Method for diagnosing invasive Candida infections

Abstract

The present invention concerns an in vitro method for diagnosing invasive candidiasis (IC) in a subject which comprises detecting the presence of a Candida glycan and detecting the presence of antibody directed against a protein selected from the group consisting of fructose bisphosphate aldolase (Fba1), enolase 1 (Eno1), heat shock protein 90 (Hsp90), hyphal wall protein (Hwp1), and mannoprotein 65 (Mp65) in a blood, plasma or serum sample of the subject. The invention also relates to a method of determining a suitable treatment regimen for a patient and to a kit for implanting the methods described herein.

Claims

1. An in vitro method for diagnosing invasive candidiasis (IC) in a subject, said method comprising the steps: a) detecting a level of a Candida mannan in a blood, plasma or serum sample of the subject using a method selected from the group consisting of an immunoassay, a chromatography, an enzyme-based chemoluminescent method, surface plasmon resonance, mass-spectrometry, and a lectin-based detection; b) detecting by an immunoassay a level of antibodies directed against a Candida protein selected from the group consisting of fructose bisphosphate aldolase (Fba1), enolase 1 (Eno1), heat shock protein 90 (Hsp90), hyphal wall protein (Hwp1), and mannoprotein 65 (Mp65) in the same sample of the subject or in another sample sequentially obtained from the same subject, essentially simultaneously or no more than 3 hours apart; c) performing a combined analysis of the level of said Candida mannan detected in step a) and the level of antibodies directed against said Candida protein detected in step b) by calculating a level of a marker Z =Σ ai ×[Marker i], wherein [Marker i] are individual levels of said Candida mannan and of the antibodies directed against said Candida protein, and ai are coefficients which values are determined in order to maximize an Area Under the Curve (AUC) of a Relative Operating Characteristic (ROC) curve for the combination of [Marker i]; and d) deducing that the subject is developing or has developed invasive candidiasis if the level of the marker Z calculated in step c) is higher than a reference level of marker Z.

2. The method according to claim 1, wherein at step b) the level of antibodies against Fba1, or antibodies against Eno1, or antibodies against Hsp90, or antibodies against Hwp1, or antibodies against Mp65, only, is detected.

3. The method according to claim 1, wherein said method does not comprise detecting an antibody directed against mannan.

4. The method according to claim 1, wherein said Candida mannan is detected by a sandwich enzyme immunoassay which uses, as capture and detection antibodies, an antibody recognizing sequences of α-linked oligomannoses comprising more than four residues.

5. The method according to claim 1, wherein said reference level of the marker Z is a single value or a range of values determined based on the level of said Candida mannan or the level of antibodies directed against said Candida protein, as appropriate, measured in a population of healthy subjects, in a population of subjects superficially infected with a Candida strain, in a population of subjects suffering from invasive candidiasis, or in a sample from the same subject obtained at an earlier time point.

6. The method according to claim 1, wherein invasive candidiasis is due to an infection with a Candida species selected from the group consisting of Candida albicans, Candida parapsilosis, Candida kruseï, Candida tropicalis, Candida glabrata, Candida lusitaniae, Geotrichum capitatum, and Candida norvegiensis.

7. The method according to claim 1, wherein in step b) the antibodies are directed against a Candida Fba1.

8. The method according to claim 7, wherein the level of the antibodies directed against the Candida Fba1 is detected using a polypeptide comprising: a) SEQ ID NO:2; or b) a variant polypeptide of the polypeptide defined in a), wherein said variant polypeptide has at least 80% sequence identity with the polypeptide defined in a), over the whole length of the polypeptide.

9. The method according to claim 1, wherein in step b) the antibodies are directed against a Candida Eno1.

10. The method according to claim 9, wherein the level of the antibodies directed against the Candida Eno1 is detected using a polypeptide comprising: a) SEQ ID NO:4; or b) a variant polypeptide of the polypeptide defined in a), wherein said variant polypeptide has at least 80% sequence identity with the polypeptide defined in a), over the whole length of the polypeptide.

11. The method according to claim 1, wherein in step b) the antibodies are directed against a Candida Hsp90.

12. The method according to claim 11, wherein the level of the antibodies directed against the Candida Hsp90 is detected using a polypeptide comprising: a) SEQ ID NO:6; b) amino acids at positions 313 to 707 of SEQ ID NO:6; c) a sequence SEQ ID NO:13, SEQ ID NO:14, SEQ ID NO:15, or SEQ ID NO:16; or d) a variant polypeptide of the polypeptide defined in a), b), or c), wherein said variant polypeptide has at least 80% sequence identity with the polypeptide defined in a), b), or c) over the whole length of the polypeptide.

13. The method according to claim 1, wherein in step b) the antibodies are directed against a Candida Hwp1.

14. The method according to claim 13, wherein the level of the antibodies directed against the Candida Hwp1 is detected using a polypeptide comprising: a) SEQ ID NO:8; b) amino acids at positions 41 to 200, or at positions 27 to 203 of SEQ ID NO:8; or c) a variant polypeptide of the polypeptide defined in a) or b), wherein said variant polypeptide has at least 80% sequence identity with the polypeptide defined in a) or b), over the whole length of the polypeptide.

15. The method according to claim 1, wherein in step b) the antibodies are directed against a Candida Mp65.

16. The method according to claim 15, wherein the level of the antibodies directed against the Candida Mp65 is detected using a polypeptide comprising: a) SEQ ID NO:10 or SEQ ID NO:12; or b) a variant polypeptide of the polypeptide defined in a), wherein said variant polypeptide has at least 80% sequence identity with the polypeptide defined in a), over the whole length of the polypeptide.

