carroll SF, Palardy JE, Parejo NA, Pribble JP,
Lemke JH: Relationship between plasma levels of lipopolysaccharide (LPS) and LPS-binding
protein in patients with severe sepsis and septic shock. J Infect Dis 1999;180:1584–1589.
4 Marshall JC, Foster D, Vincent JL Cook D, Cohen J, Dellinger P, Opal SM: Diagnostic and prog-
nostic implications of endotoxemia in critical illness: results of the Medic trial. J Infect Dis
2004;190:527–534.
5 Mokart D, Merlin M, Sannini A, Brun JP, Delpero JR, Houvenaeghel G, Moutardier V, Blache JL:
Procalcitonin, interleukin 6 and systemic inflammatory response syndrome (SIRS): early markers
of postoperative sepsis after major surgery. Br J Anaesth 2005;94:767–773.
6 Fraunberger P, Want Y, Holler E, Parhofer KG, Nagel D, Walli AK, Seidel D: Prognostic value of
interleukin 6, procalcitonin, and C-reactive protein levels in intensive care unit patients during
first increase of fever. Shock 2006;26:10–12.
7 Hotchkiss RS, Karl IE: The pathophysiology and treatment of sepsis. N Engl J Med 2003;348:
138–150.
8 Ng PC, Li G, Chui KM, Chu WC, Li K, Wong RP, Fok TF: Quantitative measurement of monocyte
HLA-DR expression in the identification of early onset neonatal infection. Biol Neonate
2006;89:75–81.
9 Le Tulzo Y, Pangault C, Amiot L, Guilloux V, Tribut O, Arvieux C, Camus C, Fauchet R, Thomas
R, Drénou B: Monocyte human leukocyte antigen-DR transcriptional downregulation by cortisol
during septic shock. Am J Respir Crit Care Med 2004;169:1144–1151.
10 Perry SE, Mostafa SM, Wenstone R, Shenkin A, McLaughlin PJ: Is low monocyte HLA-DR
expression helpful to predict out come in severe sepsis? Intensive Care Med 2003;29:1245–1252.
11 Toh CH, Samis J, Downey C, et al: Biphasic transmittance waveform in the APTT coagulation
assay is due to the formation of Ca(⫹⫹)-dependent complex of C-reactive protein and very-low
density lipoprotein and is a novel marker of impending disseminated intravascular coagulation.
Blood 2002;100:2522–2529.
12 Levi M, Opal SM: Coagulation abnormalities in critically ill patients. Crit Care 2006;10:222–229.
13 Chopin N, Floccard B, Sobas F, et al: Activated partial thromboplastin time waveform analysis: a
new tool to detect infection? Crit Care Med 2006;34:1654–1660.
14 Aikawa N, Fujishima S, Endo S, Sekine I, Kogawa K, Yamamoto Y, Kushimoto S, Yukioka H, Kato
N, Totsuka K, Kikuchi K, Ideda T, Ideda K, Harada K, Satomura S: Multicenter prospective study
of procalcitonin as an indicator of sepsis. J Infect Chemother 2005;11:152–159.
15 Christ-Crain M, Stolz D, Bingisser R, et al: Procalcitonin guidance of antibiotic therapy in com-
munity-acquired pneumonia: a randomized trial. Am J Respir Crit Care Med 2006;174:84–91.
16 Gibot S, Kolopp-Sarda MN, Benie MC, et al: Plasma level of a triggering receptor expressed on
myeloid cells-1: its diagnostic accuracy in patients with suspected sepsis. Ann Intern Med 2004;
141:9–15.
Steven M. Opal
Infectious Disease Division, Memorial Hospital of Rhone Island
111 Brewster Street
Pawtucket, RI 02860 (USA)
Tel. ⫹1 401 729 2545, Fax ⫹1 401 729 2795, E-Mail Steven_Opal@brown.edu
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 227–235
Quantifying Dynamic Kidney Processes
Utilizing Multi-Photon Microscopy
Bruce A. Molitoris, Ruben M. Sandoval
Department of Medicine, Division of Nephrology, Indiana University School of
Medicine, and Indiana Center for Biological Microscopy, Indianapolis, Ind., USA
Abstract
Multi-photon microscopy and advances in optics, computer sciences, and the available
labeling fluorophores now allow investigators to study the dynamic events within the func-
tioning kidney with subcellular resolution. This emerging technology, with improved spatial
and temporal resolution and sensitivity, enables investigators to follow complex hetero-
genous processes in organs such as the kidney. Repeated determinations within the same ani-
mal are possible minimizing their use and inter-animal variability. Furthermore, the ability to
obtain volumetric data (3D) makes quantitative 4D (time) analysis possible. Finally, use of up
to three fluorophores concurrently allows three different or interactive processes to be
observed simultaneously. Therefore, this approach compliments existing molecular, bio-
chemical, and pharmacologic techniques by advancing data analysis and interpretation to
subcellular levels for molecules without the requirement for fixation.
Copyright © 2007 S. Karger AG, Basel
New imaging technologies, such as multi-photon microscopy, have equipped
researchers with extremely powerful tools to uniquely address biologically
important questions that can only be accomplished in whole organ studies
[1–3]. In parallel with this, advances in fluorophores with increased quantum
yields and ease of labeling [4], molecular and transgenic approaches, and new
delivery techniques [5] have enabled the development of intravital studies that
can follow and quantify events with enhanced spatial and temporal resolution.
Furthermore, exponential developments in computer sciences, specifically with
applications to imaging, have removed many of the obstacles previously limit-
ing the ability to utilize microscopy to study and quantify dynamic cellular
processes. These imaging technologies enable the measurement of dynamic
4-dimensional (3D plus time) structure and function in organs and tissues [6],
Molitoris/Sandoval 228
the measurement of chemical and biochemical composition of tissues, the
expression of fluorescently labeled molecular agents including drugs and pro-
teins, quantification of the rates of physiological processes such as microvascu-
lar perfusion rates, glomerular permeability and the mechanism of cellular
uptake and intracellular trafficking [1, 2, 7].
The kidney is an extremely complex heterogeneous organ consisting of
vascular and epithelial components functioning in a highly coordinated fashion
that enables the regulation of a myriad of interdependent processes. Over the
years talented and creative individuals have developed novel experimental
approaches that enable the isolation, understanding and integrating the unique
structure–function relationships that occur. Investigators have developed model
systems to enable enhanced manipulation and isolation of specific variables of
interest. This has lead to a mechanistic understanding of cellular processes
and the identification of alterations in these processes under defined conditions
that attempt to mimic either physiologic or disease states. However these
models often lack the organ-specific complexity, the dynamic nature of cellular
processes and cell–cell interactions necessary for adequate understanding of the
process under study. For example, proximal tubule cells in cell culture undergo
dedifferentiation resulting in reduced metabolic rates and alterations in many
cellular processes. This has limited our ability to test therapeutic approaches
and has resulted in difficulties translating preclinical data into therapeutic
advances.
Multi-photon microscopy offers the investigator a minimally invasive high
resolution technique, with increased depth of penetration and markedly reduced
phototoxicity, for visualization of cell–cell and intracellular events intravitally.
The reduction in phototoxicity with multi-photon microscopy results as fluores-
cence excitation occurs only at the focal point, thereby eliminating out-of-focus
fluorescent excitation within the tissue as would occur with confocal microscopy.
The genesis of these advances was covered in a previous article [3] and in our
previous publications [1, 2, 8]. Second, improved detectors with increased sen-
sitivity, enhanced software and faster hardware, and new computational algo-
rithms for 3D analysis and quantification have enabled more rapid, sensitive
and accurate data gathering, visualization and interpretation [9]. Finally, the
revolution in fluorophores capable of reporting on a growing number of cellular
processes has markedly improved the capabilities available to the investigator.
This is especially true for the biotech industry where small proteins and
oligonucleotides can be labeled without affecting the pharmacokinetics or phar-
macologic effects of the agent. In addition, fluorescent labeling of proteins,
either by genetic or chemical means of attachment using a wide spectrum of col-
ors, makes simultaneous multi-colored imaging of different cellular processes
possible [1–7].
In vivo Renal Imaging via Multi-Photon Microscopy 229
Applications
Table 1 lists the types of data that can be obtained using multi-photon
microscopy of the kidney. Both dynamic structural and functional observations
are possible. Therefore, one can observe and correlate cell–cell interactions,
structural changes, glomerular filtration, permeability, reabsorption, cellular
metabolism, microvascular flow and the functional effects of a substance being
administered. Since tissue fixation is not necessary these structural-functional
studies can be undertaken with a large number of small as well as large mole-
cules that cannot be fixed in tissues. In figure 1a we show the use of multiple
fluorescent molecules to follow several processes simultaneously. Glomerular
capillary blood flow is easily seen and can be quantified in the capillary net-
work of Munich Wistar rat surface glomeruli. At the same time filtration of a
small molecular weight (MW) Texas Red dextran is seen in Bowman’s space.
We have used this approach to quantify glomerular permeability, show that
Table 1. Investigational uses for multi-photon microscopy
Glomerular
•
Size/volume
•
Permeability/filtration
•
Fibrosis/sclerosis
Microvasculature
•
Blood flow rate
•
Endothelial permeability
•
WBC adherence/rolling
•
Vasoconstriction
Cellular uptake
•
Cell type specific uptake
•
Site – apical vs. basolateral membrane
•
Mechanism – endocytosis vs. carrier/transporter mediated
Cellular trafficking
•
Intracellular organelle distribution
•
Cytosol localization
Cellular metabolism
•
Fluorescence decay over time
Cell toxicity
•
Cell injury in necrosis, apoptosis
•
Surface membrane/blebbing
•
Mitochondrial function
Molitoris/Sandoval 230
S1
Glm
DT
BS
S1
a
b
c
d
Fig. 1. Physiologic and morphologic parameters in the superficial rat kidney identified
by intravital 2-photon microscopy. a Glomerular filtration and movement along a nephron
segment. A large, non-filtering 500,000 MW fluorescein dextran (green) is retained within the
capillary loops in the glomerulus (center) and the microvasculature (arrowheads). Within both
of these structures circulating red blood cells (RBCs) appear as black oblong streaks because
of exclusion of the large MW dextran. A small, freely filtered 3,000 MW Texas Red dextran
(red) is seen filtering into Bowman’s Space (BS) around the glomerulus and down into the S1
segment. A variation in nuclear morphology between cells types, labeled with Hoechst 33342
(cyan), can be seen between proximal tubule cells (S1), distal tubule cells (DT), and podocytes
around the glomerulus (center). b Endocytic uptake by proximal tubule cells. A small, freely
filtered 10,000 MW Cascade Blue dextran (blue) was given 24 h prior to imaging. The bulk of
In vivo Renal Imaging via Multi-Photon Microscopy 231
molecular charge is not a determinant of filtration and the dissociation of
protein-bound molecules prior to filtration [2, 10]. In figure1b we show the
cell-specific uptake of a Cascade Blue-labeled dextran probe within the kidney.
This agent is not bound to proteins, is freely filtered and rapidly taken up by
proximal tubule cells across their apical membrane via endocytosis. No other
cell type within the kidney either bound or internalized the dextran as shown by
the lack of fluorescence in endothelial cells or in distal tubule cells. Within sec-
onds of intravenous injection there was filtration and rapid binding to the apical
membrane of proximal tubule cells. With increased time there was enhanced
cellular accumulation, especially in lysosomal structures. Using total integrated
fluorescence it is possible to quantify the extent of cellular uptake and to even
partition it into apical binding and cellular accumulation [1, 2, 7]. Additional
studies with folic acid (FA)-FITC revealed rapid loss of intracellular fluores-
cence following cellular uptake. These data did not indicate that FA was rapidly
catabolized, but rather that FA was taken up into acidic compartments resulting
in FITC quenching and loss of its fluorescent signal [11]. In fact, using the
R-FA probe we documented transcytosis as a mechanism for reclamation of the
reabsorbed FA [11]. Finally, recent data indicate that serum albumin undergoes
glomerular filtration at a much greater level than previously believed [12].
Thereafter, it is rapidly reabsorbed by PTC and a fraction undergoes transcyto-
sis, a process of reclaiming without intracellular catabolism.
the dextran can be seen localized within lysosomes (large punctate structures) at the basal por-
tion of proximal tubule cells. The leading S1 proximal tubule segment with the open connec-
tion to Bowman’s space is seen here adjacent to an unlabeled glomerulus (Glm). The arrow
indicates the direction of flow. c Microvascular injury following exposure to endotoxin. With
the same dyes used in (a), alterations in microvascular dynamics, primarily RBC flow are
seen. The oblong streaks seen in (a) become more defined as RBCs due to the reduced flow
and the presence of obstruction causing Rouleaux formations (arrows), and white cells adher-
ing to the microvascular walls (arrowheads). The heterogeneity of this alteration is seen by the
color profile of the dextrans in the blood. On the left half of the image, the blood, although
slow, has circulated sufficiently that the small 3,000 MW dextran (red) has been cleared and
only the large non-filtering 500,000 MW dextran (green) remains in the plasma. On the right
half of the image, stagnation of flow in those vessels has prevented that pool of blood from
reaching a glomerulus to filter out the 3,000 MW dextran (red). As a result, the color profile
is yellow, a combination of the large (green) and small (red) dextrans occupying the same
space in the plasma. Also visible is the vasoconstriction of the vessels to some degree, and
small MW dextran leaking into the interstitial space around the microvasculature. d A 3D ren-
dering of a surface glomerulus. Using VOXX software (developed at the Indiana Center for
Biological Microscopy), focal planes taken at 1-m intervals were rendered to produce a solid
appearing composite. The complex inter-weaving of the capillary loops is readily seen along
with the surrounding microvasculature. Bar ⫽ 20 m.
Molitoris/Sandoval 232
Figure 1c shows the renal cortical microvasculature following endotoxin
injection. Again, quantitation of erythrocyte flow, permeability alterations, and
white blood cell (WBC) rolling, adherence and infiltration are possible. WBC
within the microvasculature can be identified using Hoechst 33342 as a nuclear
marker. This is also true for all nucleated cells within the tissue. Since the stain-
ing intensity is different for PTC and distal tubule nuclei we can use this to
identify tubular segments. Rouleau formation can be noted. This is a common
occurrence in several types of acute kidney injury including ischemia, sepsis,
lipopolysaccharide and radiocontrast (unpublished observations, B.A.M.).
Figure 1d shows a 3D volume reconstruction of multiple z-axis sections of
a cortical section containing surface glomeruli. This enables quantitative analy-
sis of structural changes over time as multiple volumes can be collected in a
time series. Finally, using powerful software programs like Amira one can seg-
ment out and quantify individual cells or areas of interest within cells [13, 14].
Numerous other approaches to understanding protein function and gene
regulation have been developed. This is a rapidly growing field that will con-
tinue to make use of molecular and transgenic advances to selectively label
individual proteins. These approaches will enable the study of specific proteins,
compartments and processes in a dynamic fashion.
The use of fluorescent rationing to study events within the kidney was
recently advanced by Yu et al. [10]. Their studies have outlined specific ways to
evaluate glomerular permeability, glomerular sieving coefficients, and tubular
reabsorption of different compounds as affected by size and charge selectivity
using a generalized polarity concept. The use of this ratiometric concept has
several advantages including minimizing errors secondary to the effect of
intensity attenuation by tissue depth or fluctuations in excitation intensity or
detector sensitivity on the overall quantitative process. As variations in fluores-
cence intensity can be a major problem in quantification, the use of ratiometric
techniques is of great importance.
Once the fluorescent compound is within a cell, it then becomes possible
to quantify its intracellular distribution and metabolism. Furthermore, it is quite
possible to follow the intracellular accumulation and subcellular distribution
over time in the same animal, and to undertake repeated observations in that
animal at varying intervals over days to weeks. Furthermore, analysis of volu-
metric data, obtained by collecting images along the z axis, with quantitative
software such as VOXX [9] or Amira [13] can yield additional information
regarding cellular uptake and intracellular distribution. These studies can be
particularly helpful in the pharmacokinetic understanding of drug delivery and
metabolism at the individual cell level. However, one must remember that the
fluorescence half life and biologic half life may vary and that specific studies
are required to relate these to important parameters.
In vivo Renal Imaging via Multi-Photon Microscopy 233
Specific intracellular organelles can be studied utilizing fluorescent dyes
that have been developed to selectively label these organelles [1]. For example
rhodamine-123 is utilized to label the mitochondria of tubular epithelial cells.
As the fluorescence of this compound is directly related to the potential differ-
ence across the mitochondrial membrane, one can then develop quantitative
assays for mitochondrial function for both acute and chronic studies. It is also
possible to selectively label the mitochondria of endothelial cells and circulat-
ing WBCs utilizing rhodamine B hexyl ester which stays within the microvas-
cular compartment. Again, quantitative assays can be developed to look at the
individual number and fluorescence potential of mitochondria. This is an area
where development of an internal standard for ratiometric imaging would be
advantageous.
It is also possible to use the DNA fluorescent marker Hoechst to specifically
evaluate intranuclear uptake of other fluorescent compounds and to identify spe-
cific cell types based upon their nuclear morphology. Identifying apoptosis in
vivo, utilizing standard nuclear condensation criteria, can also be done following
an acute injection of Hoechst 33342 [15]. As is shown in figure 1, different cellu-
lar nuclei have different morphologies allowing one to identify podocyte nuclei,
distal tubule nuclei, proximal tubule nuclei and endothelial nuclei.
Challenges and Future Opportunities
Many challenges remain to maximize the ability to study and quantify
cell–cell and subcellular processes at the cellular level within the kidney using
multi-photon microscopy. Two major areas include image acquisition rates and
the depth of penetration within the kidney. The studies presented in this review
were recorded at approximately one frame per second. One can increase the
acquisition rate by limiting the area of study, but this often sacrifices other
important data. The depth of imaging possible, although 4–5 times greater than
confocal imaging, remains limited to less than 200 m for the kidney. Thus, we
are unable to visualize the cortical-medullary area from the surface of the
kidney. Perhaps new external detectors, access to longer wave length light
sources, and lenses specifically designed for multi-photon microscopes will
allow enhanced depth of penetration. Phototoxicity does remain a potential
problem at the focal point of excitation and this must always be considered dur-
ing study design. Quantifying the recorded results is also a major area under
development. Continuing improvement in software and hardware has a goal of
automation of data collection, segmentation and analysis. Finally, cost remains
an obstacle for the individual PI, with core imaging facilities and expertise gen-
erally required [1].
Molitoris/Sandoval 234
In summary, recent developments in intravital multi-photon studies within
the kidney now allow investigators to utilize unique techniques and fluorescent
probes to visualize the functioning kidney and characterize cellular and subcel-
lular events in a dynamic fashion. This approach will lead to enhanced under-
standing of renal physiology and the pathophysiology of disease processes and
their therapy. This will result in more efficient and effective translation of pre-
clinical data into therapeutic advances.
Acknowledgements
This work was made possible by National Institutes of Health grants P50 DK-61594,
PO1 DK-53465 and RO1 DK-069408, a Veterans Affairs Merit Review (to B.A.M.), and an
INGEN (Indiana Genomics Initiative) grant from the Lilly Foundation to the Indiana
University School of Medicine.
References
1 Molitoris BA, Sandoval RM: Intravital multiphoton microscopy of dynamic renal processes. Am J
Physiol Renal Physiol 2005;288:F1084–F1089.
2 Molitoris BA, Sandoval RM: Pharmacophotonics: utilizing multi-photon microscopy to quantify
drug delivery and intracellular trafficking in the kidney. Adv Drug Deliv Rev 2006;58:809–823.
3 Zipfel WR, Williams RM, Webb WW: Nonlinear magic: multiphoton microscopy in the bio-
sciences. Nat Biotechnol 2003;21:1369–1377.
4 Miyawaki A, Sawano A, Kogure T: Lighting up cells: labelling proteins with fluorophores. Nat
Cell Biol 2003;suppl:S1–S7.
5 Ashworth SL, Tanner GA: Fluorescent labeling of renal cells in vivo. Nephron Physiol 2006;103:
p91–p96.
6 Gerlich D, Ellenberg J: 4D imaging to assay complex dynamics in live specimens. Nat Cell Biol
2003;suppl:S14–S19.
7 Peti-Peterdi J: Multiphoton imaging of renal tissues in vitro. Am J Physiol Renal Physiol 2005;288:
F1079–F1083.
8 Dunn KW, Sandoval RM, Kelly KJ, Dagher PC, Tanner GA, Atkinson SJ, Bacallao RL, Molitoris
BA: Functional studies of the kidney of living animals using multicolor two-photon microscopy.
Am J Physiol Cell Physiol 2002;283:C905–C916.
9 Clendenon JL, Phillips CL, Sandoval RM, Fang S, Dunn KW: Voxx: a PC-based, near real-time
volume rendering system for biological microscopy. Am J Physiol Cell Physiol 2002;282:
C213–C218.
10 Yu W, Sandoval RM, Molitoris BA: Quantitative intravital microscopy using a Generalized
Polarity concept for kidney studies. Am J Physiol Cell Physiol 2005;289:C1197–C1208.
11 Sandoval RM, Kennedy MD, Low PS, Molitoris BA: Uptake and trafficking of fluorescent conju-
gates of folic acid in intact kidney determined using intravital two-photon microscopy. Am J
Physiol Cell Physiol 2004;287:C517–C526.
12 Russo LM, Sandoval RM, McKee M, Osicka TM, Collins AB, Brown D, Molitoris BA, Comper
WD: The normal kidney filters nephrotic levels of albumin retrieved by proximal tubule cells:
Retrieval is disrupted in nephrotic states. Kidney Int 2007; Epub ahead of print.
13 Clendenon JL, Byars JM, Hyink DP: Image processing software for 3D light microscopy.
Nephron Exp Nephrol 2006;103:e50–e54.
In vivo Renal Imaging via Multi-Photon Microscopy 235
14 Phillips CL, Gattone VH 2nd, Bonsib SM: Imaging glomeruli in renal biopsy specimens. Nephron
Physiol 2006;103:p75–p81.
15 Kelly KJ, Sandoval RM, Dunn KW, Molitoris BA, Dagher PC: A novel method to determine speci-
ficity and sensitivity of the TUNEL reaction in the quantitation of apoptosis. Am J Physiol Cell
Physiol 2003;284:C1309–C1318.
Bruce A. Molitoris, MD
Indiana Center for Biological Microscopy, Indiana University School of Medicine
950 W. Walnut St., R2–202C
Indianapolis, IN 46202 (USA)
Tel. ⫹1 317 274 5287, Fax ⫹1 317 274 8575, E-Mail bmolitor@iupui.edu
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 236–249
Diuretics in the Management of Acute
Kidney Injury: A Multinational Survey
Sean M. Bagshaw
a,g
, Anthony Delaney
b,c
, Daryl Jones
d
, Claudio Ronco
f
,
Rinaldo Bellomo
a,e
a
Department of Intensive Care, Austin Hospital, Melbourne, Vic.,
b
Department of
Intensive Care, Royal North Shore Hospital, Sydney,
c
Northern Clinical School,
University of Sydney, St. Leonards,
d
Department of Epidemiology and Preventive
Medicine, Monash University, Melbourne, Vic.,
e
Department of Medicine, Melbourne
University, Melbourne, Vic., Australia, and
f
Department of Nephrology, St. Bortolo
Hospital, Vicenza, Italy,
g
Division of Critical Care Medicine, University of Alberta
Hospital, University of Alberta, Edmonton, Alta., Canada
Abstract
Background: Diuretics are a common intervention in critically ill patients with acute
kidney injury (AKI). However, there is no information that describes the practice patterns of
diuretic use by clinicians. Methods: Multinational, multicenter survey of intensive care and
nephrology clinicians that utilized an 18-question self-reported questionnaire. Results: The
survey generated 331 responses from 16 countries. Academic institutions comprised 77.5%,
with the remaining being from metropolitan, regional or private hospitals. The use of
furosemide was most common (67.1%), delivered primarily intravenously (71.9%) and by
bolus dosing (43.3%). Other diuretics were infrequently used. The majority rated current
serum creatinine (73.6%) and urine output (73.4%), blood pressure (59.7%), central venous
pressure (65.2%) and risk of toxicity (62.4%) important when deciding on a dose. Pulmonary
edema was a prime physiologic indication for diuretic use (86.3%). Diuretic use was also
common with rhabdomyolysis (55.6%), major surgery (56%), and cardiogenic shock (56.2%),
and sepsis (49.5%). Diuretic use was most commonly given either prior to (57.7%) or during
recovery (33.9%) after renal replacement therapy (RRT). Most (76.6%) targeted a diuresis of
ⱖ0.5–1.0 ml/kg/h. The majority did not believe that diuretics could reduce mortality (74.3%),
reduce need for (50.8%) or duration of RRT (57.8%) or improve renal recovery (68.2%), how-
ever, many stated uncertainty. Most (85.1%) would be willing to participate in a randomized
trial (RCT) of diuretics in AKI with 72.4% believing it ethically acceptable to allocate patients
to placebo. Conclusion: Diuretics are frequently used in AKI. Clinicians are most familiar
with furosemide given intravenously and titrated to a physiologic endpoint of urine output.
Most clinicians believe an RCT on diuretic use in AKI is justified and ethical. This survey
confirms clinical agreement and a need for higher quality evidence on diuretic use in AKI.
Copyright © 2007 S. Karger AG, Basel
A Multinational Survey of Diuretic Use in Acute Kidney Injury 237
Acute kidney injury (AKI) is a common complication in critical illness
which portends a poor prognosis [1, 2]. Few interventions have been proven to
impact the clinical course and outcome once AKI is established [3–5].
Diuretics, in particular loop diuretics, are a common intervention used in
the management of critically ill patients with AKI [6, 7]. Loop diuretics act at the
medullary thick ascending loop of Henle to inhibit the Na
⫹
/K
⫹
/Cl
–
pump on the
luminal cell membrane surface and reduce oxygen demand [8, 9]. Theoretically,
the timely use of loop diuretics might attenuate the severity of AKI. Diuretics
may further play a vital role in managing extravascular volume overload by aug-
menting urine output and aid in acid-base and potassium homeostasis.
Small clinical studies have suggested that diuretics might diminish the
severity of injury by converting ‘oliguric’ to ‘non-oliguric’ AKI, shorten the
duration of AKI, improve the rate of renal recovery, and perhaps delay or ame-
liorate the need for renal replacement therapy (RRT) [10–15]. However,
improvements in survival or renal recovery have yet to be shown by high qual-
ity research evidence [6, 16]. Accordingly, there is controversy regarding
whether diuretics can impact clinical outcomes in AKI [17–20]. Thus, while
there appears to be a biological rationale for their use, there is a limited under-
standing of how and when diuretics are used. Similarly, there are no data describ-
ing the beliefs, attitudes and practice patterns (i.e. indications, dosing regimens
and end points for efficacy) used by clinicians routinely involved in caring for
critically ill patients with AKI. Lastly, there is uncertainty on whether clinicians
have a genuine need for a randomized controlled trial (RCT) assessing diuretics
in AKI.
Therefore, as part of a larger initiative to understand the potential thera-
peutic role of diuretics in AKI, we have conducted a multinational multicenter
survey of intensive care specialists, nephrologists and other clinicians who rou-
tinely manage critically ill patients with AKI in order to gain insight into the
current patterns of practice and potentially aid in the future design of a RCT.
Methods
Target Population
The target population for this survey were intensivists, nephrologists and other clini-
cians who routinely provide care for critically ill patients with AKI.
Survey Objectives
This survey was undertaken first to explore the beliefs, attitudes and practice patterns
of clinicians regarding the use of diuretics in the management of AKI. Specifically, the sur-
vey was designed to investigate several aspects about the use of diuretics in AKI including:
Bagshaw/Delaney/Jones/Ronco/Bellomo 238
(1) frequency; (2) timing in the course of AKI; (3) indications; (4) class of diuretic (i.e. loop,
proximal, potassium sparing, osmotic); (5) methods of administration; (6) dosing regimens;
(7) what endpoints are used by clinicians to determine a response, and (8) whether clinicians
believe the use of diuretics in AKI can impact outcome. In addition, the survey inquired on:
(1) whether clinicians need a future RCT on diuretic use in AKI; (2) the clinical/physiologic
characteristics on patients they would enroll, and (3) whether they had interest in participating
in such a trial.
Questionnaire Development
The initial questionnaire included 18 questions. The initial 3 questions were focused on
obtaining basic demographic information from the respondents (i.e. specialty, years of expe-
rience, practice type). The next 14 questions were designed to be uni-dimensional and
included closed format questions that used either a 5-point Likert-type agreement scale or a
dichotomous response (i.e. yes/no). Finally, the last question was open-ended and provided
space for additional comments.
Pilot Testing of the Survey
The questionnaire was initially piloted using a focus group, comprised of clinicians
with different levels of experience, to evaluate for clarity, comprehension and interpretation.
The focus group was moderated by the principal investigators (S.M.B., D.J. and R.B.). All
participants of the focus group contributed and provided feedback. During this process, the
wording of several questions was modified. The survey was then piloted in a small sample of
intensivists from two centers for additional feedback.
Survey Administration
The survey was intended to capture responses across a broad range of practices, in mul-
tiple centers and in multiple countries. The survey was sent to member centers of the
Australia and New Zealand Intensive Care Society Clinical Trials Group, to member centers
of a large multinational consortium of investigators, the Beginning and Ending Supportive
Therapy for kidney investigators with an interest in AKI, to members of the Australia and
New Zealand Society of Nephrology, and several member centers of the Canadian Critical
Care Trials Group. Contacts at each center were then requested to circulate the survey to all
members of their respective departments. A second reminder notice was sent to the same
groups and centers at 4 weeks.
Clinicians were provided with three methods to respond to the survey. The first was
electronic completion of the survey and return to the principle investigator (S.M.B.) as an
e-mail attachment. The second was paper completion and return by either mail or fax. The
final method was completion of an online version of the survey posted on Survey Monkey.
com (available at the website: http://www.surveymonkey.com). Clinicians were provided
with a direct link via e-mail to the survey and were able to complete the survey online. A total
of 169 responses (51%) were completed online. The survey was conducted from September
to November 2006.
Data Management and Analysis
Questionnaires returned by e-mail/fax/letter were manually entered into a master Excel
spreadsheet (Microsoft Corp, Richmond, USA). Questionnaires completed on SurveyMonkey.com
were exported in another Excel spreadsheet, reconfigured, and merged with the master Excel
A Multinational Survey of Diuretic Use in Acute Kidney Injury 239
database. No assumptions were made about missing data and missing fields were not replaced.
Categorical data are presented as proportions and compared using Fisher’s exact test. Continuous
data are presented as means (range) and compared using a Student’s t test. Analysis was performed
using STATA 8.2 (Stata Corp, College Station, USA).
Ethical Approval
This survey was approved by the Human Research Ethics Committee at the Austin
Hospital prior to commencement.
Results
Characteristics of the Sampling Frame
The survey generated 331 responses. Of these, 14 (4.2%) were incomplete
and provided no usable data. All these responses were generated from the online
version of the survey. These 14 were excluded from the analysis.
The surveys were generated from 63 cities in 16 countries (table 1). Australia,
New Zealand, Canada and the United States comprised 79.4% of responses. Most
were from academic or tertiary institutions (77.5%) with smaller proportions
from metropolitan, regional, rural, or private hospitals (fig. 1).
Table 1. Distribution of countries and cities surveyed
Country Cities Responses
Australia 21 155 (48.9%)
Canada 7 46 (14.5%)
United States 3 32 (10.1%)
New Zealand 5 18 (5.7%)
China 7 14 (4.4%)
Japan 2 13 (4.1%)
Netherlands 3 11 (3.5%)
Belgium 3 11 (3.5%)
Italy 5 8 (2.5%)
Germany 1 2 (0.6%)
Brazil 1 1 (0.3%)
Czech Republic 1 1 (0.3%)
Russia 1 1 (0.3%)
Sweden 1 1 (0.3%)
United Kingdom 1 1 (0.3%)
Uruguay 1 1 (0.3%)
Unknown ⫺ 1 (0.3%)
Total 16 63 317 (100%)
Bagshaw/Delaney/Jones/Ronco/Bellomo 240
Demographics
Intensivists or those with combined ICU training represented 68.9%,
whereas nephrology comprised 29.5% (fig. 2). All had generally been in prac-
tice for a median (intraquartile range) of 10 (5–18) years. The vast majority
(94.9%) practiced adult medicine.
Details of How Diuretics Are Administered
Loop diuretics, specifically furosemide, was the most commonly used
diuretic, with 67.1% using it ‘frequently’ or ‘almost always’. Additional loop
ICU
Nephrology
ICU combined
Anesthesia
Internal medicine
63%
29%
4%
2%
2%
Fig. 1. Summary of center types sampled for the survey.
78%
12%
7%
3%
Academic/tertiary
Metropolitan
Regional/rural
Private
Fig. 2. Summary of clinical specialties sampled for the survey.
A Multinational Survey of Diuretic Use in Acute Kidney Injury 241
diuretics such as torsemide, ethacrynic acid, and bumetanide were rarely used.
Use of other classes/types of diuretics was also much less common. The major-
ity responded either ‘almost never’ or ‘infrequently’ to use of hydrochlorozide
(79.5%), spirolactone (79.7%), metolazone (81.1%), acetazolamide (88.8%),
and mannitol (84.7%).
The majority of respondents deemed several factors as ‘important’ or ‘very
important’ when determining the dose of diuretic to administer (table 2). These
factors included current serum creatinine (73.6%), current urine output (73.4%),
blood pressure (59.7%), central venous pressure (65.2%), and risk of toxicity
(62.4%). Only a few respondents commented that patient age, baseline kidney func-
tion, and serum albumin also influenced the dose of diuretic to be administered.
Diuretics are primarily administered by the intravenous (IV) route (71.9%
responded ‘frequently’ or ‘almost always’). Diuretics are rarely given by the
oral route in the setting of AKI (75.7% reporting ‘almost never’ or ‘infre-
quently’). While both IV bolus and infusion are common, IV bolus appears to
be used more frequently than an IV infusion.
A protocol to guide diuretic therapy was reported by only 5.3% (n ⫽ 16) of
the respondents.
Indications and Timing of Administration
Pulmonary edema was the only reported physiologic indication where the
vast majority (86.3%) responded to use of diuretics either ‘frequently’ or
Table 2. Summary of responses pertaining to factors influencing the dose of diuretic administered
Factor Very Unimportant Uncertain Important Very
unimportant, % % % % important, %
Patient weight (n ⫽ 311) 11.3 32.2 17 32.8 6.8
Serum creatinine (n ⫽ 314) 6.4 13.4 6.7 54.8 18.8
Urine output (n ⫽ 302) 3.2 14.6 8.9 55.6 17.8
Toxicity (n ⫽ 314) 4.1 20.1 13.4 49.7 12.7
Cardiac output (n ⫽ 315) 7.6 25.4 20.3 39.4 7.3
Blood pressure (n ⫽ 315) 6.7 21 12.7 46.7 13
CVP (n ⫽ 313) 5.8 16.6 12.5 47.9 17.3
PAOP (n ⫽ 313) 11.2 24.3 22.7 32 9.9
MVO
2
(n ⫽ 314) 11.5 31.9 32.2 21 3.5
PaO
2
/FiO
2
ratio (n ⫽ 313) 8 29.1 30.4 29.4 3.2
MV (n ⫽ 312) 10.9 38.8 31.1 18.3 1
CVP ⫽ Central venous pressure; MV ⫽ mechanical ventilation; MVO
2
⫽ mixed venous oxygen saturation;
PAO P ⫽ pulmonary artery occlusion pressure.
Bagshaw/Delaney/Jones/Ronco/Bellomo 242
‘almost always’ (table 3). An increasing serum creatinine, oliguria where the
serum creatinine is yet to be determined, and metabolic acidosis were seldom
indications for diuretic use for most respondents. For both oliguria where the
serum creatinine is increasing and hyperkalemia, diuretic use appears common,
with most largely reporting such use either ‘sometimes’ or ‘frequently’.
Additionally, a few respondents also commented on the use of diuretics for
metabolic alkalosis (post-hypercapnic), hypercalcemia, hypertension due to
hypervolemia, as well as a single challenge to assess diuretic responsiveness.
The majority responded either ‘infrequently’ or ‘almost never’ to adminis-
tering diuretics for AKI associated with hemorrhagic shock (80%), contrast-
induced nephropathy (62.2%), and abdominal compartment syndrome (65.5%;
Table 3. Summary of the responses pertaining to physiologic indications for use of diuretics in the
management of ARF
Factor Almost Infrequently Sometimes Frequently Almost
never, % % % % always, %
Increasing SCr (n ⫽ 313) 29.4 21.1 31.6 14.4 3.5
Oliguria when SCr not known (n ⫽ 314) 29.9 22 22.9 19.1 6.1
Oliguria when SCr is increasing (n ⫽ 313) 16 19.8 32.3 21.4 10.5
Pulmonary edema (n ⫽ 314) 0.32 1 12.4 38.9 47.5
Metabolic acidosis (n ⫽ 314) 34.7 31.2 29.3 4.1 0.64
Hyperkalemia (n ⫽ 315) 7 16.5 38.1 25.1 13.3
SCr ⫽ Serum creatinine.
Table 4. Summary of the responses pertaining to the clinical scenarios for use of diuretics in the man-
agement of ARF
Factor Almost Infrequently Sometimes Frequently Almost
never, % % % % always, %
CIN (n ⫽ 315) 36.2 26 22.2 13 8
Rhabdomyolysis (n ⫽ 313) 22 15.7 28.8 26.8 6.7
ACS (n ⫽ 313) 39 26.5 22.4 8.6 3.5
Postoperative (n ⫽ 314) 18.5 21 28.3 27.7 4.5
Sepsis (n ⫽ 315) 26.4 23.2 29.2 16.5 4.8
Cardiogenic shock (n ⫽ 315) 19.1 16.8 28.3 27.9 7.9
Hemorrhagic shock (n ⫽ 314) 56.7 23.3 13.1 5.4 1.6
ACS ⫽ Abdominal compartment syndrome; CIN ⫽ contrast-induced nephropathy.
A Multinational Survey of Diuretic Use in Acute Kidney Injury 243
table 4). On the contrary, their use was more common in AKI associated with
rhabdomyolysis, major surgery, and cardiogenic shock with respondents report-
ing ‘sometimes’ or ‘frequently’ in 55.6, 56, and 56.2%, respectively. In septic
AKI, 49.6% responded either ‘infrequently’ or ‘almost never’, however 45.7%
reported using diuretics ‘sometimes’ or ‘frequently’. Other reported clinical
indications also included hepatorenal syndrome and post-prostatectomy for clot
retention.
The majority administer diuretics most commonly prior to initiation of
RRT (57.7% responded ‘frequently’ or ‘almost always’) whereas only 8%
responded similarly for use during RRT. Diuretics are also commonly used dur-
ing the recovery phase with 33.9% having responded ‘frequently’ or ‘almost
always’ (fig. 3).
Assessment of Clinical Response
A urine output of ⱖ0.5 or ⱖ1.0 ml/kg/h in response to diuretics was tar-
geted by 76.6% of respondents. Only 11.5% set a target urine output of
ⱖ1.5–2.0 ml/kg/h. A minority (8.3%) reported not using urine output to assess
the response to diuretics in AKI. An additional 13.7% reported a variety of
other targets (i.e. 80–100 ml/h, ⬎1 l over 2 h, ⬎1–1.5 l/day, any improvement in
urine output and a sufficient increase to maintain a target fluid balance) and
numerous other factors that contribute to their determination of a response to
diuretics in AKI (i.e. volume status, serum electrolytes, pulmonary edema).
0
10
20
30
40
50
60
%
Prior to
RRT
During
RRT
During
recovery
Almost never
Infrequently
Sometimes
Frequently
Almost always
Fig. 3. Summary of responses pertaining to the timing of diuretic use in the manage-
ment of ARF.
Bagshaw/Delaney/Jones/Ronco/Bellomo 244
While 17.8% responded that they do not specifically target a given fluid
balance when using diuretics in AKI, 16.2% reported that they aim for a neutral
balance and 35.9% targeted a negative daily balance in the range of 0.5–1 l.
Another 30.2% reported that fluid balance goals were dependent on additional
factors (i.e. present fluid balance status, presence of pulmonary or peripheral
edema, non-renal organ dysfunction, presence of oliguria, and phase of AKI).
The majority of respondents (65.8%) do not routinely use the change in serum
creatinine or urea to assess whether there has been a response to diuretics.
A few respondents used serum creatinine as a marker for toxicity or commented
that a rise in serum creatinine coupled with a negative fluid balance may indi-
cate either over-diuresis or relative hypovolemia.
Clinical Outcomes
The majority of respondents did not believe (either ‘strongly disagree’ or
‘disagree’) that the use of diuretics in AKI could reduce mortality (74.3%) or
improve renal recovery (68.2%; fig. 4). However, 23.2 and 27.3% still reported
uncertainty about whether diuretics could impact these outcomes. Similarly,
most reported disagreement that diuretics could reduce the need for RRT
(50.8%) or the duration of RRT (57.8%), however again, a significant propor-
tion (26 and 26.7%) reported uncertainty.
In total, 72.4% of respondents would consider it ethical to give patients a
placebo in an RCT of diuretics in AKI, whereas 15.7% believed it would be
0
10
20
30
40
50
%
Reduce
need for
RRT
Reduce
duration
RRT
Reduce
mortality
Improve
recovery
Increase
CrCl
Strongly disgree
Disagree
Uncertain
Agree
Strongly agree
Fig. 4. Summary of responses pertaining to the beliefs about outcomes with the use of
diuretics in the management of ARF.
A Multinational Survey of Diuretic Use in Acute Kidney Injury 245
unethical. Another 11.9% provided additional comments such as it would
depend on the protocol or trial design, and on the presence or absence of dan-
gerous fluid overload or pulmonary edema. Overall, 85.1% stated they would
be interested in participating in an RCT of diuretics in critically ill patients with
AKI. Most responded that they would be willing to enroll patients with a vari-
ety of clinical features including: oliguria for 2 (57.3%) or 6 h (84%); an ele-
vated serum creatinine (84.1%); septic shock (65.9%), and those likely to need
RRT (88.3%).
Additional Comments
Additional comments were made by 78 (24.6%) respondents. While the
comments varied, they were grouped into 4 broad themes. The majority of com-
ments (73.1%) provided further details on beliefs about the clinical impact of
diuretics, and when and for what indications diuretics are should be used in
AKI. The majority mentioned diuretics in the context of fluid overload, improv-
ing urine output, enabling adequate nutrition, or simply that they do not use
diuretics in AKI.
Another 19.2% provided feedback on issues related to RCT design. For
example, remarks suggested that a trial protocol should include: provisions for
exclusion of post-renal etiologies; provisions to ensure patients were not vol-
ume depleted; provisions to allow attending clinicians to place patients on RRT
or give a rescue dose of diuretics if indicated, and finally suggestions for the
enrollment of particular subgroups (i.e. cardiac surgical patients).
There were 3 (3.9%) comments that focused on the ethics of potentially
withholding diuretics (i.e. allocated to placebo) and uncertainty about clinical
equality in the context of the available literature. Finally, only 3 (3.9%) respon-
dents remarked on the survey, in particular on the wording of questions and how
the questions appeared to focus on the potential benefit of diuretics in AKI
rather than harm.
Discussion
We have conducted a multinational multicenter survey of intensive care
specialists, nephrologists and other clinicians routinely involved in the manage-
ment of critically ill patients with AKI. Our principal objective was to gain
insight into the existing beliefs, attitudes and self-reported practice patterns of
diuretic use in patients with AKI.
There are a number of important findings from this survey. First, we found
that furosemide is by far the most common and universally administered
diuretic and that diuretics are generally given intravenously as a bolus, however,
Bagshaw/Delaney/Jones/Ronco/Bellomo 246
they are also commonly given by continuous IV infusion. A urine output in the
range of 0.5–1.0 ml/kg/h is most commonly targeted. When deciding on the
dose, clinicians consider several factors important, in particular, the baseline
serum creatinine, current urine output, hemodynamics and the risk for toxicity.
Second, we found that the presence of pulmonary edema was the most widely
accepted indication for diuretics use. Moreover, clinicians would be reluctant to
potentially enroll patients into a trial where placebo would be allocated to a
patient with evidence of pulmonary edema unless there were clear provisions in
the study protocol for rescue therapy if necessary. Third, the majority of respon-
dents do not believe that use of diuretics in ARF will directly contribute to
improved outcomes such as reduced mortality or need for RRT or increased
renal recovery. On the other hand, a significant proportion of respondents were
unsure, in particular on whether diuretics could reduce either the need for or the
duration of RRT. Finally, we found that most believed that enrollment in an
RCT, where patients could be allocated to placebo, would be considered ethical.
What is more, most would be willing to participate in such an RCT and enroll
patients with a variety of clinical presentations.
So far clinical trials of loop diuretics have been small, confounded by
co-interventions (i.e. mannitol and dopamine), and characterized by delayed or
late intervention (i.e. prolonged periods of oligo-anuria or already on RRT at
enrollment). This is potentially important, as delay to appropriate therapy in
AKI has been identified as a potential contributor to increased mortality and
reduced the likelihood of renal recovery [21, 22]. Likewise, in these trials,
furosemide was often given in large IV bolus doses, where no specific titration
of therapy to physiologic endpoints such as urine output was performed. In
addition, these trials failed to include critically ill patients. The evidence from
these trials largely forms the basis for the prevailing view on whether diuretics
can potentially impact outcome in AKI. Yet, interestingly, the pattern of practice
as described in this survey would appear somewhat contradictory with this evi-
dence, in particular when considering how and when diuretics are given.
The majority of respondents in this survey administered diuretics early in the
course of AKI, and 76.6 and 52.1% titrated them to urine output and fluid
balance targets.
While only a minority reported not using urine output or fluid balance tar-
gets nor using diuretics at all during the course of AKI, many further clarified
that they reserved the use of diuretics for patients with clinically important fluid
overload. This may become a greater issue now considering that the incidence
of AKI has recently been shown to be as high as 67% in critically ill patients
[1]. More importantly, such AKI regularly occurs in the context of additional
organ failure such as acute lung injury or sepsis that prompts the need for
mechanical ventilation or vasoactive therapy [2, 23].
A Multinational Survey of Diuretic Use in Acute Kidney Injury 247
The risk of toxicity was identified by many respondents as an important
determinant for dosing of diuretics in AKI. Prior trials have generally adminis-
tered diuretics by large IV bolus doses [10–14, 24]. Such a pattern of administra-
tion may predispose patients to toxicity that can persist indefinitely, especially
deafness, tinnitus and vertigo [10, 13, 14, 24]. Recently, a small pilot RCT was
performed that compared intermittent bolus with continuous infusion protocols
of furosemide titrated to urine output in critically ill patients with AKI (Ostermann
et al., personal commun.). This trial found that while both protocols were able
to realize a similar diuresis and fluid balance, use of a continuous infusion
needed approximately half the total daily dose, thus had greater effectiveness
(milliliters urine per milligram furosemide) when compared with intermittent
bolus dosing. This would suggest that there is a rationale for the use of a con-
tinuous infusion that is titrated to hourly urine output and fluid balance goals to
improve efficacy while reducing the potential for toxicity.
The findings of this survey, taken with the fact that most respondents are
willing to participate and believe an RCT is ethical, suggest that there is not
only clinical agreement, but also a need for higher quality and more definitive
evidence on early diuretic use in AKI. The limitations to the available trial data
on diuretics in AKI may, perhaps, account for the significant proportion of
clinicians who are uncertain of whether diuretics can influence outcomes, and
which outcomes in particular, for critically ill patients with AKI. Recent
insights have suggested that timely attention to early AKI is warranted and that
intervention with diuretics may improve not just short-term physiology but have
the potential to impact clinically important measures in terms of both kidney
and non-kidney organ function. Importantly, many clinicians identified the
presence of pulmonary edema as a potential barrier to enrollment, where allo-
cation to placebo would be considered unethical unless, however, the study pro-
tocol clearly allowed for rescue therapy if deemed necessary. Otherwise, the
majority reported a willingness to enroll AKI patients with an array of criteria
signifying that a broad critically ill population could be represented in such a
trial.
This study has several strengths and limitations. First, we used an uncon-
ventional method for sampling, however we believe this was the most efficient
and expedient method to support the objective of capturing a broad range of
practice across several centers in several countries. Second, the sampling frame
of this survey, while comprised of nearly 30% from nephrologists, was largely
focused on intensivists. However, we believe this is justified when considering
that AKI now largely occurs as a complication of critically illness where not
only AKI, but all accompanying care is performed in an intensive care setting.
Similarly, sampling occurred largely in tertiary/academic centers. While this
may not completely reflect practice patterns overall, our rationale was to
Bagshaw/Delaney/Jones/Ronco/Bellomo 248
capture responses from centers with potential interest in participation in a clin-
ical trial.
Conclusions
We have conducted a multinational multicenter survey of intensive care
specialists, nephrologists and other clinicians routinely involved in the man-
agement of critically ill patients with AKI to better understand the current
viewpoints and practice patterns of diuretic use associated with AKI. We have
shown that clinicians are most familiar and comfortable with use of furosemide
given intravenously and titrated to the physiologic endpoint of urine output and
fluid balance. While many clinicians either do not believe or are uncertain that
diuretics can directly contribute to improved outcomes in AKI, most clinicians
believe an RCT on diuretic use in AKI is justified and ethical. Moreover, most
would be willing to participate and enroll patients with a variety of clinical
presentations. This survey confirms both clinical agreement and the need for
higher quality and more definitive evidence on diuretic use in AKI.
Acknowledgements
S.M.B. is supported by clinical fellowships from the Canadian Institutes for Health
Research and the Alberta Heritage Foundation for Medical Research Clinical Fellowship.
References
1 Hoste EA, Clermont G, Kersten A, et al: RIFLE criteria for acute kidney injury are associated with
hospital mortality in critically ill patients: a cohort analysis. Crit Care 2006;10:R73.
2 Uchino S, Kellum JA, Bellomo R, et al: Acute renal failure in critically ill patients: a multina-
tional, multicenter study. JAMA 2005;294:813–818.
3 Abel RM, Beck CH Jr, Abbott WM, Ryan JA Jr, Barnett GO, Fischer JE: Improved survival from
acute renal failure after treatment with intravenous essential L-amino acids and glucose. Results of
a prospective, double-blind study. N Engl J Med 1973;288:695–699.
4 Ronco C, Bellomo R, Homel P, et al: Effects of different doses in continuous veno-venous haemofil-
tration on outcomes of acute renal failure: a prospective randomised trial. Lancet 2000;356:26–30.
5 Schiffl H, Lang SM, Fischer R: Daily hemodialysis and the outcome of acute renal failure. N Engl
J Med 2002;346:305–310.
6 Uchino S, Doig GS, Bellomo R, et al: Diuretics and mortality in acute renal failure. Crit Care Med
2004;32:1669–1677.
7 Van Biesen W, Yegenaga I, Vanholder R, et al: Relationship between fluid status and its manage-
ment on acute renal failure (ARF) in intensive care unit (ICU) patients with sepsis: a prospective
analysis. J Nephrol 2005;18:54–60.
8 DeTorrente A, Miller PD, Cronin RE, Paulsin PE, Erickson AL, Schrier RW: Effects of furosemide
and acetylcholine in norepinephrine-induced acute renal failure. Am J Physiol 1978;235:F131–F136.
A Multinational Survey of Diuretic Use in Acute Kidney Injury 249
9 Kramer HJ, Schuurmann J, Wassermann C, Dusing R: Prostaglandin-independent protection by
furosemide from oliguric ischemic renal failure in conscious rats. Kidney Int 1980;17:455–464.
10 Cantarovich F, Fernandez JC, Locatelli A, Perez Loredo J: Frusemide in high doses in the treat-
ment of acute renal failure. Postgrad Med J 1971;47(suppl):13–17.
11 Cantarovich F, Rangoonwala B, Lorenz H, Verho M, Esnault VL: High-dose furosemide for estab-
lished ARF: a prospective, randomized, double-blind, placebo-controlled, multicenter trial. Am J
Kidney Dis 2004;44:402–409.
12 Karayannopoulos S: High-dose frusemide in renal failure. Br Med J 1974;2:278–279.
13 Kleinknecht D, Ganeval D, Gonzalez-Duque LA, Fermanian J: Furosemide in acute oliguric renal
failure. A controlled trial. Nephron 1976;17:51–58.
14 Shilliday IR, Quinn KJ, Allison ME: Loop diuretics in the management of acute renal failure: a
prospective, double-blind, placebo-controlled, randomized study. Nephrol Dial Transplant 1997;12:
2592–2596.
15 Vargas Hein O, Staegemann M, Wagner D, et al: Torsemide versus furosemide after continuous
renal replacement therapy due to acute renal failure in cardiac surgery patients. Ren Fail 2005;27:
385–392.
16 Mehta RL, Pascual MT, Zhuang S, McDonald BR, Gabbai FB, Pahl MV: Effect of diuretic use on
outcomes from acute renal failure (ARF) in the intensive care unit (ICU) (abstract). J Am Soc
Nephrol 2001;12:229A.
17 Kellum JA: The use of diuretics and dopamine in acute renal failure: a systematic review of the
evidence. Crit Care 1997;1:53–59.
18 Lameire N, Vanholder R, Van Biesen W: Loop diuretics for patients with acute renal failure: help-
ful or harmful? JAMA 2002;288:2599–2601.
19 Noble DW: Acute renal failure and diuretics: propensity, equipoise, and the need for a clinical
trial. Crit Care Med 2004;32:1794–1795.
20 Schetz M: Should we use diuretics in acute renal failure? Best Pract Res Clin Anaesthesiol 2004;
18:75–89.
21 Gettings LG, Reynolds HN, Scalea T: Outcome in post-traumatic acute renal failure when contin-
uous renal replacement therapy is applied early vs. late. Intensive Care Med 1999;25:805–813.
22 Bagshaw SM, Mortis G, Godinez-Luna T, Doig CJ, Laupland KB: Renal recovery after severe
acute renal failure. Int J Artif Organs 2006;29:1023–1030.
23 Bagshaw SM, Laupland KB, Doig CJ, et al: Prognosis for long-term survival and renal recovery in
critically ill patients with severe acute renal failure: a population-based study. Crit Care 2005;9:
R700–R709.
24 Brown CB, Ogg CS, Cameron JS: High dose frusemide in acute renal failure: a controlled trial.
Clin Nephrol 1981;15:90–96.
Dr. Sean M. Bagshaw
Division of Critical Care Medicine
University of Alberta Hospital, University of Alberta
3C1.16 Walter C. Mackenzie Centre
8440-112 Street, Edmonton, Alta. T6G 2B7 (Canada)
Tel. 001 780 407 6755, Fax 001 780 407 1228
E-Mail bagshaw.sean@gmail.com
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 250–258
Stem Cells In Acute Kidney Injury
Benedetta Bussolati, Giovanni Camussi
Renal Vascular Pathophysiology Laboratory, Department of Internal Medicine,
University of Turin, Turin, Italy
Abstract
The susceptibility of developing acute renal failure depends on the ability of the kidney
to recover from acute injury and regain normal function. Recently, the possible contribution
of stem cells (SCs) to the regeneration of acute tubular injury has been investigated. There is
evidence indicating that, under pathophysiological conditions, SCs derived from bone mar-
row are able to migrate in the injured kidney but they seem to play a minor role in tubular
regeneration in regard to the resident cells. However, the administration of ex vivo expanded
bone marrow-derived mesenchymal SCs has proven to be beneficial in various experimental
models of acute renal failure. The mechanism underlining this beneficial effect is still matter
of debate. The transdifferentiation or fusion of SCs to repopulate tubules is considered to
play a minor role. The administered SCs may, however, modify the microenvironment by
inducing dedifferentiation and proliferation of tubular cells surviving to injury or by allow-
ing expansion of resident SCs. The recent identification of resident progenitor/SC popula-
tions in the adult kidney supports the hypothesis that resident SCs may play a critical role in
the repair of renal injury. Therefore, therapeutic strategies to exploit the regenerative poten-
tial of SCs may be based on the administration of ex vivo expanded SCs or on stimulation of
expansion and differentiation of local progenitor/SC populations.
Copyright © 2007 S. Karger AG, Basel
The ability of the kidney to recover following episodes of acute injury has
a critical impact on patient morbidity and mortality in the hospital setting.
Renal tubular cells are particularly susceptible to injury when exposed to
endogenous cytokines as in sepsis, to endogenous or exogenous toxins as myo-
globin or aminoglycosides and radiocontrast agents, or to episodes of renal
ischemia [1]. The susceptibility of developing acute renal failure (ARF)
depends on the ability of renal tubules to regenerate and regain normal func-
tion. The age of the patient and the severity of injury may condition recovery.
After severe or repeated episodes of renal injury, recovery can be impaired or
Stem Cells In Acute Kidney Injury 251
even fail leading to the need for long-term dialysis and to an increase in patient
mortality [2]. Necrosis and loss of tubular epithelial cells is the most common
event in ARF [3] and the recovery of renal function following ARF is dependent
on the replacement of necrotic tubular cells with functional tubular epithelium
[4, 5]. The absence or reduction of epithelial and endothelial regeneration may
predispose to tubulo-interstitial scarring and chronic renal disease [6]. Studies
on the physiological response to renal injury indicate that, after the insult has
occurred, tubular cells dedifferentiate and acquire a mesenchymal phenotype.
Dedifferentiated cells then migrate into the regions where tubular cells undergo
necrosis, apoptosis or detachment with denudation of the tubular basement
membrane. This process is followed by cell proliferation and eventually by their
subsequent differentiation into functional epithelial cells with restoration of tis-
sue integrity. The process of dedifferentiation, migration, proliferation and
eventual redifferentiation is thought to be orchestrated by the local release of
growth factors such as hepatocyte growth factor, epidermal growth factor, and
insulin-like growth factor-1 [for review see, 7, 8].
It has been also suggested that the interstitium of the kidney contains adult
renal stem cells (SCs) capable of contributing to renal repair [9]. In addition,
SCs from bone marrow have been implicated in renal repair either by secretion
of factors that protect the tubular cells, or by migration to the injured tubule and
transdifferentiation into tubular epithelial cells [9].
SCs are characterized by their ability to self-renew and differentiate into a
variety of cell types. Several studies have established the plasticity of bone mar-
row-derived SCs as they have the ability to cross lineage boundaries forming
components of different tissues [10]. In bone marrow the hematopoietic SCs
(HSCs) are present that give rise to the majority of the cellular components of
blood and mesenchymal SCs (MSCs) which are able to differentiate into several
cell types including osteoblasts, chondrocytes, adipocytes and myocytes [11].
Recently, resident adult SCs have been isolated from several tissues, including
the central nervous system [12], retina [13], skeletal muscle [14] and liver [15].
Tissue SCs preferentially generate differentiated cells of the same lineage as
their tissue of origin; however, it has been shown that they can also generate
cells of different embryonic lineages [11]. It has been suggested that resident
SCs play a relevant role in the postnatal growth of organs and in the physiolog-
ical turnover of tissues. This is particularly relevant for epithelial organs with a
relatively high rate of cellular turnover, such as the intestine, skin and kidney. In
addition, resident SCs could play a relevant role in the replacement of injured
epithelial cells and in the tissue regeneration after injury.
Although the clinical management of ARF patients has significantly
improved over recent years, we lack specific therapies to improve the rate or effi-
ciency of the repair process. In the present review, we focused on the possible
Bussolati/Camussi 252
contribution and therapeutic implication of SCs both derived from the bone mar-
row or resident in the kidney in the repair of acute renal injury.
Contribution of Bone Marrow-Derived SCs in Renal Regeneration
The possibility that bone marrow-derived SCs might functionally contribute
to renal tubule regeneration is still controversial. Several studies have demon-
strated the presence of Y chromosome-bearing cells in female kidneys trans-
planted into male recipients, suggesting that SCs derived from male bone marrow
migrate and differentiate into the transplanted organ [16–18]. However, only a
small percentage of the total cells present in the kidney was derived from bone
marrow. Using whole bone marrow transplantation in the mouse, Poulsom et al.
[16] demonstrated that bone marrow-derived cells could contribute to the regen-
eration of the renal tubular epithelium. Bone marrow-derived cells were also
shown to ameliorate renal disease in a mouse model of Alport syndrome [19, 20].
In particular, a partial restoration of the expression of type IV collagen ␣
3
chain
was related to the recruitment of bone marrow-derived progenitor cells within the
damaged glomeruli and to their differentiation in podocytes and mesangial cells.
Recent studies limited the role of bone marrow-derived SCs to the regener-
ation of tubular epithelial cells. Using accurate detection of Y-bearing cells, the
percentage of SCs in the renal epithelium was considered extremely low, approx-
imately 0.1% [21, 22]. On the other hand, Lin et al. [23] showed that renal tubu-
lar resident cells provide major contribution to renal repair after ischemia-
reperfusion injury. However, the studies confirmed an effective improvement in
renal function provided by the bone marrow-derived SCs, comparing irradiated
animals and non-irradiated animals or irradiated animals with reconstituted bone
marrow [22, 23]. The hypothesis supported by these data is that the role of bone
marrow-derived SCs in renal repair appears mainly to be limited to the support
of growth factors that may contribute to stimulation of resident mature or SCs to
proliferate and differentiate [24]. In addition, it has been shown that MSCs pos-
sess an immunomodulatory function [25]. It is conceivable that MSCs recruited
by the kidney and present in the renal vessels or interstitium may limit the proin-
flammatory reaction, thus favoring tissue survival [24].
Effect of Administration of in Vitro Expanded SCs in the
Treatment of Acute Renal Failure
Several studies agree with the finding that the administration of in vitro
expanded SCs may protect and reverse ARF [21–23, 26, 27]. The experiments
Stem Cells In Acute Kidney Injury 253
from Morigi et al. [26] demonstrated that the beneficial effect of the adminis-
tration of bone marrow-derived SCs has to be ascribed to MSCs rather than to
HSCs. In addition, the experiments of bone marrow mobilization suggest that
HSCs have a potential detrimental effect on kidney regeneration due to mobi-
lization of cells that may enhance inflammation [28].
The infusion of in vitro expanded MSCs protected and improved the recov-
ery from acute tubular injury induced by cis-platinum and glycerol [26, 27]. In
these models, localization of MSCs within regenerating tubules was observed.
Both these models were characterized by extensive necrosis of proximal and
distal tubules that may favor migration of MSCs within regenerating tubules.
We found that injection of transgenic GFP MSCs rapidly induced functional
recovery and GFP⫹ cells were detectable in tubules after regeneration [27]. The
observed increase in proliferating tubular cells after MSC administration sug-
gest a trophic effect of these cells on resident tubular cells that survived the
injury (fig. 1). The beneficial effect of MSC infusion in acute renal injury
induced by ischemia-reperfusion was also reported by Duffield et al. [22].
However, they only found an interstitial localization of SCs without incorpora-
tion into tubules or endothelium. The discrepancy in the results of MSC inte-
gration into tubules may depend on the different severity of the models used.
However, there is a general agreement on the beneficial effect of MSC infusion
in ARF due either to the ability of these cells to determine a microenvironment
favoring proliferation of dedifferentiated epithelial cells or to stimulate expan-
sion of resident SCs [24].
Renal Resident SCs
The presence of organ-specific progenitor cells capable of differentiating
into epithelia, myofibroblasts and smooth muscle cells has been described in
the embryonic rat metanephric mesenchyme, indicating the presence of embry-
onic renal SCs [29].
In the adult rat kidney, Oliver et al. [30] identified slow cycling SCs, based
on the assumption that, in tissues, SCs are characterized by a low proliferating
rate that maintains the self-renewal of such a population. Cells with a slow
cycling time could be identified by the retention of bromodeoxyuridine (BrdU),
which is incorporated in cell DNA during synthesis. BrdU-retaining cells were
mainly found to be present in renal papilla [30]. Based on these observations, the
authors proposed that the renal papilla is a niche for the adult kidney SCs in rats.
Maeshima et al. [31] also described the presence of BrdU-labelled cells in
the renal tubules of adult rats. These cells, termed renal progenitor-like tubular
cells, were shown to be able to re-enter in mitosis in response to renal injury.
Bussolati/Camussi 254
Kitamura et al. [32] identified a population of renal progenitor cells from
the S3 segment of the nephron in the rat adult kidney. These cells, which
express the renal embryonic markers Pax2, Wtn4 and Wtn1, were shown to be
able to self-renew and to differentiate into mature epithelial cells expressing
aquaporin 1 and 2, and responsive to parathyroid hormone and vasopressin.
These cells contributed to tubular regeneration in rats with ischemia-reperfu-
sion injury. Recently, Gupta et al. [33] demonstrated the presence in the adult
rat kidney of a renal resident population expressing MSC markers, capable of
self-renewal and multipotent differentiation. These cells expressed embryonic
SC markers such as Octa-4 and Pax-2 and, when injected in injured kidneys
Fig. 1. Role of MSCs in the repair of acute tubular injury. MSCs derived from bone
marrow may enter the circulation and reach the sites of tissue injury. MSCs can be recov-
ered from bone marrow, expanded in vitro and administered for therapeutic purposes. The
beneficial effects of MSCs on ARF may depend on their ability to inhibit apoptosis of tubu-
lar cells and recruitment of inflammatory cells. In several studies, the possibility has also
been shown that some MSCs may transdifferentiate into mature epithelial cells. In addition,
MSCs may have a paracrine effect on tubular cells surviving injury by stimulating their ded-
ifferentiation, proliferation, migration and eventually redifferentiation into mature epithelial
cells. In addition, the administered MSCs may modify the microenvironment by allowing
expansion and differentiation of resident SCs that may contribute to the repopulation of
injured tubules.
Inhibition inflammatory cell
recruitment
Inhibition of apoptosis
Injury
Repair
Surviving
cells
Mature
tubular
cells
Paracrine
effect
Resident stem cells
Differentiation
Proliferation
Dedifferentiation
Bone marrow
Mesenchymal
stem cells
Epithelial differentiation
Integration
Stem Cells In Acute Kidney Injury 255
after ischemia-reperfusion, contributed to tubular regeneration. Using a trans-
genic model of Octa-4-X-gal, it was found that Octa-4-positive cells were asso-
ciated with the proximal tubules and were absent in the medulla. A MSC
population has also been identified in the glomeruli of mice [34]. These cells
were shown to be multipotent and to express smooth muscle myosin, like
mesangial cells.
We recently demonstrated the presence of resident progenitors/SCs in the
human adult kidney using CD133 as a SC marker [35]. CD133-positive cells
were found in the interstitium in the proximity of proximal tubules and the
glomerular capsule, and within tubular cells. These cells lacked the expression
of hematopoietic markers (CD34 and CD45) and expressed some MSC mark-
ers, such as CD29, CD90, CD44 and CD73. Moreover, they expressed Pax-2,
an embryonic renal marker, suggesting their renal origin. These cells were
shown to undergo epithelial differentiation when cultured in vitro in the pres-
ence of hepatocyte growth factor and fibroblast growth factor-4. The differenti-
ated epithelial cells expressed proximal and distal tubular markers and, when
cultured in transwell filters, formed a polarized layer showing apical microvilli
and junctional complexes. In vivo, undifferentiated CD133⫹ cells, injected
subcutaneously in Matrigel into SCID mice, spontaneously differentiated in
tubular structures expressing proximal and distal tubular epithelial markers. In
addition, these cells were shown to undergo endothelial differentiation when
cultured in vitro in the presence of VEGF. In vivo, when injected subcuta-
neously in Matrigel after the endothelial differentiation, CD133⫹ cells formed
vessels connected with the mouse vasculature. Moreover, when injected into
mice with ARF induced by glycerol, CD133⫹ renal progenitors homed to the
kidney and integrated into proximal and distal tubules during repair, indicating
the self-renewal ability of this population and the potential contribution to the
repair of acute tubular injury.
Sagrinati et al. [36] recently reported the presence of a population of
CD133⫹ CD24⫹ cells within Bowman’s capsule of adult human kidneys
which exhibited multipotent differentiation capabilities. Indeed these cells were
able to generate not only mature tubular epithelial cells but also osteogenic
cells, adipocytes and neuronal cells. CD133⫹ CD24⫹ SCs were shown to con-
tribute to tubular regeneration in the model of glycerol-induced acute renal
injury.
Recently, in mouse embryonic and adult kidneys, Challen et al. [37]
described the presence of a SC population defined as a ‘side’ population. The
side population is characterized by the unique ability to extrude Hoechst dye.
This side population was found in particular in the tubular compartment and it
was a heterogeneous population [37]. Some progenitor cells within this popula-
tion displayed the characteristics of resident renal SCs and showed a multipotent
Bussolati/Camussi 256
differentiative ability and contributed to the repair of renal injury induced by
adriamycin.
Conclusion
There is evidence that SCs may contribute to the regeneration of acute
tubular injury. Under pathophysiological conditions, SCs derived from bone
marrow, despite their ability to migrate in the injured kidney, play a minor role
in tubular regeneration in regard to resident SCs. However, the administration
of ex vivo expanded bone marrow-derived SCs was proved to be beneficial in
various experimental models of ARF. The mechanism underlining this benefi-
cial effect still remains unclear as transdifferentiation has been demonstrated
only for a minimal portion of repopulated tubules. However, the administered
SCs may modify the microenvironment by allowing expansion of resident SCs
or by inducing dedifferentiation and proliferation of tubular cells surviving
injury (fig. 1). The studies on the potential application of SCs may open new
perspectives for a therapeutic approach to renal injury. This may be achieved
either by administration of ex vivo expanded SCs or by strategies aimed to
expand and differentiate local progenitor/SC populations.
References
1 Thadhani R, Pascual M, Bonventre JV: Acute renal failure. N Engl J Med 1996;334:1448–1460.
2 Star RA: Treatment of acute renal failure. Kidney Int 1998;54:1817–1831.
3 Kelly KJ, Molitoris BA: Acute renal failure in the new millennium: time to consider combination
therapy. Semin Nephrol 2000;20:4–19.
4 Nash K, Hafeez A, Hou S: Hospital-acquired renal insufficiency. Am J Kidney Dis 2002;39:
930–936.
5 Sutton TA, Molitoris BA: Mechanisms of cellular injury in ischemic acute renal failure. Semin
Nephrol 1998;18:490–497.
6 Tobrak FG: Regeneration after acute tubular necrosis. Kidney Int 1992;41:226–246.
7 Lameire N: The pathophysiology of acute renal failure. Crit Care Clin 2005;21:197–210.
8 Abouna GM, Al Adnani MS, Kremer GD, et al: Reversal of diabetic nephropathy in human cadav-
eric kidneys after transplantation into non diabetic recipients. Lancet 1983;2:1274–1276.
9 Cantley LG: Adult stem cells in the repair of the injured renal tubule. Nat Clin Pract Nephrol
2005;1:22–32.
