The Efficacy of Procalcitonin as a Biomarker in the Management of Sepsis: Slaying Dragons or Tilting at Windmills?

Abstract

Background:

Sepsis is defined as systemic inflammatory response syndrome (SIRS) in the context of an underlying infectious process, and is associated with high rates of morbidity and mortality, particularly when initial therapy is delayed. Numerous biomarkers, including but not limited to cytokines (interleukins-2 and −6 [IL-2, IL-6] and tumor necrosis factor-α [TNF-α]), leukotrienes, acute-phase proteins (C-reactive protein [CRP]), and adhesion molecules, have been evaluated and rejected as unsuitable for the diagnosis of sepsis, predicting its severity, and guiding its treatment. Most recently, procalcitonin (PCT) has been suggested as a novel biomarker that may be useful in guiding therapeutic decision making in the management of sepsis. This article assesses critically the published literature on the clinical utility of PCT concentrations for guiding the treatment of sepsis in adult patients.

Methods:

A comprehensive search of all published studies of the use of serum concentrations of PCT to guide the treatment of sepsis in adult patients (1996 to 2011) was conducted with PubMed and Google Scholar. The search focused on the value of PCT concentrations to guide the diagnosis, prognosis, monitoring, and escalation and de-escalation of antbiotic therapy in these patients. Keywords searched included “procalcitonin,” “sepsis,” “sepsis biomarker,” “sepsis diagnosis,” “sepsis prognosis,” “sepsis mortality,” “antibiotic escalation,” “antibiotic de-escalation,” “antibiotic duration,” and “antimicrobial stewardship.”

Results:

Forty-six trials evaluating the efficacy of PCT concentrations in diagnosing sepsis have been published, with 39 of these trials yielding positive results and 7 yielding negative results. Wanner et al. published the largest study (n=405) demonstrating that peak PCT concentrations occur early after injury in both patients with sepsis and those with multiple organ dysfunction syndrome (MODS). Among 17 trials assessing the prognostic value of PCT concentrations with regard to clinical outcome and morbidity, 12 trials yielded positive results and five showed negative or equivocal results. Reith et al. published the largest study of the prognostic use of PCT concentrations (n=246), demonstrating that median PCT values on post-operative days (POD) one, four, and 10 were predictive of mortality in patients with abdominal sepsis (p<0.01). Among 14 trials of the utility of PCT concentrations for establishing an infectious cause of sepsis, 13 yielded positive results and only one yielded negative results. The largest study of this use of PCT concentrations, conducted by Baykut et al. (n=400), evaluated these concentrations in post-operative patients with infection, and demonstrated that concentrations of PCT remained elevated until POD 4, with a second increase observed between POD 4 and POD 6. In uninfected patients, PCT concentrations began to decrease on POD 2. Only a single study has assessed the utility of PCT concentrations in guiding the escalation of antibiotic therapy, and its results were negative. Specifically, Jensen et al. (n=1,200) compared a PCT-guided antibiotic escalation strategy with the standard of care for sepsis and found no difference in outcomes. They also found that the PCT group had a longer average stay in the intensive care unit (ICU), greater rates of mechanical ventilation, and a decreased estimated glomerular filtration rate (eGFR). Among four trials focusing on PCT concentrations and antibiotic de-escalation, all showed positive results with the measurement of PCT concentrations. The largest such study, by Bouadma et al. (n=621), demonstrated a four-day decrease in antibiotic duration when PCT concentrations were used to guide therapy relative to the study arm given the standard of care, with no increase in mortality (p=0.003).

Conclusions:

The diagnostic value of serum PCT concentrations for discriminating among SIRS, sepsis, severe sepsis, and septic shock remains to be established. Although higher PCT concentrations suggest a systemic bacterial infection as opposed to a viral, fungal, or inflammatory etiology of sepsis, serum PCT concentrations do not correlate with the severity of sepsis or with mortality. At present, PCT concentrations are solely investigational with regard to determining the timing and appropriateness of escalation of antimicrobial therapy in sepsis. Nevertheless, serum PCT concentrations have established utility in monitoring the clinical response to medical and surgical therapy for sepsis, and in surveillance for the development of sepsis in burn and ICU patients, and may have a role in guiding the de-escalation of antibiotic therapy.

Sepsis has represented historically a severe threat to public health and inpatient safety, as well as a leading cause of morbidity and mortality in hospitalized patients [1]. Sepsis was defined initially by the American College of Chest Physicians (ACCP)/Society of Critical Care Medicine (SCCM) Consensus Conference (1991) as consisting of measured changes in body temperature, heart rate, respiratory rate, and white blood cell count in the setting of proved concomitant blood-borne infection (Table 1) [2]. In 2001, the SCCM, European Society of Intensive Care Medicine (ESICM), ACCP, American Thoracic Society (ATS), and Surgical Infection Society (SIS) held the SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference and published a modified definition of sepsis [3]. Although this conference made no changes to the diagnostic criteria for sepsis, it created a new list of symptoms associated with sepsis, along with a list of prognostic factors and a rudimentary staging system for sepsis. This staging system, designated the predisposition, infection, response, and organ failure (PIRO) system, was designed along the lines of the tumor, node, metastasis (TNM) staging system for malignant tumors, with the goal of stratifying patients into risk groups (Table 2) [3]. The members of the International Sepsis Definitions Conference stated clearly that this new system was unsuitable for guiding clinical decision making, but noted that the new criteria might be used in future iterations, including tests for microbial products and biochemical markers [3].

Table 1.

American College of Chest Physicians and Society of Critical Care Medicine Criteria for Diagnosis of Sepsis [2]

Parameter	Value
Body temperature	• <36 °C (97 °F); or
	• >38 °C (100 °F)
Heart rate	• >90 beats/min
Respiratory rate	• >20 breaths/min; or
	• Blood gas PaCO₂<32 mm Hg (4.3 kPa)
WBC count	• <4,000 cells/mm³ (4x10⁹ cells/L); or
	• >12,000 cells/mm³ (12x10⁹ cells/L); or
	• >10% band forms (immature WBCs)
Infection	• Clinically suspected; or
	• Proved (by stain, culture, PCR)

PaCO₂=arterial blood partial pressure of carbon dioxide; PCR=polymerase chain reaction; WBC=white blood cell.

Table 2.

Society of Critical Care Medicine/European Society of Intensive Care Medicine/American College of Chest Physicians/American Thoracic Society/Surgical Infection Society International Sepsis Definitions Conference on the Predisposition, Infection, Response, and Organ-failure Staging System for Sepsis [3]

PIRO component	Current considerations	Future considerations	Justification
Predisposition	Pre-existing illness resulting in lower probability of survival. Cultural/religious beliefs, age, gender.	Genetic variability in the inflammatory response (IL-1, TNF, TLR, CD14); Interaction between pathogens and diseases harbored by host.	Current: Role of pre-existing illness in predicting morbidity and mortality in event of acute.Future: Negative impact of acute insult dependent on an individual's genetic makeup.
Insult/infection	Pathogenic culture and therapeutic sensitivity of pathogen; disease discovery and control of disease source.	Laboratory analysis of pathogen components/products (LPS, mannan, bacterial DNA); genetic transcription profiles.	Identification and characterization of an acute insult or disease process allows for the creation of specific therapeutic modalities aimed at counteracting the insult.
Response	SIRS, sepsis, septic shock, biomarker elevation (CRP).	Nonspecific inflammatory markers (PCT, IL-6), reduced response in host (HLA-DR); identification of unique therapeutic targets (protein C, TNF, PAF).	Risk of mortality and probability that the insult will respond to therapy are highly variable, and may be extrapolated from non-specific disease measures (disease severity metrics); Identification and understanding of pathophysiology of disease mediators allows for individually-targeted therapy against specific mediators.
Organ dysfunction	Organ dysfunction, based on number of failed organs or scoring systems (MODS, LODS, PEMOD, PELOD, SOFA, APACHE II).	Measurements reflecting response at cellular level: Apoptosis, hypoxia, stress.	Preemptive and early therapies (targeted therapies) are ineffectual if organ and cellular damage has already occurred; Identification of deleterious cellular processes allows for targeting of these specific processes.

APACHE II=Acute Physiology and Chronic Health Evaluation, CRP=C-reactive protein; HLA-DR=human leukocyte antigen-DR; IL=interleukin; LODS=logistic organ dysfunction system; LPS=lipopolysaccharide; MODS=multiple organ dysfunction syndrome; PAF=platelet-activating factor; PCT=procalcitonin; PELOD=pediatric logistic organ dysfunction; PEMOD=pediatric multiple organ dysfunction; PIRO=predisposition, infection, response, and organ failure; SIRS=systemic inflammatory response syndrome; SOFA=sepsis-related organ failure assessment; TLR=Toll-like receptor; TNF=tumor necrosis factor-α.

In the past several decades, epidemiologic studies have demonstrated a marked increase in the incidence of sepsis, especially in terms of rates of mortality in hospitalized patients following traumatic injury or surgery [1]. Martin et al. identified sepsis as the second leading cause of non-cardiac mortality in the ICU and as being among the top ten causes of mortality in the United States from 1979 to 2000. These authors also noted that the incidence of sepsis and sepsis-related mortality has increased steadily, especially in the ICU setting [1]. More than one-half of all cases of sepsis require treatment in the ICU, and 28-day mortality rates from sepsis in the ICU range from 20%–50%, with even higher rates in severe sepsis or septic shock [1]. The Agency for Healthcare Research and Quality reported that approximately $14.6 billion was spent on hospitalizations for sepsis from 1997 to 2008, with the cost of this increasing steadily at an average of 11.9% per year [4]. Moreover, hospitalizations for sepsis doubled from 2000 to 2008, whereas overall rates of hospitalization did not change, and most alarmingly, patients hospitalized for sepsis were eight times more likely than average to die during their hospital stay [5].

Despite the well-established increase in the incidence of and mortality from sepsis, numerous barriers to its effective management and treatment exist [1,6,7]. Perhaps the most daunting obstacle to its treatment has been establishing the diagnosis and differentiating sepsis from systemic inflammatory response syndrome (SIRS), because current diagnostic methods lack specificity for distinguishing bacterial infection from inflammatory states [6,8,9]. Treatment of sepsis has also been contentious, with differing opinions about the antibiotic regimen to be used, duration of treatment, and criteria for deciding when to start and stop antibiotic therapy [7]. Monitoring patients with sepsis and gauging their improvement is also problematic, as is the surveillance of non-infected patients for the development of sepsis [7,10].

Given the clinical importance of sepsis, considerable research directed at identifying a suitable biomarker with high sensitivity and specificity for its diagnosis and prognosis, as well as the capacity to guide therapeutic intervention for it in critically ill patients, continues to be a high priority [11]. The past few decades have witnessed the emergence of many potential biomarkers for sepsis, including but not limited to cytokines (i.e., interleukins-2, −6, and −8 [IL-2, IL-6, and IL-8], tumor necrosis factor-α [TNF-α]); cell markers; receptors; coagulation markers; markers of vascular endothelial damage, vasodilation and organ damage; leukotrienes; acute-phase proteins (i.e., C-reactive protein [CRP]); and adhesion molecules [12 –14]. Although initial studies of each of these various biomarkers suggested that they had great promise, subsequent studies indicated that nearly all of these biomarkers had insufficient sensitivity or specificity to make them useful clinically [12]. Pierrakos et al. reviewed 3,370 references encompassing 178 different biomarkers, and noted that nearly all of them have been used for prognostication rather than diagnosis, and that few if any have sufficient sensitivity and specificity for sepsis to permit their routine clinical use in its diagnosis and management [12].

