Markers of Microbial Infection LEADING AR AB Johan Groeneveld , Annelies Tacx

AB Johan Groeneveld1, Annelies Tacx1, Remco Peters2, Michiel van Agtmael2,
and C Erik Hack3
Department of Intensive Care, 2Internal Medicine, and 3Clinical Chemistry, Vrije Universiteit Medical Centre,
and 3Department of Immunopathology, Sanquin Research at the CLB, Amsterdam, The Netherlands
Sepsis is the host response to microbial infection, and is diagnosed on the basis of clinical signs and
symptoms and demonstration of an infectious focus. The latter includes microbiological confirmation,
which requires time. Indeed, in suspected sepsis, Gram-stains of secretions may not be helpful, and
results of cultures and susceptibility tests in the microbiology laboratory may take a few days to obtain.
There is an urgent need for a rapid bedside diagnosis of infection, i.e. of the type and load of
microorganisms involved, since clinical signs and symptoms of infection lack appropriate sensitivity and
specificity. Molecular markers for infection, such as bacterial DNA (polymerase chain reaction) or
circulating bacterial products, and markers related to the first line of defense (i.e. the nonspecific innate
immune response to microbial infection) have been studied to determine their diagnostic value for
infection. The innate immune response involves the release of first pro-inflammation and later of antiinflammatory factors by macrophages, neutrophils, and the endothelium, upon stimulation by bacterial
products. The aims of the innate immune response are the compartmentalization of the infection,
prevention of spread, and to kill the invading organisms through a localized proinflammatory response.
A counteracting anti-inflammatory response serves to limit the systemic effects of proinflammatory
factors. The invasiveness and microbial load are possible determinants of the host response to infection
and its secondary sequelae, such as sepsis. Circulating markers of the host response, such as interleukin-6
and procalcitonin, are helpful to predict the presence of microbial infection, particularly bacteremia, at an
early stage in patients with clinical manifestations of an inflammatory response, such as fever. Use of
infection markers, derived from both the pathogens and the host, could help to achieve a fast and
accurate likely diagnosis of infection and thereby help to decide on antimicrobial therapy and other
measures in a patient with (suspected) sepsis. Advances in Sepsis 2004;3(3):83–90.
Sepsis is classically defined as the overt and severe clinical
manifestation of microbial infection. An infection clinically
documented by imaging techniques, fever (or hypothermia),
tachycardia, tachypnea, and leukocytosis (or leukopenia)
may point to probable sepsis. A microbiologically confirmed
infection, by Gram-stain or culture, points to definite sepsis
(Table 1). Fever (or hypothermia), tachycardia, tachypnea,
and leukocytosis (or leukopenia) constitute the criteria for
the systemic inflammatory response syndrome (SIRS), the
supposed host response to microbial infection [1–3].
However, there is no gold standard for diagnosing
definite sepsis as cultures can be negative in cases of poor
Address for correspondence: ABJ Groeneveld, Department of Intensive
Care, Vrije Universiteit Medical Centre, De Boelelaan 1117, 1081 HV,
Amsterdam, The Netherlands. Email: [email protected]
sampling or timing, or (pre)treatment with antimicrobial
agents. Indeed, >30% of patients with classical clinical signs
of sepsis may have negative cultures. The mortality rate in
these patients may be somewhat lower than in those with
confirmed local infection or bacteremia [4–6]; nevertheless,
culture-negative sepsis and shock may carry a dismal
prognosis [4]. The absence of cultured bacteria in patients
with suspected sepsis, as opposed to definite sepsis with
positive cultures, may relate to occult infection rather than a
systemic inflammatory response to noninfectious disease.
The use of specific infection markers, if available, could help
to establish a more definite diagnosis. This should not
circumvent a diligent search for microbes in any patient with
suspected sepsis.
The demonstration of microbes in stains and cultures in
specimens that are sterile by nature remains the basis for
Markers of Microbial Infection
Table 1. Defining the clinical host response to infection.
