Anda di halaman 1dari 10



T ® IC
Sepsis diagnosis and management:
work in progress


Severe sepsis is a common disease process
in the critically ill and is associated with sub-
stantial morbidity and mortality. Continuing
Deparment of Intensive Care,
Erasme Hospital
Free University of Brussels, Brussels, Belgium
research has provided considerable insight
into the pathophysiology of sepsis over
recent years, enabling various aspects of the

sepsis response to be targeted. Discoveries

related to the link between coagulation and Definitions of sepsis
inflammation have been particularly excit-

ing, leading to the development of recombi- Sepsis vs infection

nant activated protein C. This review will dis-
cuss current definitions of sepsis, describe The words “sepsis“ and “infection“ are
new approaches to classification and diag- often used as synonyms, but they are dis-

nosis of patients with sepsis, present rec-


tinct entities. Infection is a microbiological

ommendations for management, and briefly
highlight areas of ongoing and future event, caused by bacteria, fungi, viruses etc.,
research. while sepsis is the host response to that infec-

Key words: Sepsis, diagnosis - Sepsis, therapy - tion, characterized by the release of many
Intensive care. mediators and by a constellation of clinical
and laboratory features (Table I), none of

which are specific for sepsis. Whenever pos-

sible, infection should be defined by the
R ecent years have seen great advances in
our understanding of the pathophysiol-
ogy of sepsis and as a result new treatments
organ primarily infected and the type(s) of
microorganisms (e.g., a lung infection due
have become available and more are being to Pneumococcus or a urinary tract infection
due to E. coli). While the presence of infec-

In addition, progress has been made in tion is essential for a diagnosis of sepsis,
techniques to facilitate the diagnosis of sep- sometimes the source of infection cannot be
sis. identified, and sometimes the bacteriology
In this article we will discuss the recent remains negative. In some patients with sep-
advances in the diagnosis and management sis, the causative infection may never be iden-
of this important and common problem. tified at all; this does not mean the patient
does not have an infection, and hence can-
not have sepsis, just that we are unable to
Address reprint requests to: Prof J. L. Vincent, Erasme locate or identify it. To help to rule out an
Hospital, Route de Lennik 808, 1070 Brussels, Belgium.
E-mail: infection, an objective measure such as the



TABLE I.—Common clinical and laboratory features of In view of these conclusions, the partici-
sepsis. pants introduced the PIRO (predisposing

— Fever/hypothermia factors, infection, response, organ dysfunc-
— Increased cardiac output/low systemic vascular resi- tion) concept,2 a suggested means of staging
stance sepsis, rather like many cancers are staged
— Increased oxygen consumption

— Unexplained tachycardia
using the tumor, nodes, metastases (TNM)
— Unexplained tachypnea/respiratory alkalosis system. Using the emerging PIRO model
— Altered white blood cell count could aid physicians in better characteriz-
— Increased C-reactive protein levels ing heterogeneous groups of septic patients,
— Increased procalcitonin levels providing insight into the continuum of sep-
— Unexplained hyperglycemia
sis, and improve the understanding and

— Unexplained lactic acidosis
— Unexplained respiratory dysfunction (acute lung management of severe sepsis and septic
injury) shock.
— Thrombocytopenia/disseminated intravascular coa-

— Unexplained disorientation or confusion PREDISPOSING FACTORS
— Unexplained alteration in liver function tests
Predisposing factors 4 include individual

— Unexplained alteration in renal function
— Capillary leak syndrome characteristics, such as age, chronic diseases,
prolonged immunodepressant medications,
etc., that may influence a patient’s response

infection probability score, which weights 6
variables [temperature, heart rate, respirato-
to infection and/or indicate which therapies
are likely to be most effective in that patient.
Increasingly, the role an individual’s genetic

ry rate, white blood cell count, C-reactive

protein (CRP), sequential organ failure assess- make-up may have on the development of
ment score (SOFA)] to make a composite sepsis and the severity of disease when it
develops are being explored and various

score, may be useful.1

polymorphisms have been shown to influ-
ence the risk of infection and/or of mortali-
New concepts ty from sepsis.5 Single nucleotide polymor-

In the most recent International Sepsis phisms, microsatellite, insertion and deletion

Consensus Conference,2 the 29 experts polymorphisms are all forms of genetic vari-
charged with providing “a conceptual and a ation that can characterize an individual’s

practical framework to define the systemic risk for sepsis, organ dysfunction, or death.6
inflammatory response to infection” came to A polymorphism of the tumor necrosis factor-α
the following conclusions: first, that the con- (TNF-α) gene, the TNF-2 allele, is associated
cepts of sepsis (the host response to an infec- with increased serum levels of TNF and a
tion), severe sepsis (sepsis associated with greater risk of mortality from septic shock.7 A
organ dysfunction) and septic shock (sepsis polymorphism within the intron 2 of the inter-
plus arterial hypotension despite adequate flu- leukin-1 receptor antagonist (IL-1ra) gene

id resuscitation), as defined in the 1991 North (IL-1RN*2) has been associated with reduced