Most recently, a novel biomarker, procalcitonin (PCT), has emerged at the forefront of studies of sepsis. Procalcitonin has been hailed by some as the “holy grail” of biomarkers for sepsis and a potential singular solution for its diagnosis, prognosis, and guidance of its treatment in adults [6]. Described initially in 1984, PCT is a polypeptide composed of 116 amino acids, with a total atomic mass of 14.5 kDa [13,14]. A precursor to calcitonin, PCT is produced primarily in the parafollicular C-cells of the thyroid gland in healthy individuals, in whom its serum concentrations typically remain under 0.1 ng/mL [13]. In the setting of a bacterial infection, cytokines (i.e., IL-6) and lipopolysaccharide (LPS) stimulate neutrophils and cells in the lungs, liver, intestine, and brain to produce PCT, resulting in its serum concentrations often exceeding 100 ng/mL [13]. By comparison with those of other markers, concentrations of PCT do not increase markedly during viral infection or other inflammatory conditions, although it has been stated that PCT can be increased in non-infectious conditions, such as major surgery, severe trauma, or burns [15]. Given its unique properties, PCT has been touted as superior to previously studied biomarkers for use in diagnostic, prognostic, and treatment algorithms of sepsis [16]. The past decade has seen the publication of a growing literature pertaining to the value of PCT in adult patients with sepsis [11]. The present review is intended to evaluate critically the existing literature on PCT and sepsis, specifically with regard to the five potential roles of serum PCT concentrations in the management of sepsis in adults of: (1) Its diagnosis, (2) prognosis, (3) monitoring, and (4) escalation and (5) de-escalation of antibiotic therapy for it. Specifically, this review examines the ability of PCT to guide therapeutic decision making in treatment of the adult patient with sepsis and to discuss the strengths, limitations, and future uses of PCT in the treatment of sepsis.

Methods

A comprehensive search was conducted of all studies published from 1996 to 2001 on the use of serum concentrations of PCT to guide the treatment of adult patients with sepsis, using PubMed and Google Scholar. The search focused on the value of PCT concentrations to guide the diagnosis, prognosis, monitoring, and escalation and de-escalation of antibiotic therapy in this population. Keywords searched included “procalcitonin,” “sepsis,” “sepsis biomarker,” “sepsis diagnosis,” “sepsis prognosis,” “sepsis mortality,” “antibiotic escalation,” “antibiotic de-escalation,” “antibiotic duration,” and “antimicrobial stewardship.”

Results

Many investigators regard PCT as the most clinically promising biomarker for the diagnosis, prognosis, monitoring, and therapeutic management of patients with sepsis. Each of these potential roles for serum concentrations of PCT has been studied extensively over the past several decades, and are addressed separately in the following sections.

The value of serum concentrations of procalcitonin in the diagnosis of sepsis

Since 1996, 46 prospective studies have been conducted of the utility of serum concentrations of PCT in the diagnosis of sepsis (Table 3) [8,9,17 –60]. Among these studies, 39 have demonstrated a statistically significant correlation between serum PCT concentrations and the diagnosis of sepsis [8,9,17 –23,28 –31,33 –45,47,48,50 –57,59,60]. Wanner et al. conducted the largest of these studies (n=405), and reported that although serum concentrations of PCT increased immediately after mechanical trauma in all patients, the highest concentrations of PCT on day 1, the day of admission, were seen in patients in whom either sepsis (6.9±2.5 ng/mL) or severe multiple organ dysfunction syndrome (MODS) (5.7±2.2 ng/mL) was diagnosed subsequently. Moreover, they noted a sustained increase in serum concentrations of PCT over the first 14 d of hospitalization in the groups with sepsis (p<0.05) as compared with patients who had an uneventful course of hospitalization (1.1±0.2 ng/mL) [17].

Table 3.

Summary of Prospective Studies of Procalcitonin Serum Levels in the Diagnosis of Sepsis in Adult Patients