Description

FIGURES

(1) FIGS. 1-5. Boxplot representation of diagnosis potential of anti-recombinant protein-antibody [RP-Ab; respectively anti-Fba1 Ab (FIG. 1); anti-Hwp1 Ab (FIG. 2); anti-Hsp90 Ab (FIG. 3); anti-Eno1 Ab (FIG. 4); anti-Mp65 Ab (FIG. 5)] or anti-mannan antibodies associated with mannanemia. Boxplot representation is a convenient way of graphically depicting groups of numerical data through their five-number summaries (the smallest observation, lower quartile (Q1), median (Q2), upper quartile (Q3), and largest observation). Boxplots can be useful to display differences between populations without making any assumptions of the underlying statistical distribution. CTRL represents controls and IC represents invasive candidiasis patients.

(2) FIGS. 6-10. ROC curves of combinations of RP-Ab [respectively anti-Fba1 Ab (FIG. 6); anti-Hwp1 Ab (FIG. 7); anti-Hsp90 Ab (FIG. 8); anti-Eno1 Ab (FIG. 9); anti-Mp65 Ab (FIG. 10)] with mannanemia. The actual diagnosis (mannan Antibody and mannan Antigen) test was represented with dotted lines and association of RP-Ab and mannanemia was represented with black line.

EXAMPLES

Example 1

Recombinant Protein Production in E. coli

(3) 1) Strains and Plasmids

(4) Candida albicans SC5314 was used as fungal DNA source to generate the different recombinant proteins. Escherichia coli strain DE3 was transformed to produce the recombinant proteins while the strain DH5α was used to amplify the plasmids.

(5) 2) Cloning

(6) Six recombinant proteins were expressed in E. coli as described previously (Fradin et al., Infect Immun (2008), 76: 4509-4517): N-terminal fragment of Hwp1 (amino acids 27 to 203), Eno1 (full length), Mp65 without its peptide signal (amino acids 1 to 22), Fba1 (full length), Sod5 without its peptide signal (amino acids 1 to 22) and its C-terminal GPI consensus sequence (last 24 terminal amino acids) and Hsp90 (full length).

(7) Six sets of primers (Table 2 below) were designed to clone the different genes or truncated genes. PCR amplified fragments with high fidelity Expand Taq polymerase (Roche) were directly cloned in pEXP5-NT/TOPO plasmid.

(8) TABLE-US-00002 TABLE 2 Set of primers used for gene cloning Recombinant proteins Primers set N-term Hwp1 5′CAAGGTGAAACAGAGGAAGCT3′ (SEQ ID NO: 17) and 5′TCAAGCAGGAATGTTTGGAGTAGT3′ (SEQ ID NO: 18) Eno1 5′ATGTCTTACGCCACTAAAATCCACGC3′ (SEQ ID NO: 19) and 5′TTACAATTGAGAAGCCTTTTGGAAATCTTTAC3′ (SEQ ID NO: 20) Mp65 5′GCTCATCAACATCATCAACAT3′ (SEQ ID NO: 21) and 5′TTAGTTAGAGTAAATACCCCAGTA3′ (SEQ ID NO :22) Fba1 5′ATGGCTCCTCCAGCAGTTTTA3′ (SEQ ID NO: 23) and 5′TTACAATTGTCCTTTGGTGTG3′ (SEQ ID NO: 24) Sod5 5′GATGCACCAATCTCAACTGAC3′ (SEQ ID NO: 25) and 3′TTAACCTTGAGGAGCAGTAGAAGC3′ (SEQ ID NO: 26) Hsp90 5′ATGGCTGACGCAAAAGTTGAA3′ (SEQ ID NO: 27) and 5′TTAATCAACTTCTTCCATAGC3′ (SEQ ID NO: 28)

Example 2

Evaluation of Diagnosis Potential of Anti-Recombinant Protein-Antibodies or Anti-Mannan Antibodies Associated with Mannanemia

(9) 1) Materials and Methods

(10) Patients

(11) Between January 2005 and December 2007, 157 serum samples were retrospectively collected in different clinical departments of Lille University Hospital (LUH), from 53 patients (24 females and 29 males [mean age, 56.78+/−23.71 years]) with proven Candida albicans candidiasis. The average number of samples per patient in this group was 2.68+/−2.13 (Table 3).

