WO2012067495A1

WO2012067495A1 - New diagnostic assay for determining severity of rsv infection

Info

Publication number: WO2012067495A1
Application number: PCT/NL2010/050765
Authority: WO
Inventors: Ronald De Groot; Peter Wilhelmus Maria Hermans
Original assignee: Stichting Katholieke Universiteit, More Particularly The Radboud University Nijmegen Medical Centre
Priority date: 2010-11-17
Filing date: 2010-11-17
Publication date: 2012-05-24

Abstract

The invention comprises a method for the prediction of the severity of a disease developing from an infection with human respiratory syncytial virus (RSV) in a subject, comprising the steps of: a. determining the plasma levels of interleukin 8 (IL-8) and RANTES (CCL5); b. determining the blood cell count of CD4+ T cells; c. making a prediction on basis of the values obtained in steps a) and b). In such a method the prediction is the development of a severe disease state of the RSV infection, if the plasma IL-8 level is more than 60 pg/mg, preferably more than 65 pg/ml, more preferably more than 67.2 pg/ml, the RANTES plasma level is less than 15 ng/ml, preferably less than 14 ng/ml, more preferably less than 13 ng/ml and the CD4+ cell count is less than 3*10 6 cells per ml blood, preferably less than 2.5*10 6 cells per ml blood, more preferably less than 2.03*10 6 cells per ml blood.

Description

Title: New diagnostic assay for determining severity of RSV infection

FIELD OF THE INVENTION

The invention relates to the field of diagnostics, more particularly diagnostics for viral diseases, especially diseases caused by Respiratory

Syncytial Virus (RSV).

BACKGROUND

Human respiratory syncytial virus (RSV) is a virus that causes respiratory tract infections. It is the major cause of lower respiratory tract infection and hospital visits during infancy and childhood. For most people, RSV produces only mild symptoms, often indistinguishable from common colds and minor illnesses. The Centers for Disease Control consider RSV to be the "most common cause of bronchiolitis (inflammation of the small airways in the lung) and pneumonia in children under 1 year of age in the United States". For some children, RSV can cause bronchiolitis, leading to severe respiratory illness requiring hospitalization and, rarely, causing death. This is more likely to occur in patients that are immunocompromised or infants born prematurely. Other RSV symptoms common among infants include listlessness, poor or diminished appetite, and a possible fever.

Recurrent wheezing and asthma are more common among individuals who suffered severe RSV infection during the first few months of life than among controls; whether RSV infection sets up a process that leads to recurrent wheezing or whether those already predisposed to asthma are more likely to become severely ill with RSV has yet to be determined.

One of the challenges for clinicians encountering children with RSV infections is to differentiate children who require hospitalization for supportive interventions from those who can be safely sent home. For instance, 35% of the children hospitalized with bronchiolitis do not receive any supportive intervention. On the other hand, clinicians do not want to discharge those children who may experience clinical detoriation. For about 4.6-6.8% of the children sent home with the diagnosis bronchiolitis it appeared that hospitalization was required later on during infection. There is thus need of diagnostic tools that can help to predict the severity of the disease, in order to help clinicians in the decision to hospitalize or not.

Nowadays the risk for severe disease upon RSV infection in children is estimated by the physician based on demographic data, medical history and clinical assessment. A disadvantage of current clinical prediction models for RSV infection is that they are mainly based on clinical signs that will be present in a late stage of clinical deterioration. Ideally, a clinical prediction model should be able to identify children before severe symptoms are clinically displayed.

One of the well known demographic risk factors, as also confirmed in the experimental part of the description, is young age. An immature immune system in combination with the lack of in utero sensitization to RSV and small airways makes that clinically most severe RSV infections are predominantly observed in young infants (age 1-3 months). The immaturity of the immune system in young children is reflected by a delay in recruitment of neutrophils and monocytes to infected tissues, less efficient antigen presentation by macrophages and dendritic cells and defects in neonatal adaptive immunity, such as an impaired production of Thl associated cytokines (IFN-γ) and diminished NK cell cytotoxicity reflect. The developing immune system and the resulting changing values of concentrations of T-cells, cytokines and other inflammatory markers over age, impair the realization of a reliable assay for diagnosing and/or predicting the severity of an infectious disease.

As said above, mainly clinical parameters have been used to predict severity of disease in RSV infection, but the use of clinical prediction models is limited. Several studies have associated severity of RSV disease with particular cytokines, such as IL-8, IL-6 and Thl and Th2 related cytokines such as IL-4 and IFN-γ (Bont, L. et al., 1999, Eur. Respir. J. 14:144-149; Hornsleth, A. et al., 1998, Pediatr. Infect. Dis. J. 17:1114-1121; Brandenburg, A.H. et al., 2000, J. Med. Virol. 62:267-277; Bermejo-Martin, J.F. et al., 2007, Eur. Cytokine Netw. 18:162-167). Hornsleth, A. et al. (2001, J. Clin. Virol. 21:163-170) described the relation between ratios of inflammatory mediators like cytokines, chemokines and receptors, and the severity of RSV infection and they found that the ratios of IL-l/RANTES, IL-8/RANTES, RANTES/IL-10 and TNF-R1/RANTES were correlated with severity.

