WO2014009480A1

WO2014009480A1 - Method for assaying sepsis and outcome in humans by detection of psp/reg

Info

Publication number: WO2014009480A1
Application number: PCT/EP2013/064696
Authority: WO
Inventors: Rolf Graf; Luregn SCHLAPBACH
Original assignee: Universitaet Bern; Zurich Universitaet Institut fuer Medizinische Virologie
Current assignee: Universitaet Bern; Zurich Universitaet Institut fuer Medizinische Virologie
Priority date: 2012-07-12
Filing date: 2013-07-11
Publication date: 2014-01-16
Anticipated expiration: 2015-01-12

Abstract

The present invention relates to a method of treatment, prediction and/or diagnosis of a systemic infection in human, in particular for prediction of the development of sepsis, based either on combining levels of pancreatic stone protein / regenerating protein (PSP) and of procalcitonin (PCT) in body fluids, or on combining levels of PSP in body fluids and results of a clinical score such as APACHEII or SAPSII.

Description

METHOD FOR ASSAYING SEPSIS AND OUTCOME IN HUMANS BY DETECTION OF

PSP/REG

FIELD OF THE INVENTION The present invention relates to a method of predicting and/or diagnosing a systemic infection in humans, in particular for predicting the development of sepsis, based either on combining levels of pancreatic stone protein / regenerating protein (PSP) and of procalcitonin (PCT) in body fluids, or on combining levels of PSP in body fluids and results of a clinical score such as APACHE II or SAPS II.

BACKGROUND OF THE INVENTION

Among critically ill patients admitted to intensive care units (ICU), at least one third to half of them is ultimately diagnosed with sepsis, which is associated with multiple organ failure and a high mortality. An early and accurate prediction/diagnosis is beneficial on the outcome of sepsis. Among the most commonly used markers for diagnosis of sepsis are leukocyte counts, C-reactive protein (CRP) and procalcitonin (PCT). In addition, cytokines such as IL-6, IL-8 and IL-10 are employed to monitor patients. However, none of the markers serves as a predictive indicator for infections including sepsis, hence treatment may lag behind the onset of sepsis. Due to the heterogeneity of both microbial agents and host inflammatory responses, therapeutic strategies should be adapted to each individual patient. The identification of patients at high risk of death and who might benefit most from early and aggressive treatment would represent a critical step towards such tailored management.

The publications and other materials, including patents, used herein to illustrate the invention and, in particular, to provide additional details respecting the practice are incorporated herein by reference. Pancreatic stone protein/regenerating protein (PSP) belongs to a family of lectin-binding proteins that were identified initially in patients with pancreatitis (L. Multigner et al., Gastroenterology 1985, 89:387-391 ). PSP has been studied predominantly in the pancreas. Under conditions of acute or chronic pancreatitis, it is highly up-regulated and may appear in the serum (W. Schmiegel et al., Gastroenterology 1990, 99:1421 -1430). Serum levels are also raised in several gastrointestinal diseases (Satomura et al., J. Gastroenterol 1995, 30:643-650). The function of PSP is still highly debated, but it is generally assumed that it is involved in promoting cell proliferation during regenerative processes (Y. Kinoshita et al., J. Gastroenterol 2004, 39:507-513).

In trauma patients, it has been shown that PSP is up-regulated in blood after trauma, and that the PSP level is related to the severity of inflammation. In particular, it is highly increased in patients during sepsis (M. Keel et al., Crit Care Med. 2009, 37(5): 1642-1648). The PSP level in blood serum has recently been proven to be a reliable indicator of systemic sepsis (US Pat. No. 8,435,755; WO 2009/030456). In the state of the art, the procalcitonin determination is described for the purpose of a study to distinguish bacterial sepsis from other disease causes (US Pat. No. 5,639,617; EP 0 656 121 ).

There is a need in the art for diagnosing a systemic infection in humans, in particular for predicting the development of sepsis.

SUMMARY OF THE INVENTION

Serum biomarkers may assist clinicians in risk stratification and decision making processes. Certain simple biomarkers may be used, especially in combination, either by combining two or more biomarkers or by combining one or more biomarkers with a clinical score such as APACHE II or SAPS II.

The present invention is, in one embodiment, directed to a method of predicting and/or diagnosing a systemic infection in humans, in particular for predicting the development of sepsis, based either on the combination of levels of pancreatic stone protein /

regenerating protein (PSP) and of procalcitonin (PCT) in body fluids, or on the

combination of levels of PSP in body fluids and a clinical score such as APACHE II or SAPS II. Furthermore, the invention is directed to a method for improving the diagnosis performance of PCT wherein the levels of PSP/reg and PCT are determined in, e.g., serum and information is combined, and wherein a high level of these two biomarkers is more indicative of early-onset sepsis in neonates and sepsis in adults than the level of PCT alone. In one aspect, the invention relates to a method, e.g., an ex-vivo method, for detecting and/or diagnosing sepsis in humans. In particular, the method may allow the diagnosis of early onset sepsis in neonates and the prognosis of sepsis in adults.

In a further aspect, the invention relates to a method of treatment of sepsis based on diagnostic results, particularly as a decision-making step to allow clinicians and other healthcare providers to take accurate treatment actions according to current practice guidelines.

The methods according to the present invention may comprise determining the levels of PSP and of PCT in an isolated body fluid sample (one sample may be used for both determinations or multiple samples from the same patient), e.g. serum or plasma, wherein said levels, in combination, are indicative of early onset sepsis. In particular, the method may comprise, subsequent to determining the levels of PSP and of PCT, comparing these levels to reference values that are either measured or provided, and then classifying / categorizing samples / patients in which levels of PSP and levels of PCT are above the respective reference values as, e.g., at risk, or above average risk, of early onset sepsis. Different risk groups may be defined.