Two or more of the following:
• Temperature >38ºC or <36ºC
• Tachycardia >90 beats/min
• Tachypnea >20/min (PaO2 <32 torr)
• White cell count >12x109 cells/L or <4x109 cells/L or >10% immature bands
Severe SIRS
SIRS plus signs of organ hypoperfusion/dysfunction such as:
• Acute mental changes
• Oxygenation disturbances (PaO2 <70–75 torr)
• Hyperlactatemia and metabolic acidemia
• Oligura (<0.5 mL/kg for at least 1 h)
• Coagulation disturbances
SIRS plus clinically presumed and/or microbiologically proven infection
Severe sepsis1/sepsis syndrome2
Sepsis plus hypotension responsive to 1 h of fluid loading, or organ hypoperfusion/dysfunction1
Sepsis plus one or more signs of organ dysfunction2
Septic shock
Sepsis and hypotension despite adequate fluid resuscitation/need for vasopressor therapy
plus one or more signs of organ dysfunction as in severe sepsis/sepsis syndrome
According to American College of Chest Physicians/Society of Critical Care Medicine (ACCP/SCCM) consensus criteria [2]. 2According to
Bone [1]. PaO2: partial pressure of arterial oxygen; SIRS: systemic inflammatory response syndrome.
microbiological proof of a clinically suspected infection, and
is therefore important for diagnosing definite sepsis. Results
of microbiological studies are often not immediately
available at the bedside, in which case sepsis may be difficult
to diagnose. In fact, it sometimes may take days before a
correct diagnosis is made. Yet the delay in making a
diagnosis of definite sepsis is of crucial therapeutic
importance. Indeed, treatment decisions, at an early stage,
are often empirical. The choice of antimicrobial therapy is
directed at the most likely causative organisms or based on
results of Gram-staining, if available [7]. However,
Gram-stainings of certain specimens (sputum, catheter
urine) do not correlate well with the final results of cultures
and are, therefore, poor guides for appropriate antimicrobial
therapy [7,8].
Many patients with clinical manifestations of sepsis do
not have a clinical focus of infection, but may still receive
antibiotics until the clinical picture becomes clear [5,9].
Trauma, burns, major surgery, pancreatitis, pulmonary
embolism, and many other conditions may elicit clinical
signs of SIRS in the absence of microbial infection. In such
cases, sepsis is often suspected but administration of
antibiotics provides no benefit. Thus, patients with fever
may suffer from SIRS by a noninfectious condition, and
withholding antimicrobial agents in these patients may be
safe. Therefore, models based on clinical information have
been developed to predict local infection (i.e. positive
microbiological results) and bacteremia/fungemia in a
variety of clinical conditions and settings [5,10]. Although
these models may have some value, the clinical and
laboratory parameters that predict bacteremia may differ
from those used to define SIRS (Table 1), may be poorly
reproducible, and may have too many false negative and
positive numbers [5,10]. False negativity implies a risk of
undertreatment with antibiotics, particularly when these
would have been prescribed on the basis of a high a priori
likelihood of microbial infection [5]. False positivity implies
overtreatment, which may have harmful effects. Although
still widely used, the SIRS criteria are both too nonspecific
and too sensitive to help in this respect [5]. Other clinical
signs, such as hypoalbuminemia and thrombocytopenia, are
good predictors of blood stream infection, but are not
included in the SIRS criteria [5]. Therefore, in order to
improve clinical decision making, it would be helpful to have
bedside parameters and determinations available for early
prediction of blood stream infection and a positive response
to appropriate antimicrobial therapy.
The innate immune response to microbial infection,
limited to bacterial infection for the purposes of this review,
is initially triggered by the membrane interference of
bacterial cell wall components, such as lipoteichoic
acid/peptidoglycans for Gram-positive infections and
lipopolysaccharide (endotoxin) for Gram-negative infections
[11–13]. CD14, Toll-like receptor-4, and MD2 together form
the receptor for endotoxin on resident macrophages and
neutrophils and are responsible for binding of bacterial
components at the site of bacterial invasion [13].
Intracellular signals trigger nuclear factor-kB (NFkB) to turn
on genes that promote the synthesis of proinflammatory
molecules, such as cytokines. The local release of these
Table 2. Markers of microbial infection.