American Consensus Conference,3 remain use- IL-1ra production and increased mortality
ful to clinicians and researchers, but that they rates.8 Polymorphisms in the Toll-like recep-
do not allow for precise staging or prognosti- tor (TLR) and interferon-γ genes have also
cation of the host response to infection. been identified influencing susceptibility to
Second, that while the concept of systemic sepsis.9, 10 Advances in genetics technology
inflammatory response syndrome (SIRS) have now allowed investigators to design
remains useful as a concept, the diagnostic glass slides (chips) with minute quantities of
criteria for SIRS published in 1992 3 are too short, gene-specific nucleotides. These gene-
sensitive and non-specific, and an expanded specific probe nucleotides, ideally one for
list of signs and symptoms of sepsis may bet- each gene in the genome, are arrayed onto
ter reflect the clinical response to infection. the chip surface to produce a DNA microar-



ray. These can be used to generate an expres- RESPONSE

sion profile, the transcriptome, for the cell The degree of host response can be

or tissue of interest.6 Genomics and the assessed according to the presence or
broader field of proteomics are likely to absence of various clinical and laboratory
become increasingly routinely used in patient features and to the degree of elevation of,

management. for example, white cell count, CRP, procalci-
tonin (PCT) etc.14 The host response will vary
INFECTION between patients and over time in the same
patient. Volk et al. described an early hyper-
Characteristics of the particular infection,
inflammatory phase followed by immuno-
such as site (e.g., lung versus urinary tract),

paralysis based on the level of HLA-DR
responsible organism(s) (e.g., Gram-nega- monocyte surface expression in septic
tive versus Gram positive, MRSA versus patients.15 Improving our assessment of the
MSSA), severity, can have an effect on the host response could help direct therapies

host response and outcome, and may influ-
ence choice of treatment.11 However, while
more appropriately.

few would argue that an E. coli urinary tract
infection is likely to be less life-threatening
than a meningococcal meningitis, classify- Organ dysfunction in severe sepsis is not

ing the relative importance of infections on
outcome is difficult. Using a systematic lit-
erature review and surveying 510 articles
a simple present or not variable, but a con-
tinuous spectrum involving many organs and
varying with time and treatment.16 The degree

describing the outcome of infections, cate- of organ involvement can be assessed with
gorized by microorganism and site of infec- various scoring systems, one of the most com-
tion, Cohen et al.12 recently generated spe- monly used being the SOFA score.17

cific risk codes for bacteremia, meningitis,

pneumonia, skin and soft tissue infections,
peritonitis, and urinary tract infections. For Diagnosis of sepsis

each infection site and organism, a two-dig-


it code was generated according to the mor- Sepsis represents a major diagnostic prob-
tality rate associated with that infection (from lem for the intensivist and considerable effort
1: ≤5% to 4: >30%), and the level of evidence is being concentrated on finding a sensitive

available to support the mortality risk (level and specific diagnostic marker; many have
A representing evidence from more than 5 been suggested 18-22 and a non-exhaustive
studies with more than 100 patients, through list of proposed markers of sepsis is shown in
Table II. Studies have shown that early ther-
to level E insufficient evidence from case
apy is critical in improving outcomes from
reports). They suggested this system be sepsis,23, 24 and the identification of a mark-
termed the grading system for site and sever- er that would allow a clear diagnosis of sep-

ity of infection (GSSSI) and although it needs sis would certainly help improve outcomes in
to be validated, this could be a useful means

these patients.
of better characterizing the different risks Procalcitonin (PCT) and CRP are 2 candi-
associated with infections caused by differ- date markers of sepsis that have received a lot
ent organisms in different sites.12 The tim- of attention. CRP is an acute phase protein
ing of onset of infection may also influence that was first described in the early 1930s.
outcomes. A recent study showed that CRP levels are widely used as a relatively
patients who developed septic shock with- non-specific marker of inflammation, and
in 24 h of ICU admission were more severe- many studies have demonstrated increased
ly ill, but had better outcomes than patients CRP levels in patients with sepsis.25-27
who developed septic shock later during Concentrations above 10 mg/dL on admis-
their ICU stay.13 sion have been associated with particularly



TABLE II.—Some of the suggested markers of sepsis. appropriateness of antimicrobial treatment

— White blood cell count

— C-reactive protein
— Procalcitonin
— Endotoxin Management
— Cytokines – IL-1, IL-2, IL-4, IL-6, IL-8, IL-10, TNF,

— TNF-receptors The management of sepsis can be consid-
— IL-1 receptor antagonist, IL-1 receptors ered in 3 sections: treatment of infection,
— Complement factors hemodynamic resuscitation and organ sup-
— Endothelin-1
— ICAM-1, VCAM-1
port, and modulation of the host response.