Many, if not all, of these studies, however, have been conducted on nasopharyngeal secretions. Taking a nasopharyngeal aspirate from young children is cumbersome and unpleasant for the patient. More importantly, quantitative diagnosis on nasopharyngeal lavage fluid is practically

impossible: the amount of fluid used for the lavage and the pressure with which it is applied is variable as is the amount of biological material that is washed out with the fluid. It would therefore be advantageous if there would be an assay that would be feasible on blood or plasma, since this can be obtained more easily and taking a blood sample will be a common procedure for diagnosing the infection anyway. It has appeared, however, that the concentrations of the various inflammatory mediators are different between mucosal tissue, such as the nasopharyngeal tissue, and the blood (see Bermejo- Martin et al., supra). Also, it has been discussed that mechanical ventilation, as is often applied in children with bronchiolitis, influences the mucosa and plasma levels of chemokines and cytokines (Schultz, C. et al., 2001, Eur.

Respir. J. 17:321-324)

There is thus still need for an assay that can reliably predict the severity of an RSV infection, especially if such an assay can be performed on a blood or plasma sample. SUMMARY OF THE INVENTION

The inventors now have found that a reliable assay can be

performed for the prediction of the severity of an RSV infection by measuring the CD4+ T cell count and the concentration of IL-8 and

RANTES.

Therefore the invention comprises a method for the prediction of the severity of a disease developing from an infection with human respiratory syncytial virus (RSV) in a subject, comprising the steps of:

a. determining the plasma levels of interleukin 8 (IL-8) and

RANTES (CCL5);

b. determining the blood cell count of CD4+ T cells;

c. making a prediction on basis of the values obtained in steps a) and b) .

In such a method the prediction is the development of a severe disease state of the RSV infection, if the plasma IL-8 level is more than 60 pg/mg, preferably more than 65 pg/ml, more preferably more than 67.2 pg/ml, the RANTES plasma level is less than 15 ng/ml, preferably less than 14 ng/ml, more preferably less than 13 ng/ml and the CD4+ cell count is less than 3*10⁶ cells per ml blood, preferably less than 2.5*10⁶ cells per ml blood, more preferably less than 2.03*10⁶ cells per ml blood.

Further in such a method, the prediction is an increased risk of the development of a severe disease state of the RSV infection, if the plasma IL- 8 level is more than 60 pg/mg, preferably more than 65 pg/ml, more preferably more than 67.2 pg/ml, the RANTES plasma level is less than 15 ng/ml, preferably less than 14 ng/ml, more preferably less than 13 ng/ml or the CD4+ cell count is less than 3*10⁶ cells per ml blood, preferably less than 2.5*10⁶ cells per ml blood, more preferably less than 2.03*10⁶ cells per ml blood.

Preferably in such a method the IL-8 and RANTES plasma levels and the CD4+ cell count are assayed by immunological assays.

Further, in such a method the subject is a child, preferably a child having an age of less than 6 months, more preferably a child having an age of less than three months.

Also part of the invention is a detection kit comprising means for the detection of IL-8 and RANTES in plasma and for determining the CD4+ cell count in blood. Preferably such a detection kit additionally comprises means for detection of RSV particles.

Also part of the invention is the use of a kit according to the invention for predicting the severity disease state upon RSV infection in a subject.

LEGENDS TO THE FIGURES

Fig. 1: Absolute cell counts of leukocytes subsets in blood from children with an RSV infection categorized by disease severity. Y-axis:

Absolute cell counts (*10⁶ cells/ml). NK is natural killer cells.

Fig. 2: a. RANTES plasma concentrations in mild, moderate and severe group with RSV infections during acute infection and for the moderate and severe group also after recovery;

b. Ratio of RANTES plasma concentrations between acute and recovery samples in the moderate and severe patient group with RSV infection.

Fig. 3: Viral counts for the number of virus particles in mono- or co- infection (left panel) or relative to disease severity (right panel).

Fig. 4: Absolute cell counts during acute infection and after recovery in mechanically ventilated children.

DETAILED DESCRIPTION

In the following description and examples a number of terms are used. In order to provide a clear and consistent understanding of the specification and claims, including the scope to be given such terms, the following definitions are provided. Unless otherwise defined herein, all technical and scientific terms used have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

IL-8 (interleukine-8, also named CXCL8) is a chemokine produced by macrophages and other cell types such as epithelial cells (any cell having toll-like receptors). It is a member of the CXC chemokine family and one of the major mediators of an inflammatory response. Its relation to bronchiolitis is well known.

RANTES (also known as CCL5) is a chemokine that plays an active role in recruiting leukocytes into inflammatory sites. It is involved in more than 100 human diseases, and also a connection with bronchiolitis has been established (see e.g. Welliver, R.C., 2000, Pediatr. Asthma Allergy Immunol. 14:93-100).

Most of the CD4 expressing cells are mature T helper cells, which play an important role in establishing and maximizing the capabilities of the immune system.

"Plasma" as used in the present application is used in the normal meaning of the word, i.e. the liquid component of the blood. However, for assaying plasma levels, the assays may be performed on (full) blood, blood serum or plasma. CD4+ cell counts are, of course, performed on full blood.

RSV infection is defined as resulting in a "severe disease" state if the patient requires hospitalization, where supportive intervention can be given. This supportive intervention can be mechanical ventilation, treatment with antiviral medicine, and measures to prevent secondary clinical effects and co- infections with other pathogens affecting the respiratory pathways. Children without hypoxia or severe feeding problems are considered to be only mildly affected, those requiring hospitalization for supplemental oxygen (oxygen saturations <93%) and/or nasogastric feeding are considered to be moderately affected and children requiring mechanical ventilation are considered to be severely affected (see: Gern J.E. et al., 2002, Pediatr. Allergy Immunol. 13:386- 393; and Wan et al., 1992, Am. Rev. Respir. Dis. 145:106-109).