The methods according to the present invention may also comprise determining the levels of PSP and of PCT in an isolated body fluid sample (one sample may be used for both determinations or multiple samples from the same patient), e.g. serum or plasma, wherein said levels, in combination, are indicative of the severity of sepsis at an early stage of the disease and/or of the complications and risk of mortality linked to said sepsis condition. In particular, the method may comprise, subsequent to determining the levels of PSP and of PCT, comparing these levels to reference values that are either measured or provided and classifying / categorizing samples / patients in which levels of PSP and levels of PCT are above the respective reference values as, e.g., at risk, or above average risk, of complications and mortality linked to said sepsis condition. Different risk groups may be defined. The method is in certain embodiments also directed to quantitatively determining the severity of the sepsis at an early, e.g., a phenotypically and/or physiologically inconspicuous, stage of the disease based on the combination of the levels of PSP and PCT. The methods according to the present invention may also comprise determining the levels of at least PSP in an isolated body fluid sample, e.g. serum or plasma, wherein said levels combined with the individual clinical score, e.g. APACHE II or SAPS II or SOFA, is indicative of the severity of sepsis at early stages of the disease and/or of the

complications and risk of mortality linked to said sepsis condition. In particular, the method may comprise, subsequent to determining the levels of PSP and determining the clinical score, comparing the values measured to reference values that may be measured or provided and classifying / categorizing samples / patients in which levels of PSP and the clinical score are above the reference values as, e.g., at riskor above average risk of complications and mortality linked to said sepsis condition. Different risk groups may be defined. The method is, in certain embodiments, also directed at quantitatively

determining the severity of the sepsis at an early, e.g., a phenotypically and/or physiologically inconspicuous, stage of the disease based on a combination of these indicators.

In a further aspect, the present invention relates to a method comprising:

a) providing a body fluid sample from said patient;

b) determining the levels of both PSP (level PSP actual) and PCT (level PCT _actuai) in said sample;

c) comparing the level of PSP determined in b) with a reference value (level PSP _ref) and comparing the level of PCT determined in b) with a reference value (level PCT _ref); d) identifying the patient as a patient in which:

(i) level PSP actual > level PSP _ref and PCT _actuai > level PCT _ref; or

(ii) level PSP actual > level PSP _ref and PCT _actuai </= level PCT _ref; or

(iii) level PSP actual </= level PSP _ref and PCT _actuai > level PCT _ref; or

(iv) level PSP actual < level PSP _ref and PCT _actuai < level PCT _ref;

and optionally

e) if the patient is a patient according to d) (i), treating the patient with an agent combating sepsis, e.g. a broad-spectrum antibiotic, and taking accurate treatment actions according to current practice guidelines.

"Current practice guidelines" are based on the recommendations of 2008 (Dellinger et al, Intensive Care Med, 2008, 34:17-60), which are an updated version of the first edition guidelines from 2001 coordinated by the International Sepsis Forum (ISF), and of the second editions from 2004 coordinated by the Society of Critical Care Medicine (SCCM), the European Society of Intensive Care Medicine (ESICM), and ISF. The body fluid sample might be one singular sample taken from a patient or might be more than one sample taken from the same patient, preferably taken not more than 3, 2, or preferably 1 hour apart.

The method of the invention without the treatment step may be an in vitro or ex-vivo diagnostic method.

The reference value may be the level of PSP and of PCT measured in a body fluid sample from a patient without known or suspected infection.

The reference value for PSP in adult patients is about 10 ng/ml or 25 ng/ml.

The reference value for PSP in neonates is about 2 ng/ml.

The reference value for PCT in adult patients is about 2 ng/ml or about 0.5 ng/ml.

The reference value for PCT in neonates is about 9 ng/ml or below. The reference values may differ for different age groups within the first three days post-partum and may be provided in form of a matrix, which plots the reference for each type of blood sample against age of the patient (Stacker et al. BMC Pediatrics 2010, 10:89).

The invention is also directed to a kit for determining early onset sepsis comprising:

an assay for determining a level of PSP in a sample from a patient, and

an assay for determining a level of PCT in a sample from said patient.

The invention is also directed to a method of identifying and treating systemic infection in neonate subjects, in particular identifying and treating early onset sepsis (EOS) in a neonate subject, wherein the level of PSP is determined in a body fluid, e.g. serum, and high level of PSP is indicative of the development of EOS.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 : Positive predictive values of different bioscores based on (A) two (PCT and PSP), (B) three (PCT, PSP and sTREM-1 ), and (C) four (PCT, PSP, sTREM-1 and CRP) sepsis markers with the corresponding 95 % confidence intervals. Dashed line: Regression line. Horizontal axis: numbers indicate the total number of marker(s) simultaneously above the cut-off value .

Vertical axis: Positive predictive value (%).

PSP = pancreatic stone protein, PCT = procalcitonin, sTREM-1 = soluble triggering receptor expressed on myeloid cells-1 , CRP = C-reactive protein.

Figure 2: Receiver operating characteristic curve of PSP in infected infants versus PSP in cord blood of healthy infants of the same gestational age (>34 weeks, n=56).

x-axis: 1 -specificity, y-axis: sensitivity

DETAILED DESCRIPTION OF VARIOUS AND PREFERRED EMBODIMENTS OF THE INVENTION

The present invention relates, in one embodiment, to a method of predicting and/or diagnosing a systemic infection in humans, in particular for predicting the development of sepsis, wherein the level of pancreatic stone protein / regenerating protein (PSP) and the level of procalcitonin (PCT) are determined individually in a body fluid sample, e.g. serum, and a high level of both, PSP and PCT, is indicative of the development of sepsis at early stages of the disease.

Furthermore the invention describes a method of improving the diagnosis performance of PCT wherein the levels of PSP and PCT are determined in, e.g., serum and their values are both considered: A high level of both these two biomarkers is more indicative of early- onset sepsis in neonates and sepsis in adults than the level of PCT alone.

The term "sepsis" as used in the present application describes a clinical picture, in which, as a rule, fever, leukocytosis, consciousness changes, a hyperdynamic circulation ("warm shock"), and a hyper-metabolic status, mainly as a consequence of the invasion of the normally sterile tissue by microorganisms, are observed. The positive detection of germs in the blood, which was previously understood to be a characteristic for a sepsis, has become less important for the diagnosis "sepsis", but may, in certain embodiments, form part of the evaluation. The term "sepsis" as used in this application corresponds to the definition of sepsis as defined in the "Definitions for Sepsis and Organ Failure and Guidelines for the Use of Innovative Therapies in Sepsis" from the ACCP/SCCM consensus conference committee (Bone et al., Chest 1992, 101 ;1644-1655). According to this definition, sepsis corresponds to a severe inflammatory response syndrome caused by the presence of an infection (either proven or probable).