Study, year	No. of patients	PCT Cutoff (ng/mL)	Sensitivity (%)	Specificity (%)	Area under ROC	Study setting and outcomes
Al-Nawas et al., 1996 [33]	337	0.5 ng/mL	60%	79%	—	Study setting: Medical ward and ICUInfection likely with PCT>0.5 ng/mL, high sensitivity, and specificity
De Werra et al., 1997 [34]	29	—	—	—	—	Study setting: ED and MICUPCT concentration markedly increased in septic shock (72–135 ng/mL) as compared with IL-6, TNF-α, nitrites/nitrates (∼1 ng/mL)PCT and nitrites/nitrates best predictors of septic shock
Benoist et al., 1998 [35]	21	5.0 ng/mL	100%	100%	—	Study setting: SICUPCT ratio (ratio of actual value to normal value) <10x in SIRS, 30x to 436x in sepsis (p=0.002)^*PCT superior to CRP in differentiating SIRS from sepsis
Whang et al., 1998 [36]	29	1.08 ng/mL	67%	80%	0.77 (p<0.05)	Study setting: MSICUPeak PCT levels higher in patients with infection documented by blood cultures than in those without infection (p<0.05)
Bossink et al., 1999 [37]	200	0.5 ng/mL	65%	58%	0.61 (p<0.05)	Study setting: Medical ward and ICUInitial PCT higher in patients with bacteremia (1.80 ng/mL [range 0.08–278.6 ng/mL]) than in patients with localized infection (0.73 ng/mL [range 0.08–34.7 ng/mL]) and patients without infection (0.33 ng/mL [range 0.08–485 ng/mL]), p<0.001Peak PCT higher in patients with bacteremia (4.10 ng/mL [range 0.08–502 ng/mL]) than in patients with localized infection (1.10 ng/mL [range 0.08–73.9 ng/mL]) and patients without infection [0.56 ng/mL [range 0.08–615 ng/mL]), p<0.001
Rothenburger et al., 1999 [38]	59	4.0 ng/mL	86%	98%	—	Study setting: SICUSignificant increase in PCT after onset of infection (10.86 ng/mL [3.28–15.13 ng/mL]) vs. baseline (0.12 ng/mL [range 0.08–0.21 ng/mL])
Ugarte et al., 1999 [24]	205	0.6 ng/mL	67.6%	61.3%	0.66 (p<0.05)	Study setting: MSICUPCT inferior to CRP in diagnosing sepsisCRP sensitivity 71.8%, specificity 66.6%, AUC 0.78 (p<0.05)
Adamik et al., 2000 [39]	83	—	—	—	>0.8 (p<0.05)	Study setting: MICUDevelopment of sepsis correlated with increasing PCT and neopterin concentrations (p<0.05)
Aouifi et al., 2000 [40]	97	1.0 ng/mL	85%	95%	0.82	Study setting: SICUSerum PCT concentrations higher in septic shock (96.98±119.61 ng/mL) than in uninfected patients (0.41±0.36 ng/mL)Best PCT cutoff for predicting infection was 1 ng/mL (sensitivity 85%, specificity 95%, positive predictive value 96%, negative predictive value 84%)PCT superior to CRP for diagnosis of sepsis (ROC AUC 0.82 for PCT vs. 0.68 for CRP)
Baykut et al., 2000 [41]	400	—	—	—	—	Study setting: SICUIn post-operative patients without infection, PCT increased until POD 2, decreasing thereafterIn patients with infection, PCT remained elevated until POD 4, with a second increase between POD 4 and POD 6 (p<0.001)
Boeken et al., 2000 [42]	74	—	—	—	—	Study setting: SICUIn post-operative patients with sepsis, mean PCT concentrations was 19.6±6.2 ng/mL, vs. 0.33±0.15 ng/mL in uninfected patients and 0.7±0.4 ng/mL in SIRS patients (p<0.05)PCT capable of reliably diagnosing sepsis and differentiating bacterial infection from fungal and viral infection
Brunkhorst et al., 2000 [18]	185	2.0 ng/mL	96%	86%	—	Study setting: MICUDay 1 PCT highest in septic shock patients (12.89±4.39 ng/mL, p<0.05)Day 1 PCT higher in severe sepsis than in sepsis (6.91±3.87 ng/mL vs. 0.53±2.9 ng/mL, p<0.001)Day 1 PCT higher in severe sepsis than in SIRS (6.91±3.87 ng/mL versus 0.41±3.04 ng/mL, p<0.001)
Cheval et al., 2000 [43]	60	5.0 ng/mL	88%	67%	—	Study setting: MSICUPCT level higher in patients with proved bacterial infection (72±153 ng/mL vs. 2.9±10 ng/mL, p=0.0003)In patients with shock, PCT higher when bacterial infection was diagnosed (89±154 ng/mL vs. 4.6±12 ng/mL, p=0.0004)PCT vs. CRP in logistic regression, PCT was unique marker of infection in patients with shock (with PCT≥5 ng/mL, OR, 6.2; 95% CI 1.1–37.0, p=0.04)
		20 ng/mL	88%	82%
Mülleret al., 2000 [28]	101	1.0 ng/mL	89%	94%	—	Study setting: MICUROC analysis showed PCT was most reliable laboratory variable for diagnosis of sepsis as compared with CRP, IL-6, lactate (p<0.01, for each comparison)
Oberhoffer et al., 2000 [44]	242	1.35 ng/mL	84%	83%	0.846 (p<0.05)	Study setting: SICUPCT levels correlated strongly with TNF-α>40 pg/mL (ROC AUC 0.846) and IL-6>500 pg/mL (ROC AUC 0.837, p<0.05)PCT superior to CRP, leukocyte count and body temperature when used in conjunction with TNF-α and IL-6 for identifying sepsis
Reith et al., 2000 [31]	312	0.8 ng/mL	—	—	—	Study setting: SICUMedian pre-operative PCT 2.05 ng/mL in surviving vs. 4.2 ng/mL in non-surviving (p=0.08)Median POD 1 PCT 4.9 in surviving vs. 13.8 in non-surviving (p<0.01)Median POD 4 PCT 4.8 in surviving vs. 13.0 in non-surviving (p<0.01)Median POD 10 PCT 0.4 in surviving vs. 13.25 in non-surviving (p<0.01)PCT useful in detecting sepsis post-operatively
Selberg et al., 2000 [29]	33	3.3 ng/mL	86%	54%	—	Study setting: MICUMedian PCT concentrations higher in sepsis (16.8 ng/mL [range 0.9–351.2 ng/mL]) than in SIRS (3.0 ng/mL [range 0.7–29.5 ng/mL], p=0.003)Combining PCT and C3a values resulted in a 91% sensitivity and 80% specificity for differentiating sepsis from SIRSPCT superior to CRP and elastase in differentiating sepsis from SIRS
Suprin et al., 2000 [25]	101	2.0 ng/mL	65%	70%	—	Study setting: MICUPCT higher in septic shock (38.5±59.1 ng/mL) than in SIRS (3.8±6.9 ng/mL), sepsis (1.3±2.7 ng/mL) and severe sepsis (9.1±18.2 ng/mL) (p=0.005)PCT had poor sensitivity and specificity for sepsis diagnosisPCT did not clearly differentiate SIRS from sepsis or severe sepsis
Wanner et al., 2000 [17]	405	1.5 ng/mL	75.5%	77%	—	Study setting: ED and trauma ICUHighest PCT concentration early after injury observed in patients with sepsis (6.9±2.5 ng/mL) or severe MODS (5.7±2.2 ng/mL) as compared with patients with an uneventful post-traumatic course (1.1±0.2 ng/mL)
Harbarth et al., 2001 [9]	78	1.1 ng/mL	97%	78%	0.92 (95% CI 0.85,1.0, p<0.001)	Study setting: MSICUPCT superior to IL-6 and IL-8 in differentiating between sepsis stages
Yukioka et al., 2001 [45]	35	—	—	—	—	Study setting: MICUMedian PCT levels were 0.6 ng/mL (range 0.1–3.4 ng/mL) in SIRS, 5.4 ng/mL (range 0.9–47.7 ng/mL) in sepsis and 73.4 ng/mL (range, 9.6–824.1 ng/mL) in septic shockGood correlation between PCT level and SOFA score (rs=0.766†, p<0.0001)PCT superior to CRP in differentiating between SIRS, sepsis, and septic shock
Baumgarten et al., 2002 [46]	35	3.0 ng/mL	55%	88%	—	Study setting: ICUPCT does not provide additional diagnostic information when sepsis determined by Bone's criteria
Giamarellos-Bourboulis et al., 2002 [8]	119	1.0 ng/mL	94.5%	64.5%	—	Study setting: MSICUMean D1 PCT levels 5.45 ng/mL (95% CI 2.11–8.81 ng/mL), 7.29 ng/mL (95% CI −1.92–14.59 ng/mL), 6.26 ng/mL (95% CI −1.32–13.85 ng/mL), and 38.76 ng/mL (95% CI 0.15–77.38 ng/mL) in patients with SIRS, sepsis, severe sepsis, and septic shock, respectivelyROC curves of sensitivity and specificity of PCT for evaluation of SIRS and sepsis were similarDifferentiation between SIRS and sepsis may not rely on a single application of PCT, but on the complete clinical and laboratory evaluation of the patient with PCT playing a role
Hausfater et al., 2002 [47]	195	0.2 ng/mL	62%	88%	0.79 (p<0.05)	Study setting: EDPCT concentration of 10.5 ng/mL and temperature of>38 °C were associated with systemic infection (CRP not associated with systemic infection).In multivariable analysis, only variable associated with systemic infection was PCT level (OR 44; 95% CI 8,1000; p=0.0004)
Meisner et al., 2002 [48]	208	2.0 ng/mL	87%	78%	0.90 (p<0.05)	Study setting: SICUMore patients with PCT>2 ng/mL on POD 1 or POD 2 (n=55) had post-operative abnormalities (95%) than patients with lower PCT (59%)Proven bacterial infection rate 9% when PCT>2 ng/mL, vs. 1% when PCT lower (ROC AUC 0.90)Suspected bacterial infection rate 24% when PCT>2 ng/mL, vs. 1% when PCT lower (ROC AUC 0.897)PCT superior to CRP in identifying post-operative infection
Ruokonen et al., 2002 [49]	208	0.8 mcg/L (on D1)	67.7%	47.8%	—	Study setting: MSICUNeopterin in febrile patients: D1 cutoff 18 pg/L, D2 cutoff 16 pg/L, D1 sensitivity 62.7%, D2 sensitivity 69.3%, D1 specificity 78.3%, D2 specificity 67.9%PCT and neopterin equally effective (but poor accuracy) in differentiating sepsis from SIRSNeopterin more specific than PCT
		0.9 mcg/L (on D2)	60.9%	63%
Tugrul et al., 2002 [50]	85	1.31 ng/mL	73%	83%	—	Study setting: MSICUPCT concentrations significantly higher in patients with severe sepsis and septic shock (19.25±43.08 ng/mL and 37.15±61.39 ng/mL, p<0.05) than patients w/ SIRS (0.73±1.37 ng/mL, p<0.05)PCT concentrations higher in septic patients (21.9±47.8 ng/mL), than in SIRS patients regardless of degree of sepsis (p<0.001)
Balci et al., 2003 [19]	33	2.42 ng/mL	85%	91%	0.969±0.016 (p<0.001)	Study setting: MSICUPCT superior to CRP, IL-2, IL-6, IL-8, TNF-α in differentiating sepsis from SIRS
		2.97 ng/mL	82%	100%
De Talance et al., 2003 [51]	108	2.0 ng/mL	91%	89%	—	Study setting: MICU
Dorge et al., 2003 [32]	80	5.0 ng/mL	63%	62%	0.720 (95% CI 0.603–0.837)	Study setting: SICUPost-operative PCT increased in non-survivors vs. survivors (34.3±7.0 ng/mL vs. 15.9±4.9 ng/mL, p<0.05), in patients with severe complications (30.3±6.7 ng/mL vs. 5.5±1.4 ng/mL, p<0.05) and in patients with infections (38.4±11.3 ng/mL vs. 10.8±1.6 ng/mL, p<0.05)ROC AUC for PCT as predictor of mortality, infections, and complications was 0.772 (95% CI, 0.651–0.894), 0.720 (95% CI 0.603–0.837) and 0.861 (95% CI 0.779–0.943), respectivelyPCT was not different with infectious compared to non-infectious complications
Du et al., 2003 [52]	51	1.6 ng/mL	80%	74%	0.923 (p<0.05)	Study setting: MSICUIL-6, PCT superior to CRP, body temperature in differentiating sepsis from SIRSSepsis score combining IL-6 and PCT best for differentiating sepsis from SIRS (ROC AUC 0.923)
Geppert et al., 2003 [53]	55	2.0 ng/mL	87%	75%	0.86 (p<0.05)	Study setting: Cardiac ICU25% of uninfected CS patients had PCT>2.0 ng/mLMedian PCT in CS patients higher than controls, but lower than septic shock (SS) patientsPCT superior to CRP in diagnosing CS and SS (PCT ROC AUC 0.86 vs. CRP ROC AUC 0.83)PCT superior to CRP in diagnosing sepsis, but PCT conc. of 10.0 ng/mL more appropriate cutoff value than 2.0 ng/mL
Hensler et al., 2003 [54]	137	1.5 ng/mL	42%	73%	—	Study setting: Trauma ICU/SICUPCT and neopterin levels D1 and D2 after major traumatic injury were unable to differentiate between patients who developed sepsis and those who did notPatients who developed MODS (vs. control) had higher PCT levels on D0 (0.60 ng/mL vs. 0.15 ng/mL), and on D1 and D2 combined (1.95 ng/mL vs. 0.32 ng/mL)High PCT concentration in the early post-traumatic phase is an independent predictor of MODS but not of sepsis
Kabir et al., 2003 [55]	15	—	—	—	—	Study setting: MSICUSerum IL-18 and PCT levels were elevated in septic patients (p<0.05), whereas TNF-α, IL-1β, IL-6, IL-10, IFN-γ, and MCP-1 levels did not changePCT is a clear candidate marker for sepsis criteria
Luzzani et al., 2003 [20]	70	2.0 ng/mL	76%	84%	0.925 (95% CI 0.899–0.952, p<0.01)	Study setting: MSICUMedian PCT in SIRS vs. sepsis 0.4 and 3.65 ng/mL (p<0.0001)PCT superior to CRP in differentiating sepsis from SIRS
Castelli et al., 2004 [30]	150	1.11 ng/mL	79%	85%	—	Study setting: MSICUPCT concentrations were 1.58 ng/mL sepsis patients, 0.38 ng/mL in SIRS patients, and 0.14 ng/mL in patients w/o SIRS (p<0.05)Different sensitivities and kinetics indicate different clinical uses for PCT and CRP
Chan et al., 2004 [56]	69	0.6 ng/mL	71%	67%	0.689 (p<0.05)	Study setting: EDCRP: cutoff 60mg/L, sensitivity 67.2%, specificity 93.9%CRP AUC 0.879 (p<0.05)CRP superior to PCT for identifying infectionPCT concentrations higher in bacteremia and septic shock
Clec'h et al., 2004 [57]	75	1.0 ng/mL	95%	54%	—	Study setting: MSICUD1 median PCT concentration significantly higher in patients with than without septic shock (14.0 ng/mL [range 0.3–767 ng/mL] vs. 1.0 ng/mL [range 0.5–36 ng/mL], p<0.01)PCT may be a valuable early diagnostic marker in patients with septic shock
Gibot et al., 2004 [58]	76	0.6 ng/mL	83%	69%	—	Study setting: MICUPlasma soluble TREM-1 level >60 ng/mL more accurate than PCT for indicating infection (sensitivity 96% [95% CI 92%–100%], specificity 89% [95% CI 82%–95%])
Aikawa et al., 2005 [59]	176	0.5 ng/mL	64%	86%	0.84 (p<0.001)	Study setting: EDPCT superior to endotoxin, IL-6, CRP in differentiating sepsis from SIRS
Mokart et al., 2005 [60]	50	1.1 ng/mL	81%	72%	—	Study setting: ICUPCT cutoff of 1.1 ng/mL had 81% sensitivity, 72% specificityWhen PCT associated with occurrence of SIRS on day 1, 100% sensitivity and 86% specificity for sepsis
Gaïni et al., 2006 [26]	194	0.087 ng/mL	79.7%	42.1%	0.75 (95% CI 0.63–0.87, p<0.01)	Study setting: Medical wardIL-6, LBP, CRP superior to PCT in differentiating sepsis from SIRSIL6: cutoff 25 pg/mL, sensitivity 81.1%, specificity 78.9%IL-6 AUC 0.87 (95% CI 0.78–0.96)LBP: cutoff 20 μg/mL, sensitivity 81.0%, specificity 68.4%LBP AUC 0.86 (95% CI 0.77–0.95)CRP: cutoff 38 mg/L, sensitivity 79.7%, specificity 57.9%CRP AUC 0.84 (95% CI 0.75–0.92)PCT superior to IL-6, LBP, CRP in differentiating sepsis from severe sepsis, AUC 0.74 (95% CI 0.61–0.87, p<0.01)
		0.5 ng/mL	45.9%	89.5%
Endo et al., 2008 [21]	82	2.0 ng/mL	94.7%	78.1%	—	Study setting: ED and ICUMedian PCT values for sepsis and severe sepsis 0.6 mg/mL and 36.1 ng/mL, respectivelyPCT concentrations could differentiate between sepsis and severe sepsis (sensitivity 94.7%, specificity 78.1%)
Tsangaris et al., 2009 [22]	50	1.0 ng/mL	70%	91%	0.85 (95% CI 0.71–0.93, p<0.0001)	Study setting: ICUMedian PCT on day of fever 1.18 ng/mL and 0.17 ng/mL in patients with and w/o proven infection, respectively (p<0.001)PCT AUC superior to AUC of CRP and WBC countPCT cutoff of 0.15 ng/mL=sensitivity 96%PCT cutoff of 1.16 ng/mL=specificity 100%Doubling of PCT between fever onset and day before was assoc. w/ proved infection (p<0.01)
Meynaar et al., 2011 [23]	76	2.0 ng/mL	97%	80%	0.95 (95% CI 0.90–0.99, p<0.05)	Study setting: MSICUPCT superior to LBP, CRP, IL-6 in differentiating sepsis from SIRS
		10.0 ng/mL	66%	93%
Sudhir et al., 2011 [27]	100	0.5 ng/mL	94%	—	—	Study setting: ICUNo association between PCT concentrations and sepsis grade (p>0.05)Association between high PCT concentrations and high SOFA score (p<0.05)

assoc.=association; AUC=area under the curve; CI=confidence interval; CRP=C-reactive protein; CS=cardiogenic shock; D=day; ED=emergency department; ICU=intensive care unit; IL=interleukin; LBP=lipopolysaccharide binding protein; MICU=medical intensive care unit; MODS=multiple organ dysfunction syndrome; MSICU=medical/surgical intensive care unit; N=number of patients; OR=odds ratio; PCT=procalcitonin; POD=postoperative day; ROC=receiver operating curve; SICU=surgical intensive care unit; SIRS=systemic inflammatory response syndrome; SOFA=Sequential Organ Failure Assessment; TNF=tumor necrosis factor; TREM-1=triggering receptor expressed on myeloid cells-1; w/o=without.