(12) TABLE-US-00003 TABLE 3 Clinical features of patients with systemic Candida albicans infection. Patients were classified according to clinical wards No. Of Date of serum sampling in Patient Sex.sup.a Age (yr) Hospital ward sera relation to blood culture (days) Candida species 1 F 34 Burn unit 7 −10, −3, 0, 4, 13, 18, 22 Candida albicans 2 M 43 Burn unit 4 −16, −9, 12, 27 Candida albicans 3 M 54 Burn unit 3 −6, −4, 10 Candida albicans 4 F 54 Burn unit 4 −10, −3, 4, 11 Candida albicans 5 F 59 Burn unit 1 17 Candida albicans 6 M 68 Burn unit 4 −10, −3, 4, 18 Candida albicans 7 F 75 Burn unit 2 7, 14 Candida albicans 8 M 81 Cardiology 1 6 Candida albicans 9 F 31 Gastroenterology 1 0 Candida albicans 10 M 53 Gastroenterology 1 2 Candida albicans 11 M 59 Gastroenterology 3 −6, −2, 6 Candida albicans 12 M 78 Heat surgery 1 1 Candida albicans 13 F 46 Clinical hematology 8 −8, −5, 1, 5, 5, 16, 22, 27 Candida albicans 14 F 62 Clinical hematology 7 −3, 4, 12, 14, 20, 27, 29 Candida albicans 15 M 70 Clinical hematology 6 3, 10, 13, 17, 19, 24 Candida albicans 16 F 86 Infectious diseases 3 1, 7, 14 Candida albicans 17 M 10 Intensive care unit 2 3, 23 Candida albicans 18 F 14 Intensive care unit 1 4 Candida albicans 19 M 32 Intensive care unit 2 14, 30 Candida albicans 20 M 35 Intensive care unit 1 −1 Candida albicans 21 F 36 Intensive care unit 2 3, 10 Candida albicans 22 M 46 Intensive care unit 2 −2, 5 Candida albicans 23 M 47 Intensive care unit 6 −3, 4, 13, 20, 27, 28 Candida albicans 24 M 49 Intensive care unit 2 −6, −2 Candida albicans 25 F 50 Intensive care unit 1 0 Candida albicans 26 M 57 Intensive care unit 6 −8, −2, 5, 12, 20, 27 Candida albicans 27 F 59 Intensive care unit 2 3, 4 Candida albicans 28 M 60 Intensive care unit 5 −10, −3, 4, 11, 25 Candida albicans 29 F 62 Intensive care unit 4 −12, −5, 2, 9 Candida albicans 30 M 64 Intensive care unit 2 2, 9 Candida albicans 31 M 66 Intensive care unit 5 2, 6, 8, 12, 13 Candida albicans 32 M 72 Intensive care unit 4 −5, 2, 6, 9 Candida albicans 33 F 75 Intensive care unit 1 3 Candida albicans 34 F 76 Intensive care unit 9 4, 7, 11, 14, 17, 22, 25, 27, 28 Candida albicans 35 F 79 Intensive care unit 9 −1, 3, 4, 5, 6, 13, 14, 20, 27 Candida albicans 36 F 82 Intensive care unit 1 5 Candida albicans 37 F 87 Intensive care unit 4 2, 7, 22, 29 Candida albicans 38 M 89 Intensive care unit 2 −7, −1 Candida albicans 39 M 48 intensive surgical care unit 1 2 Candida albicans 40 M 72 intensive surgical care unit 1 0 Candida albicans 41 M 72 intensive surgical care unit 3 −3, 4, 16 Candida albicans 42 F 73 intensive surgical care unit 2 −5, 25 Candida albicans 43 M 73 intensive surgical care unit 1 11 Candida albicans 44 M 78 intensive surgical care unit 1 2 Candida albicans 45 F 83 intensive surgical care unit 2 −9, 19 Candida albicans 46 M 15 Oncology 2 −5, 1 Candida albicans 47 M 6 Paediatrics 6 15, 17, 18, 19, 24, 26 Candida albicans 48 F 14 Paediatrics 1 −1 Candida albicans 49 F 17 Paediatrics 3 −15, −8, 6 Candida albicans 50 M 53 Pneumology 2 −11, −10 Candida albicans 51 F 80 Pneumology 1 4 Candida albicans 52 M 64 Transplantation 1 −1 Candida albicans 53 F 49 Traumatology 1 0 Candida albicans .sup.aM, male; F, female

(13) A second group of 142 serum samples was also collected in different departments of LUH from 40 patients (10 females and 30 males [mean age, 58.00+/−21.97]) with proven invasive candidiasis determined by non-albicans yeast species (Table 4). The average number of samples per patient in this group was 3.59+/−2.66. This group contains 7 different yeast species: Candida parapsilosis (17 patients; 49 sera), Candida kruseï (3 patients; 10 sera), Candida tropicalis (5 patients; 20 sera), Candida glabrata (12 patients; 40 sera), Geotrichum capitatum (1 patient; 12 sera), Candida norvegiensis (1 patient; 3 sera) and Candida lusitaniae (1 patient; 8 sera).

(14) TABLE-US-00004 TABLE 4 Clinical features of patients with systemic Candida infection. Patients were classified according to yeast species involved in IC and to clinical wards No. of Date of serum sampling in Patient Sex .sup.a Age (yr) Hospital ward sera relation to blood culture (days) Candida species 54 M 21 Burn unit 2 −1, 12 Candida parapsilosis 55 M 43 Burn unit 4 −6, 0, 7, 21 Candida parapsilosis 56 F 82 Clinical hematology 2 2, 7 Candida parapsilosis 57 M 24 Intensive care unit 4 −11, −6, 0, 1 Candida parapsilosis 58 M 51 Intensive care unit 1 25 Candida parapsilosis 59 M 54 Intensive care unit 5 −10, −3, 4, 10, 18 Candida parapsilosis 60 M 65 Intensive care unit 1 −11 Candida parapsilosis 61 M 67 Intensive care unit 4 −11, −4, 3, 17 Candida parapsilosis 62 F 75 Intensive care unit 1 23 Candida parapsilosis 63 M 76 Intensive care unit 5 −15, −12, 2, 6, 9 Candida parapsilosis 64 M 78 Intensive care unit 3 −8, −3, 4 Candida parapsilosis 65 M 87 Intensive care unit 7 −14, −7, 0, 7, 14, 21, 22 Candida parapsilosis 66 F 87 Intensive care unit 2 −5, 8 Candida parapsilosis 67 M 65 Intensive surgical care unit 2 0, 1 Candida parapsilosis 68 M 57 Intensive surgical care unit 4 −5, 1, 9, 23 Candida parapsilosis 69 M 90 Intensive surgical care unit 1 3 Candida parapsilosis 70 F 10 Paediatrics 1 3 Candida parapsilosis 71 M 47 Clinical hematology 3 −11, −4, 3 Candida kruseï 72 M 68 Clinical hematology 6 −9, −7, −1, 4, 12, 19 Candida kruseï 73 M 75 Oncology 1 4 Candida kruseï 74 M 39 Burn unit 5 −15, −8, 10, 13, 28 Candida tropicalis 75 M 51 Clinical hematology 1 3 Candida tropicalis 76 F 62 Clinical hematology 10 −15, −6, −2, −1, 0, 1, 3, 5, 7, 12 Candida tropicalis 77 M 8 Intensive care unit 3 1, 8, 18 Candida tropicalis 78 M 67 Oncology 1 −5 Candida tropicalis 79 M 24 Hyperbare 2 −2, 5 Candida glabrata 80 M 31 Intensive care unit 1 3 Candida glabrata 81 M 57 Intensive care unit 5 0, 6, 13, 20, 27 Candida glabrata 82 M 63 Intensive care unit 4 −10, 2, 10, 11 Candida glabrata 83 M 63 Intensive care unit 8 −5, −4, −1, 0, 1, 3, 7, 12 Candida glabrata 84 F 69 Intensive care unit 1 14 Candida glabrata 85 F 76 Intensive care unit 5 −9, −5, 2, 9, 16 Candida glabrata 86 F 87 Intensive care unit 3 −7, 7, 12 Candida glabrata 87 F 24 Intensive surgical care unit 5 −11, −7, −5, 0, 10 Candida glabrata 88 M 71 Intense surgical care unit 2 1, 15 Candida glabrata 89 M 85 intensive surgical care unit 3 2, 4, 11 Candida glabrata 90 F 60 Oncology 1 8 Candida glabrata 91 M 64 Clinical hematology 12 −15, −11, −9, −6, −3, −1, 1, 4, 6, 11, 13, 15 Geotrichum capitatum 92 M 45 Clinical hematology 3 1, 7, 12 Candida norvegensis 93 M 76 Clinical hematology 8 −10, −3, −1, 1, 4, 6, 8, 11 Candida lusitaniae .sup.a M, male; F, female