As used herein, "severe sepsis" corresponds to sepsis in association with organ dysfunction, as defined in the "Definitions for Sepsis and Organ Failure and Guidelines for the Use of Innovative Therapies in Sepsis" from the ACCP/SCCM consensus conference committee (Bone et al., Chest 1992, 101 ;1644-1655).

As defined herein, "early onset sepsis" (EOS) refers to sepsis in humans within 72 hours of age.

As defined herein, "systemic infection" is an infection in which the pathogen is in the bloodstream, and, as a result, the infection is widespread throughout the body. As defined herein, "peritonitis" is an inflammation of the peritoneal membrane caused by either infectious or non-infectious reasons. The vast majority of clinically significant peritonitis is of infectious origin caused by bacteria.

As defined herein, "intra-abdominal infection" denotes peritonitis caused by bacteria (e.g., a local inflammatory process initiated by bacteria and their toxins) or virus. Infectious peritonitis may be regarded as the localized equivalent of systemic sepsis. Because of the heterogeneous pattern of abdominal sepsis (the term generally given to sepsis that originates from the abdomen), it is often considered difficult to define the disease precisely or appraise its severity, thereby impeding the diagnostic progress. It is important for physicians to discriminate patients suffering from a local intra-abdominal infection from those suffering from an intra-abdominal infection with a systemic reaction.

Infectious peritonitis can be classified as primary, secondary, or tertiary. In primary peritonitis (also called spontaneous bacterial peritonitis), the source of infection does not arise from the gastrointestinal tract, and there is no identifiable anatomical derangement of the intra-abdominal viscera. Primary peritonitis is mostly caused by a chronic liver disease, such as cirrhosis. In contrast, secondary peritonitis is due to an infection of the abdominal viscera, and may arise from a defect in the wall of abdominal organs as a consequence of perforation, ischemic necrosis, or penetrating injury. Secondary peritonitis may be local (often abscesses) or diffuse. Tertiary peritonitis is defined as peritonitis that persists or recurs after more than one failed source control procedure, and is highly frequent in patients requiring intensive care unit admission for severe abdominal infections.

As defined herein, "patient" refers to any adult or neonate mammalian animal including human, dog, cat, cattle, goat, pig, swine, sheep and monkey. Patients are preferably humans.

As defined herein, a sepsis treatment or a therapy against sepsis corresponds to the medical care of a sepsis patient or an early onset sepsis patient according to current practice guidelines, based on the recommendations of 2008 (Dellinger et al., Intensive Care Med, 2008, 34:17-60), which are an updated version of the first edition guidelines from 2001 coordinated by the International Sepsis Forum (ISF), and of the second editions from 2004 coordinated by the Society of Critical Care Medicine (SCCM), the European Society of Intensive Care Medicine (ESICM), and ISF (Dellinger et al., Intensive Care Med, 2004, 30:536-555).

Body fluids useful for determination of PSP and of PCT include, but are not limited to, serum, blood plasma, whole blood, urine, sputum, cerebrospinal fluid, tear fluid, sweat, milk, or extracts from solid tissue or from fecal matter.

In a preferred aspect, the method of the invention is for the diagnosis and/or prognosis of sepsis or of early onset sepsis.

In a further aspect, the present invention relates to an ex-vivo method of prognosis and/or diagnosis of sepsis or early onset sepsis in a patient, or to an ex-vivo method for predicting the development of infections after abdominal surgery in a patient, comprising: a) providing a body fluid sample from said patient;

b) determining the levels of both PSP and PCT in said sample;

c) comparing the level of PSP determined in step b) with a reference value and comparing the level of PCT determined in step b) with a reference value;

wherein higher levels of PSP and of PCT determined in step b), compared to their respective reference value, is indicative of the development and of the severity of sepsis and is predictive of the outcome.

Furthermore the invention relates to a method of treatment of sepsis or early onset sepsis, or infections after abdominal surgery, comprising steps a) to c) as explained in the preceding paragraph, and a further step d), using results of step c) to take accurate treatment actions according to current practice guidelines.

In a further aspect, the invention relates to a method for identifying and treating a patient with sepsis or early onset sepsis, comprising:

a) providing a body fluid sample from said patient;

b) determining the levels of both PSP and PCT in said sample;

c) comparing the level of PSP determined in step b) with a reference value of PSP and comparing the level of PCT determined in step b) with a reference value of PCT ;

d) combining the levels of PSP and of PCT wherein combined higher levels of PSP and of PCT determined in step b), compared to their respective reference value, is indicative of the development and of the severity of sepsis and is predictive of the outcome.

e) administering a therapy against sepsis to the patient identified as having sepsis or early onset sepsis

The invention is also directed at a method of identifying and treating systemic infection in neonate subjects, in particular identifying and treating early onset sepsis (EOS) in an neonate subject, wherein the level of PSP is determined in a body fluid, e.g. and high level of PSP is indicative of the development of EOS

In a further aspect, the invention relates to a method of treating sepsis or early onset sepsis in a neonate patient, comprising:

a) providing a body fluid sample from said neonate patient;

b) determining the level of PSP in said body fluid sample;

c) comparing the level of PSP determined in step b) with a reference value of PSP; wherein a higher level of PSP determined in step b), compared to a reference value, is indicative of the development and of the severity of early onset sepsis and is predictive of the outcome, and

d) administering a therapy against sepsis to the neonate patient identified as having early onset sepsis.

According to another aspect of the invention, the method of the invention is an in vitro diagnostic method for identifying sepsis or early onset sepsis in a neonate patient, comprising:

a) providing a body fluid sample from said neonate patient; b) determining the level of PSP in said sample;

c) comparing the level of PSP determined in step b) with a reference value of PSP; wherein a higher level of PSP determined in step b), compared to a reference value, is indicative of the development and of the severity of early onset sepsis and is predictive of the outcome.

In a specific aspect of the invention, the reference values are the levels of PSP and of PCT measured in a body fluid sample from a patient without known or suspected infection.

In a specific aspect of the invention, in neonates, the reference values of PSP and of PCT are the levels of PSP and of PCT in cord blood of healthy neonates of the same gestational age. In another aspect of the invention, the reference value for PSP in adult patients is about 10 ng/ml. In a further aspect of the invention, the reference value in adult patients is 25 ng/ml.

In another aspect of the invention, the reference value for PSP in neonates is about 9 ng/ml. In a further aspect, the reference value in neonates is 6.5 ng/ml, determined in cord blood from healthy neonates without infection.