— Phospholipase A2
— PGE2 Treatment of infection
— Nitrates/nitrites
— Lactoferrin The source of infection must be identified

— Elastase wherever possible and eliminated by surgical
— Neopterin
removal of an infected focus when applica-
ble, and appropriate antimicrobial therapy

M must be started without delay. If the causative
high mortality rates, and increasing or per- microorganism(s) is unknown, empiric antibi-

sistently high levels indicated a poor prog-
nosis, while declining values were associat-
ed with a more favorable prognosis.28 PCT
was described more recently as a potential
otics should be started based on the likely
culprit and local antibiotic patterns of preva-
lence and resistance. Patients who receive

appropriate antibiotics have a better outcome

marker of infection and is not yet routinely than those who are initially treated with an
measured in many hospital laboratories. PCT ineffective antibiotic,33-35 yet a recent study of
may be superior to CRP in discriminating

septic ICU patients found that in as many as

infectious from other inflammatory diseases. 1/3 of patients the first line choice of antibi-
In addition, PCT has more rapid kinetics, otic was inappropriate.36 Infectious disease
being produced and cleared more rapidly specialists who are familiar with local hospi-

than CRP, so may be better able to identify


tal and community pathogens and resistance

infection early and to follow disease patterns should be involved in antimicrobial
progress.29, 30 Indeed several studies have selection decisions in all patients with severe

shown that PCT levels are correlated with sepsis.34

the severity of sepsis as measured by the
APACHE II or sequential organ failure assess- Hemodynamic resuscitation
ment scores.20, 29, 31, 32
Importantly, none of the markers current- Another basic aspect of the management of
ly available is specific to sepsis, and a diag- the patient with severe sepsis or septic shock
nosis of sepsis cannot conclusively be made is optimal resuscitation, essentially involving

on the basis of the presence of any one item; administration of sufficient fluids, and vasoac-
tive agents when required. Usually the crite-

however, the presence of several features

can help increase the diagnostic likelihood in ria for starting resuscitation include: mean
a patient with a suggestive clinical picture. arterial pressure <65 mmHg, mixed venous
The results of ongoing studies into markers oxygen saturation (SvO2) <70%, urine out-
of sepsis, such as the The Genetic and put <0.5 mL/kg/h, blood lactate concentration
Inflammatory Markers of Sepsis (GenIMS) >1.5 mEq/L, or a worsening of the peripher-
study, are eagerly awaited. The rapid and al capillary circulation. The choice of one
accurate identification of the nature of the fluid type over another and the ideal end-
infection is also of critical importance. As an point of fluid resuscitation have generated
example, identifying fungal or resistant organ- considerable debate and a discussion of these
ism more rapidly would notably improve the aspects is beyond the remit of this paper.



Fluid resuscitation should be commenced as 16 patients treated. Drotrecogin α (activat-

early as possible in the course of septic shock ed)-treated patients also showed significant-

(even before ICU admission).23 Requirements ly faster resolution of cardiovascular and res-
for fluid infusion may be difficult to deter- piratory dysfunction and significantly slower
mine, and such decisions can be facilitated by onset of hematological organ dysfunction

repeated fluid challenges. A fluid challenge than patients who received placebo.42 Apart
comprises 4 components: the type of fluid from an increased risk of bleeding, mostly
to be administered (e.g., natural or artificial associated with invasive procedures, there
colloids, crystalloids); the rate of fluid infusion were no other safety concerns with drotreco-
(e.g., 500 to 1 000 mL over 30 min); the end- gin α (activated) in the PROWESS study. A
further prospective study was conducted in

points (e.g., mean arterial pressure >70
mmHg, heart rate <110 beats/min); the safe- 361 centers to confirm these findings and
ty limits (e.g., central venous pressure not provide additional safety data. This study,
higher than 15 mmHg). ENHANCE, enrolled 2 375 adult patients and

The choice of vasoactive agent has also
been the subject of considerable debate with
treated them with the same dose of drotreco-
gin α (activated) as in the PROWESS study.24
The results confirmed the efficacy and safe-

conflicting opinions in particular regarding
the supremacy of dopamine or norepineph- ty of drotrecogin α (activated). A further
rine.37, 38 Current guidelines recommend either analysis of the safety profile of drotrecogin α

drug as a first-line agent,39, 40 and a random-
ized, controlled, double-blinded clinical trial
comparing the 2 drugs is currently under-
(activated) in more than 6 500 patients who
had received the drug in clinical trials or com-
mercially reported that 43% of bleeding

way in Europe. events that occurred in patients who received

drotrecogin α (activated) were procedure-
related, and that severe thrombocytopenia
Modulation of the host response

and meningitis may be risk factors for serious

bleeding events in patients receiving the
DROTRECOGIN ALFA (ACTIVATED) drug.43 As a result of the inherent risk of
bleeding, drotrecogin α (activated) use is

The development and licensing of contraindicated in patients with active inter-


drotrecogin α (activated), recombinant acti- nal bleeding, recent hemorrhagic stroke,

vated protein C, marked a turn-up on the intracranial or intraspinal surgery, or severe

rather downward sloping history of anti-sep- head trauma, trauma with an increased risk of
sis therapeutics, in which multiple immuno- life-threatening bleeding, presence of an
modulatory agents have fallen to the way- epidural catheter, and intracranial neoplasm
side as one after another clinical trial failed to or mass lesion or evidence of cerebral her-
show that they had any beneficial effect on niation. If a surgical procedure is necessary
outcome. With indications that the coagula- during treatment, the infusion should be
tion system was keenly involved in the patho- stopped 2 h prior to the intervention and

genesis of sepsis, the spotlight fell on medi- restarted 12 h after adequate hemostasis has
ators of the coagulation pathway and their been obtained. Drotrecogin α (activated) is

potential ability to influence sepsis outcomes. expensive, but in cost-effectiveness analy-