Bacterial DNA/rDNA
Bacterial products
Innate immune response
Polymerase chain reaction
Fluorescent in situ hybridization
Other cell wall products
Neutrophils (left shift), cytokines, interleukin-6, complement activation
products, C-reactive protein, procalcitonin, neopterin, neutrophil products,
elastase, lactoferrin, secretory phospholipase A2, indicators of
substances results in neutrophil attraction and an
inflammatory response, which helps to compartmentalize
and finally eradicate the infection [12–17]. A spillover of
these factors into the blood is, conceivably, a function of
bacterial virulence and invasion and the associated host
defense, while bacteremia would trigger such responses
directly in the circulating blood and cells [14–16,18].
Bacterial products as markers of microbial infection
To avoid delay in treatment due to the time needed for
microbiological studies of secretions and blood from patients
with a suspected infection, investigators have tried to
pursue rapid molecular tests for microbial products. Used
together with imaging studies, these tests may help in the
diagnosis of microbial infection and the recognition of sepsis
(Table 2).
Bacterial products
The detection of circulating endotoxin has been pursued for
decades as a means to rapidly diagnose Gram-negative
bacteremia [19,20]. However, the bioassay used — the
chromogenic limulus amebocyte lysate test — does not
have 100% sensitivity and specificity for Gram-negative
bacteremia. Nevertheless, endotoxemia (>5 pg/mL)
measured with this assay, has been shown to be associated
with severe sepsis, shock, and organ failure, and therefore
seems of prognostic importance. A novel chemiluminescent
assay — the endotoxin activity assay — may have a higher
diagnostic performance [20]. However, their diagnostic
value in terms of patient management and outcome is still
unclear. This may also apply to the detection of fungal
antigenemia and products of Gram-positive infections,
including (antibodies to) exotoxins, enzymes and lipid S,
derived from the lipoteichoic acid component of the cell
membrane [21].
Bacterial DNA and RNA
Polymerase chain reaction (PCR) allows rapid amplication of
trace amounts of specific bacterial DNA or the common,
nonspecific, bacterial 16S rRNA gene from body secretions
as well as from blood [22–25]. There are some variants of
this method, i.e. multiplex PCR and real-time PCR, that can
be helpful in diagnosing specific infections, such as those
caused by pneumococci, meningococci, staphylococci and
Pseudomonas aeruginosa, and are also used as a broadrange test in the search for a likely causative organism in
nonspecific infections [23]. Some of these systems are more
sensitive than blood cultures [22,23,25] and may even
detect the presence of local infections and bacteria killed by
antibiotics. Real-time broad-range PCR should enable
quantification of bacterial load as a marker for the severity
of infection, which could help in deciding the dose and
duration of antimicrobial therapy in future studies. The
fluorescent in situ hybridization (FISH technique), using
rRNA-targeted oligonucleotide probes labeled with
fluorescent dyes, can rapidly identify >95% of the common
microorganisms in blood cultures, with almost 100%
sensitivity and specificity [26]. An advantage of this is the
short time needed for analysis (only 2.5 h) once growth has
been signaled in the blood culture machine. However, the
significance of these techniques for patient managment and
outcome remains unclear, although an early and more
precise diagnosis helps in clinical decision-making,
particularly when diseases with a high mortality, such as
meningococcal sepsis, are clinically suspected.
Proinflammatory factors as markers of microbial infection
There are numerous reports suggesting that the levels of
pro- and anti-inflammatory cytokines, and other mediators
associated with the innate immune response, are of
prognostic significance at various stages of infection and
suspected sepsis, and can predict organ dysfunction and
death [18,27,28]. Indeed, activated complement products
(C3a), interleukin-6 (IL-6) and IL-8, and secretory
phospholipase A2 (sPLA2), for instance, may be elevated,
particularly, in nonsurviving patients [18]. Indeed, both in
long-term survivors and nonsurvivors of sepsis, highly
elevated levels of circulating IL-6 may initially decline upon
treatment, although less so in the nonsurvivors than in
survivors. Secondary complications, such as superimposed
nosocomial infection, may re-elevate circulating levels as a
secondary proinflammatory response. For instance,
C-reactive protein (CRP) released as an acute phase protein
by the liver upon stimulation by IL-6, may elicit a secondary
Complement 3a (nmol/L)
IL-6 (pg/mL)
Proximal markers of the innate immune response:
Complement activation products and cytokines
Numerous reports suggest that circulating levels of pro- and
anti-inflammatory cytokines, associated with the innate
immune response, have prognostic value at various stages
of infection and sepsis. In fact, it is widely believed that an
exaggerated proinflammatory cytokine response, followed
by an anti-inflammatory response, is responsible, in part, for
the progression from infection and sepsis to shock and
death, even though trials on various immunomodulating
and anticytokine therapies have largely failed [18,27].