(15) The following criteria were applied as retrospective selection rules when the laboratory and clinical files were examined: (i) positive blood culture from Candida species; (ii) availability of serum samples obtained within a range of 3 weeks before and 1 month after positive cultures, (iii) the presence of risk factors (cancer and chemotherapy, abdominal surgery, AIDS, major health problems requiring hospitalization in intensive care units -ICUs-, and use of broad-spectrum antibiotics, indwelling intravascular catheters, and hyperalimentation; and (iv) the presence of an infectious syndrome (namely, fever) that did not respond to antibacterial therapy but that did respond to antifungal therapy. In order to evaluate performances of biomarkers for the early diagnosis of IC, selection of serum samples was restricted to the period ranging from 2 weeks before to 1 month after the date of isolation of yeast species from blood culture. After blood sampling blood samples were centrifuged and serum aliquots were stored at −80° C. until required.

(16) Control Group.

(17) Two groups of control sera were included in this study: (i) Group 1 comprised 90 serum specimens from 90 hospitalized patients (32 females and 58 males; mean age, 64+/−15.5 years) without evidence of invasive candidiasis. This group of patients was enrolled in a prospective study conducted in an ICU of LUH for 6 months. The study was designed for the assessment of risk factors for nosocomial candidiasis. These patients were under clinical and mycological survey for periods ranging from 1 to 74 days (mean, 12 days). Samples of blood, oral swabs, urine, and stools were collected biweekly. Among them 90 patients, 71 were colonized by yeast species with evaluation of number of colonized body sites: 1 site (16.9%), 2 sites (26.8%), 3 sites (26.8%), 4 sites (25.3%) and 5 sites (4.2%). (ii) Group 2 consisted of 80 serum samples from healthy blood donors.

(18) EIA Detection of Anti-C. albicans Mannan Antibodies in Human Sera.

(19) Antibodies to Candida albicans mannan were detected using the Platelia Candida antibody (Ab) Plus test (Bio-Rad Laboratories, Marnes La-Coquette, France) according to manufacturer's instructions. Briefly, Enzyme ImmunoAssay (EIA) was performed with BEP III automate (Behring Laboratories, Paris, France). For individual sera, 100 μl of serum diluted 1/400 was applied to each well, and the plate was incubated for 1 h at 37° C. After washing, 100 μl of horseradish peroxidase-conjugated anti-human immunoglobulins was then added, and the plates were incubated for 1 h at 37° C. After intensive washing, the reaction was revealed by 30 min of incubation in darkness with 200 μl of tetramethylbenzidine solution. The absorbance at a λ of 450/620 nm was measured. The results were reported in arbitrary units (AU) in relation to the results on the standard curve.

(20) Detection of Mannanemia.

(21) Circulating mannan was detected using the Platelia Candida antigen Plus (Ag) test (Bio-Rad Laboratories, Marnes-La-Coquette, France) as described previously (Sendid et al., J Med Microbiol. 2002 May; 51(5):433-42). Briefly, microtiter plates were sensitized in an industrial setting with monoclonal antibody (monoclonal antibody EBCA1 of Platelia Candida antigen Plus (Ag) test). 300 μl of patient sera was denatured with 100 μl of EDTA treatment solution, and the mixture was boiled for 3 min and centrifuged at 10,000 g for 10 min. 50 μl of supernatant, obtained from patient serum and treated as described above, was mixed in a plate well with 50 μl of horseradish peroxidase-conjugated EBCA1. After incubation for 90 min at 37° C., the plates were washed intensively and the reaction was revealed by 30 min of incubation in darkness with 200 μl of tetramethylbenzidine solution. The optical density was read at a λ of 450/620 nm on a PR2100 reader (Sanofi Pasteur Diagnostics). Reactions were performed in duplicate. Each experiment included a calibration curve for a pool of normal human sera supplemented with concentrations of mannan of 0.1 to 27 ng/ml.

(22) EIA Detection of Anti-C. albicans Recombinant Proteins Antibodies in Human Sera

(23) Recombinant proteins were coated on ELISA plates (NUNC IMMUNO-MODULE 468680) at a concentration of 2 μg/ml for Mp65 and Eno1, and at a concentration of 4 μg/ml for Fba1, Hwp1 and Hsp90 produced in Escherichia coli, with carbonate solution pH 9.5+/−0.2 filtered 0.22 μm for all antigens. These preparations were incubated at room temperature overnight.

(24) After coating, the wells were blocked by 100 μl of Bovine Serum Albumin/Saccharose solution pH 7.2+/−0.2 filtered 0.22 μm, emptied and filled with 200 μl of the same solution. Plates were then frozen at −20° C. Protocols for antigens coating, preparation of diluent solutions, dilution of patient sera, and conjugate solutions were based on preliminary experiments performed with a pool of sera from IC patients known to display high titers of anti-Candida antibodies.