In another aspect of the invention, the reference value for PCT in adult patients is about

2 ng/ml. In a further aspect of the invention, the reference value in adult patients is 0.5 ng/ml.

In another aspect of the invention, the reference value for PCT in neonates is about 9 ng/ml or below. In a further aspect of the invention, the reference value in neonates is

3 ng/ml.

The PSP and PCT levels indicative of development and outcome of sepsis are dependent on the body fluid chosen for the determination and on the age of the patient. For blood serum and blood plasma in adults, these respective levels are 60 to 130 ng/ml for PSP and 0.5 ng/ml for PCT, on days 0, 1 , 2, 3, 4 or 5 after abdominal surgery or after trauma, or intubation for mechanical ventilation. Hence, more specifically, the invention relates to a method of prediction and/or diagnosis of the development of sepsis, wherein the levels of both pancreatic stone protein / regenerating protein (PSP) and procalcitonin (PCT) are determined in serum or plasma, and the combination of a level of PSP of 60 ng/ml or more, in particular a level of 130 ng/ml or more, and a level of PCT of 0.5 ng/ml or more on days 0, 1 , 2, 3, 4 or 5 is indicative of the development of sepsis, of the severity of sepsis and of the complications and risk of mortality linked to said sepsis.

For blood serum and blood plasma in neonates, these respective levels are 9 ng/ml or more for PSP and 3 ng/ml or more for PCT, within 72 hours of age. Hence, more specifically, the invention relates to a method of prediction and/or diagnosis of the development of early onset sepsis, wherein the level of PSP and PCT is determined in serum, and the combination of a level of PSP of 9 ng/ml or more, and of a level of PCT of 3 ng/ml or more, within 72 hours of age is indicative of the development of sepsis.

The detection and quantification of serum or plasma PSP and of PCT are accomplished e.g. by a sandwich ELISA. In an adult, a level of PSP of 130 ng/ml or more together with a level of PCT of 0.5 ng/ml or more, at day 0 or 1 after abdominal surgery is indicative of post-surgery sepsis, in particular of the risk of developing severe complications and the risk of mortality post surgery. Any known method may be used for the determination of the level of PSP and of PCT in body fluids. Methods include, but are not limited to, ELISA, RIA, EIA, mass spectrometry, or microarray analysis. Such methods when used for diagnosis of systemic infection, e.g. sepsis, are a further object of the invention. In order to determine the level of PSP, a sandwich ELISA may be used on the basis of a guinea pig antiserum raised against recombinant human PSP and a rabbit antiserum against the same protein. Alternatively, another sandwich ELISA may be used on the basis of a pair of mouse anti-human PSP monoclonal antibodies. A preferred pair of mouse anti-human monoclonal antibodies is constituted of (i) the monoclonal antibody produced by the hybridoma 2D8-2C5-3G8-3H 12 (ECACC Deposit Reference 12082202) and (i) the monoclonal antibody produced by the hybridoma 6E5-3B2 (ECACC Deposit Reference: 12082201 ).

Other preferred methods of PSP detection and of PCT detection are radioimmunoassay or competitive immunoassay using a single antibody and chemiluminescence detection on automated commercial analytical robots. Microparticle enhanced fluorescence, fluorescence polarized methodologies, or mass spectrometry may also be used. Detection devices, e.g. microarrays, are useful components as readout systems for PSP and PCT.

It is shown that in secondary peritonitis patients, a score combining PSP and PCT improves of the diagnostic accuracy for severity of sepsis (e.g. organ failure) and predictive value for death in comparison to PCT alone or to PSP alone or to traditional sepsis markers.

It is shown that in secondary peritonitis patients, combining PSP and clinical score such as APACHE II or SOFA significantly improves of the diagnostic accuracy for severity of sepsis and predictive value for death in comparison to PSP or to said clinical score taken.

It is also shown that the combination of PSP and PCT levels determined in patients within 72 h of birth improves the positive predictive performance of PSP and PCT taken individually for the early diagnosis of sepsis. Combining PSP and PCT significantly improves the diagnosis of EOS in comparison to PCT alone or to PSP alone or to traditional sepsis markers. The diagnostic performance of PSP and PCT is far superior compared to traditional sepsis markers. A combination score clearly improves diagnosis of sepsis. PSP is a valuable biomarker to diagnose EOS, which when used in combination with PCT, improves the diagnostic performance of PCT.

The present invention is unexpected and not obvious, since in most cases, the diagnostic or predictive performance of the combination of two given biomarkers just equals the performance of the most performing of the two given biomarkers. In other words, given two biomarkers A and B, and A being superior to B in terms of performance, the combination A + B is generally expected to be comparable to A alone, and is not improved. For instance, the performance of the combination of PSP with other

inflammatory biomarkers such as CRP (or IL-6) is not superior to the individual performance of either PSP or CRP (or IL-6). However, surprisingly, the performance of the combination of PSP and PCT is significantly superior to the individual performance of PSP and PCT, respectively.

The methods of the invention are useful for sorting the patients according to groups of risks regarding severity and outcome of sepsis. As a consequence, regular and intensive follow-up and care, possibly including antibiotic treatment and prolonged care, should be applied to patients belonging to the group(s) of high risk(s). Importantly, better biomarkers and improved application of biomarkers to clinical practice can result in a significant reduction of unnecessary antibiotic treatment and reduction in hospital stay.

EXAMPLES

Example 1: Combining PSP and PCT improves the diagnostic performance of EOS Test patients

This study included 155 neonates born after 34 0/7 weeks admitted within the first 72 hours of life to the Neonatal Intensive Care Unit with suspicion of sepsis that were empirically treated for EOS. Blood was sampled at admission together with routine laboratory evaluation for neonatal infection (including CRP, PCT, Immature by total (l/T) ratio and white blood cell (WBC) count). Children with congenital malformations were excluded. In 18 infants, the amount of sampled blood was insufficient to perform additional laboratory analyses, leaving 137 infants for whom PSP and PCT were determined. One infant required vasopressor support for septic shock and one infant required invasive ventilation due to sepsis. No infant died. Three infants developed positive blood cultures, all with group B streptococcus (GBS). In total, 33/137 infants were classified as infected (including 3 with proven infection and 30 classified as probable infection), in contrast to 104 uninfected infants which were used as control group (52 classified as possible infection and 52 as unlikely infection). Baseline characteristics of the infected versus the uninfected group are shown in Table 1.