Authors have even used initial blood IL-6 levels
>1000 pg/mL, determined in a bedside test, as an inclusion
criterion for immunomodulating therapy with antitumor
necrosis factor (TNF) antibodies [27]. Unfortunately, this
approach has also failed to show clinical benefit [27].
Some circulating cytokines may help to predict microbial
infection and sepsis in less selected patient populations.
Circulating IL-6 levels (>54 pg/mL, normal value <10 pg/mL),
C3a levels (>14 nmol/L, normal value <5 nmol/L), and sPLA2
levels (>368 ng/mL, normal value <5 ng/mL) in hospitalized
adult febrile patients, of whom the overwhelming majority
had SIRS, predicted the presence of infection/bacteremia with
equal success, which was microbiologically confirmed within
1 week after inclusion (Fig. 1) [5,29]. These factors thus
predicted definite sepsis, and in this regard performed better
than the SIRS criteria of fever and leukocytosis [29].
Circulating C3a, IL-6, and IL-8 levels have diagnostic value for
definite sepsis in the critically ill, exceeding that of the SIRS
criteria [27,30–32]. In febrile children aged 0–36 months, a
high circulating level of IL-6 predicted occult bacteremia
better than clinical signs or traditional tests, while TNF-a and
IL-1 levels did not [33]. In patients with SIRS in an emergency
Figure 1. Circulating levels of complement activation
product C3a (normal value <5 nmol/L), IL-6 (normal
value <10 pg/mL), and sPLA2 (normal value <5 ng/mL)
in febrile medical adult patients without (▲) or with (•)
microbiologically proven infection, on days 1, 2, and 3 [29].
sPLA2 (ng/ml)
wave of complement activation, which further damages the
tissues and is associated with a poor outcome [18].
Although many mediators, such as IL-10 and several other
cytokines, have been thought to reflect the pro- and
anti-inflammatory status during sepsis, the value of these
factors in predicting outcome has not been agreed upon,
and is dependant on aspects such as the patient
populations and the factors studied. Some authors have
shown that cytokine levels are of primary prognostic
importance in local infections with sepsis, and that
complement activation is particularly associated with a
worse outcome, such as shock and death, in bacteremic
patients [18].
The value of pro- and perhaps anti-inflammatory factors
in predicting or contributing to the diagnosis of microbial
infection and its response to therapy with antimicrobial
agents is less clear, and is discussed below.
IL-6: interleukin-6; sPLA2: secretory phospholipase A2.
department, IL-6 and TNF-a were of predictive value for
bacteremia [28]. An IL-6 level >35 pg/mL on day 6 after
major trauma was a better predictor of nosocomial infection
than an increased CRP (>130 mg/L) [30]. IL-8, a potent
chemoattractant, may have similar diagnostic value to IL-6 in
febrile neutropenia: a value >2000 pg/mL (normal value
<20 pg/mL) had a positive predictive value for Gramnegative bacteremia of 73% and a negative predictive value
of 94%, normal levels allowed antibiotic treatment to be
limited safely in these patients [34]. Finally, the pattern of
release of interleukins may help to differentiate Gramnegative from Gram-positive infections [35], since the latter
are associated with greater IL-1b and IL-18 release.
Macrophage products: Procalcitonin and neopterin
Originally described in neonatal and pediatric sepsis,
elevated calcitonin precursor level in the blood
(procalcitonin, [PCT]) is evolving as a relatively new and
particularly early sepsis marker, although its role is still
controversial [36]. The precise mechanism involved in the
synthesis and release of this protein by macrophages upon
infectious stimuli is still uncertain. A rapid bedside test
(immunochromatography or immunofluorescence) has been
developed and gives a reliable estimate of circulating PCT
within 30 min. Although PCT has been shown to be of
value in predicting/diagnosing severe sepsis and an adverse
outcome, there are less data as to its early diagnostic value
for predicting localized or systemic infection [5,36–45].