The Protein C Worldwide Evaluation in Severe ses, it has been shown to be comparable to
Sepsis (PROWESS) study was a multicenter, other accepted ICU interventions.44, 45
randomized, controlled trial involving 1 690 Children were not included in the PROWESS
patients from 164 centers in 11 countries.41 study, suggested that results from open label
Given at a dose of 24 µg/kg of body weight/h studies and although drotrecogin α (activat-
for 96 h, drotrecogin α (activated) reduced ed) has the same pharmacokinetics, phar-
mortality rates from 30.8% in the placebo macodynamic effects, and safety profile in
group to 24.7% in the treatment group, which the pediatric population,46 a phase III ran-
equated to one additional life saved for every domized controlled trial in children with



severe sepsis has recently been discontinued mechanisms of cell-to-cell signaling and the
for lack of efficacy. Drotrecogin α (activated) importance of apoptosis, more potential ther-

is currently licensed for use in adult patients apeutic targets become apparent. Here we
with severe sepsis and a high risk of mortal- will briefly mention just some of the many
ity (e.g., as assessed by the APACHE II score). agents under research, focusing on those
This decision is supported by the recent

that are, or are almost, at the clinical trial
Administration of Drotrecogin alfa [activat- stage.
ed] in Early Severe Sepsis (ADDRESS) study
that failed to show effectiveness of drotreco- New anti-endotoxin agents
gin α (activated) in patients with less severe
disease. Endotoxin (lipopolysaccharide, LPS) is a

component of the Gram-negative bacterial
cell wall and is a key initiator of sepsis. Once
in the circulation, endotoxin binds to LPS

In a randomized controlled trial conduct- binding protein (LBP), and the LPS-LPB com-
ed in 19 ICUs across France and including plex interacts with CD14 on the surface of
300 patients with septic shock, patients with monocytes and macrophages resulting in cel-

relative adrenal insufficiency (as assessed by lular activation. The LPS-LPB complex can
non-response to a corticotrophin test) were also interact with soluble CD14 and this can
treated with hydrocortisone (50 mg i.v. every then interact with another receptor on the

6 h) and fludrocortisone (50 µg per os daily)
or placebo for 7 days. Treated patients had a
reduced mortality compared to non-respon-
endothelial cell, which itself lacks surface
CD14. LPS can also form complexes with
serum lipoproteins, including low density

ders treated with placebo (53% vs 63%, haz- lipoproteins (LDL), high density lipoproteins
ard ratio 0.67, 95% CI 0.47-0.95, P=0.02).48 (HDL), and apolipoprotein A, which provide
In a recent meta-analysis of randomized stud- a means of eliminating LPS from the circula-

ies in patients with severe sepsis or septic tion.51 In acute illness, lipoprotein levels are
shock treated with corticosteroids, overall reduced,52, 53 this reducing LPS clearance.
use of corticosteroids did not significantly Recent studies have indicated that intensive

affect mortality.49 However, with longer cours- insulin therapy to maintain blood glucose

es (>5 days) of lower dose corticosteroids levels less than 6.1 mmol/L improved out-
(≤300 mg hydrocortisone or equivalent), mor- comes predominantly by reducing deaths

tality at 28 days and hospital mortality were from sepsis-induced multiple organ failure.54
reduced.49 Another meta-analysis reached sim- The incidence of nosocomial infections was
ilar conclusions.50 Nevertheless, some ques- also reduced. A further study from Van den
tions remain unanswered, including the opti- Berghe et al. noted that intensive insulin ther-
mal test for adrenal insufficiency and whether apy also increases HDL and LDL levels and
the results can be extrapolated to patients with multivariate analysis suggested that it was
severe sepsis not in shock. It is not entirely this change in lipid levels, rather than an

clear whether steroids treat sepsis per se, or effect on glucose levels, that contributed to
enhance hemodynamic support. Indeed, stud-

the beneficial effects of this approach on

ies have indicated that steroids can increase the mortality and morbidity.55 Replacing lipopro-
response to adrenergic agents. At present, teins may therefore represent a novel method
steroid use in sepsis should be administered of treating sepsis. Experimental studies in
only in the presence of shock. human volunteers and animal models did
indeed show that HDL reduced flu-like symp-
toms during endotoxemia, and blocked the
The future endotoxin-induced release of TNF, IL-6 and
IL-8. Interestingly, HDL administration also
As our understanding of the pathophysi- reduced CD14 expression on monocytes.56
ology of sepsis improves, in particular the In a porcine model of sepsis, an emulsion



of phospholipid, the predominant lipid in rationale that by removing inflammatory

