AU2012215436B2

AU2012215436B2 - Combination of biomarkers for forecasting a response or non-response to an anti-HCV treatment

Info

Publication number: AU2012215436B2
Application number: AU2012215436A
Authority: AU
Inventors: Mohammad Afshar; Tarik Asselah; Isabelle Catherine Batxelli; Ivan Bieche; Nathalie JULLIAN; Nadine Lambert; Patrick Marcellin; Eve Laure Mathieu; Michel Vidaud; Benedicte Watelet
Original assignee: Bio Rad Europe GmbH; Centre National de la Recherche Scientifique CNRS; Assistance Publique Hopitaux de Paris APHP; Institut National de la Sante et de la Recherche Medicale INSERM; Ariana Pharmaceuticals SA
Current assignee: Bio Rad Europe GmbH; Centre National de la Recherche Scientifique CNRS; Assistance Publique Hopitaux de Paris APHP; Institut National de la Sante et de la Recherche Medicale INSERM; Ariana Pharmaceuticals SA
Priority date: 2011-02-09
Filing date: 2012-02-09
Publication date: 2015-07-09
Anticipated expiration: 2032-02-09
Also published as: WO2012107529A1; AU2012215436A1; FR2971257B1; FR2971257A1

Abstract

The application relates to means for forecasting whether an individual infected with one or more HCVs has a strong probability of responding to an anti-HCV treatment that will comprise the administration of interferon and of ribavirin, or whether, on the contrary, this individual has a strong probability of not responding to this anti-HCV treatment. The means of the invention implement in particular the assay of the expression levels of selected genes, said selected genes being: - at least two genes among HERC5, IL8 and STAT2; and - at least one gene among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IFI27, IFI35, IFI44, IFIT1, IFIT4, IFITM1, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18.

Description

PCT/EP2012/052232 1 TITLE COMBINATION OF BIOMARKERS FOR PREDICTING A RESPONSE OR NON-RESPONSE TO AN ANTI-HCV TREATMENT 5 FIELD OF THE INVENTION The application relates to means for establishing a prediction of a high probability of a response or non-response to an anti-hepatitis C virus (HCV) treatment. Advantageously, the means of the invention can be used to establish this prediction before the anti-HCV 10 treatment has even commenced. BACKGROUND TO THE INVENTION In the vast majority of cases, an infection with the hepatitis C virus (HCV) leads to chronic hepatitis C. Chronic hepatitis C can develop into cirrhosis of the liver with portal 15 hypertension complications, and can also develop into hepatocellular carcinoma. One of the aims of treatment against an infection by HCV, more particularly against chronic hepatitis C, is to arrive at the stage where the attacks on the liver tissue induced by the viral infection regress or are even eliminated, or at least that they do not progress. In particular, this means that the risk which arises of complications and hepatocellular carcinoma can be reduced or 20 eliminated. Currently available treatments for achieving this aim are treatments which are aimed at eradicating the virus. In the first place, these treatments have to induce a significant reduction in the viral HCV load, so as to be able to obtain an undetectable level at the end of treatment. Current anti-HCV treatments comprise the administration of a combination of pegylated interferon and ribavirin. These treatments are of long duration: they are generally 25 administered over a period of at least 24 weeks and may last up to 48 weeks or even longer. However, anti-HCV treatments cause major side effects for the patient. Regarding interferon, the side effects are frequent and numerous. The most frequent side effect is that of influenza-like syndrome (fever, arthralgia, headaches, chills). Other possible 30 side effects are: asthenia, weight loss, moderate hair loss, sleep problems, mood problems and irritability, which may have repercussions on daily life, difficulties with concentrating and skin dryness. Certain rare side effects, such as psychiatric problems, may be serious and have to be anticipated. Depression may occur in approximately 10% of cases. This has to be PCT/EP2012/052232 2 identified and treated, as it can have grave consequences (attempted suicide). Dysthyroidism may occur. Furthermore, treatment with interferon is counter-indicated during pregnancy. Regarding ribavirin, the principal side effect is haemolytic anaemia. Anaemia may lead to 5 treatment being stopped in approximately 5% of cases. Decompensation due to an underlying cardiopathy or coronaropathy linked to anaemia may arise. Neutropenia is observed in approximately 20% of patients receiving a combination of pegylated interferon and ribavirin, and represents the major grounds for reducing the 10 pegylated interferon dose. The cost of these treatments is also very high. In order to be able to predict, before having even commenced administration of the anti-HCV treatment, whether a given patient will or will not respond to treatment is thus of major clinical and economic importance. 15 Research into predictive means of this type has led to various clinical, biological and viral factors being analysed. Certain clinical factors of the patient, such as age, weight, ethnic origin and hepatic fibrosis 20 score are known to influence the efficacy of anti-HCV treatment. As an example, the number of patients responding to anti-HCV treatment is lower among patients with a hepatic fibrosis score of F3 or F4 compared with those for whom the hepatic fibrosis score is Fl or F2 (scores using the Metavir F score system). Of themselves, 25 however, these clinical factors cannot be used to reliably predict, prior to starting a treatment, whether a given patient will or will not respond to an anti-HCV treatment. Thus, of themselves, these factors are not good pre-therapeutic prognostic indicators. In order to attempt to predict, before administering any treatment, whether a patient will or 30 will not respond to an anti-HCV treatment, in fact it is viral factors which are currently being used. It has in fact been shown that patients who are infected with an HCV of genotype 2 or 3 respond better to anti-HCV treatment than those who are infected with HCV of genotype 5 or PCT/EP2012/052232 3 6, who in turn respond better to anti-HCV treatment than those who are infected with an HCV of genotype 1 or 4. However, the distribution of the various genotypes is not homogeneous with respect to geographical locations, and thus simply discerning the viral genotype does not provide a predictive solution which can be applied to all patients. What is 5 more, there are differences between the viral sub-types. In fact, knowledge of the nature of the viral genotype can essentially be used to adjust the posology and/or duration of treatment, but cannot per se be used to establish a reliable prediction before starting treatment. Various combinations of biological and/or clinical and/or viral factors have also been tested 10 in order to attempt to predict, before administering any treatment, whether a patient will or will not respond to an anti-HCV treatment. However, the combinations which have been tested up to now have not achieved satisfactorily predictive performances. As an example, Hidetsugu Saito et al. 2010 succeeded in identifying combinations of 15 biological, clinical and viral factors which gave reliable predictive performances when they were applied during treatment, but they were not at all able to identify a combination which was sufficiently reliable when applied before starting anti-HCV treatment. Chen et al. 2005 and Chen et al. 2010 proposed a transcriptome signature for predicting, 20 before any anti-HCV treatment was administered, whether a patient would be a responder or non-responder to this treatment. That signature combined the levels of expression of eighteen genes (G1P2, OAS2, G1P3, OAS3, RPLP2, CEBI, IFITI, VIPERIN, RPS28, PI3KAP1, MX1, DUSPI, ATF5, LAP3, USP18, LGPl, ETEFI and STXBP5). 25 Further, at least two of those genes code for proteins which are exclusively membrane proteins (G1P3 and VIPERIN); thus, the product of the expression thereof cannot be detected in the bloodstream. Asselah et al. 2008 analysed the level of expression of fifty-eight genes before applying 30 anti-HCV treatment to forty patients with chronic hepatitis C, fourteen of whom were non responders to anti-HCV treatment. They thus identified two signatures which might be able to predict, before administering any anti-HCV treatment, whether a patient would be a non-responder to that treatment.