—: unreported value

†

rs value: Spearman rank order correlation.

p<0.05

In contrast to the study of Wanner et al., seven prospective studies (Ugarte et al. [24], Suprin et al. [25], Sudhir et al. [49], Dorge et al. [32], Baumgarten et al. [44], Ruokonen et al. [49], and Gibot et al. [58]) reported that serum concentrations of PCT had poor diagnostic value for sepsis [24,25,27,32,46,49,58]. Gaïni et al. reported mixed results, finding that IL-6, lipopolysaccharide-binding protein (LBP), and CRP were superior to PCT for differentiating between SIRS and sepsis (area under the receiver-operating characteristic curve [AUC] 0.84, p<0.01), whereas PCT was better for discriminating between severe sepsis and sepsis (AUC 0.74, p<0.01) [26]. A review by Pierrakos et al. compiled the results in 3,370 references to studies of 178 different biomarkers including PCT, and concluded that no biomarker had sufficient sensitivity and specificity to justify its use in clinical practice. [12]

The value of serum concentrations of procalcitonin in prognostication in sepsis

Since 1998, 17 prospective studies have been conducted of the prognostic utility of serum concentrations of PCT in sepsis, of which 12 have shown a statistically significant correlation between its serum concentrations and disease severity/outcome (Table 4) [22,24,27 –31,36,37,39,43,45,47,57,61 –63]. Reith et al. conducted the largest of these studies (n=246), and reported higher median PCT concentrations in patients with abdominal sepsis who did not survive post-operatively than in patients who did survive (p<0.01 on POD 1, POD 4, and POD 10), concluding that median PCT concentrations were predictive of increased mortality from sepsis [31]. By contrast, five prospective studies (Sudhir et al. [27], Whang et al. [36], Selberg et al. [29], Castelli et al. [30], and Dahaba et al. [63]) found poor or no correlation between serum concentrations of PCT and prognosis in sepsis [27,29,30,36,63]. Assicot et al. reported a correlation between serum PCT concentrations and severity of infection, noting PCT values of up to 200 ng/mL in patients with septic shock, but no data were shown in support of this conclusion [64]. Lastly, a review by Becker et al. noted that high serum concentrations of PCT correlated poorly with prognosis in both the presence and absence of infection, but did observe certain exceptions, pointing out the unreliability of borderline values as predictors of outcome, as well as the poor prognostic value of early measurements of serum PCT [65]. Becker et al. also reported that serial measurements of PCT and PCT kinetics were more useful in determining disease severity and prognosis [65].

Table 4.

Summary of All Prospective Studies of Procalcitonin Serum Levels and Dynamics in Predicting Sepsis Severity or Mortality

Study, Year	No. of patients	Mortality	Study parameters	Study setting and outcomes
Whang et al., 1998 [36]	29	13 (45%)	PCT ROC AUC vs. APACHE II score ROC AUC	Study setting: MSICUAPACHE II score>24 associated with 100% mortalityAPACHE II score<21 associated with 90% survivalAPACHE II score ROC AUC showed 21 was best cutoff for mortality (sensitivity 92%, specificity 94%)APACHE II ROC AUC 0.94±0.05PCT ROC AUC 0.77±0.09 APACHE II score superior to PCT as predictor of mortality
Bossink et al., 1999 [37]	200	17 (9%)	Peak PCT vs. survival	Study setting: Medical ward and ICUPeak PCT concentration in survivors 0.81 (range 0.08–615) vs. 2.65 (range 0.08–47.2) in non-survivors (p<0.05)^*AUC for mortality 0.79 for respiratory rate, 0.69 for elastase-α₁-antitrypsin level, 0.65 for heart rate, 0.61 for PCT level, and 0.60 for WBC count
Ugarte et al., 1999 [24]	111	53 (28%)	Median PCT vs. survival	Study setting: MSICUMedian PCT concentrations in survivors (first 96 h) 0.9 ng/mL vs. 3.4 ng/mL in non-survivors (p<0.05)
Adamik et al., 2000 [39]	83	—	PCT ROC AUC vs. nitrite/nitrate level AUC	Study setting: MICUPCT ROC AUC and neopterin ROC AUC>0.8 (p<0.05)Nitrite/nitrate ROC AUC<0.8PCT and neopterin superior to nitrite/nitrate as predictors of mortality
Cheval et al., 2000 [43]	60	—	PCT vs. survival	Study setting: MSICUPCT conc. in survivors 24.0 g/mL vs. 71.3 g/mL in non-survivors (p=0.006)
Müller et al., 2000 [19]	101	23 (23%)	Mean PCT vs. survival	Study setting: MICUMean PCT concentration in survivors 17.4±41.4 ng/mL vs. 33.5±55.1 ng/mL in non-survivors (p=0.01)
Reith et al., 2000 [31]	246	59 (24%)	Median PCT vs. survival	Study setting: SICUMedian preoperative PCT 2.05 ng/mL in surviving vs. 4.2 ng/mL in non-surviving (p=0.08)Median POD 1 PCT 4.9 in surviving vs. 13.8 in non-surviving (p<0.01)Median POD 4 PCT 4.8 in surviving vs. 13.0 in non-surviving (p<0.01)Median POD 10 PCT 0.4 in surviving vs. 13.25 in non-surviving (p<0.01)Median PCT predictive of mortality in postoperative abdominal infection patients
Selberg et al., 2000 [29]	101	16 (16%)	Median PCT vs. survival	Study setting: MICUNo difference in median PCT between survivors and non-survivors (data NS)
Yukioka et al., 2001 [45]	35	—	Median PCT vs. SOFA score	Study setting: MICUGood correlation between PCT level and SOFA score (rs=0.766^†, p<0.0001)PCT concentration is good indicator of sepsis severity and organ failure
Hausfater et al., 2002 [47]	195	6 (3%)	Mean PCT vs. survival	Study setting: EDMean PCT concentration in survivors 1.5 ng/mL vs. 17.8 ng/mL in non-survivorsPatients with PCT>0.5 ng/mL more likely to die of systemic infection within 1 mo after admission (OR 31.6; 95% CI 2.9,1575, p=0.001)
Castelli et al., 2004 [30]	150	29 (19%)	Mean PCT vs. SOFA score	Study setting: MSICUWeak correlation (r=0.254^††, p<0.001) between PCT and SOFA scoreHigher mean PCT values in patients with higher SOFA scores (p<0.05)
Clec'h et al., 2004 [57]	75	—	Median PCT vs. survival	Study setting: MSICUPCT levels in septic patients higher in non-survivors than in survivors (16.0 ng/mL [0.15–767 ng/mL] vs. 6.0 ng/mL [0.2–123 ng/mL] on D1, p=0.045; 6.5 ng/mL [0.3–135 ng/mL] vs. 1.05 ng/mL [0.11–53 ng/mL] on D10, p=0.02)A D1 PCT cutoff value of 6 ng/mL separated patients who died from those who survived with 87.5% sensitivity and 45% specificity.PCT may be a valuable prognostic marker in patients with septic shock
Dahaba et al., 2006 [63]	69	18 (26%)	PCT ROC AUC vs. SOFA score ROC AUC in predicting mortality	Study setting: SICUMedian PCT in survivors 1.8 (95% CI 0.4–4.4) ng/mL vs. 16 (95% CI 6.1–32.9) ng/mL in non-survivors (p<0.005)PCT AUC 0.78 (day 3), 0.90 (day 6), vs. SOFA score AUC 0.85 (95% CI 0.78–0.94) (day 3)SOFA score better predictor of mortality than PCT
Charles et al., 2009 [61]	180	51 (28%)	PCT dynamics vs. survival	Study setting: MICUDay 3 PCT in survivors 21.3 ng/mL vs. non-survivors 40.8 ng/mL, p=0.04≥30% decrease in PCT between D2 and D3 is an independent predictor of survival (OR 2.94, 95% CI 1.22–7.09, p<0.02)
Tsangaris et al., 2009 [27]	50	22 (44%)	Median PCT vs. survival	Study setting: ICUMedian PCT concentration <0.5 ng/mL on D3 after advent of fever associated survival (p<0.01)
Karlsson et al., 2010 [62]	242	62 (26%)	PCT dynamics vs. survival	Study setting: ICU>50% decrease in PCT between D0 and D3 associated with decreased mortality (p=0.007)
Sudhir et al., 2011 [27]	100	23 (23%)	PCT vs. SOFA score	Study setting: ICUHigher SOFA score associated with higher serum PCT (p<0.05)No association between PCT at presentation and mortality (p>0.05)

APACHE II=Acute Physiology and Chronic Health Evaluation II; AUC=area under the curve; CI=confidence interval; D=day; ED=emergency department; ICU=intensive care unit; MICU=medical intensive care unit; MSICU=medical/surgical intensive care unit; N=number of patients; NS=not significant; OR=odds ratio; PCT=procalcitonin; ROC=receiver operating curve; SICU=surgical intensive care unit; SOFA=Sequential Organ Failure Assessment; WBC=white blood cell.

—: unreported value.

†

rs value: Spearman rank order correlation.

††

r value: Pearson correlation.

Significant at p<0.05

The value of serum procalcitonin concentrations in the monitoring of sepsis

Fouteen prospective trials assessing the utility of serum PCT concentrations in monitoring for complications of sepsis have been conducted since 1998 [17,31,32,35,38,41,42,48,54,60,66 –69]. Thirteen of these trials identified a statistically significant benefit of using serum PCT concentrations in monitoring for sepsis (Table 5) [17,31,35,38,42,48,54,60,66 –69]. Baykut et al. conducted the largest of these studies (n=400), and reported that serum concentrations of PCT increased until POD 2 and decreased thereafter in patients without post-operative infectious complications, whereas serum concentrations of PCT increased until POD 4–6 in cases of post-operative infection (p<0.001) [41]. Baykut et al. concluded that serum concentrations of PCT were useful in diagnosing post-operative infection after cardiac surgery [41]. In contrast, only a single study, conducted by Dorge et al., found no useful correlation between serum concentrations of PCT and monitoring for events related to sepsis [32].

Table 5.