HDL, lowered serum endotoxin and TNF-α mediators, this strategy could improve out-

levels significantly, preserved cardiac output comes. However, hemoperfusion per se
and ejection fraction, and attenuated increas- removes all mediators both good and bad
es in systemic and pulmonary vascular resis- and may not be beneficial in all patients at all
times. Some studies have suggested that cou-

tances.57 A clinical trial with this phospho-
lipid emulsion is ongoing. pled plasma filtration adsorption (CPFA)
Other anti-endotoxin strategies currently improves blood pressure and restores
undergoing clinical testing include anti-CD14 immune function in patients with septic
antibodies,58 extracorporeal endotoxin ab- shock 69, 70 However, further study is needed
to determine which technique may be most

sorption,59 and lipid A analogs.60
beneficial, and at present hemofiltration is
Macrophage migration inhibitory factor (MIF) not recommended unless there is co-exist-
ing renal failure.39

By modulating the expression of TLR4, the
signal-transducing molecule in the LPS recep-
tor complex, macrophage migration inhibito-

ry factor (MIF) induces the production of var-
ious pro-inflammatory mediators.61 MIF lev- Severe sepsis and septic shock are still

els are raised in patients with sepsis 62, 63 and
correlate with outcome,63 and animal studies
have demonstrated improved survival with
MIF neutralization.62, 64
associated with high mortality rates, but
progress is being made. Early diagnosis is
vital in improving outcomes and develop-

ment of more effective markers of sepsis

will help in allowing treatments to be insti-
High-mobility group B-1 protein tuted as early as possible in the disease

High-mobility group B-1 protein (HMGB1) process. Appropriate infection control strate-
acts as a late mediator of systemic inflamma- gies and adequate haemodynamic stabi-
tion, and is released from endotoxin-stimu- lization remain essential, but must now be
combined with drotrecogin alfa (activated)

lated macrophages some 8-12 h after the

if there is no contraindication, and other

release of the early cytokines. HMGB1 induces

the activation of transcription factors and the immunomodulatory strategies as they
release of pro-inflammatory mediators by become available. It is unlikely that one

monocytes and endothelial cells.65, 66 In animal agent will ever be discovered to cure all
models of sepsis, anti-HMGB1 antibodies, patients with sepsis, just as one antibiotic
even when given late after sepsis induction, cannot treat all infections, and one
improved survival and reduced sepsis-induced chemotherapy drug is not active against all
organ injury.67 Ethyl pyruvate, which inhibits cancers. Sepsis treatment must be seen as a
HMGB1 production in vivo, also improved package, with different patients requiring
different packages, and the package con-

survival in a mouse model of sepsis.68 Unlike

tents varying with time in the same patient.
many animal models when the onset of sep-

Attempts to characterize patients, such as

sis is determined artificially, in patients the
the PIRO system, will help in identifying
onset is often difficult to define and many
which therapies are best given to which
patients will be diagnosed relatively late, mak-
patients and when, and as new interven-
ing the longer time-frame of HMGB1 of par-
tions undergo clinical testing, treatment pro-
ticular interest as it may be effective even
tocols must be adapted accordingly. This is
when targeted later in the sepsis process.
a rapidly moving field, very much a work in
progress, and the success of drotrecogin α
Hemoperfusion strategies
(activated) has given new impetus to the
Blood purification systems have been pro- search for effective diagnostic and thera-
posed for the treatment of sepsis based on the peutic strategies in patients with sepsis.



Riassunto 13. Roman-Marchant O, Orellana-Jimenez CE, De Backer

D, Melot C, Vincent JL. Septic shock of early or late
onset: does it matter? Chest 2004;126:173-8.

Diagnosi di sepsi e sua gestione: attuali progressi
14. Gerlach H, Dhainaut JF, Harbarth S, Reinhart K, Marshall
La sepsi grave è un processo patologico comune JC, Levy M. The PIRO Concept: R is for response. Crit
nel paziente criticamente ammalato ed è associata Care 2003;7:256-9.
ad una sostanziale morbidità e mortalità. In questi 15. Volk HD, Reinke P, Krausch D, Zuckermann H,

Asadullah K, Müller JM et al. Monocyte deactivation -
ultimi anni la continua ricerca ha fornito notevoli dati rationale for a new therapeutic strategy in sepsis.
sulla fisiopatologia della sepsi, evidenziando vari Intensive Care Med 1996;22:S474-81.
aspetti della risposta alla sepsi che devono essere 16. Vincent JL, Wendon J, Groeneveld J, Marshall JC, Streat
colpiti. Le scoperte relative alla correlazione tra coa- S, Carlet J. The PIRO Concept: O is for organ dys-
gulazione e infiammazione sono state particolarmente function. Crit Care 2003;7:260-4.
eccitanti, consentendo lo sviluppo della proteina C 17. Vincent JL, de Mendonça A, Cantraine F, Moreno R,
Takala J, Suter P et al. Use of the SOFA score to assess

attivata ricombinante. Questa revisione discuterà le the incidence of organ dysfunction/failure in intensive
attuali definizioni di sepsi, descriverà i nuovi approc- care units: results of a multicentric, prospective study.
ci alla classificazione e alla diagnosi dei pazienti con Crit Care Med 1998;26:1793-800.
sepsi e presenterà le raccomandazioni per la gestio- 18. Pinsky MR, Vincent JL, Deviere J, Alegre M, Schandene

ne, inoltre discuterà brevemente le attuali aree di L, Kahn RJ et al. Serum cytokine levels in human sep-
tic shock: relation to multiple-systems organ failure
ricerca e quelle future. and mortality. Chest 1993;103:565-75.
Parole chiave: Sepsi, diagnosi - Sepsi, terapia - Terapia 19. Marty C, Misset B, Tamion F, Fitting C, Carlet J, Cavaillon