PCT/EP2012/052232 4 The first signature was based on the levels of expression of two genes, namely IF127 and CXCL9, which were analysed using the KNN method (k-nearest neighbour method). The second signature was based on the levels of expression of three genes, namely IF127, 5 CXCL9 and IFI-6-16, which were analysed using the WV method (weighted voting method). For each of these two signatures, Asselah et al. 2008 indicated that the fact of adding supplemental genes did not allow the accuracy of the classification to be improved. Thus, there is still a need for means which could be used to predict, even before 10 commencing to administer the anti-HCV treatment, whether the patient has a high probability of responding or, in contrast, a high probability of not responding to treatment. SUMMARY OF THE INVENTION 15 The application relates to means which can be used to establish a high probability prediction of response or non-response to an anti-HCV treatment. Advantageously, the means of the invention can be used to make this prediction before the anti-HCV treatment has even begun. 20 The inventors have identified genes the levels of expression of which are predictive biomarkers of a response or non-response to anti-HCV treatment. More particularly, the inventors propose establishing the expression profile of these genes, and of using this profile as a predictive signature of response or non-response to anti-HCV treatment. 25 The application provides means which are especially suited to this purpose. In particular, the means of the invention implement a measurement or assay of the levels of expression of selected genes, said selected genes being selected from the following list of genes: HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, 30 IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18. More particularly, said selected genes comprise: - at least two genes from among HERC5, IL8 and STAT2; and PCT/EP2012/052232 5 - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT 1, STMN2 and USP 18. 5 Optionally, the means of the invention may further employ a measurement or assay of one or more clinical factors and/or one or more other virological factors and/or one or more other biological factors. In particular, the means of the invention comprise: 10 - methods which comprise the measurement or assay of the levels of expression of selected genes; - products or reagents which are specially adapted to the measurement or assay of these levels of gene expression; - manufactured articles, compositions, pharmaceutical compositions, kits, tubes or 15 solid supports comprising such products or reagents, as well as - computer systems (in particular a computer program product and computer device) which are specially adapted to implementing the means of the invention. DETAILED DESCRIPTION OF THE INVENTION 20 The stage of liver tissue damage, more particularly the nature and extent of hepatic tissue lesions, is evaluated by a hepatic fibrosis score, in particular using the Metavir F score system, which comprises five stages from FO to F4. 25 When the hepatic fibrosis score is at most F1, the clinician may optionally decide not to administer anti-HCV treatment, but when the score is at least F2, the current recommendation is to administer an anti-HCV treatment irrespective of the level of necrotico-inflammatory activity. 30 Since anti-HCV treatments are of very long duration (generally 6 to 12 months, or even longer), they induce particularly serious side effects and are very expensive; the present application proposes means for assisting in the decision as to whether or not to administer anti-HCV treatment.

PCT/EP2012/052232 6 The means of the invention can be used to provide a prediction of a high probability of a response or non-response to anti-HCV treatment. Advantageously, the means of the invention can be used to establish this prediction before this treatment has even begun. 5 The means of the invention comprise measuring or measuring the levels of expression of selected genes. They concern subjects who are infected with one or more hepatitis viruses, at least one of which is an HCV, and more particularly those of these subjects who have a hepatic fibrosis score of at least F1, more particularly at least F2, using the Metavir F score system. 10 In the application, unless otherwise specified, or unless the context indicates otherwise, all of the terms used have their usual sense in the domain(s) concerned. The expression "anti HCV treatment", "hepatitis C treatment" or an equivalent expression or the shortened term "treatment" signifies a treatment for therapeutic purposes which is intended to induce a 15 reduction in the HCV load of the patient such that at the end of the treatment, an undetectable level of HCV load, or even eradication of the HCV or HCVs, is obtained. Clinically, the desired therapeutic intention is to stop or cause to regress or even to eliminate liver tissue lesions, i.e. at the very least to prevent the hepatic fibrosis score from increasing, or even for that score to reduce, preferably to a score of at most F 1. 20 The anti-HCV treatment comprises at least one administration of interferon, more particularly alpha interferon, in particularly alpha-2a interferon or alpha-2b interferon, or a prodrug of interferon. 25 This interferon is generally a version produced by genetic engineering of natural human cytokine. However, this interferon may be an interferon which derives from Chinese hamster ovary cells (CHO cells), such as omega interferon (for example, omega interferon available from Intarcia Therapeutics, Hayward, California, USA). 30 This interferon may in particular be associated with other chemical compounds, groups or molecules, in particular polyethylene glycol (for example, PEG-INTRON@ supplied by Schering Plough Corporation, Kenilworth, New Jersey, USA, or PEGASYS@ supplied by F. Hoffmann-La Roche Ltd.; Basel, Switzerland).

PCT/EP2012/052232 7 The pegylated form of interferon has a longer lifetime in the human body, which means that the frequency of administration can be limited to a single administration per week (in the event, a single injection per week) instead of three administrations per week for the non-pegylated form. 5 The pegylated form of interferon is thus currently the preferred form of interferon. A pegylated interferon may, for example, be administered: - in a dose of approximately 1.5 g/kg/week for pegylated alpha-2b interferon (such as PEG-INTRON@), - at a concentration of 180 g/kg/week for pegylated alpha-2a interferon (such as 10 PEGASYS@). In addition to interferon, an anti-HCV treatment generally includes administering at least one other antiviral agent. 15 In addition to interferon, current anti-HCV treatment generally includes administering ribavirin. Ribavirin is a nucleoside analogue of guanosine. 20 In the context of the application, and in accordance with a particular embodiment of the invention, the anti-HCV treatment comprises administering interferon and administering: - ribavirin (for example, the ribavirin REBETOL@ supplied by Plough Corporation, Kenilworth, New Jersey, USA, or the ribavirin COPEGUS@ supplied by Roche 25 Corporation; F. Hoffmann-La Roche Ltd.; Basel, Switzerland), or - an analogue of ribavirin, or - a prodrug of ribavirin or one of its analogues. Ribavirin prodrugs in particular include taribavirin (for example, the taribavirin which is 30 available from Valeant, Aliso Viejo, California, USA). The ribavirin is preferably administered daily. The ribavirin may, for example, be administered in an amount of 800 to 1200 mg/kg/day.

PCT/EP2012/052232 8 An anti-HCV treatment may, for example, comprise the administration of: - pegylated alpha-2b interferon (such as PEG-INTRON@) in a dose of approximately 1.5 g/kg/week, and ribavirin in a dose of 800 to 1 200 mg/kg/day (if the hepatopathy 5 involves an HCV of genotype 2 or 3, a dose of approximately 800 mg/kg/day is generally advised), or - pegylated alpha-2a interferon (such as PEGASYS@) in a concentration of 180 g/kg/week and ribavirin in an amount of 1000 to 1200 mg/kg/day. 10 In addition to interferon, or interferon and ribavirin, the anti-HCV treatment may also comprise administration of at least one other generic or specific HCV antiviral agent, such as: - at least one HCV protease inhibitor, such as an NS3 protease inhibitor, and/or - at least one HCV polymerase inhibitor, such as a NS5B polymerase inhibitor, 15 more particularly at least one HCV protease inhibitor, such as an NS3 protease inhibitor. Said NS3 protease inhibitor may, for example be telaprevir (VX-950; Vertex, Cambridge, Massachusetts, USA) or boceprevir (SCH-503034; Schering-Plough, Kenilworth, New Jersey, USA). The combination of interferon (or an analogue or a prodrug of interferon), 20 ribavirin (or an analogue or a prodrug of ribavirin) and an HCV protease inhibitor such as telaprevir or boceprevir (or an analogue or a prodrug of this protease inhibitor) is a tritherapy which is in particular envisaged for the treatment of patients who are infected with at least one HCV of genotype 1 or 4. 25 Said NS5B polymerase inhibitor may, for example, be a nucleoside analogue such as R1479, or its prodrug R1626 (Roche, Basel, Switzerland), or the nucleoside analogue PSI 6130, or its prodrug R7128 (Pharmasset, Princeton, NJ, U.S.A.; Roche, Basel, Switzerland). 30 In addition to the antiviral agent or agents, the anti-HCV treatment may also comprise administering at least one other product with no direct antiviral activity, such as a drug adjuvant, for example a hormone which stimulates the production of erythrocytes and/or leukocytes, such as erythropoietin (EPO).