(25) Patient sera were diluted 1/100 in Tris saline buffer and BSA pH 7.6 and were incubated for 1 h at 37° C. on coated plate in dry incubator, 3 times washed with 800 μl of Tris saline buffer and Tween 20 (0.1%) and proclin (0.07%), incubated 1 h at 37° C. with a secondary anti-total immunoglobulin antibody conjugated to peroxidase (Bio-Rad, Marnes La Coquette, France), washed 3 times as previously described and incubated 30 minutes with 200 μl of tetramethyl benzidine (TMB) and stored at room temperature in dark. The reaction was stopped by the addition of 100 μl of stopping solution containing 2M H.sub.2SO.sub.4 and ODs were measured at 450/620 nm. In parallel, all sera were tested at the same time with the home-made EIA tests involving recombinant proteins, the Platelia™ Candida Ab Plus and the Platelia™ Candida Ag.

(26) Statistical Analysis

(27) Statistical analysis was performed in collaboration with SysDIAG: Systèmes Complexes pour le Diagnostic (UMR3145 CNRS/Bio-Rad, Montpellier, France). All statistics and figures were computed with the “R/Bioconductor” statistical open source software (Ge et al. Test 2003; 12:1-77; Gentleman et al. Genome Biol 2004; 5:R80) or SAS software v9.2 (SAS institute Inc). A differential analysis was carried out with the non-parametric Wilcoxon rank sum test and the Welch test. With the multiple testing methodologies, it is important to adjust the p-value of each marker to control the False Discovery Rate (FDR). The Benjamini and Hochberg (BH) procedure (Benjamini et al. Behav Brain Res 2001; 125:279-84) was applied on all statistical tests with the “multitest package” and an adjusted p-value below 0.05 was considered as statistically significant. A logarithmic transformation (log 10) was applied on the biomarker expression levels to ensure the data normality. All data distributions are illustrated as medians and boxplots for each biomarker. A Pearson test correlation was applied to identify biomarker correlation for all patient groups.

(28) The marker diagnostic performance could be characterised by sensitivity, which represents its ability to detect the IC population, and specificity which represents its ability to detect the control population.

(29) The results of the evaluation of a diagnostic test can be summarised in a 2×2 contingency table comparing these two well-defined populations. By fixing a cut-off the two populations could be classified into categories according to the results of the test, categorised as either positive or negative. Given a particular marker, we can identify a number of subjects with a positive test result among the “cases” population (the “True Positive”: TP) and b subjects with a positive test result among the “controls” population (the “True Negative”: TN). In the same fashion, c subjects with a negative test result among the cases (the “False Positive”: FP) and d subjects with a negative test result among the controls (the “False Negative”: FN) are observed. Sensitivity is defined as TP/(TP+FN); which is herein referred to as the “true positive rate”. Specificity is defined as TN/(TN+FP); which is herein referred to as the “true negative rate”.

(30) The accuracy of each marker and its discriminatory power was evaluated using a Receiving Operating Characteristics (ROC) analysis. ROC curves are the graphical visualization of the reciprocal relation between the sensitivity (Se) and the specificity (Sp) of a test for various values.

(31) In addition, all markers were combined with Mannan antigen (Ag) to evaluate the potential increase in sensibility and specificity using several approaches as mROC program (Kramar et al. Computer Methods and Programs in Biomedicine 2001; 66:199-207), logistic regression (Kleinbaum, D. G., Kupper, L. L., Muller, K. E. (1988) Applied Regression Analysis and other Multivariate Methods. Duxbury Press, Belmont, Calif.) and with two supervised learning algorithms, CART (Breiman L. Classification and regression trees. Wadsworth International Group, 1984.) and wKNN (Hechenbichler K, Schliep K. Weighted k-Nearest-Neighbor Techniques and Ordinal Classification. Volume 399, 2004).

(32) mROC is a program developed by Kramar et al. (Comput Methods Programs Biomed, 2001, 66:199-207) which is dedicated to identify the linear combination which maximizes the AUC (Area Under the Curve) of ROC curves. The use of this program was described for instance in Staack et al. BMC Urol 2006; 6:19. This program implements an algorithm for maximising rank correlation estimation which is also an estimate for the area under the ROC curve (Su and Liu. Journal of the American Statistical Association 1993; 88:1350-1355; Wang, Computational Statistics and Data Analysis 2007; 51:2803-2812). The equation for the respective combination is provided and can be used as a new virtual marker Z, as follows:
Z=a×Marker.sub.1+b×Marker.sub.2+c×Marker.sub.3,

(33) where a, b, c are calculated coefficients and Marker.sub.1,2,3 are individual level of markers.

(34) A logistic regression model was also applied for univariate and multivariate analysis to estimate the relative risk of IC at different biomarkers values. We analyzed biomarkers as both continuous (data not shown) and categorical (using the quartile values as cutpoints) variables. In the last cases, the odds ratio (OR) and their 95% confidence interval are computed.

(35) A CART (Classification And Regression Trees) approach was also applied to assess (biomarker+Mannan Ag) combinations. This decision tree approach allows to produce a set of classification rules, represented by a hierarchical graph easily understandable for the user. At each node of the tree, a decision is made. By convention, the left branch corresponds to a positive response to the question of interest and the right branch corresponds to a negative response to the question of interest. The classification procedure can then be translated as a set of rules ‘IF-THEN’.

(36) A wKNN (weighted k-nearest neighbours) approach was applied as previously to assess (biomarker+Mannan Ag) combinations. The wKNN algorithm is one of the variations of KNN method which uses the K nearest neighbours, regardless of their classes, but then uses weighted votes from each sample rather than a simple majority or plurality voting rule. Given a patient x, each of the K samples provides a weighted vote that is usually equal to some decreasing function of its distance from the unknown sample x. These weighted votes are then summed for each neighbour, and the class with the largest total vote is attributed to x.