Summary of All Prospective Studies Evaluating Procalcitonin Serum Concentrations in Sepsis Monitoring

Study, year	No. of patients	Study parameters	Study setting and outcomes
Benoist et al., 1998 [35]	21	PCT as identifier of sepsis in trauma patients	Study setting: SICUPCT ratio (actual value/normal value) <10x in SIRS, 30x to 436x in sepsisMedian PCT higher during sepsis than in SIRS or non-inflammatory state in surgical ICU trauma patients (p=0.002)
Reith et al., 1998 [66]	70	PCT as early predictor of post-operative infectious complication	Study setting: SICUColon Surgery:Pre-operative PCT did not differ between group with complications (0.36 ng/mL) and group w/o complications (0.19 ng/mL)POD 1 PCT differed between group with complications (6.9 ng/mL) and group w/o complications (1.2 ng/mL) (p<0.001^)Differences between two groups from POD 1 to POD 10 significant (p<0.001)Aortic Surgery*:Pre-operative and operative day PCT did not differ between group with complications and group w/o complicationsDifferences between two groups from POD 1 to POD 10 significant (p<0.05)PCT useful early predictor of post-operative infectious complication
Rothenburger et al., 1999 [38]	59	PCT as indicator of sepsis in postoperative cardiac patients	Study setting: SICUPCT concentrations increased significantly in post-operative systemic infection, as compared with baseline values (10.86 ng/mL [3.28–15.13 ng/mL] vs. 0.12 ng/mL [0.08–0.21 ng/mL]), and decreased during treatment to still significantly elevated values on D5 (0.56 ng/mL [0.51–0.85 ng/mL])PCT useful in discriminating between acute phase response and systemic infectious complications following cardiac surgery
Baykut et al., 2000 [41]	400	PCT as predictor of sepsis in postoperative cardiac patients	Study setting: SICUIn patients with post-operative infection, PCT concentrations elevated until POD 4, with second increase between POD 4 and POD 6 (p<0.001)In patients without post-operative infectious complications, PCT concentrations elevated until POD 2, decreasing thereafterPCT useful in predicting bacterial infection after cardiac surgery
Boeken et al., 2000 [42]	74	PCT as indicator of sepsis in postoperative cardiac patients	Study setting: SICUIn post-operative patients with sepsis, mean PCT concentration was 19.6±6.2 ng/mlIn uninfected patients, mean PCT concentration was 0.33±0.15 ng/mLIn SIRS patients, mean PCT concentration was 0.7±0.4 ng/mLPCT useful in diagnosis of postoperative infectious complications (p<0.05)
Reith et al., 2000 [31]	246	PCT as indicator of peritonitis in septic patients s/p abdominal surgery	Study setting: SICU59 sepsis-related mortalities (24%)Pre-operative PCT did not differ between group with complications and group w/o complicationsMean PCT levels on POD 1: 4.9±2.8 ng/mL in surviving group, 13.8±8.9 ng/mL in non-surviving group (p<0.01)Mean PCT levels on POD 4: 4.8±3.1 ng/mL in surviving group, 13.0±7.5 ng/mL in non-surviving group (p<0.01)Mean PCT levels on POD 10: 0.4±0.1 ng/mL in surviving group, 13.25±5.8 ng/mL in non-surviving group (p<0.01)PCT sensitive for detection of infectious complications
Wanner et al., 2000 [17]	133	PCT as early indicator of septic complication in trauma patients	Study setting: ED and Trauma ICUMechanical trauma led to increased PCT concentrations dependent on the severity of injury, with peak values on D1 and D3 (p<0.05) and a continuous decrease within 21 d after traumaHighest PCT concentration early after injury observed in patients with sepsis (6.9±2.5 ng/mL on D1) or severe MODS (5.7±2.2 ng/mL on D1) w/ sustained increase (p<0.05) for 14 d compared with patients with an uneventful post-traumatic course (1.1±0.2 ng/mL)Increased PCT levels during the first 3 d after trauma predicted severe SIRS, sepsis, MODS (p<0.0001, logistic regression)
Meisner et al., 2002 [48]	208	PCT as indicator of sepsis in post-operative cardiac patients	Study setting: SICUMore patients with PCT>2 ng/mL on POD 1 or POD 2 (n=55) had post-operative abnormalities (95%) than patients with lower PCT (59%)Incidence of 3+SIRS criteria 45% when PCT>2 ng/mL, vs. 4% when PCT lower (ROC AUC 0.866)Proven bacterial infection rate 9% when PCT>2 ng/mL, vs. 1% when PCT lower (ROC AUC 0.900)Suspected bacterial infection rate 24% when PCT>2 ng/mL, vs 1% when PCT lower (ROC AUC 0.897)Elevated PCT correlates with evidence of systemic inflammation and other complications early postoperatively after cardiac surgery
Dorge et al., 2003 [32]	80	PCT as early predictor of post-operative complications	Study setting: SICUPost-operative PCT increased in patients w/severe complications (30.3±6.7 ng/mL vs. 5.5±1.4 ng/mL, p<0.05) and in patients with infections (38.4±11.3 ng/mL vs. 10.8±1.6 ng/mL, p<0.05)ROC AUC for PCT as predictor of infections and complications was 0.720 (95 % CI 0.603–0.837) and 0.861 (95% CI 0.779–0.943), respectivelyHigh PCT concentrations associated with infections and severe complications early after cardiac surgeryPCT was no different with infectious than with non-infectious complications
Hensler et al., 2003 [54]	138	PCT as early predictor of sepsis in major trauma patients	Study setting: Trauma ICU/SICUPCT and neopterin concentrations POD 1 and POD 2 after major traumatic injury did not differentiate patients who developed sepsis from those who did notPatients who developed MODS (vs. control) had higher PCT concentrations on D0 (0.60 ng/mL vs. 0.15 ng/mL), and on POD 1 and POD 2 combined (1.95 ng/mL vs. 0.32 ng/mL)High PCT concentration in the early post-traumatic phase is an independent predictor of MODS but not of sepsis
Mokart et al., 2005 [60]	50	PCT as early predictor of post-operatve infectious complication	Study setting: ICUPCT and IL-6 concentrations on POD 1 significantly higher in group with complications than in group w/o complications (p=0.003 for PCT, p=0.006 for IL-6), but CRP concentrations similar in both groups PCT and IL-6 early markers of post-operative sepsis in patients undergoing major cancer surgery
Castelli et al., 2006 [67]	255	PCT in monitoring for sepsis in ICU patients	Study setting: MSICUAdmission PCT higher in trauma patients with eventual infection complications (3.4 ng/mL, 95% CI 2.63–12.71) than in patients with good outcomes (1.2 ng/mL, 95% CI 0.5–5.2, p<0.05)Trauma pt had quick rise in PCT from 1d prior to sepsis diagnosis, from 0.85 ng/mL (95% CI 0.45–1.14) to 2.1 ng/mL (95% CI 1.01–6.14, p<0.05)
Lavrentieva et al., 2007 [68]	43	PCT as early indicator of septic complication in burn patients	Study setting: Burn ICUPCT concentrations higher in sepsis (11.8±15.8 ng/mL) than in SIRS (0.63±0.43) in burn patients (p<0.001)WBC count, temperature and neutrophil count did not differ between sepsis and SIRS (p>0.05)PCT ROC AUC 0.975 (95% CI 0.91–1.035, p=0.002)Serum PCT concentrations are an early indicator of septic complication in patients with severe burn injury
Novotny et al., 2009 [69]	104	PCT for determining if repeat laparotomy needed in patients with prior laparotomy for infective source	Study setting: SICUPCT ratio based on changes between POD 1 and POD 2, cutoff 1.03 ng/mL (>1.03 ng/mL=elimination of infection focus in original laparotomy;<1.03 ng/mL=need for repeat laparotomy)PCT sensitivity and specificity for detecting an unsuccessful initial laparotomy were 63% and 95%, respectivelyPCT useful for deciding if initial laparotomy sufficient, or if repeat laparotomy needed

AUC=area under the curve; CI=confidence interval; ED=emergency department; ICU=intensive care unit; MODS=multiple organ dysfunction syndrome; MSICU=medical/surgical intensive care unit; n=number of patients; PCT=procalcitonin; POD=post-operative day; ROC=receiver operating curve; SICU=surgical intensive care unit; SIRS=systemic inflammatory response syndrome; s/p=status post; WBC=white blood cell; w/o=without.

Significant at p<0.05

The value of serum procalcitonin concentrations in guiding antibiotic escalation in sepsis

To date, only a single randomized study has assessed the ability of serum PCT concentrations to guide antibiotic escalation, and this study found no clinical utility in a protocol of antibiotic escalation based on these concentratios [70,71]. Jensen et al. (n=1,200) studied whether broadening the spectrum of antimicrobial coverage, based on serial daily serum PCT concentration measurements, could reduce the lead time to administration of an appropriate antibiotic regimen consequently reducing patient mortality [70]. They reported 28-d mortality rates in the standard-of-care arm and PCT arm of their study as respectively being 32.0% (191 deaths among 596 patients) and 31.5% (190 deaths among 604 patients) (absolute risk reduction 0.6%; 95% CI 4.7%–5.9%). Jensen et al. reported that the ICU length of stay in the PCT arm of their study was greater by one day (p=0.004), the use of mechanical ventilation per day in the ICU was increased by 4.9% (95% CI 3.0%–6.7%), and the relative risk of an estimated glomerular filtration rate (eGFR) <60 mL/min/1.73m² was 1.21 (95% CI 1.15–1.27). The study concluded that escalation of antimicrobial therapy based on measurements of serum PCT concentration did not improve survival, and resulted in higher rates of organ-related injury and a longer duration of stay in the ICU [70].

The value of serum concentrations of procalcitonin in guiding antibiotic de-escalation in sepsis

Since 2007, four randomized studies have been done of the utility of serum PCT concentrations in guiding antibiotic de-escalation in sepsis, with all of them reporting these concentrations to be useful (Table 6) [72 –75]. All four studies used a similar design, comparing a PCT intervention arm with a standard-of-care arm. However, each used different criteria in determining the optimal time frame for discontinuation of antibiotic therapy. Nobre et al. conducted the first of these studies, involving 68 patients, and reported a 4-d reduction in the duration of antibiotic therapy in the PCT arm (p=0.003) [72]. Similarly, Hochreiter et al. demonstrated that patients in the PCT arm of their study required two fewer days of antibiotics (p<0.001) than those the control arm, whereas Schroeder et al. found that the PCT intervention group in their study required 1.7 fewer days of antibiotic therapy than did the control group (p<0.001) [73,74]. Lastly, Bouadma et al. who completed the largest such study, involving 621 patients, found a 2.7-d reduction in antibiotic utilization (p<0.0001) in the PCT arm of their study [75]. None of the four studies showed a significant difference in mortality in their control and PCT intervention arms (p>0.05) [72 –75].

Table 6.