PCT/EP2012/052232 9 The anti-HCV treatment period is generally at least approximately 24 weeks, very generally approximately 24 to 48 weeks, but sometimes longer. As an example, it may be: - approximately 24 weeks for hepatopathy due to HCV of genotype 2 or 3, - approximately 48 weeks for hepatopathy due to HCV of genotype 1, 4 or 5, or for a 5 patient who is not responsive to treatment after 24 weeks. The expressions "responder" or "non-responder" should be understood to have the meanings which are usually attributed to them in the medical field. The expressions "responder" or "non-responder" should be understood to mean "responder to anti-HCV 10 treatment" or "non-responder to anti-HCV treatment", respectively. A subject is considered to be: - a subject who is a responder to treatment (patient classified as R) when the viral load of HCV has become undetectable in the blood of the patient at the end of an anti-HCV 15 treatment associating the administration of interferon and the administration of ribavirin (or a prodrug or an analogue of these active principles) and that this viral load remains undetectable 6 months after that treatment is stopped; - a subject who is a non-responder to treatment (patient classified as NR) when the viral load of HCV remains undetectable in the blood of the patient at the end of this anti 20 HCV treatment; - a responder-relapser (patient classified as RR) when the viral load of HCV becomes undetectable in the blood of the patient at the end of this anti-HCV treatment, but it becomes detectable again 6 months after stopping this anti-HCV treatment. 25 This anti-HCV treatment is generally administered: - over approximately 24 weeks for hepatopathy due to HCV of genotype 2 or 3, - over approximately 48 weeks for hepatopathy due to HCV of genotype 1, 4, 5 or 6. The treatment may be one of the treatments mentioned above, in particular such as a 30 treatment comprising or consisting of administering ribavirin (or a prodrug or an analogue of this active principle) and alpha-2a interferon or alpha-2b interferon, more particularly pegylated interferon (more particularly, pegylated alpha-2a interferon or pegylated alpha 2b interferon), or a prodrug or an analogue of this active principle. The interferon is PCT/EP2012/052232 10 usually administered at a frequency of once a week, while the ribavirin is usually administered at a frequency of twice a day. Particular examples of treatment include the following: 5 - treatment by administration: o of pegylated alpha-2b interferon (PEG-INTRON@; Schering Plough Corporation; Kenilworth, NJ; U.S.A.) in a dose of 1.5 g/kg/week, and o of ribavirin (REBETOL@; Schering Plough Corporation; Kenilworth, NJ; U.S.A.), as a function of the patient's weight and the HCV genotype(s), in 10 adoseof: - 800 to 1200 mg/kg/day for those patients who have been infected with at least one genotype 1 and/or 4 and/or 5 and/or 6 of HCV, or in a dose of - 800 mg/kg/day for those patients who have been infected with at 15 least one genotype 2 and/or 3 of HCV, or - treatment by administration: o of pegylated alpha-2a interferon (PEGASYS@; Roche Corporation; F. Hoffmann-La Roche Ltd.; Basel, Switzerland) in a dose of 180 g/kg/week, 20 and o of ribavirin (COPEGUS@; Roche Corporation; F. Hoffmann-La Roche Ltd.; Basel, Switzerland) in a dose of 1000 to 1200 mg/kg/day. One or other of these two examples of treatment can be administered for 24 weeks for 25 those of the subjects who have been infected with at least one genotype 2 and/or 3 of HCV, and for 48 weeks for those subjects who have been infected with at least one genotype 1 and/or 4 and/or 5 and/or 6 of HCV. The viral load of HCV can be considered to be undetectable in the blood of a subject when 30 the measurement of HCV RNA in the serum of a subject has given a value of less than 12 International Units (IU) per mL of serum, as measured in a test for the quantification of HCV RNA, for example as measured in a quantification test carried out with the aid of a VERSANT@ HCV-RNA 3.0 (bDNA) ASSAY kit from Siemens Healthcare Diagnostics PCT/EP2012/052232 11 (quantification limit = 615 - 7 690 000 IU/mL), following the recommendations of the manufacturer of this kit. The inventors have identified genes the level of expression of which constitute biomarkers 5 which, when taken in combination, are pertinent to the determination of the status of "responder" (R) or "non-responder" (NR) of a subject. The inventors have also observed that, depending on these expression level combinations, the population of responder-relapser (RR) subjects is very strongly segregated from that of 10 the responders (R): RR subjects are mainly classified as R (see Examples below). The genes identified thereby are the following twenty-eight genes: HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, 15 RSAD2, STAT1, STMN2 and USP18. The majority of these genes code for proteins with a localization which is not transmembrane in nature, or at least not exclusively so. The majority of these genes thus code for proteins which are susceptible of being detected in a biological fluid of a subject 20 such as the blood, serum or the plasma. This is in fact the case with the twenty-eight genes listed above, with the exception of the following seven genes which code for strictly membrane proteins: CLDN1, G1P3, IF127, IFITMI, ITGA2, OCLN and PLSCR1. The inventors have also identified that the most pertinent combinations comprise: 25 - at least two genes from among HERC5, IL8 and STAT2; and - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT 1, STMN2 and USP 18. 30 Each of these genes is individually known to the skilled person and should be understood to have the meaning given to it in this field. An indicative reminder of their respective identities is presented in Table 1 below.

PCT/EP2012/052232 12 Table 1: Identity of genes Name (in French) of coded NM accession Symbol protein Name (in English) of coded protein Alias number prot6ine 5 contenant un domaine HECT domain and RCCl1-like HERC5 HECT and des domaines de type CBE1, CEBP1 NM_016323 RCC1domain-containing protein 5 RCC1 CXCL8, GCP-1, GCP1, LECT, IL8 interleukine 8 interleukin 8 LUCT, LYNAP, MDNCF, MONAP, NM_000584 NAF, NAP-1, NAPI STAT2 transducteur de signal et activateur Signal Transducer and Activator of P113, ISGF-3, STAT113, MGC59816 NM005419 de transcription 2 Transcription 2 CCL21 ligand 21 i ch6mokine (motif C-C) chemokine (C-C motif) ligand 21 ECL, SLC, SCYA21 NM_002989 CLDN1 claudine 1 Claudin-1 CLD&, SEMPI, ILVASC NM_021101 CXCL6 ligand 6 i ch6mokine (motif CXC) chemokine (CXC motif) ligand 6 CKA-3, GCP-2, GCP2, SCYB6 NM_002993 protein 01 en boite i tate de FOXO1 Forkhead box protein 01 FKH1, FKHR, FOXO1A NM_002015 fourchette PCT/EP2012/052232 13 Table 1 (continued): Identity of genes Name (in French) of coded NM accession Symbol protein Name (in English) of coded protein Alias number G1P2 prot6ine inductible par l'interf6ron interferon alpha inducible protein ISG15, IFIl5 NM005101.3 alpha (clone IFI-15K) (clone IFI-15K) prot6ine 6 inductible par 6-16, FAM14C, IF1616, IFI-6-16, G1P3 interferon alpha-inducible protein 6 NM_022873 l'interf6ron alpha IF16 IF127 protine 27 inductible par interferon alpha-inducible protein 27 P27, ISG12, FAM14D, ISG12A NM_005532 l'interf6ron alpha IF35 protine de 35kDa induite par interferon-induced 35kDa protein IFP35, FLJ21753 NM_005533 l'interf6ron IF144 prot6ine 44 induite par l'interf6ron interferon-induced protein 44 P44, MTAP44 NM_006417 prot~ine d r~p~titions interferon-induced protein with G1OPi, IF156, ISG56, IFI-56, IFITI t6tratricopeptidiques 1 induite par .