10 Quesenberry PJ, Colvin GA, Abedi M, Dooner G, Dooner M, Aliotta J, Keaney P, Luo L, Demers D,
Peterson A, Foster B, Greer D: The stem cell continuum. Ann NY Acad Sci 2005;1044:228–235.
11 Jiang Y, Jahagirdar BN, Reinhardt RL, Schwartz RE, Keene CD, Ortiz-Gonzalez XR, Reyes M,
Lenvik T, Lund T, Blackstad M, Du J, Aldrich S, Lisberg A, Low WC, Largaespada DA, Verfaillie
CM: Pluripotency of mesenchymal stem cells derived from adult marrow. Nature 2002;418:41–49.
12 Reynolds BA, Weiss S: Generation of neurons and astrocytes from isolated cells of the adult mam-
malian central nervous system. Science 1992;255:1707–1710.
13 Tropepe V, Coles BL, Chiasson BJ, Horsford DJ, Elia AJ, McInnes RR, van der Kooy D: Retinal
stem cells in the adult mammalian eye. Science 2000;287:2032–2036.
Stem Cells In Acute Kidney Injury 257
14 Jackson KA, Mi T, Goodell MA: Hematopoietic potential of stem cells isolated from murine
skeletal muscle. Proc Natl Acad Sci USA 1999;96:14482–14486.
15 Herrera MB, Bruno S, Buttiglieri S, Tetta C, Gatti S, Deregibus MC, Bussolati B, Camussi G:
Isolation and characterization of a stem cell population from adult human liver. Stem Cells
2006;24:2840–2850.
16 Poulsom R, Forbes SJ, Hodivala-Dilke K, Ryan E, Wyles S, Navaratnarasah S, Jeffery R, Hunt T,
Alison M, Cook T, Pusey C, Wright NA: Bone marrow contributes to renal parenchymal turnover
and regeneration. J Pathol 2001;195:229–235.
17 Lin F, Cordes K, Li L, Hood L, Couser WG, Shankland SJ, Igarashi P: Hematopoietic stem cells
contribute to the regeneration of renal tubules after renal ischemia-reperfusion injury in mice. J
Am Soc Nephrol 2003;14:1188–1199.
18 Gupta S, Verfaillie C, Chmielewski D, Kim Y, Rosenberg ME: A role for extrarenal cells in the
regeneration following acute renal failure. Kidney Int 2002;62:1285–1290.
19 Prodromidi EI, Poulsom R, Jeffery R, Roufosse CA, Pollard PJ, Pusey CD, Cook HT: Bone mar-
row derived-cells contribute to podocyte regeneration and amelioration of renal disease in a mouse
model of Alport syndrome. Stem Cells 2006;24:2448–2455.
20 Sugimoto H, Mundel TM, Sund M, Xie L, Cosgrove D, Kalluri R: Bone-marrow-derived stem
cells repair basement membrane collagen defects and reverse genetic kidney disease. Proc Natl
Acad Sci USA 2006;103:7321–7326.
21 Kale S, Karihaloo A, Clark PR, Kashgarian M, Krause DS, Cantley LG: Bone marrow stem cells
contribute to repair of the ischemically injured renal tubule. J Clin Invest 2005;112:42–49.
22 Duffield JS, Park KM, Hsiao LL, Kelley VR, Scadden DT, Ichimura T, Bonventre JV: Restoration
of tubular epithelial cells during repair of the postischemic kidney occurs independently of bone
marrow-derived stem cells. J Clin Invest 2005;115:1743–1755.
23 Lin F, Moran A, Igarashi P: Intrarenal cells, not bone marrow-derived cells, are the major source
for regeneration in postischemic kidney. J Clin Invest 2005;115:1756–1764.
24 Krause D, Cantley LG: Bone marrow plasticity revisited: protection or differentiation in the kid-
ney tubule? J Clin Invest 2005;115:1705–1708.
25 Le Blanc K: Immunomodulatory effects of fetal and adult mesenchymal stem cells. Cytotherapy
2003;5:485–489.
26 Morigi M, Imberti B, Zoja C, Corna D, Tomasoni S, Abbate M, Rottoli D, Angioletti S, Benigni A,
Perico N, Alison M, Remuzzi G: Mesenchymal stem cells are renotropic, helping to repair the kid-
ney and improve function in acute renal failure. J Am Soc Nephrol 2004;15:1794–1804.
27 Herrera MB, Bussolati B, Bruno S, Fonsato V, Mauriello-Romanazzi G, Camussi G: Mesenchymal
stem cells contribute to the renal repair of acute tubular epithelial injury. Int J Mol Med 2004;14:
1035–1041.
28 Togel F, Isaac J, Westenfelder C: Hematopoietic stem cell mobilization-associated granulocytosis
severely worsens acute renal failure. J Am Soc Nephrol 2004;15:1261–1267.
29 Oliver JA, Barasch J, Yang J, Herzlinger D, Al-Awqati Q: Metanephric mesenchyme contains
embryonic renal stem cells. Am J Physiol Renal Physiol 2002;283:799–809.
30 Oliver JA, Maarouf O, Cheema FH, Martens TP, Al-Awqati Q: The renal papilla is a niche for adult
kidney stem cells. J Clin Invest 2004;114:795–804.
31 Maeshima A, Yamashita S, Nojima Y: Identification of renal progenitor-like tubular cells that par-
ticipate in the regeneration processes of the kidney. J Am Soc Nephrol 2003;14:3138–3146.
32 Kitamura S, Yamasaki Y, Kinomura M, Sugaya T, Sugiyama H, Maeshima Y, Makino H:
Establishment and characterization of renal progenitor like cells from S3 segment of nephron in
rat adult kidney. FASEB J 2005;19:1789–1797.
33 Gupta S, Verfaillie C, Chmielewski D, Kren S, Eidman K, Connaire J, Heremans Y, Lund T,
Blackstad M, Jiang Y, Luttun A, Rosenberg ME: Isolation and characterization of kidney-derived
stem cells. J Am Soc Nephrol 2006;17:3028–3040.
34 da Silva Meirelles L, Chagastelles PC, Nardi NB: Mesenchymal stem cells reside in virtually all
post-natal organs and tissues. J Cell Sci 2006;119:2204–2213.
35 Bussolati B, Bruno S, Grange C, Buttiglieri S, Deregibus MC, Cantino D, Camussi G: Isolation of
renal progenitor cells from adult human kidney. Am J Pathol 2005;166:545–555.
Bussolati/Camussi 258
36 Sagrinati C, Netti GS, Mazzinghi B, Lazzeri E, Liotta F, Frosali F, Ronconi E, Meini C, Gacci M,
Squecco R, Carini M, Gesualdo L, Francini F, Maggi E, Annunziato F, Lasagni L, Serio M,
Romagnani S, Romagnani P: Isolation and characterization of multipotent progenitor cells from
the Bowman’s capsule of adult human kidneys. J Am Soc Nephrol 2006;17:2443–2456.
37 Challen GA, Martinez G, Davis MJ, Taylor DF, Crowe M, Teasdale RD, Grimmond SM, Little
MH: Identifying the molecular phenotype of renal progenitor cells. J Am Soc Nephrol 2004;15:
2344–2357.
Prof. G. Camussi
Cattedra di Nefrologia, Dipartimento di Medicina Interna
Corso Dogliotti 14
IT–10126 Torino (Italy)
Tel. ⫹39 11 633 6708, Fax ⫹39 11 663 1184, E-Mail giovanni.camussi@unito.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 259–266
Anticoagulation Options for Patients with
Heparin-Induced Thrombocytopenia
Requiring Renal Support in the Intensive
Care Unit
Andrew Davenport
Centre for Nephrology, Division of Medicine, Department of Medicine, Royal Free
and University College Medical School, London, UK
Abstract
World wide, heparins are the most commonly used anticoagulants for renal replacement
therapy (RRT). In the intensive care unit (ICU) keeping the RRT circuit patent is more difficult
than during routine outpatient hemodialysis, as ICU patients typically have sepsis and/or
inflammation resulting in activation of the procoagulant pathways, with reduced antithrombin.
One important cause of repeated RRT circuit clotting is heparin-induced thrombocytopenia
(HIT), which should not be overlooked in patients with a reduced platelet count. If HIT is clin-
ically suspected then all heparins should be withdrawn, and the patient systemically anticoagu-
lated with either a direct thrombin inhibitor, such as argatroban and/or hirudin, or the heparinoid
danaparoid. The availability and licensing of these alternative anticoagulants varies from coun-
try to country. Argatroban has to be continuously infused, which is an advantage for continuous
RRT, but not for intermittent RRT, and can be monitored by activated partial thromboplastin
time. Hirudin has a prolonged half life, which is extended by hirudin antibodies, and requires
specialist monitoring to prevent over anticoagulation. Although the half life of danaparoid is
increased in renal failure, it can be given as boluses for intermittent and continuous RRT, or by
continuous infusion during continuous RRT, but requires factor Xa monitoring.
Copyright © 2007 S. Karger AG, Basel
Immune-mediated heparin-induced thrombocytopenia (HIT) is uncommon
in general medical and/or surgical intensive care units (ICUs), with an inci-
dence of ⬍2% in those using low molecular weight heparins [1]. The incidence
may be greater using unfractionated heparins, particularly bovine, and post-
cardiac surgery [2]. In the ICU, critically ill patients have been reported to have
up to 46% ‘all cause’ thrombocytopenia and many receive heparin (table 1) [1].
Practice Patterns for RRT in the ICU
Davenport 260
Table 1. Causes of thrombocytopenia in the ICU patient
Artefactual Inadequate anticoagulation Platelet clumping
EDTA-activated agglutination Platelet clumping
Dilutional Massive blood transfusion
Distributional Hypersplenism Sequestration
↓ Platelet production Viral infections Parvo virus
Human immunodeficiency virus
Epstein-Barr virus
Drugs Chemotherapy
Radiotherapy
Toxins Alcohol
Nutritional deficiencies Folic acid
Vitamin B
12
Liver disease Acute liver failure
Bone marrow disorders Severe sepsis
Aplasia/hypoplasia
Myelodysplastic syndromes
Myeloproliferative syndromes
↑ Platelet destruction Immune-mediated Idiopathic ITP
Drug Heparin
Sodium valproate
Quinine
Infection Severe sepsis
Human immunodeficiency virus
Epstein-Barr virus
Cytomegalovirus
Allo-immune destruction ABO incompatible transfusion
ABO incompatible organ
transplantation
Nonimmune infection Malaria
Viral hemorrhagic fevers
Disseminated intravascular
coagulation
Thrombotic thrombocytopenia ↓ Von Willebrand factor
Cleaving protease
Hemolytic uremic syndrome Idiopathic
Factor H deficiency/mutation
Membrane complement-
binding protein
deficiency/mutation
Drug induced
Anticoagulation for Patients with HIT 261
As there are many other causes of thrombocytopenia in the ICU patient other
than HIT, a pre-test probability score, known as the ‘4 Ts’ scoring system (table 2)
has been developed based on both clinical and laboratory findings to help pre-
dict the likelihood that a patient will have HIT [1].
If HIT is clinically suspected, then all exposure to heparins (both unfrac-
tionated and low molecular weight heparin) must be stopped, as even heparin
catheter locks and flushes can result in fatal reactions [3].
HIT develops due to the binding of heparin to platelet factor 4 (PF4), which
is released from the ␣ granules of activated platelets. When there is a key stoi-
chimetric ratio of heparin molecules to PF4, binding leads to a conformational
change in the PF4 molecule, allowing so-called neo-epitopes to be exposed,
resulting in the formation of autoantibodies. Typically these are IgG antibodies
against the PF4-heparin complex, although occasionally the autoantibodies are
Table 1. (continued)
Antiphospholipid syndrome
Pregnancy HELLP syndrome
Drug-induced Anti-platelet agents
abciximab, eptifibatide
Mechanical destruction Cardiopulmonary bypass
Table 2. The ‘4 Ts’ scoring system to estimate the pretest probability of heparin-
induced thrombocytopenia
Score
2 points 1 point zero
Thrombocytopenia 20–100 or 10–19 or ⬍10 or
⫻10
9
/l fall ⬎ 50% fall 30–50% fall ⬍ 30%
Timing of onset in 5–10 days ⬎10 days or ⬍1 day heparin
fall platelets heparin Rx timing not evident exposure
Thrombosis or Proven thrombosis Progressive, recurrent, none
acute systemic Skin necrosis or silent thrombosis or
symptoms acute systemic erythematous skin
reaction lesions
Other etiology for None evident possible probable
thrombocytopenia
Low probability ⱕ 3; intermediate probability 4–6; high probability ⱖ 6.
Davenport 262
of other isotypes and sometimes directed against other platelet-derived chemokines.
Once formed the antibodies bind to platelets, causing platelet, monocyte and
endothelial cell activation with increased potential for both arterial and venous
thrombosis. The peripheral platelet count falls due to sequestration of the acti-
vated platelets (fig. 1). These antibodies can now be detected using commer-
cially available ELISA kits and gel particle agglutination assays.
In patients with a high probability of HIT based on the ‘4 Ts’ scoring sys-
tem, then all heparin therapy should be withdrawn whilst awaiting confirmatory
laboratory testing. If patients are subsequently found to be HIT positive, then
despite heparin withdrawal there is an increased risk of thrombosis, even when
the peripheral platelet count has normalized [4], and therefore there is a need for
an alternative anticoagulant. Whereas in the chronic dialysis patient a regional
anticoagulant may suffice to allow routine hemodialysis, in the ICU patient sys-
temic anticoagulation is usually required. Currently two main groups of antico-
agulants are available, the heparinoids and the direct thrombin inhibitors.
The Heparinoids
Danaparoid is a heparinoid composed of a mixture of 84% heparin sulfate,
12% dermatan sulfate and 4% chondroitin sulfate. Its mechanism of action
Fig. 1. The temporal change in peripheral platelet count in a patient who developed
HIT during continuous renal replacement therapy (CRRT) and was anticoagulated with
heparin, low molecular weight heparin (LMWH) and danaparoid. UFH ⫽ Unfractionated
heparin.
300
UFH
CRRT
LMWH Danaparoid
252015
Time (days)
1050
0
50
100
200
150
250
Peripheral platelet count × 10
9
/l
Anticoagulation for Patients with HIT 263
remains to be fully elucidated, but it inhibits factor Xa and, to a lesser extent,
thrombin [5]. Although in about 5–20% of cases danaparoid exhibits in vitro
cross-reactivity with the HIT antibodies [2, 6], in vivo cross-reactivity is rare,
with the occasional clinical case reported [7]. In non-renal patients, a bolus of
2,500 anti-Xa units is recommended followed by a continuous infusion reduc-
ing from 400 IU/h for 4 h, to 300 IU/h for a further 4 h, then to 200 IU/h as a
maintenance dose [2]. However as danaparoid has a prolonged half life in renal
failure, and is not cleared by renal replacement therapy (RRT), this dosing reg-
imen has to be reduced in renal failure. Most centers use a bolus of 750 IU for
continuous RRT, followed by a maintenance infusion starting at 1–2 U/kg ⭈ h,
then adjusting the dose to maintain an anti-Xa of 0.2 to ⬍0.35 IU/ml. For inter-
mittent RRT, a bolus of 2,000–2,500IU is used depending upon patient weight,
and then adjusted according to the anti-Xa level. The recent development of
bedside testing for anti-Xa will allow easier monitoring.
Dermatan sulfate has been used as a sole anticoagulant, with reports of a
loading dose of 150 mg followed by an infusion of 15mg/h. Dermatan sulfate acts
through heparin cofactor II and inhibits thrombin and fibrin-bound thrombin. As
such, the activated partial thromboplastin time (aPTT) has been used to monitor
anticoagulation, aiming for a ratio of 1.0–1.4 over the laboratory baseline [6].
Fondaparinux is a synthetic analogue of the antithrombin-binding pen-
tasaccharide found in heparins. Fondaparinux binds to antithrombin, predomi-
nantly inhibiting factor X, and so requires anti-Xa monitoring, but also has
some action by inhibiting factor IXa [8]. It is renally excreted, so the plasma
half life is markedly increased in renal failure [4], and single case reports have
suggested an alternate day dose of 2.5mg for hemodialysis. Although HIT anti-
bodies have been reported with fondaparinux, these only poorly recognize fon-
daparinux, and are not thought to be clinically relevant [8].
Unlike unfractionated heparin, there is no specific antidote for the hepari-
noids in cases of hemorrhage, due to over anticoagulation. However recombi-
nant factor VIIa has been used successfully to control hemorrhage in combination
with tranexamic acid [9].
The Direct Thrombin Inhibitors
Argatroban directly inhibits free and clot bound thrombi. The half live is
modestly prolonged in renal failure, to around 35 min, but it accumulates in
liver failure [10]. Thus a bolus dose of argatroban (250 g/kg) is followed by an
infusion, starting at 0.5–2.0 g/kg ⭈ min, then titrating the dose according to the
aPTT, aiming for a ratio of 1.0–1.4. Currently, the clinical experience with arga-
troban is limited to North America.
Davenport 264
Recombinant hirudin irreversibly inhibits free and clot-bound thrombin.
The commercially available r-hirudin, lepirudin, has a markedly increased half
life in renal failure, and accumulates during RRT. Although some lepirudin is
removed by high flux membranes, in around 60% of patients antibodies develop
and these prevent lepirudin removal during RRT. The relationship between
aPPT and plasma hirudin is not linear, and as the aPTT ratio increases to 1.5
and above, the rate of rise in aPTT is much lower than the increase in the
hirudin concentration, so increasing the risk of bleeding. Thus the risks of hem-
orrhage with lepirudin are associated with renal failure and increased aPPT
ratios [11]. To overcome this problem, a direct thrombin test using viper venom,
called the ecarin clotting time (ECT), has been introduced.
Bolus lepirudin doses of 0.2–0.5 mg/kg for intermittent hemodialysis have
been suggested [2], but these are then often adjusted downwards in clinical
practice, aiming for a hirudin concentration of 0.6–1.4mg/l or an ECT of
80–100 s. For continuous RRT, hemorrhage has frequently been reported when
a bolus was followed by a continuous infusion, thus an infusion starting at
0.005–0.01 mg/kg ⭈ h is recommended, and then adjusting the dose downwards.
If the aPTT, ECT and or plasma hirudin start to increase, then it is probably best
to stop the infusion and administer small boluses (0.007–0.05 mg/kg) [6], once
the aPTT ratio has fallen to 1.0. As there is no specific antidote in cases of hem-
orrhage, activated factor VIIa has been used to control bleeding, and in cases
without hirudin antibodies, hemodiafiltration can also be used to help clear
lepirudin.
In addition with lepirudin, a small number of cases of anaphylaxis have
been reported, with an estimated incidence of 0.015% on the first and 0.16% on
the second exposure [2].
Conclusion
Extracorporeal anticoagulation for patients with acute kidney injury
requiring RRT in the ICU is different from that of chronic kidney failure
patients on routine outpatient hemodialysis. In the ICU, the majority of patients
have some degree of sepsis and/or inflammation leading to activation of the
procoagulant pathways shown by reduced levels of proteins S, C, antithrombin,
and tissue factor pathway inhibitor, with increased plasminogen activator
inhibitor-1 and tissue factor [12].
Although thrombocytopenia is relatively common in the ICU, an important
cause not to be overlooked is HIT, as these patients are at risk of major throm-
botic events, sudden cardiorespiratory collapse following bolus heparin admin-
istration, and persistent clotting of RRT circuits and vascular access catheters.
Anticoagulation for Patients with HIT 265
If HIT is clinically suspected then all heparin administration should cease, and
patients systemically anticoagulated with a heparinoid or direct thrombin
inhibitor, until laboratory testing for antibodies can be completed. Of the cur-
rently available anticoagulants argatroban has the shortest half life and has to be
continued as a infusion, adjusted according to the aPPT ratio. However, arga-
troban is not universally available, and also accumulates in liver failure. Both
the heparinoids, danaparoid and fondaparinux, and lepirudin have prolonged
half lives in renal failure, and have no specific antidote in cases of over antico-
agulation and hemorrhage. The heparinoids require anti-Xa monitoring, which
has become easier with the development of bedside testing. Whereas hirudin
should be monitored by either measuring the plasma hirudin concentration or
ECT. After a few days of hirudin exposure, the majority of patients develop
antihirudin antibodies, which further prolong the half life, and increase the risk
of hemorrhage.
Argatroban is currently the systemic anticoagulant of choice for HIT in
patients without severe liver disease but, if not available, then danaparoid is an
effective alternative, and should be started if HIT is suspected and continued
until the results of appropriate laboratory tests are available.
References
1 Wartentin TE, Cook DJ: Heparin, low molecular weight heparin, and heparin-induced thrombocy-
topenia in the ICU. Crit Care Clin 2005;21:513–529.
2 Keeling D, Davidson S, Watson H; Haemostasis and Thrombosis Task Force of the British
Committee for Standards in Haematology: The management of heparin-induced thrombocytope-
nia. Br J Haematol 2006;133:259–269.
3 Davenport A: Sudden collapse during haemodialysis due to immune mediated heparin induced
thrombocytopenia. Nephrol Dial Transplant 2006;21:1721–1724.
4 Arepally GM, Ortel TL: Heparin induced thrombocytopenia. N Engl J Med 2006;355:809–817.
5 Davenport A: Management of heparin induced thrombocytopenia during renal replacement ther-
apy. Hemodial Int 2001;5:81–85.
6 Oudemans van straaten HM, Wester JPJ, de Pont ACJM, Schetz MRC: Anticoagulation strategies
in continuous renal replacement therapy: can the choice be evidence based? Intensive Care Med
2006;32:188–202.
7 Keng TB, Chong BH: Heparin-induced thrombocytopenia and thrombosis syndrome: in vivo
cross-reactivity with danaparoid and successful treatment with r-Hirudin. Br J Haematol 2001;114:
394–396.
8 Weitz JI: Emerging anticoagulants for the treatment of venous thromboembolism. Throm Haemostat
2006;96:274–284.
9 Huvers F, Slappendel R, Benraad B, van Hellemondt G, van Kraaij M: Treatment of postoperative
bleeding after fondaparinux with rFVIIa and tranexamic acid. Neth J Med 2005;63:184–186.
10 Davenport A: Heparin induced thrombocytopenia during renal replacement therapy. Hemodial Int
2004;8:295–303.
11 Tardy B, Lecompte T, Boelhen F, Tardy-Poncet B, Elalamy I, Morange P, Gruel Y, Wolf M,
Francois D, Racadot E, Camarasa P, Blouch MT, Nguyen F, Doubine S, Dutrillaux F, Alhenc-Gelas M,
Martin-Toutain I, Bauters A, Ffrench P, de Maistre E, Grunebaum L, Mouton C, Huisse MG,
Davenport 266
Gouault-Heilmann M, Lucke V; GEHT-HIT Study Group: Predictive factors for thrombosis and
major bleeding in an observational study in 181 patients with heparin-induced thrombocytopenia
treated with lepirudin. Blood 2006;108:1492–1496.
12 Russell JA: Management of sepsis. N Engl J Med 2006;19:1699–1713.
Dr. A. Davenport
UCL Centre for Nephrology, Royal Free and University College Medical School
Rowland Hill Street
London NW3 2PF (UK)
Tel. ⫹44 20 783 02291, Fax ⫹44 20 783 02125, E-Mail Andrew.Davenport@royalfree.nhs.uk
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 267–274
Nutritional Support during Renal
Replacement Therapy
R.Chioléro, M.M. Berger
Department of Adult Intensive Care Medicine, CHUV, Lausanne, Switzerland
Abstract
Background/Aims: Malnutrition is common in critically ill patients with acute renal
failure. The aim of this review is to describe the basis for nutritional support during renal
replacement therapy. Methods: Review of the literature. Results: Techniques of nutritional
support and nutritional requirements are described. Conclusion: Early aggressive enteral,
parenteral or combine nutritional support is required in critically ill patients on replacement
therapy.
Copyright © 2007 S. Karger AG, Basel
The prevalence of malnutrition is elevated in critically ill patients with
acute renal failure (ARF), particularly in those requiring renal replacement
therapy (RRT). Malnutrition is promoted by numerous factors related to the
critical illness, renal failure, RRT and the nutritional support per se. Sepsis is
commonly present in ARF and constitutes another risk factor of malnutrition in
critically ill patients with ARF. The aim of this review is to describe and discuss
the basis of nutritional support in such a condition.
Numerous data demonstrate the negative impact of malnutrition on clinical
outcome in critically ill patients. There are few published data regarding ARF
patients on RRT in the intensive care unit (ICU), but they also suggest that the
occurrence of malnutrition throughout the stay is associated with poor outcome.
In a prospective cohort of 309 patients requiring intermediate care, i.e. with
moderately severe acute illness (mean APACHE II score 23.1), nutritional
assessment was performed using the subjective global assessment, anthropo-
metric, biochemical and immunology markers [1]. RRT was performed in 67%
of the patients (intermittent hemodialysis in 56%, continuous in 11%). In-
hospital mortality amounted to 39%. The overall prevalence of malnutrition
Chioléro/Berger 268
was high: 42% of patients had a normal nutritional status, 16% were moderately
malnourished and 42% had severe malnutrition. In-hospital mortality (OR 7.2,
p 0.001) and the hospital length of stay (p ⬍ 0.001) were significantly increased
in patients with malnutrition. Both septic and non-septic complications were
significantly increased in severely malnourished patients, including the overall
septic morbidity, septic shock, hemorrhage, intestinal complications, cardio-
genic shock, arrhythmias, and acute respiratory failure. In multiple regression
analysis, malnutrition was also an independent predictor of in-hospital mortal-
ity. These data confirm the high prevalence of malnutrition in ARF patients and
its strong impact on clinical outcome and resource utilization.
Early and aggressive nutritional support is therefore a high therapeutic pri-
ority in such patients.
Metabolic Responses, Nutrient Losses
Extensive metabolic changes are induced by the critical illness: hyper-
metabolism, accelerated protein catabolism, and hyperglycemia related to
insulin resistance constitute the hallmarks of the metabolic response to major
stress. Published data suggest that ARF per se does not significantly influence
the level of resting metabolism in critically ill patients, which is usually
increased in surgical, trauma and septic patients. This contrasts with protein
catabolism which is further enhanced by renal failure in comparison to criti-
cally ill patients without ARF [2].
RRT, particularly when continuous, also influences energy metabolism and
substrate utilization. There is no direct effect of continuous RRT (CRRT) on
energy metabolism [2]. A significant effect is only observed when the use of
this supportive treatment is associated with thermal losses and temperature
changes. Studies show that hypothermia is caused by CRRT in 5–50% of
patients: this causes a proportional decrease in resting metabolic rate, except
in patients with insufficient sedations, in whom shivering may induce a sub-
stantial elevation in the metabolic rate. The mean body temperature was
decreased by 2.8⬚C in critically ill patients after CRRT initiation, while the
mean VO
2
was decreased by 26% [3]. This will proportionally increase the
caloric requirements.
Protein catabolism is markedly increased in most ICU patients requiring
CRRT [2, 4]. In addition, substantial amino acid and protein losses, amounting to
10–20 g/day, are caused by CRRT according to the type of membrane and to the
technique of replacement. Recent studies in ICU patients on CRRT suggest that
protein and amino acid supply should be increased up to 2.0–2.5 kg/kg ⭈ day, to
compensate for the accelerated protein catabolism and for the losses in the CCRT
CRRT an Nutrition 269
device [5, 6]. In one study, improved nitrogen balance and clinical outcome were
associated with a high protein supply.
Hyperglycemia is present in 60–90% of fasted and nourished critically ill
patients. In addition to the multiple factors which decrease insulin sensitivity in
such a condition, glucose loss in CRRT may be substantial. Glucose loss is
dependent on the glucose concentration of the replacement solutions: the use of
low-glucose solutions (0.1–0.15%) is associated with small glucose loss (about
4% net glucose loss), whereas solutions containing 1% or more glucose induce
net absorption of glucose [2]. In 8 patients receiving CRRT and replacement
solution without glucose, we observed significant glucose loss in the effluent,
amounting to 60 g/day [7]. Such loss obviously requires increased supply to
insure sufficient carbohydrate administration.
Micronutrients
The critically ill are exposed to an increased oxidative stress, partly due
to their illness, but also to the inflammatory response, and to the oxidative side
effects of ICU treatments (high FiO
2
mechanical ventilation, transfusion, CRRT).
A series of micronutrients are essential for the endogenous antioxidant defenses,
mainly selenium, manganese, zinc, vitamins C and E, while copper and iron
have both antioxidant and prooxidant properties [8]. A German randomized
supplementation study in 42 infected ICU patients using moderate doses of
selenium (500 g/day) showed that 9 days of supplementation were associated
with a reduction in severe ARF and a trend to better survival [9]: the higher
plasma levels resulting from the supplements were associated with higher glu-
tathione peroxidase activity – a major antioxidant selenoenzyme.
Trace element and vitamin metabolism is altered in both chronic renal fail-
ure and ARF, mostly due to losses through the membranes but also to insuffi-
cient nutritional intakes. Low plasma concentrations of selenium and zinc have
repeatedly been shown, as well as vitamin C and vitamin E [10–12]. Balance
studies have been difficult to realize due to the low detection limits of the ana-
lytical methods used until recently. Story et al. [11] analyzed micronutrient
losses in the ultrafiltrate: they were small or undetectable except for vitamin C,
chromium, and copper. Of note, the lipid-soluble vitamins were not detectable.
Our group [12], using inductively coupled plasma mass spectrometry showed
that CRRT causes significant losses and negative micronutrient balances what-
ever the buffer solution: the 4 micronutrients analyzed, selenium, copper, zinc
and thiamine, were lost in large quantities (table 1) and contributed to low
plasma concentrations. While zinc was lost in the effluent, the amounts pro-
vided by the replacement solutions counteracted the losses causing modestly
Chioléro/Berger 270
positive balances. Considering these results, it is likely that all water-soluble
micronutrients are lost in the effluent due to the membrane characteristics,
resulting in rapid depletion of the patient’s pools.
Nutritional Support in CRRT Patients
Prevention of malnutrition is a high priority goal in critically ill patients with
ARF. This requires the administration of early nutritional support, i.e. within the
first 24–48 h after ICU admission: such a strategy avoids the accumulation of
markedly negative daily energy balances in patients with frequent pre-ICU mal-
nutrition. Recent studies show that septic and nonseptic complications are
increased in fasted critically ill patients or in patients receiving hypocaloric feed-
ing: observations in critically ill patients show that the incidence of bacteremia
was significantly increased after only 2 days of hypocaloric feeding [13].
The use of the enteral route should be promoted in all ICU patients, includ-
ing those with ARF: numerous studies show that full or partial enteral feeding is
possible in a large proportion of these patients. This strategy was recently sup-
ported by the implementation of the Canadian guidelines for nutritional support
Table 1. Micronutrients at risk during CRRT: comparison of micronutrient losses in
the effluent, recommended daily parenteral nutrition intakes, and quantities provided by
industrial intravenous supplements
Micronutrient Losses/24 h Daily PN, recommended Range of doses provided
mean values intakes by industrial PN
supplements
1
Chromium 25 mol 15 g 10–15 g
Copper 0.41 mg 1.0–1.2 mg 0.48–1.3 mg
Selenium 110 g60g 24–70g
Zinc 0.2 mg 6.5 mg 3.3–10mg
Vitamin B1 4.1 mg 3 mg 3.0–3.51 mg
Vitamin C 10 mg 100 mg 100–125 mg
Vitamin E ND 10 IU 10–10.2 IU
ND ⫽ Not determined; PN ⫽ parenteral nutrition.
Data from Story et al. [11] and Berger et al. [12].
1
Trace elements: Tracutil
®
, BBraun; Addamel-N
®
/Tracitrans
®
, Fresenius-Kabi; Décan
®
,
Aguettant; vitamins: Soluvit
®
/Vitalipid
®
, Fresenius-Kabi; Cernevit
®
, Baxter. Note: Decan
®
had the highest content for Cr-Se-Zn, lowest for Cu-Fe, and highest vitamin content was in
Cernevit
®
.
CRRT an Nutrition 271
in critically ill patients on mechanical ventilation, showing that full enteral sup-
port or combined enteral and parenteral feeding were possible in 89% of criti-
cally ill patients [14].
In a multicenter cohort study including 17,126 ICU patients requiring
RRT, the use of enteral feeding was associated with an improved probability of
survival [15]. Combined enteral and parenteral or uncommonly exclusive par-
enteral nutrition should be administered to patients totally or partially intolerant
to enteral nutrition to avoid prolonged hypocaloric feeding.
There are no fixed rules to determine the nutritional requirements of all
critically ill patients with ARF, since they are not a homogenous group of
patients: as previously stated, the type and severity of critical illness, severity of
ARF and use of CRRT (technique and dose) are the main determinants of the
nutrition requirements [16]. Keeping such limitations in mind, general recom-
mendations can be provided for ARF patients on CRRT, which constitute a
more homogenous group.
Provision of energy should be normal to high, amounting to 20–30 kcal/
kg ⭈ day. It should be increased in patients with major burns, severe multiple
injury or severe sepsis. It should be decreased in morbidly obese patients in
whom energy supply should be based on ideal body weight. Monitoring of
energy balance is mandatory in patients with prolonged and complicated evolu-
tion to avoid dangerous accumulation of energy deficits or inappropriate hyper-
caloric feeding. Indirect calorimetry, whenever available, is useful to provide
the targeted caloric supply.
Protein supply should be increased during CRRT. In the recent ESPEN rec-
ommendations, they are set at 1.7 g/kg ⭈ day or below [16]. As explained previ-
ously, values up to 2.0–2.5 g/kg ⭈ day may improve both the nitrogen economy
and the clinical outcome [5, 6, 17].
Fat supply should be limited to 20–30% of total energy supply, as in all
critically ill patients. The literature does not furnish data to provide specific rec-
ommendation concerning the type of fat that should be administered. Recent
data in critically ill patients without CRRT suggest that n–3 polyunsaturated
fatty acids, derivate from fish oil, may be useful in patients with acute respira-
tory distress syndrome or sepsis/septic shock.
Carbohydrates should be the main source of caloric supply in CRRT
patients, as in other critically ill patients.
Micronutrient Supply
Micronutrient requirements are increased in the vast majority of critically
ill patients [8], and particularly in those with increased losses and acute deficiencies
Chioléro/Berger 272
such as ARF patients. Therefore early supplementation/substitution is required
to avoid aggravating the oxidative stress and altering macronutrient metabo-
lism. There are no ready mixtures for this purpose on the market, as industrial
intravenous supplements are indeed intended for stable parenteral nutrition
patients, and not to match for additional losses (table 1). Considering that mod-
estly positive zinc balances are not deleterious, the simplest option is to provide
a double dose of one of the existing trace element preparations, even when the
patients are on enteral nutrition. Selenium and thiamine appear from actual data
to be the micronutrients at highest risk of depletion. Therefore additional
100 g of selenium and 100 mg of thiamine should be delivered daily intra-
venously while on CRRT.
Glucose Control
Studies in critically ill patients during the last decade suggest that a tight
control of the blood glucose concentration should be used in all critically ill
patients. Recent studies show that such metabolic control decreases the occur-
rence of renal failure and improves clinical outcome in patients with or with-
out ARF. In a first randomized controlled study performed by van den Berghe
et al. [18] in 1,548 surgical critically ill patients, the probability of developing
ARF and requiring RRT was decreased by 35 and 41%, respectively, by a strict
control of glycemia (4.4–6.1 vs. ⬍12.0 mmol/l). This was associated with a
significant decrease in mortality, length of stay, prolonged mechanical ventila-
tion and septic morbidity. Another study performed by the same authors in
medical critically ill patients again found a decreased occurrence of ARF in
patients receiving strict glycemic control but there was no effect on survival
[19]. Despite such promising results, there is presently hot debate concerning
the optimal target of blood glucose control in ICU patients. It may well be that
different targets are required in different populations. Focusing on patients
with ARF or requiring RRT there is no doubt that avoiding hyperglycemia is
a high priority therapeutic goal, although the precise target remains yet
controversial.
Conclusion
ARF patients have a high prevalence of malnutrition. Providing adequate
nutritional support is a high priority goal in those critically ill patients requiring
CRRT. Provision of early enteral nutrition, adequate amounts of energy, protein
and micronutrients constitute the most important points with special accent on
CRRT an Nutrition 273
the elevated protein and micronutrient requirements. Metabolic monitoring is
mandatory, particularly for blood glucose and energy balance control.
References
1 Fiaccadori E, Lombardi M, Leonardi S, Rotelli C, Tortorella G, Borghetti A: Prevalence and clin-
ical outcome associated with preexisting malnutrition in acute renal failure: a prospective cohort
study. J Am Soc Nephrol 1999;10:581–593.
2 Wooley JA, Btaiche IF, Good KL: Metabolic and nutritional aspects of acute renal failure in criti-
cally ill patients requiring continuous renal replacement therapy. Nutr Clin Pract 2005;20:
176–191.
3 Matamis D, Tsagourias M, Koletsos K, et al: Influence of continuous haemofiltration-related
hypothermia on haemodynamic variables and gas exchange in septic patients. Intensive Care Med
1994;20:431–436.
4 Frankenfield D, Reynolds H: Nutritional effect of continuous hemodiafiltration. Nutrition
1995;11:388–393.
5 Scheinkestel CD, Adams F, Mahony L, et al: Impact of increasing parenteral protein loads on
amino acid levels and balance in critically ill anuric patients on continuous renal replacement ther-
apy. Nutrition 2003;19:733–740.
6 Scheinkestel CD, Kar L, Marshall K, et al: Prospective randomized trial to assess caloric and pro-
tein needs of critically Ill, anuric, ventilated patients requiring continuous renal replacement ther-
apy. Nutrition 2003;19:909–916.
7 Bollmann MD, Revelly JP, Tappy L, et al: Effect of bicarbonate and lactate buffer on glucose and
lactate metabolism during hemodiafiltration in patients with multiple organ failure. Intensive Care
Med 2004;30:1103–1110.
8 Berger M, Shenkin A: Update on clinical micronutrient supplementation studies in the critically
ill. Curr Opin Clin Nutr Metab Care 2006;109:711–716.
9 Angstwurm M, Schottdorf J, Schopohl J, Gaertner R: Selenium replacement in patients with
severe systemic inflammatory response syndrome improves clinical outcome. Crit Care Med
1999;27:1807–1813.
10 Richard M, Ducros V, Foret M, et al: Reversal of selenium and zinc deficiencies in chronic
hemodialysis patients by intravenous sodium selenite and zinc gluconate supplementation – time-
course of glutathione peroxidase repletion and lipid peroxidation decrease. Biol Trace Elem Res
1993;39:149–159.
11 Story D, Ronco C, Bellomo R: Trace element and vitamin concentrations and losses in critically
ill patients treated with continuous venovenous hemofiltration. Crit Care Med 1999;27:
220–223.
12 Berger M, Shenkin A, Revelly J, et al: Copper, selenium, zinc, and thiamine balances during
continuous venovenous hemodiafiltration in critically ill patients. Am J Clin Nutr 2004;80:
410–416.
13 Rubinson L, Diette GB, Song X, Brower RG, Krishnan JA: Low caloric intake is associated with
nosocomial bloodstream infections in patients in the medical intensive care unit. Crit Care Med
2004;32:350–357.
14 Heyland DK, Dhaliwal R, Day A, Jain M, Drover J: Validation of the Canadian clinical practice
guidelines for nutrition support in mechanically ventilated, critically ill adult patients: results of a
prospective observational study. Crit Care Med 2004;32:2260–2266.
15 Metnitz P, Krenn C, Steltzer H, et al: Effect of acute renal failure requiring renal replacement ther-
apy on outcome in critically ill patients. Crit Care Med 2002;30:2051–2058.
16 Cano N, Fiaccadori E, Tesinsky P, et al: ESPEN Guidelines on Enteral Nutrition: adult renal fail-
ure. Clin Nutr 2006;25:295–310.
17 Bellomo R, Tan H, Bhonagiri S, et al: High protein intake during continuous hemodiafiltration:
impact on amino acids and nitrogen balance. Int J Artif Organs 2002;25:261–268.
Chioléro/Berger 274
18 Van den Berghe G, Wouters P, Weekers F, et al: Intensive insulin therapy in critically ill patients.
N Engl J Med 2001;345:1359–1367.
19 Van den Berghe G, Wilmer A, Hermans G, et al: Intensive insulin therapy in the medical ICU.
N Engl J Med 2006;354:449–461.
René Chioléro, MD
Department of Intensive Care Medicine, CHUV, BH 08.610
CH–1011 Lausanne (Switzerland)
Tel. ⫹41 21 314 20 02, Fax ⫹41 21 314 30 45, E-Mail rene.chiolero@chuv.ch
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 275–286
Vascular Access for HD and CRRT
Miet Schetz
Department of Intensive Care Medicine, University Hospital Gasthuisberg,
Leuven, Belgium
Abstract
A good functioning vascular access is an essential component for adequate renal
replacement therapy (RRT) in acute kidney injury. Tunneled, cuffed catheters are preferred if
the anticipated duration of RRT is more than 3 weeks. The right jugular vein is the preferred
insertion site for the temporary dialysis catheter (TDC), with ultrasound-guided insertion
reducing the risk of mechanical complications. The femoral vein is the second choice,
whereas the subclavian vein should be avoided. The most important complications of a TDC
are acute malfunction and infection. Intraluminal thrombosis, fibrin sleeve formation, mal-
positioning and kinking result in acute malfunction. Recirculation can be reduced by correct
placement of the catheter and is more an issue for intermittent hemodialysis than for contin-
uous RRT. Strict adherence to simple preventive strategies reduces catheter-related blood-
stream infection. In selected patients more sophisticated strategies such as the use of
antibiotic/antiseptic impregnated catheters and antibiotic/antiseptic lock solutions may be
useful.
Copyright © 2007 S. Karger AG, Basel
The delivery of an adequate dialysis dose is important for survival both in
acute and chronic renal failure [1–3]. An essential component to perform this
treatment is a good functioning vascular access, the ideal access being able to
deliver an adequate blood flow without associated complications. Whereas in
chronic dialysis patients a native forearm arteriovenous fistula is the optimal
access [4], renal replacement therapy (RRT) for acute kidney injury (AKI) is
mostly performed with a temporary dialysis catheter (TDC), which is easily
inserted at the bedside and immediately usable. The importance of an adequate
access is illustrated by the fact that lower than prescribed blood flows contribute
to inadequate dialysis doses during intermittent hemodialysis in patients with
AKI [5]. In continuous RRT vascular access is a major determinant of filter
life span, with access malfunction contributing to 3–25% of filter exchanges
Schetz 276
[6–10]. Literature data on vascular access for acute dialysis are relatively
scarce. Studies on TDCs are mostly performed in chronic dialysis patients,
whereas studies in critically ill patients are mostly on central venous catheters
(CVCs). TDCs have a different design than CVCs and are manipulated differ-
ently, and chronic dialysis patients differ in many respects from critically ill
patients, raising the question whether results from these studies can be extra-
polated to dialysis catheters in critically ill patients.
Catheter Characteristics
The material of TDCs should be sufficiently rigid to allow insertion and
maintain lumen patency, it should be flexible to prevent kinking, thromboresis-
tant and resistant to bacterial invasion. Modern TDCs are made of polyurethane
or silicone which are less thrombogenic than older materials [11]. Polyurethane
catheters are semi-rigid and easier to insert with the Seldinger technique but are
associated with a higher risk of endothelial damage or venous perforation.
Polyurethane has thermoplastic properties, being rigid during placement but
softening when soaked at body temperature. Silicone is mostly used for tun-
neled catheters that are soft and flexible, but are more difficult to insert and are
associated with more mechanical failure due to compression of the lumen. On
the other hand the danger of perforation is lower. These catheters can therefore
be placed in the right atrium, allowing better blood flow and a reducing the risk
of recirculation and malfunction [12, 13].
Although some centers use two separate single lumen catheters inserted
into two different large veins or into the same vein, most acute RRT is per-
formed with a double-lumen catheter inserted into a large vein. The outer diam-
eter of the double-lumen TDC varies between 11 and 14 french. The two lumens
can be arranged side-by-side (double-D or double-O configuration) or can be
concentric (coaxial). The arterial port ends about 3 cm proximal to the venous
port. The optimal length depends on the location of insertion: 12–15 cm for the
right internal jugular vein; 15–20 cm for the left internal jugular vein, and
19–24 cm for the femoral vein [4, 14]. Shorter catheters at the femoral site do
not reach the inferior vena cava and are associated with more malfunction and
recirculation and consequently reduced dialysis efficiency [15, 16]. Some jugu-
lar catheters have a curved extension allowing proper fixation and increasing
the comfort of the patient.
For short-term RRT non-tunneled non-cuffed catheters can be used, but
tunneled (5–10 cm subcutaneous course) and cuffed (fixed to the tissue by a
Dacron or silver impregnated collagen cuff) catheters should be used if the
duration of RRT is anticipated to be more than 3 weeks [4, 17].
Vascular Access 277
Where to Insert a Dialysis Catheter?
The choice of the appropriate insertion site depends on the patient’s char-
acteristics (coagulopathy, morbid obesity, previous surgery, local infection,
altered local anatomy, cardiopulmonary reserve capacity, prolonged immobi-
lization), the availability of the insertion site in the often ‘heavily catheterized’
critically ill patients, the operator’s skills and experience, and the risk of com-
plications associated with the different insertion sites.
In order to prevent atrial perforation, semi-rigid catheters in the superior
vena cava (SVC) should have their tip in the SVC or at the junction of the SVC
and the right atrium. Prior to their use the correct position should be verified
with a chest X-ray [4]. Silicone catheters function optimally with their tip in
the right atrium [12, 13]. Insertion into the right internal jugular vein is pre-
ferred for both tunneled and non-tunneled catheters because it is a more direct
route to the caval-atrial junction than a left-sided placement that results in a
lower blood flow and more complications [4, 18]. With left-sided insertion, the
access or removal lumen should be inside the catheter curve to prevent
obstruction by sucking of the vessel wall. In critically ill patients the subcla-
vian vein is the preferred insertion site for a CVC because it is associated with
the lowest risk of infectious complications [17]. However, in a patient who
may need permanent vascular access for dialysis, the use of the subclavian
vein is discouraged for insertion of a TDC because it is associated with higher
rates of central venous stenosis, precluding the ipsilateral arm for future dialy-
sis access [19–21]. The subclavian vein should therefore be reserved for very
short-term use or if there is no other alternative [4, 17]. The femoral vein is
often preferred for patients with critical respiratory conditions or for urgent
access because of the speed with which it can be performed. On the other hand,
femoral catheters are associated with the highest risk of infection, both for
CVC in critically ill patients [22] and for non-tunneled, non-cuffed TDC [23, 24].
In addition, femoral TDCs reduce patient mobilization and result in higher
recirculation rates [15, 16].
Catheter Insertion and Care
Compared with a non-tunneled, non-cuffed catheter, the insertion of a tun-
neled cuffed catheter has a higher failure rate, requires more skill and time, and
is associated with more tissue trauma. On the other hand, compared with non-
cuffed catheters, cuffed catheters are less susceptible to infection (less extralu-
minal contamination). The mean incidence of catheter-related bloodstream
infection (CRBSI) is ⫾5 per 1,000 catheter days for non-cuffed non-tunneled
Schetz 278
dialysis catheters versus 1.6–3.5 for tunneled and cuffed catheters [25, 26].
According to the NF/KDOQI guidelines, non-cuffed dialysis catheters should
therefore have a finite use-life, not exceeding 3 weeks for internal jugular or
subclavian catheters and 5 days for femoral catheters [4]. A recent randomized
comparison of tunneled and non-tunneled femoral catheters in AKI patients
showed a lower incidence of vein thrombosis and catheter-related infection
and a higher blood flow with greater delivered dialysis dose for the tunneled
catheters [27].
A meta-analysis of 18 prospective randomized trials showed that, com-
pared with the landmark method (blind insertion), the use of ultrasound guid-
ance for insertion of a CVC was associated with a significantly lower failure
rate for cannulation of the internal jugular vein, whereas the effect was limited
for the subclavian and femoral vein [28]. In addition, a recent prospective ran-
domized trial in 900 patients not only showed an increased overall success rate
but also a reduction in hemothorax, pneumothorax and catheter-related infec-
tion with ultrasound guidance for internal jugular vein catheterization [29]. Use
of ultrasound guidance is therefore advocated both for placement of CVCs [17]
and for (tunneled and non-tunneled) TDCs [4].
Maximal sterile barrier precautions (surgical scrub, sterile gloves, long-
sleeved sterile gowns, mask, cap and large sterile sheet drapes) reduce the
rate of infection. Skin disinfection should preferably be performed with 2%
chlorhexidine, but a tincture of iodine or 70% alcohol is also acceptable. For
CVC the application of an antibiotic ointment is not advocated because it pro-
motes fungal infections and antimicrobial resistance. However, in chronic dial-
ysis patients this procedure has been shown to be effective [4, 17]. Whether it
should be applied in critically ill patients with AKI is not clear. It should be
borne in mind that some of these ointments may adversely affect the integrity of
dialysis catheters [30]. The catheter exit site should be covered with a sterile dry
gauze (changed every 2 days) or transparent dressing (changed every 7 days)
[17].
Complications
Insertion-related complications include inadvertent arterial puncture,
hemothorax, pneumothorax, pericardial tamponade, arrhythmias, air embolism
and retroperitoneal hemorrhage [11, 14]. Early mechanical complications can
be reduced by insertion under ultrasound guidance [29]. In addition, routine
chest X-ray control allows early detection of hemo- and pneumothorax. Late
complications include infection, central vein thrombosis/stenosis, and catheter
dysfunction.
Vascular Access 279
Infection
Intravascular catheters put patients at risk of infection. Exit site infection is
defined as localized infection of the skin and soft tissue around the exit site,
often without systemic signs of infection. In ICU patients CRBSIs are associ-
ated with increased morbidity, mortality and duration of hospitalization and
additional medical costs [31]. Microorganisms that most commonly cause CRBSI
include (in decreasing order of frequency): coagulase-negative staphylococci,
Staphylococcus aureus, enterococci, gram-negative bacteria and yeasts. A sub-
set of patients with CRBSI develop metastatic infection including endocarditis,
thrombophlebitis, septic arthritis, osteomyelitis and epidural abcesses, espe-
cially when S. aureus or Candida albicans is the causative organism [32].
A recent systematic review showed an average rate for CVC-associated blood-
stream infection of 2.7 per 1,000 catheter days for non-tunneled and 1.7 for tun-
neled catheters [25]. For dialysis catheters the incidence is ⫾5 per 1,000
catheter days for untunneled TDCs and 1.6–3.5 for tunneled TDCs [25, 26]. In
ICU patients with AKI colonization rates for TDCs do not appear to be signifi-
cantly higher than for CVCs (9.1 vs. 5.9 per 1,000 catheter days) [33].
Different mechanisms may lead to catheter-related infections. Extraluminal
contamination results from the migration of skin flora along the external sur-
face of the catheter into the bloodstream, is the most common cause of CRBSI
of short-term percutaneously inserted non-cuffed CVCs [34], and can be
reduced by the use of cuffed and tunneled catheters. A less common cause of
extraluminal colonization is hematogenous seeding from another focus of
infection. Intraluminal colonization is the dominant mechanism in longer-
dwelling catheters and mostly results from contamination of the catheter hub,
whereas infusate contamination is rare. A bacterial biofilm forms rapidly in the
lumen of most CVCs and is the major source of both catheter-related bac-
teremia and thrombosis [35]. In ICU patients risk factors for CRBSI include:
catheter material (more thrombogenic catheters are associated with high infec-
tion rates), the number of infusion ports, the frequency of manipulation, urgent
versus elective insertion, the operator’s experience (traumatic insertion proce-
dures increase the risk of infection), insertion site (see above), indwelling time,
and the patient’s illness severity [11, 36]. In chronic dialysis patients nasal car-
riage of S. aureus, previous bacteremia, peripheral atherosclerosis, diabetes,
site of insertion, indwelling time, and the number of dialyses performed have
been identified as risk factors [23, 37, 38].
Catheter colonization can be defined as the presence of ⱖ15 colony-forming
units (CFU) on semiquantitative (roll plate) or ⱖ100 CFU on quantitative (vor-
tex or sonication method) catheter culture in the absence of clinical signs of
infection. CRBSI should be suspected when a patient with a catheter devel-
ops fever or chills and does not have clinical evidence for another source of
Schetz 280
infection. Several methods can be used to diagnose CRBSI. Methods requiring
device removal include: (1) a qualitative catheter segment culture; (2) a semi-
quantitative catheter segment culture (growth of ⱖ15 CFU), or (3) quantitative
catheter segment culture (ⱖ1,000 CFU), all of them combined with a positive
blood culture yielding the same microorganism. Methods not requiring catheter
removal include: (1) qualitative blood culture through the device; (2) quantita-
tive blood culture through the device (ⱖ100 CFU/ml); (3) paired quantitative
blood culture (3–5 times more organisms grown in blood cultures drawn
through the catheter than in a blood culture drawn peripherally); (4) differential
time to positivity (blood culture drawn through the catheter positive ⱖ2h ear-
lier than blood culture drawn peripherally), and (5) acridine orange leukocyte
cytospin (visualization of any organism). Paired blood culture is the most accu-
rate estimate. However most other methods show acceptable sensitivity and
specificity [39]. In the absence of concurrent blood cultures from a peripheral
vein, a clinical diagnosis of catheter-related bacteremia requires the exclusion
of alternate sources of infection. If S. aureus or C. albicans are the causative
organism or in the case of persistent bacteremia after catheter removal, aggres-
sive evaluation for metastatic infections, including a transesophageal echocar-
diography, should be performed [32].
Guidelines for the prevention of intravascular CRBSIs were published in
2002 [17] and include: education and training of healthcare workers, surveil-
lance of catheter-related infections, adequate hand hygiene, aseptic technique
during catheter insertion and care and selection of the catheter, the insertion site
and insertion technique with the lowest risk of complications for the anticipated
time and duration of catheterization. CVCs should not be replaced routinely,
and catheters should be removed when no longer needed. For hemodialysis,
cuffed and tunneled catheters are indicated if the period of temporary access is
anticipated to be prolonged (⬎3 weeks). TDCs should be used exclusively for
RRT. Strict adherence to these evidence-based catheter insertion and mainte-
nance policies reduces CRBSI [40–42].
More sophisticated preventive measures include the use of antibiotic-
impregnated catheters and antibiotic lock solutions. Antimicrobial impregnated
CVCs (chlorhexidine/silver sulfadiazine or minocycline/rifampin) are indicated
if the anticipated duration is more than 5 days and if, after implementing a com-
prehensive strategy to reduce rates of CRBSI, the rate of infection remains too
high [17]. The advantage of antibiotic impregnation has also been shown for
TDCs [43]. On the other hand, allergic reactions and the emergence of resis-
tance remain an important concern when antimicrobial-impregnated catheters
are used. Together with the fear of systemic toxicity, the concern about anti-
microbial resistance also applies to the prophylactic use of antibiotic/antiseptic
locks (combinations of gentamicin, taurolidine, isopropyl alcohol, minocycline
Vascular Access 281
or a cephalosporin with heparin, citrate or EDTA), which have been shown to
reduce the incidence of CRBSI in chronic dialysis patients [44]. A recent ran-
domized trial demonstrated that, compared with heparin, the use of a citrate
lock reduces catheter-related infections [45]. With other long-term central
venous access devices, vancomycin locks have been shown to reduce the risk of
CRBSI and may be useful in high-risk patients [46]. Another strategy consists
of using hubs that contain an antiseptic chamber [47].
Empiric systemic antibiotics appropriate for the suspected organisms should
be started on clinical suspicion of CRBSI after cultures have been taken. In
centers with a high prevalence of methicillin-resistant staphylococci this empiric
regimen should include vancomycin. In other centers a penicillinase-resistant
penicillin should be used. Whether broader coverage (with e.g. a third- or fourth-
generation cephalosporin) is required depends on the severity of sepsis, the
patient’s immune system and known previous infections, and on the spectrum of
microorganisms in the unit. As soon as the culture results are available, antibiotic
treatment should be tailored to the specific organism. The duration of antibiotic
treatment depends on the infecting organism (shorter for coagulase-negative
staphylococci), on the presence of an immunocompromised state, valvular heart
disease, intravascular prosthetic devices or metastatic infections, and whether or
not the catheter has been removed [32].
When infection of the catheter is suspected, the catheter can be exchanged
over a guide-wire, with further insertion at a new site if the culture of the
catheter tip is found to be positive. Exit-site infection, pocket infection of tun-
neled catheters and CRBSI with clinical signs of sepsis are an indication for
immediate catheter removal. A new non-tunneled catheter may be inserted after
antibiotic treatment has started, whereas reinsertion of a tunneled catheter
should ideally be postponed until after completion of the antibiotic treatment
[32]. When symptoms of CRBSI are mild without suspicion of metastatic com-
plications, especially when Staphylocccus epidermidis is the causative organ-
ism, attempts can be undertaken to salvage the catheter in patients with
tunneled catheters and limited access sites. If fever does not resolve over the
first 24–48 h the catheter should still be removed. If fever resolves the sus-
pected catheter may be exchanged over a guide-wire [48–50]. Another method
allowing catheter salvage is the use of an ‘antibiotic lock’ or the instillation of a
concentrated antibiotic solution (combined with an anticoagulant) into the
catheter [44]. This method is mostly reserved for long-term catheters such as
tunneled dialysis catheters or long-term devices used in oncology or chronic
parenteral nutrition. It allows obtaining local antibiotic concentrations that are
several orders of magnitude higher than the blood concentration and is a solu-
tion for the inability of therapeutic concentrations of most antibiotics to kill
microorganisms growing in a biofilm.
Schetz 282
Acute Catheter Malfunction
Acute catheter malfunction may result from complete or partial intralumi-
nal thrombosis, from a fibrin sheet around the catheter or from malpositioning
(sucking the vein wall) and kinking of the catheter, and is indicated by a
decrease in the attainable blood flow and increased arterial and/or venous pres-
sures during hemodialysis. For continuous RRT blood flows between 150 and
200 ml/min are able to deliver an adequate dialysis dose. However, intermittent
hemodialysis mostly requires higher blood flows (at least 250–300 ml/min).
These flows can only be delivered with an adequately functioning access.
In order to prevent intraluminal thrombosis the catheter is filled with an
anticoagulant (heparin or citrate) during the interdialytic period. Endoluminal
thrombosis can be treated with mechanical (brush) or chemical (thrombolytics)
methods, but this is mostly reserved for tunneled catheters [51, 52].
Improper catheter tip placement is a common cause of access malfunction.
Femoral catheters should be in the inferior vena cava, and jugular and subcla-
vian catheters should be at the junction of the SVC and the right atrium.
Malfunction of catheters in the SVC is further decreased when their tip is
located in the right atrium [12, 13, 53], which, however, is only safe with sili-
cone catheters. For non-tunneled, non-cuffed catheters malfunction is more fre-
quent with femoral compared with jugular catheters [54].
Access recirculation (recirculation of blood from the outflow to the inflow
part of the catheter) occurs whenever pumped flow through the extracorporeal
circuit exceeds flow in the vein, reduces the effective clearance, and depends on
the design, length and insertion site of the catheter and on the blood flow. Well-
functioning and nonreversed internal jugular and subclavian venous catheters
have, in general, recirculation rates of ⬍5%. Because of the higher blood flows,
recirculation is more important with intermittent than with continuous RRT.
Femoral catheters result in higher recirculation rates than subclavian or jugular
catheters, especially when they are too short [15, 16]. For catheters in the SVC
recirculation can be minimized by placement close to the right atrium. Inversion
of the connecting lines also increases recirculation from 3 to 12% [55]. Two
observational trials report contradictory results with regard to the delivered dial-
ysis dose when comparing femoral and internal jugular access [56, 57].
Central Vein Stenosis and Thrombosis
Catheter-related thrombosis can manifest itself as the formation of a fibrin
sleeve around the catheter or a thrombus adherent to the vessel wall. The inci-
dence may be as high as 33–67% and is largely dependent on the diagnostic
method. Risk factors for catheter-related thrombosis are the insertion site (vein
diameter, local hemodynamics), technical problems during insertion, the catheter
material, the indwelling time, and the hypercoagulability or increased blood
Vascular Access 283
viscosity of the patient [11]. In chronic dialysis patients, catheter-related central
venous stenosis and thrombosis are most frequent in the subclavian vein
[19–21], although incidences as high as 25% have also been reported for the
internal jugular vein [58] and the femoral vein [59]. In AKI patients treated
with femoral catheters the incidence appears to be lower with a tunneled com-
pared with a non-tunneled catheter [27]. In ICU patients the thrombotic risk
associated with CVC is higher for femoral [60] and jugular [61] than subclavian
catheters. The administration of anticoagulants [62], and also the use of less
thrombogenic material or anticoagulant-bonded catheters [63, 64] has been
shown to decrease the risk of catheter-related thrombosis and subsequently the
risk of catheter-related infection.
References
1 NF/KDOQI clinical practice guidelines for hemodialysis adequacy: update 2000. Am J Kidney Dis
2001;37(suppl 1):S7–S164.
2 Ronco C, Bellomo R, Homel P, Brendolan A, Dan M, Piccinni P, La Greca G: Effects of different
doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective
randomised trial: Lancet 2000;356:26–30.
3 Schiffl H, Lang SM, Fischer R: Daily hemodialysis and the outcome of acute renal failure. N Engl
J Med 2002;346:305–310.
4 NF/KDOQI clinical practice guidelines for vascular access: update 2000. Am J Kidney Dis
2001;37(suppl 1):S137–S181.
5 Evanson JA, Himmelfarb J, Wingard R, Knights S, Shyr Y, Schulman G, Ikizler TA, Hakim RM:
Prescribed versus delivered dialysis in acute renal failure patients. Am J Kidney Dis 1998;32:
731–738.
6 Monchi M, Berghmans D, Ledoux D, Canivet JL, Dubois B, Damas P: Citrate vs. heparin for anti-
coagulation in continuous venovenous hemofiltration: a prospective randomized study. Intensive
Care Med 2004;30:260–265.
7 Kutsogiannis DJ, Gibney RT, Stollery D, Gao J: Regional citrate versus systemic heparin anti-
coagulation for continuous renal replacement in critically ill patients. Kidney Int 2005;67:
2361–2367.
8 Palsson R, Laliberte KA, Niles JL: Choice of replacement solution and anticoagulant in continu-
ous venovenous hemofiltration. Clin Nephrol 2006;65:34–42.
9 Mehta RL, McDonald BR, Ward DM: Regional citrate anticoagulation for continuous arteriove-
nous hemodialysis. An update after 12 months; in Sieberth HG, Mann H, Stummvoll HK (eds):
Continuous Hemofiltration. Contrib Nephrol. Basel, Karger, 1991, vol 93, pp 210–214.
10 Holt AW, Bierer P, Bersten AD, Bury LK, Vedig AE: Continuous renal replacement therapy in crit-
ically ill patients: monitoring circuit function. Anaesth Intensive Care 1996;24:423–429.
11 Polderman KH, Girbes AJ: Central venous catheter use. Part 1: mechanical complications.
Intensive Care Med 2002;28:1–17.
12 Jean G, Chazot C, Vanel T, Charra B, Terrat JC, Calemard E, Laurent G: Central venous catheters
for haemodialysis: looking for optimal blood flow. Nephrol Dial Transplant 1997;12:1689–1691.
13 Petersen J, Delaney JH, Brakstad MT, Rowbotham RK, Bagley CM Jr: Silicone venous access
devices positioned with their tips high in the superior vena cava are more likely to malfunction.
Am J Surg 1999;178:38–41.
14 Oliver MJ: Acute dialysis catheters. Semin Dial 2001;14:432–435.
15 Leblanc M, Fedak S, Mokris G, Paganini EP: Blood recirculation in temporary central catheters
for acute hemodialysis. Clin Nephrol 1996;45:315–319.
Schetz 284
16 Little MA, Conlon PJ, Walshe JJ: Access recirculation in temporary hemodialysis catheters as
measured by the saline dilution technique. Am J Kidney Dis 2000;36:1135–1139.
17 O’Grady N, Alexander M, Dellinger P, Gerberding JL, Heard SO, Maki DG, Masur H,
McCormick RD, Mermel LA, Pearson ML, Raad II, Randolph A, Weinstein RA; the Healthcare
Infection Control Practices Advisory Committee: Guidelines for the prevention of intravascular
catheter-related infections. Infect Control Hosp Epidemiol 2002;23:759–769.
18 Oliver MJ, Edwards LJ, Treleaven DJ, Lambert K, Margetts PJ: Randomized study of temporary
hemodialysis catheters. Int J Artif Organs 2002;25:40–44.
19 Ciminowski GE, Worley E, Rutherford WE, Sartain J, Blondin J, Harter H: Superiority of the
internal jugular over the subclavian access for temporary dialysis. Nephron 1990;54:154–161.
20 Schillinger F, Schillinger D, Montagnac R, Milcent T: Post catheterisation vein stenosis in
haemodialysis: comparative angiographic study of 50 subclavian and 50 internal jugular accesses.
Nephrol Dial Transpl 1991;6:722–724.
21 Trerotola SO, Kuhn-Fulton J, Johnson MS, Shah H, Ambrosius WT, Kneebone PH: Tunneled infu-
sion catheters: increased incidence of symptomatic venous thrombosis after subclavian versus
internal jugular venous access. Radiology 2000;217:89–93.
22 Lorente L, Henry C, Martin MM, Jiménez A, Mora ML: Central venous catheter-related infection
in a prospective and observational study of 2595 catheters. Crit Care 2005;9:R631–R635.
23 Oliver MJ, Callery SM, Thorpe KE, Schwab SJ, Churchill DN: Risk of bacteremia from temporary
hemodialysis catheters by site of insertion and duration of use: a prospective study. Kidney Int
2000;58:2543–2545.
24 Kairaitis LK, Gottlieb T: Outcome and complications of temporary dialysis catheters. Nephrol
Dial Transplant 1999;14:1710–1714.
25 Maki DG, Kluger DM, Crinch C: The risk of bloodstream infection in adults with different
intravascular devices: a systematic review of 200 published prospective studies. Mayo Clin Proc
2006;81:1159–1171.
26 Saxena AK, Panhotra BR: Haemodialysis catheter-related infections: current treatment options
and strategies for prevention. Swiss Med Wkly 2005;135:127–138.
27 Klouche K, Amigues L, Deleuze S, Beraud JJ, Canaud B: Complications, effects on dialysis dose,
and survival of tunneled femoral dialysis catheters in acute renal failure. Am J Kidney Dis 2007;49:
99–108.
28 Hind D, Calvert N, Mc Williams R, Davidson A, Paisley S, Beverley C: Ultrasonic locating
devices for central venous cannulation: meta-analysis. BMJ 2003;327:361–367.
29 Karakitsos D, Labropoulos N, De Groot E, Patrianakos AP, Kouraklis G, Poularas J, Samonis G,
Tsoutsos DA, Konstadoulakis MM, Karabinis A: Real-time ultrasound-guided catheterisation of
the internal jugular vein: a prospective comparison with the landmark technique in critical care
patients. Crit Care 2006;10:R162–R169.
30 Riu S, Ruiz CG, Martinez-Vea A, Peralta C, Oliver CA: Spontaneous rupture of polyurethane peri-
toneal catheter: a possible deleterious effect of mupirocin ointment. Nephrol Dial Transplant
1998;13:1870–1871.
31 Pittet D, Tamara D, Wenzel RP: Nosocomial bloodstream infection in critically ill patients: excess
length of stay, extra costs and attributable mortality. JAMA 1994;271:1598–1601.
32 Mermel LA, Farr BM, Sherertz RJ, raad II, O’Grady N, Harris JS, Craven DE: Guidelines for the
management of intravascular catheter-related infections. Clin Infect Dis 2001;32:1249–1272.
33 Souweine B, Liotier J, Heng AE, Isnard M, Ackoundou-N’Guessan C, Deteix P, Traore O:
Catheter colonization in acute renal failure patients: comparison of central venous and dialysis
catheters. Am J Kidney Dis 2006;47:879–887.
34 Safdar N, Maki DG: The pathogenesis of catheter-related bloodstream infection with noncuffed
short-term central venous catheters. Intensive Care Med 2004;30:62–67.
35 Costerton JW, Stewart PS, Greenberg EP: Bacterial biofilms: a common cause of persistent infec-
tions. Science 1999;284:1318–1322.
36 Polderman KH, Girbes AR: Central venous catheter use. Part 2: infectious complications.
Intensive Care Med 2002;28:18–28.
37 Jean G, Charra B, Chazot C, Vanel T, Terrat JC, Hurot JM, Laurent G: Risk factor analysis for
long-term tunneled dialysis catheter-related bacteremias. Nephron 2002;91:399–405.
Vascular Access 285
38 Naumovic RT, Jovanovic DB, Djukanovic LJ: Temporary vascular catheters for hemodialysis: a
3-year prospective study. Int J Artif Organs 2004;27:848–854.
39 Safdar N, Fine JP, Maki DG: Meta-analysis: methods for diagnosing intravascular device-related
bloodstream infection. Ann Intern Med 2005;142:451–466.
40 Pronovost P, Needham D, Berenholtz S, Sinopoli D, Chu H, Cosgrove S, Sexton B, Hyzy R, Welsh R,
Roth G, Bander J, Kepros J, Goeschel C: An intervention to decrease catheter-related bloodstream
infections in the ICU. N Engl J Med 2006;355:2725–2732.
41 Warren DK, Cosgrove SE, Diekema DJ, Zuccotti G, Climo MW, Bolon MK, Tokars JI, Noskin GA,
Wong ES, Sepkowitz KA, Herwaldt LA, Perl TM, Solomon SL, Fraser VJ: Prevention Epicenter
Program. A multicenter intervention to prevent catheter-associated bloodstream infections. Infect
Control Hosp Epidemiol 2006;27:662–669.
42 Young EM, Commiskey ML, Wilson SJ: Translating evidence into practice to prevent central
venous catheter-associated bloodstream infections: a systems-based intervention. Am J Infect
Control 2006;34:503–506.
43 Chatzinikolaou I, Finkel K, Hanna H, Boktour M, Foringer J, Ho T, Raad I: Antibiotic-coated
hemodialysis catheters for the prevention of vascular catheter-related infections: a prospective,
randomized study. Am J Med 2003;115:352–357.
44 Manierski C, Besarab A: Antimicrobial locks: putting the lock on catheter infections. Adv Chronic
Kidney Dis 2006;13:245–258.
45 Weijmer MC, van den Dorpel MA, Van de Ven PJ, ter Wee PM, van Geelen JA, Groeneveld JO,
van Jaarsveld BC, Koopmans MG, le Poole CY, Schrander-Van der Meer AM, Siegert CE, Stas KJ;
CITRATE Study Group: Randomized, clinical trial comparison of trisodium citrate 30% and heparin
as catheter-locking solution in hemodialysis patients. J Am Soc Nephrol 2005;16: 2769–2777.