Groeneveld et al. showed that in febrile medical patients,
most of whom had SIRS [5], PCT was elevated in those
patients who later had positive local and/or blood cultures,
and that bacteremia was particularly well predicted by
hyperprocalcitonemia rather than by the major SIRS criteria
of fever or leukocytosis [37]. Conversely, a normal PCT
(<0.4 mg/mL by immunoluminometric assay) may virtually
rule out bacteremia in febrile adults, while CRP may
discriminate less well [37,42].
A relationship between the severity of infection and
circulating PCT level has also been observed. Levels were
found to be lowest in patients without sepsis symptoms or
infection, and to be increasingly high in patients with
infection only, patients with sepsis symptoms and infection,
and patients with septic shock [32,45,46].
Although not yet beyond doubt, PCT may indeed be a
better marker than circulating CRP for severe
infection/bacteremia, definite sepsis and its severity in
critically ill patients and children [31,32,38–41,45–47].
However, the use of both parameters combined may
increase the predictive value. Furthermore, in critically ill
patients, circulating PCT predicted definite sepsis better than
IL-6, IL-8, or lactate levels [32,40]. PCT may even be helpful
in diagnosing new-onset microbial infection in the critically
ill [44]. In children, PCT/CRP seems a better diagnostic
marker for severe bacterial infections than IL-6, IL-8, or IL-1
receptor antagonist [41]. However, PCT, albeit sensitive,
may not be entirely specific for infection, since elevations
occur after, for example, surgery, trauma, burns, and
myocardial infarction, independent of microbial infection
[9,39,48–51]. Nevertheless, PCT elevations after cardiac
surgery were associated wth bacterial infection and less so
with rejection or viral infection [39,52]. Additionally,
circulating PCT seemed helpful in differentiating bacterial
infection from viral infection and from flares in systemic
lupus erythematosus [53].
Neopterin is produced by macrophages stimulated by
interferon-a, for instance, but its function has not been fully
elucidated. Plasma levels may be helpful in diagnosing
microbial infection in the critically ill patient, except after
trauma [44,51].
Neutrophil products
Elastase and lactoferrin are proteases released from the
azurophilic and specific granules, respectively, of neutrophils
during activation throughout the course of microbial
infection. Circulating levels of these markers are usually
elevated in sepsis, even in febrile patients, before infection
can be demonstrated by cultures [37]. In fact, circulating
elastase (measured as a complex with a1-antitrypsin) and
lactoferrin may have predictive value for culture-proven
infection and bacteremia in febrile hospitalized patients with
SIRS (Fig. 2). The predictive value of elevated levels of
elastase-a1-antitrypsin for bacteremia and mortality is
comparable with that of PCT [37]. These variables are thus
indicative of sepsis. Elastase may perform less well, however,
in the critically ill [32]. CD64, a membrane Fcg, is
upregulated on neutrophils during infections, and
membrane expression on these cells could be helpful in
discriminating between infections and flares of autoimmune
disease [54].
Coagulation markers
Raaphorst et al. demonstrated that early activation of
fibrinolysis, as determined by levels of circulating tissue
plasminogen activator (tPA), predicted microbial infection/
sepsis in febrile medical patients with SIRS, while levels of
coagulation indicators did not [55].
Acute-phase proteins
CRP is produced by the liver when stimulated by IL-6 [56].
It is one of the acute-phase proteins, the function of
which is not completely known. It has been suggested
to be an anti-inflammatory protein counteracting the
proinflammatory phase [12,56]. Recently, it has been
demonstrated that CRP, together with circulating sPLA2,
may help to activate complement on damaged membranes,
thereby helping the immune system to tag and eradicate
irreversibly damaged cells in the body [56]. It has been
suggested that, even after trauma and associated SIRS, a
high (or increasing) CRP level in the critically ill indicates
infection; the SIRS criteria, except for leukocytosis, further
Figure 2. Circulating levels of procalcitonin (normal value
<0.5 ng/mL), elastase-a1-antitrypsin (normal value
<100 ng/mL), and lactoferrin (normal value <400 ng/mL)
in febrile medical adults with (▲) or without (•)
microbiologically proven infection on days 0, 1, and
2 after inclusion [37].