Summary of all Prospective Randomized Controlled Trials Relating to Procalcitonin-guided Antibiotic De-Escalation in Adult Patients with Sepsis (2007 to 2010)

		PCT Arm			Control Arm
Study, year	Total no. of patients	No. of patients	Treatment duration (d)	Mortality^†n (%)	No. of patients	Treatment duration (d)	Mortality^†, n (%)	^p value (treatment duration)*	Study setting and outcomes
Nobre et al., 2007 [72]^a	68	31	6 (95% CI (days) 4–16)	5 (16%)	37	10 (95% CI 3,33) d	6 (16%)	=0.003	Study setting: MSICUDuration of antibiotic therapy decreased by 4 d in PCT arm
Hochreiter et al., 2009 [73]^b	110	57	5.9±1.7 d	15 (26%)	53	7.9±0.5 d	14 (26%)	<0.001	Study setting: SICUSuration of antibiotic therapy decreased by 2 d in PCT arm
Schroeder et al., 2009 [74]^c	27	14	6.6±1.1 d	3 (21%)	13	8.3±0.7 d	3 (23%)	<0.001	Study setting: SICUDuration of antibiotic therapy decreased by 1.7 d in PCT arm
Bouadma et al., 2010 [75]^d	621	307	14.3±9.1 d w/o ABx	65 (21%)	314	11.6±8.2 d w/o Abx	64 (20%)	<0.0001	Study setting: MSICUDuration of antibiotic therapy decreased by 2.7 d in PCT arm

Abx=antibiotics; MSICU=medical/surgical intensive care unit; N=number of patients; PCT=procalcitonin; SD=standard deviation; SICU=surgical intensive care unit; Tx=treatment; w/o=without.

†

: 28-d mortality.

Statistically significant at p<0.05

Antibiotic discontinuation when PCT levels decreased by 90+% from initial value, but not prior to day 3 (if baseline PCT measured<1 μg/L) or day 5 (if baseline PCT measured≥1 μg/L).

Antibiotic discontinuation when PCT decreased to <1 ng/mL, or to 25-35% of initial value over three days if value was >1 ng/mL

Antibiotic discontinuation when clinical signs of infection improved and PCT value decreased to <1 ng/mL or to <35% of initial value within 3 d.

Antibiotic discontinuation when PCT concentration<80% of peak value, or absolute PCT concentration<0.5 mcg/L.

Meta-analyses of the role of serum concentrations of procalcitonin in guiding the management of sepsis

Since 2006, seven meta-analyses have been reported of the role of serum concentrations of PCT in the management of sepsis (Table 7) [10,76 –81]. Of these, three pertain to the diagnostic value of PCT in adult patients with sepsis and four examined the use of PCT-based algorithms in reducing antibiotic use and the duration of antibiotic exposure [10,76 –81]. Among the three meta-analyses of the diagnostic value of PCT, that done by Uzzan et al. reported resoundingly positive results, Jones at al. reported mixed results and a moderate diagnostic yield, and Tang et al. reported poor diagnostic performance. All four meta-analyses of the use of PCT-based algorithms (Kopterides et al. [81], Heyland et al. [79], Schuetz et al. [80], and Wilke et al. [81]) yielded positive results [78 –81].

Table 7.

Summary of Meta-Analyses of the Role of Serum Procalcitonin Concentrations in Adult Patients with Sepsis

Meta-analysis, year	No. of studies reviewed	No. of patients	Outcomes
Uzzan et al., 2006 [10]	33	3,943	PCT sensitivity: 42%–100%PCT specificity: 48%–100%Optimal PCT cutoff: 0.6–5 ng/mLRisk of positive PCT test in patients with infection 16x higher than in non-infected patientsPCT global OR for sepsis diagnosis: 15.7 (95% CI 9.1–27.1)CRP global OR for sepsis diagnosis: 5.4 (95% CI 3.2–9.2)PCT ROC better than CRP ROCPCT good biomarker for diagnosing sepsis, severe sepsis, septic shock; PCT superior to CRP
Jones et al., 2007 [77]	17	2,008	Disease prevalence: 4%–54% (mean 19.4%)PCT cutoff: 0.4–0.5 ng/mL (13 studies), 0.9–1.0 ng/mL (2 studies), 2.0 ng/mL (2 studies)PCT ROC AUC: 0.84 (95% CI 0.75–0.9)PCT OR for sepsis diagnosis: 9.86 (95% CI 5.72–17.02)PCT sensitivity (at PCT cutoff 0.4–0.5 ng/mL): 76% (95% CI 66%–84%)PCT specificity (at PCT cutoff 0.4–0.5 ng/mL): 70% (95% CI 60%–79%)PCT had moderate diagnostic ability for detection of bacteremia, but further research is needed
Tang et al., 2007 [76]	18	2,097	PCT sensitivity and specificity: 71% (95% CI 67%–76%)PCT positive LR: 3.03 (95% CI 2.51–3.65)PCT negative LR: 0.43 (95% CI 0.37–0.48)PCT ROC AUC: 0.78 (95% CI 0.73–0.83)PCT OR for sepsis diagnosis: 5.71 (95% CI 3.62–9.03)PCT had poor diagnostic performance in differentiating sepsis from SIRS
Kopterides et al., 2010 [78]	7	1,131	PCT-guided algorithms vs. routine practice:Duration ABx for first episode of infection reduced by approx. 2 d (WMD −2.36 d [95% CI −3.11, −1.61])Total duration ABx reduced by 4 d (WMD −4.19 d [95% CI −4.98, −3.39])No increase in 28-d mortality (OR 0.93 [95% CI 0.69–1.26])No increase in ICU LOS [WMD −0.49 days [95% CI −1.55–0.57])No increase in relapsed/persistent infection rate (OR 0.97 [95% CI 0.56–1.69])PCT-guided algorithms reduced antibiotic exposure, with no increase in adverse clinical outcomes
Heyland et al., 2011 [79]	5	947	PCT-guided algorithms vs. routine practice:Significant reduction in total ABx duration (WMD −2.14 d [95% CI −2.51, −1.78, p<0.00001^*])No increase in hospital mortality (RR 1.06 [95% CI 0.86–1.30, p=0.59], RD 0.01 [95% CI −0.04–0.07, p=0.61])No increase in 28-day mortality (RR 0.98 [95% CI 0.75–1.29, p=0.91], RD 0.00 [95% CI −0.06–0.05, p=0.88])No increase in ICU LOS (WMD −1.05 [95% CI −4.5–1.05, p=0.25])No increase in hospital LOS (WMD −1.86 [95% CI −4.75–1.04, p=0.21])No increase in recurrent or relapsing infections (RR 1.26 [95% CI 0.68–2.35, p=0.46])
Schuetz et al., 2011 [80]	14	4,467	Consistent reduction observed in antibiotic prescription and/or duration of therapy across all trialsNo significant difference in mortality between patients in the PCT groups (172 of 2,227 [7.7%]) compared with the control group [186 of 2,240 [8.3%]), summary OR 0.91 (95% CI 0.73–1.14)PCT-guided algorithms vs. routine practice:Primary care (2 trials, n=1,008): Low mortality rate in both groups (OR 0.13 [95% CI 0.6–640])ED (6 trials, n=2,449): Mortality in PCT groups 5.2% (63 of 1,214), vs. 5.4% (67 of 1,235) in control groups (OR 0.95 [95% CI 0.67–1.36])ICU (2 trials, n=1,010): Mortality in PCT groups 21.5% (109 of 506) vs. 23.4% (118 of 504) incontrol groups (OR, 0.89 [95% CI 0.66–1.20])PCT-guided algorithms effective at reducing use of Abx without diminishing patient safety
Wilke et al., 2011 [81]	6	486	PCT-guided algorithms reduced ABx duration by 4 dPCT-guided algorithms reduced ICU LOS by 1.8 dWeighted mean duration of ABx therapy 7.3 dUse of PCT-guided algorithms resulted in estimated savings of €1,163,000 for ICU-patients and €36,512 for non-ICU patients for the sample population (per patient, €886.4 average for ICU-patients and €136.2 for non-ICU patients)

Abx=antibiotics; AUC=area under the curve; CI=confidence interval; CRP=C-eeactive protein; ED=emergency department; ICU=intensive care unit; LOS=length of stay; LR=likelihood ratio; N=number of patients; OR=odds ratio; PCT=procalcitonin; RD=risk difference; ROC=receiver operating characteristic curve; RR=risk ratio; SIRS=systemic inflammatory response syndrome; WMD=weighted mean difference.

Statistically significant at p<0.05

Discussion

Current evidence supporting and refuting the clinical utility of measuring serum concentrations of PCT in the diagnosis of sepsis is controversial and conflicting. Although most diagnosis-oriented clinical studies, and those assessing the utility of serum concentrations of PCT in prognosticating outcomes of sepsis and its severity, have yielded positive results, differences in measurements and set points of serum PCT concentrations, and divergent study designs, make it all but impossible to establish evidence-based recommendations about how to use serum concentrations of PCT in differentiating sepsis from SIRS, severe sepsis, and septic shock, or in predicting outcomes and mortality in sepsis. Given this, most clinical sites that rely heavily on the measurement of serum concentrations of PCT have developed their own protocols for this or have relied on absolute PCT concentrations or trends in PCT concentrations.

A notable shortcoming of all studies of the diagnostic and prognostic use of PCT in sepsis is a small sample size. Besides the study by Wanner et al., involving 405 patients, only three additional studies had a sample size of ≥250 patients [31,33,41]. With regard to prognostic studies of the use of PCT, Reith et al. published the largest study of this, involving 246 patients [31]. Small sample sizes, combined with poor statistical power reported among the published studies reporting positive effects of the measurement of PCT, support the conclusions of Tang et al. that current studies of PCT have shown “significant heterogeneity because of variability in sample size, with diagnostic performance upwardly biased in smaller studies, but moving towards a null effect in larger studies” [76]. Although meta-analyses effectively eliminate the study limitation of small sample size, the three published meta-analyses of the role of serum PCT concentrations in the diagnosis of sepsis had mixed results, despite an overlap in the studies included in each of these meta-analyses [10,76,77]. Perhaps more importantly, no standardized minimum cutoff concentration of PCT for determining sensitivity and specificity has yet been established, although the most common cutoff values used have been 2.0 ng/mL, 1.0 ng/mL, or 0.5 ng/mL [8,18,20 –23,25 –28,33,37,40,48,51,53,57,59]. It is possible that a higher cutoff value will have to be established for use in surgical, trauma, and burn patients, in whom serum concentrations of PCT may be modestly increased, even in the absence of ongoing infectious processes [15]. Variations in the choice of comparator reference biomarkers (CRP, IL-2, IL-6, IL-8, TNF-α) and sepsis scoring systems (Acute Physiology and Chronic Health Evaluation II [APACHE II], Sepsis-related Organ Failure Assessment [SOFA]), as well as a lack of standardization of the type of metric assessed (with different studies using mean or median serum PCT concentration, area under the curve of the receiver operator characteristic [ROC] curve for PCT, PCT ratio, and PCT dynamics as means of PCT measurement) make meaningful interpretation of current data difficult. In most studies, cutoff values of serum PCT concentration for stratifying patients according to disease severity were generated arbitrarily and varied widely, creating a lack of uniformity in both study design and endpoints, such that these cannot be compared directly.

The difficulty described above in interpreting the results of reported trials of the diagnostic and prognostic use of serum PCT concentrations in sepsis are not unique to PCT, and have plagued similar such studies in the past. As a result, it is incumbent upon investigators involved currently in studies of PCT and of other biomarkers of sepsis to establish standard protocols and data endpoints if their results are to be meaningfully interpreted, and to do so before conducting large-scale randomized controlled trials (RCT). Such an approach would help to reduce the possibility of a small sample size resulting in bias toward positive outcomes, and would provide unified study design and outcome measures with a high level of statistical power [76]. Until such time as the results of newer trials become available, the current literature on the role of PCT in the diagnosis and prognosis of sepsis in adult patients is insufficient to permit wholehearted support for accepting the measurement of its serum concentrations as a sole determinant for these two purposes. Consequently, if clinicians are to utilize serum PCT concentrations, it is incumbent upon them to validate the use of these in their own clinical practice settings.