46 Safdar N, Maki DG: Use of vancomycin-containing lock or flush solutions for prevention of
bloodstream infection associated with central venous access devices: a meta-analysis of prospec-
tive, randomized trials. Clin Infect Dis 2006;43:474–484.
47 Leon C, Alvarez-Lerma F, Ruiz-Santana S, Gonzalez V, de la Torre MV, Sierra R, Leon M,
Rodrigo JJ: Antiseptic chamber-containing hub reduces central venous catheter-related infection:
a prospective, randomized study. Crit Care Med 2003;31:1318–1324.
48 Robinson D, Suhocki P, Schwab SJ: Treatment of infected tunneled venous access hemodialysis
catheters with guidewire exchange. Kidney Int 1998;53:1792–1794.
49 Beathard GA: Management of bacteremia associated with tunneled-cuffed hemodialysis catheters.
J Am Soc Nephrol 1999;10:1045–1049.
50 Tanriover B, Carlton D, Saddekni S, Hamrick K, Oser R, Westfall AO, Allon M: Bacteremia asso-
ciated with tunneled dialysis catheters: comparison of two treatment strategies. Kidney Int 2000;57:
2151–2155.
51 Clase CM, Crowther MA, Ingram AJ, Cina CS: Thrombolysis for restoration of patency to
haemodialysis central venous catheters: a systematic review. J Thromb Thrombolysis 2001;11:
127–136.
52 Hilleman DE, Dunlay RW, Packard KA: Reteplase for dysfunctional hemodialysis catheter clear-
ance. Pharmacotherapy 2003;23:137–141.
53 Abidi SM, Khan A, Fried LF, Chelluri L, Bowles S, Greenberg A: Factors influencing function of
temporary dialysis catheters. Clin Nephrol 2000;53:199–205.
54 Hryszko T, Brzosko S, Mazerska M, Malyszko J, Mysliwiec M: Risk factors of nontunneled non-
cuffed hemodialysis catheter malfunction. A prospective study. Nephron Clin Pract 2004;96:
c43–c47.
55 Level C, Lasseur C, Chauveau P, Bonarek H, Perrault L, Combe C: Performance of twin central
venous catheters: influence of the inversion of inlet and outlet on recirculation. Blood Purif 2002;
20:182–188.
56 Liangos O, Sakiewicz PG, Kanagasundaram NS, Hammel J, Pajouh M, Seifert T, Paganini EP:
Dialyzer fiber bundle volume and kinetics of solute removal in continuous venovenous hemodial-
ysis. Am J Kidney Dis 2002;39:1047–1053.
57 du Cheyron D, Bouchet B, Bruel C, Daubin C, Ramakers M, Charbonneau P: Antithrombin sup-
plementation for anticoagulation during continuous hemofiltration in critically ill patients with
septic shock: a case-control study. Crit Care 2006;10:R45.
Schetz 286
58 Wilkin TD, Kraus MA, Lane KA, Trerotola SO: Internal jugular vein thrombosis associated with
hemodialysis catheters. Radiology 2003;228:697–700.
59 Maya ID, Allon M: Outcomes of tunneled femoral hemodialysis catheters: comparison with inter-
nal jugular vein catheters. Kidney Int 2005;68:2886–2889.
60 Merrer J, De Joughe B, Golliot F, lefrant JY, Raffy B, barre E, Rigaud JP, Casciani D, Misset B,
Bosquet C, Outin H, Brun-Buisson C, Nitenberg G: Complications of femoral and subclavian
venous catheterisation in critically ill patients. JAMA 2001;286:700–707.
61 Timsit JF, Farkas JC, Boyer JM, Martin JB, Misset B, Renaud B, Carlet J: Central vein catheter-
related thrombosis in intensive care patients: incidence, risk factors and relationship with catheter-
related sepsis. Chest 1998;114:207–213.
62 Randolph AG, Cook DJ, Gonzales CA, Andrew M: Benefit of heparin in central venous and pul-
monary artery catheters: a meta-analysis of randomized controlled trials. Chest 1998;113:165–171.
63 Baumann M, Witzke O, Dietrich R, Haug U, Deppisch R, Lutz J, Philipp T, Heemann U:
Prolonged catheter survival in intermittent hemodialysis using a less thrombogenic micropat-
terned polymer modification. ASAIO J 2003;49:708–712.
64 Long DA, Coulthard MG: Effect of heparin-bonded central venous catheters on the incidence of
catheter-related thrombosis and infection in children and adults. Anaesth Intensive Care 2006;34:
481–484.
Miet Schetz, MD, PhD
Department of Intensive Care Medicine, University Hospital Gasthuisberg
Herestraat 49
BE–3000 Leuven (Belgium)
Tel. ⫹32 16 344 021, Fax ⫹32 16 344 015, E-Mail marie.schetz@uz.kuleuven.ac.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 287–296
Dialysate and Replacement Fluid
Composition for CRRT
Filippo Aucella, Salvatore Di Paolo, Loreto Gesualdo
Department of Nephrology, Dialysis and Transplantation, University of Foggia,
Foggia, Italy
Abstract
Continuous renal replacement therapies (CRRTs) are increasingly used in order to
maintain normal or near-normal acid-base balance in intensive care unit (ICU) patients.
Acid-base balance is greatly influenced by the type of dialysis employed and by the adminis-
tration route of replacement fluids. In continuous veno-venous hemofiltration, buffer balance
depends on losses with ultrafiltrate and gain with replacement fluid, while in techniques such
as continuous veno-venous hemodiafiltration, clinicians should balance the role of the
dialysate. The type of buffer greatly influences not only acid-base correction, but also clini-
cal outcome. Lactate or bicarbonate fluids are currently used, but recent studies suggest that
bicarbonate-buffered replacement fluids can improve acid-base status and reduce cardiovas-
cular events better than lactate fluids. The buffer concentration should exert a buffer load that
may compensate for deficits, for losses in the buffer process, and for extracorporeal losses
and should therefore usually be supraphysiological. However, the dialysate buffer or elec-
trolyte concentration need always to be balanced with that of the replacement fluids
employed. Both fluids should contain electrolytes in concentrations aiming for a physiologic
level and taking into account preexisting deficits or excess and all input and losses.
Clinicians should be aware that in CRRTs the quality control for sterility, physical properties,
individualized prescription and balance control are vitally important.
Copyright © 2007 S. Karger AG, Basel
Major changes have occurred in the medical management of intensive care
unit (ICU) patients over the past two decades, and the role of extracorporeal
blood treatments has broadened from conventional renal function replacement
to a series of non-renal conditions.
Continuous renal replacement therapies (CRRTs) have found widespread
use and acceptance because they incorporate several advantages, such as
improved hemodynamic stability, gradual urea removal without fluctuations,
Aucella/Di Paolo/Gesualdo 288
and optimal fluid balance. Further, they obviate the need of nutritional restric-
tions in terms of both the volume and composition of the nutrients administered
to critically ill patients [1].
In this setting, we need to keep in mind that patients with acute renal fail-
ure (ARF) depend on dialysis to maintain fluid and electrolyte balance. One
crucial goal of CRRT in the treatment of ARF patients is to achieve and main-
tain normal or near-normal acid-base balance over time, thereby preventing the
detrimental effects of acidemia on cardiovascular performance, hepatic metab-
olism and hormonal response [2]. Over the course of the procedure solutes are
ultrafiltered or diffused between blood and dialysate such that the plasma com-
position is restored toward normal values. Obviously, the makeup of the
dialysate and the replacement fluid, in combination with the features of the
dialysis technique [3], are of paramount importance to accomplish this goal.
Choice of Buffer and Acid-Base Balance
Acid-base balance is greatly influenced by the type of dialysis technique
employed, diffusion and convection being differently balanced. Moreover, the
type of buffer in both replacement fluid and dialysate greatly influences not
only acid-base correction, but also clinical outcome.
The dialysis techniques have a clear influence on buffer kinetics. Despite a
comparable blood and dialysate/replacement fluid flow rate, different techniques
are associated with varying effects on electrolyte and acid-base homeostasis.
In continuous veno-venous hemofiltration (CVVH) the buffer balance
depends on buffer losses with the ultrafiltrate and buffer gain with the replace-
ment fluid. If patients undergo isolated ultrafiltration, without infusing any
replacement fluid, for example in chronic cardiac failure, bicarbonate losses are
compensated by the reduction in the distribution volume of the buffer, such that
bicarbonate serum levels do not change significantly [3]. When replacement
fluids are infused along with fluid ultrafiltration, bicarbonate losses in the ultra-
filtrate need to be balanced by equal amounts in the infusion solution. In the
clinical setting, metabolic acidosis is the most frequent condition occurring in
these patients and a positive buffer balance is commonly required. Moreover,
clinicians need to be aware that the bicarbonate concentration in the ultrafiltrate
is higher than the plasma level, because of a sieving coefficient of ⬎1.
In continuous veno-venous hemodialysis (CVVHD), a pure diffusive tech-
nique, buffer balance is a function of the concentration gradient between the
dialysate and blood, and a feedback between the base balance and blood bicar-
bonate also occurs. Thus, the correction of acidosis depends only on the buffer
concentration in the dialysate [1, 3].
Fluids Composition for CRRT 289
In techniques with both diffusive and convective processes, such as contin-
uous veno-venous hemodiafiltration (CVVHDF), the buffer gain depends on
both the dialysate and the replacement fluid, with the inherent risk of alkalemia.
A comparison of CVVHDF and CVVH showed that they had a significantly
different impact on bicarbonate and sodium control [4]. Specifically, CVVH
was associated with a lower incidence of metabolic acidosis and a higher inci-
dence of metabolic alkalosis [4], possibly due to small differences in the total
amount of buffer infused during CVVH.
Which Buffer Should Be Preferred in CRRT?
Lactate, acetate, and bicarbonate have all been used as buffers during
CRRT. Citrate has been used as a buffer and anticoagulant. Although bicar-
bonate is the natural buffer, bicarbonate-based solutions have not been avail-
able until recently, because of the higher risk of bacterial contamination and
the instability of this buffer in the presence of calcium and magnesium ions.
Moreover, the rate of administration of buffer solutions during CRRT is defi-
nitely lower than during discontinuous treatments, because the accumulation
of acetate or lactate has rarely been reported, although this may be an impor-
tant point in high-volume treatments [5]. It is now clear that lactate or bicar-
bonate fluids offer a better control of acid-base balance and improved
cardiovascular stability compared to acetate fluids [5]. In clinical practice,
both lactate and bicarbonate ions are used in replacement fluids and in
dialysate for CRRT. These buffer have shown similar efficacy in correcting
metabolic acidosis [6]. Obviously, lactate infusion increases serum lactate lev-
els and this might lead to a misleading interpretation of the clinical situation.
Solutions containing lactate are contraindicated in patients with concomitant
lactic acidosis and in those with lactate intolerance, defined as a rise of
5 mmol/l during CRRT: these patients are at a high risk of worsening acidosis
because of insufficient conversion of lactate into bicarbonate in the face of
ongoing bicarbonate losses. Currently, lactate concentrations vary from
45 mmol/l (chloride 103 mmol/l) down to 35 mmol/l (chloride 110 mmol/l),
and these differences in solute composition can lead to clinical consequences,
namely to the development of hypochloremic metabolic alkalosis after several
days of CRRT with high lactate/low chloride solutions, or hyperchloremic
metabolic acidosis when a low lactate/high chloride fluid is used [5].
Moreover, several studies suggest that lactate-buffered replacement fluids can
exert negative effects on different metabolic and hemodynamic parameters [2].
In ICU patients suffering from multiple organ dysfunction, the conversion of
lactate to bicarbonate is frequently impaired and the resulting increase in blood
Aucella/Di Paolo/Gesualdo 290
lactate concentration may exert multiple negative effects. On the other hand,
the administration of bicarbonate in patients with lactic acidosis has also been
questioned, although CRRT makes it possible to overcome some of the
unwanted effects of bicarbonate infusion, e.g. volume overload, hyperosmolar-
ity, and a decrease in ionized calcium.
Bicarbonate-buffered replacement fluids are presently considered a valu-
able approach to improve the prognosis of critical ill patients with ARF. Several,
but not all, studies support the view that bicarbonate-buffered replacement flu-
ids can improve acid-base status and reduce the incidence of cardiovascular
events in ICU patients, when compared with lactate-buffered fluids [7]. This
suggestion should be viewed in light of the peculiar milieu of the critically ill
patient, the specific features of which (systemic inflammation, increased
energy expenditure and metabolic activation, cardiovascular stress, hyperdy-
namic cardiovascular response) cause a clinical picture characterized by an
excess of lactate production, even in the absence of an increase in blood lactate
levels, increased oxygen consumption and elevated protein catabolism. Blood
lactate, if not further metabolized, acts as a strong anion, which has the same
acidifying effect of chloride. Accordingly, iatrogenic hyperlactatemia can
causes a state of metabolic acidosis. When oxidizable anions are used in
replacement fluids, these anions, acetate, lactate or citrate, need to be com-
pletely oxidized to CO
2
and H
2
O in order to generate bicarbonate and then the
buffering capacity equals that of bicarbonate solutions. However, on the con-
trary, if metabolic conversion is not adequate, the increased blood concentration
of the anions leads to an acidotic condition. The type and extent of these acid-
base changes are governed by the intensity of plasma water exchange/dialysis,
by the buffer content of the replacement fluids, and by the actual metabolic rate
of the above-mentioned anions.
The advantage of using a buffer-free replacement solution is that the dose
of bicarbonate can be titrated according to a given target value of base excess,
which may vary according to the clinical picture of the ICU patient, as well as
to specific treatment modalities, such as permissive hypercapnia. The applica-
tion of a lung-protective strategy with reduced tidal volumes, effective lung
recruitment, adequate positive end-expiratory pressure to minimize alveolar
collapse during expiration, and permissive hypercapnia has been shown to be
advantageous in adult patients who have acute respiratory distress syndrome
[8]. Indeed, strategies such as permissive hypercapnia or permissive hypoxemia
have been reported to favor the onset of ARF by compromising renal blood flow
[9]. Therefore, the correction of acidosis, as suggested by the NIH protocol
Acute Respiratory Distress Syndrome Network [10], has been questioned as
experimental evidence and preliminary clinical data indicate that buffering
hypercapnic acidosis abrogates its protective effects [11, 12]. Nevertheless,
Fluids Composition for CRRT 291
patients with permissive hypercapnia may require huge amounts of buffer to
correct acidosis, the infusion of the buffer itself leading to a further increase in
CO
2
production, thus leading to a vicious circle which, in the presence of lim-
ited CO
2
elimination, may worsen the acidotic state of the patient. The infusion
of lactate in this condition does not appear appropriate because relatively more
lactate than bicarbonate is required to achieve the same base excess target. In
conclusion, the most reliable method to control respiratory acidosis of permis-
sive hypercapnia, is the administration of lactate-free fluids and bicarbonate
titration according to the requirements of individual patients [13].
The paradigmatic CRRT technique using bicarbonate-buffered replace-
ment solutions is acetate-free CVVH (AF-CVVH) [14]. The technique
is based on separate infusion of water and electrolytes administered pre-
filtration, and isotonic sodium bicarbonate administered post-filtration. The
setting of the technique is based on a model predicting the bicarbonate infu-
sion rate for a targeted plasma bicarbonate level. Indeed, AF-CVVH allows
fast control of acidosis and has the main advantage of separately controlling
urea retention and metabolic acidosis in patients with severe ARF and cardio-
vascular instability. This method offers the possibility of titrating the amount
of the buffer given more precisely, and has been used to treat severe metabolic
acidosis.
A further particular point is the use of citrate for regional anticoagulation
in CRRT. One absolute requirement for CRRT is anticoagulation which can
expose patients to the risk of bleeding. Citrate anticoagulation may limit such
risk, but the prevention of citrate side effects requires meticulous monitoring.
Then, since citrate is metabolized to bicarbonate, each citrate ion producing
three bicarbonate ions, no additional anionic base is required to control meta-
bolic acidosis [5]. When adopting citrate anticoagulation, normal saline is
infused as a predilution fluid, while a special hyponatremic dialysate is required
because of the sodium load due to trisodium citrate and saline infusion. The
amount of citrate to infuse is titrated on the basis of coagulation parameters, or
on surrogate parameters, such as the total calcium-ionized calcium rate, and not
on a target base excess. Consequently, the use of citrate may be associated with
both metabolic alkalosis and acidosis [1]. Metabolic alkalosis is more common
in patients with hepatic dysfunction or in those requiring support with large
amounts of blood products containing acid citrate dextrose as anticoagulant. In
this setting, alkalosis may be prevented using a high chloride load, i.e. large vol-
umes of normal saline as predilution fluid, and infusing calcium chloride. When
required, bicarbonate losses may be enhanced by increasing the dialysate flow
or the predilution saline infusion. Metabolic acidosis may also occur in the
presence of hypercitratemia and hypercalcemia, usually in the setting of ARF
with rhabdomyolysis.
Aucella/Di Paolo/Gesualdo 292
Fluid Composition
Electrolyte balance during ARF treated with CRRT is largely dependent
on the electrolyte plasma concentration available for ultrafiltration, the ultra-
filtration rate and the composition of the replacement solution. As CRRT
works continuously, serious derangement in fluid and electrolyte homeostasis
may occur in the absence of careful prescription and extremely vigilant moni-
toring [15].
Replacement fluids for CVVH should contain the following concentrations
of electrolytes and glucose: sodium 140 mmol/l; chloride 108–112 mmol/l;
potassium 0–4 mmol/l; calcium 1.5–1.75 mmol/l; magnesium 0.5–0.75 mmol/l,
and glucose 0–15.00 mg/l [1].
Sodium
Sodium balance can be rather different according to the different CRRT
techniques. In CVVH, a totally convective therapy, the value of ultrafilterable
sodium must be taken into account to choose the correct sodium concentration
in the substitution fluid [15]. It is worth remembering that the sodium concen-
tration in the ultrafiltrate is around 7 mEq/l lower than the plasma water sodium
concentration, because a certain amount of the total ion is complexed with
anions, particularly with bicarbonate and proteins. When using techniques
which combine diffusive and convective processes, such as CVVHD or
CVVHDF, a dialysis solution is added to the system. Even with these tech-
niques sodium transport is mainly due to the convective process; the diffusive
flux is definitely lower and is strongly influenced by the sodium concentration
of the dialysate. A retrospective study showed that CVVHDF was more likely to
achieve serum sodium concentrations within the normal range than CVVH,
thereby supporting the role of dialysate in the maintenance of an adequate
sodium concentration [4]. It has been suggested that a supraphysiological
sodium concentration in the dialysate or replacement solutions would improve
the hemodynamic stability or prevent the increase in intracranial pressure, but
there are no conclusive data concerning this issue [6].
Potassium
Hyperkalemia is a typical feature of ARF, particularly in the presence of
tissue breakdown (crush syndrome, hemolysis, hypercatabolic syndromes).
Unless important hyperkalemia is present, 3–4 mEq/l potassium needs to be
added to the replacement fluid in CVVH in order to avoid hypokalemia. In tech-
niques comprising a diffusive process, namely CVVHD or CVVHDF, the con-
tent of potassium in the dialysate must be titrated to about 2–4 mEq/l, to avoid
the risk of hypokalemia [15]. The very slow flux of potassium from cells to the
Fluids Composition for CRRT 293
extracellular space would possibly contribute to the high hemodynamic toler-
ance of CRRT.
Divalent Ions
About 60% of total plasma calcium is ultrafilterable, then substitution
fluid in CVVH must contain about 3mEq/l of calcium in order to prevent
hypocalcemia. When mixed techniques, such as CVVHD, are used, the
dialysate concentration of calcium usually ranges from 3 to 3.5 mEq/l. Patients
undergoing an exchange of very large volumes of ultrafiltrate during CRRT
may be prone to the risk of a positive calcium balance, and therefore may
require a lower content of calcium in replacement fluids. The same cautions are
required to keep an appropriate magnesium balance.
Intensive schedules of CRRT can easily induce hypophosphatemia in the
critically ill patient, the risk being magnified by prolonged parenteral nutrition,
malnourishment or concomitant metabolic alkalosis. This condition may
require oral or parenteral supplementation with phosphate salts, as well as sup-
plementation of dialysate solutions. The addition of phosphate salts to the
dialysate and/or the replacement solutions facilitates phosphate handling, while
the risk of phosphate precipitation in the presence of calcium ions has recently
been excluded [16].
Glucose
The recent publication of a randomized trial showing a significant survival
advantage with strict glucose control in critically ill patients has important con-
sequences for CRRT with glucose-containing dialysate or substitution fluids
[17]. The use of physiologic concentrations of glucose in the dialysate and in
replacement fluids is advised to prevent or compensate extracorporeal losses,
while glucose-free solutions might be used when an adequate nutritional regi-
men has been established [6].
A special case is that of critically ill children weighing ⬍10 kg who require
blood as a priming solution for the extracorporeal circuit before initiating
CRRT to prevent hemodilution and to maintain adequate oxygenation [18]. Blood
preparations usually contain supraphysiological electrolyte concentrations and
a nonphysiological acid-base balance that may exacerbate the critical condition
of the small patient. In such cases, the pretreatment of blood bank-derived
blood aimed to normalize pH and electrolyte concentration yields more physio-
logical blood priming.
In conclusion, due to its continuous nature, CRRT need to be carefully
monitored for the composition of dialysate and replacement fluids to avoid dan-
gerous electrolyte imbalance. Moreover, it is possible that in the near future we
Aucella/Di Paolo/Gesualdo 294
may see more sophisticated CRRT solutions containing nutrients and antioxi-
dants that may improve the outcome of ICU patients.
Physical Properties
All lactate-based or buffer-free solutions are acidic, while bicarbonate-
based fluids have a physiological pH. Experimental findings suggest that acidic
fluids cause intracellular acidosis and reduced activity of macrophages and
blood mononuclear cells [5].
The use of replacement fluids or dialysate solutions at room temperature as
well as continuous blood flow through the extracorporeal system cause an aver-
age 2⬚C reduction in body temperature, and an energy loss of about
1,000 kcal/day [1]. Although the clinical significance of this effect is not clear at
this point, the energy loss measured during CVVHD has been suggested to be
important in hemodynamic stability or even for patient prognosis [19]. Heat loss
and consequent hypothermia may also affect immune functions and increase the
risk of clotting of the CRRT circuit. The extent of these effects depends on the
length of the circuit, on the flow rate of blood, replacement fluids and dialysate,
on body weight and finally on the presence or absence of an intact autoregula-
tory mechanism to preserve core temperature. A heating device may avoid the
above-cited side effects; on the other hand a lower body temperature may be
desirable in patients with excessive oxygen consumption and low systemic vas-
cular resistance. At present, it is not clearly known in which patients the net
effect of CRRT-induced hypothermia is useful or harmful.
Obviously substitution fluids need to be sterile, and the same is true for the
dialysate, not only in high-flux dialysis schedules when backfiltration is most
likely to occur, but also in other continuous dialytic treatments. As in intermit-
tent treatments, the feasibility of online production of substitution fluids in
CRRT has recently been evaluated.
Administration Route
The CRRT prescription directly affects electrolyte and acid-base balance
[5, 6, 20]. In diffusive or mixed dialysis schedules, the concentration of elec-
trolytes in the blood is dependent on the corresponding concentration in the
dialysate and in replacement fluids. In pure convective treatments most ions
pass freely across the membrane, showing a sieving coefficient of around one.
In this case the ionic concentration of the substitution fluids should be very
similar to that of normal plasma, with an alkaline pH.
Fluids Composition for CRRT 295
The different features of predilution (lower solute clearance, higher ultra-
filtration rate and larger filter surface area) and of postdilution (higher filtration
fraction) are well known [1, 5, 6, 20]. Predilution may be preferred to reduce
the need for heavy anticoagulation or to enhance the ultrafiltration rate, espe-
cially relevant for HV-CVVH. Possibly, pre- and postdilution may be combined
when extracorporeal clearance is limited by an inadequate blood flow, but this
point deserves future studies.
Conclusions
The great complexity of current CRRT schedules, exchanging up to
50 l/day, has been made possible by the development of specialized dialysate
and replacement fluid solutions. These strategies have been supported by stud-
ies reporting an improvement in the survival of the critically ill patient follow-
ing the adoption of technical approaches providing the exchange of large
amounts of fluids, and/or the use of high dialysate flows. However, with the
increase in exchanged volumes, the control of fluid sterility, physical properties
and composition, and the choice of individualized prescriptions are becoming
increasingly crucial in the single ICU patient.
References
1 Manns M, Sigler MH, Teehan BP: Continuous renal replacement therapies. An update. Am J
Kidney Dis 1998;32:185–207.
2 Macias WL: Choice of replacement fluid/dialysate anion composition in continuous renal replace-
ment therapy. Am J Kidney Dis 1996;28(suppl):S15–S20.
3 Feriani M, Dell’Aquila R: Acid-base balance and replacement solutions in continuous renal
replacement therapy. Kidney Int 1998;53(suppl 66):S156–S159.
4 Morimatsu H, Uchino S, Bellomo R, Ronco C: Continuous renal replacement therapy: does tech-
nique influence electrolyte and bicarbonate control? Int J Artif Organs 2003;26:289–296.
5 Davemport A: Dialysate and substitution fluids for patients treated by continuous forms of renal
replacement therapy; in Ronco C, Bellomo R, La Greca G (eds): Blood Purification in Intensive
Care. Contrib Nephrol. Basel, Karger, 2001, vol 132, pp 313–322.
6 Schetz M, Leblanc M, Murray PT: The Acute Dialysis Quality Initiative – Part VII: Fluid compo-
sition and management in CRRT. Adv Renal Replace Ther 2002;9:282–289.
7 Barenbrock M, Hausberg M, Matzkies F, De La Motte S, Shaefer RM: Effects of bicarbonate and
lactate buffered replacement fluids on cardiovascular outcome in CVVH patients. Kidney Int
2000;58:1751–1757.
8 O’Croinin D, Ni Chonghaile M, Higgins B, Laffey JG: Bench-to-bedside review: permissive
hypercapnia. Crit Care 2005;9:51–59.
9 Kuiper JW, Groeneveld AB, Slutsky AS, Plotz FB: Mechanical ventilation and acute renal failure.
Crit Care Med 2005;33:1408–1415.
10 Acute Respiratory Distress Syndrome Network: 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 2000;342:1301–1308.
Aucella/Di Paolo/Gesualdo 296
11 Gehlbach BK, Schmidt GA: Bench-to-bedside review: treating acid-base abnormalities in the
intensive care unit – the role of buffers. Crit Care 2004;8:259–265.
12 Rotta AT, Steinhorn DM: Is permissive hypercapnia a beneficial strategy for pediatric acute lung
injury? Respir Care Clin N Am 2006;12:371–387.
13 MacLean AG, Davemport A, Cox D, Sweny P: Effects of continuous haemodiafiltration against
lactate-buffered and lactate-free dialysate in patients with and without liver dysfunction. Kidney
Int 2000;58:1765–1772.
14 Wynckel A, Wuillai A, Bene B, Cornillet J, Chanard J: Assessment of acetate free continuous
veno-venous hemofiltration in acute renal failure. ASAIO J 1998;44:M606–M609.
15 Locatelli F, Pontoriero G, Di Filippo S: Electrolyte disorders and substitution fluid in continuous
renal replacement therapy. Kidney Int 1998;53(suppl 66):S151–S155.
16 Troyanov S, Geadah D, Ghannoum M, Cardinal J, Leblanc M: Phosphate addition to hemodiafil-
tration solutions during continuous renal replacement therapy. Intensive Care Med 2004;30:
1662–1665.
17 Van Der Berghe G, Wouters P, Weekers F, et al: Intensive insulin therapy in critically ill patients.
N Engl J Med 2001;345:1359–1367.
18 Pasko DA, Mottes TA, Mueller BA: Pre dialysis of blood prime in continuous hemodialysis
normalizes pH and electrolytes. Pediatr Nephrol 2003;18:1177–1183.
19 Yagi N, Leblanc M, Sakai K, Wright EJ, Paganini EP: Cooling effect of continuous renal replace-
ment therapy in critically ill patients. Am J Kidney Dis 1998;32:1023–1030.
20 Leblanc M: Fluid composition for CRRT; in Ronco C, Bellomo R, Brendolan A (eds): Sepsis, Kidney
and Multiple Organ Dysfunction. Contrib Nephrol. Basel, Karger, 2004, vol 144, pp 222–227.
Prof. Loreto Gesualdo
Department of Nephrology, Dialysis and Transplantation, University of Foggia
Viale Pinto, 1
IT–71100 Foggia (Italy)
Tel. ⫹39 0881 732 054, Fax ⫹39 0881 736 001, E-Mail l.gesualdo@unifg.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 297–303
Results from International Questionnaires
Zaccaria Ricci
a
, Sergio Picardo
a
, Claudio Ronco
b
a
Department of Pediatric Cardiology and Cardiac Surgery, Bambino Gesù Hospital,
Rome, and
b
Department of Nephrology, Dialysis and Transplantation, St. Bortolo
Hospital, Vicenza, Italy
Abstract
Background: The practice of renal replacement therapy (RRT) has reached an optimal
standard of care worldwide. Nevertheless, some aspects of acute renal failure treatment and
support still present wide variability between different centers. This is especially true for the
mode and dose of RRT. This review describes the epidemiology of dialysis prescription and
delivery around the world based on recent observational studies and international surveys.
Results: Continuous RRT is delivered in 80% of intensive care units around the world.
Since a certain consensus has been achieved on the adequacy of 35 ml/kg/h of clearance in
continuous therapies, recent observations based on questionnaires and surveys demonstrated
that such adequate therapy was only prescribed in the minority of patients. The number of
centers prescribing adequate dialysis dose is increasing, but there are still many institutions
where prescription is made with no specific adequacy targets and effective delivery is not
measured. Several barriers to reaching adequacy targets have been identified including the
lack of a high evidence multicentric trial, logistics, costs, personnel and technical difficul-
ties. Conclusion: A trend to continuous therapies and increased RRT dosage over the last 10
years is shown by the surveys presented, even if scientific evidence is now very necessary as
far as definitive RRT indications and prescriptions are concerned.
Copyright © 2007 S. Karger AG, Basel
The practice of renal replacement therapy (RRT) for acute renal failure
(ARF) is extremely variable including different techniques from intermittent
hemodialysis (IHD) to continuous renal replacement therapies (CRRT) [1].
There is a general consensus that optimal strategies to improve patient outcome
in ARF should include delivery of an adequate treatment dose [2]. The Acute
Dialysis Quality Initiative (ADQI) [3, 4] has identified several areas where con-
sensus is lacking and recommendations for good clinical practice are very
much needed. In this setting, the practice of CRRT seems to be a specific and
important area where criteria for starting, the modality of therapy, and treatment
Ricci/Picardo/Ronco 298
prescription are still not carried out on solid bases of previous experience or
evidence, but rather on local protocols [5]. However, ADQI has brought about a
large consensus, and also encouraged possible studies and potential analyses
that might be useful to generate not yet available evidence. One aspect that has
been noted for urgent definition is the actual practice pattern in the choice and
conduction of renal replacement techniques. Recently, several international sur-
veys have tried to depict routine clinical practice in ARF management in order
to provide knowledge on ‘real-world’ issues, on physicians’ compliance to prac-
tice guidelines, on educational needs and research objectives. Here we review
and discuss some results of these observations.
Beginning and Ending Supportive Therapy
With the intention of determining the association between outcome and dif-
ferent epidemiological parameters (period prevalence of ARF, etiology, illness
severity, and clinical management of ARF), the Beginning and Ending Supportive
Therapy for the Kidney (BEST Kidney) investigators conducted the largest multi-
national, multicenter, prospective, epidemiological survey of ARF in intensive
care unit (ICU) patients [6] who either were treated with RRT or fulfilled at least
1 of the predefined criteria for ARF. ARF criteria were defined a priori in order to
achieve a standardized definition that was easily reproducible for each center:
oliguria, defined as a urine output of ⬍200 ml in 12 h, and/or marked azotemia,
defined as a blood urea nitrogen level of ⬎30 mmol/l. The data were collected
from September 2000 to December 2001 at 54 hospitals in 23 countries. Data
about 29,269 critically ill patients were recorded during the 16-month study
period. The median age of patients with ARF was 67 years, the median SAPS II
score was 48, the median body weight was 74 kg. Approximately 30% of the
patients had chronic renal dysfunction but were not receiving dialysis treatment.
The estimated creatinine clearance on admission to the ICU was 35 (interquartile
range 20–59) ml/min. 1,738 (5.7%) had ARF during their ICU stay, including
1,260 (4.3%) who were treated with RRT. Overall hospital mortality was 60.3%,
significantly higher than the SAPS II predicted mortality. The most common con-
tributing factor to ARF was septic shock (47.5%). Continuous RRT was the most
common initial modality used (80.0%), followed by intermittent RRT (16.9%),
peritoneal dialysis and slow continuous ultrafiltration (3.2%). 86.2% survivors
were dialysis-independent at hospital discharge. Independent risk factors for hos-
pital mortality included use of vasopressors, mechanical ventilation, septic shock,
cardiogenic shock, and hepatorenal syndrome.
From the findings of this study we can reasonably finally conclude that:
the prevalence of ARF among the ICU population is between 5 and 6%; more
Results from International Questionnaires 299
than 70% of these patients require RRT; hospital survival among ARF patients
is still disappointing (about 40%), and renal recovery among survivors is very
high. A systematic literature review [7] recently reported that no evidence of a
substantial improvement in outcome from ARF over the last 50 years has been
observed. The mortality rates remain superior to 50%, and it is likely that it will
remain unchanged in the next decade or more: this mortality rate represents a
level of adequate performance of the healthcare system but it should not be
used to obtain comparisons with other periods. In other words, as therapeutic
capability improves and the system continues to achieve a mortality of 50% for
these very sick patients, the healthcare system will presumably admit and treat
sicker and sicker patients with ARF. Future epidemiologic studies will need to
take into account this confounder in order to appreciate the continuing change
in illness severity [8].
CRRT through the New Millennium
Following such a practice-related approach, we have distributed a ques-
tionnaire on specific issues about practice patterns in the field of RRT during
the 1st (1998) [5] and the 3rd (2004) International Course on Critical Care
Nephrology [1]. The analysis comparing the answers from these 2 surveys
covering 6 years presented many interesting results, especially with regard to
the steep improvement in the standard of care for critical care nephrology.
Responders were nephrologists (60%) and intensivists (40%) from all 5 conti-
nents who attended both meetings. 345 questionnaires in 1998 and 560 in 2004
were correctly completed. The two sets of completed questionnaires were col-
lected into an access database and the results examined. Percentage values are
reported in order to compare the two uneven populations. Continuous arterio-
venous hemofiltration (HF), representing one of the available options for more
than 70% of candidates in 1998, was abandoned by the responders in 2004.
Continuous veno-venous RRTs were also considered by more than 90% candi-
dates in both surveys. In 2004 intermittent techniques, available in only 20%
centers in 1998, were routinely available in more than 80% institutions, admin-
istered as intermittent hemodialysis (two thirds) or as slow extended daily dial-
ysis (one third). Peritoneal dialysis was a rare option (5%) in ICUs in both 1998
and 2004. HF was and remained the preferred RRT modality throughout the
evaluated period: it was selected by a large majority of responders in 2004,
whereas it had appeared to be less clearly predominant before: perhaps recent
literature and new dedicated machines facilitated HF prescription and delivery
[9]. Apparently, in the 6-year interval, ICU physicians prescribing RRT without
further nephrological counseling increased from about 15% to almost 30%.
Ricci/Picardo/Ronco 300
Anticoagulation management did not change significantly, and patient bleeding
remained one of the most selected complains during both meetings, together
with circuit clotting. Responders notably modified their RRT prescriptions
from 1998 to 2004: urea clearance increased from 1–1.5 (range 0.5–2) to 2–3
(range 1–10) l/h. A more in-depth analysis of CRRT prescription in 2004
showed interesting results: surprisingly only about 50% of RRT are prescribed
upon a standard protocol. Furthermore, a large part of our responders seemed to
be uncertain about treatment prescription: this could mean that delivery is not
personalized to the patient and the clinical setting. In 2004 participants mostly
reported prescribing a dose of 35 ml/kg/h or 2–3 l/h as the urea efficiency tar-
get, with a range from 1 to more than 5 l/h which, in our opinion, is consistent
with a trend to increased RRT dosage over the last 10 years. In 1998 CRRT dose
prescription only ranged from 0.5 to 2 l/h. Different from that survey, in 2004
low treatment efficiency was not a matter of complaint anymore, whereas filter
clotting and catheter dysfunction still represented a problem for operators in the
field of RRT. As a matter of fact, from 1998 to 2004 heparin infusion remained
the preferred anticoagulation technique and anticoagulation side effects (bleed-
ing and hematoma) are still a problem. Less dangerous alternatives or more
effective molecules are still being evaluated [10, 11]. Similarly, great techno-
logical improvements were evident between the two sets of questionnaires:
more than 50% of the available equipment in 1998 consisted of continuous arte-
rio-venous HF kits, or adapted machines from chronic therapies, whereas in
2004, 100% of the participants declared using dedicated integrated monitors.
Exactly as in 1998, as far as non-renal indications are concerned (congestive
heart failure, sepsis, anasarca, systemic inflammatory disease), 90% of respon-
ders stated that they agree with non-renal indications: two thirds in 1998 and
only half in 2004 declared prescribing RRT for extended indications even in the
absence of acute renal failure (fig. 1). The lack of scientific evidence is the pri-
mary reason for skepticism with regard to adopting extracorporeal treatment:
presumably for this reason the number of the responders who in 1998 declared
starting RRT in case of septic shock, even in the absence of ARF, decreased sig-
nificantly in the last survey (fig. 2). This contradiction could indicate that cur-
rent RRT practice might not completely apply to evidence-based medicine and
that studies with a high level of evidence in the field of non-renal RTT indica-
tions are strongly needed. In the case of a non-renal indication, most of the
meeting participants would prescribe routine treatment without changing the
usual machines or settings. Nonetheless, our audit selected a number of alterna-
tive techniques as being feasible treatments during sepsis syndrome (namely,
high-volume HF, continuous plasma filtration adsorption and hemoperfusion),
showing that constant attention is paid to the most recent technical possibilities
offered by extracorporeal treatments.
Results from International Questionnaires 301
DO-RE-MI
The Dose Response Multicenter International collaboration (DO-RE-MI)
[12] is currently seeking to address the issue of how practice patterns are cur-
rently chosen and performed. DO-RE-MI is an observational, multicenter study
conducted in ICUs. The primary aim is to study the dialysis dose delivered,
which will then be compared with ICU mortality, 28-day mortality, hospital
mortality, ICU length of stay, and the number of days of mechanical ventilation.
It is hoped that this international collaboration will provide a clearer picture of
9%
2%
8%
90%
0%
91%
2004
1998
Yes
No
No response
Fig. 1. Answers of participants to the 1st and 3rd International Course on Critical Care
Nephrology to the question whether extracorporeal therapies are feasible treatments for
‘non-renal’ indications.
60
50
40
30
20
Yes
No
No
response
10
0
%
1998
2004
Fig. 2. Answers of participants to the 1st and 3rd International Course on Critical Care
Nephrology to the question whether extracorporeal therapies can be used with ‘non-renal’
indications in the absence of acute renal failure.
Ricci/Picardo/Ronco 302
how RRT is chosen, prescribed and delivered, and how such delivery may affect
outcome. The preliminary results of this survey are presented in detail else-
where in the present issue.
Conclusion
The syndrome known as ‘acute renal failure’ is common in the ICUs and
may affect from 1 to 25% of the patients [6, 13]. This wide range might depend
on the different patient populations present in different centers, and also on the
different criteria used to define its presence. When severe ARF occurs in
patients with severe systemic illness, septic shock and multiorgan dysfunction
[14], it considerably complicates patient management, increases the cost of
care, and is associated with a high level of morbidity and mortality [6]. Starting
from the definition of ARF itself, many controversies surround its management
[15]. Surveying routine clinical practice may provide precious knowledge on
‘real-world’ issues, on physicians’ compliance to practice guidelines, on educa-
tional needs and research objectives. Participants who attend this kind of meet-
ing are obviously a self-selected population and their answers cannot reasonably
reflect the worldwide daily reality of patient care. Nonetheless, especially for
the BEST study, the group of respondents was indeed quite large and a broad
distribution of participants was evident.
Analysis of available techniques in different institutions showed a certain
prevalence of continuous techniques. Nonetheless, in many institutions intermit-
tent techniques are present together with continuous ones, thus showing the
availability of different prescriptions and practices. Surprisingly, according to
our survey, only in about 50% of cases is RRT managed as a standard protocol.
Furthermore, a large number of responders seemed to be uncertain with regard
to treatment prescription: this could mean that delivery is not personalized on a
patient and clinical setting. In the late 1990s dose prescription was mostly
described in terms of liters of effluent per day. Very seldom did the dose amount
exceed 30l/day leading to an average dose delivery in the range of
15–20 ml/kg/h. After 2000 a dose-related discrimination was made. Furthermore
the respondents seemed not only to have an increased awareness of the impor-
tance of adequate doses in ARF, but also on the potential effect of higher doses in
septic patients. Nonetheless, in the last survey, a large number of respondents
still admitted to ignoring how prescription was made in their center.
A trend to continuous therapies and increased RRT dosage with respect to
the last 10 years was shown by the presented surveys, even if scientific evi-
dence is now more necessary than ever as far as definitive RRT indications and
prescriptions are concerned [16].
Results from International Questionnaires 303
References
1 Ricci Z, Ronco C, D’Amico G, et al: Practice patterns in the management of acute renal failure in
the critically ill patient: an international survey. Nephrol Dial Transplant 2006;21:690–696.
2 Ricci Z, Bellomo R, Ronco C: Dose of dialysis in acute renal failure. Clin J Am Soc Nephrol
2006;1:380–388.
3 Bellomo R, Ronco C, Kellum A, Mehta RL, Palevsky P; ADQI Workgroup: Acute renal failure –
definition, outcome measures, animal models, fluid therapy and information technology needs:
the Second International Consensus Conference of Acute Dialysis Quality Initiative (ADQI)
group. Crit Care 2004;8:R204–R212.
4 ADQI: Acute Dialysis Quality Initiative. http://www.adqi.net.
5 Ronco C, Zanella M, Brendolan A, Milan M, Canato G, Zamperetti N, Bellomo R: Management
of severe acute renal failure in critically ill patients: an international survey in 345 centres.
Nephrol Dial Transplant 2001;16:230–237.
6 Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C,
Macedo E, Gibney N, Tolwani A, Ronco C; Beginning and Ending Supportive Therapy for the
Kidney (BEST Kidney) Investigators: Acute renal failure in critically ill patients: a multinational,
multicenter study. JAMA 2005;294:813–818.
7 Ympa IP, Sakr Y, Reinhart K, Vincent JL: Has mortality from acute renal failure decreased? A sys-
tematic review of the literature. Am J Med 2005;118:827–832.
8 Bellomo R: The epidemiology of acute renal failure: 1975 versus 2005. Curr Opin Crit Care 2006;
12:557–560.
9 Ronco C, Bellomo R, Homel P, Brendolan A, Maurizio D, Piccini P, La Greca G: Effects of
different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a
prospective randomised trial. Lancet 2000;355:26–30.
10 Fiaccadori E, Maggiore U, Rotelli C, Minari M, Melfa L, Cappe G, Cabassi A: Continuous
haemofiltration in acute renal failure with prostacyclin as the sole anti-haemostatic agent.
Intensive Care Med 2002;28:586–593.
11 Monchi M, Berghmans D, Ledoux D, Canivet JL, Dubois B, Damas P: Citrate vs. heparin for anti-
coagulation in continuous venovenous hemofiltration: a prospective randomized study. Intensive
Care Med 2004;30:260–265.
12 Kindgen-Milles D, Journois D, Fumagalli R, et al: Study protocol: The Dose Response
Multicentre International collaborative initiative (DO-RE-MI). Crit Care 2005;9:R396–R406.
13 Liano F, Junco E, Madero R, Pascual J, Verde E; Madrid Acute Renal Failure Study Group: The
spectrum of acute failure in the intensive care unit compared with that seen in other settings.
Kidney Int 1998;53:S16–S24.
14 Kleinknecht D: Risk factors for acute renal failure in critically ill patients; in Ronco C, Bellomo R
(eds): Critical Care Nephrology. Dordrecht, Kluwer Academic, 1998, pp 143–152.
15 Kellum J, Palevsky P: Renal support in acute kidney injury. Lancet 2006;368:344–345.
16 Bellomo R: Do we know the optimal dose for renal replacement therapy in the intensive care unit?
Kidney Int 2006;70:1202–1204.
Zaccaria Ricci
Department of Pediatric Cardiology and Cardiac Surgery
Bambino Gesù Hospital, Piazza S. Onofrio
IT–00100 Rome (Italy)
Tel. ⫹39 06 6859 3333, E-Mail z.ricci@libero.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 304–308
Intermittent Hemodialysis for Renal
Replacement Therapy in Intensive Care:
New Evidence for Old Truths
W. Van Biesen, N. Veys, R. Vanholder
University Hospital Ghent, Ghent, Belgium
Abstract
Acute renal failure requiring dialysis is a frequent complication in critically ill
patients with a high morbidity and mortality. Until recently, no evidence-based guide-
lines on the optimal treatment modality for renal replacement in the ICU could be issued
because of a lack of well-performed randomized controlled trials (RCT). Over the last
years however, some important new concepts and RCTs have been published on this topic.
An important concept is the understanding that ‘chronic dialysis strategies’ are not suitable
for acute renal failure patients in the ICU. From this understanding the necessity of daily
dialysis followed, and later on, the need for flexible treatments related to the patients’ need,
using slow long extended daily dialysis (SLEDD). Several recent papers compared contin-
uous renal replacement therapy and intermittent hemodialysis (IHD) in ICU patients,
pointing to a lack of differences in outcome, but there were less practical problems using
IHD, even in unstable patients. In conclusion, it can be stated that all patients can be
treated with IHD when available, without jeopardizing their outcome. Slow extended daily
dialysis emerged as a hybrid renal replacement therapeutic modality and has promising
features because it combines the advantages of both continuous renal replacement therapy
and IHD, but until now, no studies evaluating whether SLEDD is superior to ‘regular IHD’
are available.
Copyright © 2007 S. Karger AG, Basel
Some decades ago, continuous renal replacement therapy (CRRT) was
developed in response to the frustration of intensivists caused by the fact that
nephrologists at that time used intermittent hemodalysis (IHD) in acute renal
failure patients in the same way as they did in their chronic patients. From the
initially very simple and inefficient continuous arteriovenous hemofiltration,
CRRT developed into a very complex treatment (fig. 1) [1], which at the end
Which Treatment for AKI in the ICU
IHD for RRT in Intensive Care 305
mimicked the highly efficient intermittent hemodiafiltration of regular dialy-
sis, but without using dialysis monitors. On the other hand, nephrologists
interested in intensive care pathology invested a lot of effort to get daily dial-
ysis accepted, and to introduce slow low-efficient daily dialysis (SLEDD) [2].
The discussion whether CRRT or IHD was the preferred modality of choice
for renal replacement therapy (RRT) in the ICU was for a long time fuelled
by a lack of randomized controlled trials (RCT) and competing interests of
nephrologists and intensivists [3]. Although many efforts were made to per-
suade physicians that CRRT was the preferred modality, it was also apparent
that most nephrologists preferred IHD [4], whereas intensivists preferred
CRRT [5]. From uncontrolled trials in centers using both CRRT and IHD, it
was apparent that, if anything, the outcome of IHD was at least as good as of
CRRT, and this independent of the severity of disease (fig. 2) [6]. The first
RCT on this topic [7] ‘unexpectedly’ showed a superior outcome for the IHD
group, an effect that was attributed to problems of randomization. A meta-
analysis, however, confirmed the lack of difference in outcome concerning
the hard endpoint ‘mortality’ [8], but the hard evidence was, until recently,
lacking.
Recent Trials
A first RCT on the topic was performed by Uehlinger et al. [9]. In this sin-
gle center study, 125 patients were randomized to either CRRT (hemodiafiltra-
tion, n ⫽ 70) or IHD (n ⫽ 55). The randomization procedure can be criticized
‘Classic IHD’ 4 h 3 times/week
‘Slow (adaptable and daily)
hemodiafiltration’
‘Classic
CRRT’
CVVHD high volume
CVVHD
CVVH
CAVHD
CAVH
Classic IHD 4 h daily
Fig. 1. Evolution of ‘pure intermittent
hemodialysis’ (IHD) and ‘pure continuous
low efficient’ dialysis towards the hybrid
treatment of slow extended daily dialysis.
Van Biesen/Veys/Vanholder 306
because not all available patients (n ⫽ 191) could be randomized due to a lack
of the availability of enough machines in each of the groups to perform the
treatment the patient was randomized to at all time points of the study.
However, this situation is quite usual in most hospitals were both modalities are
offered, and should therefore be accepted as a real life condition, increasing the
generalizability of the results to everyday clinical practice. Randomization of
the included patients was, however, quite successful, as the two groups were
comparable at the start of RRT with respect to age [62 ⫾ 15 vs. 62 ⫾ 15 years,
continuous venovenous hemodiafiltration (CVVHDF) vs. IHD], gender (66 vs.
73% male sex), number of failed organ systems (2.4 ⫾ 1.5 vs. 2.5 ⫾ 1.6),
Simplified Acute Physiology Scores (57 ⫾ 17 vs. 58 ⫾ 23), septicemia (43 vs.
51%), shock (59 vs. 58%) or previous surgery (53 vs. 45%). Both modalities
were comparable with regard to the primary endpoint (mortality rates) in the
hospital (47 vs. 51%, CVVHDF vs. IHD, p ⫽ 0.72) or in the ICU (34 vs. 38%,
p ⫽ 0.71). Also hospital length of stay and duration of RRT required in the sur-
vivors was comparable in patients on CVVHDF [median (range) 20 days
(6–71), n ⫽ 36] and in those on IHD [30 days (2–89), n ⫽ 27, p ⫽ 0.25].
A second large RCT was published by the Hemodiafe group [10]. In this
large, multicentric trial, patients were randomized to either CVVHDF (n ⫽ 175)
with ⫾ 1 l/h dialysate flow and 1.3 l/h hemofiltration with predilution, with a
IHD
CRRT
% on CRRT
All
Survival (%)
0
10
20
30
40
50
60
70
⬍20 ⬎3320–25
Apache score
***
***
**
***
*
25–33
Fig. 2. CRRT vs. IHD. Ghent University Hospital 1995–1999 (N ⫽ 557) (*p ⬍ 0.05,
**p ⬍ 0.01, ***p ⬍ 0.001).
IHD for RRT in Intensive Care 307
blood flow of 150 ml/min or to IHD (n ⫽ 184), using a blood flow of 500 ml/min
and a dialysate flow of 500 ml/min. IHD was performed daily for a duration of
5.2 h. Both IHD as CRRT were performed using the same type of membranes,
excluding the fact that eventual differences in outcome might be attributable to
differences in biocompatibility. Comorbidity scores were equally distributed
among both groups. There was no difference between the two treatment modalities
concerning the primary endpoint (patient mortality) at any moment of observa-
tion. In addition, there were no differences observed between the groups for occur-
rence of hypotension or of bleeding episodes. Also the length of ICU and hospital
stay and the duration of the need for RRT were not different between the groups.
Interestingly, although it was recommended to use a low dialysate temperature
(35.5⬚C) in the IHD group, as it has indeed been suggested that the better hemo-
dynamic tolerance in CRRT was mainly due to lowering the central body temper-
ature, the number of cases suffering from hypothermia was significantly higher in
the CRRT group (p ⫽ 0.0005). Despite the lack of differences in outcome, there
was, however, a higher number of patients that was switched from CRRT to IHD
than vice versa, the major reason for transfer being bleeding-related problems. In
conclusion, the authors postulated that, if all efforts are made to optimize treat-
ment, outcome of CRRT and IHD are comparable for all patient categories.
The results of these two trials are compatible with the recently published
observational study published by the PICARD group [11]. Among 398 patients
who required dialysis, the risk of death within 60 days was examined by
assigned initial dialysis modality [CRRT (n ⫽ 206) vs. IHD (n ⫽ 192)]. Although
the study was, by definition, not randomized, differences in comorbidity were
accounted for by a propensity score approach. Crude survival rates were lower
for patients who were treated with CRRT than IHD (survival at 30 days 45 vs.
58%; p ⫽ 0.006). Adjusted for age, hepatic failure, sepsis, thrombocytopenia,
blood urea nitrogen, and serum creatinine and stratified by site, the relative risk
of death associated with CRRT was 1.82 (95% confidence interval 1.26–2.62).
Further adjustment for the propensity score did not materially alter the associa-
tion (relative risk 1.92; 95% confidence interval 1.28–2.89). Although these
results could still reflect residual confounding by the severity of illness, as it
was not a randomized trial, the results at least do not support a survival benefit
afforded by CRRT.
Conclusion
According to current knowledge, the outcome for patients needing RRT in
the ICU is not influenced by the modality used. IHD is, however, much cheaper
than CRRT, and therefore, it should be the first RRT of choice in centers where
Van Biesen/Veys/Vanholder 308
it is available. It is likely that newer hybrid techniques (SLEDD) offer an addi-
tional advantage but this remains to be proven.
References
1 Sever MS, Vanholder R, Lameire N: Management of crush-related injuries after disasters. N Engl
J Med 2006;354:1052–1063.
2 Vanholder R, Van Biesen W, Lameire N: What is the renal replacement method of first choice for
intensive care patients? J Am Soc Nephrol 2001;12(suppl 17):S40–S43.
3 Lameire N, Van Biesen W, Vanholder R, Colardijn F: The place of intermittent hemodialysis in the
treatment of acute renal failure in the ICU patient. Kidney Int Suppl 1998;66:S110–S119.
4 Abdeen O, Mehta RL: Dialysis modalities in the intensive care unit. Crit Care Clin 2002;18:
223–247.
5 Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, et al: Acute renal failure in
critically ill patients: a multinational, multicenter study. JAMA 2005;294:813–818.
6 Swartz RD, Messana JM, Orzol S, Port FK: Comparing continuous hemofiltration with hemodial-
ysis in patients with severe acute renal failure. Am J Kidney Dis 1999;34:424–432.
7 Mehta RL, McDonald B, Gabbai FB, Pahl M, Pascual MT, Farkas A, et al: A randomized clinical
trial of continuous versus intermittent dialysis for acute renal failure. Kidney Int 2001;60:
1154–1163.
8 Tonelli M, Manns B, Feller-Kopman D: Acute renal failure in the intensive care unit: a systematic
review of the impact of dialytic modality on mortality and renal recovery. Am J Kidney Dis 2002;
40:875–885.
9 Uehlinger DE, Jakob SM, Ferrari P, Eichelberger M, Huynh-Do U, Marti HP, et al: Comparison of
continuous and intermittent renal replacement therapy for acute renal failure. Nephrol Dial
Transplant 2005;20:1630–1637.
10 Vinsonneau C, Camus C, Combes A, Costa de Beauregard MA, Klouche K, Boulain T, et al:
Continuous venovenous haemodiafiltration versus intermittent haemodialysis for acute renal fail-
ure in patients with multiple-organ dysfunction syndrome: a multicentre randomised trial. Lancet
2006;368:379–385.
11 Cho KC, Himmelfarb J, Paganini E, Ikizler TA, Soroko SH, Mehta RL, et al: Survival by dialysis
modality in critically ill patients with acute kidney injury. J Am Soc Nephrol 2006;17:3132–3138.
W. Van Biesen
Renal Division, Department of Internal Medicine
University Hospital Ghent
BE–9000 Ghent (Belgium)
E-Mail wim.vanbiesen@ugent.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 309–319
Continuous Renal Replacement in
Critical Illness
Claudio Ronco
a
, Dinna Cruz
a
, Rinaldo Bellomo
b
a
Department of Nephrology, San Bortolo Hospital, Vicenza, Italy;
b
Department of Intensive Care, Austin Hospital, Melbourne, Vic., Australia
Abstract
Acute renal failure in the intensive care unit is usually part of the multiple organ dys-
function syndrome, and the complexity of illness in patients with this complication has risen
in recent years. Continuous renal replacement therapy (CRRT) was introduced in the late
1970s and early 1980s to compensate for the inadequacies of conventional intermittent
hemodialysis (IHD) in the treatment of these patients. IHD was considered aggressive and
unphysiological, often resulting in hemodynamic intolerance and limited efficiency.
Although CRRT has been shown to be physiologically superior with respect to IHD in both
observational and randomized studies, it is not clear whether this physiological superiority
translates into clinically important gains. A number of recent studies have tried to address
this issue, and with these, there is a lack of evidence to suggest improved survival and major
clinical outcomes with CRRT. However, these studies are generally underpowered and have
certain aspects which may influence the interpretation of their results. In addition, the devel-
opment of hybrid techniques, such as slow extended daily dialysis, makes this a dynamic area
of study where the terms of comparison are constantly changing. This article reviews recent
trials comparing CRRT and IHD, and discusses their results and limitations.
Copyright © 2007 S. Karger AG, Basel
Acute renal failure (ARF) affects 5–7% of all hospitalized patients [1].
Apparently, ARF continues to be associated with poor outcomes in spite of
technological advances [2]. Deeper analysis suggests that pathophysiological
mechanisms and clinical pictures are today more complex, including a large
proportion of sepsis-related cases and significant involvement of other organs.
ARF is today frequent in the intensive care unit (ICU) with a reported incidence
of 1–25% [3] depending on the population studied and the criteria used for def-
inition. Thus we observe two separate syndromes: uncomplicated ARF can be
managed outside the ICU and usually carries good prognosis with mortality
Ronco/Cruz/Bellomo 310
rates less than 10%. In contrast, complicated ARF in the ICU is part of multiple
organ dysfunction syndromes and is associated with mortality rates of 50–70%
[4]. The average outcome of the two syndromes has maintained mortality of
ARF falsely constant over the last decades. In fact, the outcome of the less
severe forms has improved while the outcome of most severe forms has
remained stable or worse. The problem is then to analyze the different case mix,
the differences in etiology of ARF and the different complexity of the syndrome
when sepsis is involved. Thus, patients who would previously have died before
they could develop ARF now survive long enough to develop ARF and complex
derangements leading to multiple organ failure and high mortality.
In this setting, careful attention is paid to the type of renal replacement
modality utilized for the treatment of ARF. Continuous (CRRT) and intermit-
tent (IHD) renal replacement modalities have often been compared without
reaching consensus on which form is optimal for the patients.
However, when comparing CRRT with IHD therapy, it is wise to remember
that the very reason for the development and introduction of CRRT into clinical
practice in the late 1970s and early 1980s was to compensate for the clear inad-
equacies of conventional IHD in the treatment of critically ill patients with mul-
tiorgan failure [5]. In the last 20 years, substantial clinical experience accumulated
in the use of CRRT and its technology evolved a great deal from adaptive arte-
riovenous techniques to dedicated CRRT machines for venovenous therapy.
During this period, multiple studies confirmed the many physiological advan-
tages of CRRT over conventional (3–4 h/day every other day as typically deliv-
ered to end-stage renal failure patients) IHD [6–11]. This was especially true in
centers which had developed appropriate and necessary physician and nursing
expertise in its application. Importantly and predictably, not a single compara-
tive study ever showed IHD to be physiologically better than CRRT.
In these years physicians have increasingly shifted from conventional IHD
to CRRT especially for critically ill patients [12]. In fact, conventional IHD has
essentially disappeared from ICUs in some countries [13]. However, the contro-
versy as to which therapy is ‘best’ from the point of view of mortality or other
major clinical outcomes (duration of ICU stay, hospital stay and rate and time to
renal recovery) remains. This is because no sufficiently powered multicenter
randomized controlled trials have yet been conducted to assess these outcomes
and because there have been major and continuing changes in what experts con-
sider optimal continuous therapy or optimal intermittent therapy.
Furthermore, all studies minimized any chance of detecting a differential
effect on outcome by allowing crossover from continuous to intermittent ther-
apy in ICU. On the other hand, looking at recent meta-analyses [14], one can
clearly see that several of these studies were performed before the year 2000
and in most of them, patients with higher severity scores were allocated to the
CRRT in Critical Illness 311
CRRT arm. Data published in the year 2000 suggested a correlation between
dose of therapy and outcome [15]. One could then speculate that populations
treated with CRRT before the year 2000 were highly underdialyzed and thus
comparisons with intermittent modalities in those times were flawed.
The concept and the objectives of RRT have evolved in these years
together with the evolution of the ARF syndrome. As intensive care patients
have become more complex, it has become clear that critically ill patients may
require specific forms of RRT. Twenty-five years ago, several of these patients
could not be dialyzed due to their hemodynamic instability or other technical
and clinical problems. In order to facilitate renal support in this setting, CRRT
techniques have been developed. Continuous arteriovenous hemofiltration was
a safe and well-tolerated treatment but efficiency was not sufficient. New
machines have subsequently made it possible to perform continuous venove-
nous hemofiltration (CVVH), hemodialysis (CVVHD) and hemodiafiltration
(CVVHDF) routinely maintaining the good clinical tolerance but allowing ade-
quate levels of blood purification [16–18]. Thus, when comparing CRRT and
IHD, one should first analyze the two therapies on the ground of physiological
endpoints.
The complexity of the patient with ARF associated with multiple organ
failure suggests that continuous therapies should be utilized as a first choice
treatment in intensive care settings. Furthermore, clinical conditions other than
ARF, such as congestive heart failure, respiratory distress syndrome, cerebral
edema and so on, may benefit from these forms of treatments when oliguria is
present or there are early associated signs of renal insufficiency.
The patient with severe hemodynamic instability often cannot tolerate
intermittent treatments such as hemodialysis or hemodiafiltration carried out
for 3–4h/day. The slow continuous fluid removal achieved with CRRT is gener-
ally well tolerated and an optimal hydration status can be reached within a rela-
tively short period of time with adequate constancy of measured hemodynamic
parameters. Direct measurements of blood volume during treatment have made
it possible to demonstrate that even in the presence of small volumes of ultrafil-
tration, a significant drop in circulating blood volume can be observed in inter-
mittent treatments. This phenomenon is not observed in continuous treatments.
This aspect may be of tremendous importance in the phase of recovery from
ARF. The recovering kidney is extremely sensitive to variations in perfusion
pressures and blood flows. Accordingly, IHD may turn out to be unphysiolog-
ical and may contribute to possible further damage to the renal parenchyma. In
contrast, CRRT may be well tolerated and may contribute to a constant and pro-
gressive recovery of the kidney without major hemodynamic alterations. These
considerations are supported by recent data from Uchino et al. [12], showing a
higher chance of renal recovery in ARF patients treated with CRRT.
Ronco/Cruz/Bellomo 312
Continuous therapies can also effectively correct various forms of acido-
sis. In fact, while IHD produces a dramatic alkalinization during treatment but
where a subsequent rebound of acidosis can be frequently observed, CRRT acts
slowly but continuously and reaches a steady-state concentration both for ure-
mic solutes and organic acids in blood.
In patients with cerebral edema, intermittent treatments may worsen the
clinical condition because of a postdialytic influx of fluid into the brain tissue.
After IHD brain density decreases due to osmotic fluid shifts. These alterations
induced by intermittent treatments are not observed with continuous therapies
that can therefore be utilized with maximal advantage in patients with or at risk
of brain edema.
Several mechanisms have been proposed to explain the improvement of
adult respiratory distress syndrome (ARDS) patients treated with CRRT. The
continuous fluid withdrawal from the interstitium due to a progressive vascular
refilling represents a major advantage. However, the modulation of the vascular
inflammation thanks to the clearance or adsorption of specific proinflammatory
substance onto the membrane has been recently hypothesized. This mechanism
has also be invoked as an interesting possibility for patients with systemic
inflammatory response syndrome or septic shock. Numerous ex vivo as well as
animal and human studies have shown that synthetic filters can extract nearly
every substance involved in sepsis to a certain degree [19]. Prominent examples
are complement factors, TNF, IL-1, IL-6, IL-8 and PAF [17]. The progressive,
continuous unselective removal of humoral mediators has been the theoretical
base to generate the ‘peak concentration hypothesis’ of sepsis and to explain the
beneficial immunomodulatory effects induced by CRRT.
Regarding plasma cytokine levels, their decrease appears minor in host
defense to infection while high levels need to be modulated by anti-inflammatory
feedback. As sepsis does not fit a one-hit model but shows the complex behav-
ior of mediator levels that change over time, neither single-mediator-directed
nor one-time interventions seem appropriate. Some studies showed no influ-
ence on cytokine plasma levels by CRRT. On the other hand, significant clinical
benefits in terms of hemodynamic improvement have been achieved even with-
out measurable decreases in cytokine plasma levels (the peak concentration
hypothesis). The removal of substances other than the measured cytokines
might have been responsible for the achieved effect. However, several media-
tors may act together to alter the functional responses of the circulating leuko-
cytes. When the response to sepsis is viewed in a network perspective, absolute
values seem to be less relevant than relative ones. Within an array of interde-
pendent mediators, even small decreases could induce major balance changes.
In spite of some encouraging results as already mentioned, the extent of achiev-
able clinical benefit with conventional CRRT (using conventional filters and
CRRT in Critical Illness 313
flow rates) in sepsis has generally been disappointing. Consequently, attempts
have been made to improve the efficiency of soluble mediator removal in sepsis
by increasing the amount of plasma water exchange, i.e. increasing ultrafiltra-
tion rates. Animal studies provide great support to this concept. These studies
established that a convection-based treatment can remove substances which can
induce hemodynamic effects resembling septic shock, when sufficiently high
ultrafiltration rates are applied. More relevant to human sepsis was the finding
that ultrafiltration dosage is correlated to outcome in critically ill patients with
ARF. In a large randomized, controlled study including 425 patients, an ultrafil-
tration dosage of 35ml/kg/h increased survival rate from 41 to 57% compared
to a dosage of 20 ml/kg/h [15]. Eleven to 14% (per randomization group) of the
patients had sepsis. In these subgroups there was a trend of a direct correlation
between treatment dosage and survival even above 35 ml/kg/h in contrast to the
whole group where a survival plateau was reached. Of note, there was no increase
in adverse effects even with the highest ultrafiltration dosage. Impressive clini-
cal results were obtained in an evaluation of short-term high-volume hemofil-
tration (HVHF) in patients with catecholamine-refractory septic shock [18]
comprising a patient cohort with very poor expected survival. A control group
was not defined. Only one 4-hour session of HVHF removing 35 l of ultrafil-
trate replaced by bicarbonate-containing fluid was applied as soon as mean
blood pressure could not be stabilized above 70 mm Hg with dopamine, norepi-
nephrine and epinephrine after appropriate volume resuscitation. HVHF was
followed by conventional CVVH.
With the extracorporeal fluxes used for CRRT, negative thermal balance of
up to –100 kJ/h can be obtained depending on the length of the blood lines, the
room temperature and the dialysate/replacement fluid temperature. This might
contribute to modulating the inflammatory response as well as oxygen demand
in several organs with the possibility of using such a mechanism for specific
clinical targets.
When electrolyte disorders are life threatening or refractory to medical
corrections, extracorporeal therapy is the treatment of choice. Because of the
slow and gentle rate of fluid exchange, the treated blood operates in continuous
equilibrium with peripheral tissues and organs, and the entire organism may
benefit from a safe and effective restoration of water, sodium and electrolyte
homeostasis. This restoration of homeostasis is particularly true for acid-base
control (as administration of bicarbonate can be easily titrated to the necessary
acid-base goals), intra-/extracellular potassium and phosphate equilibrium and
water fluxes between the interstitium and the intracellular space.
After evaluating the physiological effects of continuous and intermittent
treatment modalities, and considering the potential advantages of continuous
therapies, one should still face the lack of evidence for improved survival and
Ronco/Cruz/Bellomo 314
major clinical outcomes of one technique over another. In this setting the lack
of evidence may induce physicians to draw false conclusions from a misleading
interpretation of the available literature. In order to clarify the real content of
the available trials, it is still useful to review some of their aspects even in the
presence of the above-mentioned shortcomings. Mehta et al. [20] randomized
166 critically ill patients with severe acute kidney injury to either CRRT or IHD
therapy. There was a significantly higher ICU mortality rate in subjects ran-
domized to CRRT (60 vs. 42%, p ⫽ 0.02). After post hoc adjustment for sever-
ity of illness, the increased risk attributed to CRRT was no longer statistically
significant (odds ratio 1.6). There was clear baseline imbalance with patients
randomized to CRRT having greater illness severity (higher APACHE III and
grater incidence of liver failure). The reasons for such imbalances remain
unclear. However, several other aspects of this study are still of interest. First,
patients were allowed to cross over making a true comparison impossible.
Second, patients with hemodynamic instability (MAP ⬍ 70 mm Hg) were
excluded. These are the very patients where the advantages of CRRT are most
evident. This selection bias was the expression of an acknowledged lack of
equipoise: in such patients intermittent therapy is physiologically inferior.
Third, despite these limitations, one very relevant observation emerged: if
patients received a sufficient trial of CRRT and survived, renal recovery was
dramatically increased (92.3 vs. 59.4%, p ⬍ 0.01). In other words, intermittent
therapy delayed or impeded renal recovery. Fourth, CRRT delivered superior
control of uremia.
In another single-center randomized trial, 125 patients were randomized to
CRRT or IHD. Inhospital mortality rates did not differ by treatment assignment
(CRRT 47 vs. 51% for IHD, p ⫽ 0.72) [21]. However, this trial suffered from
extraordinary logistic constraints in that patients could not be randomized on a
1:1 basis because of untrained staff in CRRT or a lack of CRRT machines. This
is hardly the environment that would create a scientific ‘level playing field’ for
the two therapies. More importantly, if the near 4% absolute decrease in mor-
tality still seen with CRRT were true, it would have taken 5,000 patients to
detect it at a beta of 0.2 and an alpha of 0.05. The investigators randomized only
125 patients! Such a 4% decrease might seem small but would be clinically
relevant because it would imply that the number needed to treat to save one life
is only 25. The number needed to treat to save one life with percutaneous coro-
nary intervention in patients with myocardial infarction with ST segment eleva-
tion is 50 [22].
In a third single-center prospective randomized trial, 80 patients were ran-
domized to treatment with CRRT or IHD [23]. Survival was 67.5% for CRRT
and 70% for dialysis. Again the study was dramatically underpowered to detect
differences in survival between the two therapies and crossover from CRRT to
CRRT in Critical Illness 315
IHD occurred in 22.5% of CRRT patients. Of interest, this trial confirmed the
hemodynamic problems associated with IHD with 40% of patients requiring an
increase in vasopressor therapy during initial treatment with IHD compared to
only 12.5% for CRRT (p ⫽ 0.005). In this study, CRRT was associated with a
6-day (close to 15%) reduction in hospital stay. Nine patients were converted
from CRRT to IHD because of frequent filter clotting, a concept that seems
strange to practitioners who have only used CRRT in the well-established era of
citrate and regional heparin/protamine anticoagulation.