JM. Circulating interleukin-8 concentrations in patients
M with multiple organ failure of septic and nonseptic ori-
gin. Crit Care Med 1994;22:673-9.
20. Ugarte H, Silva E, Mercan D, de Mendonça A, Vincent

T References
1. Peres-Bota D, Melot C, Lopes Ferreira F, Vincent JL.
Infection probability score (IPS): a simple method to
JL. Procalcitonin as a marker of infection in the inten-
sive care unit. Crit Care Med 1999;27:498-504.
21. Gibot S, Kolopp-Sarda MN, Bene MC, Cravoisy A, Levy
B, Faure GC et al. Plasma level of a triggering recep-

tor expressed on myeloid cells-1: its diagnostic accu-

help assess the probability of infection in critically racy in patients with suspected sepsis. Ann Intern Med
patients. Crit Care Med 2003;31:2579-84. 2004;141:9-15.
2. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, 22. Marshall JC, Foster D, Vincent JL, Cook DJ, Cohen J,
Cook D et al. 2001 SCCM/ESICM/ACCP/ATS/SIS

Dellinger RP et al. Diagnostic and prognostic implica-

International Sepsis Definitions Conference. Crit Care tions of endotoxemia in critical illness: results of the
Med 2003;31:1250-6. MEDIC study. J Infect Dis 2004;190:527-34.
3. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, 23. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A,
Knaus WA et al. Definition for sepsis and organ failure Knoblich B et al. Early goal-directed therapy in the

and guidelines for the use of innovative therapies in treatment of severe sepsis and septic shock. N Engl J

sepsis. Chest 1992;101:1644-55. Med 2001;345:1368-77.

4. Angus DC, Burgner D, Wunderink R, Mira JP, Gerlach 24. Vincent JL, Bernard GR, Beale R, Putersen C, Dhairant
H, Wiedermann CJ et al. The PIRO concept: P is for pre- JF et al. Drotrecogin alfa (activated) treatment in severe
disposition. Crit Care 2003;7:248-51.

sepsis from the global open-label trial ENHANCE: fur-

5. Lin MT, Albertson TE. Genomic polymorphisms in sep- ther evidence for survival and safety and implications
sis. Crit Care Med 2004;32:569-79. for treatment. Crit Care Med 2005;33:2266-77.
6. Cobb JP, O’Keefe GE. Injury research in the genomic 25. Póvoa P, Almeida E, Moreira P, Fernandes A, Mealha R,
era. Lancet 2004;363:2076-83. Aragão A et al. C-reactive protein as an indicator of
7. Appoloni O, Dupont E, Andrien M, Duchateau J, sepsis. Intensive Care Med 1998;24:1052-6.
Vincent JL. Association of TNF2, a TNFα promoter 26. Yentis SM, Soni N, Sheldon J. C-reactive protein as an
polymorphism, with plasma TNFα levels and mortali- indicator of resolution of sepsis in the intensive care
ty in septic shock. Am J Med 2001;110:486-8. unit. Intensive Care Med 1995;21:602-5.
8. Arnalich F, Lopez-Maderuelo D, Codoceo R, Lopez J, 27. Smith RP, Lipworth BJ, Cree IA, Spiers EM, Winter JH.

Solis-Garrido LM, Capiscol C et al. Interleukin-1 receptor C-reactive protein. A clinical marker in community-
antagonist gene polymorphism and mortality in patients acquired pneumonia. Chest 1995;108:1288-91.

with severe sepsis. Clin Exp Immunol 2002;127:331-6. 28. Lobo SM, Lobo FR, Bota DP, Lopes-Ferreira F, Soliman
9. Lorenz E, Mira JP, Cornish KL, Arbour NC, Schwartz DA. HM, Melot C et al. C-reactive protein levels correlate
A novel polymorphism in the toll-like receptor 2 gene with mortality and organ failure in critically ill patients.
and its potential association with staphylococcal infec- Chest 2003;123:2043-9.
tion. Infect Immun 2000;68:6398-401. 29. Meisner M, Tschaikowsky K, Palmaers T, Schmidt J.
10. Stassen NA, Leslie-Norfleet LA, Robertson AM, Comparison of procalcitonin (PCT) and C-reactive pro-
Eichenberger MR, Polk HC Jr. Interferon-gamma gene tein (CRP) plasma concentrations at different SOFA
polymorphisms and the development of sepsis in scores during the course of sepsis and MODS. Crit
patients with trauma. Surgery 2002;132:289-92. Care (Lond) 1999;3:45-50.
11. Vincent JL, Opal S, Torres A, Bonten M, Cohen J, 30. Castelli GP, Pognani C, Meisner M, Stuani A, Bellomi D,
Wunderink R. The PIRO concept: I is for infection. Crit Sgarbi L. Procalcitonin and C-reactive protein during
Care 2003;7:252-5. systemic inflammatory response syndrome, sepsis and
12. Cohen J, Cristofaro P, Carlet J, Opal S. New method of organ dysfunction. Crit Care 2004;8:R234-42.
classifying infections in critically ill patients. Crit Care 31. Claeys R, Vinken S, Spapen H, ver Elst K, Decochez K,
Med 2004;32:1510-26. Huyghens L et al. Plasma procalcitonin and C-reactive