NM_001548 l'intef~rontetratricopeptide repeats 1 IFNAI1, RNM561 l'interf~ron prot~ine d r~p~titions interferon-induced protein with IRG2, IF160, ISG60, RIG-G, CIG-49, IFIT4 t6tratricopeptidiques 3 induite par tetratricopeptide repeats 3 GARG-49, IFIT3 NM_001549 l'interf6ron PCT/EP2012/052232 14 Table 1 (continued): Identity of genes Name (in French) of coded NM accession Symbol protein Name (in English) of coded protein Alias number IFITM1 prot6ine 1 transmembranaire interferon-induced transmembrane 9-27, CD225, IF1l7, LEUl3 NM003641 induite par l'interf6ron protein 1 CD49B, BR, GPla, CD49B, VLA-2, ITGA2 integrin alpha 2 integrin alpha-2 NM_002203 prot6ine se liant i LGALS3 lectin, galactosidase-binding, LGALS3BP (lectine, se liant i la galactosidase, soluble, 3 binding protein005567.3 soluble, 3) MDK midkine midkine NEGF2 NM_001012334 prot6ine de r6sistance 1 du Myxovirus (Influenzae virus) MX1 resistance protein 1 IF178 NM_002462 (prot~ine 78 induite par (interferon induced protein 78) l'interf6ron) OAS1 synth6tase 1 2'-5'-oligoad6nylate 2'-5'-oligoadenylate synthetase 1 OIAS, IFI-4, OIASI NM_016816 OAS2 synth6tase 2 2'-5'-oligoad6nylate 2'-5'-oligoadenylate synthetase 2 MGC78578 NM_016817 OAS3 synth6tase 3 2'-5'-oligoad6nylate 2'-5'-oligoadenylate synthetase 3 p100, MGC133260 NM_006187 PCT/EP2012/052232 15 Table 1 (continued to end): Identity of genes Name (in French) of coded NM accession Symbol protein Name (in English) of coded protein Alias number OCLN occludine occludin NM_002538 PLSCR1 scramblase phospholipide 1 phospholipid scramblase 1 MMTRA1B NM_021105 prot6ine 2 contenant un domaine radical S-adenosyl methionine RSAD2 de radical de m6thionine de S- 2 Cig5, vig1, cig33, 2510004LO1Rik NM_080657 ad6nosyl transducteur de signal et activateur Signal Transducer and Activator of STAT1 ISGF-3, STAT91, DKFZp686B04100 NM_007315 de transcription 1 alpha/beta Transcription 1 -alpha/beta STMN2 stathmine 2 Stathmin-2 SCG1O, SCGN1O, SGC1O NM_007029 USP18 hydrolase carboxy-terminale Ubl Ubl carboxyl-terminal hydrolase 18 ISG43, UBP43 NM_017414 RPLP0 phosphoprot6ine ribosomale acide human acidic ribosomal 36B4 NM001002 PO humaine phosphoprotein PO TBP prot~ine se liant d une boite TATA TATA box binding protein - NM_003194 PCT/EP2012/052232 16 None of these genes is a gene of the hepatitis virus. They are mammalian genes, more particularly human genes. 5 In addition to the levels of expression of genes selected from the list of the twenty-eight genes of the invention (HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OAS 1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP 18), the means of the invention may further comprise the measurement of other factors, in particular one 10 or more clinical factors and/or one or more virological factors and/or one or more biological factors other than the level of expression of the genes selected from said list of twenty-eight genes. More particularly, in addition to the levels of expression of the genes selected from said 15 list of twenty-eight genes of the invention, the means of the invention may optionally comprise: - measuring the level of expression of mammalian genes (more particularly human genes) other than those from said list of twenty-eight genes, for example to measure the level of transcription of genes which are listed below as "other biological factors", 20 such as the gene coding for alkaline phosphatase (ALP) see Example 2d) below, and/or - measuring intracorporal metabolites (for example, cholesterol), and/or - measuring elements occurring in the blood (for example platelets), and/or - measuring the quantity of iron which is circulating. 25 However, these measurements are optional. In accordance with the application, the number of mammalian genes (more particularly human genes) the level of expression of which is measured and which are not genes 30 selected from said list of twenty-eight genes of the invention (for example the gene coding for gamma glutamyl transpeptidase and/or the gene coding for alkaline phosphatase), is preferably a maximum of 18, more particularly 14 or fewer, more particularly 11 or fewer, more particularly 6 or fewer, more particularly 4 or 3 or 2 or 1 or 0, more particularly 3 or 2 or 1 or 0, in particular 2 or 1 or 0.

PCT/EP2012/052232 17 It follows that counting these "other" mammalian genes (more particularly these "other" human genes) the level of expression of which may optionally be measured, as well as the maximum number of twenty-eight genes which may be the genes selected in accordance 5 with the invention, the total number of genes the level of expression of which is measured in a method in accordance with the application is preferably 3 to 46 genes, more particularly 3 to 42, more particularly 3 to 39, more particularly 3 to 34, more particularly 3 to 32, more particularly 3 to 31, more particularly 3 to 30, more particularly 3 to 29, in particular 3 to 28, more particularly 3 to 27, more particularly 3 to 26, more particularly 3 10 to 25, more particularly 3 to 24, more particularly 3 to 23, more particularly 3 to 22, more particularly 3 to 21, more particularly 3 to 20, more particularly 3 to 19, more particularly 3 to 18, more particularly 3 to 17, more particularly 3 to 16, more particularly 3 to 15, more particularly 3 to 14, more particularly 3 to 13, more particularly 3 to 12, more particularly 3 to 11, more particularly 3 to 10, more particularly 3 to 9, more particularly 3 15 to 8, more particularly 3 to 7 (for example 4, 5, 6 or 7), more particularly 3 to 6 (for example 4, 5 or 6). Further, as will be presented in more detail below, and as illustrated in the Examples, the number of genes selected from the list of twenty-eight genes of the invention (HERC5, 20 IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STATI, STMN2 and USP18) may advantageously be 3, 4, 5 or 6, more particularly 3, 4 or 5, in particular 4 or 5. 25 In one embodiment, the total number of genes the level of expression of which is measured, is: - 3, 4, 5 or 6 genes selected from said list of twenty-eight genes of the invention, and - 0, 1, 2, 3 or 4 "other" mammalian genes, more particularly 0, 1, 2 or 3 "other" mammalian genes, more particularly 0, 1 or 2 "other" mammalian genes. 30 In one embodiment, the total number of mammalian genes the level of expression of which is measured in a method in accordance with the application is thus 3 to 10, more particularly 3 to 9, more particularly 3 to 8, more particularly 3 to 7, more particularly 3 to 6, more particularly 3 to 5, more particularly 3 to 4.

PCT/EP2012/052232 18 The means of the invention may optionally comprise measuring the expression product (RNA or protein) of one or more non-human genes, more particularly one or more viral genes, more particularly one or more genes of the hepatitis virus, more particularly one or 5 more genes of HCV. The means of the invention may optionally comprise determining the genotype or genotypes of the HCV or HCVs with which the subject is infected. 10 The means of the invention may optionally comprise determining one or more clinical factors of said subject, such as the viral load before treatment (VLbeforeTTT in the examples below). The means of the invention may optionally comprise determining one or more biological 15 factors of said subject other than the level of expression of a gene of said subject, such as measuring the cholesterol of said subject, for example. One feature of the means of the invention is that they include the fact of measuring (or assaying) the level to which the selected genes are expressed in the organism of said 20 subject. The expression "level of expression of a gene" or equivalent expression as used here designates both the level to which this gene is transcribed into RNA, more particularly into mRNA, and also the level to which a protein encoded by that gene is expressed. 25 The term "measure" or "assay" or equivalent term is to be construed as being in accordance with its general use in the field, and refers to quantification. The level of transcription (RNA) of each of said genes or the level of translation (protein) of 30 each of said genes or the level of transcription for certain of said selected genes and the level of translation for the others of these selected genes can be measured. In accordance with one embodiment of the invention, either the level of transcription or the level of translation of each of said selected genes is measured.