Procalcitonin (ng/mL)
Elastase-a-antitrypsin (ng/mL)
Early in the management of the patient with suspected
infection, new molecular techniques for fast identification of
the pathogen, including monitoring of bacterial load,
together with specific markers of the local and circulating
inflammatory host response, will increasingly be used to
help identify microbial infection and determine its severity.
In fact, in febrile patients, levels of circulating cytokines,
complement activation products, PCT, CRP, lipid mediators,
and neutrophil degranulation products, have predictive
value for microbial infection and bacteremia and could be
used in future studies, when rapid determinations are
possible, to select patients for adjunctive therapeutic studies
with a higher a priori chance of infection than criteria
based on clinical signs and symptoms only [28]. Future
research should also focus on the value of various
combinations of factors to enhance their diagnostic value.
Finally, the value of these mediators in timing and selecting
antimicrobial therapy and predicting its effectiveness
deserves further evaluation.
The authors have no relevant financial interests to disclose.
1. Bone RC. Sepsis, the sepsis syndrome, multi-organ failure:
A plea for comparable definitions. Ann Intern Med
2. American College of Chest Physicians/Society of Critical Care
Medicine Consensus Conference: Definitions for sepsis and organ
Lactoferrin (ng/mL)
failure and guidelines for the use of innovative therapies in sepsis.
Crit Care Med 1992;20:864–74.
3. Levy MM, Fink MP, Marschall JC et al., for the International Sepsis
Definitions Conference. 2001 SCCM/ESICM/ACCP/ATS/SIS.
International sepsis definitions conference. Intensive Care Med
4. Rangel-Frausto MS, Pittet D, Cistigan M et al. The natural
history of the systemic inflammatory response syndrome (SIRS).
A prospective study. JAMA 1995;273:117–23.
5. Bossink AWJ, Groeneveld ABJ, Hack CE et al. The clinical host
response to microbial infection in medical patients with fever.
Chest 1999;116:380–90.
6. Alberti C, Brun Buisson C, Martin C et al. Epidemiology of sepsis
aid this prediction [9,28,48,57]. Hence, the course of CRP
may be a sensitive, rather than a specific tool to predict the
course of infection and sepsis in critically ill patients
[39,44,46,47,49,56,57]. Like CRP, sPLA2 is produced by the
liver as an acute-phase reactant upon stimulation by
IL-6 [29]. As with C3a and IL-6, circulating sPLA2 levels may
have some predictive value for microbial infection/
bacteremia in febrile patients [29].
and infection in ICU patients from an international cohort study.
Intensive Care Med 2002;28:108–21.
7. Huemer G, Graninger W, Maurotz W. Bed-side infection screening
of ICU patients using gram stained smears. Eur J Anaethesiol
8. Croce MA, Fabian TC, Waddle-Smith L et al. Utility of Gram’s
stain and efficacy of quantitative cultures for posttraumatic
pneumonia. Ann Surg 1998;227:743–55.
9. Miller PR, Munn DD, Meredith JW et al. Systemic inflammatory
25. Harris KA, Hartley JC. Develpoment of broad-range 16S rDNA PCR
response syndrome in the trauma intensive care unit: Who is
for use in the routine diagnostic clinical microbiolpogy service.
infected? J Trauma 1999;47:1004–8.
J Med Microbiol 2003;52:685–91.
10. Bates DW, Sands K, Miller E et al. for the Academic Medical
Center Consortium Sepsis Project Working Group. Predicting
bacteremia in patients with sepsis syndrome. J Infect Dis
26. Kempf VAJ, Trebesius K, Autenrieth IB. Fluoresecent in situ
hybridization allows rapid identification of microorganisms in
blood cultures. J Clin Microbiol 2000;38:830–8.
27. Reinhart K, Menges T, Gardlund B et al. The AFELIMOMAB
11. Heumann D, Glauser MOP, Calandra T. Molecular basis of
sepsis study group. Randomized, placebo-controlled trial of the
host-pathogen interaction in septic shock. Curr Opin Microbiol
anti-tumor necrosis factor antibody fragment afelimomab in
hyperinflammatory response during severe sepsis: The RAMSES
12. Oberholzer A, Oberholzer C, Moldawer LL. Sepsis syndromes:
Understanding the role of innate and acquired immunity.
Shock 2001;16:83–96.
study. Crit Care Med 2001;29:765–9.