protein in acute septic shock: clinical and biological cor- of hydrocortisone and fludrocortisone on mortality in
relates. Crit Care Med 2002;30:757-62. patients with septic shock. JAMA 2002;288:862-71.
32. Luzzani A, Polati E, Dorizzi R, Rungatscher A, Pavan R, 49. Annane D, Bellissant E, Bollaert PE, Briegel J, Keh D,

Merlini A. Comparison of procalcitonin and C-reactive Kupfer Y. Corticosteroids for severe sepsis and septic
protein as markers of sepsis. Crit Care Med 2003;31: shock: a systematic review and meta-analysis. BMJ
1737-41. 2004;329:480. Epub 2004 Aug 02.
33. Hanon FX, Monnet DL, Sorensen TL, Molbak K, 50. Minneci PC, Deans KJ, Banks SM, Eichacker PQ,

Pedersen G, Schonheyder H. Survival of patients with Natanson C. Meta-analysis: the effect of steroids on
bacteraemia in relation to initial empirical antimicrobial survival and shock during sepsis depends on the dose.
treatment. Scand J Infect Dis 2002;34:520-8. Ann Intern Med 2004;141:47-56.
34. Byl B, Clevenbergh P, Jacobs F, Struelens MJ, Zech F, 51. Lynn WA. Anti-endotoxin therapeutic options for the
Kentos A et al. Impact of infectious diseases specialists treatment of sepsis. J Antimicrob Chemother 1998;41
and microbiological data on the appropriateness of Suppl A:71-80.
antimicrobial therapy for bacteremia. Clin Infect Dis 52. Gordon BR, Parker TS, Levine DM, Saal SD, Wang JC,
1999;29:60-6. Sloan BJ et al. Low lipid concentrations in critical illness:

35. Valles J, Rello J, Ochagavia A, Garnacho J, Alcala MA. implications for preventing and treating endotoxemia.
Community-acquired bloodstream infection in criti- Crit Care Med 1996;24:584-9.
cally ill adult patients: impact of shock and inappro- 53. van Leeuwen HJ, Heezius EC, Dallinga GM, van Strijp
priate antibiotic therapy on survival. Chest 2003;123: JA, Verhoef J, van Kessel KP. Lipoprotein metabolism

1615-24. in patients with severe sepsis. Crit Care Med 2003;31:
36. Malacarne P, Rossi C, Bertolini G. Antibiotic usage in 1359-66.
intensive care units: a pharmaco-epidemiological mul- 54. Van den Berghe G, Wouters P, Weekers F, Verwaest C,
ticentre study. J Antimicrob Chemother 2004;54:221-4. Bruyninckx F, Schetz M et al. Intensive insulin therapy

37. Vincent JL, De Backer D. My initial vasopressor agent in the critically ill patient. N Engl J Med 2001;345:
in septic shock is dopamine rather than norepinephrine. 1359-67.
Crit Care 2003;7:6-8. 55. Mesotten D, Swinnen JV, Vanderhoydonc F, Wouters PJ,
38. Sharma VK, Dellinger RP. The International Sepsis Van den Berghe G. Contribution of circulating lipids to

39. T
Forum’s controversies in sepsis: my initial vasopres-
sor agent in septic shock is norepinephrine rather than
dopamine. Crit Care 2003;7:3-5.
Dellinger RP, Carlet JM, Masur H, Gerlach H, Calandra 56.
the improved outcome of critical illness by glycemic
control with intensive insulin therapy. J Clin Endocrinol
Metab 2004;89:219-26.
Pajkrt D, Doran JE, Koster F, Lerch PG, Arnet B, van der

T, Cohen J et al. Surviving sepsis campaign guidelines Poll T et al. Antiinflammatory effects of reconstituted
for management of severe sepsis and septic shock. high-density lipoprotein during human endotoxemia.
Crit Care Med 2004;32:858-73. J Exp Med 1996;184:1601-8.
40. Hollenberg SM, Ahrens TS, Annane D, Astiz ME, Chalfin 57. Goldfarb RD, Parker TS, Levine DM, Glock D, Akhter

DB, Dasta JF et al. Practice parameters for hemody- I, Alkhudari A et al. Protein-free phospholipid emulsion
namic support of sepsis in adult patients: 2004 update. treatment improved cardiopulmonary function and
Crit Care Med 2004;32:1928-48. survival in porcine sepsis. Am J Physiol Regul Integr
41. Bernard GR, Vincent JL, Laterre PF, LaRosa SP, Dhainaut Comp Physiol 2003;284:R550-7.
JF, Lopez-Rodriguez A et al. Recombinant human pro- 58. Reinhart K, Gluck T, Ligtenberg J, Tschaikowsky K,

tein C Worldwide Evaluation in Severe Sepsis Bruining A, Bakker J et al. CD14 receptor occupancy

(PROWESS) study group. Efficacy and safety of recom- in severe sepsis: results of a phase I clinical trial with
binant human activated protein C for severe sepsis. N a recombinant chimeric CD14 monoclonal antibody
Engl J Med 2001;344:699-709. (IC14). Crit Care Med 2004;32:1100-8.