Many factors have been identified that play a role in RSV infection and the development of bronchiolitis. Among these are the not exhaustive list of Thl cytokines, like IL-16, 11-2, IL-12p70, IFNy, TNFa), Th2 cytokines, like IL-13, IL-4, IL-6, 11-10, chemokines, like IP-10, IL-8, MlPla (CCL3), MIP-16 (CCL4), RANTES, growth factors, like FGFb, PDGFbb, GCSF, VEGF, and IL- 1, IL-1RA and IL-17 (see e.g. Bermejo-Martin, supra).

The inventors now have found that plasma levels of IL-8, RANTES and the plasma CD4+ cell count can be used to reliably predict the severity of disease developing from an RSV infection in young children. First of all, it is surprising that such a prediction can be made by assaying plasma levels of the listed compounds. The RSV infection localizes in the upper respiratory tract of the body and predominantly affects the mucosal tissues that are present there. The immune reaction that is triggered by the virus thus has a strong local, mucosal character of which the local inflammation (bronchiolitis) is one of the most characteristic phenomena. This thus means that most of the components that are involved in the immune reaction (such as the above listed cytokines and chemokines) will be formed and will act locally. It is thus only logical that the local levels of those compounds will be predominantly influenced by the viral infection, while plasma levels may react more slowly or even remain virtually unaltered. It has also been discussed in the scientific literature that there is a difference between mucosal levels and plasma levels of these compounds. Pitrez et al. (Ann. Allergy Asthma Immunol., 92:659-662, 2004) found that cytokines produced by in vitro PBMCs may not necessarily reflect the concurrent cytokine pattern production at the mucosal surface in the respiratory tract of infants with acute bronchiolitis. Vieira et al. (J. Bras.

Pneumol., 36:59-66, 2010) found a positive correlation of disease severity with mucosal (nasopharyngeal) increased levels of sICAM-1 and IL-10, while in the serum IL-6 levels were found to be predictive. There is thus a recognized difference in cytokine response between mucosal tissue and serum levels. One further illustration of this can be found in the experimental results presented in the present description, where it was found for IL-6 that the level in nasopharyngeal aspirates was 10-20 fold higher than in plasma, but also that the average value in plasma was highest in severely diseased subjects, while for the nasopharyngeal level of IL-6 this was found to be the highest in mild and moderately affected subjects (see Table 3). These data thus also confirm the findings of Vieira et al. that IL-6 levels in serum are related with disease severity, while the levels in nasopharyngeal secretions are not.

A second surprising effect is that, amongst the many factors that are effected by RSV infection, only the three factors mentioned in the claims of the invention are needed to give a clinically useful prediction of the severity of the disease. Whereas in the past many factors have been deemed or illustrated to be correlated, it appears that the information conferred by the plasma levels of IL-8, RANTES and the CD4+ cell count is sufficient to give a clinically useful prediction.

IL-8, RANTES and the CD4+ cell count in the plasma may be analyzed according to standard procedures. For IL-8 and RANTES various kits are commercially available from a large number of vendors. Most of these kits are based on immunoassays (such as ELISA), but also other methods may be used. In the experimental part both IL-8 and RANTES are measured in an assay using flow cytometry. But also for RANTES various other immunoassays (based on EIA or ELISA) are commercially available. For measuring the CD4+ cell count also immune based assays are commercially available, such as based on anti-CD4 antibodies obtainable through eBioscience Ltd. (Hatfield, UK).

As can be seen from the experimental results (below) absolute levels can be given for all three parameters, which are used to determine whether the specific parameter qualifies as predictive or not. If all three parameters are found to be 'predictive' this is a reliable indication that the patient from which the samples are taken will develop a severe form of the RSV infection. This means that if the plasma level of RANTES is found to be below 15 ng/ml, preferably below 14 ng/ml, more preferably below 13 ng/ml, the plasma level of IL-8 is found to be over 60 pg/ml, preferably over 65 pg/ml and more preferably over 67.2 pg/ml and the number of CD4+ cells is found to be below 3*10⁶/ml, preferably below 2.5*10⁶/ml, more preferably below 2.03*10⁶/ml (see also Table 4), such blood levels are predictive for a severe development of the infection and treatment and hospitalization of the patient may be adjusted to this prediction. Treatment may comprise applying oxygen and/or mechanical ventilation, administration of bronchodilators and/or administrations of antiviral drugs, such as ribavirin and/or anti-RSV antibodies.

This prediction on basis of the three different inflammatory markers, IL-8, RANTES and CD4+ cell count, had a high sensitivity of 96%, and a specificity of 93% in our experiments. If only one or two of the

parameters reach the above- defined threshold, there is a suspicion that the child will develop a severe form or at least a moderate form of the disease and close monitoring of the child's condition is advised.