PCT/EP2012/052232 19 The fact of measuring (or assaying) the level of transcription of a gene includes the fact of quantifying the RNAs transcribed from that gene, more particularly of determining the concentration of RNA transcribed by that gene (for example the quantity of those RNAs with respect to the total quantity of RNA initially present in the sample, such as a value for Ct 5 normalized by the 2 -Act method; see below). The fact of measuring (or assaying) the level of translation of a gene includes the fact of quantifying proteins encoded by that gene, more particularly of determining the concentration of proteins encoded by this gene, (for example the quantity of that protein 10 per volume of biological fluid). Certain proteins encoded by a mammalian gene, in particular a human gene, may occasionally be subjected to post-translation modifications such as, for example, cleavage into polypeptides and/or peptides. If appropriate, the fact of measuring (or assaying) the 15 level of translation of a gene may then comprise the fact of quantifying or determining the concentration, not of the protein or proteins themselves but of one or more post-translational forms of this or these proteins, such as, for example, polypeptides and/or peptides which are specific fragments of this or these proteins. 20 In order to measure or assay the level of expression of a gene, it is thus possible to quantify: - the RNA transcripts of that gene, or - proteins expressed by this gene or post-translational forms of such proteins, such as polypeptides or peptides which are specific fragments of these proteins, for example. 25 The application pertains to the subject matter defined in the claims as filed, the subject matter described below and the subject matter illustrated in the "Examples" section. In particular, the application concerns means for predicting whether a subject infected with one or more HCVs has a high probability of responding to an anti-HCV treatment which will 30 comprise administering interferon and ribavirin or whether, in contrast, that subject has a high probability of not responding to said anti-HCV treatment. In particular, the means of the invention comprise: PCT/EP2012/052232 20 - methods which include measuring or assaying the levels of expression of selected genes (level of transcription or translation); - products or reagents which are specifically adapted to measuring or assaying these levels of expression of the genes; 5 - manufactured articles, compositions, pharmaceutical compositions, kits, tubes or solid supports comprising such products or reagents; as well as - computer systems (in particular, a computer program product and computer device) which are specially adapted to implementing the means of the invention. The means of the invention, more particularly the method of the invention, are deployed 10 before treating the HCV infection, and advantageously may be deployed before the anti HCV treatment has been commenced, more particularly before any anti-HCV treatment has been commenced. In accordance with one aspect of the invention, the application thus relates to a method, 15 more particularly an in vitro method, for predicting whether a subject infected with one or more hepatitis C viruses has a high probability of responding to an anti-HCV treatment which will comprise administering interferon and ribavirin or whether, in contrast, this subject has a high probability of not responding to this anti-HCV treatment. 20 The method includes the fact of measuring the levels to which the selected genes are transcribed or translated, said selected genes being genes selected from the following list of genes: HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18. 25 More particularly, the predictive method of the application comprises the fact of measuring the levels to which the selected genes are transcribed or translated, said selected genes being: - at least two genes from among HERC5, IL8 and STAT2, and 30 - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT 1, STMN2 and USP 18.

PCT/EP2012/052232 21 These measurements may be carried out in a sample which has been obtained from said subject. In the predictive method of the invention, the total number of genes selected may be 3, 4, 5 5 or 6, more particularly 3, 4 or 5, especially 4 or 5. This being the case, as is presented and illustrated in more detail below, the predictive method of the invention may also comprise measuring or assaying one or more factors, in particular one or more virological factors and/or one or more clinical factors and/or one or 10 more biological factors other than the levels of expression of genes selected from HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18. 15 A predictive method of the invention can thus be defined by the fact that it comprises the step of carrying out measurements which comprise or are constituted by the following measurements: - in a sample which has already been obtained from said subject, measuring the levels to which the selected genes have been transcribed or translated, said selected 20 genes being: - at least two genes from among HERC5, IL8 and STAT2; and - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STATIC, 25 STMN2 and USP18, - optionally, assaying or determining, for said subject, the value of one or more clinical factors and/or of one or more virological factors and/or of one or more biological factors other than the levels of expression of genes selected from HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, 30 IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT 1, STMN2 and USP18. The application also relates to an anti-HCV therapy method which comprises the fact of predicting the response of a subject to an anti-HCV treatment with the aid of the predictive PCT/EP2012/052232 22 method of the invention. If said subject is predicted to be a non-responder, the clinician may elect not to administer a treatment which comprises (more particularly which is essentially constituted by) administering interferon and administering ribavirin (or their prodrugs), more particularly of not administering such a treatment as a first line treatment. 5 In such a situation, the clinician may, for example, elect to administer an anti-HCV treatment which does not include (or is not essentially constituted by) administering interferon and administering ribavirin (or their prodrugs), more particularly of administering such a treatment as a first line treatment. The clinician may alternatively elect not to administer the anti-HCV treatment, at least as a first line treatment. If said 10 subject is predicted to be a responder, the clinician may elect to administer an anti-HCV treatment, in particular a treatment which comprises (more particularly is essentially constituted by) administering interferon and administering ribavirin (or their prodrugs), more particularly of administering a first line treatment which comprises (more particularly which is essentially constituted by) administering interferon and administering 15 ribavirin (or their prodrugs). Measuring (or assaying) the level of expression of said selected genes may be carried out in a sample which has been obtained from said subject, such as: - a biological sample removed from or collected from said subject, or 20 - a sample comprising nucleic acids (in particular RNAs) and/or proteins and/or polypeptides and/or peptides of said biological sample, in particular a sample comprising nucleic acids and/or proteins and/or polypeptides and/or peptides which have been or are susceptible of having been extracted and/or purified from said biological sample, or 25 - a sample comprising cDNAs which have been or are susceptible of having been obtained by reverse transcription of said RNAs. A biological sample collected or removed from said subject may, for example, be a sample removed or collected or susceptible of being removed or collected from: 30 - an internal organ or tissue of said subject, in particular from the liver or the hepatic parenchyma, or - a biological fluid from said subject, in particular an intracorporal fluid such as the blood, serum, plasma or urine.

PCT/EP2012/052232 23 A biological sample collected or removed from said subject may, for example, be a sample comprising a portion of tissue from said subject, in particular a portion of hepatic tissue, more particular a portion of the hepatic parenchyma. 5 A biological sample collected or removed from said subject may, for example, be a sample comprising cells which have been or are susceptible of being removed or collected from a tissue of said subject, in particular from a hepatic tissue, more particularly hepatic cells. A biological sample collected or removed from said subject may, for example, be a sample 10 of biological fluid such as a sample of blood, serum, plasma or urine, more particularly a sample of intracorporal fluid such as a sample of blood or serum or plasma. This is the case in particular when said genes selected from said list of twenty-eight genes of the invention code for proteins which have an extra-cellular localization, or at least for those of said selected genes which code for proteins having such a localization. 15 In accordance with one embodiment of the invention, said biological sample is thus a sample of a biological fluid from said subject, such as a sample of intracorporal fluid, such as a blood, serum, plasma or urine sample, and the levels of expression of said selected genes which are measured may be levels of protein translation. 20 Said biological sample may be removed or collected by inserting a sampling instrument, in particular by inserting a needle or a catheter, into the body of said subject. This instrument may, for example be inserted: - into an internal organ or tissue of said subject, in particular into the liver or into the 25 hepatic parenchyma, for example: - to remove a sample of liver or hepatic parenchyma, said removal possibly, for example, being carried out by hepatic biopsy puncture (HBP), more particularly by transjugular or transparietal HBP, or - to remove or collect cells from the hepatic compartment (removal of cells 30 and not of tissue), more particularly from the hepatic parenchyma, in particular to remove hepatic cells, this removal or collection possibly being carried out by hepatic cytopuncture; and/or PCT/EP2012/052232 24 - into a vein, an artery or a vessel of said subject in order to remove a biological fluid from said subject, such as blood. The means of the invention are not limited to being deployed on a tissue biopsy, in 5 particular hepatic tissue. They may be deployed on a sample obtained or susceptible of being obtained by taking a sample with a size or volume which is substantially smaller than a tissue sample, namely a sample which is limited to a few cells. In particular, the means of the invention can be deployed on a sample obtained or susceptible of being obtained by hepatic cytopuncture. 10 The quantity or the volume of material removed by hepatic cytopuncture is much smaller than that removed by HBP. In addition to the immediate gain for the patient in terms of reducing the invasive nature of the technique and reducing the associated morbidity, hepatic cytopuncture has the advantage of being able to be repeated at distinct times for the same patient (for example to determine the change in the hepatic fibrosis 15 between two time periods), while HBP cannot reasonably be repeated on the same patient. Thus, in contrast to HBP, hepatic cytopuncture has the advantage of allowing clinical changes in the patient to be monitored. Thus, in accordance with the invention, said biological sample may advantageously be: 20 - cells removed or collected from the hepatic compartment (removal or collection of cells and not of tissue), more particularly from the hepatic parenchyma, i.e. a biological sample obtained or susceptible of being obtained by hepatic cytopuncture; and/or 25 - biological fluid removed or collected from said subject, such as blood or urine, in particular blood. The measurement (or assay) may be carried out in a biological sample which has been collected or removed from said subject and which has been transformed, for example: 30 - by extraction and/or purification of nucleic acids, in particular RNAs, more particularly mRNAs, and/or by reverse transcription of said RNAs, in particular of said mRNAs, or PCT/EP2012/052232 25 - by extraction and/or purification of proteins and/or polypeptides and/or peptides, or by extraction and/or purification of a protein fraction such as serum or plasma extracted from blood. 5 As an example, when the collected or removed biological sample is a biological fluid such as blood or urine, before carrying out the measurement or the assay, said sample may be transformed: - by extraction of nucleic acids, in particular RNA, more particularly mRNA, and/or by reverse transcription of said RNAs, in particular of said mRNAs (most generally by 10 extraction of RNAs and reverse transcription of said RNAs), or - by separation and/or extraction of the seric fraction or by extraction or purification of seric proteins and/or polypeptides and/or peptides. Thus, in one embodiment of the invention, said sample obtained from said subject 15 comprises (for example in a solution), or is, a sample of biological fluid from said subject, such as a sample of blood, serum, plasma or urine, and/or is a sample which comprises (for example in a solution): - RNAs, in particular mRNAs, which are susceptible of having been extracted or purified from a biological fluid such as blood or urine, in particular blood; and/or 20 cDNAs which are susceptible of having been obtained by reverse transcription of said RNAs; and/or - proteins and/or polypeptides and/or peptides which are susceptible of having been extracted or purified from a biological fluid, such as blood or urine, in particular blood, and/or susceptible of having been encoded by said RNAs, 25 preferably - proteins and/or polypeptides and/or peptides which are susceptible of having been extracted or purified from a biological fluid, such as blood or urine, in particular blood, and/or susceptible of having been encoded by said RNAs. 30 When said sample obtained from said subject comprises a biological sample obtained or susceptible of being obtained by sampling a biological fluid such as blood or urine, or when said sample obtained from said subject is obtained or susceptible of having been obtained from said biological sample by extraction and/or purification of molecules PCT/EP2012/052232 26 contained in said biological sample, the measurement is preferably a measurement of proteins and/or polypeptides and/or peptides, rather than measuring nucleic acids. When the biological sample which has been collected or removed is a sample comprising a 5 portion of tissue, in particular a portion of hepatic tissue, more particularly a portion of the hepatic parenchyma such as, for example, a biological sample removed or susceptible of being removed by hepatic biopsy puncture (HBP), or when the biological sample collected or removed is a sample comprising cells obtained or susceptible of being obtained from such a tissue, such as a sample collected or susceptible of being collected by hepatic 10 cytopuncture, for example, said biological sample may be transformed: - by extraction of nucleic acids, in particular RNA, more particularly mRNA, and/or by reverse transcription of said RNAs, in particular said mRNAs (most generally by extraction of said RNAs and reverse transcription of said RNAs), or - by separation and/or extraction of proteins and/or polypeptides and/or peptides. 15 A step for lysis of the cells, in particular lysis of the hepatic cells contained in said biological sample, may be carried out in advance in order to render nucleic acids or, if appropriate, proteins and/or polypeptides and/or peptides, directly accessible to the analysis. 20 Thus, in one embodiment of the invention, said sample obtained from said subject is a sample of tissue from said subject, in particular hepatic tissue, more particularly hepatic parenchyma, or is a sample of cells of said tissue and/or is a sample which comprises (for example in a solution): 25 - hepatic cells, more particularly cells of the hepatic parenchyma, for example cells obtained or susceptible of being obtained by dissociation of cells from a biopsy of hepatic tissue or by hepatic cytopuncture; and/or - RNAs, in particular mRNAs, which are susceptible of having been extracted or purified from said cells; and/or 30 - cDNAs which are susceptible of having been obtained by reverse transcription of said RNAs; and/or - proteins and/or polypeptides and/or peptides which are susceptible of having been extracted or purified from said cells and/or susceptible of having been coded for by said RNAs.