(37) CART and wKNN are supervised learning methods. These methods require the use of a training set used to construct the model and a test set to validate it. So, we have shared our data set: ⅔ of the dataset are used for the learning phase and ⅓ are used for the validation phase. This sharing has been randomized and respect the initial proportion of the various statutes in each sample. To estimate the errors prediction of these two classifiers, we used the 10-fold cross-validation method, repeated 10 times in order to avoid overfitting problems. For these approaches, we used the “ipred package”, the “rpart package” and the “kknn package” of the R software.

(38) Hierarchical Ascendant Clustering Analysis (HAC) is a method of cluster analysis, based on a pairwise distance matrix, which builds a hierarchy of clusters with sequentially agglomerative and divisive approaches. We have used this method to organize the map and to group the sample according to the nearest level of biomarker intensity. For this analysis, raw data were mean-centred and Pearson correlation matrix and average linkage were chosen as parameters.

(39) 2) Results

(40) Standardization of Tests.

(41) For each experiment, 3 serum controls were used. Negative control collected from healthy subject and 2 positive sera consisted of 1 pool of sera with known reactivity against Candida albicans mannan and one serum collected from patient belonging to IC group selected from previous series of experiments. All these controls allowed us to reduce inter-experiments variations.

(42) A study of diagnosis potential of different recombinant proteins was performed by comparison of medians. When serological data were analyzed, antibodies against Fba1 (Fba1 Ab), Hwp1 (Hwp1 Ab), Hsp90 (Hsp90 Ab), Eno1 (Eno1 Ab) and Mp65 (Mp65 Ab) are the best biomarkers to discriminate IC patients from controls, however humoral response against Sod5 was less discriminant for both groups.

(43) Comparison of Serological Reactivity Against a Panel of Antigens within IC and Control Groups

(44) Using all group of patients (IC versus controls), boxplots were performed for each combination of mannanemia test and EIA tests involving recombinant proteins (FIGS. 1-5). When antibody response against each recombinant protein associated to mannanemia test was compared to mannanemia and anti-mannan antibody tests, significant differences were observed for Fba1 (p<0.0001; FIG. 1), Hwp1 (p<0.0001; FIG. 2), Hsp90 (p<0.0001; FIG. 3), Eno1 (p<0.0001; FIG. 4) and Mp65 (p<0.0001; FIG. 5).

(45) Analysis of Discriminatory Potential of Each RP-Ab/Mannanemia Combination in Compared with the Mannanemia and Anti-Mannan Antibody Association

(46) Combinations of more than 2 markers were tested however none have significantly improved the results obtained combining mannanemia and one of anti-protein recombinant antibodies.

(47) ROC curves show the improvement of IC diagnostic with combination of RP-Ab and mannan antigen compared with combination of mannan antigen and anti-mannan antibody (FIGS. 6-10).

(48) Comparison of Sensitivity and Specificity of Mannanemia and RP-Ab Combined Analysis.

(49) Retrospective analysis of the cohort allowed sensitivity and a specificity of 26.6% and 99.4% respectively for mannanemia alone when combination of Platelia™ Candida Ag and Ab showed an improvement of sensitivity (80.0%) and decrease of specificity (61.7%).

(50) The ROC curves obtained from combination of RP-Ab/mannanemia (combined marker analysis by sera) showed a significant improvement of the diagnostic performances as reveled by AUCs: 0.902, 0.886, 0.884, 0.872, and 0.853 for Fba1, Hwp1, Hsp90, Eno1, and Mp65 respectively versus 0.769 for mannanemia and anti-mannan antibody combination (Table 5).

(51) Furthermore, the association of RP-Ab and mannanemia increases significantly sensitivity and specificity. With a specificity arbitrarily fixed at 80.0% for the combination of mannan Ag+mannan Ab, sensitivities of mannan Ag+Hsp90, mannan Ag+Fba1 Ab, mannan Ag+Hwp1 Ab, mannan Ag+Eno1 Ab, and mannan Ag+Mp65 Ab were 80.9%, 83.8%, 83.8%, 79.1%, and 75.5%, respectively while the current serological diagnosis tests that combined mannan Ag+mannan Ab has a sensitivity of 61.7% (Table 5).

(52) TABLE-US-00005 TABLE 5 Diagnosis potential (mROC approach) of RP-Ab associated with mannanemia for IC diagnosis. Combination of 2 Cut- biomarkers AUC off* Se(%) Sp(%) PPV(%) NPV(%) CI95% Mannan Ag* + Fba1 Ab* 0.902 −1.975 83.8 80.0 86.9 75.0 [0.871; 0.927] Mannan Ag* + Hwp1 Ab* 0.886 −2.629 83.8 80.0 87.2 75.2 [0.851; 0.915] Mannan Ag* + Hsp90 Ab* 0.884 −2.031 80.9 80.0 86.8 72.0 [0.848; 0.911] Mannan Ag* + Eno1 Ab* 0.872 −2.209 79.1 80.0 86.6 70.1 [0.835; 0.901] Mannan Ag* + Mp65 Ab* 0.853 −1.483 75.5 80.0 86.4 66.8 [0.815; 0.885] Mannan Ag* + Mannan Ab* 0.769 −0.869 61.7 80.0 83.8 56.3 [0.723; 0.809] *Markers normalized by a log10 transform. AUC: area under the curve; Se: sensibility; Sp: specificity; PPV: positive predictive value (measures the proportion of subjects with positive test results who are correctly diagnosed); NPV: negative predictive value (measures the proportion of subjects with negative test results who are correctly diagnosed); CI 95%: 95% confidence interval. The cut-off value was set in order to specificity to 80%.