Also part of the invention is a detection kit comprising means for the detection of the plasma levels of IL-8 and RANTES, further comprising means for the assessment of the CD4+ cell count. Such a detection kit can be used by the pediatrician to predict the severity status of the disease that the child will develop after RSV infection. In order to speed up the clinical diagnosis, it is preferred that the detection kit also comprises means to detect viral particles of the virus. Normally, the test for identifying the viral cause of the disease is done separately, but of course it is more efficient to confirm the diagnosis of RSV infection concomitantly with the assay for predicting the severity of the disease that will develop. However, to be able to accurately predict the severity of the disease, the infection should be well established and local and systemic immune responses should already have developed. Again, for the detection of RSV particles in samples from the patient, many tests are commercially available, such as the Sure-Vue® test kit of Fisher Healthcare, the NOW RSV assay of Meridian Bioscience, the SimulFluor® RSV/Para 3 assay of Millipore, and many others. The tests to detect the presence of RSV can be based on immunological techniques, but may also be based on nucleic acid detection, such as (RT-)PCR.

The invention will be illustrated in the following Example(s), which is for illustrative purpose and not deemed to be limiting the invention as claimed.

EXAMPLES

Methods

Study design

Children younger than 2 years of age with laboratory confirmed RSV bronchiolitis were prospectively included during three consecutive winter seasons (from November to April in the years 2006-2009). Bronchiolitis was defined as an acute infection of the lower airways, characterized by increased respiratory effort (tachypnea and/or use of accessory respiratory muscles) and expiratory wheezing and/or crackles and/or apnea). Written informed consent was obtained from all parents and the study was approved by the Committee on Research involving Human Subjects of the University Nijmegen Medical Centre. Within 24 hours after admission a blood sample and nasopharyngeal aspirate was collected and parents from hospitalized children were asked permission to draw a second blood sample and nasopharyngeal aspirate 4 to 6 weeks after admission. Medical history, demographics and clinical parameters were collected from questionnaires and medical records. Patients were classified into three different groups based on severity of disease. Children without hypoxia or severe feeding problems were allocated in the mild group, those requiring hospitalization for supplemental oxygen (oxygen saturations <93%) and/or nasogastric feeding in the moderate group and children requiring mechanical ventilation in the severe group.

Sample collection

A nasopharyngeal aspirate was collected by introducing a catheter, connected to a collection tube and an aspiration system, into the

nasopharyngeal cavity. Then, 1.5 ml of saline was instilled into the catheter and, while slowly retracting the catheter, the nasopharyngeal fluid was aspirated in a collection tube. Afterwards the catheter was flushed with 1 ml of saline and added to the collection fluid. The samples were kept cold and immediately transferred to the laboratory. The nasopharyngeal aspirate was centrifuged at 500 g for 10 minutes at 4°C and the supernatant was frozen at - 80°C.

Five ml blood was collected into sodium heparin tubes and directly transferred to the laboratory. A thin blood smear was prepared and stained with (May-Grunwald-)Giemsa to determine the percentages of granulocytes. Direct immunophenotyping was performed on samples that were collected during working hours. From samples collected during the evening and weekends indirect immunophenotyping was performed on viably frozen peripheral mononuclear blood cells (PBMCs). PBMCs were obtained by density gradient centrifugation (Lymphoprep®, Axis Shield, Norway) and stored in liquid nitrogen after cryopreservation. Plasma samples were stored at -80°C for cytokine analyses.

Virus detection in nasopharyngeal secretions

The presence of RSV was confirmed by multiplex RT-PCR on nasopharyngeal samples as previously described (Templeton, K.E. et al., 2004, J. Clin. Microbiol. 42:1564-1569). The multiplex real-time polymerase chain reaction (RT-PCR) assay had the capability to detect 15 different viral pathogens; influenza virus type A and B, coronavirus 229E and OC43, human bocavirus, enterovirus, adenovirus, parechovirus, parainfluenza virus types 1- 4, human metapneumovirus, rhinovirus (RV) and RSV. The amount of virus was recorded semi-quantitatively based on the Ct value (cycle threshold value).

Immunophenotyping and intracellular cytokine staining

For direct phenotyping 500 ul of whole blood was stained with different extracellular monoclonal antibodies (MAbs) The following

combinations of markers and MAbs were used: CD4+ T cells (CD3+CD4+CD8- ), CD8+ T cells (CD3+CD4-CD8+), CD4+8+ T cells (CD3+CD4+CD8+), NK cells (CD3-CD56+), B cells (CD45+/SS-/CD19+), monocytes (CD14+CDla-), granulocytes (CD45+ / SS+ / CD 14- /CD la-).

Indirect phenotyping of PBMCs and intracellular cytokine staining was performed on thawed cryopreserved PBMCs. The following combinations of markers and fluorescent antibodies were used: CD14-FITC, CD16.56-PE, IL- 4-PE, CD3-PerCP, CD19-APC, IFN-y-APC, CD4-PE-Cy7, CD8-APC-Cy7.

Extracellular staining of surface markers CD14, CD16.56, CD3, CD4, CD8 was performed in 96-wells microtiter plates with the aforementioned fluorescent antibodies. For intracellular cytokine staining of IL-4 and IFN-γ, PBMCs were stimulated with β-Mercapto-Ethanol (6-ME), phorbol myristate acetate (PMA) (10 ng ml^-1), golgistop and ionomycin at 37°C for 4 hours. Then, staining of surface markers CD3, CD4, CD8 was performed. Thereafter, cells were fixed, permeabilized with 0.5% saponin and 0.5% BSA in PBS and stained for intracellular IL-4, IFN-7. Samples were acquired immediately after staining on a BD FACSCanto and analyzed using FlowJo analyses 7.6.