Most recently, the HEMODIAF Study Group reported the results of a
randomized multicenter study comparing the results of CRRT to IHD in 360
critically ill patients with acute kidney injury [24]. Overall, there was no differ-
ence in the primary endpoint of 60-day survival (33% with CRRT vs. 32% with
IHD). However, the authors noted an unexpected and significant increase in
survival rates within the IHD group over time (relative risk 0.67/year), an effect
not seen in the CRRT group. This creates a major bias. How was IHD changed?
Who changed it? Was advice offered to trial sites? This is of major concern
because if it had not happened and the initial trend in outcome seen with IHD
had continued, the conclusions would have been diametrically opposite. In this
regard, it is of interest to see how the average duration of dialysis became 5.2 h
(significantly longer than conventional IHD), while CRRT continued to deliver
a calculated creatinine clearance of 25 ml/kg/h (not 29 ml/kg/h as reported by
the authors who did not correct for the effects of predilution). Such clearance is
much lower than what is considered an ‘optimal’ CRRT dose [15]. Finally,
again, this trial reported a 10% crossover from CRRT to IHD.
Recently, the PICARD group compared the outcomes of different renal
replacement therapy modalities [25]. This analysis incorporated five sites in the
USA and used multivariable regression analysis, and a propensity score
approach to address the effect of confounding variables. Within the PICARD
cohort, using this methodology, the provision of CRRT in comparison to IHD
was associated with a significantly higher relative risk of mortality. However,
patients with CRRT were obviously sicker: they had more organ failure (CNS,
liver, hematologic and respiratory), higher mechanical ventilation rate, more
sepsis, lower blood pressure, higher total bilirubin and lower platelet count.
Despite this, none of the respiratory and circulatory variables, like mechanical
ventilation requirement, ARDS, heart rate or blood pressure, were included in
the multiple regression analysis for mortality, although these variables had uni-
variate p values of ⬍0.0001. It is quite hard to understand why these variables
were not selected as independent variables. These observations cast serious
doubt on the accuracy of this analysis.
Thus, there is clear lack of suitably designed multicenter randomized con-
trolled trials where all ICU patients with ARF are randomized to either CRRT
Ronco/Cruz/Bellomo 316
or IHD (doses to be defined) from start to finish. Until such a trial is done, the
question of clinical superiority cannot be answered.
Whatever future trials might wish or be able to compare, there is clear evi-
dence that RRT is not like a ‘tablet’. Its effects are modified by the expertise of
those who prescribe it and guide it (physicians) and those who execute it
(nurses). For example, it is clear from the HEMODIAF study that the quality of
IHD can be improved and better outcomes follow. Another study has recently
confirmed that priming the dialysis circuit with isotonic saline, setting dialysate
sodium concentration above 145 mmol/l, discontinuing vasodilator therapy, and
setting dialysate temperature below 37⬚C [26] improves IHD-related outcomes.
Finally, the introduction of slow extended daily dialysis (SLEDD) introduces the
final step in the rehabilitation of dialysis in ICU [27]. Indeed as IHD becomes
more and more like CRRT through SLEDD, the protagonists of CRRT will be
filled with delight: their battle was not with dialysis per se, but rather with the
mindless application of it in conventional mode to critically ill patients. Once
IHD in ICU is reformed, adjusted to take into account the needs of the critically
ill, and extended to 6 h (or 8 or maybe even 12 h) so that fluid removal is per-
formed safely and uremic control optimized, little of the controversy will remain.
As a continuous extracorporeal therapy, CRRT frequently requires continu-
ous anticoagulation, which may increase the bleeding risk. Thus, it also needs
‘reform’. Citrate and regional heparin/protamine anticoagulation are safe and
effective, but underused. This might represent a chance of improvement for the
future. Machine operation and safety should be given more attention especially
providing more education and personnel training [28]. Finally and more impor-
tantly, the dose of CRRT might well require readjustment with two recent ran-
domized controlled trials showing that an increasing dose of CRRT improves
survival [29]. All studies comparing IHD to CRRT have used much lower
CRRT doses than those shown to improve survival. The optimal weekly Kt/V
for CRRT in an 80-kg man would appear to be close to 10. If multicenter trials
currently under way confirm this observation, CRRT will also have to be reformed.
More importantly, by implication, its twin (IHD) will have to be reformed fur-
ther as well to be able to deliver such an optimal dose intermittently. Both tech-
niques also have to look further than the dose. The issue of timing of
intervention is likely very important and yet not well studied so far [30]. Timing
of cessation may be equally important but has not yet been studied. Much work
needs to be done in the field of acute RRT.
In conclusion, there is a reason for the human kidney working 24 h/day.
The reason is to maintain homeostasis and to keep biological parameters within
a tight steady-state control. Intermittent forms of renal replacement therapies
are aggressive and unphysiological since they must accomplish their task in a
short fraction of the day trying to correct in a few minutes derangements that
CRRT in Critical Illness 317
have developed over hours or days. The result is a severe hemodynamic intoler-
ance, a limited efficiency for waste product elimination, a high risk of brain
edema, worsening of inflammatory conditions and a complete failure to keep
acid-base and electrolytes in steady condition.
In contrast, continuous therapies perform a gentle and slow correction of
derangements leading to a steady-state condition close to that provided by
native kidneys. Most of the claimed drawbacks or complications can be pre-
vented if adequate prescription and monitoring are carried out and the right
treatment dose is delivered.
All these considerations apply to critically ill ICU patients with ARF. In
these patients, intermittent dialysis can be problematic or even impossible to
perform. The opposite may be true for patients in renal wards with uncompli-
cated ARF where daily (possibly long) hemodialysis sessions are perfectly
capable to accomplish the required tasks. Even in these cases however, an inter-
mediate form between fully intermittent and fully continuous dialysis regimes
is strongly advisable since ARF patients are still very different from typical
chronic hemodialysis patients.
The evidence that CRRT is physiologically superior to conventional IHD is
clear. The evidence that this physiological superiority can be translated into
clinically important gains is not. Appropriately powered, designed, conducted
and analyzed studies have simply not yet been done. In addition the evolution of
both therapies presents a fluid environment where the terms of comparison con-
stantly change. The correct focus for clinicians might actually be not so much to
compare the two, but to make sure that whatever therapy is applied, it is ‘done
right’. They need to ensure that patients receive the best therapy for their condi-
tion at a given time in the course of their illness in ICU, receive it safely, in a
timely fashion, with the correct dose and for long enough.
References
1 DuBose TD Jr, Warnock DG, Mehta RL, Bonventre JV, Hammerman MR, Molitoris BA, Paller MS,
Siegel NJ, Scherbenske J, Striker GE: Acute renal failure in the 21st century: recommendations for
management and outcomes assessment. Am J Kidney Dis 1997;29:793–799.
2 Clermont G, Acker CG, Angus DC, Sirio CA, Pinsky MR, Johnson JP: Renal failure in the ICU:
comparison of the impact of acute renal failure and end-stage renal disease on ICU outcomes.
Kidney Int 2002;62:986–996.
3 Liano F, Junco E, Pascual J, Madero R, Verde E: The spectrum of acute renal failure in the inten-
sive care unit compared with that seen in other settings. The Madrid Acute Renal Failure Study
Group. Kidney Int 1998;66:S16–S24.
4 De Mendonca A, Vincent J-L, Suter PM, et al: Acute renal failure in the ICU: risk factors and out-
come evaluated by the SOFA score. Intensive Care Med 2000;26:915–921.
5 Kramer P, Wigger W, Rieger J, Matthaei D, Scheler F: Arteriovenous haemofiltration: a new and
simple method for treatment of over-hydrated patients resistant to diuretics. Klin Wochenschr
1977;55:1121–1122.
Ronco/Cruz/Bellomo 318
6 Bellomo R, Boyce N: Continuous veno-venous hemodiafiltration compared with conventional
dialysis in critically ill patients with acute renal failure. ASAIO J 1993;39:M794–M797.
7 Bellomo R, Martin H, Parkin G, Love J, Boyce N: Continuous arteriovenous haemodiafiltration in
the critically ill. Influence on major nutrient balances. Intensive Care Med 1991;17:399–402.
8 Bellomo R, Farmer M, Bhonagiri S, Porceddu S, Ariens M, M’pisi D, Ronco C: Changing acute
renal failure treatment from intermittent hemodialysis to continuous hemofiltration: impact on
azotemic control. Int J Artif Organs 1999;22:145–150.
9 Ronco C, Bellomo R, Brendolan A, Pinna V, La Greca G: Brain density changes during renal
replacement therapy in critically ill patients with acute renal failure. Continuous hemofiltration
versus intermittent hemodialysis. J Nephrol 1999;12:173–178.
10 Tan HK, Bellomo R, M’pisi DA, Ronco C: Phosphatemic control during acute renal failure: inter-
mittent hemodialysis vs. continuous hemodiafiltration. Int J Artif Organs 2001;24:186–191.
11 Manns M, Sigler MH, Teehan BP: Intradialytic renal haemodynamics – potential consequences
for the management of the patient with acute renal failure. Nephrol Dial Transplant 1997;12:
870–872.
12 Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C,
Macedo E, Gibney N, Tolwani A, Ronco C: Acute renal failure in critically ill patients: a multina-
tional, multicenter study. JAMA 2005;294:813–818.
13 Silvester W, Bellomo R, Cole L: The epidemiology, management and outcome of severe acute
renal failure of critical illness in Australia. Crit Care Med 2001;29:1910–1915.
14 Kellum JA, Angus DC, Johnson JP, Leblanc M, Griffin M, Ramakrishnan N, Linde-Zwirble WT:
Continuous versus intermittent renal replacement therapy: a meta-analysis. Intensive Care Med
2002;28:29–37.
15 Ronco C, Bellomo R, Homel P, et al: Effects of different doses in continuous veno-venous
haemofiltration on outcomes of acute renal failure: a prospective randomised trial. Lancet 2000;356:
26–30.
16 Ronco C, Bellomo R: Continuous versus intermittent renal replacement therapy in the treatment
of acute renal failure. Nephrol Dial Transplant 1998;13/6:79–85.
17 Ronco C, Tetta C, Lupi A, et al: Removal of platelet-activating actor in experimental continuous
arteriovenous hemofiltration. Crit Care Med 1995;23:99–107.
18 Honore PM, Jamez J, Wauthier M, et al: Prospective evaluation of short-term, high-volume isov-
olemic hemofiltration on the hemodynamic course and outcome in patients with intractable circu-
latory failure resulting from septic shock. Crit Care Med 2000;28:3581–3587.
19 Bellomo R, Tipping P, Boyce N: Continuous veno-venous hemofiltration with dialysis removes
cytokines from the circulation of septic patients. Crit Care Med 1993;21:522–526.
20 Mehta RL, McDonald B, Gabbai FB, et al: A randomized clinical trial of continuous versus inter-
mittent dialysis for acute renal failure. Kidney Int 2001;60:1154–1163.
21 Uehlinger DE, Jakob SM, Ferrair P, et al: Comparison of continuous and intermittent renal
replacement therapy for acute renal failure. Nephrol Dial Transplant 2005;20:1630–1637.
22 Keeley EC, Hillis LD: Primary PCI for myocardial infarction with ST-segment elevation. N Engl J
Med 2007;356:47–54.
23 Augustine JJ, Sandy D, Seifert TH, Paganini EP: A randomized controlled trial comparing inter-
mittent with continuous dialysis in patients with ARF. Am J Kidney Dis 2004;44:1000–1007.
24 Vinsonneau C, Camus C, Combes A, et al: Continuous venovenous haemodiafiltration versus
intermittent haemodialysis for acute renal failure in patients with multiple-organ dysfunction syn-
drome: a multicentre randomised trial. Lancet 2006;368:379–385.
25 Cho KC, Himmelfarb J, Paganini E, et al: Survival by dialysis modality in critically ill patients
with acute kidney injury. J Am Soc Nephrol 2006;17:3132–3138.
26 Schortgen F, Soubrier N, Delclaux C, et al: Hemodynamic tolerance of intermittent hemodialysis
in critically ill patients: usefulness of practice guidelines. Am J Respir Crit Care Med 2000;162:
197–202.
27 Marshall MR, Ma T, Galler D, et al: Sustained low-efficiency daily diafiltration (SLEDD-f) for
critically ill patients requiring renal replacement therapy towards an adequate therapy. Nephrol
Dial Transplant 2004;19:877–884.
CRRT in Critical Illness 319
28 Schultz DG: FDA Updated Public Health Notification: Gambro Prisma
®
Continuous Renal
Replacement System. September 22, 2006. www.fda.gov/cdrh/safety/022706-gambro.html
29 Bellomo R: Do we know the optimal dose for renal replacement therapy in the intensive care unit?
Kindey Int 2006;70:1202–1204.
30 Bent P, Tan HK, Bellomo R, Buckmaster J, Doolan L, Hart G, Silvester W, Gutteridge G, et al:
Early and intensive continuous veno-venous hemofiltration for severe acute renal failure after car-
diac surgery. Ann Thorac Surg 2001;71:832–837.
Claudio Ronco
Department of Nephrology, San Bortolo Hospital
Viale Rodolfi 37
IT–36100 Vicenza (Italy)
Tel. ⫹39 0444 753650, Fax ⫹39 0444 753973, E-Mail cronco@goldnet.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 320–324
Sustained Low-Efficiency Dialysis
Ashita J. Tolwani, Thomas S. Wheeler, Keith M. Wille
University of Alabama at Birmingham, Birmingham, Ala., USA
Abstract
Sustained low-efficiency dialysis (SLED) is an increasingly popular form of renal
replacement therapy for patients with renal failure in the intensive care unit. Advantages of
SLED are efficient clearance of small solutes, good hemodynamic tolerability, flexible treat-
ment schedules, and reduced costs. Studies comparing outcomes of SLED with those of
other dialysis modalities are being performed.
Copyright © 2007 S. Karger AG, Basel
Hybrid hemodialysis, also known as sustained low-efficiency dialysis
(SLED), extended daily dialysis (EDD), or slow continuous dialysis, has
emerged as an alternative to intermittent hemodialysis (IHD) and continuous
renal replacement therapy (CRRT) in the treatment of renal failure in the
intensive care unit (ICU) setting. Hybrid hemodialysis is utilized in an
increasing number of hospitals in the US, Europe, South America, Asia, New
Zealand and Australia [1]. Pesacreta et al. [2] interviewed 131 physicians in
27 medical centers and found SLED to be the primary therapy for ICU renal
replacement therapy in 7% of patients compared to 56% for IHD and 35% for
CRRT.
Hybrid hemodialysis, which was first described in 1988, utilizes an IHD
machine to perform dialysis for extended periods of time using low blood flow
and dialysate rates of 100–300 ml/min. In contrast to CRRT, hybrid modalities
are not continuous and usually run for 6–12 h/day. Hybrid hemodialysis
employs characteristics of both CRRT and IHD, hence, the use of the term
‘hybrid’. The rationale for its use is its ability to provide improved hemody-
namic stability as in CRRT and high solute clearances and flexible scheduling
as in IHD without the need for expensive CRRT machines, costly customized
solutions, and trained staff.
SLED 321
Technical Issues
Machinery
In the US, the Fresenius 2008H/K
®
model hemodialysis machines are
capable of delivering the low dialysate and blood flow rates required of hybrid
modalities [3, 4]. Outside of the US, the Fresenius 4008S ArRT Plus
®
,
Fresenius Genius
®
, and Gambro 200S Ultra
®
are used to deliver hybrid therapy
[3, 4]. The machinery can be categorized as single pass machines or batch
machines. Single pass machines use dialysate generated on-line from reverse
osmosis purified water and a bicarbonate proportioning system. The dialysate
in batch machines is generated from prepackaged salts and sterile water and is
stored within the machine.
Prescription
Duration and Timing of Treatment
The duration of SLED therapy can be individualized according to the
needs of patients. Treatment durations may range from 6 to 18 h/day. In a study
of 20 patients on SLED vs. 19 patients on CRRT, Kielstein et al. [5] found that
the urea reduction ratio was similar between the two groups [continuous ven-
ovenous hemofiltration (CVVH) therapy 53 ⫾ 2%; EDD therapy 52 ⫾ 3%].
Although urea reduction ratio was equivalent, EDD patients were dialyzed for
11.7 ⫾ 0.1 h compared with 23.3 ⫾ 0.2 h for CRRT patients. This suggests
that the effect of 12 h of SLED is comparable to 23 h of CRRT.
Since SLED is not a continuous modality, the timing of therapy can be tai-
lored to the benefit of the patient. Many centers now perform nocturnal SLED
so that patients may be available during the day for various diagnostic and ther-
apeutic procedures, thereby avoiding interruptions of therapy [4].
Dialysate and Ultrafiltration Rates
Published studies use dialysate flow rates ranging from 100–300 ml/min,
depending on the dialysis machine specifications, treatment duration, and toler-
ance of ultrafiltration. Higher dialysate flows of 300 ml/min are generally used
with treatments of less than 8h, while lower dialysate flow rates are used with
longer treatments. The ultrafiltration rate is varied according to the patient’s
clinical need and hemodynamic stability.
Dialysate Composition
The composition of the dialysate varies according to clinical needs. Typical
dialysate baths consist of 3.0–4.0 mEq/l potassium, 1.5–2.5 mEq/l calcium, and
24–35 mmol/l bicarbonate. Phosphate removal with hybrid therapies can be
extensive, and phosphorus can be repleted intravenously or by adding 45ml of
Tolwani/Wheeler/Wille 322
Fleets Phosphasoda to 9.5 l of bicarbonate concentrate, giving a final concen-
tration of 0.81 mmol/l [6].
A potential concern with on-line dialysate generation as used in the single
pass machine is the possibility of backfiltration of endotoxin from the dialysate
compartment into the patient. However, insufficient data exist as to the recom-
mended purity of the dialysate.
Anticoagulation
The reported incidence of SLED circuit clotting without anticoagulation is
26–46% using single pass machines and much less using batch systems [7–9].
If anticoagulation is required, unfractionated heparin is most commonly given
as a 1,000- to 2,000-unit bolus followed by a maintenance dose of 500–1,000units/h
with a goal APTT of 1.5 times baseline. With heparin, the reported incidence of
clotting is 17–26% [10].
Several regional citrate anticoagulation protocols have been published.
These utilize a zero calcium dialysate with intravenous calcium replacement or
a low calcium dialysate without intravenous calcium replacement [11–13].
Case reports using prostacyclin and argatroban for anticoagulation have also
been published [10, 14].
Clinical Outcomes
Solute Control
SLED, compared to conventional IHD, offers greater small solute clearance
(Kt/V 1.3–1.5), less small solute disequilibrium (single pool urea kinetic model-
ing), and, with high-flux dialyzers, greater large solute clearance [5]. Kinetic
models indicate that both SLED and CRRT provide effective control of azotemia
in hypercatabolic acute kidney injury (AKI) patients [15]. SLED, however, is
less effective than CRRT for larger solute control [5]. Sustained low efficiency
diafiltration (SLEDD-f), a variation of SLED, combines diffusive and convec-
tive solute transport to improve clearance of larger molecular weight solutes [9].
Of note, while albumin removal is minimal with SLED, amino acid losses
are significant, and protein supplementation of 0.2 g/kg/day should be adminis-
tered on treatment days.
Hemodynamic Tolerance
Reports of hybrid therapy have consistently shown that ultrafiltration is gen-
erally well tolerated. Only a minority (0–7%) of patients in all published reports
had to discontinue hybrid treatment because of refractory hypotension. Kumar
et al. [6] prospectively compared SLED (n ⫽ 25) with CVVH (n ⫽ 17) and found no
difference in mean arterial pressure or net daily ultrafiltration rates but significant
SLED 323
differences in treatment duration (7.5 h with SLED vs. 19.5 h with CVVH) and
anticoagulation requirement (median heparin dose 4,000 U/day with SLED vs.
21,000 U/day with CVVH). Similarly, a prospective controlled trial, in which
39 ICU patients were randomized to either hybrid therapy or CRRT, found no sig-
nificant difference in hemodynamic parameters, inotrope dose, or outcome [5].
Mortality
Marshall et al. [8] performed SLED using a Fresenius 2008H
®
IHD
machine at a dialysate flow rate of 100 ml/min. Sufficient solute removal and
ultrafiltration were achieved in most patients, and hospital mortality (62%) did
not exceed the predicted mortality rate based on the acute physiology and
chronic health evaluation (APACHE II) score.
Currently, the VA/NIH Acute Renal Failure Trial Network (ATN) Study, a
prospective, multicenter study, is recruiting ICU patients with AKI and randomiz-
ing them to high- vs. low-dose dialysis [16]. Since patients may undergo intermit-
tent hemodialysis, CRRT, or hybrid therapy, this study may allow for outcome
comparisons using the different modalities of therapy. The Stuivenberg Hospital
Acute Renal Failure (SHARF) study, which is ongoing, is a prospective, random-
ized multicenter clinical trial comparing hybrid therapy with CRRT for patients
with AKI. No significant outcome differences were observed following an interim
analysis of 996 patients, but continued patient recruitment is underway [17].
Cost
Hybrid therapies are less expensive than CRRT. This is due in part due to
avoidance of expenses associated with CRRT machinery and preparation of
specialized fluids. Studies have found that the daily cost of SLED may be up to
8 times less expensive than CRRT [18–20].
Conclusion
Hybrid hemodialysis has proven to be a viable option in the treatment of
renal failure since it combines the relatively low cost and complexity of IHD
with the advantages of gradual fluid and solute removal of CRRT. In addition,
its intermittent nature allows for scheduling of other diagnostic and therapeutic
procedures between treatments.
References
1 Fliser D, Kielstein JT: Technology insight: treatment of renal failure in the intensive care unit with
extended dialysis. Nat Clin Pract Nephrol 2006;2:32–39.
Tolwani/Wheeler/Wille 324
2 Pesacreta M, Overberger P, Palevsky PM, the VA/NIH Acute Renal Failure Trial Network:
Management of renal replacement therapy in acute renal failure: a survey of practitioner prescrib-
ing practices. J Am Soc Nephrol 2004;15:350A.
3 Fliser D, Kielstein JT: A single-pass batch dialysis system: an ideal dialysis method for the patient
in intensive care with acute renal failure. Curr Opin Crit Care 2004;10:483–488.
4 Lonnemann G, Floege J, Kliem V, Buckhurst R, Koch K: Extended daily veno-venous high-flux
haemodialysis in patients with acute renal failure and multiple organ dysfunction syndrome using
a single path batch dialysis system. Nephrol Dial Transplant 2000;15:1189–1193.
5 Kielstein J, Kretschmer U, Ernst T, Hafer C, Bahr M, Haller H, Fliser D: Efficacy and cardiovas-
cular tolerability of extended dialysis in critically ill patients: a randomized controlled study. Am J
Kidney Dis 2004;43:342–349.
6 Kumar VA, Yeun JY, Depner TA, Don BR: Extended daily dialysis vs. continuous hemodialysis for
ICU patients with acute renal failure: a two-year single center report. Int J Artif Organs 2004;27:
371–379.
7 Kumar V, Craig M, Depner T, Yeun J: Extended daily dialysis: a new approach to renal replacement
for acute renal failure in the intensive care unit. Am J Kidney Dis 2000;36:294–300.
8 Marshall M, Golper T, Shaver M, Alam M, Chatoth D: Sustained low-efficiency dialysis for criti-
cally ill patients requiring renal replacement therapy. Kidney Int 2001;60:777–785.
9 Marshall M, Tianmin M, Galler D, Rankin A, Williams A: Sustained low-efficiency daily diafil-
tration (SLEDD-f) for critically ill patients requiring renal replacement therapy: towards an ade-
quate therapy. Nephrol Dial Transplant 2004;19:877–884.
10 Marshall M, Golper T: Sustained low efficiency or extended daily dialysis. Up-to-date 2006,
online version 13.3. http://www.uptodate.com
11 Finkel K, Foringer J: Safety of regional citrate anticoagulation for continuous sustained low effi-
ciency dialysis (C-SLED) in critically ill patients. Ren Fail 2005;27:541–545.
12 Marshall MR, Ma TM, Eggleton K, Ferencz A: Regional citrate anticoagulation during simulated
treatments of sustained low efficiency diafiltration. Nephrology 2003;8:302–310.
13 Morgera S, Scholle C, Melzer C, Slowinski T, Liefeld L, Baumann G, Peters H, Neumayer H:
A simple, safe, and effective citrate anticoagulation protocol for the Genius dialysis system in
acute renal failure. Nephron Clin Pract 2004;98:35–40.
14 Fiaccadori E, Maggiore U, Parenti E, Giacosa R, Picetti E, Rotelli C, Tagliavini D, Cabassi A:
Sustained low-efficiency dialysis (SLED) with prostacyclin in critically ill patients with acute
renal failure. Nephrol Dial Transplant 2006;22:529–537.
15 Liao Z, Zhang W, Hardy PA, et al: Kinetic comparison of different acute dialysis therapies. Artif
Organs 2003;27:802.
16 Palevsky PM, O’Connor T, Zhang JH, Star RA, Smith MW: Design of the VA/NIH Acute Renal
Failure Trial Network (ATN) Study: intensive versus conventional renal support in acute renal fail-
ure. Clin Trials 2005;2:423.
17 Malbrain M, Elseviers M, Van der Niepen P, et al: Interim results of the SHARF4 Study: outcome
of acute renal failure with different modalities (abstract). Crit Care 2004;8(suppl 1):153.
18 Alam M, Marshall M, Shaver M, Chatoth D: Cost comparison between sustained low efficiency
hemodialysis (SLED) and continuous venovenous hemofiltration (CVVH) for ICU patients with
ARF (abstract). Am J Kidney Dis 2000;35:A9.
19 Ma T, Walker R, Eggleton K, Marshall M: Cost comparison between sustained low efficiency
dialysis/diafiltration (SLEDD) and continuous renal replacement therapy for ICU patients with
ARF (abstract). Nephrology 2002;7:A54.
20 Berbece A, Richardson R: Sustained low-efficiency dialysis in the ICU: cost, anticoagulation, and
solute removal. Kidney Int 2006;70:963–968.
Ashita J. Tolwani, MD
ZRB 604
1530 3rd Ave. S.
Birmingham, AL 35294-0007 (USA)
Tel. ⫹1 205 975 2021, Fax ⫹1 205 996 2156, E-Mail atolwani@uab.edu
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 325–332
The Role of the International Society of
Nephrology/Renal Disaster Relief Task Force
in the Rescue of Renal Disaster Victims
R. Vanholder
a
, W. Van Biesen
a
, N. Lameire
a
, M.S. Sever
b
a
Renal Division, Department of Internal Medicine, University Hospital, Ghent,
Belgium;
b
Departments of Nephrology and Internal Medicine, Istanbul School of
Medicine, Istanbul, Turkey
Abstract
Disasters are a major cause of distress and material as well as corporal damage. Next to
direct trauma, the crush syndrome inducing multiorgan problems as a consequence of mus-
cle compression and the release of muscular contents into the bloodstream is the most impor-
tant cause of death; this is to a large extent related to the induction of severe acute kidney
injury, for which dialysis is a life-saving therapy. The practical means (both hardware and
personnel) to do so are, however, often lacking in disaster conditions. The Renal Disaster
Relief Task Force (RDRTF) offered support for renal problems in the aftermath of several
disasters, e.g. the Marmara earthquake (1999) in Turkey, the Bam earthquake (2003) in Iran,
and the Kashmir earthquake (2005) in Pakistan. A preconceived intervention plan is followed
with adaptations according to local conditions. Material and personnel are dispatched to the
disaster areas. These interventions have been life-saving for a substantial number of victims.
The current article describes the structure and approach of the RDRTF.
Copyright © 2007 S. Karger AG, Basel
The Concept of Disaster
Disasters are sudden calamities provoking substantial damage, loss and
distress, and most often strike extended areas. At first instance, natural disas-
ters, e.g. earthquakes, come to mind, but they can be man-made as well. If a nat-
ural or man-made disaster results in the collapse of buildings or other solid
structures, this might result in the entrapment of numerous victims under the
rubble, which in turn may be the origin of the crush syndrome (CS; see below).
In some disasters, a substantial number of victims suffering from the CS will
Vanholder/Van Biesen/Lameire/Sever 326
subsequently develop acute kidney injury (AKI). This sequence of events has
been defined as renal disaster [1]. During the past few years, several earth-
quakes could be classified as renal disasters: the Armenian earthquake in 1988
[2], the Marmara earthquake in Turkey (1999) [3], the Bam earthquake in Iran
(2003) [4], and the Kashmir earthquake in Pakistan and surrounding countries
(2005) [5].
In this review, we define the CS and its pathophysiology; we then describe
the concept of renal disaster and of the Renal Disaster Relief Task Force
(RDRTF), and finally the measures to be taken to potentially limit the number
of renal victims.
The CS and Rhabdomyolysis
Crush injury develops due to entrapment under the rubble and the ensuing
compression of the muscular system. If these events give rise to systemic man-
ifestations, such as the systemic inflammatory reaction syndrome, the adult res-
piratory distress syndrome, diffuse intravascular coagulation or shock, the term
CS is justified. One of the most devastating epiphenomena of the CS is AKI.
The latter is to a large extent the consequence of rhabdomyolysis, a term which
refers to the disintegration of striated muscle as a consequence of muscle
compression.
It might seem as if the CS affects only a minor fraction of the total poten-
tial number of disaster victims, e.g. if compared to the overall number of fatali-
ties. However, no support can be given to those who die immediately and no
substantial help is needed for those with minor injuries; this leaves the propor-
tionally small group of subjects with CS as the one in whom intervention, espe-
cially by dialysis, might increase survival chances. Actually, CS is the second
most frequent cause of death after direct trauma, and the only potentially fatal
complication for which life-saving therapy is available.
Pathophysiology of AKI in Rhabdomyolysis
One of the central pathophysiological events in rhabdomyolysis is the
sequestration of large quantities of body water in the damaged muscles [6],
which can easily amount to a total volume of 5 l or more. This process induces
intravascular dehydration and is one of the main elements at the origin of AKI.
Together with water, calcium is also attracted into the damaged muscles.
This may sometimes give rise to metastatic calcification [7]. The subsequent
hypocalcemia is a potential cause of heart failure and/or cardiac arrhythmias
ISN and RDRTF in Renal Disaster Victims 327
which provoke further ischemic kidney damage. Simultaneously, several other
compounds with toxic impact are released from the damaged muscles. The
most important one here is myoglobin, which induces tubular obstruction, espe-
cially together with dehydration. At the same time, hepatic metabolism of myo-
globin generates bilirubin which also has a nephrotoxic potential.
Release of potassium out of the damaged muscle cells induces hyper-
kalemia. Its deleterious effect is potentialized by the hypocalcemia. Other com-
pounds released from the muscles are phosphate aggravating hypocalcemia,
acids enhancing hyperkalemia, and nucleic acids, which are metabolized to uric
acid, another obstructive nephrotoxin.
The Concept of Renal Disaster
In many of the areas recently struck by major earthquakes, the potential to
perform dialysis was present, but the facilities to deal appropriately with a large
number of crush victims were restricted for socioeconomic and/or logistic reasons.
The first awareness of a major kidney problem subsequent to an earth-
quake occurred after the 1988 Armenian earthquake in Spitak. Although a vast
quantity of dialysis material and personnel was transported into the damaged
area, the initiative to start support was taken without appropriate preparation
several days after the disaster, so that adequate help became possible only when
most renal victims had either died or recovered.
Together with the development of the concept of ‘renal disaster’ [1] grew
the awareness that if such support were needed in the future, a preconceived
plan would be necessary, with volunteer lists and stocks of material.
The Renal Disaster Relief Task Force
The International Society of Nephrology (ISN) installed the RDRTF in
1989, planning to organize rescue structures for three areas (Northern, Central
and South America, South-East Asia and Europe). For the time being, the
European branch is the most operational one. Lists of volunteers (physicians,
nurses, technicians) available to leave at short notice are registered in the head-
quarters of the RDRTF. Materials, such as drugs or dialysis devices, are avail-
able in the pharmacy of the hospital where the actions are coordinated
(University Hospital Ghent) or in the warehouses of Médecins Sans Frontières
(MSF – Doctors without Borders), the nongovernmental relief organization
under which the RDRTF operates. Further needs are supplemented by contacts
with pharmaceutical and dialysis companies.
Vanholder/Van Biesen/Lameire/Sever 328
A summary of the main activities of the RDRTF over the last few years is
given in table 1.
Organizational Aspects
Severity Assessment
One of the first tasks in renal rescue is to make a severity assessment so as
to know how many renal victims, especially those potentially in need of dialysis,
should be anticipated. The number of AKI patients in need of dialysis and their
proportion to the number of fatalities might be largely different from disaster to
disaster (table 2), and depends on the intensity of the earthquake, its location,
the time of occurrence, the quality of the buildings, the body constitution of the
potential victims, and the quality of rescue activities.
The general perception is that disasters in densely populated areas with
good rescue facilities, medium quality buildings (not good enough to avoid col-
lapse but solid enough to cause a lot of crushing after collapse), efficient pri-
mary care immediately after extrication, sufficient transport possibilities and
adequate hospital infrastructure in safe areas at a reasonable distance of the
affected zone result in a high prevalence of CS cases with kidney damage.
Fluid Administration
One of the mainstays of therapy and prevention of dialysis is a timely and
appropriate administration of fluids. In the Bingöl earthquake in Turkey 2003,
Table 1. Main interventions of the ISN/RDRTF, European branch
• Iran, March 1997: material support
• Moldova, March 1999: material support
• Macedonia, May 1999: evacuation of chronic patients
• Macedonia/Kosova, July 1999: material support
• Turkey, August 1999: major intervention
• Kosova, February 2000: educational support
• India, January 2001: scouting
• Turkey, May 2003: material support
• Algeria, May 2003: scouting
• Iran, December 2004: major intervention
• Luisiana, August 2005: advisory role
• Pakistan, October 2005: major intervention
• Poland, January 2006: advisory role
• Indonesia, May 2006: scouting
• Lebanon, July 2006: material support
ISN and RDRTF in Renal Disaster Victims 329
which was a disaster of limited extent, 12 victims with CS who ultimately
needed no dialysis treatment, received twice as much fluids (on average
50.6 vs. 24.0 l over the first 3 days following their extrication) as compared to 4
other victims needing dialysis [8]. This indicates that appropriate fluid adminis-
tration might obviate the need for dialysis. In table 3, we describe a schematic
approach to fluid administration as previously summarized [9].
Table 2. Ratio dialyzed/deaths (⫻1,000)
Location Country Year Ratio
Spitak Armenia 1988 9.0–15.4
Northern Iran Iran 1990 3.9
Kobe Japan 1995 24.6
Marmara Turkey 1999 28.1
Chi-Chi Taiwan 1999 13.3
Gujarat India 2001 1.7
Boumerdès Algeria 2003 6.6
Bam Iran 2003 3.7
Kashmir Pakistan 2005 2.4
Yogyakarta Indonesia 2006 ⬍0.1
Table 3. Fluid administration
Early fluid resuscitation (first 6 h, preferably starting before extrication)
While still under the rubble
1 l/h isotonic saline
After rescue
1 l/h half isotonic saline
50 mEq Na bicarbonate to be added to each second or third liter of half isotonic saline
5 g/h mannitol to be added if urinary flow ⬎20ml/h
After hospitalization
Under well-controlled conditions
Target an urinary flow ⬎300 ml/h
Due to muscular sequestration daily infusion may be needed to exceed urinary output by
ⱖ5 l/day during the first 3 days
Same combination of half isotonic saline complemented with Na bicarbonate and
mannitol in the same proportions as above
To be continued until myoglobinuria disappears
Central venous pressure measurements may offer additional guidance
In chaotic mass disasters (follow-up inadequate)
Restrict to 6 l/day, especially in the elderly and the presumed anuric
Vanholder/Van Biesen/Lameire/Sever 330
Hospital Infrastructure
Hospitals in the affected area are often damaged to an extent that treatment
of severely wounded victims becomes difficult. In the case of aftershocks, the
damage may be extended further, leading to the total collapse of hospital infra-
structure. It is therefore preferable to transport severe casualties with CS with a
short delay to a safer area which is not damaged. Transportation at a later stage
when patients have become dependent on ventilation or other intensive care
therapies may be fatal. Hence, early transport should be organized. Dialyzing
patients in local field hospitals near the damaged areas might be considered, but
because these can seldom be embedded into appropriate intensive care or other
hospital infrastructures, this option is less preferable.
Transport
As roads are often damaged, alternative means such as boats, air bridges or
helicopters should be taken into account [10]. Since hyperkalemia is a major
and potentially fatal problem, all severely affected victims with crush injury
should receive kayexalate salt before transportation.
Dialysis Treatment
Dialysis treatment may become necessary in an overwhelming number of
crush victims.
Intermittent Standard Hemodialysis
This option offers the possibility to treat several patients at the same posi-
tion, is relatively efficient in correcting the concentration of small water soluble
compounds such as potassium, and can be applied with minimal or no antico-
agulation. Due to heavy electrolyte disturbances, it may sometimes be neces-
sary to repeat hemodialysis several times per day.
Continuous Renal Replacement Therapy
Continuous renal replacement therapy (CRRT) offers the theoretical
advantage of maintaining better hemodynamic stability, although up to now no
controlled studies have shown superiority as to the clinical outcome [11, 12].
Only one patient can be treated per position and the need for continuous antico-
agulation may be a handicap in patients with bleeding or at risk of bleeding. The
strategy may be useful for areas where no traditional hemodialysis infrastructure
is available, but necessitates the availability of bulky amounts of substitution
fluid.
ISN and RDRTF in Renal Disaster Victims 331
Peritoneal Dialysis
Also peritoneal dialysis offers the theoretical advantage of maintaining
better hemodynamic stability than intermittent standard hemodialysis. Removal
of small molecules is less efficient than with intermittent standard hemodialy-
sis which may be a handicap in case of hyperkalemia. Peritoneal dialysis is
less appropriate in subjects with abdominal or thoracic surgery or trauma, and
in subjects with respiratory distress. Like CRRT, this strategy may be useful
for the rare areas where no traditional hemodialysis infrastructure is available
but similar to CRRT, it also necessitates the availability of bulky amounts of
fluid.
Conclusions
The experience with the RDRTF has shown that preplanned structures
for renal support in severe disaster conditions may be life saving for a sub-
stantial number of victims affected by the CS. This approach necessitates the
availability of volunteer lists, a well-defined action plan, the possibility to
deploy logistic support, and easy access to stocks of medication and dialysis
material.
References
1 Solez K, Bihari D, Collins AJ, et al: International dialysis aid in earthquakes and other disasters.
Kidney Int 1993;44:479–483.
2 Collins AJ: Kidney dialysis treatment for victims of the Armenian earthquake. N Engl J Med
1989;320:1291–1292.
3 Vanholder R, Sever MS, De Smet M, Erek E, Lameire N: Intervention of the Renal Disaster Relief
Task Force in the 1999 Marmara, Turkey earthquake. Kidney Int 2001;59:783–791.
4 Hatamizadeh P, Najafi I, Vanholder R, et al: Epidemiologic aspects of the Bam earthquake in Iran:
the nephrologic perspective. Am J Kidney Dis 2006;47:428–438.
5 Vanholder R, van der Tol A, De Smet M, et al: Earthquakes and crush syndrome casualties: lessons
learned from the Kashmir disaster. Kidney Int 2007;71:17–23.
6 Vanholder R, Sever MS, Erek E, Lameire N: Rhabdomyolysis. J Am Soc Nephrol 2000;11:1553–1561.
7 Thyssen EP, Hou SH, Alverdy JC, Spiegel DM: Temporary loss of limb function secondary to soft
tissue calcification in a patient with rhabdomyolysis-induced acute renal failure. Am J Kidney Dis
1990;16:491–494.
8 Gunal AI, Celiker H, Dogukan A, et al: Early and vigorous fluid resuscitation prevents acute renal
failure in the crush victims of catastrophic earthquakes. J Am Soc Nephrol 2004;15:1862–1867.
9 Sever MS, Vanholder R, Lameire N: Management of crush-related injuries after disasters. N Engl
J Med 2006;354:1052–1063.
10 Sever MS, Erek E, Vanholder R, et al: The Marmara earthquake: epidemiological analysis of the
victims with nephrological problems. Kidney Int 2001;60:1114–1123.
11 Tonelli M, Manns B, Feller-Kopman D: Acute renal failure in the intensive care unit: a systematic
review of the impact of dialytic modality on mortality and renal recovery. Am J Kidney Dis
2002;40:875–885.
Vanholder/Van Biesen/Lameire/Sever 332
12 Vinsonneau C, Camus C, Combes A, et al: Continuous venovenous haemodiafiltration versus
intermittent haemodialysis for acute renal failure in patients with multiple-organ dysfunction syn-
drome: a multicentre randomised trial. Lancet 2006;368:379–385.
R. Vanholder
Renal Section, 0K12, University Hospital
De Pintelaan, 185
BE–9000 Ghent (Belgium)
Tel. ⫹32 9240 4525, Fax ⫹32 9240 4599, E-Mail raymond.vanholder@ugent.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 333–339
Renal Replacement Therapy for the Patient
with Acute Traumatic Brain Injury and
Severe Acute Kidney Injury
Andrew Davenport
Centre for Nephrology, Division of Medicine, Department of Medicine,
Royal Free and University College Medical School, London, UK
Abstract
Fortunately with improvements in initial medical resuscitation, such as the avoidance
of nephrotoxins, the incidence of acute kidney injury requiring renal support in patients with
acute traumatic brain injury remains low. However the incidence of cerebral hemorrhage in
patients on chronic dialysis programs appears to be increasing. By carefully adapting renal
replacement to minimize cardiovascular instability and reduce the rate of change of serum
osmolality, patient survival in this group of critically ill patients is increasing and starting to
approach that of patients with traumatic brain injury without kidney injury.
Copyright © 2007 S. Karger AG, Basel
Patients with acute traumatic brain injury (TBI) may develop acute kidney
injury (AKI) requiring renal support [1], due to ischemic AKI following hypop-
erfusion as a complication of the original injury, or due to subsequent sepsis
and/or administration of nephrotoxins. Fortunately the number of such patients
requiring renal replacement therapy (RRT) each year is small, as they are typi-
cally young with previous good health and receive high quality resuscitation at
the accident scene. However, TBI can also occur in patients on dialysis for
established chronic kidney disease. Coumarins are prescribed for atrial fibrilla-
tion, severe cardiac failure, previous ischemic stroke, and also in dialysis units
to prevent clotting of dialysis catheters, and in patients with central venous
thrombosis. The risk of bleeding in hemodialysis (HD) patients prescribed
coumarins is greater than that of the general population [2], resulting in an
increased risk of subdural/intracranial hemorrhage.
Extracorporeal Treatment for Specific Indications
Davenport 334
Standard Medical Management of Patients with TBI
Over the last decade the optimal management of patients with TBI has
moved to a strategy designed to maintain cerebral perfusion pressure (CPP),
with controlled ventilation and limited use of hyperventilation to treat surges in
intracranial pressure (ICP) [3–5]. However, some centers continue to focus their
management on attempts to control cerebral volume [6]. The brain contains
more glial cells than neurons, and in particular the astrocytes are involved in
maintaining the integrity of the cerebral capillary endothelial blood brain barrier
(BBB) and homeostasis of the cerebral extracellular matrix (fig. 1). The brain
lies in a fixed volume cranium, so any volume expansion due to acute hemor-
rhage/hematoma will result in an increase in ICP unless the expansion can be
compensated by a reduction in cerebrospinal fluid or cerebral blood volume.
If cerebral autoregulation is preserved, sustaining the mean arterial blood
pressure will reduce the intracranial blood volume by autoregulatory cerebral
vasoconstriction and improve cerebral perfusion [3]. However if autoregulation
fails, higher pressure will lead to increased cerebral capillary hydrostatic
Fig. 1. Drawing of the intact blood brain barrier, which prevents free movement of pro-
teins and solutes into the brain when transcapillary hydrostatic pressure (Pc) and plasma
oncotic pressure (Pop) are balanced so that there is no net water. Crystalloid osmotic pressure
(Cop) is equal in all three compartments. Active solute transport by carrier transport systems
is vital for brain nutrition.
Astrocyte
PO
4
⫺
Na
⫹
K
⫹
Cl
⫺
Cop 5,500mmHg
Water only
Cerebral capillary
Flow
Cerebral interstitium
Cop 5,500mmHg
Cop 5,500mmHg
Pop 20–25mmHgPc 20–25mmHg
RRT for Patients with TBI and Severe AKI 335
pressure with increased transcapillary leak causing further cerebral volume
expansion and higher ICP. Thus control of cerebral volume, advocated by the
Lund group [6], by attempting to reduce cerebral capillary hydrostatic pressure,
whilst maintaining cerebral capillary plasma osmotic pressure, may improve
clinical outcome [7]. Although there may appear to be some marked differences
in the management strategy of the Brain Trauma Foundation compared to the
Lund group, the key similarity is to determine the optimum CPP and ICP for
any individual patient to maximize potential recovery [8].
Effect of Renal Replacement Therapies on ICP and CPP
In AKI, urea and other solutes increase and patients may develop meta-
bolic acidosis. The BBB breaks down in cases of TBI allowing solute passage
into the brain. This influx is initially compensated by astrocytes taking up addi-
tional ions and water, and generating idiogenic osmolarity.
The goals of any RRT are not only to clear urea and other solutes, but also
to correct metabolic acidosis and restore sodium and water balance. Urea and
water do not simply diffuse across the BBB, but pass through urea transporters
and aquaporin channels, respectively. As the relative speed of passage through
aquaporin channels is faster than that of urea transporters, this leads to the
development of an osmotic gradient. RRT can impact on the brain in patients
with TBI [9] by rapidly altering plasma solute concentrations. In addition, the
rapid infusion of bicarbonate leads to a rapid correction of plasma pH.
Bicarbonate cannot readily cross cell membranes, but reacts with hydrogen ions
to form H
2
CO
3
which can dissociate into water and carbon dioxide which can
readily cross cell membranes. Thus the fluxes of HCO
3
⫺
, H
2
CO
3
, and carbon
dioxide lead to changes in pH, according to the Henderson-Hasselbach equa-
tion, resulting in intracellular acidosis and the compensatory production of
intracellular osmolarity, with consequent water movement into the brain.
Hypotension due to excessive ultrafiltration may reduce cerebral perfusion and
CPP, with a consequent increase in ICP. Anticoagulation may be required for
some RRT modalities, and this might increase the risk of intracerebral hemor-
rhage, depending upon the nature of the TBI and the ICP monitor used [10].
Peritoneal Dialysis
Although the rate of change in serum urea and osmolality is slower during
peritoneal dialysis (PD) compared to intermittent HD, the dialysis disequilib-
rium syndrome has been reported during PD [11]. PD solutions are hypona-
tremic and this leads to a fall in serum sodium and water retention. This is
contrary to current neurosurgical intensive care practice designed to maintain
Davenport 336
plasma osmolality as advocated by the Lund hypothesis, and many centers now
infuse hypertonic saline to maintain a high or supraphysiological serum
sodium. In addition performing large volume exchanges of hypertonic glucose
can adversely impact cerebral perfusion [12] and CPP by reducing right atrial
filling [13]. Thus PD prescriptions should use the lowest glucose concentrations
possible and avoid major swings in intraperitoneal volumes.
PD may well provide adequate clearances in patients with TBI and AKI
alone, but may not be so effective in cases complicated by systemic sepsis.
Intraperitoneal infection remains the commonest complication of PD, although
in cases of TBI nosocomial infection, in particular ventilator-associated pneu-
monia, is a well-recognized complication and may be increased by PD.
Intermittent Hemodialysis/Hybrid Therapies
Brain edema occurs during routine outpatient thrice weekly HD. Thus
intermittent HD increases brain swelling in the patient with TBI [14]. ICP
increases not only due to changes in osmotic gradients, but also due to abrupt
falls in mean arterial pressure and thus cerebral perfusion and perfusion pres-
sure [15]. Intermittent RRT should be prescribed to maximize cardiovascular
stability using a high sodium dialysate concentration, cooled dialysate, and
minimizing changes in effective blood volume. As there are now HD machines
with blood volume monitoring and/or blood volume control, these should be
used preferentially. In addition the rate of change in serum osmolality should be
minimized by increasing the frequency of RRT to a daily schedule with an
extended session time, utilizing slower blood pump flows, small surface area
dialyzer membranes and even reducing dialysate flow. There are no randomized
prospective trials which have investigated the optimum predialysis urea to min-
imize changes in ICP during dialysis, however clinical practice suggests that a
predialysis urea of ⬍15 mmol/l (preferably ⬍12 mmol/l or BUN ⬍30 mg/dl)
reduces the risk of ICP increasing during treatment.
Continuous Dialysis/Hemofiltration Therapies
Continuous arteriovenous hemofiltration was shown to have a much lesser
effect on ICP than intermittent RRT [16]. However the introduction of pumped
venous RRT and, in particular, high volume exchanges in patients with critical
cerebral perfusion is associated with changes in ICP [17] (fig. 2). Thus when
performing continuous RRT in critically ill patients with AKI, circuit design is
important in that filtration is preferable to dialysis as this provides a slower rate
of change in serum urea and other small solutes. Sodium balance during
hemofiltration is positive as sodium sieving is ⬍1.0, but even so a replacement
fluid with a sodium concentration of ⬎140 mmol/l should be used initially. At
the start of treatment, small volume exchanges of 1.0 l/h should be used, and
RRT for Patients with TBI and Severe AKI 337
only when the patient has been shown to be stable should the exchange volume
be increased.
Anticoagulation Strategies
Patients with AKI are at a high risk of hemorrhage, due either to hemorrhage
associated with the initial TBI, or immediately post-neurosurgery to remove hem-
orrhage, or due to the presence of ICP-monitoring devices [10]. There is a greater
risk of hemorrhage associated with intraventricular compared to brain parenchy-
mal and subdural monitors/catheters. Thus ideally RRT should be anticoagulant
free or a regional anticoagulant [17] such as citrate, nafamostat, and vasodilatory
prostanoids used. In patients with compromised CPP, the vasodilatory effect of
the potent prostanoids, prostacyclin and epoprostenol, may reduce CPP and
thereby increase ICP [18]. Prior to using these agents, it is important to ensure
that hypovolemia is corrected and that CPP is adequate and, if necessary, pressor
support is temporarily increased.
Management of Sustained Surges in ICP during RRT
If sustained surges in ICP occur during RRT then standard medical man-
agement should be instituted, first checking that patients are adequately oxy-
genated with controlled hyperventilation (PaO
2
⬎11 kPa (82.5 mm Hg) with a
PaCO
2
of 4.5–5 kPa (49.5–55 mm Hg)) [4, 5]. Then if the CPP is high, slow
boluses of propofol and/or thiopentone could be administered, whereas if the
CPP is normal or low, then hypertonic saline and/or mannitol should be given.
Fig. 2. Change in intracerebral pressure in a patient with cerebral edema who was
treated by both continuous arteriovenous hemofiltration (CAVHF) and intermittent high vol-
ume veno-venous hemofiltration (4.0-liter exchange/h for 4 h; IVVHF). Mannitol boluses:
100 ml ⫻ 20% mannitol.
40
Mannitol
Before 1 2 3
Time (h)
4
35
30
Mean intracranial pressure (mmHg)
25
20
15
10
5
0
CAVHF
IVVHF
Davenport 338
Individual centers differ in concentrations of hypertonic saline used (ranging
from 3 to 10%), and these can be infused during RRT aiming for a serum
sodium of 145, up to a maximum of 155 mmol/l, depending upon the ICP
response. 100 ml of a 20% mannitol solution can also be given over 10–15 min,
although most units now favor hypertonic saline. There is debate as to whether
colloid solutions are deleterious because, when the BBB has broken down, col-
loids could enter the cerebral extracellular space and, due to their oncotic pres-
sure, attract water and increase cerebral edema. The Lund group advocate
albumin to maintain a normal serum albumin [6], whereas the SAFE study
reported an adverse effect of colloid resuscitation in neurotrauma [19].
In refractory cases, other medical strategies include a short period of
hyperventilation, deliberate hypothermia, CSF removal if a ventricular drain is
in place [20], and surgical craniotomy as a last resort.
Conclusion
In patients with TBI requiring RRT, the RRT should be tailored to
minimize abrupt changes in serum osmolality and to maintain cardiovascular
stability. Continuous therapies are therefore advantageous for short intermittent
RRT. Similarly anticoagulation free or regional anticoagulants are to be pre-
ferred to reduce the risk of further cerebral hemorrhage and bleeding around
ICP-measuring devices.
References
1 Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, Levin A: Acute Kidney
Injury Network (AKIN): report of an initiative to improve outcomes in acute kidney injury. Crit Care
2007;11:R31.
2 Lo DS, Rabbat CG, Clase CM: Thromboembolism and anticoagulant management in haemodial-
ysis patients: a practical guide to clinical management. Thromb Res 2006;118:385–395.
3 Rosner MJ, Rosner SD, Johnson AH: Cerebral perfusion pressure. Management protocol and clin-
ical results. J Neurosurg 1995;83:949–962.
4 Brain Trauma Foundation. American Association of Neurological Surgeons. Joint Section on
Neurotrauma and Critical Care. Methodology. J Neurotrauma 2000;17:561–562.
5 Brain Trauma Foundation. American Association of Neurological Surgeons. Joint Section on
Neurotrauma and Critical Care. Guidelines for the management of severe head injury. J Neurotrauma
1996;13:641–734.
6 Grande PO: The ‘Lund Concept’ for the treatment of severe head trauma – physiological principles
and clinical applications. Intensive Care Med 2006;32:1475–1484.
7 Howells T, Elf K, Jones PA, Ronne E, Piper I, Nilsson P, Andrews PJD, Enblad P: Pressure reac-
tivity as a guide in the treatment of cerebral perfusion pressure in patients with brain trauma.
J Neurosurg 2005;102:311–317.
8 Andrews PJD, Citerio G: Lund Therapy – pathophysiology based therapy or contrived over inter-
pretation of limited data? Intensive Care Med 2006;32:1461–1463.
RRT for Patients with TBI and Severe AKI 339
9 Gondo GK, Fujitsu T, Kuwabara Y, Mochimatsu Y, Ishiwata H, Oda N, Takagi T, Yamashita H,
Fujino I, Kim I, et al: Comparison of five modes of dialysis in neurosurgical patients with renal
failure (in Japanese). Neurol Med Chir (Tokyo) 1989;29:1125–1131.
10 Blei AT, Olafsson S, Webster S, Levy R: Complications of intracranial pressure monitoring in ful-
minant hepatic failure. Lancet 1993;341:157–158.
11 Pai MF, Hsu SP, Peng YS, Hung KY, Tsai TJ: Hemorrhagic stroke in chronic dialysis patients. Ren
Fail 2004;26:165–170.
12 Davenport A: Is there a role for continuous renal replacement therapies in patients with liver and
renal failure? Kidney Int Suppl 1999;72:S62–S66.
13 Selby NM, Fonseca S, Hulme L, Fluck RJ, Taal MW, McIntyre CW: Automated peritoneal dialysis
has significant effects on systemic hemodynamics. Perit Dial Int 2006;26:328–335.
14 Davenport A: Renal replacement therapy in the patient with acute brain injury. Am J Kidney Dis
2001;37:457–466.
15 Davenport A, Will EJ, Davison AM: Early changes in intracranial pressure during haemofiltration
treatment in patients with grade 4 hepatic encephalopathy and acute oliguric renal failure. Nephrol
Dial Transplant 1990;5:192–198.
16 Davenport A, Will EJ, Losowsky MS, Swindells S: Continuous arteriovenous haemofiltration in
patients with hepatic encephalopathy and renal failure. Br Med J (Clin Res Ed) 1987;295:1028.
17 Davenport A: Is there a role for continuous renal replacement therapies in patients with liver and
renal failure? Kidney Int Suppl 1999;72:S62–S66.
18 Davenport A, Will EJ, Davison AM: The effect of prostacyclin on intracranial pressure in patients
with acute hepatic and renal failure. Clin Nephrol 1991;35:151–157.
19 Finfer S, Bellomo R, Boyce N, French J, Myburgh J, Norton R; SAFE Study Investigators: A com-
parison of albumin and saline for fluid resuscitation in the intensive care unit. N Engl J Med
2004;350:2247–2256.
20 Morris KP, Forsyth KP, Parslow RC, Tasker RC, Hawley CA; UK Paediatric Traumatic Brain
Injury Study, and Paediatric Intensive Care Society Study Group: Intracranial pressure complicat-
ing severe traumatic brain injury in children: monitoring and management. Intensive Care Med
2006;32:1606–1612.
Andrew Davenport, MD
Centre for Nephrology, Division of Medicine, Department of Medicine
Royal Free and University College Medical School
Rowland Hill Street
London NW3 2PF (UK)
Tel. ⫹44 20 783 022 91, Fax ⫹44 20 783 021 25, E-Mail andrew.davenport@royalfree.nhs.uk
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 340–353
Cardiopulmonary Bypass-Associated
Acute Kidney Injury: A Pigment
Nephropathy?
Michael Haase
a,b
, Anja Haase-Fielitz
a,b
, Sean M. Bagshaw
a
,
Claudio Ronco
c
, Rinaldo Bellomo
a
a
Department of Intensive Care, Austin Health, Melbourne, Vic., Australia;
b
Department
of Nephrology, Charité University Medicine, Berlin, Germany;
c
Department of
Nephrology, St. Bortolo Hospital, Vicenza, Italy
Abstract
Acute kidney injury (AKI) is a common and serious postoperative complication fol-
lowing exposure to cardiopulmonary bypass (CPB). Several mechanisms have been pro-
posed by which the kidney can be damaged and interventional studies addressing known
targets of renal injury have been undertaken in an attempt to prevent or attenuate CPB-
associated AKI. However, no definitive strategy appears to protect a broad heterogeneous
population of cardiac surgery patients from CPB-associated AKI. Although the association
between hemoglobinuria and the development of AKI was recognized many years ago, this
idea has not been sufficiently acknowledged in past and current clinical research in the con-
text of cardiac surgery-related AKI. Hemoglobin-induced renal injury may be a major con-
tributor to CPB-associated AKI. Accordingly, we now describe in detail the mechanisms by
which hemoglobinuria may induce renal injury and raise the question as to whether CPB-
associated AKI may actually be, in a significant part, a form of pigment nephropathy where
hemoglobin is the pigment responsible for renal injury. If CPB-associated AKI is a pigment
nephropathy, alkalinization of urine with sodium bicarbonate might protect from: (1) tubular
cast formation from met-hemoglobin; (2) proximal tubular cell necrosis by reduced endocy-
totic hemoglobin uptake, and (3) free iron-mediated radical oxygen species production and
related injury. Sodium bicarbonate is safe, simple to administer and inexpensive. If part of
AKI after CPB is truly secondary to hemoglobin-induced pigment nephropathy, prophylactic
sodium bicarbonate infusion might help attenuate it. A trial of such treatment might be a rea-
sonable future investigation in higher risk patients receiving CPB.
Copyright © 2007 S. Karger AG, Basel
Dr. Haase holds a postdoctoral Feodor-Lynen Research Fellowship from the Alexander
von Humboldt-Foundation, Germany.
Renal Pigment Injury after Cardiac Surgery 341
Epidemiology and Risk Factors
With over one million operations a year worldwide, cardiac surgery is one
of the most common major surgical procedures [1]. Acute kidney injury (AKI)
is a common postoperative complication following exposure to cardiopul-
monary bypass (CPB) [2–5]. AKI requiring dialysis occurs in up to 5% of
patients undergoing elective cardiac surgery. An additional 8–15% of patients
have moderate AKI with an increase in serum creatinine level of 1.0 mg/dl
(88.4 mol/l). A lesser degree of AKI with a greater than 25% increase in
serum creatinine from baseline to a postoperative peak level may affect more
than 50% of patients [5, 6].
Some patients are at particular risk of developing CPB-related acute renal
failure (ARF) such as those with an increased duration of CPB, a preoperative
serum creatinine level of 1.2 mg/dl, insulin-dependent diabetes mellitus, age
70 years, reduced left ventricular function, valve surgery, preoperative atrial
fibrillation, and vascular disease [7]. Interestingly, there is evidence that a
longer duration of CPB is associated with an increased likelihood of more
severe AKI [4, 8, 9].
In these patients, reducing the incidence and severity of post-CPB AKI
might prevent expenditure and morbidity and improve other outcomes.
Costs and Outcomes
AKI carries a significant cost and is a serious postoperative complication
[10]. Also AKI leads to a significant increase in hospital expenditure especially
if complicated by the need for dialysis [11].
Adverse outcomes of AKI after cardiac surgery include prolonged inten-
sive care unit and hospital stay and discharge to extended-care facilities
[12, 13]. After adjustment for comorbidities and intraoperative variables, all
degrees of AKI are associated with increased mortality [2, 4, 5]. Even minimal
increments in serum creatinine are associated with an independent increase in
mortality [14, 15].
Pathophysiological Mechanisms and Prevention
Multiple causes of AKI following cardiac surgery have been proposed
including perioperative hemodynamic instability and impaired renal blood flow,
ischemia-reperfusion injury, and CPB-induced activation of inflammatory path-
ways and the generation of reactive oxygen species (ROS) [5, 16–19]. Other
Hasse/Haase-Fielitz/Bagshaw/Ronco/Bellomo 342
less common sources of renal injury include atheroembolism into the renal
arteries and exogenous nephrotoxins such as nephrotoxic antibiotics, nons-
teroidal anti-inflammatory drugs and anesthetics, all of which may contribute to
AKI in selected patients [5].
However, ischemia-reperfusion injury and the generation of oxido-
inflammatory stress represent two conventionally accepted major mechanisms
in the pathogenesis of CPB-related AKI. The evidence supporting such mecha-
nisms, however, is indirect and weak. In particular no randomized controlled
trials (RCTs) seeking to prophylactically affect such pathways has yet to be
shown effective.
Several RCTs have attempted to prevent or attenuate AKI, but most of these
interventions have been found to be ineffective [6, 20–29] or inconclusive, and/or
have only been studied in specific cardiac surgery subpopulations [30, 31].
Renal Hypoperfusion and Hypoxia/Ischemia-Reperfusion Injury
Although the kidneys receive more blood flow per gram of tissue than
other major organs, they are also the most susceptible to ischemic injury.
Metabolic demands from active tubular reabsorption and the oxygen diffusion
shunt characteristic of the renal circulation contribute to the vulnerable physiol-
ogy of renal perfusion including low medullary oxygen tension (10–20 mm Hg)
[32]. Therefore, maintenance of physiological blood flow and near normal
mean arterial pressure (MAP) before, during and after CPB is believed to be
important for the prevention of postoperative AKI [33, 34].
Perioperative hemodynamic instability, which exceeds the autoregulatory
reserve of the renal circulation may contribute to renal hypoperfusion and
hypoxia in the renal medulla. Renal hypoxia may be increased by anemia result-
ing from hemodilution during and after CPB. Any oxygen supply-demand
imbalances may play a crucial role in the development of AKI. However, at pre-
sent, it remains unclear whether hypoxia per se or rather re-oxygenation (possi-
bly through ROS) or [35] an impaired blood flow regulation (possibly through
nitric oxide) cause AKI [36].
Ischemia-reperfusion injury of the kidney frequently occurring during car-
diac surgery, for example due to cross clamping and reopening of the aorta, may
be another important factor contributing to postoperative AKI [37, 38]. Decreased
tissue oxygen tension promotes mitochondrial generation of ROS [39].
Unfortunately, there is a shortage of clinical trials randomizing patients
according to hemodynamic targets (e.g. high versus low systemic MAP/CI/
blood flow or renal blood flow) or high versus low hemoglobin levels to inves-
tigate renal outcomes. The only RCT available studied 21 cardiac surgery
patients and found no influence of intraoperative systemic MAP on postopera-
tive renal function [20]. In the absence of any RCT, a retrospective study of
Renal Pigment Injury after Cardiac Surgery 343
1,760 patients evaluating the effect of low on-pump hematocrit on postopera-
tive renal outcome found that CPB hemodilution to hematocrit 24% was
associated with an increased likelihood of renal injury and worse operative out-
comes [40].
Several RCTs have investigated pharmacological interventions, which
were believed to improve renal perfusion, to increase renal blood flow or to
decrease cortical oxygen consumption. None of these medications has been
consistently found to prevent or attenuate AKI including fenoldopam, dopamine,
clonidine, diltiazem, nesiritide [21–23, 31], pentoxifylline [24] and angiotensin-
converting enzyme inhibitors and diuretics [25–28]. Alternatively, these inter-
ventions have only been studied in specific cardiac surgery subpopulations
[30, 31].
Oxido-Inflammatory Stress
CPB has been shown to stimulate neutrophils and to induce the generation
of ROS and inflammatory mediators [41–44]. Increased levels of serum lipid
peroxidation products and an intra- and postoperatively decreased total serum
antioxidative capacity have also been found [16, 17]. There is evidence indicat-
ing that the generation of ROS may contribute to the initiation and maintenance
of acute tubular necrosis [45]. Oxidative stress is considered to be an important
cause of renal injury in patients exposed to CPB [16–19].
CPB is proinflammatory and activates components of the nonspecific
immune system. The inflammatory response to CPB generates cytokines (e.g.
TNF and IL-6), both systemically, locally and in the kidney, that have major
effects on the renal microcirculation and may lead to tubular injury [46–48].
Several RCTs aimed to achieve a reduction in inflammation and ROS, and
to prevent AKI following CPB. However, N-acetylcysteine, which directly scav-
enges ROS, regenerates the glutathione pool, and reduces oxidative stress dur-
ing CPB [49], did not provide renal protection [29, 50, 51]. Also,
dexamethasone and enoximone [52–54] failed to demonstrate renal protection
despite a reduction in the inflammatory response.
In summary, although diverse mechanisms exist by which the kidney can
be damaged during cardiac surgery, no key mechanism of CPB-associated AKI
and no corresponding preventive strategies have yet been identified.
A Pigment Nephropathy?
Pigment nephropathy is known to result from hemoglobinuria and myoglo-
binuria [55–58]. A wide range of causative factors are involved in the release of
free hemoglobin or free myoglobin into the serum including hemolysis from
Hasse/Haase-Fielitz/Bagshaw/Ronco/Bellomo 344
extracorporeal circulation (e.g. CPB), rhabdomyolysis, chemical agents and
also various venoms, malaria infection, mechanical destruction occasioned by
valvular prosthesis or a transfusion reaction, heat stroke, burns as well as some
genetic defects predisposing to reduced erythrocyte membrane stability [55].
Hemoglobin and myoglobin have a similar chemical core structure called
heme protein. At the center of the heme group is the iron metal ion (Fe
2
).
Hemoglobin consists of four protein chains and four heme groups whereas
myoglobin only consists of a single protein chain and one heme group. Given
the ability of both molecules to release free iron, which can act as nephrotoxin,
similar pathogenetic mechanisms in the development of hemoglobinuric and
myoglobinuric AKI can be assumed [55, 56]. However, due to the involvement
of CPB in the release of free hemoglobin and not myoglobin, in the following
we will focus on hemoglobinuria.
Although the association between hemoglobinuria and the development of
AKI was recognized many years ago, this idea has not been sufficiently
acknowledged in past and current clinical research focusing on cardiac surgery-
related AKI [56, 57, 59–64]. Hemoglobin-induced AKI may be a clinically rel-
evant cause for CPB-associated renal injury. Here we refer to hemoglobin-induced
AKI as pigment nephropathy with free serum hemoglobin as the toxic pigment
(fig. 1).
Hemolysis induced by mechanical destruction of erythrocytes through the
CPB circuit releases free hemoglobin into the plasma, where it combines with
haptoglobin to form a complex, which is carried to the liver, bypassing the kid-
ney, and metabolized [65]. Thus, under normal circumstances hemoglobin does
not exist in the serum in its free form. However, when the quantity of free serum
hemoglobin exceeds the binding capacity of haptoglobin, free serum hemoglo-
bin is able to scavenge endothelium-derived nitric oxide [36], but it will also
pass through the glomerulus, appear in urine, release iron which is involved in
the generation of ROS, and cause occlusion of renal tubules with hemoglobin
casts and necrosis of tubular cells [65, 66].
In summary, hemolysis by the extracorporeal circulation may be an impor-
tant contributor to AKI after cardiac surgery [67]. Accordingly, we now
describe the mechanisms of injury proposed for hemoglobinuria in detail and
we raise the question of whether CPB-associated AKI may eventually be, in
part, a form of pigment nephropathy [57, 61, 68].
Sources and Magnitude of Hemolysis during CPB
CPB is involved in causing hemolysis by mechanical destruction of the
erythrocytes thus generating free hemoglobin [59, 60, 67]. Many sources of
hemolysis contribute to increased plasma levels of free serum hemoglobin dur-
ing the use of CPB and in the early postoperative period [69].
Renal Pigment Injury after Cardiac Surgery 345
Shear stress on erythrocytes resulting from contact with foreign surfaces of
the bypass circuit (boundary layer of oxygenator, filters, tubing), cross-
sectional area, the number of circuit connectors, blood aspiration by cardiotomy
suction, and the roughness of surface of the pump, all of which are aggravated
by high flow and high pressure conditions, are important determinants of
hemolysis [70–72].
Free hemoglobin levels of 150 mg/dl, which is about 10-fold the upper
physiological range, have been observed during the use of CPB until several hours
Formation of tubular
Met-Hb casts
Conversion of Hb to
Met-Hb promoted under
aciduric conditions
Precipitation of
Met-Hb within the
distal tubular segments
Tubular obstruction
Filtration failure Endocytic uptake of
free Hb into PTC
PTC necrosis
Reperfusion after
myocardial ischemia
Generation
of ROS
Release of iron
from free Hb
Iron catalyses the
Haber-Weiss
reaction and
promotes
hydroxyl radical
formation and
lipid peroxidation
Formation of
tubular casts
Hb-ischemia
synergy
CPB-induced
hemolysis
Fig. 1. Possible mechanisms of injury in pigment nephropathy induced by cardiopul-
monary bypass (CPB). Hb Hemoglobin; PTC proximal tubular cell; ROS reactive
oxygen species.
Hasse/Haase-Fielitz/Bagshaw/Ronco/Bellomo 346
postoperatively [69] despite the short duration of CPB (85 min). However, the
detrimental effect of CPB on red cell destruction is accentuated by prolongation
of CPB time [73, 74]. Thus, the longer the duration of CPB the more hemolysis
occurs and the more free hemoglobin is generated. This may be of importance
to the current clinical situation where complex surgery of the aortic arch and
aortic valve is performed, and an increasing number of cardiac surgical centers
have implemented time-consuming arterial coronary revascularization aiming
to improve long-term results. Interestingly, there is evidence that a longer dura-
tion of CPB is associated with an increased likelihood of a more severe AKI [8,
9]. In addition, the use of CPB appears to have a close relation to hemolysis-
induced gallstone formation after open cardiac surgery, which can be prevented
by ursodeoxycholic acid [75, 76].