(53) Noteworthy, the contribution of mannanemia and RP-Ab association was significantly higher for Candida parapsilosis infected patients where RP-Ab reached a sensitivity of 67.2% versus 21.3% for anti-mannan antibodies. Such an improvement was also observed for episodes determined by Candida albicans, Candida kruseï, Candida glabrata, Candida tropicalis and Candida lusitaniae.

(54) As compared with the combination of mannanemia with anti-mannan antibodies, mannanemia and RP-Ab combination improved specificity of detection of IC associated with Geotrichum capitatum or Candida norvegiensis.

(55) Analysis of RP-Ab/Mannanemia for the Precocity of IC Diagnosis

(56) The contribution of RP-Ab/mannanemia association to early diagnosis of IC was performed by considering only serum samples collected during the period day −15 and the day of isolation of yeasts species from blood culture (day of positivation of blood culture).

(57) All combinations were able to significantly differentiate IC from Controls (p<0.0001) with higher values of AUC than mannan Ag/Ab mannan (Table 6).

(58) TABLE-US-00006 TABLE 6 Analysis of biomarkers association performance in two weeks before isolation of yeasts from blood samples. Nfold Marker combinations pWILCOX pWILCOX_FDR pWELCH pWELCH_FDR Median AUC Mannan Ag + Fba1 Ab 0.0001 0.00011 0.00010 0.00012 2.06 0.892 Mannan Ag + Hwp1 Ab 0.0001 0.00011 0.00010 0.00012 1.63 0.872 Mannan Ag + Hsp90 Ab 0.0001 0.00011 0.00010 0.00012 1.64 0.871 Mannan Ag + Eno1 Ab 0.0001 0.00011 0.00010 0.00012 1.48 0.863 Mannan Ag + Mp65 Ab 0.0001 0.00011 0.00010 0.00012 1.63 0.826 Mannan Ag + Mannan Ab 0.0001 0.00011 0.00010 0.00012 1.19 0.719

(59) Determination of mean day of positivity of different biomarkers was performed between day −15 and the day of isolation of yeast from blood samples. For all of these markers, the mean of positivity is between −5 and −6 days before the isolation of yeasts from blood samples.

(60) TABLE-US-00007 TABLE 7 Determination of mean day of positivity of different RP-Ab and mannanemia association in comparison with the isolation day of yeasts in blood samples. Marker combinations Mean (positive Day) Mannan Ag + Mannan Ab −5.05 Mannan Ag + Fba1 Ab −5.59 Mannan Ag + Hsp90 Ab −5.28 Mannan Ag + Eno1 Ab −5.48 Mannan Ag + Hwp1 Ab −5.41 Mannan Ag + Mp65 Ab −5.31

(61) Accordingly, all RP-Ab/mannanemia remained discriminant for IC even if no significant difference with mannanemia/anti-mannan antibody in terms of mean delay of positivity before blood culture (5-6 days before positive blood culture).

(62) Determination of Diagnostic Odds Ratio of Different Antibody and Antigen Combination For the Diagnosis of IC

(63) Odds ratio reflect the scale of risk for developing IC according to the intensity of antibody response against RP and mannanemia levels. Knowing that the prognosis of IC is closely correlated with the delay of initiation of antifungal therapy, this ratio could help to identify patients that need an early antifungal treatment.

(64) For all the associations of markers, 3 modalities of response intensity were determined in function of the repartition of values obtained in the cohort (using the quartile values as cutpoints). So, for each combination of markers the values of modality1 (mod1) are relative to the interval [Min, T1[(1.sup.st tertile), the values of modality2 (mod2) are relative to the interval [T1, T2[(2.sup.nd tertile) and the values of modality3 (mod3) are relative to the interval [T2, Max[(3.sup.rd tertile). The more the intensity of response is high, the more the risk of being IC is important.

(65) The Anti-mannan antibody and mannanemia combination was associated with significant and adjusted Odds Ratios (OR) varying between 2,4 and 17.5. In comparison, the (Mp65 Ab and mannanemia) combination was associated with significant and adjusted Odds Ratios varying between 5.5 and 47.9. The (Eno1 Ab and mannanemia) combination was associated with significant and adjusted Odds Ratios varying between 7.7 and 59.4. The (Hwp1 Ab and mannanemia) combination was associated with significant and adjusted Odds Ratios varying between 7.1 and 65.5. The (Hsp90 Ab and mannanemia) combination was associated with significant and adjusted Odds Ratios varying between 6.8 and 77.5. The (Fba1 Ab and mannanemia) combination was associated with significant and adjusted Odds Ratios varying between 8.8 and 108.2 for (Tables 8-13).

(66) TABLE-US-00008 TABLE 8 Determination of Odds Ratios on Mannan Ag + Mannan Ab according to the intensity of signals Odds Ratio Estimates 95% Wald Point Confidence Effect Modality Interval Estimate Limits Mannan Ag + Mannan [0.05; 0.64[ vs 2.414 1.519 3.838 Ab mod2 vs mod1 [−2.19; 0.05[ Mannan Ag + Mannan [0.64; 8.65[ vs 17.485 9.041 33.817 Ab mod3 vs mod1 [−2.19; 0.05[ Mannan Ag + Mannan [0.64; 8.65[ vs 7.242 3.763 13.936 Ab mod3 vs mod2 [0.05; 0.64[

(67) TABLE-US-00009 TABLE 9 Determination of Odds Ratios on Mannan Ag + Mp65 Ab according to the intensity of signals Odds Ratio Estimates Point 95% Wald Effect Modality Interval Estimate Confidence Limits Mannan Ag + Mp65 [−0.09; 1.43[ vs 5.456 3.324 8.956 Ab mod2 vs mod1 [−6.89; −0.09[ Mannan Ag + Mp65 [1.43; 10.37[ vs 47.887 21.578 106.276 Ab mod3 vs mod1 [−6.89; −0.09[ Mannan Ag + Mp65 [1.43; 10.37[ vs 8.777 3.979 19.360 Ab mod3 vs mod2 [−0.09; 1.43[