Cytokine concentrations in plasma and nasopharyngeal aspirates

Concentrations of the cytokines IL-16, IL-13, IL-10, IL-4, IL-12p70, TNF-a, IFN-Y, GM-CSF, IL-17, G-CSF, IL-6, IL-2, and chemokines IL-8

(CXCL-8), RANTES (CCL-5), MCP-1 and IP- 10 (CXCL-10) were measured by flow cytometry using the BD CBA Human Soluble Flex Set system (Becton Dickinson, Heidelberg, Germany) according to the manufacturer's instructions. In brief, cytokine-specific antibody-coated beads were incubated for 1 h with 25 pL plasma, supernatant of nasopharyngeal aspirates or standard solution. Thereafter, samples were incubated with the corresponding PE-labelled detection antibodies for 2 h. After one washing step, samples were measured by flow cytometry. Analysis of data and quantification of cytokines were performed using FCAP Array™ software (Becton Dickinson).

The detection limits were 2.3 pg/ml for IL-16, 0.6 pg/ml for IL-13, 0.13 pg/ml for IL-10, 1.3 pg/ml for MCP-1, 1.4 pg/ml for IL-4, 0.6 pg/ml for IL12p70, 1.2 pg/ml for TNF-a, 0.8 pg/ml for IFN-γ, 0.2 pg/ml for GM-CSF 0.3 pg/ml for IL-17a, 0.5 pg/ml for IP-10, 1.6 pg/ml for G-CSF, 1.6 pg/ml for IL-6, 11.2 pg/ml for IL-2, 0.002 pg/ml for RANTES and 1.2 pg/ml for IL-8.

Statistical methods

Values are expressed as percentages for categorical variables and as mean and standard error (SE) or median and interquartile range (IQR) for continuous variables. When variables were not normally distributed, Kruskal- Wallis tests was performed to compare continuous variables in the mild, moderate and severe group. When significant, Mann-Whitney U test was used for individual comparisons. Chi-squared tests were performed to compare categorical data. A two sided value of p< 0.05 was considered statistically significant.

A ROC curve was made for those markers that were statistically different between severe and mild RSV infection. An optimal cut-off value for individual markers was then determined with a sensitivity approaching 100% and specificity >85%. If the diagnostic marker was unable to meet the above mentioned criteria, the optimal cut-off value was adjusted so that both the sensitivity and specificity approached 75%. With these optimal cut-off values the diagnostic utilities (sensitivity, specificity, and positive and negative predictive values of these markers) were calculated and the best marker or combination of markers for diagnosing severity of infection in RSV infected children were selected. A combination of tests was considered positive in case any one of the selected markers exceeded their respective cut-off values. All statistical tests were performed by SPSS for Windows (Release 16; SPSS Inc., Chicago, IL). The level of significance was set at 5% in all comparisons.

Results

Severity of RSV infection is associated with a young age

Demographics and clinical features of the 52 included children are presented in Table 1. Children in the severe group were significantly younger than those in the mild and moderate groups (1.0 vs. 2.0 and 5.3 months;

p=0.003 and p=0.04, respectively). No significant differences were observed in the other clinical parameters.

Severe RSV infection is not associated with higher viral load nor with infection with more than one virus.

Multiplex RT-PCR confirmed the presence of RSV in all nasopharyngeal samples. In 21 of the 53 samples (40%) two or more viruses were detected, of which RV was most frequently found (17 samples). Infection with more than one was more often observed in children with mild RSV infection (16%) compared to those with moderate (46%) and severe disease (73%; p=0.007). No differences in the semi quantitatively measured RSV load were found between the different patient groups or between the children infected with RSV alone and those infected with two or more viruses

(supplementary Fig 1).

Severity of RSV infection is associated with lymphopenia Fig 1 shows the absolute cell numbers of various leukocyte subsets in blood. Lower CD4+ T-cells (p=0.000 and p=0.03) and NK cells (p=0.008 and p=0.012) were found in children with severe RSV infection compared to those with mild and moderate disease. CD8 T-cell numbers were significantly lower in children with severe RSV infection compared to those with mild disease (p=0.008). In the recovery phase, the number of NK cells, CD4 and CD8+ T cells in children with severe RSV infection were increased to normal values for age (supplementary Fig 2).

Severity of RSV infection is associated with high IL-8, IL-6 and G-CSF, and low RANTES plasma concentrations

The chemokine and cytokine concentrations in plasma and nasopharyngeal aspirates in the three patient groups are summarized in Table 2. Analyses of cytokines measured in plasma were restricted to IL-6, IL-8, IL- 10, IP10, MCP-1 and G-CSF, as the other analyzed cytokines were detected in only 8-14% of the acute samples. For nasopharyngeal samples, IL-16 and TNF- a were added to this set of cytokines. The other analyzed cytokines were present in only 5-31% of the acute nasopharyngeal samples.

Higher IL-8 and IL-6 plasma concentrations were found in children with severe disease compared to those with mild and moderate disease (p < O.OOland p=0.025, respectively) and G-CSF plasma concentrations were higher in children with severe disease compared to those with moderate disease

(p=0.021). In contrast, RANTES plasma concentrations were lower in children with severe disease compared to those with mild and moderate disease (p < 0.01). While increased RANTES concentrations were found during acute infection compared to recovery (p=0.034) in the moderate group, RANTES plasma levels were lower during acute infection compared to recovery in the severe group (p=0.009). In addition, RANTES plasma concentrations in the recovery phase were significantly higher in children with moderate RSV infection compared to those with severe symptoms (Fig 2) (p=0.025).