Table 1 : Baseline characteristics of infected vs. uninfected patients

Characteristic Infected n=33 Uninfected n=104 p-value†

Number (range) or (%) Number (range)

Male sex 18 (55%) 68 (65%) 0.26

Gestational age [weeks] 39.9 (34.0 to 41 .6) 38.9 (34.0 to 42.0) 0.07 Birth weight [grams] 3335 (1950 to 4400) 3060 (1630 to 4750) 0.52

Cesarean section 23 (70%) 54 (59%) 0.30 Apgar 1 min 7 (1 to 9) 7 (2 to 9) 0.64 Apgar 5 min 8 (3 to 10) 8 (2 to 10) 0.77 Apgar 10 min 9 (3 to 10) 9 (2 to 10) 0.51

Umbilical artery pH 7.24 (6.89 to 7.42) 7.25 (6.80 to 7.42) 0.77

Age at study entry (h) 12 (2 to 67) 8 (0 to 72) 0.21

Median values (range), or number (percentage) are given where appropriate.

† p-value Mann-Whitney test Diagnosis of infection

The likelihood of infection was assessed at 24 to 72 hours after admission into four categories based on cultures, perinatal sepsis risk factors, clinical signs of sepsis and laboratory examinations (Stocker et al. BMC Pediatrics 2010, 10:89): 1. proven infection (positive blood or cerebrospinal fluid cultures), 2. probable infection (negative cultures, ≥3 abnormal findings), 3. possible infection (negative cultures, two abnormal findings) and 4. unlikely infection (negative cultures, single abnormal finding).

The following factors were considered as abnormal findings: i) maternal risk factors (group B streptococcus (GBS) positivity, prolonged rupture of membranes >18 hours, chorioamnionitis), ii) clinical symptoms (respiratory distress / apnea, tachycardia / bradycardia, arterial hypotension / poor perfusion, seizures / floppy infant, irritability / lethargy / poor feeding, vomiting / feeding intolerance / ileus) and iii) conventional laboratory examinations (WBC, CRP, l/T ratio) and iv) blood and cerebrospinal fluid cultures.

For this marker study, an infected group (proven or probable infection), versus an uninfected group (possible or unlikely infection) was defined a priori. Sepsis was defined according to the ACCP/SCCM consensus conference committee of 1992: "Definitions for Sepsis and Organ Failure and Guidelines for the Use of Innovative Therapies in Sepsis" from the ACCP/SCCM consensus conference committee (Bone et al., Chest 1992, 101 ;1644-1655).

PSP and infection markers in infected versus uninfected infants

In the 137 children included in the EOS study, median PSP concentration was 8.75 ng/ml ranging from 2.02 to 99.40 ng/ml (IQR 5.18 - 13.33). PSP concentration was weakly correlated with birth weight (Spearman's Rho 0.16, p=0.08), gestational age (Rho 0.18, p=0.03) and with postnatal age at study sampling (Rho 0.17, p=0.04). PSP at admission correlated significantly with CRP at admission (Rho 0.35, p<0.001 ), maximal CRP during hospitalisation (Rho 0.46, p=<0.001 ), PCT (Rho 0.23, p=0.028), sTREM-1 (Rho 0.21 , p=0.025), and WBC count (Rho -0.23, p=0.008), but not with l/T ratio (Rho 0.12, p=0.16). PSP was significantly higher in infected infants (median 1 1.3 ng/ml) compared to uninfected infants (median 7.45 ng/ml, Mann-Whitney p=0.001 ). Infected infants had significantly higher PCT (median 20.33 versus 2.0 ng/ml, p<0.001 ) higher CRP (median 6 vs. 1 mg/l, p=0.004), and trendwise higher sTREM-1 levels (median 1436 vs 1234 pg/ml, p=0.055), while MIF, IT ratio and WBC were not significantly different between the groups (p>0.1 ). Receiver-operating characteristics (ROC) curve analysis to diagnose EOS:

PSP and PCT performed best among all parameters studied to distinguish infected infants from uninfected infants (Table 2): The area under the curve on ROC curve analysis resulted at 0.69 (95%-CI 0.59 - 0.80, p<0.001 ) for PSP, and at 0.77 (95%-CI 0.66 - 0.87, p<0.001 ) for PCT. The ROC curve analysis for CRP showed an area under the curve of 0.66 (p=0.006), and of 0.62 for sTREM-1 (p=0.055) whereas other studied markers were not significant. Based on ROC analyses, optimal cut-offs for each sepsis marker were defined. Table 2 shows the resulting sensitivity, specificity, and positive and negative predictive value (PPV, NPV). A cut-off of PSP above 9 ng/ml resulted in a sensitivity of 79% and a specificity of 63% with a NPV of 90%.

Table 2: Receiver-operating characteristic curve analysis and diagnostic performance for different infection markers and bioscores^*.

Individual ROC

Biomarkers AUC p-value Cutoff Sens Spec PLR NLR PPV NPV

% % % %

PSP 0.69 0.001 >9.0 ng/ml 79 62 2.1 0.35 39 90

PCT 0.77 0.001 >3 ng/ml 80 58 1 .9 0.35 41 89 sTREM-1 0.62 0.055 >1 .25 ng/ml 75 52 1 .6 0.48 34 87

MIF 0.54 0.54 >50 ng/ml 84 30 1 .2 0.54 29 85

CRP 0.66 0.006 >20 mg/l 36 89 3.4 0.7 52 82

IT ratio 0.56 0.28 >0.2 70 40 1 .2 0.75 27 80

WBC 0.48 0.69 >15^*10(9)/l 48 51 1 1 24 76

Bioscore ROC

models AUC p-value Score Sens Spec PLR NLR PPV NPV

% % % %

2 markers

PSP + PCT 0.81 <0.001 1 (1/2 posi) 96 35 1 .5 0.12 35 96

2 (2/2 posi) 64 90 6.3 0.4 70 87

3 markers

PSP+PCT+

sTREM-1 0.81 <0.001 1 (1/3 posi) 100 24 1 .3 0 33 100

2 (2/3 posi) 82 58 1 .9 0.32 42 90

3 (3/3 posi) 55 93 8 0.49 75 85 4 markers

PSP+PCT+

sTREM+CRP 0.81 <0.001 1 (1/4 posi) 100 22 1 .3 0 32 100

2 (2/4 posi) 86 54 1 .9 0.25 41 91

3 (3/4 posi) 55 90 5.4 0.51 67 84

4 (4/4 posi) 32 97 9.4 0.71 78 79

^*ROC AUC, receiver-operating curve area under the curve; Sens, sensitivity; Spec, specifity; PLR, positive likelihood ratio; NLR, negative likelyhood ratio; PPV, positive preditive value; NPV, negative predictive value; posi, positive. Performance of bioscores to predict EOS:

Based on ROC curve analyses, PSP, PCT, sTREM-1 and CRP were used for the bioscore models. Table 2 shows the diagnostic performance of different bioscores based on two, three or four of the sepsis markers. The bioscore based on PSP and PCT alone performed similar to more complex bioscores and was superior to PCT or PSP alone. The combined PSP/PCT score had a NPV of 96% if one at least one marker was above cutoff and a PPV of 70% if both were positive.