(68) TABLE-US-00010 TABLE 10 Determination of Odds Ratios on Mannan Ag + Eno1 Ab according to the ntensity of signals Odds Ratio Estimates Point 95% Wald Effect Modality Interval Estimate Confidence Limits Mannan Ag + Eno1 [−0.30; 1.61[ vs 7.739 4.613 12.983 Ab mod2 vs mod1 [−3.75; −0.30[ Mannan Ag + Eno1 [1.61; 20.74[ vs 59.421 26.538 133.049 Ab mod3 vs mod1 [−3.75; −0.30[ Mannan Ag + Eno1 [1.61; 20.74[ vs 7.678 3.466 17.011 Ab mod3 vs mod2 [−0.30; 1.61[

(69) TABLE-US-00011 TABLE 11 Determination of Odds Ratios on Mannan Ag + Hwp1 Ab according to the intensity of signals. Odds Ratio Estimates Point 95% Wald Effect Modality Interval Estimate Confidence Limits Mannan Ag + Hwp1 [−0.34; 1.90[ vs 7.065 4.237 11.781 Ab mod2 vs mod1 [−4.45; −0.34[ Mannan Ag + Hwp1 [1.90; 11.66[ vs 65.464 28.095 152.538 Ab mod3 vs mod1 [−4.45; −0.34[ Mannan Ag + Hwp1 [1.90; 11.66[ vs 9.266 4.018 21.365 Ab mod3 vs mod2 [−0.34; 1.90[

(70) TABLE-US-00012 TABLE 12 Determination of Odds Ratios on Mannan Ag + Hsp90 according to the intensity of signals. Odds Ratio Estimates 95% Wald Point Confidence Effect Modality Interval Estimate Limits Mannan Ag + Hsp90 [−0.35; 1.89[ vs 6.777 4.068 11.289 Ab mod2 vs mod1 [−7.45; −0.35[ Mannan Ag + Hsp90 [1.89; 12.40[ vs 77.458 31.551 190.161 Ab mod3 vs mod1 [−7.45; −0.35[ Mannan Ag + Hsp90 [1.89; 12.40[ vs 11.430 4.705 27.771 Ab mod3 vs mod2 [−0.35; 1.89[

(71) TABLE-US-00013 TABLE 13 Determination of Odds Ratios on Mannan Ag + Fba1 Ab according to the intensity of signals. Odds Ratio Estimates Point 95% Wald Effect Modality Interval Estimate Confidence Limits Mannan Ag + Fba1 [−0.39; 2.17[vs 8.762 5.180 14.822 Ab mod2 vs mod1 [−8.08; −0.39[ Mannan Ag + Fba1 [2.17; 11.96[vs 108.250 40.808 287.149 Ab mod3 vs mod1 [−8.08; −0.39[ Mannan Ag + Fba1 [2.17; 11.96[vs 12.354 4.745 32.165 Ab mod3 vs mod2 [−0.39; 2.17[

(72) Thus, according to this scale, association of anti-mannan antibody and mannanemia allows a risk at a maximum of 17.5 while the risk obtained with the association with mannanemia and RP-Ab reaches 47.9, 59.4, 65.5, 77.5 and 108.2 for respectively Mp65, Eno1, Hwp1, Hsp90 and Fba1.

(73) Performance of the Combined Interpretation of the Separated Biomarker Assays

(74) The performance of the diagnosis method based on the combined interpretation of the separated biomarker assays, i.e. based on determining if an elevated level of said Candida glycan, and/or an elevated level of antibody directed against said Candida protein selected from the group consisting of Fba1, Eno1, Hsp90, Hwp1, and Mp65, relative to their respective reference level is detected, was evaluated both on sera and patients. In the evaluation on sera, it is evaluated if the biomarker levels measured in each serum sample of any patient correctly led to the identification of the patient as having or not having invasive candidiasis. In the evaluation on patients, it is evaluated if altogether the biomarker levels measured in sera samples of a given patient correctly led to the identification of the patient as having or not having invasive candidiasis.

(75) The performances on sera using combined interpretation of separated assays, as detailed in Table 14, show that for a specificity set at 79.9% (about 80%) the levels of sensitivity obtained for the different biomarker combinations Mannan Ag+protein Ab, although not identical to those obtained on sera using the combined analysis of the biomarkers (results shown in Table 5), are all improved as compared with the reference test Mannan Ag+Mannan Ab.

(76) TABLE-US-00014 TABLE 14 Performances on sera using combined interpretation of separated assays Markers Sensitivity (%) Specificity (%) Mannan Ag + Hsp90 Ab 74.8 79.9 Mannan Ag + Mp65 Ab 75.6 79.9 Mannan Ag + Fba1 Ab 83.5 79.9 Mannan Ag + Eno1 Ab 78.3 79.9 Mannan Ag + Hwp1 Ab 81.5 79.9 Mannan Ag + Mannan Ab* 62.6 75.1 *sensitivity and specificity values of the current Platelia Candida antigen (Ag) and Platelia Candida Ab Plus tests (Bio-Rad Laboratories, Marnes-La-Coquette, France).

(77) The comparison of Table 14 and Table 15 shows that the performances on sera or on patients using combined interpretation of separated assays are similar.

(78) TABLE-US-00015 TABLE 15 Performances on patients using combined interpretation of separated assays Markers Sensitivity (%) Specificity (%) Mannan Ag + Hsp90 Ab 80.6 79.9 Mannan Ag + Mp65 Ab 83.9 79.9 Mannan Ag + Fba1 Ab 84.9 79.9 Mannan Ag + Eno1 Ab 82.8 79.9 Mannan Ag + Hwp1 Ab 84.9 79.9 Mannan Ag + Mannan Ab* 61.3 75.1 *sensitivity and specificity values of the current Platelia Candida antigen (Ag) and Platelia Candida Ab Plus tests (Bio-Rad Laboratories, Marnes-La-Coquette, France).