In nasopharyngeal samples, higher IL-6 concentrations were found in children with moderate and severe disease compared to those with mild disease (p < 0.05). In both plasma and nasopharyngeal samples almost all cytokine concentrations were higher during acute infection than in the recovery phase (Table 5 and 6).

No significant differences in the percentage or absolute numbers of IL-4 and IFN-γ producing CD4+ and CD8+ T-cells were observed in children with severe RSV infection compared to those with mild and moderate RSV infection (Table 7).

Cytokine patterns of severe RSV infection in infants younger than 3 months are similar to those in older children

Since age is a well known risk factor for severe RSV infection and children in the severe group were significantly younger than those in the mild and moderate group we also performed analyses on a subgroup of children younger 3 months. Because of the small numbers we categorized these infants in two groups based on mechanical ventilation. In this young age group we still observed significantly higher IL-8, IL-6 and G-CSF plasma concentrations, lower RANTES plasma concentrations and lower NK and CD4 T-Cell counts in ventilated infants compared to non-ventilated infants (Table 3).

Co-infections did not influence differences in cytokine patterns between children with mild and severe RSV infection

We also compared cytokine patterns of single RSV infections in 16 ventilated children to those in 15 non- ventilated children. There were no significant differences in age and other clinical parameters. Higher IL-8, IL-6 and G-CSF and, although not significant (p=0.058), lower RANTES plasma concentrations were observed in ventilated children compared to non- ventilated children. In addition, CD4 T-cell and NK-cell counts were lower in ventilated children compared to non- ventilated children.

The combination of IL-8, RANTES and CD4 T cell count discriminates severe RSV infection from mild RSV infection

Based on significant differences between the different severity groups in the aforementioned analyses a set of markers was selected for further analyses (Table 4). As described in statistical methods we calculated cut-off values of respectively 67 pg/ml, 13 ng/ml and 2.0 *10^E6 cells/ml for IL-8 and RANTES plasma levels and CD4 T-cell counts, respectively. Both IL-8 plasma levels and CD4 T cell counts showed high sensitivity (89% and 94%, respectively) and specificity (77% and 79%, respectively), followed by RANTES with a sensitivity of 79% and specificity of 74%. When 2 out of 3 cut-off values were exceeded in the patient a sensitivity of 93% and specificity of 96% was reached.

The above results thus show that the combination of IL-8 and RANTES plasma concentrations and CD4 T-cell counts represents a set of markers to predict severity of disease in children with RSV bronchiolitis.

Table 1. Demographics of children diagnosed with an RSV infection categorized by severity of disease

Mild Moderate Severe P-value

(N=ll) (N=22) (N=19)

Age in months, median [IQ] 5.3 [2.0-8.9] 2.0 [1.4-6.7] 1.0[0.7-3.9] p=0.006*

Age < 3 months 4 (36) 14 (64) 14 (74) NS

Male (%) 7 (64) 15 (68) 14 (74) NS

Birth weight (g), meaniSE 3312+186 3262±157 3167±191 NS

Prematurity <35 wks 1 (9) 2 (9) 6 (32) NS

Breastfeeding 8 (73) 12 (63) 8 (44) NS

Smoking during pregnancy 1 (9) 3 (15) 5 (28) NS

Congenital heart disease 0 1 (5) 1 (5.3) NS

Atopic disease 2 (18) 2 (9) 2 (11) NS

Siblings 7 (64) 13 (59) 17 (90) NS

Day care 3 (38) 3(14) 1 (5) NS

Passive smoking 2 (20) 2 (10) 3 (18) NS

Family history of atopy 7 (64) 13 (65) 12 (67) NS onset of symptoms in days, 6 [4-8] 4[3-5.25] 5[3-6] NS median [IQ]

Ct value RSV, meaniSE 29.7±1.6 28.5±0.9 28.7±0.9 NS

Co-infection 8 (73) 10 (46) 3 (16) p=0.007** Data are presented as number (%), unless otherwise specified. *mild vs. mod, p=0.04; mild vs. severe, p=0.003 **mild vs. severe, p=. IQ= interquartiles 25^th and 75^th percentile, g=grams, SE= standard error, Ct= cycle time.

Table 2a. Cytokine and chemokine plasma concentrations (pg/ml) in the different severity groups of children with RSV infections.

Mild (N=ll) Moderate (N=20) Severe (N=19) p-value

IL-8 48.1 [36.0-76.8] 55.7 [40.5-106.6] 127.8 [79.0-321.9] p=0.000*

RANTES 25507 [19971-35204] 25235 [11264-45161] 8632 [3895-12155] p=0.007**

IL-6 20.4 [11.0-35.6] 10.1 [0.0-69.8]* 72.9 [21.9-122.4] * p=0.025†

IL-10 21.7 [0.0-37.0] 6.0 [0.0-38.9] 0.0 [0.0-23.6]

MCP-1 220.0 [168.1-296.4] 138.4 [106.7-229.9] 174.6 [103.9-481.0]

GCSF 51.9 [21.2-180.8] 47.0 [12.9-111.1]* 129.9 [85.6-362.7]* p=0.021$

IP- 10 575.6 [418.9-902.8] 395.7 [331.9-666.5] 660.4 [347.3-1184.9]