28. Terrigino CA, Lopez BL, Karras DJ et al. Endogenous mediators
in emergency department patients with presumed sepsis:
13. Cohen J. The immunopathogenesis of sepsis. Nature
Are levels associated with progression to severe sepsis and death?
Ann Emerg Med 2000;35:26–34.
14. Schein M, Wittmann DH, Holzheimer R et al. Hypothesis:
29. Groeneveld ABJ, Bossink AWJ, Van Mierlo GJ et al.
Compartmentalization of cytokines in intraabdominal infection.
Circulating inflammatory mediators in patients with fever:
Surgery 1996;119:694–700.
Predicting bloodstream infection. Clin Diagn Lab Immunol
15. Schütte H, Lohmeyer J, Rosseau S et al. Bronchoalveolar and
systemic cytokine profiles in patients with ARDS, severe
30. Fassbender K, Pargger H, Müller W et al. Interleukin-6 and
pneumonia and cardiogenic pulmonary oedema. Eur Respir J
acute-phase protein concentrations in surgical intensive care
unit patients: Diagnostic signs in nosocomial infection.
16. Montón C, Torres A, El-Ebiary M et al. Cytokine expression in
severe pneumonia: A bronchoalveolar lavage study. Crit Care Med
Crit Care Med 1993;21:1175–80.
31. Selberg O, Hecker H, Martin M et al. Discrimination of sepsis
and systemic inflammatory response syndrome by determination
17. Muehlstedt SG, Richardson CJ, West MA et al. Cytokines and the
pathogenesis of nosocomial pneumonia. Surgery 2001;130:602–11.
18. Groeneveld ABJ, Tacx AN, Bossink AWJ et al. Circulating
inflammatory mediators predict shock and mortality in febrile
patients with microbial infection. Clin Immunol 2003;106:106–15.
19. Venet C, Zeni E, Viallon A et al. Endotoxaemia in patients with
severe sepsis or septic shock. Intensive Care Med 2000;26:538–44.
20. Marshall JC, Walker PM, Foster DM et al. Measurement of
endotoxin activity in critically ill patients using whole blood
neutrophil dependent chemiluminescence. Crit Care 2002;6:342–8.
21. Lambert PA, Worthington T, Tebbs SE et al. A novel
of circulating plasma concentrations of procalcitonin, protein
complement 3a, and interleukin-6. Crit Care Med
32. Harbarth S, Holeckova K, Froidevaux C et al. The Geneva sepsis
network. Diagnostic value of procalcitonin, interleukin-6, and
interleukin-8 in critically ill patients admitted with suspected sepsis.
Am J Respir Crit Care Med 2001;164:396–402.
33. Strait RT, Kelly KJ, Kurup VP. Tumor necrosis factor-a,
interleuklin-b, and interleukin-6 levels in febrile, young children
with and without occult bacteremia. Pediatrics 1999;104:1321–6.
34. Kern WV, Heiss M, Steinbach G. Prediction of Gram-negative
Staphylococcus epidermidis exocellular antigen with potential
bacteremia in patients with cancer and febrile neutropenia by
for the serodiagnosis of infections. FEMS Immunol Med Microbiol
means of interleukin-8 levels in serum: Targeting empirical
monotherapy versus combination therapy. Clin Infect Dis
22. Kane TD, Alexander JW, Johannignan JA. The detection of
microbial DNA in the blood. A senstive method for diagnosing
35. Freezor RJ, Oberholzer C, Baker HV et al. Molecular
bacteremia and/or bacterial translocation in surgical patients.
characterization of the acute inflammatory response to infections
Ann Surg 1998;227:1–9.
with Gram-negative versus Gram-positive bacteria. Infect Immun
23. Sleigh J, Cursons R, La Pine M. Detection of bacteraemia
in critically ill patients using 16S rDNA polymerase chain
reaction and DNA sequencing. Intensive Care Med
36. Reinhart K, Karzai W, Meisner M. Procalciton as a marker of the
systemic inflammatory response to infection. Intensive Care Med
24. Christensen JE, Stencil JA, Reed KD. Rapid identification of
37. Bossink AWJ, Groeneveld ABJ, Thijs LG. Prediction of microbial
bacteria from positive blood cultures by terminal restriction
infection and mortality in medical patients with fever: Plasma
fragment length polymorphism profile analysis of the 16S rRNA
procalcitonin, neutrophilic elastase-a1-antitrypsin, and lactoferrin
gene. J Clin Microbiol 2003;41:3790–800.
compared with clinical variables. Clin Infect Dis 1999;29:398–407.