The monitoring of inpatients for infectious complications constitutes another potential use for measuring serum concentrations of PCT. Hospital inpatients, post-operative patients, and most notably trauma and burn patients in the ICU are at considerable risk for contracting health-care-associated infections (HAI) during their hospitalization [82]. Such infections represent an important public health topic, especially in the context of antibiotic resistance, because multi-drug-resistant (MDR) pathogens represent a serious cause of death in the ICU and among inpatient populations [83]. The overwhelming majority of PCT-centered prospective studies designed to differentiate infectious from non-infectious causes of SIRS have had positive outcomes for the measurement of PCT, and despite differences in study design, its serum concentrations currently play a substantial role in guiding early intervention for sepsis and reducing sepsis-related mortality in many ICUs worldwide. Early measurements of serum PCT concentration s may allow the detection of potential and ongoing infectious complications, permitting prompt administration of empiric antibiotic therapy [84]. many consensus guidelines have endorsed the importance of early antibiotic administration, such as the Surviving Sepsis Campaign, which stresses the importance of rapid and early antimicrobial therapy whenever the diagnosis of sepsis is entertained.

To date, only one randomized trial of the utility of measuring serum PCT concentrations to guide antibiotic escalation in sepsis has been published. In this trial, Jensen et al. reported that serum concentrations of PCT were not suitable for directing antibiotic escalation in the clinical setting because they failed to reduce mortality below that with standard care, extended the duration of care in the ICU, increased the need for mechanical ventilation, and decreased eGFR (representative of end-organ damage) [70]. Although one negative trial does not entirely negate a future role for PCT in the setting of sepsis, further studies are necessary to determine whether PCT can be used reliably to direct antibiotic dosage and expansion of the spectrum of antibiotic coverage while also reducing mortality and minimizing adverse drug reactions and complications in this setting.

The rapid development of antibiotic resistance in the past decade has highlighted the critical need for stewardship programs in antimicrobial therapy [82,83]. Such programs aim in part to decrease selective pressure on resistant bacterial strains via responsible selection of the type of antibiotic used, the dose used, and the duration of treatment on a per-patient basis [85]. De-escalation of antibiotic therapy, through the reduction of its duration or of the spectrum of coverage, is of central importance to stewardship programs, and four recent studies have highlighted a potential role for the measurement of serum PCT concentrations as an effective means of guiding such de-escalation [72 –75,86]. These results are extremely important and provide more than sufficient data to justify a large-scale, multi-center RCT focusing on the benefits of PCT-guided early discontinuation of antibiotic therapy and its impact on selective pressures that give rise to strains of MDR organisms.

In summary, the results of studies of the use of serum PCT concentrations for the diagnosis of and prognosis in sepsis are contradictory and are plagued by small sample sizes and lack of uniformity in study design. Seven of the 46 published studies of the role of serum PCT concentrations in establishing a diagnosis of sepsis, encompassing 805 (approximately 15%) of the total of 5,583 patients enrolled in these 46 trials, have yielded negative results for a benefit of measuring concentrations of PCT as an indicator of sepsis, and two reviews have concluded that PCT has poor diagnostic performance [12,24,25,27,30,32,46,49,58]. Nearly one-third of all published trials of the ability of serum PCT concentrations to predict outcomes of sepsis have also yielded negative results, and one published review concluded that serum PCT concentrations lack any prognostic value [27,29,30,36,63,65]. To date, the only published trial of the role of serum PCT concentrations in the escalation of antibiotic therapy yielded negative results and found poorer patient outcomes with their use [39]. The measurement of serum PCT concentrations has shown the most consistent promise with regard to the monitoring of sepsis and de-escalation of antibiotic therapy for it. Nearly all studies of the use of serum PCT concentrations in monitoring the response to treatment for sepsis have yielded positive results, and all four studies of serum PCT concentrations in antibiotic de-escalation found that their use was associated with reductions in the duration of antibiotic therapy without a significant increase in mortality [17,31,35,38,41,42,48,54,60,66 –69,72 –75], Although additional studies of the various clinical roles for the measurement of serum PCT concentrations in sepsis are needed, it is becoming imminently clear that PCT has the greatest potential to reshape the clinical management of adult patients with sepsis, specifically with respect to their response to antibiotic therapy and its duration and de-escalation.

Emerging data on the clinical utility of measurement of the serum concentration of PCT suggest that this biomarker may have promise for use in patient surveillance for infectious complications, and in the de-escalation of antibiotic therapy in the ICU setting. The value of serum PCT concentrations in the diagnosis of and prognosis in sepsis, specifically in predicting disease severity and mortality, as well as in the escalation of antibiotic therapy, is currently unproved. In the imaginary world of Don Quixote, the current clinical utility of serum PCT concentrations in sepsis cannot be equated to slaying dragons, but neither is it merely thrusting at windmills. The results of ongoing clinical trials notwithstanding, routine measurement of serum PCT concentrations in sepsis cannot currently be endorsed unequivocally, although future research should further clarify its role. Until such time as these trials are completed, the measurement of serum PCT concentrations should be used as only one criterion in the diagnosis and management of sepsis, in conjunction with measurements of other biomarkers and sepsis scoring systems, to provide the most accurate means of assessment and therapeutic intervention in adult patients with sepsis.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Martin

, Mannino

, Eaton

, et al. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med, 2003; 348:1546–1554

Bone

, Balk

, Cerra

, et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest, 1992; 101: 1644–1655.

Levy

, Fink

, Marshall

, et al. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med, 2003; 31:1250–1256.

Agency for Healthcare Research and Quality. HCUP facts and figures: Statistics on hospital-based care in the United States, 2008. Available from: http://www.hcup-us.ahrq.gov/reports/factsandfigures/2008/TOC_2008.jsp.

Hall

, Williams

, DeFrances

, et al. Inpatient care for septicemia or sepsis: A challenge for patients and hospitals. NCHS Data Brief, 2011; 62. Available from: www.cdc.gov/nchs/data/databriefs/db62.pdf.

Carrigan

, Scott

, Tabrizian

. Toward resolving the challenges of sepsis diagnosis. Clin Chem, 2004; 50:1301–1314.

Levy

. The challenge of sepsis. Crit Care, 2004; 8:435–436.

Giamarellos-Bourboulis

, Grecka

, Scarpa

, et al. Procalcitonin: A marker to clearly differentiate systemic inflammatory response syndrome and sepsis in the critically ill patient?. Intensive Care Med, 2002; 28:1351–1356.

Harbarth

, Holeckova

, Froidevaux

, et al. 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.

10.

Uzzan

, Cohen

, Nicolas

, et al. Procalcitonin as a diagnostic test for sepsis in critically ill adults and after surgery or trauma: A systematic review and meta-analysis. Crit Care Med, 2006; 34:1996–2003.

11.

Rivers

, McIntyre

, Morro

, et al. Early and innovative interventions for severe sepsis and septic shock: Taking advantage of a window of opportunity. CMAJ, 2005; 173:1054–1065.

12.

Pierrakos

, Vincent

. Sepsis biomarkers: A review. Crit Care, 2010; 14:R15.

13.

Bozza

, Bozza

, Neto

. Beyond sepsis pathophysiology with cytokines: What is their value as biomarkers for disease severity?. Mem Inst Oswaldo Cruz, 2005; 100(Suppl I):217–221.

14.

Ivady

, Beres

, Szabo

. Recent advances in sepsis research: Novel biomarkers and therapeutic targets. Curr Med Chem, 2011; 18:3211–3225.

15.

Gonsalves

, Sakr

. Early identification of sepsis. Curr Infect Dis Rep, 2010; 12:329–335.

16.

Wolff

, Bouadma

. What procalcitonin brings to management of sepsis in the ICU. Crit Care, 2010; 14:1007.

17.

Wanner

, Keel

, Steckholzer

, et al. Relationship between procalcitonin plasma levels and severity of injury, sepsis, organ failure, and mortality in injured patients. Crit Care Med, 2000; 28:950–957.

18.

Brunkhorst

, Wegscheider

, Forycki

, et al. Procalcitonin for early diagnosis and differentiation of SIRS, sepsis, severe sepsis, and septic shock. Intensive Care Med, 2000; 26(Suppl 2):S148–152.

19.

Balci

, Sungurtekin

, Gürses

, et al. Usefulness of procalcitonin for diagnosis of sepsis in the intensive care unit. Crit Care 200;7:85–90.

20.

Luzzani

, Polati

, Dorizzi

, et al. Comparison of procalcitonin and C-reactive protein as markers of sepsis. Crit Care Med, 2003; 31:1737–1741.

21.

Endo

, Aikawa

, Fujishima

, et al. Usefulness of procalcitonin serum level for the discrimination of severe sepsis from sepsis: A multicenter prospective study. J Infect Chemother, 2008; 14:244–249.

22.

Tsangaris

, Plachouras

, Kavatha

, et al. Diagnostic and prognostic value of procalcitonin among febrile critically ill patients with prolonged ICU stay. BMC Infect Dis, 2009; 9:213.

23.

Meynaar

, Droog

, Batstra

, Vreeded

, Herbrink

. In critically ill patients, serum procalcitonin is more useful in differentiating between sepsis and SIRS than CRP, Il-6, or LBP. Crit Care Res Pract, 2011; 2011:594–645.

24.

Ugarte

, Silva

, Mercan

, et al. Procalcitonin used as a marker of infection in the intensive care unit. Crit Care Med, 1999; 27:498–504.

25.

Suprin

, Camus

, Gacouin

, et al. Procalcitonin: A valuable indicator of infection in a medical ICU?. Intensive Care Med, 2000; 26:1232–1238.

26.

Gaïni

, Koldkjær

, Pedersen

, et al. Procalcitonin, lipopolysaccharide-binding protein, interleukin-6 and C-reactive protein in community-acquired infections and sepsis: A prospective study. Crit Care, 2006; 10:R53.

27.

Sudhir

, Venkatachalaiah

, Kumar

, et al. Significance of serum procalcitonin in sepsis. Indian J Crit Care Med, 2011; 15:1–5.

28.

Müller

, Becker

, Schächinger

, et al. Calcitonin precursors are reliable markers of sepsis in a medical intensive care unit. Crit Care Med, 2000; 28:977–983.

29.

Selberg

, Hecker

, Martin

, et al. Discrimination of sepsis and systemic inflammatory response syndrome by determination of circulating plasma concentrations of procalcitonin, protein complement 3a, and interleukin-6. Crit Care Med, 2000; 28:2793–2798.

30.

Castelli

, Pognani

, Meisner

, et al. Procalcitonin and C-reactive protein during systemic inflammatory response syndrome, sepsis and organ dysfunction. Crit Care, 2004; 8:R234–242.

31.

Reith

, Mittelkötter

, Wagner

, et al. Procalcitonin (PCT) in patients with abdominal sepsis. Intensive Care Med, 2000; 26(Suppl 2):S165–169.

32.

Dorge

, Schondube

, Dorge

, et al. Procalcitonin is a valuable marker in cardiac surgery but not specific for infection. Thorac Cardiovasc Surg, 2003; 51:322–326.

33.

Al-Nawas

, Krammer

, Shah

. Procalcitonin in diagnosis of severe infections. Eur J Med Res, 1996; 1:331–33.

34.