Relationship of Free Hemoglobin and CPB-Associated
Pigment Nephropathy
It is well recognized that free serum hemoglobin may be a major contribu-
tor to AKI [57, 60, 61]. Several mechanisms for the development of pigment
nephropathy have been suggested including the formation of tubular hemoglo-
bin cast, free iron-induced generation of ROS, and scavenging of endothelium-
derived nitric oxide [36, 56, 57, 63, 64].
In animal studies, infusion of free hemoglobin causes ARF [56–58]. The
conversion of hemoglobin to met-hemoglobin is thought to be an important
pathophysiological step in the induction of pigment nephropathy. An acid envi-
ronment typical of tubular urine facilitates this conversion. Met-hemoglobin
precipitates within the distal tubular segments forming casts, producing tubular
obstruction and hence, filtration failure [77]. Such heme pigment casts have
been described in virtually every study of pigment nephropathy [56–58, 77].
Also, under aciduric conditions, tubular obstruction may allow greater time for
endocytic uptake of free hemoglobin into proximal tubular cells, which is asso-
ciated with proximal tubular cell necrosis [57].
In addition, pigment nephropathy is drastically accentuated by additional
ischemia-reperfusion (renal artery occlusion) resulting in widespread met-
hemoglobin cast formation and proximal tubular cell necrosis [56, 57]. One
mechanism identified that may help to explain a degree of hemoglobin-
ischemia synergy is that ischemia-triggered cast formation enhances tubular
obstruction, which facilitates proximal tubular cell hemoglobin uptake [57].
Infusion of hemoglobin under alkalinuric conditions causes virtually no renal
injury and urine alkalinization attenuates renal failure in animal models [57, 64].
There is significantly reduced conversion of hemoglobin to met-hemoglobin, less
met-hemoglobin cast formation, hence, less tubular obstruction. Proximal tubular
cell necrosis appears to be extremely rare under these circumstances [57].
Renal Pigment Injury after Cardiac Surgery 347
Another cast formation-independent pathogenetic mechanism of pigment
nephropathy has been suggested: reperfusion following myocardial ischemia
generates oxygen-derived free radicals (e.g. in cardiac surgery patients), which
may release iron from free hemoglobin [78]. Iron promotes hydroxyl radical
formation and lipid peroxidation [56]. In fact, several recent studies have
demonstrated the potential for iron [78, 79] to catalyze the Haber-Weiss reac-
tion (fig. 2) whereby superoxide radical (O
2
) and hydrogenperoxide (H
2
O
2
)
yield hydroxyl radical (OH
•
) [80].
At neutral or alkaline pH, free ferric ions precipitate as insoluble ferric
hydroxide, reducing the production of injurious hydroxyl radicals [81].
It is also known that an acid environment typical of tubular urine enhances
the formation of reactive hydroxyl radicals as the Haber-Weiss reaction is
pH-dependent with a right shift when pH decreases. There is little argument
that hydroxyl radicals are injurious in a wide variety of settings [80].
Accordingly, the beneficial effect of higher proximal tubular pH by urinary
alkalinization, achieved for example with the use of sodium bicarbonate infu-
sion, was protective in a rat model of ARF [64].
Finally, free serum hemoglobin is able to scavenge endothelium-derived
nitric oxide 600-fold faster than erythrocytic hemoglobin [36]. This may lead to
vasoconstriction, decreased blood flow, platelet activation, increased endothelin-1
expression and AKI [36].
Potential Strategies for the Prevention of
Pigment Nephropathy following CPB
Several potential strategies for the prevention of pigment nephropathy have
been proposed [56, 57, 60, 64, 82–85]. Administration of haptoglobin has been
shown to have prophylactic and therapeutic effects on renal injury secondary to
hemolysis [60, 82–84].
Also, iron chelation with deferoxamine has been found to be protective
against pigment nephropathy in some animal models [56, 57, 85].
O
2
2Fe
3
2Fe
2
O
2
2O
2
2H
H
2
O
2
O
2
H
2
O
2
Fe
2
OH
•
OH
Fe
3
Fig. 2. The superoxide-driven Haber-Weiss describes one possible mechanism in the
generation of hydroxyl radicals that is catalyzed by free iron ions and most active at acid pH.
Hasse/Haase-Fielitz/Bagshaw/Ronco/Bellomo 348
Taking into account the pathogenetic mechanisms proposed above for the
notion of pigment nephropathy following exposure to CPB, there is sufficient
biological rationale to target the urine pH as a means of attenuating AKI after
CPB. An acid urine pH appears to be harmful in this setting by promoting
increased formation of distal met-hemoglobin casts and tubulo-toxic ROS
including hydroxyl radicals. An alkaline urine should be protective.
Sodium bicarbonate might potentially be able to target several injurious
mechanisms of pigment nephropathy following CPB and might therefore repre-
sent a promising approach (fig. 3).
Urinary alkalinization with sodium bicarbonate might protect from: (1) tubu-
lar cast formation by reduced conversion of hemoglobin to met-hemoglobin;
(2) proximal tubular cell necrosis by reduced endocytotic hemoglobin uptake;
(3) oxidant injury by slowing pH-dependent Haber-Weiss radical production;
(4) oxidant injury by direct scavenging of peroxynitrite and other reactive
species generated from nitric oxide [86], and (5) free iron-mediated radical
injury by precipitation of free ferric ions as insoluble ferric hydroxide at neutral
or alkaline pH [81].
In addition, there is already evidence from a double-blind RCT that bicar-
bonate might attenuate AKI in patients undergoing infusion of contrast media
[87]. The reduction of contrast-induced AKI by bicarbonate infusion would also
Urinary alkalinization with
Sodium bicarbonate
may protect from
Tubular cast
formation
by
reduced
conversion of
Hb to Met-Hb
Proximal tubular
cell necrosis
by
reduced
endocytotic
Hb uptake
Oxidant injury
by
slowing pH-dependent
Haber-Weiss radical
production
AND
scavenging peroxynitrite
and other reactive species
generated from NO
Iron-mediated
radical injury
by
precipitation of
free ferric ions
as insoluble
ferric hydroxide
at neutral or
alkaline pH
Fig. 3. Potential targets for sodium bicarbonate to protect the kidney from pigment-
induced injury. Hb Hemoglobin; NO nitric oxide.
Renal Pigment Injury after Cardiac Surgery 349
be consistent with the hypothesis that, in part, AKI following exposure to con-
trast media is from ROS generated within the acid environment of the renal
medulla [88, 89].
Conclusion
It is estimated that up to 50% of patients develop AKI following cardiac
surgery. CPB-associated AKI may, at least in part, be a pigment nephropathy.
Urine alkalinization might protect the kidney from such pigment nephropathy.
Sodium bicarbonate is known to be a safe, simple to administer, and inexpen-
sive intervention, which can alkalinize urine. It might, therefore, effectively
protect the kidney from injury in this setting. Given these theoretical and exper-
imental premises, it would seem reasonable to consider conducting a pilot RCT
to test whether prophylactic sodium bicarbonate infusion is indeed able to pre-
vent or attenuate AKI in patients receiving CPB.
References
1 Albert MA, Antman EM: Preoperative evaluation for cardiac surgery; in Cohn LH, Edmunds LH Jr
(eds): Cardiac Surgery in the Adult. New York, McGraw-Hill, 2003, pp 235–248.
2 Chertow GM, Levy EM, Hammermeister KE, et al: Independent association between acute renal
failure and mortality following cardiac surgery. Am J Med 1998;104:343–348.
3 Loef BG, Epema AH, Smilde TD, et al: Immediate postoperative renal function deterioration in
cardiac surgical patients predicts in-hospital mortality and long-term survival. J Am Soc Nephrol
2005;16:195–200.
4 Conlon PJ, Stafford-Smith M, White WD, et al: Acute renal failure following cardiac surgery.
Nephrol Dial Transplant 1999;14:1158–1162.
5 Stafford-Smith M, Podgoreanu M, Swaminathan M, et al: Association of genetic polymorphisms
with risk of renal injury after coronary bypass graft surgery. Am J Kidney Dis 2005;45:519–530.
6 Bove T, Landoni G, Calabro MG, et al: Renoprotective action of fenoldopam in high-risk patients
undergoing cardiac surgery: a prospective, double-blind, randomized clinical trial. Circulation
2005;111:3230–3235.
7 Thakar CV, Arrigain S, Worley S, et al: A clinical score to predict acute renal failure after cardiac
surgery. J Am Soc Nephrol 2005;16:162–168.
8 Boldt J, Brenner T, Lehmann A, et al: Is kidney function altered by the duration of cardiopul-
monary bypass? Ann Thorac Surg 2003;75:906–912.
9 Fischer UM, Weissenberger WK, Warters RD, et al: Impact of cardiopulmonary bypass manage-
ment on postcardiac surgery renal function. Perfusion 2002;17:401–406.
10 Landoni G, Zangrillo A, Franco A, et al: Long-term outcome of patients who require renal
replacement therapy after cardiac surgery. Eur J Anaesthesiol 2006;23:17–22.
11 Fischer MJ, Brimhall BB, Lezotte DC, et al: Uncomplicated acute renal failure and hospi-
tal resource utilization: a retrospective multicenter analysis. Am J Kidney Dis 2005;46:
1049–1057.
12 Mangano CM, Diamondstone LS, Ramsay JG, et al: Renal dysfunction after myocardial revascu-
larization: risk factors, adverse outcomes, and hospital resource utilization. The Multicenter Study
of Perioperative Ischemia Research Group. Ann Intern Med 1998;128:194–203.
Hasse/Haase-Fielitz/Bagshaw/Ronco/Bellomo 350
13 Antunes PE, Prieto D, Ferrao de Oliveira J, Antunes MJ: Renal dysfunction after myocardial
revascularization. Eur J Cardiothorac Surg 2004;25:597–604.
14 Lassnigg A, Schmidlin D, Mouhieddine M, et al: Minimal changes of serum creatinine predict
prognosis in patients after cardiothoracic surgery: a prospective cohort study. J Am Soc Nephrol
2004;15:1597–1605.
15 Zanardo G, Michielon P, Paccagnella A, et al: Acute renal failure in the patient undergoing cardiac
operation. Prevalence, mortality rate, and main risk factors. J Thorac Cardiovasc Surg 1994;107:
1489–1495.
16 Starkopf J, Zilmer K, Vihalemm T, et al: Time course of oxidative stress during open-heart surgery.
Scand J Thorac Cardiovasc Surg 1995;29:181–186.
17 McColl AJ, Keeble T, Hadjinikolaou L, et al: Plasma antioxidants: evidence for a protective role
against reactive oxygen species following cardiac surgery. Ann Clin Biochem 1998;35:616–623.
18 Doi K, Suzuki Y, Nakao A, et al: Radical scavenger edaravone developed for clinical use amelio-
rates ischemia/reperfusion injury in rat kidney. Kidney Int 2004;65:1714–1723.
19 McCord JM: Oxygen-derived free radicals in postischemic tissue injury. N Engl J Med 1985;312:
159–163.
20 Urzua J, Troncoso S, Bugedo G, et al: Renal function and cardiopulmonary bypass: effect of per-
fusion pressure. J Cardiothorac Vasc Anesth 1992;6:299–303.
21 Butler J, Emerman C, Peacock WF, et al: The efficacy and safety of B-type natriuretic peptide
(nesiritide) in patients with renal insufficiency and acutely decompensated congestive heart fail-
ure. Nephrol Dial Transplant 2004;19:391–399.
22 Wang DJ, Dowling TC, Meadows D, et al: Nesiritide does not improve renal function in patients
with chronic heart failure and worsening serum creatinine. Circulation 2004;110:1620–1625.
23 Sackner-Bernstein JD, Skopicki HA, Aaronson KD: Risk of worsening renal function with nesiri-
tide in patients with acutely decompensated heart failure. Circulation 2005;111:1487–1491.
24 Boldt J, Brosch C, Piper SN, et al: Influence of prophylactic use of pentoxifylline on postoperative
organ function in elderly cardiac surgery patients. Crit Care Med 2001;29:952–958.
25 Colson P, Ribstein J, Mimran A, et al: Effect of angiotensin converting enzyme inhibition on blood
pressure and renal function during open heart surgery. Anesthesiology 1990;72:23–27.
26 Ryckwaert F, Colson P, Ribstein J, et al: Haemodynamic and renal effects of intravenous enalapri-
lat during coronary artery bypass graft surgery in patients with ischaemic heart dysfunction. Br J
Anaesth 2001;86:169–175.
27 Lassnigg A, Donner E, Grubhofer G, et al: Lack of renoprotective effects of dopamine and
furosemide during cardiac surgery. J Am Soc Nephrol 2000;11:97–104.
28 Sirivella S, Gielchinsky I, Parsonnet V: Mannitol, furosemide, and dopamine infusion in postoper-
ative renal failure complicating cardiac surgery. Ann Thorac Surg 2000;69:501–506.
29 Burns KE, Chu MW, Novick RJ, et al: Perioperative N-acetylcysteine to prevent renal dysfunction
in high-risk patients undergoing cabg surgery: a randomized controlled trial. JAMA 2005;294:
342–350.
30 Sward K, Valsson F, Odencrants P, et al: Recombinant human atrial natriuretic peptide in ischemic
acute renal failure: a randomized placebo-controlled trial. Crit Care Med 2004;32:1310–1315.
31 Mentzer RM Jr, Oz MC, Sladen RN, et al: Effects of perioperative nesiritide in patients with left
ventricular dysfunction undergoing cardiac surgery. J Am Coll Cardiol 2007;49:716–726.
32 Brezis M, Rosen S: Hypoxia of the renal medulla – its implications for disease. N Engl J Med
1995;332:647–655.
33 Bhat JG, Gluck MC, Lowenstein J, Baldwin DS: Renal failure after open heart surgery. Ann Intern
Med 1976;84:677–682.
34 Abel RM, Buckley MJ, Austen WG, et al: Acute postoperative renal failure in cardiac surgical
patients. J Surg Res 1976;20:341–348.
35 Rosenberger C, Rosen S, Heyman SN: Renal parenchymal oxygenation and hypoxia adaptation in
acute kidney injury. Clin Exp Pharmacol Physiol 2006;33:980–988.
36 Gladwin MT, Crawford JH, Patel RP: The biochemistry of nitric oxide, nitrite, and hemoglobin:
role in blood flow regulation. Free Radic Biol Med 2004;36:707–717.
37 Andersson LG, Bratteby LE, Ekroth R, et al: Renal function during cardiopulmonary bypass:
influence of pump flow and systemic blood pressure. Eur J Cardiothorac Surg 1994;8:597–602.
Renal Pigment Injury after Cardiac Surgery 351
38 Abbott WM, Austen WG: The reversal of renal cortical ischemia during aortic occlusion by man-
nitol. J Surg Res 1974;16:482–489.
39 Dada LA, Chandel NS, Ridge KM, et al: Hypoxia-induced endocytosis of Na,K-ATPase in alveo-
lar epithelial cells is mediated by mitochondrial reactive oxygen species and PKC-zeta. J Clin
Invest 2003;111:1057–1064.
40 Habib RH, Zacharias A, Schwann TA, et al: Role of hemodilutional anemia and transfusion during
cardiopulmonary bypass in renal injury after coronary revascularization: implications on operative
outcome. Crit Care Med 2005;33:1749–1756.
41 Chello M, Mastroroberto P, Patti G, et al: Simvastatin attenuates leucocyte-endothelial interac-
tions after coronary revascularisation with cardiopulmonary bypass. Heart 2003;89:538–543.
42 Boyle EM Jr, Lille ST, Allaire E, et al: Endothelial cell injury in cardiovascular surgery: athero-
sclerosis. Ann Thorac Surg 1997;63:885–894.
43 Paparella D, Yau TM, Young E: Cardiopulmonary bypass induced inflammation: pathophysiology
and treatment. An update. Eur J Cardiothorac Surg 2002;21:232–244.
44 Partrick DA, Moore EE, Fullerton DA, et al: Cardiopulmonary bypass renders patients at risk for
multiple organ failure via early neutrophil priming and late neutrophil disability. J Surg Res 1999;86:
42–49.
45 Nath KA, Norby SM: Reactive oxygen species and acute renal failure. Am J Med 2000;109:
665–678.
46 Cunningham PN, Dyanov HM, Park P, et al: Acute renal failure in endotoxemia is caused by TNF
acting directly on TNF receptor-1 in kidney. J Immunol 2002;168:5817–5823.
47 Segerer S, Nelson PJ, Schlondorff D: Chemokines, chemokine receptors, and renal disease:
from basic science to pathophysiologic and therapeutic studies. J Am Soc Nephrol 2000;11:
152–176.
48 Heyman SN, Rosen S, Darmon D, et al: Endotoxin-induced renal failure. II. A role for tubular
hypoxic damage. Exp Nephrol 2000;8:275–282.
49 Sucu N, Cinel I, Unlu A, et al: N-Acetylcysteine for preventing pump-induced oxidoinflammatory
response during cardiopulmonary bypass. Surg Today 2004;34:237–242.
50 Ristikankare A, Kuitunen T, Kuitunen A, et al: Lack of renoprotective effect of i.v. N-acetylcysteine
in patients with chronic renal failure undergoing cardiac surgery. Br J Anaesth 2006;97:611–616.
51 Hynninen MS, Niemi TT, Poyhia R, et al: N-Acetylcysteine for the prevention of kidney injury in
abdominal aortic surgery: a randomized, double-blind, placebo-controlled trial. Anesth Analg
2006;102:1638–1645.
52 Morariu AM, Loef BG, Aarts LP, et al: Dexamethasone: benefit and prejudice for patients under-
going on-pump coronary artery bypass grafting: a study on myocardial, pulmonary, renal, intesti-
nal, and hepatic injury. Chest 2005;128:2677–2687.
53 Boldt J, Brosch C, Lehmann A, et al: The prophylactic use of the beta-blocker esmolol in combi-
nation with phosphodiesterase III inhibitor enoximone in elderly cardiac surgery patients. Anesth
Analg 2004;99:1009–1017, table of contents.
54 Boldt J, Brosch C, Suttner S, et al: Prophylactic use of the phosphodiesterase III inhibitor enoxi-
mone in elderly cardiac surgery patients: effect on hemodynamics, inflammation, and markers of
organ function. Intensive Care Med 2002;28:1462–1469.
55 Evenepoel P: Acute toxic renal failure. Best Pract Res Clin Anaesthesiol 2004;18:37–52.
56 Paller MS: Hemoglobin- and myoglobin-induced acute renal failure in rats: role of iron in nephro-
toxicity. Am J Physiol 1988;255:F539–F544.
57 Zager RA, Gamelin LM: Pathogenetic mechanisms in experimental hemoglobinuric acute renal
failure. Am J Physiol 1989;256:F446–F455.
58 Tam SC, Wong JT: Impairment of renal function by stroma-free hemoglobin in rats. J Lab Clin Med
1988;111:189–193.
59 Takami Y, Makinouchi K, Nakazawa T, et al: Effect of surface roughness on hemolysis in a pivot
bearing supported Gyro centrifugal pump (C1E3). Artif Organs 1996;20:1155–1161.
60 Tanaka K, Kanamori Y, Sato T, et al: Administration of haptoglobin during cardiopulmonary
bypass surgery. ASAIO Trans 1991;37:M482–M483.
61 Feola M, Simoni J, Tran R, Canizaro PC: Nephrotoxicity of hemoglobin solutions. Biomater Artif
Cells Artif Organs 1990;18:233–249.
Hasse/Haase-Fielitz/Bagshaw/Ronco/Bellomo 352
62 Lehotsky J, Kaplan P, Matejovicova M, et al: Ion transport systems as targets of free radicals dur-
ing ischemia reperfusion injury. Gen Physiol Biophys 2002;21:31–37.
63 Halliwell B, Gutteridge JM: Role of free radicals and catalytic metal ions in human disease: an
overview. Methods Enzymol 1990;186:1–85.
64 Atkins JL: Effect of sodium bicarbonate preloading on ischemic renal failure. Nephron 1986;44:
70–74.
65 Keene WR, Jandl JH: The sites of hemoglobin catabolism. Blood 1965;26:705–719.
66 Loebl EC, Baxter CR, Curreri PW: The mechanism of erythrocyte destruction in the early post-
burn period. Ann Surg 1973;178:681–686.
67 Kanamori Y, Tanabe H, Shimono T, et al: The effects of administration of haptoglobin for hemoly-
sis by extracorporeal circulation (in Japanese). Rinsho Kyobu Geka 1989;9:463–467.
68 Clyne DH, Kant KS, Pesce AJ, Pollak VE: Nephrotoxicity of low molecular weight serum pro-
teins: physicochemical interactions between myoglobin, hemoglobin, bence-jones proteins and
tamm-horsfall mucoprotein. Curr Probl Clin Biochem 1979;9:299–308.
69 Loef BG, Epema AH, Navis G, et al: Off-pump coronary revascularization attenuates transient
renal damage compared with on-pump coronary revascularization. Chest 2002;121:1190–1194.
70 Tamari Y, Lee-Sensiba K, Leonard EF, et al: The effects of pressure and flow on hemolysis caused
by Bio-Medicus centrifugal pumps and roller pumps. Guidelines for choosing a blood pump.
J Thorac Cardiovasc Surg 1993;106:997–1007.
71 Crane KA, Brown D, Anderson R, Begelman KM: Further decrease in subclinical hemolysis uti-
lizing 12.7 mm tubing in the arterial roller head. Ann Thorac Surg 1983;35:463–465.
72 Skrabal CA, Khosravi A, Westphal B, et al: Effects of poly-2-methoxyethylacrylate (PMEA)-
coating on CPB circuits. Scand Cardiovasc J 2006;40:224–229.
73 Yamaguchi H, Shimizu T, Akutsu H, et al: Hemolysis and red cell deformability during cardiopul-
monary bypass – the effect of prostaglandin E1 for prevention of hemolysis (in Japanese). Nippon
Kyobu Geka Gakkai Zasshi 1990;38:625–629.
74 Hirayama T, Herlitz H, Jonsson O, Roberts D: Deformability and electrolyte changes of erythro-
cytes in connection with open heart surgery. Scand J Thorac Cardiovasc Surg 1986;20:253–259.
75 Azemoto R, Tsuchiya Y, Ai T, et al: Does gallstone formation after open cardiac surgery result
only from latent hemolysis by replaced valves? Am J Gastroenterol 1996;91:2185–2189.
76 Ai T, Azemoto R, Saisho H: Prevention of gallstones by ursodeoxycholic acid after cardiac
surgery. J Gastroenterol 2003;38:1071–1076.
77 Jaenike JR: The renal lesion associated with hemoglobinemia: a study of the pathogenesis of the
excretory defect in the rat. J Clin Invest 1967;46:378–387.
78 Das DK, Engelman RM, Liu X, et al: Oxygen-derived free radicals and hemolysis during open
heart surgery. Mol Cell Biochem 1992;111:77–86.
79 Gutteridge JM: Iron promoters of the Fenton reaction and lipid peroxidation can be released from
haemoglobin by peroxides. FEBS Lett 1986;201:291–295.
80 Halliwell B, Gutteridge JM: Oxygen toxicity, oxygen radicals, transition metals and disease.
Biochem J 1984;219:1–14.
81 Cohen G: The Fenton reaction; in Greenwald RA (ed): CRC Handbook of Methods for Oxygen
Radical Research. Boca Raton, CRC Press, 1985, pp 55–64.
82 Ohshiro T, Kosaki G, Funakoshi S: Haptoglobin therapy: effect on prevention and treatment of
hemoglobinuria. Med J Osaka Univ 1978;29:269–279.
83 Ohshiro TU, Mukai K, Kosaki G: Prevention of hemoglobinuria by administration of haptoglobin.
Res Exp Med (Berl) 1980;177:1–12.
84 Yoshioka T, Sugimoto T, Ukai T, Oshiro T: Haptoglobin therapy for possible prevention of renal
failure following thermal injury: a clinical study. J Trauma 1985;25:281–287.
85 Shah SV, Walker PD: Evidence suggesting a role for hydroxyl radical in glycerol-induced acute
renal failure. Am J Physiol 1988;255:F438–F443.
86 Caulfield JL, Singh SP, Wishnok JS, et al: Bicarbonate inhibits N-nitrosation in oxygenated nitric
oxide solutions. J Biol Chem 1996;271:25859–25863.
87 Merten GJ, Burgess WP, Gray LV, et al: Prevention of contrast-induced nephropathy with sodium
bicarbonate: a randomized controlled trial. JAMA 2004;291:2328–2334.
Renal Pigment Injury after Cardiac Surgery 353
88 Bakris GL, Lass N, Gaber AO, et al: Radiocontrast medium-induced declines in renal function: a
role for oxygen free radicals. Am J Physiol 1990;258:F115–F120.
89 Katholi RE, Woods WT Jr, Taylor GJ: Oxygen free radicals and contrast nephropathy. Am J
Kidney Dis 1998;32:64–71.
Prof. Rinaldo Bellomo
Department of Intensive Care, Austin Hospital
Melbourne, Vic. 3084 (Australia)
Tel. 61 3 9496 5992, Fax 61 3 9496 3932, E-Mail rinaldo.bellomo@austin.org.au
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 354–364
CRRT Technology and Logistics:
Is There a Role for a Medical
Emergency Team in CRRT?
Patrick M. Honoré
a
, Olivier Joannes-Boyau
b
, Benjamin Gressens
a
a
St-Pierre Para-University Hospital, Ottignies-Louvain-la-Neuve, Belgium, and
b
Haut Leveque University Hospital, University of Bordeaux, Pessac, France
Abstract
Implementing continuous renal replacement therapy (CRRT) in a intensive care unit
(ICU) is a somewhat difficult issue and quiet different from starting a new ventilation mode
or a new hemodynamic device. It may indeed require an on-call medical emergency CRRT
team as expertise in this field is really a key issue to success. Education for the nursing team
is another key point, especially as ongoing or continuous education is changing very quickly.
Uniformity of the type of device used is another crucial part in the organization process with
regard to CRRT implementation in the ICU. Involvement of both the ICU and nephrology
teams is another key to success especially when different modes and higher exchange rates
are used. Also, a nursing group devoted to the ongoing implementation and education of the
ICU team is very useful in order to attain the goals that have been set. Already in 1984 acute
renal failure was described as one of the remaining and challenging problems in the ICU.
Hemodialysis was not always feasible then because of the hemodynamic instability of criti-
cally ill patients. Under those circumstances continuous arteriovenous hemofiltration
(CAVH) was advocated as an efficient alternative method with less detrimental hemody-
namic effects. At the time it was thought that CAVH would be found to be an effective ‘arti-
ficial kidney’ (control of body fluid, electrolyte and acid-base homeostasis and uremia) and
this without serious side effects. But already nearly 25 years ago, it was found that continu-
ous anticoagulation was a major problem that could cause life-threatening complications in
posttraumatic and surgical patients. At the time, it was thought that running a protamine infu-
sion on the venous line would help to diminish these complications. CRRT has been carried
out in our ICU since 1985, first with CAVH and since 1989 with some early forms of con-
tinuous veno-venous hemofiltration (CVVH). The unit has used BSM 22, BM 25 and Prisma
for nearly 10 years, and Aquarius since the end of 2001. The educational process started at
the beginning of 1990 with the implementation of CVVH using BSM 22 and BM 25. Very
soon it was realized that a new strategy implementing pulse high-volume hemofiltration
(pulse-HVHF) was really needed. Therefore, a nursing group composed of 5–8 nurses who
would be taught beforehand was started, and this dedicated group would then teach the rest
CRRT Technology and Logistics 355
of the staff nurses. This group exists today and has at least 6–8 meetings/year in which all
the problems that must be faced in the implementation of CRRT are dealt with. Here all the
steps made by our and other units in this field will be discussed, including an overview of the
various protocols implemented and a description of our dedicated nursing group with regard
to CRRT.
Copyright © 2007 S. Karger AG, Basel
Global Approach to the Problem in the ICU
Implementing continuous renal replacement therapy (CRRT) in an inten-
sive care unit (ICU) is a somewhat difficult issue and quiet different from start-
ing a new ventilation mode or a new hemodynamic device. It may indeed require
an on-call medical emergency CRRT team (MECT) as expertise in this field is a
key issue for success. Education for the nursing team is another key point, espe-
cially as is it rapidly changing. Uniformity of the type of device used is another
crucial part in the organization process with regard to implementing CRRT in
the ICU. The involvement of both the ICU and nephrology teams is another key
to success especially when different modes and higher exchange rates are being
used. Also a nursing group devoted to the ongoing implementation and educa-
tion of the ICU team is very useful in order to achieve the goals set as we have
done at Saint-Pierre Hospital [1].
Already in 1984 (nearly a quarter of a century ago), Schetz et al. [2] des-
cribed acute renal failure as one of the remaining and challenging problems in
the ICU because hemodialysis was not always feasible due to the hemodynamic
instability of critically ill patients. Under these circumstances continuous arteriove-
nous hemofiltration (CAVH) was advocated as an efficient alternative method
with less detrimental hemodynamic effects. At the time Schetz et al. [2] were
also convinced that CAVH would be found to be an effective ‘artificial kidney’
(control of body fluid, electrolyte and acid-base homeostasis and uremia) and
this without serious side effects. But already nearly 25 years ago, they empha-
sized that a major problem was continuous anticoagulation, which could cause
life-threatening complications in posttraumatic and surgical patients. At the
time, they were in favor of running a protamine infusion on the venous line in
order to diminish these complications [2].
As many experts have explained, the key to developing a successful CRRT
course is the involvement of clinical experts – the staff nurses. It is an ongoing
process that requires continual improvement. A devoted course on CRRT can
greatly help to assure the competence of the ICU nurse in caring for the criti-
cally ill patient while on CRRT as Clevenger [3] has elegantly shown by his
own experience.
Honoré/Joannes-Boyau/Gressens 356
As Giuliano and Pysznik [4] have highlighted, implementing a program
as complex as continuous veno-venous hemofiltration (CVVH) without the
involvement of nephrology nurses is a real challenge. However, with proper
planning, appropriate staff support, and the ability to make changes as imple-
mentation proceeds, a successful program can be developed. Indeed, our reward
will be that we will be able to offer therapy that is important and potentially life-
saving to those critically ill patients with renal failure who are unable to tolerate
intermittent hemodialysis [4].
More and more units are becoming convinced that a clinical educator is the
right answer to this problem, and we tend to agree. Harvey et al. [5] described
the new role of a renal critical care educator based in a regional pediatric renal
unit with an ‘outreach’ to three pediatric ICUs (PICUs) within the region.
Harvey et al. [5] demonstrated that after 18 months the training objectives for
the PICU staff progressed in all aspects of hemofiltration and all units used
flexible training programs which, at the same time, were under constant evalua-
tion using questionnaires. This type of renal critical care educator also worked
alongside staff whilst CVVH was being performed to further instill confidence.
During the same period of time, equipment and protocols have largely been
standardized throughout the region, and an ongoing survey of CVVH use was
initiated which could help to inform audit standards such as complication rates.
Harvey et al. [5] concluded that the renal critical care educator was the catalyst
for the formation of a regional hemofiltration group which shared in the devel-
opment of guidelines and protocols and discussion of clinical data. Since CRRT
is used infrequently in many PICUs the development of a renal critical care
educator could serve as a model for the development and maintenance of skills
in other regions [5]. This latest part could be seen in some ways as a nurse
emergency team for CRRT in ICU.
Policies concerning the choice of vascular access are also a crucial. As out-
lined by East and Jacoby [6] complications related to central venous line use are
known to increase patient morbidity and mortality and increase costs and length
of hospital stay. Education programs to promote best central line practice have
been shown to reduce central line complications. This was especially demon-
strated with regard to central line care policy in the pediatric cardiovascular
ICU [6].
In our unit, we were able to demonstrate that, when reaching 35 ml/kg/h of
exchange, a blood flow of 300 ml/min was required and this could not be sus-
tained for more than 24 h other than using the right internal jugular approach
[7] as the femoral access is not reliable in more than 50% of the patients due to
the very high incidence of abdominal compartment syndrome in the ICU [8].
Guidelines and written instructions as protocols are extremely important in
this setting. This has been very well illustrated by the work of Kingston et al. [9]
CRRT Technology and Logistics 357
who introduced the concept of patient group direction (PGD) which is a specif-
ically written instruction for supplying or administrating named medicines in an
identified clinical situation. The introduction of a PGD must demonstrate a ben-
efit for patients. They were able to easily show that hemofiltration was a very
good candidate for this type of specifically written instruction. Using a hemofil-
tration PGD, patient care was improved by providing standardization in the
administration of fluids and electrolytes and as well enabling nurses to respond
rapidly to changes in biochemistry during the procedure [10]. In our unit we use
similar types of written instructions [11], and in other units similar guidelines
also exist, as shown in the literature [12].
The Saint-Pierre Way
CRRT has been performed in our ICU since 1985, first with CAVH and
after 1989 with some early forms of CVVH. The unit has used BSM 22, BM 25
and Prisma for nearly 10 years and Aquarius since the end of 2001. The educa-
tion process really started from the beginning of 1990 with the implementation
of CVVH using BSM 22 and BM 25. At that time, a medical expert was teach-
ing the nursing staff directly. Very soon after, it was found that with the imple-
mentation of pulse high-volume hemofiltration (pulse-HVHF) [13, 14], a new
strategy was needed. Therefore, a nursing group, composed of 5–8 nurses, was
formed in order that they be taught beforehand, and this dedicated group would
then teach the remainder of the nursing staff. This group still exists and has at
least 6–8 meetings/year in which all problems regarding the implementation of
CRRT are dealt with.
In the early phase, pulse-HVHF was performed using Gambro AK-10
Ultra. This technique was initially developed by our nephrologist team who,
like us, were on call 24 h/day throughout the year. Indeed, they have been doing
HVHF in almost 50% of their chronic renal failure patients for more than
25 years [15]. Later this technique was taken over by us but still we work in
close relationship with the nephrology team. Indeed, the both the ICU and
nephrology units are very close allowing easy access to every box in the ICU
that is equipped with an internal circuit of ultrapure water as part of the big
nephrology circuit.
The biggest challenge that we faced was the change from BSM 22 and
BM 25 to Prisma at the end of 1993, and further changes from Prisma to
Aquarius at the end of 2001. The latest change occurred at the end of 2001
when the 35-ml/kg/h rule was implemented in our ICU. In 2001 we also took
over the pulse-HVHF technology. The next challenge will be the implementa-
tion of citrate in our unit [1, 7].
Honoré/Joannes-Boyau/Gressens 358
These frequent and important changes require a MECT to be on call,
sometimes just for technical problems. This team allows very rapid implemen-
tation of a new technique/machine in the unit. Undoubtedly, this type of team
enables the use of various approaches with regard to CRRT in the unit including
pulse-HVHF. Thereby filter lifespan has also greatly increased.
It appears that the integration of a dedicated medical team plus a nursing
group is really a condition for success with this kind of therapy. A close relation-
ship with the nephrology team is also a key importance. Continuous education is
also a crucial point regarding this technique which needs continuous review of all
existing protocols [1, 7]. So, the world ‘continuous’ is not only for the technique
but also for ‘educational’ process itself as shown by other groups as well [16, 17].
Our Strategy for Implementing the 35-ml/kg/h Rule
All the rigorous education processes performed inside the nursing and the
medical teams in order to enable the change from 20 to 35 ml/kg/h [18] will
now be described.
At the beginning of 2002, we decided to implement 35 ml/kg/h as a stan-
dard CRRT dose for all our patients according to the study of Ronco et al. [19].
For the past 15 years, we have been using CRRT with BSM 22, BM 25, Prisma
and finally Aquarius.
A nursing group specially dedicated to CRRT was created in our unit more
than 10 years ago. This group allows us to have nurses dedicated to CRRT in
every shift. This is also very helpful when new machines or techniques are imple-
mented. Such a change needs a rigorous education process, first in the nursing
group and then in the whole nursing team as well physicians and colleagues.
Our hospital is a regional general hospital with 455 beds, serving a popu-
lation of 150,000 inhabitants; and it has a medical-surgical general ICU with
15 beds and 1,000 admissions a year. The unit performs CRRT on between 50
and 70 patients per year.
All the changes made in daily practice are summarized below and need
seven crucial changes (fig. 1).
(1) Vascular access. A 14-french coaxial catheter is needed. The catheter
site must exclusively be by a right internal jugular approach (posterior), 20 cm
length, with the tip of the catheter placed in the right atrium (fig. 2). In order to
perform 35 ml/kg/h in each patient regardless of the body weight (from 50 up to
150 kg), a blood flow of 300 ml/min is required in order to keep the filtration
fraction around and below 30% (fig. 3).
For a 50-kg patient 35 50 1,750l/h which gives an filtration fraction
of less than 15% but for a 150-kg patient 35 150 5,250 l/h, this gives an
CRRT Technology and Logistics 359
filtration fraction of 29%. Obviously the blood flow cannot keep changing
dependent on the body weight of the patient. Very strict rules must be applied.
The choice of the right internal jugular approach needs a strong consensus
among all the ICU consultants, and again very strict rules have to be applied.
(2) Pre- and post-dilution policy. According to the latest literature, we
chose 33% pre-dilution and 66% post-dilution. This policy allows us to prevent
some clotting and some clogging by using a certain amount of pre-dilution.
This policy also permits a good convection rate. Indeed, the loss of convection
by the concomitant use of pre-dilution reduces the risk of fouling and protein
cake formation and finally preserves further loss of convection.
(3) Pre-determination of the exchange rate according to the body weight.
In order to reduce the workload at the beginning of the CRRT, we introduced
full tables which automatically give the amount of pre- and post-dilution for a
given body weight, eliminating the need for sophisticated calculation that can
lead to further mistakes (fig. 4). Additional tables exist also for the adaptation
of antibiotics to 35 ml/kg/h and other drugs as well. Control of delivery is also
very Important and therefore, we use special dedicated forms to monitor nurs-
ing (fig. 5).
(4) No further need of associated dialysis. During the last 3 years, even in
very severe rhabdomyolysis cases, no need of associated dialysis was required
no matter what the release of potassium was. This enabled us to use the
35ml/kg/h
Vascular access (type and location)
→ (RIJ 20 cm!)
Blood flow (filtration fraction)
→ 300 ml/min
Pre and post %
→ 33% and 66%
Associate dialysis?
→ no!
Adaptation to body weight
→ Tables …
Increased cost …
→ not that much!
‘Nursing cell’
Fig. 1. The seven arms of the hemofiltration tree.
Honoré/Joannes-Boyau/Gressens 360
Aquarius with combined pre- and post-dilution and also to save money regard-
ing additional cost of fluids.
In conclusions, the implementation of new guidelines concerning the
35-ml/ kg/h rule cannot be decided in one day. This will need a rigorous long-standing
educational process for the nursing staff as well as the medical team. A nursing
group can be of help in order to achieved this. We will soon embark upon the imple-
mentation of citrate using the same implementation and educational process.
Medical Emergency CRRT and Nursing Emergency CRRT Teams
As said earlier, very big challenges occurred with the change from BSM 22
and BM 25 to Prisma at the end of 1993, and further changes from Prisma to
Aquarius at the end of 2001. The latest change was at the end of 2001 when the
Need of a co-axial catheter
of 14 french. The catheter
site had to be exclusively a
Right Internal Jugular
Approach (Posterior one)
with an 20cm length and
the tip of the catheter
placed in the right atrium.
Longitudinal view
Longitudinal view
Axial view
Schematic view
Vascular access
RIJ >
300ml/min
LIJ
L Scl
Swan-Ganz?
Role of IAH and ACS in ICU
for femoral approaches
LF
< 300ml/min
RF
< 300ml/min
Axial view
• Length
– Right internal
jugular: 20cm.
– Femoral: 25cm.
• Diameter: 14 French.
• Type coaxial: 360 degree
arterial intake
R Scl
Multiple
lumen
central line?
Choice of vascular
access
length, diameter
and type of catheter
Fig. 2. Vascular access for CRRT at 35 ml/kg/h.
CRRT Technology and Logistics 361
Blood flow and filtration fraction
FF
• 35ml/kg/h Policy need higher blood flow 300ml/min.
• When applied at a patient of 50kg 1.750l/h FF 10%.
• When applied at a patient of 150kg 5.250l/h FF 29%.
• You need a fixed blood flow.
FR
Barriers to Achieve Prescribed Treatment Dose
Adequate blood access and flow
predilution ratio
QP weight loss
FR
QUF
QBF QUF QP
Fig. 3. Ideal blood flow for hemofiltration at 35 ml/kg/h.
Pre-determination of the
exchange rate according
the body weight: in order to
reduce the workload at the
beginning of the CRRT, we
did introduce full tables
Pre-determination of the exchange rate according the body weight
Use of tables for ‘Adaptation’ to body weight
Weight
50
60
70
80
90
100
120 4,200
4,000
3,500
3,300
3,200
3,000
2,700
2,700
2,600
2,400
2,300
2,100
1,900
1,800 600
600
700
800
800
900
900
1,000
1,100
1,100
1,200
1,200
1,300
1,400
1,300
1,400
1,500
1,600
1,700
1,800
2,000
2,100
2,200
2,300
2,800
115
85
95
55
65
75
Pre-dilution
1/3 dose T
Post-dilution
2/3 dose T
Dose T
35ml/kg/h
Fig. 4. Predetermined tables for pre- and post-dilution ratio at 35ml/kg/h.
35-ml/kg/h rule was implemented in our ICU. Pulse-HVHF technology was
also taken over in 2001. The next challenge will be the implementation of
citrate in our unit [1, 7].
Honoré/Joannes-Boyau/Gressens 362
These frequent and important changes require a MECT, on call sometimes
just for technical problems. This team allows a very quick implementation of a
new technique/machine inside the unit. Undoubtedly, this type of team makes pos-
sible the use of various approaches regarding CRRT in the unit including pulse-
HVHF. Filter lifespan has also greatly increased using this MECT approach.
Also other similar approaches have occurred in the pediatric world this time
using nursing emergency CRRT teams (NECT). In the PICU, Harvey et al. [5]
have described the new role of a renal critical care educator based in a regional
pediatric renal unit reaching out to three PICUs within the region. They demon-
strated that, after 18 months, the training objectives for the PICU staff in all
aspects of hemofiltration had progressed in all units using flexible training
Monitoring
Control of treatment dose and dose delivery
In practice: The St-Pierre way
Date : ..............
à ...............
à ...............
Vignette D’identification
Heures Bicaflac/
physio
TCA Dilution
Anticoagulation
Doses Substitution Modification
déplétion
Data
Déplet/3h Déplet/tot Art Vein
Pressions
PTM
Perte
de
charge
Fraction
de
filtration
<30%
n du set utilisé : ...............
Filtre placé : Ie ...............
changé : Ie ...............
Anticoagulation Bolus OUI/NON
Continu OUI/NON
Thérapie : CWHF
Nb J d’ E.E.R : .............
Débits
Sang ............... ml/min
Substitution totale ............... ml/h
2/3 Post-dilution ............... ml/h
1/3 Pré-dilution ............... ml/h
Déplétion Horaire ............... ml/h
Surveillance hemofiltration
continue
Aquarius Edwards Clinique St Pierre Ottignies
Pressions de références
Débit de sang : 300 ml/min
Pression artérielle : 170 à 190
Pression veinéuse : 200 à 220
Pres. trans. men : 30 à 400
Fig. 5. Nursing monitoring data form.
CRRT Technology and Logistics 363
programs which were simultaneously under constant evaluation using question-
naires. This type of renal critical care educator also worked alongside staff whilst
CVVH was being performed to further instill confidence. During the same time
period, equipment and protocols have been largely standardized throughout the
region, and an ongoing survey of CVVH use was initiated which could help to
inform audit standards such as complication rates. It was concluded that the
renal critical care educator was the catalyst for the formation of a regional
hemofiltration group which shared in the development of guidelines and proto-
cols and discussion of clinical data. Indeed, since CRRT is used infrequently in
many PICUs the development of the renal critical care educator could serve as a
model for the development and maintenance of skills in other regions [5]. This
could be seen in some ways by the nurse emergency team for CRRT in ICU.
Conclusions and Perspectives
The implementation of guidelines, MECT, NECT, and educational
processes is really the key issue for the future of CRRT in the ICU. The imple-
mentation of new guidelines with regard to the 35-ml/kg/h rule cannot be
decided in one day. This will need a rigorous longstanding educational process
for the nursing staff as well the medical team. A nursing group can be of help in
order to achieved this. We will soon embark upon the use of citrate using the
same implementation and educational processes.
First education processes and ongoing education are the keys to success in
the use of CRRT. Local experts are needed in the nursing team, the medical ICU
team and the nephrology team as well.
References
1 Renard D, Douny N, Verly Ch, Lourtie A-M, Honoré PM: Nursing monitoring of continuous
haemofiltration; in Robert R, Honoré PM, Bastien O (eds): Extra-Corporeal Circulation in ICU.
The French SRLF Collection Europe, Collection Director: Prof F Saulnier. Paris, Elsevier, 2006,
pp 179–194.
2 Shetz M, Lauwers P, Ferdinande P, Van de Walle J: The use of continuous arteriovenous haemofil-
tration in intensive care medicine. Acta Anaesthesiol Belg 1984;35:67–78.
3 Clevenger K: Setting up a continuous venovenous hemofiltration educational program. A case
study in program development. Crit Care Nurs Clin North Am 1998;10:235–244.
4 Giuliano KK, Pysznik EE: Renal replacement therapy in critical care: implementation of a unit-
based continuous venovenous hemodialysis program. Crit Care Nurs 1998;18:40–51.
5 Harvey B, Watson AR, Jepson S: A renal critical care educator: the interface between paediatric
intensive care and nephrology. Intensive Crit Care Nurs 2002;18:250–254.
6 East D, Jacoby K: The effect of a nursing staff education program on compliance with central line
care policy in the cardiac intensive care unit. Pediatric Nurs 2005;31:182–184.
Honoré/Joannes-Boyau/Gressens 364
7 Honoré PM, Piette V, Galloy A-C, Almpanis C, Pelgrim J-P, Dugernier Th: High volume haemofil-
tration: general review; in Robert R, Honoré PM, Bastien O (eds): Extra-Corporeal Circulation in
ICU. The French SRLF Collection Europe, Collection Director: Prof F Saulnier. Paris, Elsevier,
2006, pp 195–220.
8 Malbrain MNG, Chiumello D, Pelosi P, Wilmer A, Brienza N, Malcangi V, et al: Prevalence of
intra-abdominal hypertension in critically ill patients: a multicentre epidemiological study.
Intensive Care Med 2004;30–822–829.
9 Kingston D, Sykes S, Raper S: Protocol for the administration of haemofiltration fluids and using
patients group electrolytes direction. Nurs Crit Care 2002;7:193–197.
10 Honoré PM, Joannes-Boyau O, Merson L, Boer W, Piette V, Galloy AC: The big bang of haemofil-
tration: the beginning of a new era in the third millennium for extra-corporeal blood purification!
Int J Artif Organs 2006;29:649–659.
11 Rahman TM, Treacher D: Management of acute renal failure on the intensive care unit. Clin Med
2002;2:108–113.
12 Winkelman C: Haemofiltration: a new technique in critical care nursing. Heart Lung 1985;14:
265–271.
13 Honoré PM, Jamez J, Wauthier M, Dugernier Th: Prospective evaluation of short time high vol-
ume isovolemic haemofiltration on the haemodynamic course and outcome of patients with
refractory septic shock. Crit Care Nephrol 1998;90:87–99.
14 Honoré PM, Jamez J, Wauthier M, Lee PA, Dugernier Th, Pirenne B, Hanique G, Matson JR:
Prospective evaluation of short term high volume isovolemic haemofiltration on the haemody-
namic course and outcome in patients with intractable circulatory failure resulting from septic
shock. Crit Care Med 2000;28:3581–3587.
15 Troch R, Van Ypersele de Strihou C: Home haemofiltration/dialysis: experience in Belgium. Acta
Clin Belg 1974;29:218–224.
16 Craig M: Continuous venous to venous haemofiltration. Implementing and maintaining a pro-
gram: examples and alternatives. Crit Care Nurs Clin North Am 1998;10:219–233.
17 Hameleers P: 24 hour CVVH treatment on the intensive care unit: a reality. EDTNA ERCA
J 1998;24:21–22.
18 Honoré PM, Joannes-Boyau O, Kotulak T, Boer W, Renard D, Verly C: Implementation of 35 ml/kg/h
of CRRT dose in ICU: a combined medical and nursing approach. Blood Purif 2006;24:261–262.
19 Ronco C, Bellomo R, Homel P, et al: Effects of different doses in continuous veno-venous haemofil-
tration in outcomes of acute renal failure: a prospective randomised trial. Lancet 2000;356:26–30.
Patrick M. Honoré, MD
St-Pierre Para-University Hospital
Avenue Reine-Fabiola, 9
BE–1340 Ottignies-Louvain-la-Neuve (Belgium)
Tel. 32 10 437 346, Fax 32 10 437 123, E-Mail Pa.honore@clinique-saint-pierre.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 365–370
Continuous Hemodiafiltration with
Cytokine-Adsorbing Hemofilter in the
Treatment of Severe Sepsis and
Septic Shock
Hiroyuki Hirasawa
a
, Shigeto Oda
a
, Kenichi Matsuda
b
a
Department of Emergency and Critical Care Medicine, Chiba University
Graduate School of Medicine, Chiba, and
b
Department of Emergency
and Critical Care Medicine, Yamanashi University School of Medicine,
Yamanashi, Japan
Abstract
Continuous hemodiafiltration (CHDF) using a polymethymethacrylate (PMMA) mem-
brane hemofilter (PMMA-CHDF) can effectively and continuously remove various
cytokines from the circulating blood. PMMA-CHDF can decrease the blood levels of various
cytokines when the blood levels of cytokines are high prior to the initiation of CHDF. The
main mechanism of cytokine removal with PMMA-CHDF is the adsorption of cytokines to
the hemofilter membrane and this characteristic was not observed in the other membrane
material. PMMA-CHDF could improve blood pressure, the depressed monocytic HLA-DR
expression, and recover the delayed neutrophil apoptosis in septic patients. Thus, cytokine
removal with PMMA-CHDF would be effective for the treatment of severe sepsis and septic
shock.
Copyright © 2007 S. Karger AG, Basel
It is widely accepted that cytokines play a pivotal role in the pathophysiol-
ogy of severe sepsis and septic shock [1–3]. However, despite recent progress in
understanding the pathophysiology of sepsis under the wide application of mol-
ecular biology, no anticytokine therapy has been effectively applied in clinical
settings in the management of severe sepsis and septic shock [4]. On the other
hand, there is ongoing controversy regarding the efficacy of cytokine removal
in septic patients with various types of continuous blood purifications such as
Hirasawa/Oda/Matsuda 366
continuous hemofiltration, continuous hemodialysis and continuous hemodi-
afiltration (CHDF) [5, 6]. Ronco et al. [7] proposed the ‘peak concentration
hypothesis’ indicating that the removal of cytokines with continuous renal
replacement therapy or continuous blood purification is effective to modulate
the inflammatory response in sepsis syndrome. Some investigators indicated
that hemofiltration is effective in the treatment of sepsis if it is performed with
high-volume filtration [8, 9], especially as salvage therapy in patients with
severe hyperdynamic septic shock [9].
We reported that CHDF using a polymethylmethacrylate (PMMA) mem-
brane hemofilter (PMMA-CHDF) could effectively and continuously remove
various cytokines from the circulating blood of a patient mainly through the
adsorption of cytokines to the hemofilter membrane, and that PMMA-CHDF
decreases the blood levels of not only proinflammatory cytokines but also anti-
inflammatory cytokines when the blood levels of those cytokines are high prior
to the initiation of CHDF [10, 11] (fig. 1). Furthermore, we also reported that
such cytokine removal with PMMA-CHDF was effective for the treatment of
hypercytokinemia-related pathophysiology [12] such as severe sepsis, septic
shock [13], septic acute respiratory distress syndrome, septic multiple organ
failure [14, 15], and severe acute pancreatitis [16].
High-level
group
(n24)
Low-level
group
(n18)
Blood level (pg/ml)
10
1
1
10
1
10
2
10
3
NS
p 0.01
Pre
Post
Pre
Post
TNF (MW: 17 kDa)
High-level
group
(n46)
Low-level
group
(n9)
Blood level (pg/ml)
p 0.01
NS
10
1
10
2
10
3
10
4
10
5
1
Pre
Post
Pre
Post
IL-6 (MW: 21 kDa)
High-level
group
(n25)
Low-level
group
(n20)
Blood level (pg/ml)
10
1
p 0.01
NS
10
1
10
2
10
3
10
4
1
Pre
Post
Pre
Post
IL-8 (MW: 8 kDa)
High-level
group
(n16)
Low-level
group
(n9)
Blood level (pg/ml)
p 0.01
NS
10
1
10
1
10
2
10
3
10
4
1
Pre
Post
Pre
Post
IL-10 (MW: 19 kDa)
Fig. 1. Changes in blood levels of cytokines with 3 days of CHDF (mean SD).
CHDF with Cytokine Adsorbing Hemofilter for Severe Sepsis 367
There are three different mechanisms for the removal of cytokines with
CHDF, namely convection, diffusion and adsorption of cytokines to the
hemofilter. As shown in figure 2, there is a significant positive correlation
between the blood level of interleukin (IL)-6 and the clearance of IL-6 with
PMMA-CHDF: the higher the blood level, the greater the clearance. These
results indicate that the main mechanism of cytokine removal with PMMA-
CHDF is through the adsorption of cytokines to the hemofilter. Furthermore,
these features of PMMA-CHDF on cytokine clearance are very much clinically
beneficial when we apply PMMA-CHDF as a cytokine remover because
PMMA-CHDF is especially effective when a patient shows a severe degree of
hypercytokinemia [10, 11, 13]. On the other hand, as shown in figure 2, there is
no significant positive correlation or even a significant negative correlation
between the IL-6 blood level and the clearance of IL-6 with CHDF when CHDF
was performed with a hemofilter made from membrane materials other than
PMMA, such as polyacrylonitrile (PAN) and polysulfone (PS). Thus the mem-
brane material of the hemofilter is crucial when we apply CHDF as a cytokine
modulator. We tried CHDF with various hemofilters made from many kinds of
membrane materials and found that the PMMA hemofilter is superior to
hemofilters made from other membrane materials when CHDF is applied as
cytokine modulator [10, 11].
Since the main mechanism of cytokine removal with PMMA-CHDF is
the adsorption of cytokine to the hemofilter membrane, we need not to apply
troublesome high-volume CHDF [8, 9] to remove cytokines as proposed by
y 11.5 log x 26.2
r 0.56, p 0.01
n 40
60
40
20
0
20
40
60
10 10
2
10
3
10
4
1
Blood level (pg/ml)
PMMA
IL-6 clearance (ml/min)
PAN
y8.51 log x 27.0
r 0.50, p0.01
n26
60
40
20
0
20
40
60
10 10
2
10
3
10
4
Blood level (pg/ml)
IL-6 clearance (ml/min)
PS
y12.3 log x 39.4
r 0.59, p0.06
n11
60
40
20
0
20
40
60
10
2
10
3
10
4
10
Blood level (pg/ml)
IL-6 clearance (ml/min)
Fig. 2. Correlation between clearance and IL-6 blood level among various hemofilters
SFI. Hemofilter 1.0 m
2
; Q
B
60 ml/min; Q
D
500 ml/h; Q
F
300 ml/h. PAN
Polyacrylonitrile; PMMA polymethylmethacrylate; PS polysulfone.
Hirasawa/Oda/Matsuda 368
others who use hemofilters made of materials other than PMMA. We also do
not need large blood flows to perform PMMA-CHDF as a cytokine modula-
tor. Actually we perform PMMA-CHDF with a filtration rate of 300–500 ml/h
and dialysate rate of 500–1,000 ml/h and blood flow of 60–100 ml/h [10, 11].
In 2000, we introduced a rapid measurement system for the IL-6 blood
level using an automated chemiluminescent enzyme immunoassay which can
measure the IL-6 blood level within 30 min in the clinical laboratory [13]. Since
then, we scientifically determine the indication and the timing of initiation and
termination of PMMA-CHDF for cytokine removal [13]. Taking the results
from previous studies in our laboratory [12, 14, 15], we now apply PMMA-
CHDF when a septic patient shows IL-6 blood levels of 1,000 pg/ml [13]. We
reported that PMMA-CHDF could improve blood pressure and could recover
decreased urinary output within 2 h following the initiation [13]. One of the
important aspects of severe sepsis and septic shock is immunoparalysis due to
an excessive anti-inflammatory response caused by anti-inflammatory cytokines,
such as IL-10. A decrease in monocytic HLA-DR expression has been impli-
cated as an indicator of immunoparalysis. We found that PMMA-CHDF was
able to remove IL-10 and that such a removal of anti-inflammatory cytokines
resulted in an improvement in the depressed monocytic HLA-DR expression in
septic patients. We also reported that delayed apoptosis of neutrophils, which is
an important pathogenesis of organ failure in sepsis, is recovered through the
removal of cytokines with PMMA-CHDF [17]. These are the mechanisms of
the efficacy of PMMA-CHDF in the treatment of severe sepsis and septic
shock.
Recently it was reported that cytokine-related gene polymorphism plays
an important role in determining the degree of inflammatory response to an
insult such as infection [18]. Therefore, we investigated the effect of cytokine-
related genetic polymorphism on septic patients and reported that septic
patients with a cytokine-related genetic polymorphism have more severe
hypercytokinemia compared to septic patients without a cytokine-related
genetic polymorphism [19–21]. More importantly we also reported that such
cytokine-related genetic polymorphisms affected the efficacy of PMMA-
CHDF on septic patients. In septic patients with a cytokine-related genetic
polymorphism, PMMA-CHDF is less effective in controlling the cytokine
storm and the survival rate is worse in those patients with genetic polymor-
phisms compared to those without cytokine-related genetic polymorphisms,
even though patients in both groups received same therapeutic approaches
including PMMA-CHDF [19, 20]. Taking these results, we recently started a
prospective study to investigate the efficacy of personalized application of
PMMA-CHDF based on the examination of cytokine-related genetic polymor-
phism in our intensive care unit population. We applied two PMMA-CHDF
CHDF with Cytokine Adsorbing Hemofilter for Severe Sepsis 369
lines at the same time to a patient to enhance the removal rate of cytokine with
PMMA-CHDF when the patient was diagnosed as having a cytokine-related
genetic polymorphism. The results of this prospective study will hopefully be
published in the near future.
References
1 Tetta C, Bellomo R, D’Intini V, et al: Do circulating cytokines really matter in sepsis? Kidney Int
2003;63(suppl):S69–S71.
2 Wang H, Czura CJ, Tracey KJ: Lipid unites disparate syndromes of sepsis. Nat Med 2004;10:
124–125.
3 Lotze MT, Tracey KJ: High-mobility group box 1 protein (HMGB1): nuclear weapon in the
immune arsenal. Nat Rev Immunol 2005;5:331–342.
4 Kox WJ, Volk T, Kox SN, et al: Immunomodulatory therapies in sepsis. Intensive Care Med
2000;26:S124–S128.
5 Schetz M, Ferdinande P, van den Berghe G, et al: Removal of pro-inflammatory cytokines with
renal replacement therapy: sense or nonsense? Intensive Care Med 1995;21:169–176.
6 Heering P, Grabensee B, Brause M: Cytokine removal in septic patients with continuous venove-
nous hemofiltration. Kidney Blood Press Res 2003;26:128–134.
7 Ronco C, Tetta C, Mariano F, et al: Interpreting the mechanisms of continuous renal replacement
therapy in sepsis: the peak concentration hypothesis. Artif Organs 2003;27:792–801.
8 Cole L, Bellomo R, Journois D, et al: High-volume haemofiltration in human septic shock.
Intensive Care Med 2001;27:978–986.
9 Cornejo R, Downey P, Castro R, et al: High-volume hemofiltration as salvage therapy in severe
hyperdynamic septic shock. Intensive Care Med 2006;32:713–722.
10 Matsuda K, Hirasawa H, Oda S, et al: Current topics on cytokine removal technologies. Ther Apher
2001;5:306–314.
11 Nakada T, Hirasawa H, Oda S, et al: Blood purification for hypercytokinemia. Transfusion Apher
Sci 2006;35:253–264.
12 Hirasawa H, Oda S, Matsuda K, et al: Clinical aspect of hypercytokinemia-induced pathophysiol-
ogy in critical care; in Ogawa M, Yamamoto T, Hirota M (eds): The Biological Response to
Planned and Unplanned Injuries: Cellular, Molecular and Genetic Aspects. Excerpta Medica. Int
Congr Ser 1255. Amsterdam, Elsevier, 2003, pp 39–40.
13 Oda S, Hirasawa H, Shiga H, et al: Sequential measurement of IL-6 blood levels in patients with
systemic inflammatory response syndrome/sepsis. Cytokine 2005;29:169–175.
14 Hirasawa H, Baue AE: Blood purification therapy to prevent or treat MOF; in Baue AE, Faist E,
Fry DE (eds): Multiple Organ Failure: Pathophysiology, Prevention, and Therapy. New York,
Springer, 2000, pp 501–504.
15 Hirasawa H, Oda S, Shiga H, et al: Endotoxin adsorption or hemodiafiltration in the treatment of
multiple organ failure. Curr Opin Crit Care 2000;6:421–425.
16 Oda S, Hirasawa H, Shiga H, et al: Management of intra-abdominal hypertension in patients with
severe acute pancreatitis with continuous hemodiafiltration using a polymethyl methacrylate
membrane hemofilter (PMMA-CHDF). Ther Apher Dial 2005;9:355–361.
17 Hirano T, Hirasawa H, Oda S, et al: Modulation of polymorphonuclear leukocyte apoptosis in the
critically ill by removal of cytokines with continuous hemodiafiltration. Blood Purif 2004;22:
188–197.
18 Ma P, Chen D, Pan J, et al: Genomic polymorphism within interleukin-1 family cytokines influ-
ences the outcome of septic patients. Crit Care Med 2002;30:1046–1050.
19 Watanabe E, Hirasawa H, Oda S, et al: Extremely high interleukin-6 blood levels and outcome in
the critically ill are associated with tumor necrosis factor- and interleukin-1-related gene polymor-
phisms. Crit Care Med 2005;33:89–97.
Hirasawa/Oda/Matsuda 370
20 Nakada K, Hirasawa H, Oda S, et al: Influence of toll-like receptor 4, CD14, tumor necrosis fac-
tor, and interleukin-10 gene polymorphism on clinical outcome in Japanese critically ill patients.
J Surg Res 2005;129:322–328.
21 Watanabe E, Hirasawa H, Oda S, et al: Cytokine-related genotypic differences in peak interleukin-6
blood levels of patients with SIRS and septic complications. J Trauma 2005;59:1181–1190.
Dr. Shigeto Oda
1-8-1 Inohana, Chuo
Chiba City
Chiba 260–8677 (Japan)
Tel. 81 43 226 2341, Fax 81 43 226 2371, E-Mail odas@faculty.chiba-u.jp
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 371–386
Blood and Plasma Treatments: High-Volume
Hemofiltration – A Global View
Patrick M. Honoré
a
, Olivier Joannes-Boyau
b
, Benjamin Gressens
a
a
St-Pierre Para-University Hospital, Ottignies-Louvain-la-Neuve, Belgium;
b
Haut Leveque University Hospital, University of Bordeaux, Pessac, France
Abstract
From the recent past, hemofiltration, particularly high-volume hemofiltration, has
rapidly evolved from a somewhat experimental treatment to a potentially effective ‘adjunc-
tive’ therapy in severe septic shock and especially refractory or catecholamine-resistant hypo-
dynamic septic shock. Nonetheless, this approach lacks prospective randomized studies
(PRTs) evaluating the critical role of early hemofiltration in sepsis. An important milestone,
which could be called the ‘big bang’ in terms of hemofiltration, was the publication of a PRT
in patients with acute renal failure (ARF). Before this study, nobody believed that hemofiltra-
tion could change the survival rate in intensive care. Since that big bang, many physicians
consider that hemofiltration at a certain dose can change the survival rate in intensive care. We
now must try to define what the exact dose in septic ARF should be. As suggested by many
studies this dose might well be higher than 35 ml/kg/h in the septic ARF group. The issue of
the dosage of continuous high-volume hemofiltration must be tested in future randomized
studies. Since the Vicenza study has shown that 35 ml/kg/h is the best dose in terms of survival
when dealing with nonseptic ARF in the intensive care unit (ICU), several studies from differ-
ent groups have shown that a higher dose might be correlated with better survival in septic
ARF. This has also been shown in some way by the Vicenza group but not with a statistically
significant value. New PRTs have just started in Europe such as the IVOIRE (hIgh VOlume in
Intensive Care) study. The RENAL study is another large study looking more basically at dose
in nonseptic ARF in Australasia. The ATN study in the USA is also testing the importance of
dose in the treatment for ARF. Nevertheless, ‘early goal-directed hemofiltration therapy’ has
to be studied in our critically ill patients. Regarding this issue, fewer studies, mainly retro-
spective, exist; but again the IVOIRE study will address this issue by studying septic patients
with acute renal injury according to the RIFLE classification. This chapter will focus on the
early application and adequate dose of continuous high-volume hemofiltration in septic shock
in order to improve not only the hemodynamics but also survival in this very severely ill
cohort of patients. This could be called the big bang of hemofiltration as one could have never
anticipated that an adequate dose of hemofiltration could markedly influence the survival rate
of septic ARF patients in the ICU. Apart from the use of an early and adequate dose of
Honoré/Joannes-Boyau/Gressens 372
hemofiltration in sepsis, a higher dose could also provide a better renal recovery rate and
reduce the risk of associate chronic dialysis in these patients. Furthermore, this presentation
will also review brand-new papers regarding the use of hemofiltration in systemic inflamma-
tory response syndrome and out-of-hospital cardiac arrest.
Copyright © 2007 S. Karger AG, Basel
Global Approach to the Problem in the ICU
Over the last two decades, hemofiltration and especially high-volume
hemofiltration (HVHF) have rapidly evolved from a somewhat experimental
treatment to a potentially effective ‘adjunctive’ therapy in severe septic shock
and refractory or catecholamine-resistant (hypodynamic) septic shock (CRSS).
Nevertheless, this approach lacks prospective randomized studies (PRTs)
evaluating the critical role of early hemofiltration in sepsis.
An important step forward, which could be called the ‘big bang’ in terms
of hemofiltration [1], was the publication of a PRT in patients with acute renal
failure (ARF) [2]. Before this study [2] nobody believed that hemofiltration
could change the survival rate in intensive care. Since that ‘big bang’, many
physicians think that ‘correctly dosed’ hemofiltration has the potential to
change the survival rate in intensive care. So the world of hemofiltration in the
intensive care unit (ICU) is not a definite world; it is still in expansion. Right
now, we have to try to define what will be the exact dose needed in septic ARF.
This dose might well be higher than 35ml/kg/h in the septic ARF group as sug-
gested by many studies [2–5].
In order to challenge this hypothesis, a continuous dose rather than a
pulsed dose of HVHF will be tested in ongoing or future randomized studies.
As the Vicenza study [2] has shown that 35 ml/kg/h is the best dose in terms of
survival, while dealing with nonseptic ARF in ICU, several studies from differ-
ent groups have shown that, in septic ARF, a higher dose might correlate with
better survival [3–5]. In some way this has also been shown by the study of
Ronco et al. [2], but unfortunately not with statistically significance.
New PRTs have just started in Europe, such as the IVOIRE (hIgh VOlume
in Intensive Care) study [6] and the RENAL study; another large study looking
more basically at dose in nonseptic ARF is ongoing in Australasia [7], as well as
the ATN study in the USA which also tests the importance of dose in the treat-
ment of ARF. Nevertheless, early goal-directed hemofiltration therapy [8] has
to be studied in critically ill patients. In this regard, fewer mainly retrospective
studies do exist, but again the IVOIRE study [6] will address this issue by
studying septic patients with acute renal injury according to the RIFLE classifi-
cation [9] rather than ARF, which is already late in the illness process.
HVHF: A Global View 373
This review will focus on the early application and adequate dosing of con-
tinuous HVHF in septic shock in order to improve not only hemodynamics but
also survival in this very severely ill cohort of patients.
This could be called the ‘big bang of hemofiltration’ as one could never
have anticipated the fact that an adequate dose of hemofiltration could markedly
influence the survival rate in the ICU while treating septic ARF patients.
As well as the use of early and adequate dose of hemofiltration in sepsis, a
higher dose could also provide better renal recovery rate and reduce the risk of
chronic dialysis dependency in these patients.
Big Bang of Hemofiltration and Hemofiltration-Derived Therapies
The ‘big bang’ of clinical hemofiltration occurred when the study of
Ronco et al. [2] was published in the Lancet in 2000. This study comprised 425
patients and prospectively assessed three dosages, 20, 35 and 45 ml/kg/h,
respectively. Exchange was exclusively realized in post-dilution in order to
maximize convection. The membrane was also changed every 24 h. This study
demonstrated that in nonseptic ARF a dose of 35 ml/kg/h was correlated with
the best survival rate. This difference was statistically significant. This can be
expressed as a big bang in terms of hemofiltration [10]. Before that study,
nobody was really thinking realistically that hemofiltration could ever change
survival in intensive care. Since that big bang, it is now believed by many that
the use of a correctly dosed hemofiltration has the potential to change the sur-
vival rate in intensive care patients. Now the exact dose needed in septic ARF
must be defined. This dose will probably be higher than 35 ml/kg/h in the septic
ARF group, as has been suggested by many studies [4, 5, 11–16].
This has been put into perspective by an interesting recently published
review paper [17] highlighting all the potentials of extracorporeal therapy with
regard to the wide spectrum of molecules that could be removed, ranging in the-
ory from 500 up to 900,000 kDa when using plasmafiltration for instance. Other
outcome-dose studies also participate in the expansion of this initial big bang,
among which the IVOIRE [6], RENAL [7] and ATN take the lead. The IVOIRE
study [6] is trying to expand the findings of the initial study by Ronco et al. [2]
to the septic ICU. It will include more than 480 patients with septic shock plus
acute renal injury as defined by the RIFLE classification in ICU [9]. Allocation
into the two arms will be determined by computerized randomization. One
group will receive 35 vs. 70 ml/kg/h in the other group. This study will try to
demonstrate that the higher dose (70 ml/kg/h) will further improve the survival
rate of septic ARF patients in the ICU at 28, 60 and 90 days, respectively, after
ICU admission.