PCT/EP2012/052232 27 In accordance with the invention, said subject is a human being or a non-human animal, in particular a human being or a non-human mammal, more particularly a human being. 5 Because of the particular selection of genes proposed by the invention, the status of responder or non-responder of said subject may be deduced or determined from values for the measurements obtained for said subject, in particular by statistical inference and/or statistical classification, for example using reference cohorts (pre)-established according to their status of responder or non-responder. 10 In addition to measuring (or assaying) the level to which the selected genes are expressed in the organism of said subject, a method of the invention may thus further comprise a step for deducing or determining the status of responder or non-responder of said subject from values for measurements obtained for said subject. This step for deduction or determination 15 is a step in which the values for the measurements obtained for said subject are analysed in order to infer therefrom the status of responder or non-responder of said subject. The status of responder or non-responder of said subject may be deduced or determined by comparing the values for measurements obtained from said subject with their values or the 20 distribution of their values, in reference cohorts which have already been set up as a function of their status as responder or non-responder to anti-HCV treatment, in order to classify said subject into that of those reference cohorts to which it has the highest probability of belonging (i.e. to attribute to said subject its status of responder or non-responder). The individuals composing those cohorts are individuals for whom it has been established that 25 they are responders or non-responders to this treatment by applying that anti-HCV treatment. The measurements made on said subject and on the individuals of the reference cohorts or sub-populations are measurements of the levels of gene expression (transcription or translation). 30 In order to measure the level of transcription of a gene, its level of RNA transcription is measured. Such a measurement may, for example, comprise assaying the concentration of transcribed RNA of each of said selected genes, either by assaying the concentration of these RNAs or by assaying the concentration of cDNAs obtained by reverse transcription PCT/EP2012/052232 28 of these RNAs. The measurement of nucleic acids is well known to the skilled person. As an example, the measurement of RNA or corresponding cDNAs may be carried out by amplifying nucleic acid, in particular by PCR. Some reagents are described below for this purpose (see Example 1 below). Examples of appropriate primers and probes are also given 5 (see, for example, Table 10 below). The conditions for amplification of the nucleic acids may be selected by the skilled person. Examples of amplification conditions are given in the "Examples" section which follows (see Example 1 below). In order to measure the level of translation of a gene, its level of protein translation is 10 measured. Such a measurement may, for example, comprise measuring the concentration of proteins translated from each of said selected genes (for example, measuring the proteins in the general circulation, in particular in the serum). Protein measurement is well known to the skilled person. As an example, the proteins (and/or polypeptides and/or peptides) may be measured by ELISA or any other immunometric method which is known to the skilled 15 person, or by a method using mass spectrometry which is known to the skilled person. The measurement values are values of concentration or proportion, or values which represent a concentration or a proportion. The aim is that within a given combination, the measurement values for the levels of expression of each of said selected genes reflect as 20 accurately as possible, at least with respect to each other, the degree to which each of these genes is expressed (degree of transcription or degree of translation), in particular by being proportional to these respective degrees. As an example, in the case of measurement of the level of expression of a gene by 25 measurement of transcribed RNAs, i.e. in the case of measurement of the level of transcription of this gene, the measurement is generally carried out by amplification of the RNAs by reverse transcription and PCR (RT-PCR) and by measuring values for Ct (cycle threshold). 30 A value for Ct provides a measure of the initial quantity of amplified RNAs (the smaller the value for Ct, the larger the quantity of these nucleic acids). The Ct values measured for a target RNA (Ctarget) are generally related to the total quantity of RNA initially present in the sample, for example by deducing, from this Cttarget, the value for a reference Ct (Ctreerence), such as the value of Ct which was measured under the same operating conditions for the PCT/EP2012/052232 29 RNA of an endogenous control gene for which the level of expression is stable (for example, a gene involved in a cellular metabolic cascade, such as RPLP0 or TBP; see Example 1 below). 5 In one embodiment of the invention, the difference (Cttarget - Ctreference), or ACt, may also be exploited by the method known as the 2 -Act method (Livak and Schmittgen 2001; Schmittgen and Livak 2008), with the form: 2 -ACt = 2 - (Ct target - Ct reference) 10 Hence, in one embodiment of the invention, the levels to which each of said selected genes is transcribed are measured as follows: - by amplification, of a fragment of the RNAs transcribed by each of said selected genes, for example by reverse transcription and PCR of these RNA fragments in order to obtain the Ct values for each of these RNAs, 15 - optionally, by normalisation of each of these Ct values with respect to the value for Ct obtained for the RNA of an endogenous control gene, such as RPLP0 or TBP, for example by the 2 -ACt method, - optionally, by Box-Cox transformation of said normalized values for Ct. 20 In the case of measuring the level of expression of a gene by measuring proteins expressed by that gene, i.e. in the case of measuring a level of translation of that gene, the measurement is generally carried out by an immunometric method using specific antibodies, and by expression of the measurements made thereby in quantities by weight or international units using a standard curve. Examples of specific antibodies are indicated 25 in Table 27 below. Examples of protein measurement means are given in Table 16 below. A value for the measurement of the level of translation of a gene may, for example, be expressed as the quantity of this protein per volume of biological fluid, for example per volume of serum (in mg/mL or in pg/mL or in ng/mL or in pg/mL, for example). 30 If desired or required, the distribution of the measurement values obtained for the individuals of a cohort may be smoothed so that it approaches a Gaussian law.