Data are presented as as median [25 -75 percentile] . Krukall Wallis test were perfomed, followed by Mann Whitney U test for one to one comparisons, P<0.05 was considered to be statistically significant. *Mild vs severe, moderate vs. severe, p<0.01, **mild vs severe, moderate vs. severe, p<0.01,† moderate vs. severe, p=0.02 i moderate vs severe, p=0.01

Table 2b. Cytokine and chemokine nasopharyngeal concentrations (pg/ml) in the different severity groups of children with RSV infections.

mild (N=9) Moderate (N=13-15) Severe (N=14-15) p- value

IL-8 9106 [7599-18947] 11908 [10672-20307] 17773 [11917-67609]

RANTES 116.8 [52.6-179.9] 74.0 [57.5-208.5] 48.0 [0.0-199.8]

IL-1B 552.4 [206.5-944.4] 1670.4 [763.5-3150.9] 1093.7 [545.3-5932.5]

IL-6 1475.20 [1136.6-1908.4] 2593.3 [1641.6-4492.8] 4801.2 [1520.3-8186.6] p=0.031*

IL-10 28.9 [0.0-49.6] 101.7 [0.0-207.8] 22.7 [0.0-139.4]

MCP-1 140.7 [60.4-251.2] 394.1 [202.6-394.1] 539.9 [89.2-1865.7]

TNF-a 216.1 [168.4-356.0] 649.3 [266.2-1990.1] 431.2 [251.1-2290.9]

G-CSF 4940.4 [3416.0-8715.8] 5001.6 [3220.8-8403.7] 3242.7 [1437.1-5780.9]

IP-10 11788 [7168-24024] 14633.9 [5289.1-24037.4] 5016.0 [775.9-11538.9

Data are presented as as median [25 -75 percentile]. Krukall Wallis test were performed, followed by Mann Whitney U test for one to one comparisons, p<0.05 was considered to be statistically significant. *Mild vs severe, moderate vs. severe, p<0.05

Table 3. Significant comparisons between moderate and severe RSV infection in children younger than 3 months.

Mild/Moderate Severe p- value (Non-ventilated) (ventilated)

age (days) 53.4 [45.0-59.7] 31.0 [19.0-48.7] 0.01 plasma 11-8 57.3 [40.2-43.9] 127.8 [87.4-137.8] < 0.01 plasma IL-6 12.9 [3.2-43.8] 66.4 [3.2-98.0] 0.029 plasma G-CSF 34.3 [6.1-144.6] 114.5 [3.2-157.3] 0.033 plasma RANTES 12529.0 [5863.3-15769.5] 8874.2 [8193.6-38503.2] 0.01 nasopharyngeal IL-6 335.1 [229.7-368.8] 291.7 [152.0-522.1] 0.031

NK cells 0.59 [0.06-1.16] 0.20 [0.15-0.24] 0.011

CD4+ Tcells 2.48 [1.40-5.56] 1.52 [1.19-1.90] 0.02

CD4/CD8 ratio 5.28 [3.31-7.91] 2.60 [2.60-3.78] 0.013

Numbers are presented as median concentrations [25 =75 percentile]. Cytokine concentrations are given in pg/ml and cell numbers in 10^E6 cells/ml. Mann- Whitney U tests were performed, p<0.05 was considered to be statistically significant.

Table 4. Comparison of sensitivity, specificity, positive predictive values and negative predictive values, and area under the curve for different markers in patients with severe vs. non-severe RSV infections.

Marker Cut off value sensitivity specificity PPV NPV AUC

RANTES < 13 ng/ml 0.79 0.74 65% 85% 0.782

IL-8 > 67.2 pg/ml 0.89 0.77 68% 92% 0.844

IL-6 > 28.7 pg/ml 0.75 0.67 60% 80% 0.718

G-CSF > 82.7 pg/ml 0.80 0.67 62% 83% 0.723 age < 2 mnd 0.65 0.57 48% 73% 0.718

CD4 cells < 2.03*10E6/ml 0.94 0.79 73% 96% 0.857

NK cells < 0.23*10E6/ml 0.71 0.68 57% 79% 0.752

PPV= positive predictive value, NPV= negative predictive value, AUC= area under the curve. Cut off values are based on optimal sensitivity and specificity for each marker

Table 5. Paired analyses of cytokine and chemokine plasma concentrations (pg/ml) during acute RSV infection and after recovery in moderate and severe disease group

moderate acute moderate recovery N p-value

IL-8 60.7 [49.7-126.2] 9.9 [6.3-14.0] 12 0.002

CCL-5 20312 [11846-45161] 12872 [9583-32677] 12 0.034

IL-6 22.6 [2.7-76.9] 0.0 [0.0-0.0] 11 0.021

IL-10 30.7 [0.0-54.7] 0.0 [0.0-0.0] 12 0.012

MCP-1 159.2 [106.7-296.1] 146.4 [87.0-198.8] 12 0.05

GCSF 47.0 [13.1-136.0] 5.8 [0.0-14.5] 11 0.05

IP10 413.5 [365.2-794.0] 191.5 [128.8-246.2] 11 0.003

severe acute severe recovery N p-value

IL-8 130.7 [90.7-326.5] 12.2 [1.4-25.0] 16 0

CCL-5 8874 [3314-20488] 34351 [21253-61944] 15 0.009

IL-6 79.4 [9.7-146.1] 0.0 [0.0-0.0] 15 0.003

IL-10 7.4 [0.0-23.5] 0.0 [0.0-0.0] 16 0.012

MCP-1 176.1 [107.1-482.4] 232.4 [148.6-358.4] 16 NS

GCSF 129.9 [85.6-371.7] 3.9 [0.0-10.5] 15 0.002

IP10 660.4 [437.3-1286.7] 187.6 [133.8-292.3] 15 0.002

Table 6. Paired analyses of cytokine and chemokine nasopharyngeal concentrations (pg/ml) during acute RSV infection and after recovery in moderate and severe disease group