Honoré/Joannes-Boyau/Gressens 374
Animal Models with Their Strengths and Pitfalls
Animal models have shown benefits in term of survival when an early and
strong hemofiltration dose was applied in septic animal models. The early use
of hemofiltration has been well applied in many animal models. The earlier
experiments by Grootendorst et al. [18–21] most of the time used hemofiltra-
tion before or just after the injection of a bolus or even infusion of endotoxin.
Only in the late 1990s, with the studies of Rogiers et al. [22, 23], did the
investigators start waiting for about 6–12 h before using HVHF, thereby allow-
ing the animals to become extremely ill, hemodynamically unstable with early
multiple organ dysfunction syndrome. In this way, the animal model was able to
mimic the clinical situation in some way. Only animal models that were submit-
ted to the early application of HVHF were shown to be very beneficial (and in
some ways with very impressive results) mainly by the fact that in addition to
early use, the investigators applied a much stronger dose of HVHF. By aggre-
gating 12 studies over the last 10 years with regard to animal models, Honoré
et al. [24] have shown that the mean dose used in those experiments was about
100 ml/kg/h whereas for humans (last 13 human studies or so) only 40 ml/kg/h
was effectively given. The most beneficial effects of HVHF have been shown
by these animal models, although the maximum delay between the septic insult
and intervention was less than 12 h. This is totally different from the clinical
situation in which the delay is rarely below 24 h and/or even below 48 h. The
literature shows that in animal models not only a stronger application of dose
was very important but also that early application was the second most impor-
tant condition to make the use of hemofiltration in sepsis beneficial in terms of
hemodynamics and survival [25] (table 1). It has also been advocated that the
best response seen in the animal model was obtained when sepsis was intravas-
cular as opposed to extravascular or when sepsis was restricted in another more
confined compartment, for instance peritonitis. This might explain why the use
of high-permeability hemofiltration (HPHF) in a sheep model of peritonitis,
described by Rogiers [26], was not able to show any beneficial effect. If we
believe the ‘mediator delivery hypothesis’ [27], we can see that the absence of
large intakes of fluids in HPHF has major consequences: no increase in the
lymphatic flow. This increased lymphatic flow is in charge of retrieving massive
amounts of cytokines and mediators from the interstitium and the tissue level
back to the blood compartment level, making them available for extra-renal
removal.
It makes more sense to think that HVHF could be efficient in an acute
model of peritonitis as demonstrated by Rogiers et al. [23], but HPHF remains
ineffective in that specific setting of this particular study [27]. It is also known
that in this kind of animal model the cytokine pattern is very different from the
HVHF: A Global View 375
Table 1. HVHF in animal studies: survival
Reference Material Membrane Hemofiltration UF/h Timing Animal LD UF/h Survival Effect
⫹ surface technique weight % indexed
kg to body % Length
size
ml/kg/h
Freeman Dog PS HVHF 600 60 min 10 100 60 142 (T) 7 days (T) N
et al. [59] Peritonitis S: NA (CAVH) after 14.2 (C) 7 days (C)
peritoneal
clot
Rogiers Sheep PS VHVHF NA000 4 h after NA 100 100 100 (T) 14 h (T) P
et al. [60] Peritonitis S: NA (CVVH) surgery 0 (C) 14 h (C)
Cecostomy
Yekebas Pig CA-SMS VHVHF NA Immediately NA 100 100 67 (T) 60 h (T) P
et al. [30] pancreatitis S: 0.6 m
2
(CVVH) after 20 0 (C) 60 h (C)
⫹ sepsis induction of
pancreatitis
Lee et al. [61] Pig PS VHVHF 1,000 Immediately 7.5 100 133 38.5 (T) 7 days (T) P
S. aureus S: NA (CAVH) after 0 (C) 7 days (C)
infusion S. aureus
infusion
Grootendorst SMA C.P VHVHF 6,000 Before 37 100 150 66 (T) 24 h (T) P
et al. [21] clamping S: 2.3 m
2
(CVVH) clamping 0 (C) 24 h (C)
C ⫽ Control; CAVH ⫽ continuous ambulatory venous hemofiltration; CP ⫽ cuprophane; CVVH ⫽ continuous veno-venous hemofiltration;
HVHF ⫽ high-volume hemofiltration; N ⫽ negative; N ⫽ no; NA ⫽ not available; P ⫽ positive; PS ⫽ polysulfone; S ⫽ surface; T ⫽ treated;
VHVHF ⫽ very high-volume hemofiltration; Y ⫽ yes.
Honoré/Joannes-Boyau/Gressens 376
human situation because the duration of the proinflammatory phase is much
longer in the animal setting and is not always followed by an secondary
immunoparalysis phase. What the animal model has brought to light is the use
of HVHF as a prophylactic measure in the second phase of sepsis, the so-called
immunoparalysis phase or the compensatory anti-inflammatory response syn-
drome phase as explained by Bone [28]. Yekebas et al. [29–31] and Wang et al.
[32] have worked to modernize this immunoparalysis post- systemic inflamma-
tory response syndrome (SIRS). Therefore, they induced traumatic pancreatitis
in healthy pigs. Hemofiltration started 12 h after pancreatic trauma but before
sepsis and shock occurred. They waited about 12 h and, after this time, bacterial
translocation occurred and very fulminant peritonitis and intravascular sepsis
occurred inducing a shock state in these pigs. By comparing different settings
of hemofiltration, especially low dose (20 ml/kg/h) plus adsorption and HVHF
at the rate of 100ml/kg/h, the authors were able to demonstrate that the prophy-
lactic use of HVHF (100 ml/kg/h) was able to reduce the immunoparalysis level
and the subsequent risk of secondary infection and, ultimately, the death rate.
So, for the first time, HVHF was able to show that it could work not only in the
proinflammatory phase but also in the secondary immunoparalysis phase per se
as a prophylactic measure.
Transposition of these findings to the human setting is even more difficult
because most of the animal models used a so-called hypodynamic septic shock
[33]. It is only in the last few years that researchers are really challenging the
so-called hyperdynamic septic shock concept, which is much closer to the
human situation. Researchers like Sun et al. [34] or other groups [35] have
nicely shown that this option is really feasible.
Human Studies and Translations or Transpositions
from Animal Models
One of the greatest remaining problems with human studies (especially
the mechanistic studies) is the fact that the number of patients is very limited
resulting from the high cost of the technique. What is important in these
human studies is to understand that HVHF applied at a continuous dose for
96 h can be compared in some ways to activated protein C (APC) [36, 37].
Obviously, we cannot rely on the same level of evidence as we can for APC.
HVHF like APC can have a pleiotropic action on sepsis in the human setting.
Indeed, it can interfere with the proinflammatory phase and, by decreasing the
so-called proinflammatory phase, it can potentially reduce the unbound part of
cytokines and reducing the corresponding remote organ-associated damages
[38, 39].
HVHF: A Global View 377
The second point is that it can also alter and reduce some cardiovascular
compounds (in the blood compartment) that are responsible for the shock state
in the human situation. Indeed, endothelin-I can be removed and is held respon-
sible for early pulmonary hypertension in sepsis, whereas endocannabinoids are
responsible for the vasoplegia and myodepressant factor responsible for the car-
diodepression seen in sepsis [40–42]. All these factors can be easily removed by
HVHF.
Thirdly, HVHF can also alter the clotting system similar to the way it
decreases platelet-activating inhibitor (PAI) factor 1, thus eventually reducing
the level of diffuse intravascular coagulopathy [43]. It is well known that the
level of PAI-1 is correlated with sepsis with a increased APACHE II score and a
higher mortality rate [44].
Fourthly, it has been shown many times in animals that HVHF can reduce
the risk of immunoparalysis after sepsis and the subsequent risk of nosocomial
or secondary infection [29–32].
It has also been shown that HVHF can reduce the level of inflammatory
cell apoptosis occurring during sepsis as it can extract caspase-3 products with
a molecular weight of about 35,000 kDa, as well as some products of the
caspase-8 pathways which are heavily involved in the setting of inflammatory
cell apoptosis, especially in macrophages and neutrophils [45].
We know that clinical studies cannot reproduce the mean 100-ml/kg/h
exchange that has been realized in animal models (only 40 ml/kg/h in human
studies vs. 100 ml/kg/h) [24]. As a consequence, many anticipated effects seen
in animal models can never been reproduced in human settings related to the
use of inadequately low doses of HVHF.
What we do know is that there is huge variability between clinical trials
concerning the range of doses applied. It can vary from 1 to 15 in terms of dose
[24] when we aggregate all the recent studies.
If we decide to show that hemofiltration can be considered as a treatment in
the ICU, it must be adapted to the body weight and it must also to be adapted to
the severity of illness of the ICU patients. If we are dealing with nonseptic ARF,
perhaps a lower dose will be optimal. On the contrary, if we are dealing with sep-
tic ARF than we might need a higher dose, close to 50 or 70 ml/kg/h. From the
data presently available, we can say that in CRSS (or refractory hypodynamic
septic shock), the use of pulse HVHF running at about 100ml/kg/h during 4 con-
secutive hours (and then back at 35 ml/kg/h) is an important adjunctive treatment
that can dramatically increase the survival (table 2) of these severely ill patients
as compared with classical treatment [4, 5, 14, 15]. The monocentric study of
Oudemans-van Straaten et al. [3], which was realized in a cohort of mainly cardiac
surgery patients with oliguria at the time of inclusion, showed that the patient
subgroup with the best improvement (in term of observed versus expected
Honoré/Joannes-Boyau/Gressens 378
Table 2. HVHF in human studies: survival
Reference Patients Diagnosis Design CRRT Mb ⫹ S Ultra- Observed Survival Timing Weight UF
n severity technique m
2
filtrate effect kg volume
ml/h indexed
to body
ml/kg/h
Sander et al. 26 SIRS R 13 CVVH CA-SMS 1,000 No effect NA NA NA 13 (E)
[62], 1997 13 Co,st 0.6 75 (E)
LV-CVVH
LVHF
Matamis et al. 20 Sepsis P, LV-CVVH PS 1,500 MAP NA NA NA 20
[63], 1994 MODS UNC LVHF 7 SVR 75 (E)
Ronco et al. 425 USI-ARF R LV-CVVH PS –1,500 ND survival NA NA 20
[2], 2000 MV-CVVH 0.7–1.3 –2,700 with doses 75 (E) 35
LVHF –3,400 45
(sepsis)
John et al. 30 Sepsis R 20 CVVH Mb ⫹ SNA SVR NA NA NA NA
[64], 1968 10 IHD ⫽NA
Grootendorst 26 Sepsis P, HV-CVVH PA 4,500 MAP survival NA NA 60 (E)
et al. [65], 1996 MODS UNC HFHV S: NA CI? with no rela-
tion to hemo-
dynamics
Oudemans- 306 USI-ARF P Intermittent Cell tri 5,000 NA survival NA NA 65 (E)
van straaten cohort HV- S: 1.9 75 (E)
et al. [3], 1999 CVVH HFHV
HVHF: A Global View 379
Cole et al. 11 Septic R HV-CVVH CA-SMS 6,000 vaso NA NA NA 80 (E)
[66], 2001 shock ⫹ Cross HFHV S: 1.6 75
MODS
Honoré et al. 20 Refractory P, Short-term PS 9,000 MAP ⫹ survival 6.5 h (S) 66.2 (S) 140
[4], 2000 septic inter, HV-CVVH S: 1.6 SVR related to 14 h (NS) 82.5 (NS) 110
shock UNC VHVHF IC hemo-
LVSWI dynamics
ARF ⫽ Acute renal failure; C ⫽ controlled; CA-SMS ⫽ copolymer acrylonitrile and sodium methal sulfate (AN 69); Cell Tri ⫽ cellulose tri-
acetate; Cross ⫽ cross-over study; E ⫽ estimated; HVHF ⫽ high-volume hemofiltration; Inter ⫽ interventional; LVHF ⫽ low-volume hemofil-
tration; NA ⫽ not available; NS ⫽ non-survivors; P ⫽ prospective; PA ⫽ polyamide; PS ⫽ polysulfone; R ⫽ randomized; S ⫽ survivors;
UNC ⫽ uncontrolled; Vaso ⫽ vasopressor; VHVHF ⫽ very high-volume hemofiltration; VLVHF ⫽ very low-volume hemofiltration.
Honoré/Joannes-Boyau/Gressens 380
mortality) was the septic subgroup of patients in this specific study. The tech-
nique used was intermittent hemofiltration at a dose of 60 ml/kg/h. Since the
publication of several positive trials dealing with CRSS [4, 5, 14, 15], it is at pre-
sent almost accepted for hemofiltration in the ICU that, when dealing with
CRSS, a short-term procedure applying a very high dose should be the preferred
procedure, whereas for classical hyperdynamic septic shock with acute renal
injury, a continuous moderate high dose (during 96 h) might be the ideal choice
in order to achieve a dose of 50–70 ml/kg/h (for 96h). In this setting work on the
pleiotropic background is needed, mainly on immunoparalysis post-septic insult,
as shown especially by Yekebas [29–31] and Wang et al. [32].
Pulsed high volume has been shown to be still very effective in septic
shock as recently outlined in the literature [46, 47]. Those studies confirmed the
initial findings of Honoré et al. [4, 14, 15] and Joannes-Boyau et al. [5].
They also showed that the threshold dose needed to improve patients with
septic ARF was about 45 ml/kg/h as suggested by Ronco et al. [2] and Piccinni
et al. [47]. Recent studies have also shown that in CRSS, but this time hyper-
dynamic, HVHF might be a salvage therapy if a protocol-guided approach is
used [48].
Recommendations for Clinical Practice, Future Research, HVHF
Evolving Technology and Drug Adaptation during
Hemofiltration and Hemofiltration-Derived Therapies
Regarding the recommendations for clinical practice: CRSS, either hypo-
dynamic [4, 5, 14, 15] or hyperdynamic [48], could be seen as an indication
(level V evidence and grade E recommendation) for experienced clinicians in
the field of HVHF.
A patient with septic shock and ARF should receive a renal replacement
dose of at least 35 ml/kg/h (level II evidence and grade C recommendation).
Despite the numerous studies published and the ongoing IVOIRE study [6],
there are no sufficient hard data yet to support a higher extended dose in this
condition. This is true for other potential indications, such as septic shock with
or without renal failure or injury or even sepsis and SIRS (with or without fail-
ure or injury). In the case of SIRS induced by out-of-hospital cardiac arrests
[49], the existing data are too scarce to allow guidelines yet.
Regarding recommendations for future research concerning CRSS, it will
be very difficult in this case to apply a PRT and we should stick to the available
data or perform small bi-centric randomized studies.
Evaluating hyperdynamic septic shock patients, more numerous and larger
PRTs are needed to detect potential interference with APC. Indeed, this potential
HVHF: A Global View 381
interference deserves more attention as the molecular weight of APC
(55,000 kDa) creates the theoretical possibility that the membrane during
HVHF can adsorb the drug. Yet the risk is really minimal.
Nevertheless we have to think also about a possible synergy between HVHF
and APC as shown by experimental work completed by Heylen et al. [50].
Regarding sepsis and SIRS, the aim is still to reduce immunoparalysis shown by
various animal studies (level II evidence). For this type of patient, more mech-
anistic studies evaluating the potential risk of the technique are needed that must
be balanced with the anticipated beneficial effects. In many cases, a short-term
procedure will not be the ideal technique in these patients because to reduce
immunoparalysis a long technique, for instance 96 h, will be required.
What is the best environment for future research?
First a safer and more efficient technique must be developed in order to
increase the clinical operability, safer application and better clinical effectiveness.
Secondly a better understanding of the mechanisms of sepsis and SIRS is
needed in order to identify the molecular as well the proteinomic targets for
HVHF.
In order to meet the specific needs of control and restoration of immune
homeostasis in sepsis and SIRS, better designed new HVHF technology must
be ensured rather than modification of already existing technology.
Thirdly an appropriately designed (and with suitable power), trial of
HVHF is necessary to test the clinical effectiveness of this therapy in patients
with SIRS and/or sepsis. Lastly this trial should be conducted by indexing the
dose to body size and paying more attention to controlling (and analyzing) the
effects of time delay between the onset of SIRS or sepsis and HVHF initiation
in order to better define the duration of treatment as well the appropriate initia-
tion window.
Conclusions and Perspectives
The so-called big bang of hemofiltration has taught us that hemofiltration
is not a definitive world. As demonstrated by Ronco et al. [2] in 2000 hemofil-
tration is still in expansion. As we are aware of this phenomenon concerning
nonseptic ARF in ICU, we are now trying to define what should be the exact
dose for continuous hemofiltration in septic ARF.
HVHF can still be seen as a potent immunomodulatory treatment in sepsis
or SIRS. Since the mediator delivery hypothesis has been unravel [27], we
know that not only is the extraction important but also the amount of fluids
exchanged, and so the intake of fluids per se can increase the lymphatic flow
dramatically up to 20- to 40-fold. As a consequence, circulatory cytokines are
Honoré/Joannes-Boyau/Gressens 382
no longer valuable players [51], with the exception perhaps of very severe
CRSS [52], and now what is important is the crucial relationship between
immunological changes at the tissue level (where mediators do harm), hemody-
namic modifications and survival. A last point is obviously the possible syner-
gism in terms of therapy between APC and HVHF as both treatments have a lot
of similarities in terms of pleiotropic effects, and recent research work has
shown that synergy is possible between these two therapies. Nevertheless, many
more studies are needed to precisely define what the exact role (and the exact
impact on survival) of HVHF is, especially in hyperdynamic septic shock with-
out acute renal injury.
Higher doses of treatment may also be important whatever the choice of
the initial therapy, as it is able to influence the rate of secondary chronic dialy-
sis dependency or conversely the rate of renal recovery [53, 54] as shown also
by the work of Schiffl et al. [55] and Ronco et al. [56] (table 3).
In other words, the cost-effectiveness of continuous hemofiltration therapy
when compared to intermittent techniques may be changing very quickly with
time.
Recent publications have shown that we have to be very careful with regard
to studies as the conclusions may not be fully supported by the data and there-
fore may be in some ways misleading [57, 58].
The expansion and the odyssey of the hemofiltration universe continues.
Table 3. Renal recovery and dose of RRT
Reference Study design Patients Interventions Recovery Main Survival Level of
n groups outcome evidence
Jacka et al. Retrospective 116 Renal recovery CRRT ⫽ 87% Better Not IV
[53], 2005 IHD ⫽ 37.5% recovery in affected
CRRT
Manns et al. Retrospective 261 Cost of acute CRRT ⫽ less Better renal Not IV
[54], 2003 cohort renal failure dialysis recovery affected
dependency CRRT
Best III Data collection 1,260 Type of CRRT CRRT ⬎75% Better renal Not III
study
1
epidemiological IHD ⫽ 60% recovery affected
(2004) CRRT
Schiffl et al. Prospective 74 DHD DHD DHD ⫽ 9 days Faster DHD ⫽ 72% II
[55], 2002 randomized 72 AHD AHD renal recovery recovery of AHD ⫽ 54%
renal
function
1
Best III study: Unpublished data 2004.
HVHF: A Global View 383
References
1 Lemaitre GH: The Big Bang Theory (Eddington’s translation). London, Royal Astronomical
Society, Monthly Notices, March 1931.
2 Ronco C, Bellomo R, Homel P, Brendolan A, Dan M, Piccini P, et al: Effects of different doses in
continuous veno-venous haemofiltration. Lancet 2000;356:26–30.
3 Oudemans-van Straaten HM, Bosman RJ, Van der Spoel JL, Zanstra DF: Outcome of critically ill
patients treated with intermittent high-volume haemofiltration: a prospective cohort analysis.
Intensive Care Med 1999;25:814–821.
4 Honoré PM, Jamez J, Wauthier M, Lee P, Dugernier T, Pirenne B, Hanique G, Matson JR:
Prospective evaluation of short-term, high volume isovolemic hemofiltration on the hemodynamic
course and outcome in patients with intractable circulatory failure resulting from septic shock.
Crit Care Med 2000;28:3581–3587.
5 Joannes-Boyau O, Rapaport S, Bazin R, Fleureau C, Janvier G: Impact of high volume hemofil-
tration on hemodynamic disturbance and outcome during septic shock. ASAIO J 2004;50:
102–109.
6 Honoré PM, Joannes-Boyau O: The IVOIRE (hIgh VOlume in Intensive caRE) study: impact of
high volume haemofiltration in early septic shock with acute renal injury. A prospective multicen-
tric randomized study. Presented for the Stoutenbeek Award of the 18th Annual Congress of ESICM
Society, Berlin, 2004.
7 Bellomo R: http//clinical trials.gov/show/NCT221013
8 Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovitch M:
Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med.
2001;345:1368–1377.
9 Bellomo R, Kellum JA, Mehta R, Palevsky PM, Ronco C: The acute dialysis II: The Vicenza
Conference. Adv Ren Replace Ther 2002;9:290–293.
10 Lemaitre GH: La théorie cosmologique du ‘Big Bang’. Cosmologie dynamique et théorie de
l’atome primitif. Ann Soc Sci Bruxelles, 1927.
11 Bellomo R, Honoré PM, Matson JR, Ronco C, Winchester J: Extracorporeal blood treatment
(EBT) methods in SIRS/Sepsis. Int J Artif Organs 2005;28:450–458.
12 Honoré PM, Joannes-Boyau O: High volume hemofiltration (HVHF) in sepsis: a comprehensive
review of rationale, clinical applicability, potential indications and recommendations for future
research. Int J Artif Organs 2004;27:1077–1082.
13 Honoré PM, Matson JR: Extracorporeal removal for sepsis: acting at the tissue level – the begin-
ning of a new era for this treatment modality in septic shock. Crit Care Med 2004;32:896–897.
14 Honoré PM, Jamez J, Wauthier M, Dugernier T: Prospective evaluation of short-time high volume
isovolemic hemofiltration on the haemodynamic course and outcome of patients with refractory
septic shock. Crit Care Nephrol 1998;90:87–99.
15 Honoré PM, Jamez J, Wittebole X, Wauthier M: Influence of high volume haemofiltration on the
haemodynamic course and outcome of patients with refractory septic shock. Retrospective study
of 15 consecutives cases. Blood Purif 1997;15:135–136.
16 Klouche K, Cavadore P, Portales P, Clot J, Canaud B, Beraud JJ: Continuous veno-venous
hemofiltration improves hemodynamic in septic shock with acute renal failure without modifying
TNF-␣ and IL-6 plasma concentrations. J Nephrol 2002;15:150–157.
17 Matson JR, Zydney RL, Honoré PM: Blood filtration: new opportunities and the implications on
system biology. Crit Care Resusc 2004;6:209–218.
18 Grootendorst AF, van Bommel AF, van der Hoeven B, van Leengoed LA, van Oosta AL: High vol-
ume haemofiltration improves right ventricular function in endotoxin induced shock in the pig.
Intensive Care Med 1992;18:235–240.
19 Grootendorst AF, van Bommel AF, van der Hoeven B: High volume haemofiltration improves
haemodynamics in endotoxin induced shock in the pigs. J Crit Care 1992;7:67–75.
20 Grootendorst AF, van Bommel AF, van Leengoed LA, van Zande AR, Huygens HJ, Groenefeld ABJ:
Infusion of ultrafiltrate from endotoxemic pigs depressed myocardial performance in normal pigs.
J Crit Care 1993;8:61–69.
Honoré/Joannes-Boyau/Gressens 384
21 Grootendorst AF, van Bommel AF, van Leengoed LA, Nabuurs M, Boumans CS, Groenefeld ABJ:
High volume haemofiltration improves haemodynamics and survival of pigs exposed to gut
ischemia reperfusion. Shock 1994;2:72–78.
22 Rogiers P, Zhang H, Smail N, Pauwels D, Vincent JL: Continuous veno-venous haemofiltration
improves right cardiac performance by mechanisms other than tumor necrosis factor alpha attenu-
ation during endotoxin shock. Crit Care Med 1999;27:1848–1855.
23 Rogiers P, Zhang H, Pauwels D, Vincent JL: Comparison of polyacrylonitrile (AN69) and poly-
sulfone membrane doing haemofiltration in canine endotoxic shock. Crit Care Med 2003;31:
1219–1225.
24 Honoré PM, Zydney AL, Matson JR: High volume and high permeability haemofiltration in sep-
sis. The evidences and the key issues. Care Crit Ill 2003;3:69–76.
25 Matson JA, Lee PA: Evolving concepts of therapy for sepsis and septic shock and the use of
hyperpermeable membranes. Curr Opin Crit Care 2000;6:431–436.
26 Rogiers P: Hemofiltration treatment for sepsis: is it time for controlled trials? Kidney Int Suppl
1999;72:99–103.
27 Di Carlo JV, Alexander SR: Hemofiltration for cytokine-driven illness: the mediator delivery
hypothesis. Int J Artif Organs 2005;28:777–786.
28 Bone RC: Sir Isaac Newton, sepsis, SIRS and CARS. Crit Care Med 1996;24:1125–1128.
29 Yekebas EF, Treede H, Knoefel WT, Bloechle C, Fink E, Isbicki JR: Influence of zero balanced
haemofiltration on the course of severe experimental pancreatitis in pigs. Ann Surg 1999;229:
514–522.
30 Yekebas EF, Eisenberger CF, Ohnnesorge H, Saalmuller, Elsmer HA, Engelhardt M, et al:
Attenuation of sepsis-immunoparalysis with continuous veno-venous haemofiltration in experi-
mental porcine pancreatitis. Crit Care Med 2001;29:1423–1430.
31 Yekebas EF, Strate T, Zolmajd S, Eisenberger CE, Erbesdobler A, Saalmuller A, et al: Impact of
different modalities of continuous veno-venous hemofiltration on sepsis-induced alterations in
experimental pancreatitis. Kidney Int 2002;62:1806–1818.
32 Wang H, Zhang ZH, Yan XW, Li WQ, Ji DX, Quan ZF, Gong DH, Li N, Li JS: Amelioration of
haemodynamics an oxygen metabolism by continuous veno-venous hemofiltration in experimen-
tal pancreatitis. World J Gastroenterol 2005;11:127–131.
33 Piper RD, Cook DJ, Bone RC, Sibbald WJ: Introducing critical appraisal to studies of animal mod-
els investigating novel therapies in sepsis. Crit Care Med 1996;24:2059–2070.
34 Sun Q, Rogiers P, Pauwels D, Vincent JL: Comparison of continuous thermodilution and bolus
cardiac output measurements in septic shock. Intensive Care Med 2002;28:1276–1280.
35 Natanson C, Hoffman WD, Suffredini AF, Eichacker PQ, Danner RL: Selected treatment strategies
for septic shock based on proposed mechanisms of pathogenesis. Ann Intern Med 1994;120:
771–783.
36 Bernard GR, Vincent JL, Laterre PF, LaRosa SP, Dhainaut JF, Lopez-Rodriguez A, Steingrub JS,
Garber GE, Helterbrand JD, Ely EW, Fischer CJ Jr; Recombinant Human Protein C Worldwide
Evaluation in Severe Sepsis (PROWESS) Study Group: Efficacy and safety of recombinant
human activated protein C for severe sepsis. N Engl J Med 2001;344:699–709.
37 Bernard GR, Margolis BD, Shanies HM, Ely EW, Wheeler AP, Levy H, Wong K, Wright TJ; for
the US Investigators: Extended evaluation of recombinant human activated protein C United
States Trial (ENHANCE US): a single-arm, phase 3, multicenter study of drotrecogin alfa (acti-
vated) in severe sepsis. Chest 2004;125:2206–2216.
38 Tetta C, Bellomo R, Kellum J, Ricci Z, Pohlmeiere R, Passlick-Deetjen J, Ronco C: High volume
hemofiltration in critically ill patients: why, when and how?; in Ronco C, Bellomo R, Brendolan A
(eds): Sepsis, Kidney and Multiorgan Dysfunction. Contrib Nephrol. Basel, Karger, 2004, vol 144,
pp 362–375.
39 Brendolan A, D’Intini V, Ricci Z, Bonello M, Ratanarat R, Salvatori G, Bordoni V, De Cal M,
Andrikos E, Ronco C: Pulse high volume hemofiltration. Int J Artif Organs 2004;27:398–403.
40 Bellomo R, Kellum JA, Gandhi CR, Pinsky MR, Ondulik B: The effect of intensive plasma water
exchange by hemofiltration on hemodynamics and soluble mediators in canine endotoxemia. Am
J Respir Crit Care Med 2000;161:1429–1436.
HVHF: A Global View 385
41 Kohro S, Imaizumi H, Yamakage M, Masuda Y, Namiki A, Asai Y: Reductions in levels of bacter-
ial superantigens/cannabinoids by plasma exchange in a patient with severe toxic shock syndrome.
Anaesth Intensive Care 2004;32:588–591.
42 Court O, Kumar A, Parrillo JE, Kumar A: Clinical review: myocardial depression in sepsis and
septic shock. Crit Care 2002;6:500–508.
43 Garcia Fernandez N, Lavilla FJ, Rocha E, Purroy A: Haemostatic changes in systemic inflamma-
tory response syndrome during continuous renal replacement therapy. J Nephrol 2000;13:
282–289.
44 Zouaoui Boudjeltia KZ, Piagnerilli M, Brohée D, Guillaume M, Cauchie P, Vincent JL, Remacle
C, Bouchaert Y, Vanhaverbeck M: Relationship between CRP and hypofibrinolysis: is this a possi-
ble mechanism to explain the association between CRP and outcome in critically ill patients.
Thromb J 2004;2:7–12.
45 Bordoni V, Balgon I, Brendolan A, Crepaldi C, Gastoldon F, D’intini V, et al: Caspase 3 and 8 acti-
vation and cytokine removal with a novel cellulose tracetate super-permeable membrane in vitro
sepsis model. Int J Artif Organs 2003;26:897–905.
46 Ratanarat R, Brendolan A, Piccinni P, Dan M, Salvatory G, Ricci Z, Ronco C: Pulse high-volume
haemofiltration for treatment of severe sepsis: effects on hemodynamics and survival. Crit Care
2005;9:294–302.
47 Piccinni P, Dan M, Barbacini S, Carraro R, Lieta E, Marafon S, et al: Early isovolaemic haemofil-
tration in oliguric patients with septic shock. Intensive Care Med 2006;32:80–86.
48 Cornejo R, Downey P, Castro R, et al: High-volume hemofiltration as salvage therapy in severe
hyperdynamic septic shock. Intensive Care Med 2006;42:713–722.
49 Laurent I, Adrie C, Vinsonneau C, et al: High-volume hemofiltration after out-of-hospital cardiac
arrest: a prospective randomized study. J Am Coll Cardiol 2005;46:432–437.
50 Heylen A, Bervoets K, Smet M, Alexander JP, Rogiers P: Combination of Drotecogin ␣ and early
hemofiltration can improve outcome in severe septic patients with acute renal failure (abstract).
Spring Meeting of the Belgium Society of ICM, June 2005.
51 Tetta C, Bellomo R, D’Intini V, et al: Do circulating cytokines really matter in Sepsis? Kidney Int
2003;63:69–71.
52 Honoré PM, Ernst Y, Jamez J, Lemaire M, Pirenne B, Lebaupin C, et al: VHVHF is accompanied
by a significant drop in cytokines only in survivors during early refractory septic shock. Crit Care
2003;7:106–107.
53 Jacka MY, Ivancinova X, Gibney RT: Continuous renal replacement therapy improves renal recov-
ery from acute renal failure. Can J Anesth 2005;52:327–332.
54 Manns B, Doig CJ, Lee H, Dean S, Tonelli M, Johnson D, Donalson C: Cost of acute renal failure
requiring dialysis in the intensive care unit: clinical and resource implications of renal recovery.
Crit Care Med 2003;31:644–646.
55 Schiffl H, Lang SM, Fischer R: Daily hemodialysis and the outcome of acute renal failure. N Engl
J Med 2002;34:305–310.
56 Ronco C, Belomo R, Homel P, Brendolan A, Dan M, Piccinni P, La Greca G: Effects of different
doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective
randomised trial. EDTNA ERCA J 2002;2:7–12.
57 Vinsonneau C, Camus C, Combes A, et al: Continuous venovenous haemofiltration versus inter-
mittent haemodialysis for acute renal failure in patients with multiple-organ dysfunction syn-
drome: a multicentre randomised trial. Lancet 2006;368:379–345.
58 Honoré PM, Joannes-Boyau O: The ‘French Hemodiafe Trial’: this study is neither decisive nor
definitive in resolving the controversy on renal replacement therapy in ICU. Int J Artif Organs
2006;29:1190–1192.
59 Freeman BD, Yatsiv I, Natanson C, Solomon MA, Quezado ZM, Danner RL, Banks SM, Hoffman WD:
Continuous arteriovenous hemofiltration does not improve survival in a canine model of septic
shock. J Am Coll Surg 1995;180:286–292.
60 Rogiers P, et al: 21st ISICEM, Brussels, 2001.
61 Lee PA, Matson JR, Pryor RW, Hinshaw LB: Continuous arteriovenous hemofiltration therapy for
Staphylococcus aureus-induced septicemia in immature swine. Crit Care Med 1993;21:914–924.
Honoré/Joannes-Boyau/Gressens 386
62 Sander A, Armbruster W, Sander B, Daul AE, Lange R, Peters J: Hemofiltration increases IL-6
clearance in early systemic inflammatory response syndrome but does not alter IL-6 and TNF
alpha plasma concentrations. Intensive Care Med 1997;23:878–884.
63 Matamis D, Tsagourias M, Koletsos K, Riggos D, Mavromatidis K, Sombolos K, Bursztein S:
Influence of continuous haemofiltration-related hypothermia on haemodynamic variables and gas
exchange in septic patients. Intensive Care Med 1994;20:431–436.
64 John et al: Abstract, Intensive Care Med 1998;S75.
65 Grootendorst AF, Bouman CS, Hoeben KH: The role of continuous replacement therapy in sepsis
and multiorgan failure. Am J Kidney Dis 1996;28:50–57.
66 Cole L, Bellomo R, Journois D, Davenport P, Baldwin I, Tipping P: High-volume haemofiltration
in human septic shock. Intensive Care Med 2001;27:978–986.
Patrick M. Honoré, MD
ST-Pierre Para-University Hospital
Avenue Reine Fabiola, 9
BE–1340 Ottignies-Louvain-La-Neuve (Belgium)
Tel. ⫹32 10 437 346, Fax ⫹32 10 437 123, E-Mail Pa.honore@clinique-saint-pierre.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 387–395
Blood and Plasma Treatments: The
Rationale of High-Volume Hemofiltration
Patrick M. Honoré
a
, Olivier Joannes-Boyau
b
, Benjamin Gressens
a
a
St-Pierre Para-University Hospital, Ottignies-Louvain-La-Neuve, Belgium, and
b
Haut Leveque University Hospital, University of Bordeaux, Pessac, France
Abstract
Since the early 1990s, experts in the field have thought that a reduction in cytokines in
the blood compartment could, in theory, reduce mortality, but this is perhaps too naïve as the
pharmacodynamics and pharmacokinetics of cytokines throughout the body are not well
known and are probably much more complicated than previously thought. This ha now led to
three leading theories and concepts. Ronco and Bellomo conceived the peak concentration
hypothesis in which clinicians concentrate their efforts to remove mediators and cytokines
from the blood compartment at the proinflammatory phase of sepsis. By reducing the amount
of free cytokines, it is hoped that the level of remote organ (associated) damages can be dra-
matically decreased and, as a consequence, the overall death rate. In this regard, it is still not
known what will happen at the interstitial and tissue level with regard to mediators and
cytokines which are obviously the most important part in terms of consequences at the tissue
level. In this setting, techniques that can more rapidly and substantially remove great amounts
of cytokines or mediators are privileged. Among these, there is high-volume and very high-
volume hemofiltration and a number of hybrid therapies encompassing high-permeability
hemofiltration, super high-flux hemofiltration, hemo-adsorption or coupled filtration and
adsorption and other types of adsorption using physical or chemical forces rather than driving
forces as used normally in hemofiltration-derived techniques. The second concept is called the
threshold immunomodulation hypothesis, also called the Honoré concept. In this concept the
view of the system is much more dynamic. In experiments when removal is occurring on the
blood compartment side, the level on the interstitial side and the tissue side is also changing
and, because not only mediators but also pro-mediators are being removed, some pathways
have really stopped when enough pro-mediators have been removed by this technique. At this
point, the cascade is blocked and this point is called the threshold point. At this level, the cas-
cade is lost and no further harm can be done to the tissue of the organism. Obviously, it is dif-
ficult to know when this point has been reached once high-volume hemofiltration is applied.
But what is known, is that hemodynamics and survival can be improved in some patients as
shown by various studies using high-volume hemofiltration without any significant drop in
mediators inside the blood compartment itself. This effect is obtained without a dramatic fall
Honoré/Joannes-Boyau/Gressens 388
in the plasma cytokine level because the cytokine or mediator levels should fall at the tissue
level and not specifically at the blood compartment level. Nevertheless, the exact mechanism
by which high-volume hemofiltration increases the flow of mediators and cytokines between
the interstitial compartment and the blood compartment (and back to the blood side) is not
known. Before the end of 2005, it was found that this missing step is perhaps well explained
by the last theory and/or concept. The third theory and concept is called the mediator delivery
hypothesis and has also been called the Alexander concept. In this theory, the use of high-vol-
ume hemofiltration and especially high intakes of incoming fluids (3–5l/h) is able to increase
the lymphatic flow 20- to 40-fold, even more so for mediators and cytokine lymphatic flow
(drag). This has been demonstrated by several reports and is obviously extremely important.
Perhaps this can explain why some very recent studies using high-permeability hemofiltration
in sepsis have not been effective in improving hemodynamics and survival in septic acute ani-
mal models. In summary various brand new theories will be reviewed here in depth.
Copyright © 2007 S. Karger AG, Basel
Rationale Revisited Based upon Recent Human
and Animal Model Publications
Since the early 1990s, it has been advocated that a reduction in cytokines
in the blood compartment could, in theory, lead to a reduction in mortality [1, 2],
but this is too naïve as the pharmacodynamics and pharmacokinetics of cytokines
throughout the body are not well know and are probably much more compli-
cated than previously thought.
This has led to three leading theories and concepts. The Ronco and
Bellomo concept of the peak concentration hypothesis [3–5] (fig. 1), in which
efforts are made to remove mediators and cytokines from the blood compart-
ment at the proinflammatory phase of sepsis. It is hoped that by reducing the
amount of free cytokines the level of remote organ (associated) damages can be
decreased and as a consequence, the overall associated death rate.
In this regard, it is still not know what will happen at the interstitial and tis-
sue level with regard to mediators and cytokines, which are obviously the most
important part in terms of consequences at the tissue level.
In this setting, techniques that can more rapidly and more substantially
remove great amounts of cytokines or mediators are privileged. Among these,
a large place has been given to high-volume hemofiltration (HVHF) and very
HVHF and quite a lot of hybrid therapies encompassing high-permeability
hemofiltration (HPHF) [6], super high-flux hemofiltration (SHFHF) [7], hemo-
adsorption [8] or coupled filtration and adsorption (CPFA) [9] and any other
type of adsorption using physical or chemical forces rather than driving forces
as normally used in hemofiltration-derived techniques.
Also with regard to this issue, semantics is very crucial. Indeed, it can be
argued that the term ‘aDsorption’ is probably not the right term because blood
HVHF: New Insights in Rationale 389
is not flooding through a semi-permeable membrane and it is not the net effect
of convection forces plus oncotic forces that results in the passage of mediators
through this kind of device.
In this type of device, it is more appropriated to use the term ‘aBsorption’
as chemical and physical forces are really engaged in that setting. So, we should
be very careful about the use of appropriate terms when describing this kind of
technique [10]. Indeed, membrane separation only occurs with ‘aDsorption’.
The second concept is called the threshold immunomodulation hypothesis
(fig. 2) and has been called the Honoré concept [11, 12]. In this concept the
view of the system is much more dynamic. In some experiments when removal
occurs on the blood compartment side, the level on the interstitial side (and also
on the tissue side) also changes and, because not only mediators but also pro-
mediators are being removed, some pathways are really stopped when enough
pro-mediators have been removed by this technique. At this point, the cascade is
blocked and when reached, is called the threshold point. At this level, the cas-
cade is lost and no further harm can be done to the tissue of the organism.
Obviously, it is difficult to know when this point is reached once HVHF is
applied at the clinical level. But hemodynamics and survival can be improved in
some patients as shown by various studies using HVHF without any significant
ConceptVascular removal and remoted organ damages
Time elapsed
‘Spill over’
Normal volume
Hemofiltration
Septic
source
Body
reaction
Paracrine
effect
Endocrine
effect
Endogenous
Reducing
the
duration of
the
endocrine
effect of
SIRS
Metabolism
cirrhosis !!
High volume
Hemofiltration
Fig. 1. Rationale for extracorporeal removal. The peak concentration hypothesis.
Honoré/Joannes-Boyau/Gressens 390
drop in mediators in the blood compartment itself [13–15]. This effect is
obtained without any dramatic fall in plasma cytokine level because where the
cytokine or mediator level should fall is at the tissue level (where they do harm)
and not specifically at the blood compartment level.
Nevertheless, the exact mechanism by which HVHF can increase the flow
of mediators and cytokines between the interstitial compartment and the blood
compartment (and back to the blood side) is not known. Before the end of 2005,
it was not known if this missing step was perhaps well explained by the last the-
ory and/or concept.
The third theory and concept is called the mediator delivery hypothesis
(fig. 3) [16] and has been called the Alexander concept. In this theory, the use
of HVHF and especially high intakes of incoming fluids (3–5 l/h) is able to
increase the lymphatic flow by 20- to 40-fold, especially for mediator and
cytokine lymphatic flow (drag) [17–19], and is obviously extremely important.
Thus, the use of exchange fluid might be very important (and not only extrac-
tion) in order to increase the flow of lymphatic transport between the interstitial
tissue and the blood compartment.
We can now understand why high-flow hemofiltration is able to dramati-
cally increase the lymphatic transport from tissue and the interstitial space
including cytokines and mediators back to the blood compartment in order to be
potentially removed afterwards.
Concept
Spill over
Tissue
Self
regulation,
loop effects
and feed
back
mechanisms
Cytokines
Pro–
mediators
Inhibition of
production
‘Lost’ cascade
‘More dynamical
approach’
; Mediator ‘delivery’
and removal
Interstitial space
Vessel
At a certain level, the cascade is lost at the tissue level
ongoing re-equilibrium
supplies cytokin to
central circulation
for removal
HF
Fig. 2. Rationale for extracorporeal removal. The threshold modulation hypothesis.
HVHF: New Insights in Rationale 391
In comparison, HPHF is able to remove larger amounts of mediators and
cytokines in the blood compartment but is not able to increase lymphatic flow
and, as a consequence, is not able to remove some very crucial cytokines and
mediators at the interstitial and tissue level (where they do harm).
Therefore, this can explain why some very recent studies have shown that
using HPHF in sepsis is not effective for improving hemodynamics and survival
in septic acute animal models [20].
Therefore, clinicians should be aware of these new insights regarding the
rationale for extracorporeal removal in severe septic shock in order to choose
the best option with regard to adjunctive treatment for severe septic shock at the
bedside.
Future of HPHF and Increased Filter Porosity
With regard to mediators and despite the increased complexity of the ratio-
nale, one should think that increasing filter porosity could be a good option
[21]. Many mediators have a greater molecular weight and could be eliminated
by using more sophisticated techniques as HPHF, SHFHF and hemo-adsorption.
These techniques are able to remove more substantial amounts and perhaps
more mediators, but the question remains about removal in the right compartment.
Spill over
‘Flow drag’
‘Interstitial protein washout’
‘More dynamic approach’ → Lymphatic flow → Bringing more cytokines
into the blood for hemofiltration, other organs metabolism ...
Lymph/serum ratio of albumin0.48/IL-89.8 but IL-640!
Concept
HF
Liver
Lymphatics
(G by 15 fold!)
Interstitial space
Reinfusion
fluid
Tissue
Self
regulation,
loop effects
and feed
back
mechanisms
Cytokines
Pro–
mediators
Vessel
Fig. 3. Rationale for extracorporeal removal. The mediator delivery hypothesis.
Honoré/Joannes-Boyau/Gressens 392
There is also the risk of losing many important nutrients, hormones, drugs,
especially antibiotics, and many unknown metabolites.
Hybrid techniques have been attempted, taking advantage simultaneously
of different techniques without having to support their drawbacks. CPFA and
cascade hemofiltration (CCHF) [22] are able to retrieve large amounts and
large molecules without the risk of losing important nutrients because part of
the so-called purified blood is going back to the patient.
With regard to filter porosity, if we stick to hybrid techniques such as
CCHF and CPFA and if a significant part of this so-called purified blood is
going back to the patients, there would be no real theoretical limits as nothing
important would be lost and only target molecules would be adsorbed.
Along these lines, it can be seen that a complete neglected domain is
HVHF, and related techniques are searching for molecules below 45 kDa and
plasma filtration looks for molecules around 900 kDa [23]. As a consequence,
molecules between these two limits are really neglected and clinicians (and as
well investigators) should pay much more attention to them. If we want to guar-
antee that all the purified blood will go back to the patient, the limits must
be widened eluding to the potential risk of losing many important blood
components.
Clinical Implications for the Intensivist in 2006
Regarding the use of extracorporeal treatment in sepsis as an adjunctive
therapy in intensive care, it can be said that the 35-ml/kg/h rule should be
applied widely in intensive care units (ICUs) as recent unpublished surveys
have shown that less than 20% of the units (at least in continental Europe) are
using them.
A recent position paper published by an Acute Dialysis Quality Initiative
(ADQI) group has stressed that HVHF could be used in catecholamine-resistant
septic shock (level V evidence and grade E recommendation) [24]. The same
ADQI position paper also widely supported the extended use of the 35-ml/kg/h
rule with level II evidence and grade C recommendation [24].
In classical hyperdynamic septic shock, especially ICU acute septic renal
failure or ICU acute septic renal injury (according to the RIFLE classification),
we are eagerly waiting the results of several outcome dose studies in the ICU.
Amongst the ongoing studies regarding the appropriate dose of hemofiltra-
tion in critically ill septic acute renal failure, the hIgh VOlume in Intensive
caRE (IVOIRE) study must be mentioned [25]. This ongoing study will poten-
tially give us very important insights for the future regarding the exact dose to
use in subgroups of septic patients with acute renal injury. The IVOIRE study
HVHF: New Insights in Rationale 393
will include more than 480 patients with septic shock plus acute renal injury
defined by the RIFLE classification in the ICU. Allocation into the two groups
will be determined by computerized randomization. One group will receive 35
versus 70 ml/kg/h in the other group. The exact and complex methodology will
be available soon [26].
This study will try to demonstrate that a higher dose (70 ml/kg/h) will fur-
ther improve the survival rate of septic acute renal failure patients in the ICU.
Ronco et al. [5] have already alluded to this with the 45-ml/kg/h subgroup,
whereas the septic subpopulation had better survival although the nonseptic one
did not improve further.
Conclusions and Perspectives
Clinicians should be aware of the new insights regarding the rationale for
extracorporeal removal in severe septic shock in order to choose the best option
for adjunctive treatment at the clinical level. The exchange volume is not only
important for removal of mediators but also for displacement of mediators
throughout the body [27–30]. Membrane porosity or system complexity can
never replace systems that just use high-volume exchange rates with simple
membrane separation technology.
As a final note, the world of hemofiltration and associated hybrid therapies
is still evolving rapidly. Not only the investigator but also the clinician should
be aware of the recent advances as several ongoing dose outcome studies may
profoundly change daily practice. The expansion and odyssey of the hemofil-
tration universe continues.
References
1 Damas P, Canivet JL, de Groote D, Vrindts Y, Albert A, Franchimont P, et al: Sepsis and serum
cytokines concentrations. Crit Care Med 1997;25:405–412.
2 Casey LC, Balk RA, Bone RC: Plasma cytokine and endotoxin levels correlate with survival in
patients with the sepsis syndrome. Ann Intern 1993;119:771–778.
3 Ronco C, Tetta C, Mariano F, Wratten ML, Bonello M, Bellomo R: Interpreting the mechanism of
continuous renal replacement therapy in sepsis. The peak concentration hypothesis. Artif Organs
2003;27:792–801.
4 Ronco C, Bellomo R: Acute renal failure and multiple organ dysfunction in the ICU: from renal
replacement therapy (RRT) to multiple organ support therapy (MOST). Int J Artif Organs
2002;25:733–747.
5 Ronco C, Ricci Z, Bellomo R: Importance of increased ultrafiltration volume and impact on
mortality: sepsis and cytokine story and the role for CVVH. EDTRA ERCA J 2002;2:13–18.
6 Lee PA, Weger G, Pryor RW, Matson JR: Effects of filter pore size on efficacy of continuous
arterio-venous haemofiltration therapy for Staphylococcus aureus-induced septicaemia in immature
swine. Crit Care Med 1998;26:730–737.
Honoré/Joannes-Boyau/Gressens 394
7 Lee WC, Uchino S, Fealy N, Baldwin I, Panagiotopoulos S, Goehl H, Morgera S, Neumayer HH,
Bellomo R: Super high flux hemodialysis at high dialysate flows: an ex vivo assessment. Int J Artif
Organs 2004;27:24–28.
8 Honoré PM, Matson JR: Hemofiltration, adsorption, sieving and the challenge of sepsis therapy
design. Crit Care 2002;6:394–396.
9 Bellomo R, Tetta C, Ronco C: Coupled plasma filtration adsorption. Intensive Care Med 2003;29:
1222–1228.
10 Bellomo R, Honoré PM, Matson JR, Ronco C, Winchester J: Extracorporeal blood treatment
(EBT) methods in SIRS/Sepsis. Consensus statement. ADQI III Conference. Electronic
Supplement Material.www.adqi.net (2005).
11 Honoré PM, Joannes-Boyau O: High volume hemofiltration (HVHF) in sepsis: a comprehensive
review of rationale, clinical applicability, potential indications and recommendations for future
research. Int J Artif Organs 2004;27:1077–1082.
12 Honoré PM, Matson JR: Extracorporeal removal for sepsis: acting at the tissue level – the begin-
ning of a new era for this treatment modality in septic shock. Crit Care Med 2004;32:896–897.
13 Honoré PM, Jamez J, Wauthier M, Dugernier T: Prospective evaluation of short-time high volume
isovolemic hemofiltration on the hemodynamic course and outcome of patients with refractory
septic shock. Crit Care Nephrol 1998;90:87–99.
14 Honoré PM, Jamez J, Wittebole X, Wauthier M: Influence of high volume haemofiltration on the
haemodynamic course and outcome of patients with refractory septic shock. Retrospective study
of 15 consecutives cases. Blood Purif 1997;15:135–136.
15 Klouche K, Cavadore P, Portales P, Clot J, Canaud B, Beraud JJ: Continuous veno-venous
hemofiltration improves hemodynamic in septic shock with acute renal failure without modifying
TNF- and IL-6 plasma concentrations. J Nephrol 2002;15:150–157.
16 Di Carlo JV, Alexander SR: Hemofiltration for cytokine-driven illness: the mediator delivery
hypothesis. Int J Artif Organs 2005;28:777–786.
17 Olszewski WL: The lymphatic system in body homeostasis: physiological condition lymph fat
rest. Lymphat Res Biol 2003;1:11–21.
18 Onarherim H, Missavage E, Gunther RA, Kramer GC, Reed RK, Laurent TC: Marked increase of
plasma hyaluronan after major thermal injury and infusion therapy J Surg Res 1991;50:259–265.
19 Wasserman K, Mayerson HS: Dynamics of lymph and plasma protein and exchange. Cardiologia
1952;21:296–307.
20 Rogiers P: High volume haemofiltration: high volume, high permeability: which target (abstract).
4th ERTIC Meeting, Nice, 2005.
21 Honoré PM, Zydney AL, Matson JR: High volume and high permeability haemofiltration in sep-
sis. The evidence and the key issues. Care Crit Ill 2003;3:69–76.
22 Valbonesi M, Carlier P, Icone A, Accorsi P, Borberg H, Schreiner T, et al: Cascade filtration: a new
filter for secondary filtration – a multicentric study. Int J Artif Organs 2004;27:513–515.
23 Matson JR, Zydney RL, Honoré PM: Blood filtration: new opportunities and the implications on
system biology. Crit Care Resusc 2004;6:209–218.
24 Bellomo R, Honoré PM, Matson JR, Ronco C, Winchester J: Extracorporeal blood treatment
(EBT) methods in SIRS/sepsis. Int J Artif Organs 2005;28:450–458.
25 Honoré PM, Joannes-Boyau O: The IVOIRE study: impact of high volume haemofiltration in
early septic shock with acute renal injury: a prospective multicentric randomized study. Presented
for the Stoutenbeek Award, 18th Ann Congr ESICM Soc, Berlin, 2004.
26 Joannes-Boyau O, Honoré PM, Boer W, et al: The IVOIRE Study. Description of the methodology
and the design used. Submitted to Crit Care 2007.
27 Joannes-Boyau O, Honoré PM, Boer W: Hemofiltration: the case for removal of sepsis mediators
from where they do harm. Crit Care Med 2006;34:2244–2246.
28 Honoré PM, Jamez J, Wauthier M, Lee PA, Dugernier Th, Pirenne B, Hanique G, Matson JR:
Prospective evaluation of short-term high-volume isovolemic hemofiltration on the hemodynamic
course and outcome in patients with intractable circulatory failure resulting from septic shock.
Crit Care Med 2000;28:3581–3587.
HVHF: New Insights in Rationale 395
29 Honoré PM, Joannes-Boyau O, Meurson L , Boer W, Piette V, Galloy AC, Janvier G: The big bang
of hemofiltration: the beginning of a new era in the third millennium for extra-corporeal blood
purification! Int J Artif Organs 2006;29:649–659.
30 Honoré PM, Joannes-Boyau O: The ‘French Hemodiafe Trial’: this study is neither decisive nor
definitive in resolving the controversy on renal replacement therapy in ICU. Int J Artif Organs
2006;29:1190–1192.
Patrick M. Honoré, MD
St-Pierre Para-University Hospital
Avenue Reine Fabiola, 9
BE–1340 Ottignies-Louvain-La-Neuve (Belgium)
Tel. 32 10 437 346, Fax 32 10 437 123, E-Mail Pa.honore@clinique-saint-pierre.be
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 396–404
Liver Support Systems
Antonio Santoro
a
, Elena Mancini
a
, Emiliana Ferramosca
a
,
Stefano Faenza
b
a
Unità Operativa di Nefrologia, Dialisi e Ipertensione, Dipartimento di Medicina
Interna, Scienze Nefrologiche ed Invecchiamento,
b
Dipartimento di Discipline
Chirurgiche, Rianimatorie e dei Trapianti, Policlinico Sant’Orsola-Malpighi,
Bologna, Italia
Abstract
Liver insufficiency is a dramatic syndrome with multiple organ involvement. A multi-
plicity of toxic substances (hydrophilic like ammonia and lipophilic like bilirubin or bile
acids or mercaptans) are released into the systemic circulation, thus altering many enzy-
matic cellular processes. Patients frequently die while on the transplantation waiting list
because of organ scarcity. Systems supporting liver function may be useful to avoid further
complications due to the typical toxic state, ‘bridging’ the patients to the transplantation, or,
in the event of an acute decompensation of a chronic liver disease, sustain liver function
long enough to permit the organ’s regeneration and functional recovery. An ideal liver sup-
port system should substitute the main functions of the liver (detoxification, synthesis and
regulation). Extracorporeal systems now available may be totally artificial or bioartificial.
While the first are only able to perform detoxification, the second may add the functions of
synthesis (plasma proteins, coagulation factors) and regulation (neurotransmitters).
Bioartificial liver working with isolated hepatocytes and a synthetic membrane in an extra-
corporeal system are however still far from being ready for clinical use. At present, liver
insufficiency may be treated with an extracorporeal support technology aimed either at
detoxification alone or at a real purification. Charcoal hemoperfusion or exchange/absorp-
tion resins may be used for blood detoxification. Blood or plasma exchange, from a theoret-
ical point of view, could be suitable for a polyvalent intoxication, such as liver failure;
however, the multicompartmental distribution of some solutes largely endangers the effi-
cacy of these procedures. Selective plasmapheresis techniques are now available for some
solutes (e.g. styrene for bilirubin) and may progressively reduce the plasma levels and pre-
sumably the deposits of the solute. Novel treatments introduced to improve detoxification,
mainly of the protein-bound substances, are the molecular adsorbent recirculation system
(MARS) and Prometheus
TM
systems. MARS performs an albumin dialysis, where albumin
is the exogenous carrier for the toxic substances, and different experiences have proved its
efficacy mainly in the treatment of hepatic encephalopathy, while data on survival are still
limited to small case series. With Prometheus, the most recent system developed for a wide
Liver Support Systems 397
detoxification, albumin-bound toxins are directly removed in two separate cartridges with
different solute affinity, without the need for exogenous albumin; plasmadsorption is then
coupled with a real dialysis process. After promising initial results, the efficacy of
Prometheus in the patients’ hard endpoints will be evaluated in a large international trial. On
the whole, liver support systems may offer, in many cases, a survival benefit. Stem cells are
however, even in this filed, the real great hope for the future of patients with end-stage liver
disease.
Copyright © 2007 S. Karger AG, Basel
In presence of liver insufficiency a multiplicity of toxic substances, both
lipophilic and hydrophilic, are released into the systemic circulation, thus altering
many enzymatic cellular processes. The lipophilic substances interfere with struc-
tural cellular processes, such as the reconstruction of cellular membranes, while
the hydrophilic ones alter and block functional processes both enzymatic and
nonenzymatic in nature. Still, the different forms of liver failure (acute, subacute,
acute-on-chronic, chronic) are characterized by a high patient mortality [1].
The only proven successfully therapeutic solution for such an extremely
critical situation is orthotopic liver transplantation. Unfortunately, organ
scarcity cannot fulfill all the needs for transplantation arising during acute and
chronic liver diseases. Furthermore, as in the case of decompensation of a
chronic liver disease, it is necessary to support the liver function long enough to
permit the regeneration and functional recovery of the organ.
An ideal liver support system has to support or substitute the main
functions of the liver, providing detoxification, synthesis and regulation. For
detoxification (e.g. removal of bilirubin, bile acids, and toxins), artificial
detoxification systems have been shown to reach varying degrees of effi-
ciency. The complex tasks of regulation (e.g. central nervous system transmit-
ter precursors) and synthesis (e.g. coagulation factors) remain to be addressed
by the use of live hepatocytes in bioartificial livers. Yet in this field, the current
research is still quite far from accomplishing the ideal bioartificial liver (BAL)
for clinical use, even though several studies have been performed on limited
patient series.
In this review, we describe the most promising extracorporeal artificial
systems for liver support, while we do not take into consideration the biological
system such as BAL and even more the stem cells that are the great hope for the
future of patients with liver disease (table 1). The initial concept of BAL was to
simply fill a hemodialyzer with cells. BAL basically integrates isolated hepato-
cytes which come into contact with patient’s blood or plasma with the interpo-
sition of a synthetic membrane. Excellent reviews on the history of BAL, and
the various systems so far used in experimental models and clinical trials are
available in the literature [2–4].
Santoro/Mancini/Ferramosca/Faenza 398
A recent consideration focuses on stem cells beyond primary hepatocytes
and cell lines. Successful generation of hepatocyte-like cells was obtained from
human embryonic stem cells [2]. Despite the richness of publications, espe-
cially in the USA, related to the huge potential of embryonic stem cells to gen-
erate different types of tissues, there is a scarcity of research focused on liver.
Recently, the European Parliament banned the use of embryonic stem cells for
research purposes in Europe [5].
Artificial Liver Support Systems
Clinical indications for an extracorporeal liver support have been steadily
evolving since the beginning of orthotopic liver transplantation in an increasing
number of centers. Table 2 provides a schematic summary of the most frequent
clinical situations which may benefit from the use of extracoporeal systems for
liver support.
Table 1. Systems for extracorporeal liver support
Blood detoxification
•
Sorbent-based therapy (charcoal, exchange or adsorption resins)
Blood purification techniques
•
Plasma exchange
•
Selective plasma adsorption (cartridge of styrene)
•
High-performance hemodiafiltration/hemofiltration
Provision of whole liver function
•
BAL
Porcine hepatocytes
Human hepatocytes
•
Extracorporeal hepatic perfusion
Table 2. Therapeutic beneficial effects of artificial support systems
•
Bridge to recovery from liver failure before liver transplantation or during
the course of severe acute or acute-on-chronic liver diseases
•
Physiological support of liver function after delayed graft function
•
Physiological support after liver surgery (for trauma or neoplasia)
Liver Support Systems 399
Conventional Systems
In 1956 Kiley et al. [3] reported the use of hemodialysis in the treatment of
5 patients with hepatic encephalopathy: in 4 patients the state of consciousness
improved, but with no improvement in long-term survival.
During the last 20 years, there has been a movement from hemodialysis to
hemodiafiltration due to the availability of highly permeable membranes capable
of removing molecules with a molecular weight of over 15,000 Da. Currently,
hemofiltration has proven particularly useful in resolving the intracranial/intra-
cerebral hypertension which accompanies hepatic encephalopathy [6].
Hemoperfusion on charcoal was introduced for the therapy of liver failure
by Yatzidis [4] in 1964, and since then has received a broad consensus, even if
the rather few controlled studies have failed to demonstrate a significant clin-
ical advantage as compared with conventional medical therapies. Combined
hemoperfusion, charcoal plus resins [7], could theoretically offer advantages
even if no controlled prospective studies have been performed yet.
Total blood exchange is definitely, at least on a theoretical level, a treat-
ment suitable for a polyvalent intoxication, such as hepatic failure. Currently,
since many toxic substances generated during liver failure, bilirubin first and
foremost, are distributed into several pools, such as the interstitial and the cel-
lular compartments beyond plasma, one treatment cannot provide a complete
therapeutic solution.
Novel Treatments
Bilirubin and the biliary salts are toxic in many respects as they can induce a
series of chain-linked toxic effects leading to multiorgan dysfunctions (table 3).
Treatments that reduce bilirubin levels and thus the toxicity of hyperbilirubinemia
may be relevant. Modern technologies now exist and include selective cascade
Table 3. Biological actions of bilirubin and biliary salts
•
Inhibit hydrolytic enzymes, dehydrogenases, and enzymes involved in the
electron transport
•
Act as an uncoupler of oxidative phosphorylation
•
Decrease the Na-K-adenosine triphosphatase activity
•
Inhibit the tyrosine uptake
•
Reduce the gluthatione-8-transferase activity of ligandine
•
Inhibit protein kinase C
Santoro/Mancini/Ferramosca/Faenza 400
plasmapheresis, the molecular adsorbent recirculating system (MARS) and the
Prometheus
TM
system.
Although plasma exchange was introduced in the 1960s after some positive
results [8], it has never attained popularity due to the risk of viral infections, high
costs and depression of specific immunity. Recently, the use of high exchange
rate plasmapheresis has provided promising results, above all in cases of drug-
induced liver intoxication. In order to overcome these disadvantages some varia-
tions have been suggested, as in the cascade plasmapheresis system [9].
Whatever the treatment, bilirubin reduction occurs according to an expo-
nential pattern with a rapid decline during the first 30 min followed by a slower
decline at different velocities depending on the initial bilirubin value and with
patient-to-patient pattern variability (fig. 1). There is a rebound phase in biliru-
bin plasmatic values following each treatment session, appearing during the
first 30 min after treatment end. Rebound velocities in the immediate posttreat-
ment phase are extremely high, while the rise of bilirubin values during the
interval between the two treatments is much slower. The initial rebound surge is
the expression of bilirubin release from the tissue-interstitial and cellular pools
towards the vascular one in order to reequilibrate the two sectors. As mentioned
for bilirubin reduction, also the rebound, at the same time intervals after treat-
ment end, has a patient-to-patient pattern variability. This variability is probably
correlated with the pattern of bilirubin pool and its dynamics of accumulation.
It is known that the cellular pool is much slower than the interstitial one.
Furthermore, the transfer constraints of the extravasal pool, with a high case to
case variability, also depend on the accumulation and production pattern.
30
min
60
min
120
min
150min
12h
Time (min)
Time posttreatment
30
32
34
36
38
40
42
44
46
48
50
0
ab
30 60 90 120 150 180
30
32
34
36
38
40
42
44
46
48
50
Total bilirubin (mg/dl)
Fig. 1. Bilirubin behavior (cascade plasmapheresis). Progressive reduction throughout
the treatment (a) and posttreatment rebound (b).
Liver Support Systems 401
MARS is a purification system aimed at removing albumin-bound toxic
molecules. The principle is that dialysis membranes are impermeable to albu-
min but able to clear toxic substances bound to albumin, such as endogenous
benzodiazepines, mercaptanes, and biliary acids when an albumin-rich dialysate
is used. The dialysate also contains electrolytes and bicarbonate as buffer, and is
regenerated by passage through an anionic-exchange resin charcoal absorption
and sequential hemodialysis.
Worldwide, various nonrandomized trials and clinical applications in more
than twenty clinical indications [10] have demonstrated some efficacy of the
system in the treatment of hepatic encephalopathy. A recently published
prospective, controlled and randomized trial, but on small numbers (8 patients
in the study and 5 in the group treated with conventional standard medical ther-
apy alone), demonstrated a greater survival for the study group (25 days as
compared with 4 in the control group), an enhanced prothrombin activity and
lower levels of bilirubin [11]. Other randomized controlled trials are, however,
needed in order to prove the efficacy of this liver support in increasing survival
over 3 months of the liver failure patients [12].
In 2001 the first Prometheus treatment was performed in a patient. This
fractionated plasma separation, adsorption and dialysis system uses dialysis
and adsorption for detoxification of water-soluble and albumin-bound toxins
like other extracorporeal liver support systems, too. However, in contrast to pre-
cursor systems like MARS and fractionated plasma separation and adsorption
system (FPSA) [13, 14] Prometheus [15] separates both procedures (fig. 2).
First of all, blood flows through a special albumin-permeable filter
AlbuFlow
TM
with a membrane cutoff close to 300,000 Da. Here the patient’s
own albumin is separated from the blood, while cells and higher molecular
weight solutes like fibrinogen are retained in the blood. The albumin is then
perfused through the prometh
TM
01 and prometh
TM
02 adsorber, whereby the
bound toxins are captured by direct contact with the high-affinity adsorbing
material. The purified albumin then reenters the bloodstream. Afterwards the
blood passes through a high-flux dialyzer.
Due to this two-stage procedure the efficiency is considerably increased.
Albumin-bound toxins are directly transported, by convection, to the site where
removal takes place. In this way, no dissociation from albumin and diffusion is
necessary, known to be the limiting factors in removal kinetics when high-flux
membranes are used for mass transfer from blood. In addition, there is no need
for exogenous albumin.
Afterwards the blood undergoes high-flux hemodialysis, one of the most
effective treatments for removal of water-soluble toxins. The use of Fresenius
Polysulfone
TM
membranes and ultrapure dialysis fluid prepared with the 4008H
dialysis machine ensures maximum biocompatibility. Maintenance and monitoring
Santoro/Mancini/Ferramosca/Faenza 402
of the extracorporeal circuit is performed by the Prometheus unit. It is a com-
prehensive ‘all-in-one’ extracorporeal liver support system based on the
Fresenius 4008H dialysis unit with FPSA and hemodialysis integrated in one
unit. Actually the Prometheus unit may not only be used for liver support ther-
apy but also for conventional dialysis alone [16].
Clinical safety and adsorber efficiency as expressed by laboratory findings
were confirmed in 11 patients treated with Prometheus on 2 consecutive days
[14]. Total bilirubin, bile acids and ammonia were reduced during 5-hour
treatments by 21, 43 and 40%, respectively. This was confirmed in a study with
14 patients with acute or acute-on-chronic liver failure where citrate was used
as anticoagulant [17]. Here bilirubin decreased by 39.9%, total bile acids by
28.1% and ammonia by 39.0% during treatments of 6 h. Successful treatment of
fulminant hepatic failure was also reported [18].
We used [19] the Prometheus system in 12 patients with acute or acute-
on-chronic liver insufficiency: 8 cirrhosis, 1 posttransplant dysfunction, and
Adsorber
02
Blood
pump
AlbuFlow
Recirculation
pump
Heparin
Adsorber
01
Filtrate
recirculation
circuit
High-flux
dialyzer
Extracorporeal
blood circuit
Dialysate
FPSA
Fig. 2. Scheme of Prometheus system circuit. At first blood flows through a special
albumin-permeable filter AlbuFlow with a membrane cutoff close to 300,000 Da. Here the
patient’s own albumin is separated from the blood, while cells and higher molecular weight
solutes like fibrinogen are retained in the blood. The albumin is then perfused through the
prometh 01 and prometh 02 adsorber. The prometh 01 adsorber cartridge contains a highly
porous neutral resin for the removal of albumin-bound toxins such as bile acids, aromatic
amino acids, and phenolic substances while the prometh 02 adsorber cartridge contains a
high-performance anion exchange resin. The purified albumin then reenters the bloodstream.
Afterwards the blood passes through a high-flux dialyzer.
Liver Support Systems 403
3 secondary liver insult (2 cardiogenic shocks and 1 rhabdomyolysis). All
the patients were severely hyperbilirubinemic, hypercholemic and hyper-
ammoniemic. Twenty-eight sessions were performed, 2.5/patient, each lasting
340 ⫾ 40 min.
Mean total bilirubin decreased from 33.6 ⫾ 20 to 22.2 ⫾ 13.6 mg/dl
(p ⬍ 0.001); the reduction ratios for cholic acid and ammonia were 48.6 and
51.6%, respectively. The pre- to postsession reduction of urea was 57.6 ⫾ 9.5%
and of creatinine 42.7 ⫾ 10%. A significant reduction was observed in the cir-
culating level of the soluble receptor for interleukin 2 (before 2,687.2 ⫾
1,434.7 IU/ml, after 1,977.1 ⫾ 602 IU/ml, p ⬍ 0.001) and in IL-6 (before
56.1 ⫾ 11.1 pg/ml, after 35.9 ⫾ 10.3 pg/ml, p ⫽ 0.05). Intratreatment hemo-
dynamics was stable.
Two patients received liver transplant. Secondary liver insult was com-
pletely overcome in all 3 patients. The overall survival at 30 days was 41.6%
(5/12 patients). Prometheus, based on FPSA, produced high clearance for protein-
bound and water-soluble markers, which resulted in a high treatment dose. The
efficacy in the patients’ outcome of this highly efficient system is expected and
we are taking part in an European multicenter trial in a larger population with
acute-on-chronic liver insufficiency.
Conclusions
For patients with acute-on-chronic liver failure and perhaps also for
patients with acute liver insufficiency the artificial liver support devices may
offer a survival benefit. However, the evidence from both the artificial systems
and the bioartificial ones is anything but conclusive and, as is always the case
with pioneering methods, there is a definite need for randomized clinical trials
that can shed light on the real survival advantages over standard medical ther-
apy for liver failure.