38. Ugarte H, Silva E, Mercan D et al. Procalcitonin used as a
marker of infection in the intensive care unit. Crit Care Med
protein during the early posttraumatic systemic inflammatory
response syndrome. Intensive Care Med 1998;24:185–8.
39. Aouifi A, Piriou V, Bastien O et al. Usefulness of procalcitonin
for diagnosis of infection in cardiac surgical patients. Crit Care Med
49. Bitkover CY, Hansson L-O, Valen G et al. Effects of cardiac surgery
on some clinically used inflammation markers and procalcitonin.
Scand Cardiovasc J 2000;34:307–14.
40. Müller B, Becker KL, Schächinger H et al. Calcitonin precursors
50. Lecharny J-B, Khater D, Bronchard R et al. Hyperprocalcitonemia
are reliable markers of sepsis in a medical intensive care unit.
in patients with perioperative myocardial infarction after cardiac
Crit Care Med 2000;28:977–83.
surgery. Crit Care Med 2001;29:323–5.
41. Lacour AG, Gervaix A, Zamora SA et al. Procalcitonin, IL-6, IL-8,
IL-1 receptor antagonist and C-reactive protein as identificators
of serious bacterial infections in children with fever withouth
localising signs. Eur J Pediatr 2001;160:95–100.
51. Hensler T, Sauerland S, Lefering R et al. Neugenzuer EAM. The
clinical value of procalcitonin and neopterin in predicting sepsis and
organ failure after major trauma. Shock 2003;20:420–6.
52. Boeken U, Feindt P, Micek M et al. Procalcitonin (PCT) in cardiac
42. Chirouze C, Schuhmacher H, Rabaud C et al. Low serum
surgery: Diagnostic values in systemic inflammatory response
procalcitonin level accurately predicts the absence of
syndrome (SIRS), sepsis and heart transplantation (HTX).
bacteremia in adult patients with acute fever. Clin Infect Dis
Cardiovasc Surg 2000;8:550–4.
53. Brunkhorst FM, Eberhard OK, Brunkhorst R. Discrimination of
43. Duflo F, Debon R, Monneret G et al. Alveolar and serum
procalcitonin. Diagnostic and prognostic value in ventilatorassociated pneumonia. Anesthesiology 2002;96:74–9.
44. Ruokonen E, Ilkka L, Takala J. Procalcitonin and neopetrin as
indicators of infection in critically ill patients. Acta Anaesthesiol
Scand 2002;46:398–404.
infectious and noninfectious causes of early acute respiratory
distress syndrome by procalcitonin. Crit Care Med
54. Allen E, Bakke AC, Purtzer MZ et al. Neutrophil CD64 expression:
Distinguishing acute inflammatory autoimmune disease from
systemic infections. Ann Rheum Dis 2002;61:522–5.
45. Luzzani A, Polati E, Dorizzi R et al. Comparison of procalcitonin
and C-reactive protein as markers of sepsis. Crit Care Med
55. Raaphorst J, Groeneveld ABJ, Bossink AWJ et al. Early inhibition of
activated fibrinolysis predicts microbial infection, shock and mortality
in febrile medical patients. Thromb Haemost 2001;86:543–9.
46. Suprin E, Camus C, Gacoui A et al. Procalcitonin: A valuable
indicator of infection in a medical ICU? Intensive Care Med
56. Póvoa P. C-reactive protein: A valuable marker of sepsis.
Intensive Care Med 2002;28:235–43.
57. Reny J-L, Vuagnat A, Ract C et al. Diagnosis and follow-up of
47. Hatherill M, Tibby SM, Sykes K et al. Diagnostic markers of
48. Mimoz O, Benoist JF, Edouard AR et al. Procalcitonin and C-reactive
infections in intensive care patients: Value of C-reactive protein
infection: comparison of procalcitonin with C reactive protein
compared with other clinical and biological variables.
and leucocyte count. Arch Dis Child 1999;81:417–21.
Crit Care Med 2002;30:529–35.