Multivariate logistic regression models showed that increased PSP was best at predicting EOS (Table 3). PSP above 9 ng/ml showed the strongest association with EOS, with an Odd's ratio of 15.7 (95%-CI 2.9 - 85.5, p<0.001 ). Again, a bioscore based on PSP and PCT performed similar to bioscores adding sTREM-1 or CRP, suggesting that neither sTREM-1 nor CRP provided additional independent information (Figure 1 ).

Table 3: Multivariate logistic regression models to predict early-onset sepsis^*.

Individual marker Odd's ratio

models Marker p-value (95%-CI)

2 individual markers PSP >9 ng/ml <0.001 13.8 (3.4-56.5)

PCT >3 ng/ml 0.004 7.0 (1.8-26.8)

3 individual markers PSP >9 ng/ml <0.001 17.3 (3.3-91.5)

PCT >3 ng/ml 0.1 3.4 (0.8-14.6)

sTREM-1 >1.25 ng/ml 0.26 2.7 (0.5-14.5)

4 individual markers PSP >9 ng/ml 0.002 15.7 (2.9-85.5)

PCT >3 ng/ml 0.15 2.9 (0.7 - 12.7)

sTREM-1 >1.25 ng/ml 0.29 2.5 (0.5-13.7)

CRP >20 mg/l 0.3 2.5 (0.4 - 14.7)

Odd's ratio Correct

Bioscore models Score p-value (95%-CI) prediction

2 markers (PSP,

PCT) 1 0.08 7.5 (0.8-70.0)

2 <0.001 81.8 (7.8-861.8)

model <0.001 9.8 (3.4-28.3) 83%

3 markers (PSP,

PCT, sTREM-1) model <0.001 5.0 (2.0 - 12.4) 84%

4 markers (PSP,

PCT, sTREM-1,

CRP) model <0.001 4.0 (1.9-8.6) 80%

^*Models are adjusted for gestational age, birth weight, sex and postnatal age at study sampling.

Comparison with PSP levels in cord blood:

In order to establish reference values of PSP in neonates, PSP was measured in cord blood of 78 preterm and term neonates without major neonatal disease. Median PSP concentration in cord blood was 4.48 ng/ml ranging from undetectable to 27.98 ng/ml (IQR 2.57 - 8.62). PSP concentration correlated significantly with birth weight (Spearman's Rho 0.52, p<0.001 ), and gestational age (Rho 0.47, p<0.001 ). When comparing PSP in infected infants versus PSP in cord blood of healthy infants of the same gestational age (>34 weeks, n=56), the AUC resulted at 0.78 (95%-CI 0.69 - 0.88, p<0.001 ) (Figure 2). A cut-off of PSP above 6.5 ng/ml resulted in a sensitivity of 80% and a specificity of 54%.

Example 2: Combining PSP and PCT improves outcome prediction in patients with peritonitis at the Intensive Care Unit

Test patients

In this prospective cohort study, 91 consecutive post-operative patients admitted to the intensive care unit (ICU) with proven diagnosis of secondary peritonitis according to the MPI score and meeting sepsis definition criteria were included in the study. For all patients, the data collected were obtained from their first admission to the ICU after their first operation for peritonitis. Age, gender or pre-existing disease was no reason for exclusion. Patients were not included in the study if blood was not taken within three hours from admission to the ICU or if patients were transferred from other hospitals. At ICU admission, blood samples were taken within three hours for analysis. The Mannheim Peritonitis Index (MPI) and APACHE II clinical scores were also determined. Table 4 shows the patient characteristics. To assess the potential peritonitis severity after surgery, this study analyzed patients that exhibited various degrees of post-operative sepsis related complications including organ failure and death.

Association of blood parameters with clinical conditions, clinical scores and outcomes The association of WCC, CRP, IL-6, PCT, and PSP with different clinical conditions, such as the severity and localization of peritonitis, the presence of organ failure and mortality in the ICU, was investigated. PSP was the only blood parameter that significantly differed among all different clinical conditions and nearly all of their subgroups on univariate analysis (Table 5). In addition, PSP best correlated with the clinical scores (MPI, APACHE II, and SOFA scores) when compared to WCC, CRP, IL6, and PCT. Similarly, PSP was the only blood parameter that significantly differed among the clinical scores when grouped according the cut-off points generated by the ROC curves.

Table 4: Patient characteristics of the 91 patients with peritonitis

Patient Characteristics n = 91

Age, median (IQR) 66 (50-72)

Gender, male/female, number (%) 53/38 (58%/42%) Diagnosis, number (%)

Benign vs. Malignant 62/29 (68%/32%)

Colon perforation 20 (22%)

Small bowel perforation 19 (21 %)

Gastric perforation 17 (19%)

Pancreatic tumor 14 (15%)

Gall bladder empyema 8 (9%)

Mesenteric ischemia 7 (8%)

Liver abscess rupture 4 (4%)

Appendicular perforation 2 (2%)

MPI^* Score, median (IQR) 30 (20-33)

APACHE II Score*, median (IQR) 18 (14-26)

SOFA Score^**, median (IQR) 6 (3-10)

Catecholamines at admission, number (%) yes/no 50/41 (55%/45%)

Dobutamine and Noradrenaline, number (%) 25 (27%)

Noradrenaline, number (%) 21 (23%)

Dobutamine, number (%) 4 (5%)

p02/fi02ratio, median (IQR) 226.7 (164.4-322.0) White Cell Count (WCC), median (IQR) 15 (1 1 -20)