References
1 Santoro A, Mancini E: The kidney in hepatorenal syndrome. I Part I. Int J Artif Organs 2004;27:
95–103.
2 Rambhatla L, Chiu CP, Kundu P, Peng Y, Carpenter MK: Generation of hepatocyte-like cells from
human embryonic stem cells. Cell Transplant 2003;12:1–11.
3 Kiley J, Welch HF, Pender JC: Removal of blood ammonia by hemodialysis. Proc Soc Exp Biol
Med 1956;91:489–490.
4 Yatzidis H: A convenient hemoperfusion microapparatus over charcoal for the treatment of
endogenous and exogenous intoxification: its use as an effective artificial kidney. Proc Eur Dial
Transplant Assoc 1964;1:83–87.
5 Commission of the European Community, March 2002.
Santoro/Mancini/Ferramosca/Faenza 404
6 Davenport A, Will EJ, Losowsky MS: Rebound surges of intracranial pressure as a consequence of
forced ultrafiltration used to control intracranial pressure in patients with severe hepatorenal fail-
ure. Am J Kidney Dis 1989;14:516–519.
7 Tetta C, Cavaillon JM, Camussi G, et al: Continuous plasma filtration coupled with sorbents.
Kidney Int 1998;66(suppl):S186–S189.
8 Trey MB, Bums DG, Saunders SJ: Treatment of hepatic coma by exchange blood transfusion.
N Engl J Med 1966;42:394–398.
9 Santoro A, Mancini E, Buttiglieri S, Krause A, Yakubovich M, Tetta C: Extracorporeal support of
liver function. Part II. Int J Artif Organs 2004;27:176–185.
10 Stange J, Hassanein TI, Mehta R, Mitzner SR, Bartlett RH: The molecular adsorbents recycling
system as a live support system based on albumin dialysis: a summary of preclinical investiga-
tions, prospective randomised, controlled clinical trial and clinical experience from 19 centers.
Artif Organs 2002;26:103.
11 Stange J, Mitzer S, Klammt S, et al: Liver support by extracorporeal blood purification: a clinical
observation. Liver Transpl 2000;6:603–613.
12 Ichai P, Samuel D: Extracorporeal liver support with MARS in liver failure: has it a role in the
treatment of severe alcoholic hepatitis? J Hepatol 2003;38:104–106.
13 Strobl W, Vogt G, Mitteregger R, et al: Das ‘Fractionated Plasma Separation and Adsorption System’
(FPSA), ein neues membran-adsorptionsgestütztes adjunktives extrakorporales Blutreinigungssystem
für das Leberversagen. Biomed Technik Ergänzungsband 1998;43/1: 168–169.
14 Falkenhagen D, Strobl W, Vogt G, et al: Fractionated plasma separation and adsorption system: a
novel system for blood purification to remove albumin bound substances. Artif Organs 1999;23:
81–86.
15 Krause A: Prometheus – a new extracorporeal liver support therapy; in Arroyo V, Forns X, Garcia-
Pagán JC, Rodés J (eds): Progress in the Treatment of Liver Diseases. Toronto, Ars Medica, 2003,
pp 437–443.
16 Rifai K, Manns PM: Clinical experience with Prometheus. Ther Apheresis Dial 2006;10:132–137.
17 Herget-Rosenthal S, Treichel U, Saner F: Citrate anticoagulated modified fractionated plasma sep-
aration and adsorption: first clinical efficacy and safety data in liver failure (poster). American
Society of Nephrology (ASN), San Diego, 2003.
18 Kramer L, Bauer E, Schenk P, et al: Successful treatment of refractory cerebral oedema in
ecstasy/cocaine-induced fulminant hepatic failure using a new high-efficacy liver detoxification
device (FPSA-Prometheus). Wien Klin Wochenschr 2003;115:599–603.
19 Santoro A, Faenza S, Mancini E, Ferramosca E, Grammatico F, Zucchelli A, Facchini MG, Pinna AD:
Prometheus system: a technological support in liver failure. Transplant Proc 2006;38: 1078–1082.
Antonio Santoro
Malpighi Nephrology, Dialysis and Hypertension Unit
Policlinico S.Orsola-Malpighi
Via P. Palagi 9
IT–40138 Bologna (Italy)
Tel. ⫹39 051 6362430, Fax ⫹39 051 6362511, E-Mail santoro@aosp.bo.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 405–410
Coupled Plasma Filtration Adsorption
Marco Formica
a
, Paola Inguaggiato
a
, Serena Bainotti
a
, Mary Lou Wratten
b
a
Nephrology and Dialysis Unit, S. Croce and Carle Hospital, Cuneo, and
b
Sorin Group Italia (Bellco), Scientific Research Department, Mirandola, Italy
Abstract
Sepsis is one of the main causes of death in critically ill patients worldwide, and in many
cases it is associated with renal and/or other organ failure. However, we do not have a unique
efficient therapy to reduce this extremely high mortality rate. In the last years interest around
the use of extracorporeal blood purification techniques has increased. One of the emerging
treatments in patients with severe sepsis and septic shock is coupled plasma filtration adsorp-
tion (CPFA), a novel extracorporeal blood purification therapy aimed at a nonselective reduc-
tion of the circulating levels and activities of both pro- and anti-inflammatory mediators. Early
experimental studies and the following clinical trials have demonstrated impressive results
regarding hemodynamics and respiratory parameters, even in patients without concomitant
acute renal injury, paralleled by a quick tapering of vasoactive drugs. Considering the still high
morbidity and mortality rates in septic shock patients, this new blood purification technique
seems to have benefits when applied early in the course of sepsis, also without renal indica-
tions, suggesting that it might be performed to prevent rather than to treat acute kidney injury.
Copyright © 2007 S. Karger AG, Basel
Epidemiology of Sepsis
Sepsis represents a crucial source of morbidity and mortality in critically
ill patients: in the United States it is the tenth most frequent cause of death in
intensive care units (ICU) [1]. As reported in the literature, patients diagnosed
with sepsis and severe sepsis or septic shock have a mortality rate ranging from
20 to 80%. This extreme variability is due to several factors: the huge disparity
of the clinical picture, the number of involved organs, the study design and the
timing and quality of treatment. The kidney is often injured during sepsis; then
renal failure alone or in association with other organ failure, in the so-called
M.L. Wratten is a full-time employee of the Sorin Group Italia (Bellco).
Formica/Inguaggiato/Bainotti/Wratten 406
multiorgan dysfunction syndrome, may be the clinical expression of the exag-
gerated host response to infection [2]. Indeed, sepsis is the leading cause of
acute renal injury (AKI) and its prevalence ranges from 19% in cases of sepsis,
to 23% in severe sepsis and up to 51% in septic shock [3].
Pathophysiology of Sepsis
Sepsis is considered an exaggerated immune response to infection, leading
to an overproduction and release of a wide array of pro- and anti-inflammatory
molecules. To summarize, there are two important pathways: the proinflamma-
tory response and the opposite immunosuppressive (or immunodysfunctional)
response. The first pathway determines the release of tumor necrosis factor-␣
(TNF-␣), interleukin-1 (IL-1), and IL-6 that have a predominantly proinflam-
matory role, whereas the second pathway leads to the release of IL-10 and IL-4,
cytokines with a predominantly anti-inflammatory activity. These two pathways
might be present at the same time, and not necessarily in sequence as previously
thought. Cytokines seem to be the principal factors causing diffuse endothelial
injury, because they are able to induce vasoparalysis and drive selective perme-
ability with important consequences on systemic hemodynamics. Moreover,
during sepsis monocytes lose their ability to synthesize and deliver cytokines,
determining the so-called ‘immunoparalytic’ state [4].
Treatment of Sepsis
The prognosis of patients admitted to ICU with septic shock and multi-
organ dysfunction syndrome today still has high mortality and all the attempts
to find a ‘magic bullet’ to restore the immune derangements have failed, due to
the complex interactions that take place between the pro- and anti-inflamma-
tory responses along with the clinical course of sepsis. At the present time no
effective therapy is available for sepsis. Several attempts have been made in
recent years, but clinical trials did not show any good therapeutic agent target-
ing the specific components of this pathological cascade and having a positive
effect on outcome [5]. Then, the attention moved to techniques able to remove
different circulating cytokines, that is extracorporeal blood purification [6, 7].
Standard continuous renal replacement therapies do not show a great ability to
remove sepsis-related molecules because of the small volumes employed and
low membrane sieving coefficient [8], whereas the use of a large-pore mem-
brane seems to have a higher clearance of cytokines because of an increased
convective transfer, but at the expense of added albumin loss [9, 10].
Coupled Plasma Filtration Adsorption 407
Coupled Plasma Filtration Adsorption: Technical Characteristics
Thus, a new extracorporeal blood purification technique was developed,
coupled plasma filtration adsorption (CPFA), which couples plasma filtration
and adsorption using a resin cartridge, along with a second hemofiltration sys-
tem that allows convective exchange. The nonselective removal of inflammatory
mediators is due to a hydrophobic styrenic resin, which has high affinity and a
large capacity for many cytokines and mediators [11]. The sorbent adsorption
allows reinfusion of the endogenous plasma after nonselective, simultaneous
removal of different sepsis-associated mediators through a specific cartridge.
CPFA is performed using a four-pump modular treatment (Lynda, Bellco
®
,
Mirandola, Italy) that consists of a plasma filter (0.45 m
2
polyethersulfone with
an approximate cutoff of 800 kDa), a nonselective hydrophobic resin cartridge
(140 ml) with a surface of about 700 m
2
/g, and a synthetic, high-permeability,
1.4-m
2
polyethersulfone hemofilter where convective exchanges may be applied
to the reconstituted blood in a postdilutional mode (fig. 1). The innovative aspect
of this technique is the application of the sorbent to plasma instead of blood [12].
This aspect has important advantages: the lower flow of plasma allows a longer
contact time with the sorbent, and biocompatibility problems are avoided. The
postdilution reinfusion rate can be set for up to a maximum of 4 liters/h. The
blood flow is usually 150–180 ml/min while the plasma filtration rate is main-
tained at a fractional filtration of the blood flow (approximately 15–20%). The
treatment is usually run for approximately 10h, after which the cartridge begins
to show saturation of the mediators.
Plasma
Reinfusate in
UF out
‘Bad molecules’
‘Good molecules’
Fig. 1. Schematic diagram of CPFA.
UF ⫽ Ultrafiltrate.
Formica/Inguaggiato/Bainotti/Wratten 408
Clinical Results
The treatment goal of CPFA is to target the excess of circulating pro- and
anti-inflammatory mediators to restore the normal immune function. Ronco et al.
[13] performed some of the first clinical CPFA studies with a prototype
machine, and showed that a 10-hour session of CPFA had a better impact on
hemodynamics, expressed as increased mean arterial pressure (MAP) and
decreased norepinephrine requirement, if compared to standard continuous
venovenous hemodiafiltration (CVVHDF). Moreover, they demonstrated that
monocytes obtained from blood treated with CPFA recovered their ability in
responding to the lipopolysaccaride challenge with TNF-␣ production to a
much greater extent than blood treated by CVVHDF. Subsequently, it was
hypothesized that CPFA might have a role in the treatment of septic patients,
not only for renal function substitution but also for immunomodulation activity.
Still employing the prototype machine, Formica et al. [14] evaluated the hemo-
dynamic performance to CPFA taking into account two unique features: (1)
repeated application of the technique (for a mean of 10 h/day) along the course
of the septic shock and (2) the use in patients without concomitant AKI. These
authors demonstrated that hemodynamic and respiratory parameters, such as
MAP, cardiac index, peripheral vascular resistances and oxygen arterial pressure/
inspired oxygen fraction ratio, were improved, so that norepinephrine adminis-
tration could be stopped after a mean of 5 CPFA sessions (fig. 2). CPFA seems
to be feasible and safe, without clinical side effects and risks, and might be reli-
able even in the absence of AKI; in patients with AKI, more special anticoagu-
lation schedules may be customized for different needs [14, 15]. The improvement
in splanchnic perfusion, evaluated by means of tonometry of gastric-mucosal
PCO
2
diffusion, is also of interest, which could further underline the resolution
of the hyperdynamic-vasoparalytic state displayed in septic shock. Thus, CPFA
is a treatment targeted to the nonselective removal of soluble mediators
involved in the septic shock scenario. We can furthermore speculate that the
association of different removal mechanisms (diffusion/convection/ adsorption)
may play a role in reestablishing a new immune balance (immune modulation)
with a significant reduction of acute phase reactants by hampering their peaks
[16, 17]. The results may be related to its ability to restore leukocyte respon-
siveness to immunoactive stimuli and this may be clinically beneficial because
of the link to hemodynamic improvement [13]. Despite the fact that some clin-
ical results appear to be quite good, it is mandatory to interpret them with cau-
tion because of the small sample size considered, but this underlines that these
good outcome results, where achieved, fueled the relationship between the
nephrologists and the intensivists in managing these very complex, critically ill
patients.
Coupled Plasma Filtration Adsorption 409
Future Developments
It is reasonable to propose to extend this technique also to former stages
of septic shock, such as severe sepsis or systemic inflammatory response syn-
drome along with pancreatitis. Further ongoing studies may provide evidence
about the potential efficacy of blood purification systems such as CPFA in
critically ill septic patients. Among these, the target of one prospective ran-
domized trial is to compare the clinical outcome of septic patients with AKI
treated either with CPFA or pulse high volume hemofiltration. Furthermore, a
large Italian multicenter study has been initiated, registered on Clinical
Trial.gov with the identifier NCT00332371 and on ISRCTN with the code
ISRCTN24534559, identified by the acronym COMPACT (Combined
Plasmafiltration and Adsorption Clinical Trial), regarding treatment of septic
shock patients with early initiation (within 6 h of diagnosis) of CPFA. The
main target of this study is to compare hospital and ICU mortality and morbid-
ity rates between patients treated with standard medical care alone and those
treated with standard medical care and CPFA. In this trial, started in December
2006, 330 patients will be recruited, in order to show a mortality difference of
25% between the two arms.
References
1 Martin GS, Mannino DM, Eaton S, Moss M: The epidemiology of sepsis in the United States
from1979 through 2000. N Engl J Med 2003;348:1546–1554.
2 Schrier RW, Wang W: Acute renal failure and sepsis. N Engl J Med 2004;351:159–169.
3 Rangel-Frausto MS, Pittet D, Costigan M: The natural history of the systemic inflammatory
response syndrome (SIRS). A prospective study. JAMA 1995;273:117–123.
Before CPFA After CPFA
60
70
80
90
100
MAP (mmHg)
Fig. 2. MAP behavior before and after
CPFA treatments. Before vs. after: 77.2 ⫾
12.5 vs. 83.3 ⫾ 14.1 mm Hg (p ⬍ 0.001).
Formica/Inguaggiato/Bainotti/Wratten 410
4 Cavaillon JM, Adib-Conquy M, Cloez-Tayarani I, Fitting C: Immunodepression in sepsis and SIRS
assessed by ex vivo cytokine production is not a generalized phenomenon: a review. J Endotoxin
Res 2001;7:85–93.
5 Zeni D, Freeman B, Nathanson C: Anti-inflammatory therapies to treat sepsis and septic shock.
Crit Care Med 1997;25:1095–1100.
6 Schetz M: Nonrenal indications for continuous renal replacement therapy. Kidney Int 1999;56
(suppl 72):S88–94.
7 Cole L, Bellomo R, Journois D, Davenport P, Baldwin I, Tipping P: High-volume hemofiltration
in human septic shock. Intensive Care Med 2001;27:978–986.
8 De Vriese AS, Vanholder RC, De Sutter JH, Colardyn FA, Lameire NH: Continuous renal replace-
ment therapies in sepsis: where are the data? Nephrol Dial Transplant 1998;13:1362–1364.
9 Tetta C, Cavaillon JM, Schulze M, Ronco C, Ghezzi PM, Camussi G, Serra AM, Curti F,
Lonnemann G: Removal of cytokines and activated complement components in an experimental
model of continuous plasma filtration coupled with sorbent adsorption. Nephrol Dial Transplant
1998;13:1458–1464.
10 Tetta C, Gianotti L, Cavaillon JM, Wratten ML, Fini M, Braga M, Bisagni P, Giavaresi GL,
Bolzani R, Giardino R: Continuous plasma filtration coupled with sorbent adsorption in a rabbit
model of endotoxic shock. Crit Care Med 2000;28:1526–1533.
11 Winchester JF, Kellum JA, Ronco C, Brady JA, Quartararo PJ, Salsberg JA, Levin NW: Sorbents
in acute renal failure and the systemic inflammatory response syndrome. Blood Purif 2003;21:
79–84.
12 Reeves JH, Butt WW, Shann F, Layton JE, Stewart A, Waring PM, Presneill JJ: Continuous
plasmafiltration in sepsis syndrome. Crit Care Med 1999;27:2096–2104.
13 Ronco C, Brendolan A, Lonnemann G, Bellomo R, Piccinni P, Digito A, Dan M, Irone M, La
Greca G, Inguaggiato P, Maggiore U, De Nitti C, Wratten ML, Ricci Z, Tetta C: A pilot study of
coupled plasma filtration with adsorption in septic shock. Crit Care Med 2002;30:1250–1255.
14 Formica M, Olivieri C, Livigni S, Cesano G, Vallero A, Maio M, Tetta C: Hemodynamic response
to coupled plasmafiltration-adsorption in human septic shock. Intensive Care Med 2003;29:
703–708.
15 Mariano F, Tetta C, Stella M, Biolino P, Miletto A, Triolo G: Regional citrate anticoagulation in
critically ill patients treated with plasmafiltration and adsorption. Blood Purif 2004;22:313–319.
16 Opal S: Hemofiltration-adsorption systems for the treatment of experimental sepsis: it is possible
to remove the ‘evil humors’ responsible for septic shock? Crit Care Med 2000;28:1681–1682.
17 Ronco C, Tetta C, Mariano F, Wratten ML, Bonello M, Bordoni V, Cardona X, Inguaggiato P,
Pilotto L, D’Intini V, Bellomo R: Interpreting the mechanisms of continuous renal replacement
therapy in sepsis: the peak concentration hypothesis. Artif Organs 2003;27:792–801.
Dr. Marco Formica
Nephrology and Dialysis Unit, S. Croce and Carle Hospital
Via Carle, 25
IT–12100 Cuneo (Italy)
Tel. ⫹39 0171 616241, Fax ⫹39 0171 616120, E-Mail formica.m@ospedale.cuneo.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 411–418
Albumin Dialysis and Plasma Filtration
Adsorption Dialysis System
Federico Nalesso
a
, Alessandra Brendolan
a
, Carlo Crepaldi
a
, Dinna Cruz
a
,
Massimo de Cal
a
, Rinaldo Bellomo
b
, Claudio Ronco
a
a
Department of Nephrology, Dialysis and Transplantation, San Bortolo Hospital, Vicenza,
Italy;
b
Department of Intensive Care, Austin Hospital, Melbourne, Vic., Australia
Abstract
Albumin-bound toxins are important in the pathophysiology of liver failure, systemic
inflammatory response syndrome, and poisoning. Due to its intrinsic ability to bind mole-
cules, albumin has been used in blood purification techniques, such as single pass albumin
dialysis, the molecular adsorbent recirculating system and the Prometheus systems. Plasma
filtration adsorption dialysis is the latest technology that can combine the best processes of
blood/plasma purification in order to determine a selective and effective purification of
molecules implicated in liver failure.
Copyright © 2007 S. Karger AG, Basel
Molecules are present in the plasma, in the plasmatic water or bound to
specific or unspecific carriers. The most important plasmatic carrier is albu-
min. According to their characteristics of solubility, plasmatic molecules are
present as solution in the plasmatic water, if water soluble, or bound to carriers,
if hydrophobic.
Techniques such as hemodialysis, hemofiltration, and their combination
are able to remove small molecules and medium molecules acting on plasmatic
water and its solutes. According to the membrane cutoff and high volume of
infusion, they can improve the total removal of molecules with high molecular
weight compared with small molecules. In order to improve the efficiency of
the removal of the molecules, it is possible to combine the convective and/or
diffusive processes with adsorption on specific materials.
The adsorption allows removing a wide range of hydrophobic and higher
molecular weight substances such as bilirubin, salt acids, cytokines, myoglobin
Nalesso/Brendolan/Crepaldi/Cruz/de Cal/Bellomo/Ronco 412
and others. The possibility of using specific physical interaction in some mole-
cule absorbers (ion exchange, chemical affinity, Van der Waals forces) allows
for the removal of specific molecule targets such as bile acids and bilirubin dur-
ing liver failure. The adsorption process acts on protein-bound substances and
high molecular weight toxins present in the plasma. Convection and diffusion
are not able to obtain good clinical clearances of high molecular weight or
hydrophobic molecules due to their theoretical and practical limitations (vol-
ume of infusion and cutoff membrane).
Summarizing all these concepts, we can understand and highlight the cen-
tral role of plasma as a transporter of toxic molecules and its potential function
in the purification of blood. Therefore, thanks to its intrinsic capacity to bind
and transport molecules, plasma is the best fluid to perform a purification
process. It seems useful to combine the physical and chemical principles of
purification (diffusion, convection and adsorption) in order to improve and
obtain the best removal of substances. According to this view, there is the possi-
bility to use the albumin as a medium of purification thanks to its capacity to
bind toxins.
A critical issue of the clinical syndrome in liver failure is the accumulation
of toxins not cleared by the failing liver. Based on this hypothesis, the removal
of lipophilic, albumin-bound substances such as bilirubin, bile acids, meta-
bolites of aromatic amino acids, medium-chain fatty acids and cytokines should
be beneficial to the clinical course of a patient in liver failure. This led to the
development of artificial filtration and adsorption devices.
Hemodialysis, hemofiltration and their combination are used in renal fail-
ure and primarily remove water-soluble toxins; however, they do not remove
toxins bound to albumin that accumulate in liver failure because of the techni-
cal limitations to use high cutoff membrane and shift and removal of albumin-
bound toxins.
Artificial detoxification devices currently under clinical evaluation include
the molecular adsorbent recirculating system (MARS), single pass albumin
dialysis and the Prometheus system. A new system is going to be developed in
the Dialysis Unit of San Bortolo Hospital, Vicenza thanks to its innovative con-
ception of patient’s plasma as substrate and medium of purification; the system
is the plasma filtration adsorption dialysis (PFAD) technology.
Molecular Adsorbent Recirculation System
The MARS, developed by Teraklin of Germany, is the best-known extra-
corporeal liver dialysis system. It consists of two separate dialysis circuits. The
first circuit consists of human albumin, is in contact with the patient’s blood
Albumin Dialysis and PFAD System 413
through a semipermeable membrane and has two special devices to clean the
albumin after it has absorbed toxins from the patient’s blood. The second circuit
consists of a hemodialysis machine and is used to purify the albumin in the first
circuit, before it is recirculated to the semipermeable membrane in contact with
the patient’s blood. The MARS system [1] can remove a number of toxins,
including ammonia, bile acids, bilirubin, copper, iron and phenols.
Single Pass Albumin Dialysis
Single pass albumin dialysis is a simple method of albumin dialysis using
standard renal replacement therapy machines without an additional perfusion
pump system [2]: the patient’s blood flows through a circuit with a high-flux
hollow fiber hemodiafilter. The other side of this membrane is cleansed with an
albumin solution in counterdirectional flow (such as a standard dialysate during
a bicarbonate dialysis), which is discarded after passing the filter. The albumin
can be used in single pass or regenerated by adsorber and reused in a closed
system.
Prometheus
The Prometheus system (Fresenius Medical Care, Bad Homburg,
Germany) is a new device based on the combination of albumin adsorption with
high-flux hemodialysis after selective filtration of the albumin fraction through
a specific polysulfone filter (AlbuFlow) [3].
Plasma Filtration Adsorption Dialysis
The PFAD technology [4] is based on a new principle of purification that
utilizes a tricompartmental dialyzer (TD) to purify the patient’s blood by a com-
bination of three sequential techniques: convection and adsorption both on
plasma followed by a process of ‘whole blood dialysis’ provided by the regener-
ated patient’s own plasma (fig. 1).
The TD is the core of this new technology (fig. 2). It is composed of hol-
low fibers like a regular dialyzer for hemodialysis. The compartments are
located in different areas, and each has its own particular function. The hollow
fibers form three compartments along the extension of the dialyzer (fig. 2): the
first compartment is formed by the inner space of hollow fibers in which
the blood goes through the length of the whole fibers thanks to a roller pump
Nalesso/Brendolan/Crepaldi/Cruz/de Cal/Bellomo/Ronco 414
(blood pump). The internal compartment of the dialyzer is divided into two
more compartments separated by a wall along the extension of the hollow
fibers: the second compartment forms a stage for filtering plasma, and the third
compartment forms a stage for dialysis. The second compartment is the delin-
eated space where the patient’s plasma can be filtered from the whole blood
across the hollow fiber membrane (fig. 1, number 1). The third compartment is
the space where the patient’s regenerated plasma performs a process of purifi-
cation based on a ‘diffusive and binding process’; in this way the regenerated
patient’s plasma is used as a dialysate in countercurrent to purify the blood
flowing in the first compartment (fig. 1, number 2). The second and third com-
partment have a specific cutoff of hollow fiber membranes according to their
specific function and their area is able to assure the processes that occur (filtra-
tion and dialysis). The second and the third compartment communicate through
Infusion
bag
b
c
d
4
3
2
e
1
a
Patient
Waste
bag
Description
a
Blood pump, blood flow ⫽ Q
B
1
Plasma separator (second compartment)
b
Plasma pump, plasma flow ⫽ Q
P
c
Infusion pump, reinfusion flow ⫽ Q
R
d
Ultrafiltration pump
3
Filter to perform the convection on plasma
4
Adsorber
e
Plasma dialysate pump, dialysate flow ⫽ Q
D
2
Dialyzer (third compartment)
Fig. 1. Schematic representation of PFAD circuits. The second (1) and third (2) com-
partment are described as separate devices in order to simplify the explanation of single
processes.
Albumin Dialysis and PFAD System 415
First compartment Second compartment
Plasma purification Plasma reinfusion
Recycling virtual systemPlasma dialysis
Plasma used
as dialysate
Blood
Compartment
for whole
blood dialysis
Communication
between the
plasma filtration
and dialysis
compartment
Filtered
plasma
Venous line
Plasma reinfused
to the patient
Plasma
adsorption
cartridge
Blood
Filtered
plasma
Plasma adsorption
cartridge
Plasma purification
by convection
and/or diffusion
Plasma purification circuit
Blood
Hollow fibers
Compartment for
plasma filtration
Tricompartmental dialyzer
Blood
Filtered plasma
Hollow fibers
Compartment for
plasma filtration
Tricompartmental dialyzer
Venous line
Plasma
reinfused
to the patient
Filtered
plasma
Plasma
adsorption
cartridge
Blood
Plasma
reinfused
to the patient
Plasma used
as dialysate
Blood
Compartment
for whole
blood dialysis
Communication
between the
plasma filtration
and dialysis
compartment
Filtered plasma
Venous line
Plasma adsorption
cartridge
Plasma
adsorption
cartridge
Plasma purification by
convection and/or diffusion
Compartment for
plasma filtration
Fig. 2. The TD and the purification processes.
Nalesso/Brendolan/Crepaldi/Cruz/de Cal/Bellomo/Ronco 416
a particular opening in the arterial extremity of dialyzer (fig. 2, not shown in
fig. 1). In the first human prototype the second and the third compartment are
separated and formed by two different devices, as shown in the figures.
The first step of the process aims to separate the plasma from whole blood
(fig. 1). This process determines the plasma filtration from the inner space of
hollow fibers to the space of the second compartment. The obtained patient’s
plasma goes from the second compartment to the plasma purification circuit
where it is purified by two different and separate methods: convection and
adsorption (fig. 1: 3 ⫽ convection process, 4 ⫽ adsorption). The plasma flow is
obtained by a roller pump (fig. 1, b).
The plasma purification circuit (fig. 1) is composed of two separate
processes in order to remove first the water-soluble and dialyzable toxic mole-
cules by convection and then hydrophobic and nondialyzable molecules by
adsorption on a specific adsorber (fig. 1, numbers 3 and 4). The convection
process is obtained by performing a high volume hemofiltration directly on
plasma. It is known that high volume ultrafiltration using a super high flux fil-
ter has achieved better cytokine clearances compared to those currently
achieved by urea during standard continuous renal replacement therapy [5]. The
convective process is able to reestablish hydroelectrolytes, acid-base equilib-
rium and fluid balance acting directly on plasmatic water.
After the convective purification, the plasma is adsorbed by a specific
adsorber to remove hydrophobic or nondialyzable molecules (fig. 1, number 4).
The adsorber is specific for the molecules implicated in the patient’s disease
(sepsis, hepatorenal syndrome, acute and chronic liver failure). The cartridge
for adsorption presents a good pressure-flow performance and excellent
mechanical and chemical stability in order to perform the best adsorption of
plasma.
After these two different processes, the purified plasma can take two sepa-
rate paths (fig. 1). In the first pathway, the purified plasma returns to the patient
through the venous line, in the other, the patient’s regenerated plasma is used as
dialysate in the suitable compartment of TD, in order to perform the dialysis
procedure based on the ‘diffusion and binding process’. In this step, the
patient’s filtrated plasma from the whole blood and the spent dialysate from the
third compartment can be purified again in the plasma circuit (fig. 1). In fact
there is a connection between the second and the third compartment at the arte-
rial extremity of TD; thus, the spent dialysate can go to the plasma circuit
through the second compartment (fig. 2) generating a virtual open loop. In the
human prototype the separation between the second and the third compartment
assures the same plasma process in the open loop.
The plasma circuit is an open circuit at the venous line; therefore, the dialy-
sis system is not working in closed recycling modality, thus, having new fresh
Albumin Dialysis and PFAD System 417
regenerated plasma from the plasma circuit every time it is fed by new plasma
from the second compartment of TD (fig. 1). The unique feature of the plasma
circuit is that the plasma flow to obtain the dialysate in the third compartment is a
virtual flow inside the open plasma loop and can exceed the plasma filtration flow
from the first compartment being a recycling flow, as shown in the drawings.
The PFAD is a technique that can be continuously performed for at least
8 h or more a day.
Discussion
The PFAD combines three different processes in order to purify the
patient’s whole blood using the patient’s plasma. With this technology the
plasma is both a medium and a substrate of purification.
This technology allows for the removal of molecules from plasmatic water
and protein-bound plasma by the sequential combination of convective treat-
ment and adsorption. According to the water solubility and molecular weight,
the PFAD can remove both diffusible and nondiffusible molecules, thanks to the
combination of specific techniques. In the first step, water-soluble molecules
can be removed by convection. This purification process can improve the selec-
tion of successive adsorption which is able to remove hydrophobic and large
molecular weight molecules. In the plasma circuit the patient’s plasma is the
substrate of purification but when it is reinfused or used as dialysate it becomes
a medium of purification. Therefore, the regenerated albumin and carriers are
able to bind toxic ligands according to the ‘law of mass action’. This is an
important physical process that allows ligands to be shifted from the whole
blood and tissues and bound to the regenerated carriers across the TD or capil-
lary membrane in the PFAD system and patient’s body, respectively.
In liver failure the accumulation of albumin-bound toxins has been demon-
strated; these toxins are responsible, to variable extents, for multiple-organ dys-
function (kidney, cardiovascular instability) [6]. The function of albumin as a
transporter and as a possible purification vector have been described in albumin
dialysis, in which the removal of these molecules improves the clinical condi-
tion of the patient. The best known and most widely used extracorporeal device
for liver function support is MARS [7], which uses albumin only to perform
purification by adsorption and by classic dialysis. The current literature demon-
strates that this approach is capable of improving patient survival [8, 9].
Moreover, this type of approach is useful for intoxications caused by exogenous
pathogens that are scarcely water-soluble but are plasma protein-bound.
In several pathologies (sepsis, hepatorenal syndrome, systemic inflamma-
tory response syndrome) there is an involvement of cytokines, and much of the
Nalesso/Brendolan/Crepaldi/Cruz/de Cal/Bellomo/Ronco 418
damage that affects the various organs and systems that are not primarily
involved in the basic pathological process are determined by molecular factors
that circulate in the blood or are present in the plasma. In the PFAD the speci-
ficity of the adsorber used for the plasma adsorption is important to remove the
toxic molecules retained or produced during liver failure.
References
1 Saliba F: The molecular adsorbent recirculating system (MARS) in the intensive care unit: a res-
cue therapy for patients with hepatic failure. Crit Care 2006;10:118.
2 Chawla LS, Georgescu F, Abell B, Seneff MG, Kimmel PL: Modification of continuous venove-
nous hemodiafiltration with single-pass albumin dialysate allows for removal of serum bilirubin.
Am J Kidney Dis 2005;45:e51–e56.
3 Rifai K, Manns MP: Clinical experience with Prometheus. Ther Apher Dial 2006;10:132–137.
4 Nalesso F: Patent of PFAD. WO2004091694, 2004.
5 Uchino S, Bellomo R, Goldsmith D, Davenport P, Cole L, Baldwin I, Panagiotopoulos S, Tipping P:
Super high flux hemofiltration: a new technique for cytokine removal. Intensive Care Med 2002;28:
651–655.
6 Sen S, Jalan R, Williams R: Liver failure: basis of benefit of therapy with the molecular adsor-
bents recirculating system. Int J Biochem Cell Biol 2003;35:1306–1311.
7 Sen S, Williams R, Jalan R: Emerging indications for albumin dialysis. Am J Gastroenterol 2005;100:
468–475.
8 Isoniemi H, Koivusalo AM, Repo H, Ilonen I, Hockerstedt K: The effect of albumin dialysis on
cytokine levels in acute liver failure and need for liver transplantation. Transplant Proc 2005;37:
1088–1090.
9 Sen S, Jalan R, Williams R: Liver failure: basis of benefit of therapy with the molecular adsor-
bents recirculating system. Int J Biochem Cell Biol 2003;35:1306–1311.
Prof. Claudio Ronco
Department of Nephrology, Dialysis and Transplantation
St. Bortolo Hospital, Viale Rodolfi 31
IT–36100 Vicenza (Italy)
Tel. ⫹39 0444 753 689, Fax ⫹39 0444 753 949, E-Mail Cronco@goldnet.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 419–427
Renal Assist Device and Treatment of
Sepsis-Induced Acute Kidney Injury in
Intensive Care Units
Naim Issa
a
, Jennifer Messer
b
, Emil P. Paganini
b
a
Department of Nephrology and Hypertension, and
b
Section of Dialysis and
Extracorporeal Therapy, Department of Nephrology and Hypertension,
Cleveland Clinic, Cleveland, Ohio, USA
Abstract
Acute kidney injury (AKI) is a frequent and serious complication of sepsis in ICU
patients and is associated with a very high mortality. Despite the advent of sophisticated
renal replacement therapies (RRT) employing high-dose hemofiltration and high-flux mem-
branes, mortality and morbidity from sepsis-induced AKI remained high. Moreover, these
dialytic modalities could not substitute for the important functions of renal tubular cells in
decreasing sepsis-induced AKI biological dysregulations. The results from the in vitro and
preclinical animal model studies were very intriguing and led to the development of a bioar-
tificial kidney consisting of a renal tubule assist device containing human proximal tubular
cells (RAD) added in tandem to a continuous venovenous hemofiltration circuit. The results
from the phase I safety trial and the recent phase II clinical trial showed that the RAD not
only can replace many of the indispensable biological kidney functions, but also modify the
natural history of sepsis-induced AKI by ameliorating patient survival.
Copyright © 2007 S. Karger AG, Basel
Sepsis-Associated Acute Kidney Injury and Available
Renal Replacement Therapies
Acute kidney injury (AKI) occurs in approximately 20% of patients with
sepsis and 51% of patients with septic shock with positive blood cultures [1].
Sepsis-induced AKI can be associated with 70% mortality as compared
with 40% mortality among patients with AKI alone [2]. Sepsis stimulates the
induction of nitric oxide synthase leading to nitric oxide-mediated arterial vasodi-
latation. The arterial vasodilatation induces a decrease in systemic vascular
Issa/Messer/Paganini 420
resistance resulting in increased sympathetic tone and the release of vasopressin
from the central nervous system along with activation of the renin-angiotensin-
aldosterone system. The resulting renal vasoconstriction induces sodium and
water retention and predisposes to AKI [3–5]. Sepsis also induces the genera-
tion of oxygen radicals that scavenge renal endothelial nitric oxide thereby
causing peroxynitrite-related acute tubular injury and necrosis [6]. Following or
more likely concomitant to this renal vasoconstrictor phase, a proinflammatory
phase involving cytokines and chemokines leads to further acute injury of the
renal endothelium and an increased rate of patient death [7]. Patients with AKI
especially when associated with sepsis are extremely hypercatabolic and fre-
quently require renal replacement therapy (RRT) in the form of intermittent or
continuous RRT. Neither form of RRT has proved to be superior in terms of sur-
vival and renal recovery. Despite the advent of RRT, mortality rates from AKI
have not changed significantly and the appropriate dose of dialysis in AKI has
not been defined to date. A recent study showed that intensive daily hemodialysis
(HD) compared with alternate-day HD reduced mortality (28 vs. 46%, p ⬍ 0.01)
without increasing hemodynamically-induced morbidity [8]. Moreover inten-
sive daily HD was associated with less systemic inflammatory response syn-
drome or sepsis (22 vs. 46%, p ⫽ 0.005) and a shorter duration of acute renal
failure (ARF; mean ⫾ SD, 9 ⫾ 2 vs. 16 ⫾ 6 days, p ⫽ 0.001). A randomized
study by Ronco et al. [9] showed that hemofiltration rates of 35 or 45 ml/kg/h
improve survival in ARF (p ⬍ 0.001) as compared with 20 ml/kg/h. To date any
survival benefit of cytokine removal by convective or diffusive mode in patients
with sepsis-induced AKI remains to be proven. In a recent study, the use of a
polyflux hemofilter with a high membrane cutoff point (approximately 60kDa)
convection had an advantage over diffusion in the clearance capacity of
cytokines, but was associated with greater plasma protein losses [10]. Neither
of these dialytic modalities could substitute for the important biological func-
tions of renal tubular cells in decreasing sepsis-induced AKI-associated mortal-
ity. Moreover, adding proximal tubular cells to the RRT circuit may offer a
more complete and physiological form of RRT.
Proximal Tubular Cell Therapy
Renal tubular epithelial cells are the main site of blood purification by
solute and fluid clearance as well as of reclamation of the essential electrolytes
and metabolites from the glomerular ultrafiltrate in order to maintain the
‘milieu interieur’ of the human body. These highly differentiated epithelial cells
actively transport electrolytes and water and perform other metabolic and
endocrinological activities as well. Tubular cells are very sensitive to ischemic
RAD and Treatment of Sepsis-Induced AKI 421
injury and sepsis [5] that can lead to acute tubular necrosis resulting in tubular
dysfunction, solute, electrolyte and fluid dysregulation [11] that may ultimately
require RRT. Fortunately, tubular cells have the capacity to regenerate and
regain their original functionality providing the underlying basement mem-
brane is not deranged. This regenerative capability led researchers to postulate
the presence of putative resident or migrating progenitor stem cells that may
furnish newly formed tubules as opposed to the same tubular cells undergoing
mitosis to form new tubular cells [12–17]. Humes et al. [18, 19] successfully
isolated human proximal tubular cells from deceased donor kidneys not suitable
for transplantation because of excessive fibrosis. These proximal tubular cells
were employed to create the renal assist device (RAD) as described later in the
text. In their preclinical animal and subsequent human clinical studies [18, 19],
the researchers demonstrated the vitality and functionality of these proximal
tubular cells seeded in the RAD cartridge in regulating active transport of elec-
trolytes and glucose, as well as glutathione metabolism, ammonia excretion and
1␣-hydroxylation of 25-dihydroxyvitamin D
3
, and regulating immune response
by decreasing proinflammatory cytokine levels. The procedure of human renal
tubular cell isolation is described in detail elsewhere [20] and is beyond the
scope of this review.
Bioartificial Kidney: in vitro Studies, Preclinical Animal Studies,
Circuit, and Phase I/II Clinical Studies
In vitro Studies
Humes et al. [21] developed the RAD that was tested in vitro for a variety
of differentiated tubular functions.
In these in vitro studies, they demonstrated that the RAD seeded with
porcine proximal tubular cells allowed vectorial transport of fluid from the intra-
luminal space to the antiluminal space through the Na
⫹
, K
⫹
-ATPase pumps of
the tubular cells. The RAD also facilitates other important metabolic and
endocrine activities of the renal proximal tubular cells analogously to a nephron
[21, 22]. These important functions included active bicarbonate and glucose
transports, intraluminal glutathione breakdown into its constituent amino acids,
ammonia production, and conversion of 25-(OH)-vitamin D
3
to the active form
1,25-(OH)
2
vitamin D
3
.
Preclinical Animal Studies
Humes et al. [18] developed animal models to support their hypothesis
that the RAD provides incremental renal replacement support, thus decreasing
morbidity and mortality rates observed in patients with AKI. They postulated
Issa/Messer/Paganini 422
that the RAD tubular cells supplemented the standard RRT circuit, conferring
benefits by adding the functional properties of the nephron such as clearance
and other important metabolic and endocrinological functions. They also pos-
tulated that the RAD can play a role in normalizing proinflammatory cytokine
imbalance that characterizes systemic inflammatory response syndrome and
multiorgan failure. Their initial preclinical animal model consisted of nephrec-
tomized dogs with endotoxin lipopolysaccharide-induced hypotension. In this
study, the RAD increased ammonia excretion, glutathione metabolism, and
1,25-dihydroxy-vitamin D
3
production in uremic dogs. Moreover, the dogs had
excellent cardiovascular stability during the extracorporeal therapy with the
RAD [18].
Further intriguing preclinical animal trials on acutely uremic animals with
induced septic shock showed that animals treated with cell RAD demonstrated
significantly better cardiovascular performance and survival times than those
treated with sham RAD lacking the tubular cells [23, 24]. Interestingly enough,
these preclinical trials showed that plasma cytokine levels were also altered dur-
ing the treatment with RAD. Specifically the IL-10 levels (p ⬍ 0.01) as well as
the mean arterial pressures (p ⬍ 0.04) were both significantly higher during
the treatment interval in the cell RAD animals compared to their sham
controls [23].
The same group performed a more realistic sepsis animal model employ-
ing nonnephrectomized pigs [25]. The pigs were administered 30 ⫻ 10
10
bacte-
ria/kg body weight of Escherichia coli intraperitoneally to induce septic shock
and thereby causing acute tubular necrosis [26]. The pigs were started on con-
tinuous venovenous hemofiltration (CVVH) along with a RAD in tandem to the
CVVH extracorporeal circuit. They were divided in two groups: the sham RAD
(without cells) or a cell RAD with renal proximal tubular cells. Cell RAD ther-
apy resulted in significantly higher cardiac outputs and renal blood flow and
was associated with significantly lower plasma circulating proinflammatory
cytokine concentrations (IL-6 and interferon-␥), resulting in a nearly twice the
average survival time in the cell-RAD-treated group compared with the sham
RAD control group.
Human Bioartificial Kidney: Hemofilter and Circuit
The hemofilter of the RAD that was used in human clinical studies con-
sisted of seeding human renal proximal tubular cells obtained from cadaveric
kidneys (deemed not suitable for kidney transplantation) into the hollow into
the hollow fiber membranes of a standard polysulfone high-flux hemofiltration
cartridge pretreated with a synthetic extracellular matrix protein [21]. This
hemofilter typically consists of confluent monolayers containing up to
1.5 ⫻ 10
9
cells. These seeded cells along the inner surface of the hollow fibers
RAD and Treatment of Sepsis-Induced AKI 423
remain immunoprotected from the patient’s blood by the semipermeable mem-
brane. The RAD is kept horizontally oriented and at a temperature of 37⬚C to
assure vitality and maximum functioning of the tubular cells.
The bioartificial kidney was then created by attaching the RAD to a con-
ventional CVVH system as shown in figure 1. The major difference between
the RAD and the conventional CVVH circuit is that the blood and ultrafiltrate
flow in opposite directions inside the hemofilter. In the RAD circuit, the ultra-
filtrate emanating from the CVVH circuit is diverted via a pump to the inside of
the hollow fibers and inundates the proximal tubular cells. On the other hand,
the blood derived from the CVVH circuit is shunted to the dialysate compart-
ment of the hemofiltration cartridge. This shunting process creates a ‘bioartifi-
cial nephron’ (fig. 2) in which the ultrafiltrate circulates inside the hollow
fibers of the dialyzer, bathes the proximal tubular cells allowing their interac-
tion, and thereby replaces the indispensable metabolic functions. Once processed,
the ultrafiltrate emanating from the RAD is discarded as urine. The blood exiting
the RAD is returned to the patient via a pump (fig. 1).
Standard dialysis system
Hemofilter
Hemofilter
RAD
Bioartificial system
Fig. 1. Schematic diagram of the extracorporeal circuit for the bioartificial kidney;
flow rates are detailed in the text.
Issa/Messer/Paganini 424
Phase I/II Clinical Studies
These promising preclinical animal studies led the US Food and Drug
Administration to approve a phase I safety clinical trial on 10 patients with ARF
and multiorgan failure on CVVH, with predicted hospital mortality rates aver-
aging above 85% according to the APACHE III score [19, 26]. A RAD, seeded
with human renal proximal tubule cells from discarded deceased donor kidneys
as described earlier, was placed in tandem with the CVVH system. The effluent
ultrafiltrate from the hemofilter was shunted to the inside of the dialyzer fibers
at a rate of 10 ml/min; the blood emanating from the hemofilter was pumped to
the outside space of the dialyzer of the RAD with the aid of a pump at
150 ml/min. The hydraulic pressure inside the RAD filter was adjusted to reab-
sorb 5 ml/min to the blood compartment so that the processed 5 ml of ultrafil-
trate was subsequently discarded. The blood was then returned to the patient
with the aid of a pump as described earlier (fig. 1).
This phase I safety study showed that the RAD therapy can be safely done
in combination with CVVH for up to 24 h under the protocol guidelines; it also
demonstrated cardiovascular stability as well as increased urine output from the
native kidneys and less vasopressor requirements with minimal side effects. No
evidence of death directly related to RAD was shown among the study population
Artificial glomerulus
Artificial tubule
Natural
glomerulus
Natural tubule
Fig. 2. Schema depicting the analogy of the bioartificial kidney with a human nephron.
RAD and Treatment of Sepsis-Induced AKI 425
with overall 30-day survival of 60% (6/10 patients). The study authors also
showed that for the subset of patients who had excessive proinflammatory
cytokine levels, RAD therapy resulted in significant declines in granulocyte
colony-stimulating factor, interleukin-6 (IL-6), IL-10 and IL-6/IL-10 ratios.
Moreover, RAD therapy evidenced metabolic and endocrinological activities
with increased glutathione degradation and conversion of 25-OH-vitamin D
3
to
1,25-(OH)
2
-vitamin D
3
analogously to the in vitro and the preclinical animal
studies. These auspicious results led the FDA to approve a randomized, open-
labeled multicenter phase II trial recently completed in which the RAD was
employed in 58 critically ill ICU patients with dialysis-dependent AKI [27]. The
major endpoint of the study was 28-day all-cause mortality. Patients were ran-
domized 2 to 1 after 6h of CVVH initiation to CVVH with or without RAD for
72 h. The impressive results showed lower 28-day mortality in patients receiv-
ing any duration of RAD therapy in comparison to patients receiving CVVH
without RAD (34.3 vs. 55.6%); these results warrant a larger clinical study.
Conclusion
All the preclinical and clinical trials demonstrate a safety profile of the
RAD therapy. RAD therapy not only replaces solute and water clearance but
also replaces active reabsorptive transport, metabolic functions, and beneficial
systemic effects that decrease the morbidity and mortality associated with AKI
in critically ill patients. This research also shows that cell therapy can be a very
promising step toward achieving better outcomes in patients with sepsis-
induced AKI by using a ‘living membrane’, thereby crossing the limitations of
diffusive and convective therapies by going beyond physics, artificial mem-
branes, and urea clearance. While the use of cell therapy has not yet been tested
in end-stage renal disease patients on chronic HD, it has the potential to become
complementary to a variety of dialytic therapies because of its ability to replace
many indispensable renal physiological functions.
References
1 Rangel-Frausto MS, Pittet D, Costigan M, Hwang T, Davis CS, Wenzel RP: The natural history of
the systemic inflammatory response syndrome (SIRS). A prospective study. JAMA 1995;273:
117–123.
2 Neveu H, Kleinknecht D, Brivet F, Loirat P, Landais P: Prognostic factors in acute renal failure due
to sepsis. Results of a prospective multicentre study. The French Study Group on Acute Renal
Failure. Nephrol Dial Transplant 1996;11:293–299.
3 Benedict CR, Rose JA: Arterial norepinephrine changes in patients with septic shock. Circ Shock
1992;38:165–172.
Issa/Messer/Paganini 426
4 Cumming AD, Driedger AA, McDonald JW, Lindsay RM, Solez K, Linton AL: Vasoactive hor-
mones in the renal response to systemic sepsis. Am J Kidney Dis 1988;11:23–32.
5 Schrier RW, Wang W: Acute renal failure and sepsis. N Engl J Med 2004;351:159–169.
6 Wang W, Jittikanont S, Falk SA, Li P, Feng L, Gengaro PE, Poole BD, Bowler RP, Day BJ, Crapo
JD, Schrier RW: Interaction among nitric oxide, reactive oxygen species, and antioxidants during
endotoxemia-related acute renal failure. Am J Physiol Renal Physiol 2003;284:F532–F537.
7 Schor N: Acute renal failure and the sepsis syndrome. Kidney Int 2002;61:764–776.
8 Schiffl H, Lang SM, Fischer R: Daily hemodialysis and the outcome of acute renal failure. N Engl
J Med 2002;346:305–310.
9 Ronco C, Bellomo R, Homel P, Brendolan A, Dan M, Piccinni P, La Greca G: Effects of different
doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective
randomised trial. Lancet 2000;356:26–30.
10 Morgera S, Slowinski T, Melzer C, Sobottke V, Vargas-Hein O, Volk T, Zuckermann-Becker H,
Wegner B, Muller JM, Baumann G, Kox WJ, Bellomo R, Neumayer HH: Renal replacement ther-
apy with high-cutoff hemofilters: impact of convection and diffusion on cytokine clearances and
protein status. Am J Kidney Dis 2004;43:444–453.
11 Schrier RW, Wang W, Poole B, Mitra A: Acute renal failure: definitions, diagnosis, pathogenesis,
and therapy. J Clin Invest 2004;114:5–14.
12 Fujigaki Y, Goto T, Sakakima M, Fukasawa H, Miyaji T, Yamamoto T, Hishida A: Kinetics and
characterization of initially regenerating proximal tubules in S3 segment in response to various
degrees of acute tubular injury. Nephrol Dial Transplant 2006;21:41–50.
13 Maeshima A, Sakurai H, Nigam SK: Adult kidney tubular cell population showing phenotypic
plasticity, tubulogenic capacity, and integration capability into developing kidney. J Am Soc
Nephrol 2006;17:188–198.
14 Yamashita S, Maeshima A, Nojima Y: Involvement of renal progenitor tubular cells in epithelial-
to-mesenchymal transition in fibrotic rat kidneys. J Am Soc Nephrol 2005;16:2044–2051.
15 Bussolati B, Camussi G: Adult stem cells and renal repair. J Nephrol 2006;19:706–709.
16 Oliver JA, Maarouf O, Cheema FH, Martens TP, Al-Awqati Q: The renal papilla is a niche for adult
kidney stem cells. J Clin Invest 2004;114:795–804.
17 Lin F, Moran A, Igarashi P: Intrarenal cells, not bone marrow-derived cells, are the major source
for regeneration in postischemic kidney. J Clin Invest 2005;115:1756–1764.
18 Humes HD, Fissell WH, Weitzel WF, Buffington DA, Westover AJ, MacKay SM, Gutierrez JM:
Metabolic replacement of kidney function in uremic animals with a bioartificial kidney containing
human cells. Am J Kidney Dis 2002;39:1078–1087.
19 Humes HD, Weitzel WF, Bartlett RH, Swaniker FC, Paganini EP, Luderer JR, Sobota J: Initial
clinical results of the bioartificial kidney containing human cells in ICU patients with acute renal
failure. Kidney Int 2004;66:1578–1588.
20 Smith PL, Buffington DA, Humes HD: Kidney epithelial cells. Methods Enzymol 2006;419:
194–207.
21 Humes HD, MacKay SM, Funke AJ, Buffington DA: Tissue engineering of a bioartificial renal
tubule assist device: in vitro transport and metabolic characteristics. Kidney Int 1999;55:
2502–2514.
22 Nikolovski J, Gulari E, Humes HD: Design engineering of a bioartificial renal tubule cell therapy
device. Cell Transplant 1999;8:351–364.
23 Fissell WH, Dyke DB, Weitzel WF, Buffington DA, Westover AJ, MacKay SM, Gutierrez JM,
Humes HD: Bioartificial kidney alters cytokine response and hemodynamics in endotoxin-
challenged uremic animals. Blood Purif 2002;20:55–60.
24 Fissell WH, Lou L, Abrishami S, Buffington DA, Humes HD: Bioartificial kidney ameliorates
gram-negative bacteria-induced septic shock in uremic animals. J Am Soc Nephrol 2003;14:
454–461.
25 Humes HD, Buffington DA, Lou L, Abrishami S, Wang M, Xia J, Fissell WH: Cell therapy with a
tissue-engineered kidney reduces the multiple-organ consequences of septic shock. Crit Care Med
2003;31:2421–2428.
RAD and Treatment of Sepsis-Induced AKI 427
26 Humes HD, Weitzel WF, Bartlett RH, Swaniker FC, Paganini EP: Renal cell therapy is associated
with dynamic and individualized responses in patients with acute renal failure. Blood Purif
2003;21:64–71.
27 Tumlin J, Wali R, Brennan K, Humes HD: Effect of the renal assist device (RAD) on mortality of
dialysis-dependent acute renal failure: a randomized, open-labeled, multicenter, phase II trial
(abstract). American Society of Nephrology (ASN) 38th Annual Meeting, Philadelphia, 2005.
Emil P. Paganini
Head, Section of Dialysis and Extracorporeal Therapy
Department of Nephrology and Hypertension Cleveland Clinic
9500 Euclid Avenue, M82
Cleveland, OH 44195 (USA)
Tel. ⫹1 216 444 5792, Fax ⫹1 216 444 7577, E-Mail paganie@ccf.org
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 428–433
Renal Replacement Therapy in Neonates
with Congenital Heart Disease
Stefano Morelli, Zaccaria Ricci, Luca Di Chiara, Giulia V. Stazi,
Angelo Polito, Vincenzo Vitale, Chiara Giorni, Claudia Iacoella, Sergio Picardo
Department of Pediatric Cardiology and Cardiac Surgery, Bambino Gesù Hospital,
Rome, Italy
Abstract
Background: The acute renal failure (ARF) incidence in pediatric cardiac surgery
intensive care unit (ICU) ranges from 5 to 20% of patients. In particular, clinical features of
neonatal ARF are mostly represented by fluid retention, anasarca and only slight creatinine
increase; this is the reason why medical strategies to prevent and manage ARF have limited
efficacy and early optimization of renal replacement therapy (RRT) plays a key role in the
outcome of cardiopathic patients. Methods: Data on neonates admitted to our ICU were
prospectively collected over a 6-month period and analysis of patients with ARF analyzed.
Indications for RRT were oligoanuria (urine output less than 0.5 ml/kg/h for more than 4 h)
and/or a need for additional ultrafiltration in edematous patients despite aggressive diuretic
therapy. Results: Incidence of ARF and need for RRT were equivalent and occurred in 10%
of admitted neonates. Eleven patients of 12 were treated by peritoneal dialysis (PD) as only
RRT strategy. PD allowed ultrafiltration to range between 5 and 20 ml/h with a negative bal-
ance of up to 200 ml over 24 h. Creatinine clearance achieved by PD ranged from 2 to
10 ml/min/1.73 m
2
. We reported a 16% mortality in RRT patients. Conclusion: PD is a safe
and adequate strategy to support ARF in neonates with congenital heart disease. Fluid
balance control is easily optimized by this therapy whereas solute control reaches acceptable
levels.
Copyright © 2007 S. Karger AG, Basel
The incidence of acute renal failure (ARF) in the neonatal intensive care
unit (ICU) ranges between 5 and 20% of patients admitted [1, 2]. It is typically
classified as prerenal disease, intrinsic renal disease (including vascular
insults), and obstructive uropathy [3]. ARF in the newborn commonly occurs in
the postnatal period because of hypoxic ischemic injury and toxic insults.
Nephrotoxic ARF in newborns is usually associated with aminoglycoside
RRT in Neonates with CHD 429
antibiotics and nonsteroidal anti-inflammatory medications used to close a
patent ductus arteriosus. Such alterations are usually reversible. Neonates with
congenital heart disease are a particular subset of critically ill patients at high
risk of ARF [2]. Low cardiac output and hypoxic/ischemic/nephrotoxic insults
can induce prerenal/intrinsic kidney injury in the early phase of ICU admission.
Neonates who undergo cardiac surgery for correction or palliation of congeni-
tal heart disease are exposed to additional risk factors for ARF. Once intrinsic
renal failure has become established, management of the metabolic complica-
tions of ARF requires appropriate management of fluid balance, electrolyte sta-
tus, acid-base balance, nutrition and, when appropriate, the initiation of renal
replacement therapy (RRT). Renal replacement may be provided by peritoneal
dialysis (PD), intermittent hemodialysis, or continuous hemofiltration. PD is
the preferred modality of therapy for ARF in the neonate since it is relatively
easy to perform, it does not require heparinization, and it can be safely admin-
istered to a hemodynamically unstable patient [1, 3].
The present chapter will discuss the epidemiology of RRT, its indication
and prescription in neonates admitted to our cardiologic intensive care unit
(CICU) with particular attention to several aspects of PD management and
administration.
Patients
Between June and December 2006 we prospectively collected clinical data
on neonates admitted to the CICU of our department, who represented 31% of
the overall patients (114 over 367). Data are presented as median and interquar-
tile range (IQR). Tests for nonparametric data were performed when necessary.
A p value ⬍0.05 was considered significant. The weight of the patients was
2.9 kg (1.8–3.9). Forty-one (36%) of these patients were premature. Eighty-two
(72%) of them underwent elective and 22 (19%) urgent surgery. Nine percent of
the admitted patents did not require any surgical procedure. Twelve patients
(10%) required RRT. Six of these patients had aortic arch obstruction (coarctation/
interruption); other children had miscellaneous diagnoses (2 cases of anom-
alous pulmonary venous return, 1 transposition of the great arteries, 1 double
outlet right ventricle and 2 cases of right ventricular obstruction). Indications
for renal support were oligoanuria (urine output less than 0.5 ml/kg/h for more
than 4 h; 8–66%), need for additional ultrafiltration in edematous patients
(2–16%) or both (2–16%), despite aggressive diuretic therapy (continuous
furosemide infusion at a dose of 10 mg/kg/day). Four patients (3%) required
preoperative RRT, whereas 9 (8%) were treated after surgery; 1 patient received
RRT before and after intervention. Eleven patients were treated by PD. One
Morelli/Ricci/Di Chiara/Stazi/Polito/Vitale/Giorni/Iacoella/Picardo 430
patient was administered predilution hemofiltration during postoperative extra-
corporeal membrane oxygenation. Indications for stopping RRT included
return of sufficient urine output to maintain or achieve negative fluid balance
and normalization of serum electrolytes and acid-base status. When PD was
prescribed, access to the peritoneal cavity was obtained through a periumbilical
Tenckhoff catheter. Commercially available 1.36 and 2.7% glucose solutions
were utilized. Solutions were lactate buffered. An exchange volume of 10 ml/kg
was infused and left to dwell for 10–15 min, then drained for 10–15 min before
the procedure was repeated; this technique allowed a median (IQR) exchange
volume of 64 (40–100) ml/h and 1,500–2,000 liters/day (about 25 ml/kg/h and
600 ml/kg/day). Creatinine clearance achieved by PD ranged from 2 to
10 ml/min/1.73 m
2
. Median (IQR) net ultrafiltration was 3 (1.5–6) ml/kg/h.
This management allowed a median (IQR) negative balance of 15 (10–60)
ml/kg/24 h. PD institution never induced hemodynamic instability neither did it
trigger inotropic infusion increase. Transient hyperglycemia (serum glucose
over 150 mg/dl) was observed on the first treatment day in 6 patients: this con-
dition was generally not observed after the first 24 h and did not require any
treatment. Persistent hyperlactatemia due to lactate-buffered PD solutions was
present in 2 patients: these neonates were already hyperlactatemic at PD institu-
tion. Peritonitis was never observed. Catheter site induration and/or leakage
from catheter insertion were observed in 2 patients but did not require any spe-
cific intervention.
Median (IQR) duration of RRT was 4 (3–8) days and in 11 (91%) patients
full renal function had recovered after 8 days from ARF onset. Median (IQR)
CICU length of stay among RRT patients was 24 (15–58) days, which was sig-
nificantly higher than in other neonates [4 (6–11); p ⬍ 0.05]. Ten patients were
transferred to the cardiology ward and 2 died. Mortality among RRT patients
(16%) was significantly higher than among other neonates (8%; p ⬍ 0.05).
Discussion
Surgical interventions in case of complex congenital cardiac pathology are
often performed in the neonatal period. With the use of improved cardiopul-
monary bypass systems, new treatment modalities in preoperative, periopera-
tive, and postoperative patient care and follow-up, morbidity and mortality rates
have decreased [4]. There is also evidence that the neonatal kidney is more vul-
nerable to conditions of hemodynamic stress, with loss of autoregulation lead-
ing to blood-pressure-dependent renal blood flow and ischemia-induced renal
injury. All of these conditions render the neonate more prone to complications
of ischemia than the older infant or child [5, 6].
RRT in Neonates with CHD 431
Hence, perioperative care of the neonate with congenital heart disease is
challenging: organ failure should be avoided and, when it occurs, failing organs
should be effectively replaced. In particular, clinical features of neonatal ARF
are essentially represented by fluid retention, anasarca and only slight creati-
nine increase. In this case, the fluid balance must be optimized and acid-base/
electrolyte equilibrium achieved. Medical interventions are often disappointing
either for prevention or for the therapy of established ARF [7]. Neonatal RRT is
mostly performed by PD [1, 3]. In the adult setting renal replacement is gener-
ally achieved by extracorporeal techniques, such as dialysis and hemofiltration,
whether continuous or intermittent. These techniques have reached a good stan-
dard of care, specific practice patterns, dedicated technology and relatively
high levels of consensus to the point that extracorporeal RRT is administered
worldwide to 99% of adult ARF patient, PD being limited to underdeveloped
countries [8]. Nonetheless, extracorporeal RRT requires an adequate vascular
access, the heparinization of the patient and a relative amount of extracorporeal
circulating blood volume; all these aspects identify PD as by far the ideal RRT
in the neonate patient with ARF [1].
PD is an RRT modality where solutes and water are transported across a
membrane that separates two compartments: the blood in the peritoneal capillaries
and the dialysis solution in the peritoneal cavity, which is rendered hyperosmolar
by a high concentration of glucose. Three transport processes are simultaneously
involved during PD: uremic solutes and potassium diffuse from the peritoneal cap-
illary blood into the PD solution whereas glucose, lactate and calcium diffuse in
the opposite direction; simultaneously, the relative hyperosmolarity of dialytic
solution leads to ultrafiltration of water (and associated solutes) across the mem-
brane; finally, water and solutes are absorbed into the lymphatic system [9].
PD prescriptions during neonatal RRT generally tend to involve short
dwell times (10–15 min) with relatively high exchange volumes (10–15ml/kg/h)
[10]. This technique enhances solute diffusion from blood to dialysate solution
because a high concentration gradient is constantly maintained between the
solutions. Ultrafiltration is optimized for the same reasons. Nonetheless, one of
the main disadvantages of PD in the setting of adult renal disease is a relative
lack of efficiency especially when the treatment of a highly catabolic patient is
required. It must be said that during neonatal kidney dysfunction this is not
often the case, and that serum creatinine levels are generally maintained below
1 mg/dl even in oligoanuric children. A long debate has been ongoing in recent
years about the beneficial effect of removing inflammatory mediators through
RRT [11]. This issue seems of outstanding importance in postoperative patients,
and some authors have measured significant levels of proinflammatory mole-
cules and cytokines on peritoneal drainage after cardiopulmonary bypass in
neonates [12]. Several studies, however, noted a statistical difference in the
Morelli/Ricci/Di Chiara/Stazi/Polito/Vitale/Giorni/Iacoella/Picardo 432
percentage of fluid overload of children with severe renal dysfunction requiring
RRT: at the time of dialysis initiation, survivors tended to have less fluid over-
load than nonsurvivors, especially in the setting of the multiorgan dysfunction
syndrome [13–15]. Prevention of volume overload prompted some authors to
deliver postoperative prophylactic PD in neonates and infants after complex
congenital cardiac surgery [16].
We acknowledged this aspect of neonatal kidney dysfunction to the point
that our ARF population substantially corresponded to patients who received
PD: urine output and fluid balance needs triggered our intervention rather than
serum creatinine levels. This led to a relatively high incidence of ARF/PD
(10%), but also to early and timely treatments: after 4 h of oligoanuria PD catheter
is inserted if not already present and RRT is started without the need for
nephrologic counseling or dedicated staff. These PD schedules allowed ultrafil-
tration to range between 5 and 20 ml/h with a negative balance of up to 200ml
over 24 h. Creatinine levels at the stop of PD were significantly lower than ini-
tial values and they never exceeded 1 mg/dl (data not shown); solute control was
presumably adequate with this strategy. We reported a 16% mortality in PD
patients and, even if significantly higher than in non-RRT patients, it appeared
lower than that reported by other authors (20–70%) [3, 17].
Conclusions
PD is a safe and adequate renal replacement technique to support ARF in
neonates with congenital heart disease. Fluid balance control is easily opti-
mized by this therapy, whereas solute control only reaches acceptable levels.
New technology has recently been made available for diuretic resistant heart
failure: this kind of miniaturized, highly accurate, slow efficiency ultrafiltration
device that utilizes peripheral vascular access and allows relatively high ultra-
filtration rates could be promising and so be adopted by pediatric critical care
nephrology and significantly impact future strategies in pediatric RRT [18].
References
1 Andreoli SP: Acute renal failure in the newborn. Semin Perinatol 2004;28:112–123.
2 Gouyon JB, Guignard JP: Management of acute renal failure in newborns. Pediatr Nephrol
2000;14:1037–1044.
3 Moghal NE, Embleton ND: Management of acute renal failure in the newborn. Semin Fetal
Neonatal Med 2006;11:207–213.
4 Feltes TF: Postoperative recovery of congenital heart disease; in Garson A, Bricker JT, Fisher DJ,
Neish SR (eds): The Science and Practice of Pediatric Cardiology, ed 2. Baltimore, Williams &
Wilkins, 1997.
RRT in Neonates with CHD 433
5 Vanpee M, Blennow M, Linne T, et al: Renal function in very low birth weight infants: normal
maturity reached during childhood. J Pediatr 1992;121:784–788.
6 Drukker A, Guignard JP: Renal aspects of the term and preterm infant: a selective update. Curr
Opin Pediatr 2002;14:175–182.
7 Kellum JA: What can be done about ARF. Minerva Anestesiol 2004;70:181–188.
8 Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman
C, Macedo E, Gibney N, Tolwani A, Ronco C; Beginning and Ending Supportive Therapy for the
Kidney (BEST Kidney) Investigators: Acute renal failure in critically ill patients: a multinational,
multicenter study. JAMA 2005;294:813–818.
9 Korbet SM, Kronfol NO: Acute peritoneal dialysis prescription; in Daugirdas JT, Blake PG, Ing
TS (eds): Handbook of Dialysis, ed 3. Lippincott, Williams & Wilkins, 2001.
10 McNiece KL, Ellis EE, Drummond-Webb JJ, Fontenot EE, O’Grady CM, Blaszak RT: Adequacy
of peritoneal dialysis in children following cardiopulmonary bypass surgery. Pediatr Nephrol
2005;20:972–976.
11 Venkataraman R, Subramanian S, Kellum JA: Clinical review: extracorporeal blood purification
in severe sepsis. Crit Care 2003;7:139–145.
12 Bokesch PM, Kapural MB, Mossad EB, Cavaglia M, Appachi E, Drummond-Webb JJ, Mee RBB:
Do peritoneal catheters remove pro-inflammatory cytokines after cardiopulmonary bypass in
neonates? Ann Thorac Surg 2000;70:639–643.
13 Goldstein SL, Currier H, Graf JM, et al: Outcome in children receiving continuous veno-venous
hemofiltration. Pediatrics 2001;107:1309–1312.
14 Goldstein SL, Somers MJ, Baum MA, et al: Pediatric patients with multi-organ dysfunction syn-
drome receiving continuous renal replacement therapy. Kidney Int 2005;67:653–658.
15 Foland JA, Fortenberry JD, Warshaw BL, et al: Fluid overload before continuous hemofiltration
and survival in critically ill children: a retrospective analysis. Crit Care Med 2004;32:1771–1776.
16 Alkan T, Akcevin A, Turkoglu H, Paker T, Sasmazel A, Bayer, Ersoy C, Askn D, Aytac A:
Postoperative prophylactic peritoneal dialysis in neonates and infants after complex congenital
cardiac surgery. ASAIO J 2006;52:693–697.
17 Sorof JM, Stromberg D, Brewer ED, Feltes TF, Fraser CD: Early initiation of peritoneal dialysis
after surgical repair of congenital heart disease. Pediatr Nephrol 1999;13:641–645.
18 Liang KV, Hiniker AR, Williams AW, Karon BL, Greene EL, Redfield MM: Use of a novel ultra-
filtration device as a treatment strategy for diuretic resistant, refractory heart failure: initial clini-
cal experience in a single center. J Card Fail 2006;9:707–714.