42. Vincent JL, Angus DC, Artigas A, Kalil A, Basson BR, 59. Reinhart K, Meier-Hellmann A, Beale R, Forst H, Boehm
Jamal HH et al. Effects of drotrecogin alfa (activated) D, Willatts S et al. Open randomized phase II trial of an
on organ dysfunction in the PROWESS trial. Crit Care extracorporeal endotoxin adsorber in suspected Gram-
Med 2003;31:834-40. negative sepsis. Crit Care Med 2004;32:1662-8.
43. Bernard GR, Macias WL, Joyce DE, Williams MD, Bailey 60. Lynn M, Rossignol DP, Wheeler JL, Kao RJ, Perdomo
J, Vincent JL. Safety assessment of drotrecogin alfa CA, Noveck R et al. Blocking of responses to endotoxin
(activated) in the treatment of adult patients with severe by E5564 in healthy volunteers with experimental endo-
sepsis. Crit Care 2003;7:155-63. toxemia. J Infect Dis 2003;187:631-9.
44. Angus DC, Linde-Zwirble WT, Clermont G, Ball DE, 61. Calandra T, Froidevaux C, Martin C, Roger T. Macro-
Basson BR, Ely EW et al. Cost-effectiveness of drotreco- phage migration inhibitory factor and host innate

gin alfa (activated) in the treatment of severe sepsis. Crit immune defenses against bacterial sepsis. J Infect Dis
Care Med 2003;31:1-11. 2003;187 Suppl 2:S385-90.

45. Neilson AR, Burchardi H, Chinn C, Clouth J, Schneider 62. Calandra T, Echtenacher B, Roy DL, Pugin J, Metz CN,
H, Angus D. Cost-effectiveness of drotrecogin alfa (acti- Hultner L et al. Protection from septic shock by neu-
vated) for the treatment of severe sepsis in Germany. tralization of macrophage migration inhibitory factor.
J Crit Care 2003;18:217-27. Nat Med 2000;6:164-70.
46. Barton P, Kalil AC, Nadel S, Goldstein B, Okhuysen- 63. Bozza FA, Gomes RN, Japiassu AM, Soares M, Castro-
Cawley R, Brilli RJ et al. Safety, pharmacokinetics, Faria-Neto HC, Bozza PT et al. Macrophage migration
and pharmacodynamics of drotrecogin alfa (activated) inhibitory factor levels correlate with fatal outcome in
in children with severe sepsis. Pediatrics 2004;113: sepsis. Shock 2004;22:309-13.
7-17. 64. Bozza M, Satoskar AR, Lin G, Lu B, Humbles AA,
47. Abraham E, Latesia PF, Garg R, Levy H, Talwar D, Gerard C et al. Targeted disruption of migration
Trzaskoma BL et al. Drotrecogin alfa (activated) for inhibitory factor gene reveals its critical role in sepsis.
adults with severe sepsis and a low risk of death. N Engl J Exp Med 1999;189:341-6.
J Med 2005;333:1332-41. 65. Andersson U, Wang H, Palmblad K, Aveberger AC,
48. Annane D, Sebille V, Charpentier C, Bollaert PE, Francois Bloom O, Erlandsson-Harris H et al. High mobility
B, Korach JM et al. Effect of treatment with low doses group 1 protein (HMG-1) stimulates proinflammatory



cytokine synthesis in human monocytes. J Exp Med Yang R et al. Ethyl pyruvate prevents lethality in mice
2000;192:565-70. with established lethal sepsis and systemic inflamma-
66. Fiuza C, Bustin M, Talwar S, Tropea M, Gerstenberger tion. Proc Natl Acad Sci U S A 2002;99:12351-6.

E, Shelhamer JH et al. Inflammation-promoting activ- 69. Ronco C, Brendolan A, Lonnemann G, Bellomo R,
ity of HMGB1 on human microvascular endothelial Piccinni P, Digito A et al. A pilot study of coupled plas-
cells. Blood 2003;101:2652-60. ma filtration with adsorption in septic shock. Crit Care
67. Yang H, Ochani M, Li J, Qiang X, Tanovic M, Harris HE Med 2002;30:1250-5.

et al. Reversing established sepsis with antagonists of 70. Formica M, Olivieri C, Livigni S, Cesano G, Vallero A,
endogenous high-mobility group box 1. Proc Natl Acad Maio M et al. Hemodynamic response to coupled plas-
Sci U S A 2004;101:296-301. mafiltration-adsorption in human septic shock. Intensive
68. Ulloa L, Ochani M, Yang H, Tanovic M, Halperin D, Care Med 2003;29:703-8.