PCT/EP2012/052232 30 To this end, the measurement values obtained for individuals of that cohort, for example the values obtained by the 2 -" method, may be transformed by a transformation of the Box-Cox type (Box and Cox, 1964; see Tables 6 and 25 below; see Example 2 below). 5 Thus, the application relates to an in vitro method for predicting whether a subject infected with one or more HCVs has a strong probability of being a responder to an anti-HCV treatment which will comprise the administration of interferon and ribavirin or whether, in contrast, that subject has a strong probability of not being a responder to this anti-HCV treatment, 10 said method comprising the following steps: i) making the following measurements: - in a sample which has previously been obtained from said subject, measuring the levels to which selected genes have been transcribed or translated, said selected genes being: 15 - at least two genes from among HERC5, IL8 and STAT2, and - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STATIC, STMN2 and USP 18, 20 - optionally, assaying or determining, for said subject, the value of one or more clinical factors and/or of one or more virological factors and/or of one or more biological factors other than the levels of expression of genes selected from HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, 25 OCLN, PLSCR1, RSAD2, STAT 1, STMN2 and USP18, ii) comparing the values for the measurements obtained for said subject in step i) with their values, or the distribution of their values, in reference cohorts which have been pre-established as a function of their status of responder or non-responder to anti-HCV treatment, in order to classify said subject into that of those reference 30 cohorts with respect to which it has the highest probability of belonging. The comparison of step ii) may in particular be made by combining the measurement (or assay) values obtained for said subject in a multivariate classification model.

PCT/EP2012/052232 31 Such a multivariate classification model compares (in a combined manner) values of measurements obtained for said subject with their values or with the distribution of their values in reference cohorts which have been pre-established as a function of their status of responder or non-responder to anti-HCV treatment, in order to classify said subject into 5 that of those reference cohorts with respect to which it has the strongest probability of belonging, for example by attributing to it an output value which indicates the status of responder or non-responder of said subject. Such a multivariate classification model may be constructed, in particular constructed in 10 advance, by making an inter-cohort comparison of the values of measurements obtained for said reference cohorts or of distributions of those measurement values. More particularly, such a multivariate classification model may be constructed, in particular constructed in advance, by measuring or assaying the levels of expression of said genes 15 selected from reference cohorts pre-established as a function of their status of responder or non-responder to anti-HCV treatment, and by analysing these measurement values or their distribution using a multivariate statistical method in order to construct a multivariate classification model which infers or determines a status of responder or non-responder to anti-HCV treatment from the values for the levels of expression of said selected genes. 20 If in addition to values for the measurement of the levels of transcription or translation of said selected genes, the values measured for said subject comprise the value or values for one or more other factors, such as one or more virological factors and/or one or more clinical factors and/or one or more other biological factors (see below and in the 25 Examples), the classification model is of course constructed, in particular constructed in advance, by measuring or assaying the same values in reference cohorts which have been pre-established as a function of whether they have the status of responder or non-responder to anti-HCV treatment, and by analysing these values or their distribution by means of a multivariate statistical method in order to construct a multivariate classification model which 30 infers or determines a status of responder or non-responder to anti-HCV treatment from these values. As an example, a model may be constructed by a mathematical function, a non-parametric technique, a heuristic classification procedure or a probabilistic predictive approach. A PCT/EP2012/052232 32 typical example of classification based on the quantification of the level of expression of biomarkers consists of distinguishing between "healthy" and "sick" subjects. The formalization of this problem consists of m independent samples, described by n random variables. Each individual i (i 1,..., m) is characterized by a vector xi describing the n 5 characteristic values: xij, i=1,...m j=1,...n These characteristic values may, for example, represent gene expression values and/or the intensities of protein data and/or the intensities of metabolic data and/or clinical data. 10 Each sample xi is associated with a discreet value yi, representing the clinical status of the individual i. By way of example, yi = 0 if the patient i has a status of non-responder to anti-HCV treatment, yi = 1 if the patient i has a status of responder to anti-HCV treatment. A model offers a decision rule (for example a mathematical function, an algorithm or a 15 procedure) which uses the information available from xi to predict yj in each sample observed. The aim is to use this model in order to predict the clinical status of a patient p, namely yp, from available biological and/or clinical values, namely xp. A variety of multivariate classification models is known to the skilled person (see Hastie, 20 Tibishirani and Friedman, 2009; Falissard, 2005; Theodoridis and Koutroumbos 2009). They are generally constructed by processing and interpreting data by means, for example, of: - a multivariate statistical analysis method, for example: 25 o a linear or non-linear mathematical function, in particular a linear mathematical function such as a function generated by the mROC method (multivariate ROC method), or o a ROC (Receiver Operating Characteristics) method; o a linear or non-linear regression method, such as the logistical regression 30 method, for example; o a PLS-DA (Partial Least Squares - Discriminant Analysis) method; a a LDA (Linear Discriminant Analysis) method; PCT/EP2012/052232 33 - a machine learning or artificial intelligence method, for example a machine learning or artificial intelligence algorithm, a non-parametric, or heuristic, classification method or a probabilistic predictive method such as: o a decision tree; or 5 o a boosting type method based on binary classifiers (example: Adaboost) or a method linked to boosting (bagging); or o a k-nearest neighbours (or KNN) method, or more generally the weighted k-nearest neighbours method (or WKNN), or o a Support Vector Machine (or SVM) method (for example an algorithm); 10 or o a Random Forest (or RF); or o a Bayesian network; or o a Neural Network; or o a Galois lattice or Formal Concept Analysis. 15 The decision rules for the multivariate classification models may, for example, be based on a mathematical formula of the type y = f(xI,x 2 ,...xn) where f is a linear or non-linear mathematical function (logistic regression, mROC, for example), or on a machine learning or artificial intelligence algorithm the characteristics of which consist of a series of control 20 parameters identified as being the most effective for the discrimination of subjects (for example, KNN, WKNN, SVM, RF). The multivariate ROC method (mROC) is a generalisation of the ROC (Receiver Operating Characteristic) method (see Reiser and Faraggi 1997; Su and Liu 1993, Shapiro, 25 1999). It calculates the area under the ROC curve (AUC) relative to a linear combination of biomarkers and/or biomarker transformations (in the case of normalization), assuming a multivariate normal distribution. The mROC method has been described in particular by Kramar et al. 1999 and Kramar et al. 2001. Reference is also made to the examples below, in particular point 2 of Example 1 below (mROC model). 30 The mROC version 1.0 software, commercially available from the designers (A. Kramar, A. Fortune, D. Farragi and B. Reiser) may, for example, be used to construct a mROC model.