Zaccaria Ricci
Department of Pediatric Cardiology and Cardiac Surgery
Bambino Gesù Hospital
Piazza S. Onofrio
IT–00100 Rome (Italy)
Tel. ⫹39 06 6859 3333, E-Mail z.ricci@libero.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 434–443
The DOse REsponse Multicentre
International Collaborative Initiative
(DO-RE-MI)
1
G. Monti
a
, M. Herrera
g
, D. Kindgen-Milles
h
, A. Marinho
i
, D. Cruz
b
,
F. Mariano
c
, G. Gigliola
d
, E. Moretti
e
, E. Alessandri
f
, R. Robert
j
, C. Ronco
b
a
Department of Anesthesiology and Intensive Care, Hospital Niguarda, Milan,
b
Department of Nephrology, Hospital San Bortolo, Vicenza,
c
Department of
Nephrology and Dialysis Unit, CTO Hospital, Turin,
d
Department of Nephrology,
Hospital Santa Croce e Carle, Cuneo,
e
Department of Anesthesiology and Intensive
Care, Hospital Riuniti di Bergamo, Bergamo, and
f
Department of Anesthesiology and
Intensive Care, Umberto I Hospital, Rome, Italy;
g
Regional Hospital, Malaga, Spain;
h
Anesthesiology Clinic, University of Düsseldorf, Düsseldorf, Germany;
i
Anesthesiology and Intensive Care Unit, Hospital Geral Sant Antonio, Porto,
Portugal;
j
I Department of Intensive Care, University of Poitier, Poitier, France
Abstract
Background: Current practices for renal replacement therapy (RRT) in ICU remain
poorly defined. The observational DOse REsponse Multicentre International collaborative
initiative (DO-RE-MI) survey addresses the issue of how the different modes of RRT are cur-
rently chosen and performed. The primary endpoint of DO-RE-MI will be the delivered dose
versus in ICU, 28-day, and hospital mortality, and the secondary endpoint, the hemodynamic
response to RRT. Here, we report the first preliminary descriptive analysis after 1-year
recruitment. Methods: Data from 431 patients in need of RRT with or without acute renal
failure (mean age 61.2 15.9) from 25 centers in 5 countries (Spain, Italy, Germany,
Portugal, France) were entered in electronic case report forms (CRFs) available via the web-
site acutevision.net. Results: On admission, 51% patients came from surgery, 36% from the
New Trials and Meta-Analyses
1
Scientific Committee: Germany: D. Kindgen-Milles; France: D. Journois (Paris),
R. Robert (Poitiers); Italy: R. Fumagalli (Milan), C. Ronco (Vicenza), S. Vesconi (Milan);
Spain: J. Maynar (Vitoria); Portugal: A. Marinho (Porto); Steering Committee: Germany:
J.A. Amman (Düsseldorf); Italy: A. Brendolan (Vicenza), G. Monti (Milano), M. Formica
(Turin), F. Mariano (Turin), M. Marchesi (Bergamo), S. Livigni (Turin), M. Maio (Turin),
D. Silengo (Turin).
DO-RE-MI 435
emergency department, and 16% from internal medicine. On admission, mean SOFA and
SAPS II were 13 and 50, respectively. The first criteria to initiate RRT was the RIFLE in 38%
(failure: 70%, injury: 25%, risk: 22%), the second the high urea/creatinine, and the third
immunomodulation. A total of 3,010 cumulative CRF were reported: continuous venovenous
hemodiafiltration (CVVHDF) 60%, continuous venovenous hemofiltration (CVVH) 15%,
intermittent hemodialysis (IHD) 15%, high-volume hemofiltration (HVHF) 7%, continuous
venovenous hemodialysis (CVVHD) 1%, and coupled plasma filtration adsorption/CVVD
2%. In 15% of cases, the patient was shifted to another modality. Mean blood flow rates
(ml/min) in the different modalities were: 145 (CVVHDF), 200 (CVVH), 215 (IHD), 283
(HVHF), and 150 (CVVHD). Downtime ranged from 8 to 28% of the total treatment time.
Clotting of the circuit accounted for 74% of treatment interruptions. Conclusions: Despite a
large variability in the criteria of choice of RRT, CVVHDF remains the most used (49%).
Clotting and clinical reasons were the most common causes for RRT downtime. In continu-
ous RRT, a large variability in the delivered dose is observed in the majority of patients and
often in the same patient from one day to another. Preliminary analysis suggests that in a
large number of cases the delivered dose is far from the ‘adequate’ 35 ml/h/kg.
Copyright © 2007 S. Karger AG, Basel
Various continuous and intermittent modalities of renal replacement ther-
apy (RRT) are currently used. In recent years remarkable advances in continu-
ous RRT (CRRT) technology have been made, driven by nephrologists
dedicated to improving efficiency and function. Today, however, intensivists are
Participating Doctors/Centers: Belgium: P. Honoré (Saint-Pierre Para-University
Hospital, Ottignies-Louvain-La-Neuve); France: R. Robert (Hôpital Jean Bernard, CHU,
Poitiers), D. Journois (Hôpital European Georges Pompidou, Paris), V. Labiotte (Hôpital de
Lens), B. Thievenin (Hôpital de Maubeuge), O. Joannes Boyau (University Hospital of
Bordeaux; Germany: A. Amman, D. Kindgen-Milles (University of Düsseldorf, Düsseldorf);
Portugal: A. Marinho (Hospital Geral Sant Antonio, Porto), A. Lafuente (Hospital de
Penafiel, Penafiel), A. Santos (Hospital Geral Sant Antonio, Porto); Italy: E. Moretti
(Ospedale Riuniti di Bergamo, Bergamo), M. Cerisara (Ospedale Maggiore, Crema), P.
Inguaggiato, G. Gigliola (Ospedale Santa Croce e Carle, Cuneo), W. Morandini (Ospedale
Valle Camonica, Esine), R. Fumagalli, R. Rona (University of Milan, Ospedale S. Gerardo,
Monza), A. Sicignano (Policlinico di Milano, Milan), G.P. Monti, S. Vesconi (Ospedale
Niguarda, Milan), G. Slaviero (IRCCS San Raffaele, Milan), F. Mariano, L. Tedeschi (CTO
Ospedale, Turin), S. Livigni, M. Maio (Ospedale G. Bosco, Turin), Z. Ricci, E. Alessandri
(Policlinico Umberto 1, Rome), A. Brendolan, D. Cruz (Ospedale San Bortolo, Vicenza), G.
Marchesi (Ospedale Bolognini di Seriate, Seriate); Spain: J. Maynar (Hopital de Vitoria),
Teresa Doñate, A. Leon (Hopital Gral. De Catalunya, Sant Cugat del Vallés), M. Herrera
(Hospital Carlos Haya, Malaga), F. Labayen, J. Maynar (Hospital Santiago Apostol, Vitoria),
Á. Montero, J. Sánchez-Izquierdo (Hospital 12 De Octubre, Madrid), J. Luño, E. Junco
(Hospital Gregorio Marañon, Madrid), J.A. Sánchez Tomero, C. Bernix (Hospital De La
Princesa, Madrid), J. Bustos (Hospital Virgen De La Salud, Toledo), J. Cruz, J. Moll (Hospital
La Fe, Valencia), R. Cabadas (Policlinico De Vigo, Vigo).
Monti/Herrera/Kindgen-Milles/Marinho/Cruz/Mariano/Gigliola/Moretti/Alessandri/ 436
Robert/Ronco
the most familiar with these techniques. Nevertheless, in some countries such
as the USA, CRRT is still infrequently employed [1]. Other modalities include
intermittent hemodialysis (IHD), slow extended daily dialysis [2], or daily
hemodialysis [3]. Some of the reasons for the considerable variability world-
wide in extracorporeal treatment of acute renal failure (ARF) include local
practice (e.g. whether management is by nephrologists or intensivists), the cen-
ter’s experience with the various techniques, organization and health resources.
Various methods of extracorporeal treatment, whether intermittent or continu-
ous, are currently being employed and no guidelines exist. This variability was
highlighted in the Beginning and Ending of Supportive Therapy for the Kidney
(BEST Kidney) trial, which collected data on ARF management in 1,743
patients in 54 ICU from 23 countries worldwide [4].
The practice of CRRT has apparently not changed, even following the
prospective studies conducted by Ronco et al. [5]. Despite the positive find-
ings of that prospective trial, the practice of a higher intensity CRRT has not
been widely adopted into routine ICU practice. The most outstanding exam-
ples are Australia and New Zealand, where almost 100% of treatments are
CRRT. A survey of several units active in the Australian and New Zealand
Intensive Care Society Clinical Trials Group (Bellomo, unpubl. data, 2002)
found that very few units had adopted the intensive CRRT regimen proposed
by Ronco and coworkers [4]. Data from such Australian units show instead that
the vast majority (90%) prescribe a ‘fixed’ standard CRRT dose of 2 l/h,
which is not adjusted for body weight. Thus, a 100-kg man would receive
20 ml/kg h – the dose shown to have the worst outcome in the study by Ronco
and coworkers [4]. In another recent study that involved several Australian
units (the BEST Kidney study), the median body weight for Australian patients
was 80 kg, thus indicating that the vast majority receive a CRRT intensity of
approximately 25 ml/kg h of effluent. Finally, although in the study con-
ducted by Ronco and colleagues [4] the technique of CRRT was uniformly in
the form of continuous venovenous hemofiltration (CVVH) with postfilter
fluid replacement, current practice includes a variety of techniques in addition
to CVVH, such as continuous venovenous hemodiafiltration (CVVHDF).
Furthermore, scarce information exists on the practice of CRRT in Europe,
particularly regarding the actually delivered dose of therapy in critically ill
patients with ARF (i.e. in those who could potentially derive more benefit
from high-volume convective therapy).
In a recent preliminary collaborative study [6] we reported that there was
no significant difference between prescribed and delivered ultrafiltration rate
(both in ml/min and in l/h), which was related to the reduced downtime associ-
ated with the technique. However, of greater relevance is that the dose of dialy-
sis was over 40 ml/kg h.
DO-RE-MI 437
If we are to understand how dialysis doses are actually delivered in rou-
tine clinical practice in ICUs, an observational clinical study is needed to con-
firm how, to what extent and with what clinical indication the different
modalities of RRT are administered. With this in mind we initiated the DOse
REsponse Multicentre International collaborative initiative (DO-RE-MI) sur-
vey. The protocol was published [7]. The survey was listed as CRG110600093
in the Cochrane Renal Group. The primary endpoint of DO-RE-MI is mortal-
ity (ICU mortality, 28-day mortality and hospital mortality), and the secondary
endpoint is the hemodynamic response to RRT, expressed as percentage reduc-
tion in noradrenaline (norepinephrine) requirement to maintain blood pressure.
Here, we present the preliminary descriptive results from the DO-RE-MI sur-
vey. The survey started on June 1, 2005 and will be terminated in December
2007.
Materials and Methods
All data from incident patients admitted to the ICU in need of RRT with or without
ARF are entered into electronic case report forms (CRFs) and downloaded via the internet
onto a server [8, 11]. The following rules are applied without exception:
First, all patient data are entered anonymously. To this aim, each center has a code, and
patients are consecutively assigned a progressive number. Under no circumstances is there
any written or oral transmission of data that may make the identification of the patient possi-
ble. Failure to adhere to this is immediately followed by cancellation of the data from the
website by the webmaster.
Second, data for each patient are entered into a separate CRF. These data may be copied
from paper CRFs in order to make the reporting of data from bed to computer station easier.
All fields may be amended at any time until the patient’s CRF is completed and closed. At
this point, one may access the patient’s CRF but it is no longer possible to amend it. In the
case of overt inconsistency, corrections must be detailed in writing (E-Mail) by the person
responsible for the data quality of the center. In no cases are corrections permitted in the
absence of an express written request. The person responsible for data quality will have
access to his or her center’s CRF. A registry collects the correspondence between the person
responsible for data quality and the center.
Third, completion of some fields in the CRF is mandatory. Failure to complete them
prevents progression to the following CRF and closure of the opened CRF. Failure to com-
plete a CRF electronically results in the patient being excluded from the study.
Finally, each center is enabled to open CRFs of its own patients but under no circum-
stances the CRFs of patients from other centers.
Definitions
‘Treatment interruption’ is defined as when the treatment is stopped and resumed
within 18 h. In the case of treatment interruption, the CRF is continued and the treatment that
follows is considered in the context of the preceding one. The only exception is when, after
RRT interruption, the modality is changed.
Monti/Herrera/Kindgen-Milles/Marinho/Cruz/Mariano/Gigliola/Moretti/Alessandri/ 438
Robert/Ronco
‘Treatment end’ is defined as when a given RRT is stopped because of clinical or
other factors for more than 12 h or when clinical or other factors have changed since the
start of RRT. Should the patient be started on another RRT, then the latter is considered a
new one.
In case the modality is changed, a new CRF is to be filled in. This is followed by a new
CRF.
Each center is asked to define the clinical/practical reasons for changing a modality.
The change of modality may be necessary after treatment is interrupted. In this case, the fol-
lowing treatment is considered a new treatment. This CRF aims to provide information on
why the modality was chosen. It is similar to the CRF.
Results
In the first year, 434 patients were recruited from a total of 37 centers
(15 in Italy, 14 in Spain, 5 in Portugal, 2 in France, 1 in Germany; mean age:
61.2 15.9 years, mean weight: 68.9 18.2kg). The most common diagnosis
for admission was septic shock. However, the assignment to admission diagnosis
is still fraught by the highest percent of missing data (56%). On admission,
patients had serum creatinine between 1 and 2 mg%, and in 16% creatinine
values were over 4 mg% with a clearcut difference between male and female
patients (fig. 1). The most common indication for initiating treatment was high
0
5
10
15
20
25
30
35
%
0.5 0.5 x 11 x 22 x 44 x 6 6
Male
Female
Missing
55
n
290
Mean
2.18
Min
0.3
Max
19.3
Serum creatinine values (mg%)
Fig. 1. Percentage of patients on admission categorized according to six ranges of
creatinine values.
DO-RE-MI 439
plasma urea/creatinine levels followed by one of the RIFLE criteria and by
hoping to achieve some kind of immunomodulation (fig. 2). The different cen-
ters remarkably differed as to the type of RRT of choice. In some centers, all
RRT modalities were available. In some cases, patients were started on high-vol-
ume hemofiltration (HVHF) to CVVH or CVHDF, in some others pulse HVHF
was initiated and finally some other centers were using almost only one type of
RRT modality either CVVHDF or IHD (fig. 3). A total of 3,010 cumulative CRF
were reported: CVVHDF 60%, CVVH 15%, IHD 15%, HVHF 7%, continuous
venovenous hemodialysis (CVVHD) 1%, and coupled plasma filtration adsorp-
tion/CVVD 2%. In 15% of cases, the patient was shifted to another modality.
Mean blood flow rates (ml/min) in the different modalities were: 145
(CVVHDF), 200 (CVVH), 215 (IHD), 283 (HVHF), and 150 (CVVHD). The
parameters to deliver RRT also appeared to differ among the different modalities
but evidently in the same modality among different centers (fig. 4). Downtime
was precisely calculated as it will allow to calculate the really delivered dose of
dialysis. Downtime ranged from 8 to 28% of the total treatment time. The major
causes of treatment interruption were clotting of the circuit in 74%, failure of the
vascular access in 11%, clinical reasons in 10%, and machine alarms in 2%. In
3% there was no specific reason. As shown in figure 5, the anticoagulation regi-
men remarkably differed in the different modalities of RRT.
1) Oliguria (urine output <200ml/12h)
2) High urea/creatinine
3) Anuria (urine output <50ml/12h)
4) What RIFLE criteria (specify as below)
5) Metabolic acidosis
6) Fluid overload
7) Immunomodulation
8) Hyperkalemia (>6.5 mmol/l or rapidly rising K
+
)
9) Hyperthermia (>41˚C)
10) Others (specify)
3rd1st2nd
0.140.360.941.10.631.660.831.590.68Mean score
435
435435435435435435435435Total
15341119841156107155743
7182860471131580292
4
2021665927965171
4093632752112881393041353150
9
87654321Score
156/43536%
155/43536%
111/43525%
Fig. 2. Criteria to initiate RRT as listed. Score 0 No priority; score 3 top priority.
Shaded fields indicate the top three criteria either as those with the highest values for score 3
(horizontal order) or as mean score (vertical order).
Monti/Herrera/Kindgen-Milles/Marinho/Cruz/Mariano/Gigliola/Moretti/Alessandri/ 440
Robert/Ronco
Discussion
The practice of CRRT has been the subject of much debate. Only a few
prospective randomized studies have been performed and published on the rela-
tionship between CRRT and outcome, and so conclusions are difficult to draw
[9, 10]. As emphasized in a recent editorial [11], in the field of artificial organs,
prospective observational studies, despite their inherent limitations, have been
performed because they are more affordable but are also capable of providing
useful information from a practical and medical standpoint.
The results of the present study were obtained from the analyses of the
CRFs and they are at the present time descriptive in nature. Calculation of the
actually delivered dose of dialysis is ongoing for each patient undergoing dif-
ferent modalities, taking into consideration the downtimes and the change of
modality if any. It is anticipated that the dose of dialysis is surprisingly much
lower than that which is recognized as being the minimally required one to
ensure enhanced survival in a randomized clinical trial [4]. The large variability
2%
2%
8%16%23%
49%
%
57264974154Total
145
1244
21236
129
127
2125
11224
119
117
238214
4413
1129
2258
52987
111176
911614
31073
2
184
29
2
315
1
3
5
1
1
3
13
1
1
87
8
13
9
42
20
27
20
44
16
161
CPFACVVHDHVHFIHDCVVHCVVHDFCenter No.
Fig. 3. Each center is listed by its identification number. Each column indicates the
number of treatments per each type of RRT modality. CPFA Coupled plasma filtration
adsorption.
DO-RE-MI 441
in current practices even in a relatively restricted geographical area allows us
already to make the hypothesis that RRT in ICU is still undefined not only con-
ceptually but also in very practical terms.
The present survey has the well-known limitations of any observational
study. First, due to the largely geographically dispersed sample of centers, the
information will not make it possible for us to sort out any practice more spe-
cific for any of the countries represented by the centers. Nevertheless, due to
the high number of patients and the design of the CRF [see 7], we feel that the
survey may throw lighten on how RRT is currently done in ICUs and should
hopefully provide tentative answers to as yet undefined questions such as: what
are the criteria for beginning and ending treatment?; what is the currently deliv-
ered dose of dialysis?; how is fluid control taken care of?; what schedules are
mostly used?; how is technology used (or not used)?, and finally what are the rea-
sons for downtime in RRT? Finally, the survey will probably show the ‘real world’
CPFA
35675217285
6%233%257%437%198%1312%10
Subclavian
catheter
20%70%29%229%155%918%15
Jugular
catheter
74%2667%414%135%1887%15071%60
Femoral
catheter
HVHFCPFACVVHDIHDCVVHDFCVVH
Vascular
access
Blood flow rates
Vascular access
0
0
53
0
100
208
Lactate
100%32CPFA
100%41CVVHD
70%125HVHF
100%291IHD
83%519CVVH
83%986CVVHDF
%BicBicarbonate
Buffer
1,20037.3146CVVHDF
4133.4175CVVHD
3251.9204
60945.1207CVVH
29146.8209IHD
17563.4250HVHF
nSDQ
B
Treatment
Fig. 4. Summary of data showing the use of the kind of buffer, the blood flow rate
and vascular access in each type of RRT modality. CPFA Coupled plasma filtration
adsorption.
Monti/Herrera/Kindgen-Milles/Marinho/Cruz/Mariano/Gigliola/Moretti/Alessandri/ 442
Robert/Ronco
in terms of the actually delivered dose outside the rigor of a randomized con-
trolled study. It will also imply that comparisons of therapies in observational
studies may be fraught by the large variability which might be at the origin of a
great background noise that possible disturbs the emergence of as yet unde-
fined advantages in outcome measures.
References
1 Mehta RL, Letteri JM: Current status of renal replacement therapy for acute renal failure. A sur-
vey of US nephrologists. The National Kidney Foundation Council on Dialysis. Am J Nephrol
1999;19:377–382.
2 Marshall MR, Golper TA, Shaver MJ, Alam MG, Chatoth DK: Sustained low-efficiency dialysis
for critically ill patients requiring renal replacement therapy. Kidney Int 2001;60:777–785.
3 Schiffl H, Lang SM, Fischer R: Daily hemodialysis and the outcome of acute renal failure. N Engl
J Med 2002;346:305–310.
4 Uchino S, Kellum JA, Bellomo R, et al: Acute renal failure in critically ill patients. A multi-
national, multicentre study. JAMA 2005;294:813–818.
5 Ronco C, Bellomo R, Homel P, Brendolan A, Dan M, Piccinni P, La Greca G: Effects of different
doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective
randomized trial. Lancet 2000;356:26–30.
6 Brendolan A, D’Intini V, Ricci Z, Bonello M, Ratanarat R, Salavtori G, Bordoni V, DeCal M,
Andrikos E, Ronco C: Pulse high volume haemofiltration. Int J Artif Organs 2004;27:398–403.
CVVH
None
48%
Heparin
45%
Other
7%
CVVHDF
Citrate
1%
Heparin
57%
None
23%
Other
17%
HVHF
None
60%
Heparin
37%
Other
3%
IHD
Heparin
48%
Other
29%
None 23%
Fig. 5. Anticoagulation regimen in the different RRT modalities.
DO-RE-MI 443
7 Kindgen-Milles D, Journois D, Fumagalli R, et al: Study protocol: the dose response Multicentre
International Collaborative Initiative (DO-RE-MI). Crit Care 2005;9:R396–R406.
8 Acute Vision. http://www.acutevision.it.
9 Metha R, McDonald B, Gabbai FB, Pahl M, Pascual MT, Farkas A, Kaplan RM; Collaborative
Group for Treatment of ARF in the ICU: A randomized clinical trial of continuous versus inter-
mittent dialysis for acute renal failure. Kidney Int 2001;60:1154–1163.
10 Tonelli M, Manns B, Feller-Kopman D: Acute renal failure in the intensive care unit: a systematic
review of the impact of dialytic modality on mortality and renal recovery. Am J Kidney Dis
2002;100:158–160.
11 Ronco C: Evidence-based medicine: can we afford it? Int J Artif Organs 2004;27:819–820.
Dr. Gianpaola Monti
Department of Anesthesiology and Intensive Care, Hospital Niguarda
IT–20162 Milan (Italy)
E-Mail Gianpaola.monti@ospedaleniguarda.it
Ronco C, Bellomo R, Kellum JA (eds): Acute Kidney Injury.
Contrib Nephrol. Basel, Karger, 2007, vol 156, pp 444–451
Clinical Effects of Polymyxin B-Immobilized
Fiber Column in Septic Patients
Dinna N. Cruz
a,b
, Rinaldo Bellomo
c
, Claudio Ronco
a
a
Department of Nephrology, Ospedale San Bortolo, Vicenza, Italy;
b
Section of
Nephrology, Department of Medicine, St. Luke’s Medical Center, Quezon City,
Philippines;
c
Department of Intensive Care, Austin Hospital, Melbourne, Vic., Australia
Abstract
Endotoxin is one of the principal biological substances that cause gram-negative septic
shock. Lack of clinical success with antiendotoxin or anticytokine therapy has shifted interest to
extracorporeal therapies to reduce circulating levels of the mediators of sepsis. Direct hemoper-
fusion with polymyxin-B-immobilized fiber (PMX-F) is a promising treatment of gram-
negative sepsis in critically ill patients. Because of the high affinity of polymyxin B for
endotoxin, the rationale underlying extracorporeal therapy would be to remove circulating endo-
toxin by adsorption, thus preventing progression of the biological cascade of sepsis. In a system-
atic review of 28 studies (pooled sample size 1,390 patients), the preliminary results of which are
described here, PMX-F therapy appeared to significantly lower endotoxin levels, improve blood
pressure, and reduce mortality. However, publication bias and lack of blinding need to be con-
sidered. These encouraging results need to be verified with large-scale controlled clinical trials.
Copyright © 2007 S. Karger AG, Basel
Sepsis is a common problem encountered in the hospital population and is
responsible for spending a large proportion of hospital resources. It involves a
complex interaction between bacterial factors and the host immune system pro-
ducing a systemic inflammatory response. When uncontrolled, it is associated
with multiple organ failure and high mortality rates.
Endotoxin, a lipopolysaccharide released from gram-negative bacteria, has
been implicated as a potent, prototypical stimulus of the immune response to
bacterial infection [1]. It causes the release of cytokines such as interleukin-1
and TNF-␣, activates complements and coagulation factors, and is an ideal
potential therapeutic target to treat septic shock. However, antiendotoxin drug
therapies, such as monoclonal antibodies or lipopolysaccharide-neutralizing
Polymyxin B-Immobilized Fiber Column in Sepsis 445
proteins, failed to demonstrate a clinical benefit in trials [1]. Polymyxin B is a
cationic cyclic polypeptide antibiotic which binds with high affinity to endo-
toxin, neutralizing its effects. Unfortunately, clinical use is limited by signifi-
cant nephro- and neurotoxicity.
Lack of clinical success with antiendotoxin or anticytokine therapy has
shifted interest to extracorporeal therapies to reduce circulating levels of the
mediators of sepsis. To overcome the toxicity issues of systemic administration
of polymyxin B yet take advantage of its ability to neutralize lipopolysaccha-
ride, polymyxin B bound and immobilized to polystyrene fibers (PMX-F) was
developed [2]. The resulting cartridge can be used in direct hemoperfusion, the
rationale being the removal of circulating endotoxin by adsorption, thus pre-
venting progression of the biological cascade of sepsis.
PMX-F has been shown to bind and neutralize endotoxin in both in vitro
and in vivo studies [1, 2]. An improvement in survival in both murine and
canine models of endotoxic shock has also been demonstrated, especially when
combined with antibiotic therapy [1]. Clinical trials, both randomized and
observational, have reported encouraging results but are limited by small sam-
ple size. To assimilate the published clinical experience with PMX-F a system-
atic review of 28 published studies was conducted [3]. Some preliminary results
of this review are presented here (table 1).
Effects on Endotoxin Level
Summary data from 19 studies [9 randomized controlled trials (RCT),
10 observational] confirmed that PMX-F was able to reduce levels of circulating
endotoxin [4–22]. Overall, endotoxin levels decreased by 22 pg/ml (95% CI
18.1–25.8 pg/ml), representing a decrease of 33–80% from baseline levels [3].
Studies which enrolled patients with higher baseline endotoxin levels tended to
Table 1. Summary of the results of a systematic review of direct hemoperfusion with
PMX-F column
Parameter Studies Patients Effect after Statistically
n n PMX-F significant
Blood pressure 12 275 ↑ yes
Dopamine/dobutamine dose 4 96 ↓ yes
PaO
2
/FiO
2
ratio 7 151 ↑ yes
Endotoxin level 17 435 ↓ yes
Mortality 15 885 ↓ yes
Cruz/Bellomo/Ronco 446
show a greater reduction after PMX-F. Work by Uriu et al. [20] suggests that
reduction in blood endotoxin concentration by PMX-F therapy positively corre-
lated with the reduction in cardiac output. In a European multicenter study, signif-
icant improvements in cardiac index and left ventricular stroke work were noted
after treatment with PMX [21]. This change was not accompanied by a change in
heart rate or in pulmonary capillary wedge pressure, and can therefore be inter-
preted as reflecting an increase in stroke volume. The clinical effects of endotoxin
removal, specifically on hemodynamics, were therefore examined in further detail.
Effects on Blood Pressure
Mean arterial pressure (MAP) data were reported in 12 studies (2 RCT,
10 observational) [11, 14, 17–19, 21–27], while systolic blood pressure (SBP)
was reported in 2 studies [14, 28]. Baseline MAP ranged from 68 to 87 mm Hg
and SBP from 80 to 108 mm Hg. All the individual studies showed an impro-
vement in blood pressure. Overall, the MAP increased by 19 mm Hg (95% CI
15–22 mm Hg, p ⬍ 0.001) [3] and SBP increased by 24 mm Hg in both studies.
This represented an increase of approximately 27% from baseline values. For
MAP, effect sizes of RCTs were smaller than of nonrandomized studies. This is
consistent with the observation in meta-analyses that trials of lower quality also
tend to show larger treatment effects [29, 30]. Similar to the results with endo-
toxin, studies which enrolled patients with lower baseline blood pressure
demonstrated a bigger change after PMX-F therapy.
In critically ill patients, it is often difficult to interpret blood pressure in
isolation, as vasoactive agents can be manipulated to alter the blood pressure. In
4 studies (1 RCT, 3 observational), there was a trend toward a decrease in the
dose of dopamine or dobutamine after PMX-F [11, 18, 20, 22]. Overall, the
dose was decreased by 1.8 g/kg/min (95% CI 0.4–3.3 g/kg/min, p ⫽ 0.01).
In these studies, there was also an increase in mean MAP after PMX-F (range
16–28 mm Hg). Only two cohort studies reported noradrenaline doses [18, 27].
The mean noradrenaline dose decreased by 0.2–0.9 g/kg/min after PMX-F
treatment, while the MAP increased by 16–26 mm Hg. It has been reported that
left ventricular stroke work index (LVSWI) is less likely to be affected by
vasopressors than is blood pressure. Therefore, LVSWI serves as a good indi-
cator of improvement in circulation, without having to take the vasopressor
dose into consideration. Unfortunately there was insufficient data on this and
other hemodynamic parameters for meta-analysis. Individual studies have
demonstrated that PMX-F therapy elevated the stroke volume and improved
LVSWI in patients with septic shock [11, 21]. In fact, an improvement in
LVSWI within 24 h of PMX therapy appeared to predict survival [11].
Polymyxin B-Immobilized Fiber Column in Sepsis 447
Effects on Mortality
Data on mortality, variably reported as 14-day [17], 28-day [10, 11, 15, 21,
24, 31], 30-day [6, 25, 32], and 60-day mortality [8], were available from
11 studies. Mortality was reported but the length of follow-up was not clearly
stated in another 4 studies [4, 7, 9, 23]. Pooled mortality for the 15 studies
(8 RCT, 7 non-RCT, 885 patients) was 61.8% in the conventional therapy group
and 34.4% in the PMX-F group. Looking specifically at the 9 studies that
reported 28- to 30-day mortality (704 patients), PMX-F therapy appeared to
significantly reduce mortality compared with conventional medical therapy
(RR 0.54; 95% CI 0.43–0.68). However, this must be viewed in the context of
the mortality of the patients under standard medical management (table 2). The
overall mortality of 61.8% seen in the conventional therapy group was compa-
rable to that seen in a French multicenter study on moderate-dose corticosteroid
therapy (63%) [33] and a Brazilian study on protective ventilation (71%) [34],
but higher than that reported in a study on early goal-directed therapy (46.5%) [35]
and activated protein C (30.8%) [36]. Moreover, mortality in the conventional
therapy group within the various studies averaged 58% (range 0–88.6%), which
is higher than the predicted mortality based on APACHE II scores (mean
44.8%; range 9.7–63.9%). Whether the apparent beneficial effect of PMX-F is
due to a reduction in deaths related to endotoxin removal by the therapy itself or
Table 2. Selected studies on sepsis and its effect on mortality
Intervention Ref. No. Group APACHE II Mortality
score %
PMX-F 3 Standard 10.6–28 61.8
PMX-F 16.7–28.5 34.4
Low-dose steroids 33 Placebo SAPS II 63.0
57 ⫾ 19
Steroids SAPS II 53.0
60 ⫾ 19
Protective ventilation 34 Conventional 24.0 ⫾ 7.0 71.0
Protective 24.0 ⫾ 6.0 38.0
EGDT 35 Standard 20.4 ⫾ 7.4 49.2
EGDT 21.4 ⫾ 6.9 33.3
APC 36 Placebo 25.0 ⫾ 7.8 30.8
APC 24.6 ⫾ 7.6 24.7
APC ⫽ Activated protein C; EGDT ⫽ early goal-directed therapy.
Cruz/Bellomo/Ronco 448
due to a higher than expected mortality in the conventional therapy group is not
clear. Nevertheless, these results are provocative, and should pave the way for
large-scale RCTs. The presence of publication bias was explored with a funnel
plot analysis. Interestingly, contrary to expectation, smaller studies (3 studies,
n ⫽ 17–35) had point estimates for mortality favoring standard medical therapy
(data not shown).
The efficacy of therapies for sepsis may be affected by the baseline sever-
ity of illness of the patients. For instance, activated protein C was deemed rela-
tively cost effective when targeted to patients with severe sepsis, greater
severity of illness (an APACHE II score of 25 or more), and a reasonable life
expectancy if they survive the episode of sepsis [37]. We attempted a supple-
mentary analysis of the pooled mortality data along these lines of thinking.
There were 9 studies (n ⫽ 678) in which the baseline APACHE scores of
enrolled patients were ⬍25, and 6 studies (n ⫽ 207) in which the baseline
APACHE scores were 25 or higher. There appeared to be a greater effect seen in
the sicker patients (RR 0.45, 95% CI 0.30–0.68) as opposed to the less severely
ill patients (RR 0.58, 95% CI 0.45–0.75). However, such exploratory analyses
are performed post hoc with the use of summary data, rather than patient level
data, and should therefore be interpreted with caution. At this point, these find-
ings remain hypothesis-generating rather than conclusive.
As with all systematic reviews, these findings are limited by the quality of
the primary studies. One third of the included studies were nonrandomized, and
the vast majority of published clinical experience comes from Japan, where
PMX-F has been in clinical use for over 10 years. Differences between intrinsic
patient characteristics and/or medical practices may limit the generalizability of
these results to a more heterogeneous septic population. In addition, because of
the small number of controlled studies, meta-analyses on blood pressure and
vasopressor dose were performed on data from single cohorts (i.e. pre- and
post-PMX-F), regardless of study design. With the use of single-arm studies,
there will tend to be a bias towards improvement since the data will tend to
overrepresent the survivors, particularly in a high-mortality disease such as sep-
sis. Nevertheless, this systematic review remains the most comprehensive sum-
mary to date of the clinical effects of direct hemoperfusion with PMX-F.
Conclusion
Polymyxin B binds endotoxin, one of the principal biological substances that
cause gram-negative septic shock, but has adverse nephro- and neurotoxin
effects. Direct hemoperfusion with PMX-F column would theoretically allow
removal of circulating endotoxin without systemic side effects. Based on
Polymyxin B-Immobilized Fiber Column in Sepsis 449
published literature, PMX-F therapy appears to effectively reduce endotoxin levels
and have some positive effects on blood pressure, use of vasoactive agents, and
mortality. Despite these encouraging results, randomized, controlled clinical trials
are necessary to definitively determine its efficacy as a form of therapy in sepsis.
Since the publication of this article, this work has been updated.
Acknowledgment
This work has been made possible by the International Society of Nephrology-funded
fellowship of Dr. Dinna Cruz, and was presented in part at the annual meeting of the
American Society of Nephrology, San Diego, Calif., November 2006.
References
1 Manocha S, Feinstein D, Kumar A, Kumar A: Novel therapies for sepsis: antiendotoxin therapies.
Expert Opin Investig Drugs 2002;11:1795–1812.
2 Shoji H: Extracorporeal endotoxin removal for the treatment of sepsis: endotoxin adsorption car-
tridge (Toraymyxin). Ther Apher Dial 2003;7:108–114.
3 Cruz D, Corradi V, Polanco N, Bellomo R, Ocampo C, de Cal M, Ronco C: Effectiveness of
polymyxin B-immobilized fiber column in sepsis: a metaanalysis (abstract). J Am Soc Nephrol
2006;17:768A.
4 Nakamura T, Ushiyama C, Sukuzi Y, Inoue T, Shoji H, Shimada N, Koide H: Combination therapy
with polymyxin B-immobilized fibre haemoperfusion and teicoplanin for sepsis due to methi-
cillin-resistant Staphylococcus aureus. J Hosp Infect 2003;53:58–63.
5 Nakamura T, Kawagoe Y, Matsuda T, Ueda Y, Koide H: Effects of polymyxin B-immobilized fiber
on urinary N-acetyl-B-glucosaminidase in patients with severe sepsis. ASAIO J 2004;50:
563–567.
6 Nakamura T, Ebihara I, Shoji H, Ishiyama C, Suzuki S, Koide H: Treatment with polymyxin
B-immobilized fiber reduces platelet activation in septic shock patients: decrease in plasma levels
of soluble P-selectin, platelet factor-4 and beta-thromboglobulin. Inflamm Res 1999;68:171–175.
7 Nakamura T, Ushiyama C, Suzuki Y, Shoji H, Shimada N, Koide H: Hemoperfusion with
polymyxin-B immobilized fiber for urinary albumin excretion in septic patients with trauma.
ASAIO J 2002;48:244–248.
8 Nakamura T, Ushiyama C, Suzuki Y, Osada S, Inoue T, Shoji H, Hara M, Shimada N, Koide H:
Hemoperfusion with polymyxin-B immobilized fiber in septic patients with methicillin-resistant
Staphylococcus aureus-associated glomerulonephritis. Nephron Clin Pract 2003;94:C33–C39.
9 Nakamura T, Kawagoe Y, Matsuda T, Koide H: Effect of polymyxin B-immobilized fiber on bone
resorption in patients with sepsis. Intensive Care Med 2004;30:1838–1841.
10 Nemoto H, Nakamoto H, Okada H, Sugahara S, Moriwaki K, Arai M, Kanno Y, Suzuki H: Newly
developed polymyxin B-immobilized fibers improve the survival of patients with sepsis. Blood
Purif 2001;19:361–369.
11 Suzuki H, Nemoto H, Nakamoto H, Okada H, Sugahara S, Kanno Y, Moriwaki K: Continuous
hemodiafiltration with polymyxin B-immobilized fiber is effective in patients with sepsis syn-
drome and acute renal failure. Ther Apher 2002;6:234–240.
12 Nakamura T, Ebihara I, Shimada N, Suzuki S, Ushiyama C, Shoji H, Koide H: Effects of hemop-
erfusion with polymyxin B-immobilized fibre on serum neopterin and soluble interleukin-2 recep-
tor concentrations in patients with septic shock. J Infect 1998;37:241–247.
Cruz/Bellomo/Ronco 450
13 Nakamura T, Ebihara I, Shimada N, Koide H: Changes in plasma erythropoietin and interleukin-6
concentrations in patients with septic shock after hemoperfusion and polymyxin B-immobilized
fiber. Intensive Care Med 1998;24:1271–1276.
14 Nakamura T, Kawagoe Y, Matsuda T, Shoji H, Ueda Y, Tamura N, Ebihara I, Koide H: Effect
of polymyxin B-immobilized fiber on blood metalloproteinase-9 and tissue inhibitor of
metalloproteinase-1 levels in acute respiratory distress syndrome patients. Blood Purif 2004;22:
256–260.
15 Nakamura T, Kawagoe Y, Sukuzi T, Shoji H, Ueda Y, Kobayashi N, Koide H: Changes in plasma
interleukin-18 by direct hemoperfusion with polymyxin B-immobilized fiber in patients with sep-
tic shock. Blood Purif 2005;23:417–420.
16 Shimada N, Nakamura T, Takayashi Y, Tanaka A, Shoji H, Sekizuka K, Ebihara I, Koide H: Effects
of polymyxin B-immobilized fiber on serum phosphate concentrations in patients with sepsis.
Nephron 2000;86:359–360.
17 Tani T, Hanasawa K, Endo Y, Yoshioka T, Kodama M, Kaneko M, Uchiyama Y, Akizawa T,
Takahashi K, Sugai T: Therapeutic apheresis for septic patients with organ dysfunction: hemoper-
fusion using a polymyxin-B immobilized column. Artif Organs 1998;22:1038–1044.
18 Tojimbara T, Sato S, Nakajima I, et al: Polymyxin B-immobilized fiber hemoperfusion after emer-
gency surgery in patients with chronic renal failure. Ther Apher Dial 2004;8:286–292.
19 Ueno T, Sugino M, Nemoto H, Shoji H, Kakita A, Watanabe M: Effect over time of endotoxin
adsorption therapy in sepsis. Ther Apher Dial 2005;9:128–136.
20 Uriu K, Osajima A, Kamochi M, Watanabe H, Aibara K, Kaizu K: Endotoxin removal by direct
hemoperfusion with an adsorbent column using polymyxin B-immobilized fiber ameliorates sys-
temic circulatory disturbance in patients with septic shock. Am J Kidney Dis 2002;39:937–947.
21 Vincent J, Laterre P, Cohen J, Burchardi H, Bruining H, Lerma F, Wittebole X, de Backer D, Brett
S, Marzo D, Nakamura H, John S: A pilot-controlled study of a polymyxin B-immobilized hemo-
perfusion cartridge in patients with severe sepsis secondary to intra-abdominal infection. Shock
2005;23:400–405.
22 Kojika M, Sato N, Yaegashi Y, Suzuki Y, Suzuki K, Nakae H, Endo S: Endotoxin adsorption ther-
apy for septic shock using polymyxin B-immobilized fibers (PMX): evaluation by high-sensitivity
endotoxin assay and measurement of the cytokine production capacity. Ther Apher Dial 2006;10:
12–18.
23 Ono S, Tsujinomoto H, Matsumoto A, Ikuta S, Kinoshita M, Michizuki H: Modulation of human
leukocyte antigen-DR on monocytes and CD16 on granulocytes in patients with polymyxin
B-immobilized fiber. Am J Surg 2004;188:150–156.
24 Tsujimoto H, Ono S, Hiraki S, Majima T, Kawarabayashi N, Sugasawa H, Kinoshita M, Hiraide H,
Mochizuki H: Hemoperfusion with polymyxin B-immobilized fibers reduced the number of
CD16⫹CD14⫹ monocytes in patients with septic shock. J Endotoxin Res 2004;10:229–237.
25 Tsushima K, Kubo K, Koizumi T, Yamamoto H, Fujimoto K, Hora K, Kan-nou Y: Direct hemo-
perfusion using a polymyxin B immobilized column improves acute respiratory distress syndrome.
J Clin Apheresis 2002;17:97–102.
26 Ikeda T, Ikeda K, Nagura M, et al: Clinical evaluation of PMX-DHP for hypercytokinemia caused
by septic multiple organ failure. Ther Apher Dial 2004;8:293–298.
27 Casella G, Monti G, Terzi V, Pulici M, Ravizza A, Vesconi S: Terapie ‘non-convenzionali’ nello
shock settico refrattario: esperienza clinica con Polymyxin B. Minerva Anestesiol 2006;72(suppl 1):
63–67.
28 Ebihara I, Nakamura T, Shimada N, Shoji H, Koide H: Effect of hemoperfusion with polymyxin
B-immobilized fiber on plasma endothelin-1 and endothelin-1 mRNA in monocytes from patients
with sepsis. Am J Kidney Dis 1998;32:953–961.
29 Moher D, Pham B, Jones A, Cook DJ, Jadad AR, Moher M, Tugwell P, Klassen TP: Does quality
of reports of randomised trials affect estimates of intervention efficacy reported in meta-analyses?
Lancet 1998;352:609–613.
30 Schulz KF, Chalmers I, Hayes RJ, Altman DG: Empirical evidence of bias. Dimensions of
methodological quality associated with estimates of treatment effects in controlled trials. JAMA
1995;273:408–412.
Polymyxin B-Immobilized Fiber Column in Sepsis 451
31 Nakamura T, Matsuda T, Suzuki Y, Shoji H, Koide H: Polymyxin B-immobilized fiber in patients
with sepsis. Dial Transplant 2003;32:602–607.
32 Tsugawa K, Koyonagi N, Hashizume M, Wada H, Ayukawa K, Akahoshi K, Tomikawa M,
Sugimachi K: Results of endotoxin absorption after a subtotal resection of the small intestine and
a right hemicolectomy for severe superior mesenteric ischemia. Hepatogastroenterology 2002;49:
1303–1306.
33 Annane D, Sébille V, Charpentier C, Bollaert PE, Francois B, Korach JM, Capellier G, Cohen Y,
Azoulay E, Troche G, Chaumet-Riffaut P, Bellissant E: Effect of a treatment with low doses of
hydrocortisone and fludrocortisone on mortality in patients with septic shock. JAMA 2002;288:
862–971.
34 Amato MB, Barbas CS, Medeiros DM, Magaldi RB, Schettino GP, Lorenzi-Filho G, Kairalla RA,
Deheinzelin D, Munoz C, Oliveira R, Takagaki TY, Carvalho CR: Effect of a protective-ventilation
strategy on mortality in the acute respiratory distress syndrome. N Engl J Med 1998;338:347–354.
35 Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich M;
Early Goal-Directed Therapy Collaborative Group: Early goal-directed therapy in the treatment of
severe sepsis and septic shock. N Engl J Med 2001;345:1368–1377.
36 Bernard GR, Vincent JL, Laterre PF, LaRosa SP, Dhainaut JF, Lopez-Rodriguez A, Steingrub JS,
Garber GE, Helterbrand JD, Ely EW, Fisher CJ Jr; Recombinant Human Protein C Worldwide
Evaluation in Severe Sepsis (PROWESS) Study Group: Efficacy and safety of recombinant
human activated protein C for severe sepsis. N Engl J Med 2001;344:699–709.
37 Manns BJ, Lee H, Doig CJ, Johnson D, Donaldson C: An economic evaluation of activated protein C
treatment for severe sepsis. N Engl J Med 2002;347:993–1000.
Dinna N. Cruz
Department of Nephrology, San Bortolo Hospital
Viale Rodolfi 37
IT–36100 Vicenza (Italy)
Tel. ⫹39 0444 753650, Fax ⫹39 0444 753973, E-Mail dinnacruzmd@yahoo.com
452
Adib-Conquy, M. 101
Alessandri, E. 434
Arroyo, V. 17
Aucella, F. 287
Bagshaw, S.M. 1, 236, 340
Bainotti, S. 405
Baldwin, I. 178, 191
Bellomo, R. 1, 10, 75,
167, 236, 309, 340,
411, 444
Berger, M.M. 267
Bonventre, J.V. 39, 213
Brendolan, A. 411
Bussolati, B. 250
Calzavacca, P. 167
Camussi, G. 250
Cano, N.J.M. 112
Cavaillon, J.-M. 101
Chioléro, R. 267
Crepaldi, C. 411
Cruz, D.N. 309, 411, 434,
444
Davenport, A. 259, 333
De Becker, W. 185
de Cal, M. 411
Delaney, A. 236
Devarajan, P. 203
Di Chiara, L. 428
Di Paolo, S. 287
Dragun, D. 75
Faenza, S. 396
Ferramosca, E. 396
Formica, M. 405
Gesualdo, L. 287
Gigliola, G. 434
Giorni, C. 428
Gressens, B. 354, 371,
387
Haase, M. 75, 340
Haase-Fielitz, A. 75, 340
Herrera, M. 434
Hirasawa, H. 365
Honoré, P.M. 354, 371, 387
Hoste, E.A.J. 32
Iacoella, C. 428
Inguaggiato, P. 405
Issa, N. 419
Joannes-Boyau, O. 354,
371, 387
Joannidis, M. 92
Jones, D. 236
Kellum, J.A. 10, 32, 158
Kindgen-Milles, D. 434
Lameire, N. 325
Langenberg, C. 1
Leverve, X.M. 112
Licari, E. 167
Mancini, E. 396
Mariano, F. 434
Marinho, A. 434
Matsuda, K. 365
Messer, J. 419
Molitoris, B.A. 227
Monti, G. 434
Morelli, S. 428
Moretti, E. 434
Nalesso, F. 411
Oda, S. 365
Opal, S.M. 220
Paganini, E.P. 419
Picardo, S. 297, 428
Pinsky, M.R. 47, 133
Polanco, P.M. 133
Polito, A. 428
Ricci, Z. 197, 297, 428
Robert, R. 434
Ronco, C. XI, 1, 10, 167,
197, 236, 297, 309, 340,
411, 434, 444
Sandoval, R.M. 227
Santoro, A. 396
Schetz, M. 275
Schmit, X. 64
Schusterschitz, N. 92
Sever, M.S. 325
Author Index
Author Index 453
Singer, M. 119
Stazi, G.V. 428
Taccone, F. 64
Tolwani, A.J. 320
Van Biesen, W. 304, 325
Vanholder, R. 304, 325
Veys, N. 304
Vincent, J.-L. 24, 64
Vitale, V. 428
Wheeler, T.S. 320
Wille, K.M. 320
Wratten, M.L. 405
Yassin, J. 119
454
N-acetyl--D-glucosaminidase (NAG),
acute kidney injury biomarker studies
216
Acid-base balance, see Anion gap; Strong
ion gap
Activated partial thromboplastin time
(aPTT), waveform analysis in sepsis-
induced renal injury 224, 225
Acute renal failure (ARF)
definitions 32, 33
epidemiology 33, 34
outcomes
end-stage kidney disease 35
length of hospital stay 35
long-term outcome 36
mortality 35, 36
recovery 250, 251
Acute tubular necrosis (ATN)
conventional view 3, 4
conversion from prerenal azotemia 4–6
sepsis association evidence 7, 8
Albumin
blood detoxification approaches
molecular adsorbent recirculation
system 412, 413, 417
overview 411, 412
plasma filtration adsorption dialysis
circuits 413, 414, 416
purification process 415–417
tricompartmental dialyzer 413, 414
Prometheus system 413
single-pass albumin dialysis 413
fluid resuscitation in sepsis-associated
acute kidney injury 171
hepatorenal syndrome management in
spontaneous bacterial peritonitis 22, 23
toxin transport 412
Alveolar oxygen (P
A
O
2
), calculation
122–124
Anion gap (AG)
calculation 158–160
interpretation 161–163
APACHE scoring
limitations 97, 98
mortality prediction 93
performance 97
Apolipoprotein E, cardiopulmonary-bypass-
associated acute kidney injury alleles
84–86
Argatroban, heparin-induced
thrombocytopenia anticoagulation for
renal replacement therapy 263, 265
Arterial pressure
catheterization indications 139
determinants 137, 138
physiological significance 136, 137
polymyxin-B-immobilized column
effects in sepsis 446
variations during ventilation 138, 139
Beginning and Ending Supportive Therapy
for the Kidney (BEST Kidney)
investigators, continuous renal
replacement therapy survey 298, 299
Bicarbonate
cardiopulmonary-bypass-associated acute
kidney injury prevention 348, 349,
436
Subject Index
Subject Index 455
continuous renal replacement therapy
buffer 290–291
Bioartificial kidney
animal studies 421, 422
clinical trials 424, 425
hemofilter and circuit 422, 423
in vitro studies 421
proximal tubular cell therapy 420, 421
sepsis-induced acute kidney injury
management rationale 419, 420
Bioartificial liver (BAL)
approaches 397, 398
prospects 397
Biomarkers
acute kidney injury
N-acetyl--
D-glucosaminidase 216
cystatin C 207, 208, 217
ideal characteristics 204
interleukin-18 208, 209, 217
kidney injury molecule-1
208, 215, 216
neutrophil gelatinase-associated
lipocalin 205–207, 216, 217
prospects 209, 217, 218
rationale 204, 214
definition 214, 215
immunodysregulation in sepsis
cellular markers 106–109
plasma biomarkers 102–106
sepsis-induced renal injury
activated partial thromboplastin time
waveform analysis 224, 225
interleukin-6 223
lipopolysaccharide assays 222, 223
monocyte function markers 223, 224
overview 220, 221
procalcitonin 225
triggering receptor expressed on
myeloid cells (TREM-1) 225
validation 215
Calcitonin, see Procalcitonin
Calcium, replacement fluid composition
293
Cardiac output (CO)
measurement 149, 150
resuscitation end point 151, 152
Cardiopulmonary-bypass-associated acute
kidney injury
costs and outcomes 341
epidemiology 82, 341
hemolysis
free hemoglobin and kidney injury
induction 346, 347
pigment nephropathy 343, 344
sources and magnitude during bypass
344–346
ischemia-reperfusion injury 342, 343
oxidative stress 343
pathophysiology 341, 342
prevention 342, 347, 349
single-nucleotide polymorphisms
82–87
Catechol-O-methyltransferase (COMT),
cardiopulmonary-bypass-associated acute
kidney injury single-nucleotide
polymorphisms 83, 86, 87
CD11b, sepsis pathophysiology 59
CD14
immunodysregulation marker in sepsis
104, 105
inflammation modulation 48
Central venous catheter (CVC)
complications
acute malfunction 282
central vein stenosis and thrombosis
282, 283
infection 279–281
insertion and care 277, 278
Central venous pressure (CVP)
determinants 140
measurement 139, 140
monitoring 140–142
Cerebral perfusion pressure (CPP), renal
replacement therapy effects
anticoagulant effects 337
continuous renal replacement therapy
336, 337
intermittent hemodialysis 336
peritoneal dialysis 335, 336
Chemokines, acute kidney injury
pathophysiology 43, 44
Citrate, continuous renal replacement
therapy buffer 290
Subject Index 456
Congenital heart disease, renal replacement
therapy in neonates 428–432
Continuous hemodiafiltration, cytokine-
adsorbing hemofilter 365–369
Continuous renal replacement therapy
(CRRT)
alarms 193
circuit patency
access catheter 180
anticoagulant use 179, 180
blood pump flow speed 181
clotting sites 179
membrane factors 181
substitution fluid administration 181,
182
venous chamber 182
computerized monitoring 188
cytokine-adsorbing hemofilter in
continuous hemodiafiltration
365–369
dialysate and replacement fluid
composition
administration route 294, 295
buffers 288–292
calcium 293
glucose 293, 294
physical properties 294
potassium 292, 293
sodium 292
emergency team 192
equipment 188
heparin-induced thrombocytopenia
patients, see Heparin-induced
thrombocytopenia
information technology applications
dialysis accuracy and safety 201
dialysis dose calculation 198–201
prospects 202
intermittent hemodialysis comparison
305–308, 310–317
intracranial pressure and cerebral
perfusion pressure effects 336, 337
medical emergency team
exchange rate 359
global approach in intensive care unit
355–357
rationale 360–363
Saint-Pierre Para-University Hospital
experience 357, 358
vascular access 358, 359
modes 311
nursing
emergency team
challenges 194, 195
overview 360–363
rationale 193, 194
models 192, 193
tasks in intensive care unit 188, 189
nutrition, see Nutrition, acute kidney
disease
program features for intensive care unit
186, 187
proinflammatory mediator removal
312
Renal Disaster Relief Task Force 330
staff training and education 182, 183,
187
standard settings 189
surveys
Beginning and Ending Supportive
Therapy for the Kidney 298, 299
Dose Response Multicenter
International 301, 302
International Course on Critical Care
Nephrology 299, 300
summary of findings 302
thermal balance 313
vascular access
catheter insertion and care 277, 278
complications
acute malfunction 282
central vein stenosis and thrombosis
282, 283
infection 279–281
temporary dialysis catheter
characteristics 276, 277
insertion sites 277
Coupled plasma filtration adsorption
(CPFA)
sepsis management
clinical trials 408
prospects 409
rationale 406
technical characteristics 407
Subject Index 457
Critical care nephrology
collaboration with nephrologists and
other specialists 29, 30
continuous renal replacement therapy, see
Continuous renal replacement therapy
general intensivist role 28, 29
overview 25, 27
Crush syndrome, see also Renal Disaster
Relief Task Force
mortality 326
rhabdomyolysis and acute kidney injury
pathophysiology 326, 327
Cystatin C, acute kidney injury biomarker
studies 207, 208, 217
Danaparoid, heparin-induced
thrombocytopenia anticoagulation
for renal replacement therapy
262, 263
Dermatan sulfate, heparin-induced
thrombocytopenia anticoagulation for
renal replacement therapy 263
Disaster, see Renal Disaster Relief Task Force
Diuretics, acute kidney injury management
clinical trials 237
survey
additional comments 245
clinical outcomes 244, 245
clinical response assessment
243, 244
demographics 240
drug administration
dosing 241
drug types 240, 241
indications and timing 241–243
route 241
sampling frame 239
study design 237–239
summary of findings 245–248
Dose Response Multicenter International
(DO-RE-MI)
participating centers 435
renal replacement therapy use patterns
301, 302, 436, 438–442
study design 437, 438
Endotoxin, see Lipopolysaccharide
Fluid resuscitation, sepsis-associated acute
kidney injury
albumin 171
animal studies 173–175
clinical trials 169–171
gelatin 172
lactate 172, 173
rationale 168, 169
saline 172
starches 171, 172
Fondaparinux, heparin-induced
thrombocytopenia anticoagulation for
renal replacement therapy 263
Furosemide, see Diuretics, acute kidney
injury management
Gelatin, fluid resuscitation in sepsis-
associated acute kidney injury 172
Glucose
control in renal replacement therapy 272
replacement fluid composition 293, 294
Heat shock proteins (HSPs), intracellular
inflammatory response in sepsis 54–57
Hemodialysis, see Continuous renal
replacement therapy; Intermittent
hemodialysis; Slow extended daily
dialysis
Hemodynamic monitoring
arterial pressure
catheterization indications 139
determinants 137, 138
physiological significance 136, 137
variations during ventilation 138, 139
cardiac output
measurement 149, 150
resuscitation end point 151, 152
central venous pressure
determinants 140
measurement 139, 140
monitoring 140, 141
mixed venous oxygen saturation
central venous oxygen saturation 150,
151
measurement 150
resuscitation end point 151, 152
overview 133, 134
Subject Index 458
physiological basis 134, 135
pulmonary artery occlusion pressure
catheterization controversies 152, 153
left ventricular function
afterload 147, 148
performance 148, 149
measurement 144, 145
pleural pressure effects 145, 146
pulmonary edema 146
pulmonary vasomotor tone 146
pulmonary artery pressure
catheterization
controversies 152, 153
indications 141
determinants 144
rationale 134
Hemofiltration, see Continuous
hemodiafiltration; Continuous renal
replacement therapy; High-volume
hemofiltration
Heparin-induced thrombocytopenia (HIT)
anticoagulant alternatives for renal
replacement therapy
direct thrombin inhibitors 263, 264
heparinoids 262, 263
monitoring 265
autoantibody detection 261, 262
incidence 259
pathophysiology 261, 262
scoring 261
Hepatorenal syndrome (HRS)
liver support approaches, see also
Bioartificial liver
artificial systems 398
conventional systems 399
novel treatments 399–403
overview 397, 398
pathogenesis of type 2 disease 18, 19
spontaneous bacterial peritonitis and
type 1 disease
pathophysiology 20, 21
prevention of syndrome 21
treatment
liver transplantation 22
vasoconstrictors and albumin 22, 23
types 17–19
High-volume hemofiltration (HVHF)
animal studies 374–376
clinical trials in acute kidney injury 372,
373, 376–380, 382, 392
dosing 377, 380
filter porosity 391, 392
indications 372, 380
mediator delivery hypothesis 390, 391
peak concentration hypothesis 388, 389
prospects for study 380–382, 391–393
rationale 388–391
threshold immunomodulation hypothesis
389, 390
Hybrid hemodialysis, see Slow extended
daily dialysis
Hypoxia
causes 129
classification 129
definition 128
Immunodysregulation, sepsis
cellular markers 106–109
definition 101, 102
plasma biomarkers 102–106
Incidence, acute kidney injury 34, 203
Inflammation, acute kidney injury
pathophysiology
innate immune response 42
intracellular inflammatory response
heat shock proteins 54–57
mitochondria role 57, 58
nuclear factor-B 51–54
leukocyte-endothelial cell interactions
42, 43
overview 40, 41
prospects for study 45
response modulation by resolvins and
protectins 44
systemic and cellular events 48–50
tubular contribution to injury 43, 44
Information technology (IT)
continuous renal replacement therapy
applications
dialysis accuracy and safety 201
dialysis dose calculation 198–201
prospects 202
definition 197
Hemodynamic monitoring (continued)
Subject Index 459
Intensive care unit (ICU), see Critical care
nephrology
Intercellular adhesion molecule-1 (ICAM-1),
acute kidney injury pathophysiology 42, 43
Interleukin-1 (IL-1), inflammatory response
49, 50
Interleukin-6 (IL-6)
cardiopulmonary-bypass-associated acute
kidney injury single-nucleotide
polymorphisms 83, 84, 86
inflammatory response 49, 50
removal by cytokine-adsorbing
hemofilter 368
sepsis-induced renal injury biomarker 223
Interleukin-10 (IL-10)
immunodysregulation marker in sepsis 105
inflammatory response 49, 50
removal by cytokine-adsorbing
hemofilter 368
sepsis-associated acute kidney injury
single-nucleotide polymorphisms
79, 81
Interleukin-18 (IL-18), acute kidney injury
biomarker studies 208, 209, 217
Intermittent hemodialysis
continuous renal replacement therapy
comparison 305–308, 310–317
dialysate and replacement fluid
composition
administration route 294, 295
buffers 288–292
calcium 293
glucose 293, 294
physical properties 294
potassium 292, 293
sodium 292
intracranial pressure and cerebral
perfusion pressure effects 336
Renal Disaster Relief Task Force 330
vascular access
catheter insertion and care 277, 278
complications
acute malfunction 282
central vein stenosis and thrombosis
282, 283
infection 279–281
temporary dialysis catheter
characteristics 276, 277
insertion sites 277
Intracranial pressure (ICP), renal
replacement therapy effects
anticoagulant effects 337
continuous renal replacement therapy
336, 337
intermittent hemodialysis 336
peritoneal dialysis 335, 336
sustained intracranial pressure surge
management 337
Ischemia-reperfusion injury,
cardiopulmonary-bypass-associated acute
kidney injury 342, 343
Kidney injury molecule-1 (KIM-1), acute
kidney injury biomarker studies 208,
215, 216
Lactate
anaerobic metabolism 120, 122
fluid resuscitation in sepsis-associated
acute kidney injury 172, 173
Left ventricular function, pulmonary artery
occlusion pressure studies
afterload 147, 148
performance 148, 149
Lepirudin, heparin-induced
thrombocytopenia anticoagulation for
renal replacement therapy 264
Lipopolysaccharide (LPS)
inflammatory response 444
polymyxin B removal, see Polymyxin-B-
immobilized column
sepsis-induced renal injury biomarker
assays 222, 223
Lipopolysaccharide-binding protein (LBP),
immunodysregulation marker in sepsis 104
Liver transplantation, hepatorenal syndrome
management in spontaneous bacterial
peritonitis 22
Mesenchymal stem cell (MSC)
acute kidney injury pathophysiology 44
plasticity 251
renal repair role 251, 252, 256
therapy in acute renal failure 252, 253
Subject Index 460
Mitochondria, intracellular inflammatory
response in sepsis 57, 58
Mixed venous oxygen saturation (SvO2)
central venous oxygen saturation
150, 151
measurement 150
resuscitation end point 151, 152
Molecular adsorbent recirculating system
(MARS)
blood detoxification 412, 413, 417
liver support 400, 401
Mortality
acute kidney injury 36, 37, 203
continuous renal replacement therapy
versus intermittent hemodialysis
314, 315
crush syndrome 326
multiple organ failure 24, 25
polymyxin-B-immobilized column
outcomes in sepsis 447, 448
prediction scores
APACHE score 93
limitations 97, 98
mortality prediction model 95, 96
performance 97
physician judgement 98
Sequential Organ Failure Assessment 96
simplified acute physiology score
94, 95
therapeutic intervention scoring system
96
sepsis 68, 167, 405
slow extended daily dialysis 323
Mortality prediction model (MPN) 95, 96
Multi-photon microscopy
applications 227–229
challenges and prospects 233, 234
kidney dynamic process studies 229,
231–233
Multiple organ failure (MOF), see also
Critical care nephrology
classification 65, 66
incidence 67, 68
mortality 24, 25
predisposing factors 66
resuscitation and reversal of organ injury
60, 61
Sequential Organ Failure Assessment
score 25, 26
NADPH oxidase, sepsis-associated acute
kidney injury single-nucleotide
polymorphisms 80, 81
Neonates, congenital heart disease renal
replacement therapy 428–432
Neutrophil-gelatinase-associated lipocalin
(NGAL), acute kidney injury biomarker
studies 205–207, 216, 217
Nuclear factor-B (NF-B)
activation 52–54
intracellular inflammatory response in
sepsis 51–54, 60
Nursing, see Continuous renal replacement
therapy
Nutrition, acute kidney disease
amino acids 115, 116
assessment 267, 268
energy 115, 268
lipids 115
malnutrition prevalence 267
pathophysiology considerations
113, 114
renal replacement therapy
considerations 114, 268
glucose control 272
hyperglycemia 269
nutritional support 270, 271
protein catabolism 268, 269
vitamin and mineral loss 269, 270
vitamins and minerals 116, 117, 271,
272
Oxygen delivery
alveolar oxygen 122–124
anaerobic metabolism and lactate excess
120, 122
capillary to pulmonary vein 126
cellular uptake 127, 128
oxidative metabolism 119, 120
red blood cells 124–126
supply dependency 130
tissue transfer 126, 127
Peritoneal dialysis (PD)
Subject Index 461
congenital heart disease renal
replacement therapy in neonates
429–432
intracranial pressure and cerebral
perfusion pressure effects 335, 336
Renal Disaster Relief Task Force 331
PIRO, sepsis staging 66
Plasma filtration adsorption dialysis
(PFAD)
circuits 413, 414, 416
purification process 415–417
tricompartmental dialyzer 413, 414
Plasmapheresis, liver support 400
Platelet factor 4 (PF4), heparin-induced
thrombocytopenia pathophysiology 261
Polymorphisms, see Single-nucleotide
polymorphisms 7
Polymyxin-B-immobilized column
blood pressure effects 446
endotoxin removal efficiency 445, 446
lipopolysaccharide binding 445
mortality studies in sepsis 447, 448
Potassium, replacement fluid composition
292, 293
Prerenal azotemia
acute tubular necrosis conversion 4–6
conventional view 3, 4
therapeutic implications 6, 7
Procalcitonin, sepsis-induced renal injury
biomarker 225
Prometheus system
blood detoxification 413
liver support 400–403
Protectins, acute kidney injury
pathophysiology 44
Pulmonary artery occlusion pressure
(Ppao)
catheterization controversies 152, 153
left ventricular function
afterload 147, 148
performance 148, 149
measurement 144, 145
pleural pressure effects 145, 146
pulmonary edema 146
pulmonary vasomotor tone 146
Pulmonary artery pressure
catheterization
controversies 152, 153
indications 141
determinants 144
Pulmonary edema, pulmonary artery
occlusion pressure measurement 146
Renal assist device, see Bioartificial kidney
Renal blood flow (RBF), sepsis-associated
acute kidney injury 168
Renal Disaster Relief Task Force
(RDRTF)
crush syndrome
mortality 326
rhabdomyolysis and acute kidney
injury pathophysiology 326, 327
dialysis
continuous renal replacement therapy
330
intermittent hemodialysis 330
peritoneal dialysis 331
earthquakes as renal disasters
326, 327
interventions 328
origins 327
tasks
field hospitals 330
fluid administration 328, 329
severity assessment 328
transport 330
Renal stem cell
distribution in kidney 255
markers 254, 255
properties 253, 254
renal repair role 251, 252, 256
Resolvins, acute kidney injury
pathophysiology 44
Rhabdomyolysis, acute kidney injury
pathophysiology 326, 327
RIFLE criteria
acute kidney injury 2, 11, 14, 15, 35, 75,
167, 168
cardiac patients 13, 14
classes 33, 34
extracorporeal membrane oxygenation
patients 14
origins 11
validation studies 12, 13
Subject Index 462
Saline, fluid resuscitation in sepsis-
associated acute kidney injury 172
Sepsis
acute kidney injury association
epidemiology 77, 406
fluid resuscitation
albumin 171
animal studies 173–175
clinical trials 169–171
gelatin 172
lactate 172, 173
rationale 168, 169
saline 172
starches 171, 172
pathophysiology 168
single-nucleotide polymorphisms
77–82
acute tubular necrosis association
evidence 7, 8
bioartificial kidney, see Bioartificial
kidney
classification 65, 66
coupled plasma filtration adsorption
clinical trials 408
prospects 409
rationale 406
technical characteristics 407
critical care nephrology 25, 27
endotoxin removal, see Polymyxin-B-
immobilized column
hot response 66
immunodysregulation
cellular markers 106–109
definition 101, 102
plasma biomarkers 102–106
incidence 67, 68
infection 66
intracellular inflammatory response
heat shock proteins 54–57
mitochondria role 57, 58
nuclear factor-B 51–54
mortality 68, 167, 405
multiple organ failure, see Multiple organ
failure
organ dysfunction 66, 67
outcome studies 68–72
predisposing factors 66
pro-inflammatory and anti-inflammatory
activity 58–60, 406
systemic cellular events 48–50
systemic inflammatory response
syndrome 47, 65
Sequential Organ Failure Assessment
(SOFA)
mortality prediction 96
multiple organ failure scoring 25, 26
outcome studies 72
Simplified Acute Physiology Score (SAP)
mortality prediction 94, 95
outcome studies 68–72
Single-nucleotide polymorphisms (SNPs)
abundance in human genome 76
cardiopulmonary-bypass-associated acute
kidney injury 82–87
clinical significance 76, 77
sepsis-associated acute kidney injury
77–82
Single-pass albumin dialysis, blood
detoxification 413
Slow extended daily dialysis (SLEDD)
advantages 316, 323
anticoagulation 322
costs 323
dialysate composition 321, 322
dialysate and ultrafiltrate rates 321
duration and timing 321
hemodynamic tolerance 322, 323
machinery 321
mortality 323
principles 320
solute control 322
Sodium, replacement fluid composition
292
Sodium/potassium ATPase, acute kidney
injury pathophysiology 40
Spontaneous bacterial peritonitis (SBP),
type 1 hepatorenal syndrome features
pathophysiology 20, 21
prevention of syndrome 21
treatment
liver transplantation 22
vasoconstrictors and albumin 22, 23
Starches, fluid resuscitation in sepsis-
associated acute kidney injury 171, 172
Subject Index 463
Stem cell, see Mesenchymal stem cell;
Renal stem cell
Strong ion gap (SIG)
etiology 163, 164
interpretation 161–163
strong ion difference calculation 160, 161
Survey, see Continuous renal replacement
therapy; Diuretics, acute kidney injury
management
Sustained low-efficiency dialysis, see Slow
extended daily dialysis
Systemic inflammatory response syndrome
(SIRS)
high-volume hemofiltration 380, 381
sepsis 47, 65
Terlipressin, hepatorenal syndrome
management in spontaneous bacterial
peritonitis 22, 23
Therapeutic intervention scoring system
(TISS), mortality prediction 96
Thrombocytopenia
causes in intensive care unit
260, 261
heparin-induced thrombocytopenia, see
Heparin-induced thrombocytopenia
Toll-like receptors (TLRs)
acute kidney injury pathophysiology
42
TLR4 as immunodysregulation marker in
sepsis 102, 104, 105, 107, 109
Transforming growth factor- (TGF-),
immunodysregulation marker in sepsis
105, 106
Traumatic brain injury (TBI)
acute kidney injury association 334
management 334, 335
renal replacement therapy effects on
intracranial pressure and cerebral
perfusion pressure
anticoagulant effects 337
continuous renal replacement therapy
336, 337
intermittent hemodialysis 336
peritoneal dialysis 335, 336
sustained intracranial pressure surge
management 337
Triggering receptor expressed on myeloid
cells-1 (TREM-1)
immunodysregulation marker in sepsis
102, 104, 106
sepsis-induced renal injury biomarker 225
Tumor necrosis factor-␣ (TNF-␣)
inflammatory response 49, 50
sepsis pathophysiology 59, 60
sepsis-associated acute kidney injury
single nucleotide polymorphisms 78,
79, 81