moderate acute moderate recovery N p-value

IL-8 14971 [11314-24860] 8758.2 [1449.8-15201.0] 7 0.028

CCL-5 74.0 [40.7-176.5] 23.3 [0.0-60.1] 5 0.043

IL-1B 1743.0 [763.5-3236.5] 68.4 [24.0-868.6] 5 0.043

IL-6 2001.6 [1616.3-3518.4] 215.9 [64.4-559.7] 7 0.018

IL-10 103.8 [21.1-412.8] 0.0 [0.0-51.2] 5 i 0.068

MCP-1 302.8 [159.0-373.9] 32.3 [7.0-65.7] 4 i 0.068

TNF-a 823.5 [200.4-1944.9] 2.4 [0.0-70.7] 4 ^j 0.068

G-CSF 6090.0 [3623.1-9296.7] 3590.1 [449.1-4922.1] 7 ΐ 0.063

IP- 10 18229 [8009-28110] 3321.0 [1897.8-9194.0] 7 0.018

severe acute severe recovery N p-value

IL-8 16072 [10196-41044] 7528.3 [2973.5-14524.0] 10 0.047

CCL-5 48.0 [3.0-435] 20.8 [0.0-107.5] 9 NS

IL-1B 1033.0 [604.5-9572.0] 204.4 [144.5-880.0] 8 0.05

IL-6 4069.3 [1159.2-9814.7] 363.5 [78.4-657.3] 10 0.007

IL-10 7.7 [0.0-355] 0.0 [0.0-80.2] 5 NS

MCP-1 781.0 [68.8-3412.2] 61.3 [0.0-162.3] 7 0.018

TNF-a 431.2 [288.5-5475.3] 144.0 [0.0-378.6] 8 0.069

G-CSF 3246.1 [1431.9-7609.3] 3038.1 [1327.2-5827.4] 10 NS

IP- 10 3930.5 [424.5-18245.6] 1326.4 [489.1-2806.1] 10 NS

Wilcoxon signs rank test Table 7 IL-4 and IFN-γ producing CD4 and CD8 positive T-lymphocytes from children with severe and non-severe RSV infection measured by intracellular cytokine staining

CD4 CD8

IFN-Y ( ) IL-4 (% ) IFN-Y/IL-4 IFN-Y ( ) IL-4 (%) IFN-Y/IL-4

Non severe 0.17 0.29 0.43 2.17 0.48 3.15

(N=9) [0.07-0.26] [0.26-0.47] [0.30-0.69] [1.15-2.94] [0.29-0.56] [2.74-6.75]

Severe 0.11 0.33 0.39 2.14 0.59 2.73

(N=10) [0.06-0.18] [0.22-0.51] [0.14-0.56] [1.04-3.24] [0.52-0.63] [1.85-5.46]

P-value* NS NS NS NS NS NS

Data are medians [interquartile range 25-75%]

*Mann Whitney U-test, NS= not significant (p>0.05)

Claims

1. Method for the prediction of the severity of a disease developing from an infection with human respiratory syncytial virus (RSV) in a subject, comprising the steps of:

a. determining the plasma levels of interleukin 8 (IL-8) and

RANTES (CCL5);

b. determining the blood cell count of CD4+ T cells;

c. making a prediction on basis of the values obtained in steps a) and b) . 2. Method according to claim 1, where the prediction is the

development of a severe disease state of the RSV infection, wherein the plasma IL-8 level is more than 60 pg/mg, preferably more than 65 pg/ml, more preferably more than 67.

2 pg/ml, the RANTES plasma level is less than 15 ng/ml, preferably less than 14 ng/ml, more preferably less than 13 ng/ml and the CD4+ cell count is less than 3*10⁶ cells per ml blood, preferably less than 2.5*10⁶ cells per ml blood, more preferably less than 2.03*10⁶ cells per ml blood.

3. Method according to claim 1, wherein the prediction is an increased risk of the development of a severe disease state of the RSV infection, if the plasma IL-8 level is more than 60 pg/mg, preferably more than 65 pg/ml, more preferably more than 67.2 pg/ml, the RANTES plasma level is less than 15 ng/ml, preferably less than 14 ng/ml, more preferably less than 13 ng/ml or the CD4+ cell count is less than 3*10⁶ cells per ml blood, preferably less than 2.5*10⁶ cells per ml blood, more preferably less than 2.03*10⁶ cells per ml blood.

4. Method according to any of claims 1 - 3, wherein the IL-8 and RANTES plasma levels and the CD4+ cell count are assayed by

immunological assays.

5. Method according to any of claims 1 - 4, wherein the subject is a child, preferably a child having an age of less than 6 months, more preferably a child having an age of less than three months.

6. Detection kit comprising means for the detection of IL-8 and

RANTES in plasma and for determining the CD4+ cell count in blood.

7. Detection kit according to claim 6, additionally comprising means for detection of RSV particles.

8. Use of a kit according to claim 6 or 7 for predicting the severity disease state upon RSV infection in a subject.