C-reactive Protein (CRP), median (IQR) 222 (143-291 ) lnterleukin-6 (IL-6), median (IQR) 88 (34-375)

Procalcitonin (PCT), median (IQR) 1 .07 (0.27-6.10)

Pancreatic Stone Protein (PSP), median (IQR) 125 (25-419)

Organ failure^***, number (%) yes/no 61/30 (67%/33%)

Organ failure, median number of organs (IQR) 2 (0-3)

Renal Replacement Therapy, number (%) yes/no 22/69 (24%/76%)

Intra-abdominal complications^***, number (%) yes/no 4/87 (4%/96%)

Re-laparotomy, number (%) yes/no 48/43 (53%/47%)

Mortality rate, number (%) 23/91 (25%)

Multi-organ failure 19 (20%)

Acute Respiratory Distress Syndrome (ARDS) 2 (2%)

Cardiac arrest 1 (1 %)

Hemorrhagic shock 1 (1 %)

$ APACHE II: "Acute Physiology and Chronic Health Evaluation II" is a severity-of -disease classification system and applied within 24 hours of admission of a patient to an intensive care unit ^*MPI Score: Mannheim Peritonitis Index. Performed on suspicion of peritonitis. The official cut-off point for a positive MPI score is≥ 26.

II Grouped more vs. less than 130. Cut-off point generated by ROC curves and the Yuden's Index.

*^*SOFA indicates: The Sequential Organ Failure Assessment score

^***Two patients developed a chronic fistula and two developed intra-abdominal adhesions causing bowel obstruction.

Table 5: Differences of PSP and PCT in the 91 patients with different clinical conditions

PCT PSP

Median Median

n P value P value (i.q.r.) (i.q.r.)

1 .45 30.6

Localisation Localised 26 0.20 0.01

(0.28-7.39) (19.7-262.17)

0.90 140.1

Diffuse 65 - - (0.18-6.00) (28.9-518.1 )

0.85 30.3

Severity Minor 30 Ref Ref

(0.17-5.04) (21.7-234.3)

1 .07 122.3

Moderate 50 0.43 0.03

(0.18-6.10) (24.0-521 .8)

1 .13 201 .3

Severe 1 1 0.93 0.061

(0.62-7.34) (136.5-514.5)

0.40 25.4

Organ failure None 30 Ref Ref

(.013-1 .42) (19.4-108.0)

1 .31 185.9

1 -3 organs 55 0.14 <0.001

0.47-8.41 ) (47.8-500.9)

14.60 721 .4

> 3 organs 6 0.38 0.047

(6.00-40.78) (514.5-830.5)

0.83 75.0

Status Alive 68 0.14 0.003

(0.15-4.84) (20.8-230.5)

1 .19 499.3

Dead 23 - - (0.62-19.52) (136.5-625.5)

^* The Tests for Several Independent Samples procedure compares two or more groups of cases on one variable. The Kruskal-Wallis H test, an extension of the Mann-Whitney U test, is the nonparametric analog of one-way analysis of variance and detects the overall differences in distribution location. Post hoc multiple comparisons determine which of the sub-group medians significantly differ from the first group. As equal variances were not assumed, the Dunnett's pairwise comparison test was chosen. For all subgroups, the first subgroup was used to compare with the remaining groups. For example, for the severity group, the sub-group "None" was used to compare with the "Minor", "Moderate", and "Severe" subgroups.

Ref: indicates the subgroup where the other two other subgroups were compared with when performing pairwise comparisons.

The localization was defined as "Localized" when the infection was contained within a single organ part of the abdominal cavity, and results in an intra-abdominal abscess while minimal or no inflammation of the overlying peritoneum and no anatomic disruption of the Gl tract was seen. The localization was defined as "diffuse" when the infection was uncontained and spread to the whole abdomen, affecting most of the peritoneum.

The severity was

"-" indicates: not applicable.

Blood parameters and clinical score prediction of death in the ICU

To evaluate the predictive value of each blood parameter, the ideal cut-off points to predict death in the ICU was first identified using ROC curve analysis. Based on the Yuden's Index (giving equal weight to both sensitivity and specificity), the cut-off points identified are listed in Table 6. As the blood samples of three patients were not available for all the measurements, these patients were not included in this analysis, which was then conducted on a total of 88 patients. The positive and negative predictive value, the positive and negative likelihood ratios, the relative risk and odds ratios were also calculated. Clearly, PSP was superior to WCC, CRP, IL6, and PCT in predicting death in the ICU. In addition, a multivariate logistic regression analysis on different clinical scores and blood parameters for the prediction of mortality shows that the combination of PSP and PCT has a significantly greater specificity than that of PSP alone, and PCT alone, respectively, and that the combination of PSP and APACHE II has a significantly greater predictive value, specificity and AUC than that of PSP alone, and APACHE II alone, respectively (Table 7). Table 6: Blood parameters and clinical scores predicting death in the intensive care unit.

Total: n=88, deaths:

Yuden's Cut-off

n=23, mortality rate: AUC Sensitivity Specificity

Index^* point^*

25%

0.13 20 0.49 0.39 0.74

C-Reactive Protein (CPR) 0.1 1 175 0.48 0.70 0.42

White cell count (WCC) 0.34 90 0.64 0.74 0.60

PCT 0.27 0.5 0.63 0.87 0.40

PSP 0.41 130 0.78 0.78 0.62

MPI 0.37 30 0.72 0.78 0.56

APACHE II 0.65 22 0.82 0.87 0.78

SOFA 0.66 10 0.88 0.83 0.84

Total: n=88, deaths: n=23,

PPV* NPV^§ PLR^ NLR** OR*** jo value^† mortality rate: 25%

White cell count (WCC) 0.32 0.78 1 .400 0.845 1 .658 0.862

C-Reactive Protein (CPR) 0.28 0.79 1 .154 0.767 1 .505 0.820 lnterleukin-6 (IL-6) 0.39 0.86 1 .179 0.446 3.988 0.520

PCT 0.33 0.90 1 .442 0.329 4.390 0.058

PSP 0.42 0.90 2.129 0.344 6.192 <0.001

MPI 0.38 0.88 1 .774 0.389 4.560 0.007

APACHE II 0.57 0.95 3.942 0.167 23.556 <0.001

SOFA 0.63 0.93 5.107 0.207 26.614 <0.001

^* Grouped more vs. less than the cut-off point generated by ROC curves Yuden's Index. ^† P values were computed by the Fischer's Exact test for categorical variables.