De Werra

, Jaccard

, Betz Corradin

, et al. Cytokines, nitrite/nitrate, soluble tumor necrosis factor receptors, and procalcitonin concentrations: Comparisons in patients with septic shock, cardiogenic shock, and bacterial pneumonia. Crit Care Med, 1997; 25:607–613.

35.

Benoist

, Mimoz

, Assicot

, et al. Serum procalcitonin, but not C-reactive protein, identifies sepsis in trauma patients. Clin Chem, 1998; 44:1778–1779.

36.

Whang

, Steinwald

, White

, et al. Serum calcitonin precursors in sepsis and systemic inflammation. J Clin Endocrinol Metab, 1998; 83:3296–3301.

37.

Bossink

, Groeneveld

, Thijs

. Prediction of microbial infection and mortality in medical patients with fever: Plasma procalcitonin, neutrophilic elastase-alpha-antitrypsin, and lactoferrin compared with clinical variables. Clin Infect Dis, 1999; 29:398–407.

38.

Rothenburger

, Markewitz

, Lenz

, et al. Detection of acute phase response and infection. The role of procalcitonin and C-reactive protein. Clin Chem Lab Med, 1999; 37:275–279.

39.

Adamik

, Kubler-Kielb

, Golebiowska

, et al. Effect of sepsis and cardiac surgery with cardiopulmonary bypass on plasma level of nitric oxide metabolites, neopterin, and procalcitonin: Correlation with mortality and postoperative complications. Intensive Care Med, 2000; 26:1259–1267.

40.

Aouifi

, Piriou

, Bastien

, et al. Usefulness of procalcitonin for diagnosis of infection in cardiac surgical patients. Crit Care Med, 2000; 28:3171–3176.

41.

Baykut

, Schulte-Herbruggen

, Krian

. The value of procalcitonin as an infection marker in cardiac surgery. Eur J Med Res, 2000; 5:530–536.

42.

Boeken

, Feindt

, Micek

, et al. Procalcitonin (PCT) in cardiac surgery: Diagnostic value in systemic inflammatory response syndrome (SIRS), sepsis and after heart transplantation (HTX). Cardiovasc Surg, 2000; 8:550–554.

43.

Cheval

, Timsit

, Garrouste-Orgeas

, et al. Procalcitonin (PCT) is useful in predicting the bacterial origin of an acute circulatory failure in critically ill patients. Intensive Care Med, 2000; 26(Suppl. 2):S153–158.

44.

Oberhoffer

, Russwurm

, Bredle

, et al. Discriminative power of inflammatory markers for prediction of tumor necrosis factor-α and interleukin-6 in ICU patients with systemic inflammatory response syndrome (SIRS) or sepsis at arbitrary time points. Intensive Care Med, 2000; 26:S170–174.

45.

Yukioka

, Yoshida

, Kurita

, et al. Plasma procalcitonin in sepsis and organ failure. Ann Acad Med Singapore, 2001; 30:28–31.

46.

Baumgarten

, Pequeriaux

, Van Puyenbroek

, et al. Diagnosis of sepsis by procalcitonin. Ned Tijdschr Klin Chem, 2002; 27:32–35.

47.

Hausfater

, Garric

, Ayed

, et al. Usefulness of procalcitonin as a marker of systemic infection in emergency department patients: A prospective study. Clin Infect Dis, 2002; 34:895–901.

48.

Meisner

, Rauschmayer

, Schmidt

, et al. Early increase of procalcitonin after cardiovascular surgery in patients with postoperative complications. Intensive Care Med, 2002; 28:1094–1102.

49.

Ruokonen

, Ilkka

, Niskanen

, et al. Procalcitonin and neopterin as indicators of infection in critically ill patients. Acta Anaesthesiol Scand, 2002; 46:398–404.

50.

Tugrul

, Esen

, Celebi

, et al. Reliability of procalcitonin as a severity marker in critically ill patients with inflammatory response. Anaesth Intensive Care, 2002; 30:747–754.

51.

De Talance

, Burlet

, Claudel

. La procalcitonine (PCT) est-elle le marqueur spécifique du choc septique?. Immunoanalyse et Biologie Spécialisée, 2003; 18:120–122.

52.

, Pan

, Chen

, et al. Serum procalcitonin and interleukin-6 levels may help to differentiate systemic inflammatory response of infectious and non-infectious origin. Chin Med J, 2003; 116:538–542.

53.

Geppert

, Steiner

, Delle-Karth

, et al. Usefulness of procalcitonin for diagnosing complicating sepsis in patients with cardiogenic shock. Intensive Care Med, 2003; 29:1384–1389.

54.

Hensler

, Sauerland

, Lefering

, et al. The clinical value of procalcitonin and neopterin in predicting sepsis and organ failure after major trauma. Shock, 2003; 20:420–426.

55.

Kabir

, Keller

, Grass

, et al. Cytokines and chemokines in serum and urine as early predictors to identify septic patients on intensive care unit. Int J Mol Med, 2003; 12:565–570.

56.

Chan

, Tseng

, Tsay

, et al. Procalcitonin as a marker of bacterial infection in the emergency department: An observational study. Crit Care, 2004; 8:R12–20.

57.

Clec'h

, Ferriere

, Karoubi

, et al. Diagnostic and prognostic value of procalcitonin in patients with septic shock. Crit Care Med, 2004; 32:1166–1169.

58.

Gibot

, Kolopp-Sarda

, Bene

. 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.

59.

Aikawa

, Fujishima

, Endo

, et al. Multicenter prospective study of procalcitonin as an indicator of sepsis. J Infect Chemother, 2005; 11:152–159.

60.

Mokart

, Merlin

, Sannini

, et al. Procalcitonin, interleukin 6 and systemic inflammatory response syndrome (SIRS): Early markers of postoperative sepsis after major surgery. Br J Anaesth, 2005; 94:767–773.

61.

Charles

, Tinel

, Barbar

, et al. Procalcitonin kinetics within the first days of sepsis: Relationship with the appropriateness of antibiotic therapy and the outcome. Crit Care, 2009; 13:R38.

62.

Karlsson

, Heikkinen

, Pettilä

, et al. Predictive value of procalcitonin decrease in patients with severe sepsis: A prospective observational study. Crit Care, 2010; 14:R205.

63.

Dahaba

, Hagara

, Fall

, et al. Procalcitonin for early prediction of survival outcome in postoperative critically ill patients with severe sepsis. Br J Anaesth, 2006; 97:503–508.

64.

Assicot

, Gendrel

, Carsin

, et al. High serum procalcitonin concentrations in patients with sepsis and infection. Lancet, 1993; 341:515–518.

65.

Becker

, Snider

, Nylen

. Procalcitonin assay in systemic inflammation, infection, and sepsis: Clinical utility and limitations. Crit Care Med, 2008; 36:941–952.

66.

Reith

, Mittelkötter

, Debus

, et al. Procalcitonin in early detection of postoperative complications. Dig Surg, 1998; 15:260–265.

67.

Castelli

, Pognani

, Cita

, et al. Procalcitonin, C-reactive protein, white blood cells and SOFA score in ICU: Diagnosis and monitoring of sepsis. Minerva Anestesiol, 2006; 72:69–80.

68.

Lavrentieva

, Kontakiotis

, Lazaridis

, et al. Inflammatory markers in patients with severe burn injury. What is the best indicator of sepsis?. Burns, 2007; 33:189–194.

69.

Novotny

, Emmanuel

, Hueser

, et al. Procalcitonin ratio indicates successful surgical treatment of abdominal sepsis. Surgery, 2009; 145:20–26.

70.

Jensen

, Hein

, Lundgren

, et al. Procalcitonin-guided interventions against infections to increase early appropriate antibiotics and improve survival in the intensive care unit: A randomized trial. Crit Care Med, 2011; 39:2048–2058.

71.

Jensen

, Lundgren

, Hein

, et al. The Procalcitonin And Survival Study (PASS)—a randomised multi-center investigator-initiated trial to investigate whether daily measurements biomarker procalcitonin and pro-active diagnostic and therapeutic responses to abnormal Procalcitonin levels, can improve survival in intensive care unit patients. Calculated sample size (target population): 1000 patients. BMC Infect Dis, 2008; 8:91.

72.

Nobre

, Harbarth

, Graf

, et al. Use of procalcitonin to shorten antibiotic treatment duration in septic patients: a randomized trial. Am J Respir Crit Care Med, 2008; 177:498–505.

73.

Hochreiter

, Köhler

, Schweiger

, et al. Procalcitonin to guide duration of antibiotic therapy in intensive care patients: A randomized prospective controlled trial. Crit Care, 2009; 13:R83.

74.

Schroeder

, Hochreiter

, Koehler

, et al. Procalcitonin (PCT)-guided algorithm reduces length of antibiotic treatment in surgical intensive care patients with severe sepsis: Results of a prospective randomized study. Langenbecks Arch Surg, 2009; 394:221–226.

75.

Bouadma

, Luyt

, Tubach

, et al. Use of procalcitonin to reduce patients' exposure to antibiotics in intensive care units (PRORATA trial): A multicentre randomised controlled trial. Lancet, 2010; 375:463–474.

76.

Tang

, Eslick

, Craig

, et al. Accuracy of procalcitonin for sepsis diagnosis in critically ill patients: systematic review and meta-analysis. Lancet Infect Dis, 2007; 7:210–217.

77.

Jones

, Fiechtl

, Brown

, et al. Procalcitonin test in the diagnosis of bacteremia: A meta-analysis. Ann Emerg Med, 2007; 50:34–41.

78.

Kopterides

, Siempos

, Tsangaris

, et al. Procalcitonin-guided algorithms of antibiotic therapy in the intensive care unit: A systematic review and meta-analysis of randomized controlled trials. Crit Care Med, 2010; 38:2229–2241.

79.

Heyland

, Johnson

, Reynolds

, et al. Procalcitonin for reduced antibiotic exposure in the critical care setting: A systematic review and an economic evaluation. Crit Care Med, 2011; 39:1792–1799.

80.

Schuetz

, Chiappa

, Briel

, et al. Procalcitonin algorithms for antibiotic therapy decisions: A systematic review of randomized controlled trials and recommendations for clinical algorithms. Arch Intern Med, 2011; 171:1322–1331.

81.

Wilke

, Grube

, Bodmann

. The use of a standardized PCT-algorithm reduces costs in intensive care in septic patients—A DRG-based simulation model. Eur J Med Res, 2011; 16:543–548.

82.

Klevens

, Edwards

, Richards

, et al. Estimating health care-associated infections and deaths in U.S. hospitals, 2002. Public Health Rep, 2007; 122:160–166.

83.

Brusselaers

, Vogelaers

, Blot

. The rising problem of antimicrobial resistance in the intensive care unit. Ann Intensive Care, 2011; 1:47.

84.

Lavrentieva

, Papadopoulou

, Kioumis

, et al. PCT as a diagnostic and prognostic tool in burn patients. Whether time course has a role in monitoring sepsis treatment. Burns, 2012; 38:356–363.

85.

Griffith

, Postelnick

, Scheetz

. Antimicrobial stewardship programs: Methods of operation and suggested outcomes. Expert Rev Anti Infect Ther, 2012; 10:63–73.

86.

Bal

, Gould

. Antibiotic stewardship: Overcoming implementation barriers. Curr Opin Infect Dis, 2011; 24:357–362.