* PPV indicates Positive Predictive Value

^§ NPV indicates Negative Predictive Value

1¹ PLR indicates Positive Likelihood Ratio, values > 2 are considered clinically significant ^** NLR indicates Negative Likelihood Ratio, values closer to 0 and less than 0.20 are considered clinically significant

*^** OR indicates the common Odds Ratio

AUC indicates: Area Under the Curve (ROC Analysis) Table 7: Receiver-operating characteristic curve analysis and mortality prediction performance for PSP, PCT and APACHE II, and bioscore.

Individual ROC

Biomarkers AUC p-value Cutoff Sens Spec PPV NPV

% % % %

PSP 0.78 0.001 >130 ng/ml 79 62 39 90

PCT 0.63 0.058 >2 ng/ml 43 66 31 77

APACHE II 0.82 0.001 >22 87 78 57 95

Bioscore ROC

model AUC p-value Score Sens Spec PPV NPV

% % % %

2 markers

PSP + PCT 0.77 <0.001 1 (1/2 positive) 78 55 38 88

2 (2/2 positive) 43 72 36 78

2 markers

PSP

+ APACHE II 0.85 <0.001 1 (1/2 positive) 91 52 40 94

2 (2/2 positive) 74 88 68 90

Blood parameters and bioscore detection of organ failure

Among the 88 patients from which blood samples were available to measure the different blood parameters, 58 patients developed post-operative complications in presence of organ failures and severe sepsis, while 30 had no organ failure. To evaluate diagnostic performance and predictive value of PSP and PCT to predict post-operative organ failure and severe sepsis complications, a ROC curve analysis was performed and the ideal cutoff were identified. The cut-off points identified for PSP and PCT were 50 ng/ml and 2 ng/ml, respectively. The specificity, sensitivity, positive and negative predictive value were calculated. The performance of PSP to detect and predict organ failure and severe sepsis was superior to that of PCT (Table 8). In addition, a multivariate logistic regression analysis for the prediction of organ failure shows that the combination of PSP and PCT has a significantly greater positive predictive value, specificity than that of PSP alone, and PCT alone, respectively (Table 8). Table 8: Receiver-operating characteristic curve analysis and organ failure diagnostic performance for PSP and PCT, and bioscores.

Individual ROC

Biomarkers AUC p-value Cutoff Sens Spec PPV NPV

% % % %

PSP 0.80 <0.001 >50 ng/ml 76 67 81 59

PCT 0.73 <0.001 >2 ng/ml 47 83 84 45

Bioscore ROC

model AUC p-value Score Sens Spec PPV NPV

% % % %

2 markers

PSP + PCT 0.81 <0.001 1 (1/2 positive) 79 60 79 60

2 (2/2 positive) 43 90 89 45

In a third example, PSP and PCT are measured in plasma samples collected from patients at admission to the emergency room or ICU, and the combination of PSP and PCT improves diagnosis of sepsis, as compared to PSP alone or PCT alone. Once given the above disclosure, many other features, modifications, and improvements will become apparent to the skilled artisan. Such other features, modifications, and improvements are therefore considered to be part of this invention, the scope of which is to be determined by summary of the invention and the disclosure as a whole.

Claims

1 . A method of predicting and/or diagnosing sepsis in a patient, wherein the level of pancreatic stone protein / regenerating protein (PSP) and the level procalcitonin (PCT) are determined in a body fluid sample from said patient and a high level of both, PSP and PCT, is indicative of the development of sepsis.

2. An ex-vivo method of prognosis and/or diagnosis of sepsis or early onset sepsis, or for predicting the development of infections after abdominal surgery in a patient, comprising: a) providing a body fluid sample from said patient;

b) determining the levels of both PSP and PCT in said sample;

c) comparing the level of PSP determined in step b) with a reference value of PSP and comparing the level of PCT determined in step b) with a reference value of PCT;

3. A method of treatment of sepsis or early onset sepsis, or infections after abdominal surgery, comprising

a) providing a body fluid sample from said patient;

b) determining the levels of both PSP and PCT in said sample;

wherein higher levels of PSP and of PCT determined in step b), compared to their respective reference value, is indicative of the development and of the severity of sepsis and is predictive of the outcome; and

d) using the outcome of step c) to take accurate treatment actions according to current practice guidelines.

4. The method of claim 2 or 3, wherein the reference value of PSP and PCT are the levels of PSP and PCT in a control sample from a patient without known or suspected infection.

5. The method of claim 2 or 3, wherein the reference value for PSP and PCT in adult patients is 25 ng/ml for PSP and 2 ng/ml for PCT.

6. The method of claim 2 or 3, wherein the reference value for PSP and PCT in neonates patients is 2 ng/ml for PSP and 9 ng/ml for PCT.

7. The method according to any one of the preceding claims wherein the body fluid sample is serum.

8. The method of any one of claims 1 to 7 wherein the levels of PSP and PCT are determined by ELISA, RIA, EIA, mass spectrometry, or microarray analysis.

9. An ex-vivo method for detecting the development of systemic infection in a neonate patient comprising determining the level of pancreatic stone protein (PSP) in an isolated body fluid sample from said patient, and a high level of PSP is indicative of the development of systemic infection.

10. The method according to claim 9, comprising:

a) providing a body fluid sample from said neonate patient;

b) determining the level of PSP in said body fluid sample;

c) comparing the level of PSP detected in step b) with a reference value of PSP; wherein a higher level of PSP determined in step b), compared to the reference value, is indicative of the development and of the severity of early onset sepsis and is predictive of the outcome.

1 1. A method of treating early onset sepsis in a neonate patient, comprising:

a) providing a body fluid sample from said neonate patient;

b) determining the level of PSP in said body fluid sample;

12. The method according to claim 10 or 1 1 , wherein the reference value is the level of PSP measured in a body fluid sample from a neonate patient without infection.

13. The method according to claims 10 to 12, wherein the body fluid is cord blood.

14. The method of claim 10 to 13, wherein the reference value is 6.5 ng/ml.

15. The method of any one of claims 9 to 14 wherein the level of PSP is determined by ELISA, RIA, EIA, mass spectrometry, or microarray analysis.