PCT/EP2012/052232 34 Andrew Kramar and Antoine Fortune can be contacted at or via the Unit6 de Biostatistique du Centre R6gional de Lutte contre le Cancer (CRLC) [Biostatistics Unit, Regional Cancer Fighting Centre], Val d'Aurelle - Paul Lamarque (208, rue des Apothicaires; Parc Eurom6decine; 34298 Montpellier Cedex 5; France). 5 David Faraggi and Benjamin Reiser can be contacted at or via the Department of Statistics, University of Haifa (Mount Carmel; Haifa 31905; Israel). The family of artificial intelligence or machine learning methods is a family of algorithms 10 which, instead of proceeding to an explicit generalization, compares the examples of a new problem with examples considered to be training examples and which have been stored in the memory. These algorithms directly construct hypotheses from the training examples themselves. A simple example of this type of algorithm is the k-nearest neighbours (or KNN) model and one of its possible extensions, known as the weighted k nearest 15 neighbours (or WKNN) algorithm (Hechenbichler and Schliep, 2004). In the context of the classification of a new observation x, the simple basic idea is to make the nearest neighbours of this observation count. The class (or clinical status) of x is determined as a function of the major class from among the k nearest neighbours of the 20 observation x. Libraries of specific KKNN functions are available, for example, from R software (http://www. R-project.org/). R software was initially developed by John Chambers and Bell Laboratories (see Chambers 2008). The current version of this software suite is 25 version 2.11.1. The source code is freely available under the terms of the "Free Software Foundation's GNU" public licence at the website http://www. R-project.org/. This software may be used to construct a WKNN model. Reference is also made to the examples below, in particular to point 2 of Example 1 below 30 (WKNN model). A Random Forest (or RF) model is constituted by a set of simple tree predictors each being susceptible of producing a response when it is presented with a sub-set of predictors PCT/EP2012/052232 35 (Breiman 2001; Liaw and Wiener 2002). The calculations are made with R software. This software may be used to construct RF models. Reference is also made to the examples below, in particular to point 2 of Example 1 below 5 (RF model). A neural network is constituted by an orientated weighted graph the nodes of which symbolize neurons. The network is constructed from examples of each class (for example F2 versus F1) and is then used to determine to which class a new element belongs; see 10 Intrator and Intrator 1993, Riedmiller and Braun 1993, Riedmiller 1994, Anastasiadis et al. 2005; see http://cran.r-project.org/web/packages/neuralnet/index.html. R software, which is freely available from http://www.r-project.org/, (version 1.3 of Neuralnet, written by Stefan Fritsch and Frauke Guenther following the work by Marc Suling) may, for example, be used to construct a neural network. 15 Reference is also made to the examples below, in particular to point 2 of Example 1 below (NN model). The comparison of said step ii) may thus in particular be carried out by using the 20 following method and/or by using the following algorithm or software: - mROC, - KNN, WKNN, more particularly WKNN, - RF, or

-

NN, 25 more particularly mROC. Each of these algorithms, or software or methods, may be used to construct a multivariate classification model from values for measurements of each of said reference cohorts, and to combine the values of the measurements obtained for said subject in this model to 30 deduce therefrom a status of responder or non-responder for the subject. In one embodiment of the invention, the multivariate classification model implemented in the method of the invention is expressed by a mathematical function, which may be linear or non-linear, more particularly a linear function (for example, a mROC model). The PCT/EP2012/052232 36 status of responder or of non-responder of said subject is thus deduced by combining said measurement values obtained for said subject in this mathematical function, in particular a linear or non-linear function, in order to obtain an output value, more particularly a numerical output value, which is an indicator of the status of responder or of non 5 responder of said subject. In one embodiment of the invention, the multivariate classification model implemented in the method of the invention is a learning or artificial intelligence model, a non-parametric classification model or heuristic model or a probabilistic prediction model (for example, a 10 WKNN, RF or NN model). The status of responder or of non-responder of said subject is thus induced by combining said measurement values obtained for said subject in a non parametric classification model or heuristic model or a probabilistic prediction model (for example, a WKNN, RF or NN model) in order to obtain an output value, more particularly an output tag, indicative of the status of responder or of non-responder of said subject. 15 Alternatively or in a complementary manner, said comparison of step ii) may include the fact of comparing the values for the measurements obtained for said subject with at least one reference value which discriminates between a status of responder or of non-responder, in order to classify said subject into the group of responder individuals or into the group of 20 non-responder individuals. As an example, the values for the measurements may be compared to their reference values in: - a sub-population of individuals of the same species as said subject, who are 25 infected with the HCV, to whom the anti-HCV treatment has been administered and who have been shown to be responders to this treatment, and/or - a sub-population of individuals of subjects of the same species as said subject, infected with the HCV, to whom the anti-HCV treatment has been administered, and who have been shown to be non-responders to this treatment, 30 or to a reference value which represents the combination of these reference values. A reference value may, for example, be: PCT/EP2012/052232 37 - the value for the measurement of the level of expression of each of said selected genes in each of the individuals for each of the sub-populations or reference cohorts, or - a positional criterion, for example the mean or median, or a quartile, or the 5 minimum, or the maximum of these values in each of these sub-populations or reference cohorts, or - a combination of these values or means, median, or quartile, or minimum, or maximum. 10 The reference value or values used must be able to allow the status of responder to be distinguished from that of non-responder. It may, for example, concern a decision or prediction threshold established as a function of the distribution of the values for the measurements in each of said sub-populations or 15 cohorts, and as a function of the levels of sensitivity (Se) and specificity (Spe) set by the user (Se = TP / (TP + FN) and Sp = TN / (TN + FP), with TP = number of true positives, FN = number of false negatives, TN = number of true negatives, and FP = number of false positives). This decision or prediction threshold may in particular be an optimal threshold which attributes an equal weight to the sensitivity (Se) and to the specificity 20 (Spe), such as the threshold maximizing Youden's index (J) defined by J = Se + Spe - 1. Alternatively or in a complementary manner, several reference values may be compared. This is the case in particular when the values for the measurements obtained for said subject are compared with their values in each of said sub-populations or reference 25 cohorts, for example with the aid of a machine learning or artificial intelligence classification method. Thus, the comparison of step ii) may, for example, be carried out as follows: - select the levels of sensitivity (Se) and specificity (Spe) to be given to the method, 30 - establish a mathematical function, linear or non-linear, in particular a linear mathematical function (for example, by the mROC method), starting from measurement values for said genes in each of said sub-populations or cohorts, and calculate the decision or prediction threshold associated with this function due to PCT/EP2012/052232 38 the choices of levels of sensitivity (Se) and specificity (Spe) made (for example, by calculating the threshold maximizing Youden's index), - combine the measurement values obtained for said subject into this mathematical function, in order to obtain an output value which, compared with said decision or 5 prediction threshold, can be used to attribute a status of responder or a status of non-responder to said subject, i.e. to classify said subject into that of these sub populations or reference cohorts to which it has the greatest probability of belonging. 10 In particular, the invention is based on the demonstration that, when taken in combination, the levels of expression of: - at least two genes from among HERC5, IL8 and STAT2; and - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OAS1, 15 OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18, are biomarkers which provide a "signature" which is predictive of the status of responder or of non-responder. The skilled person having available a combination of genes described by the invention is 20 in a position to construct a multivariate classification model, in particular a multivariate statistical analysis model (for example a linear or non-linear mathematical function) or a machine learning or artificial intelligence model (for example, a machine learning or artificial intelligence algorithm), with the aid of his general knowledge in the field of statistical techniques and means, in particular in the domain of statistical processing and 25 interpretation of data, more particularly biological data. A multivariate classification model may, for example, be constructed, in particular constructed in advance, as follows: a) for a population of individuals of the same species as said subject, and who 30 are infected with one or more HCVs, determining for each of these individuals whether or not that individual responds to an anti-HCV treatment which comprises the administration of interferon and of ribavirin, and classifying these individuals into distinct sub-populations depending on whether they have been shown to be responders or non-responders to this treatment, thus PCT/EP2012/052232 39 constituting reference cohorts established as a function of the response or non response of these individuals to anti-HCV treatment; b) in at least one sample which has already been obtained from each of said individuals, the nature of which is identical to that of the sample from said 5 subject, carrying out the same measurements as those carried out for said subject in said step i); c) carrying out an inter-cohort comparison of the values of the measurements obtained in step b), or the distribution of these values, in order to construct a multivariate classification model which infers a status of responder to anti-HCV 10 treatment or a status of non-responder to said treatment, starting from the combination of said values of the measurements obtained in step b). If said subject or subjects for whom the status of responder or non-responder is to be determined present this fibrosis due to a particular known chronic hepatic disease, for 15 example due to an infection with a particular HCV genotype, then advantageously, individuals with a comparable clinical situation are used. The individuals are also selected so as to constitute a statistically acceptable cohort having no particular bias, in particular no particular clinical bias. The aim is to construct a multivariate classification model which is as relevant as possible from a statistical point of view. 20 Preferably, the cohorts or sub-populations of individuals comprise as many individuals as possible. If the number of individuals is too low, the comparison or the constructed model might not be sufficiently reliable and generalizable in view of the envisaged medical applications. 25 In particular, cohorts or sub-populations will be selected which each comprise at least 30 individuals, for example at least 40 individuals, preferably at least 50 individuals, more particularly at least 70 individuals, and still more particularly at least 100 individuals. 30 Preferably, a comparable number of individuals is present in each cohort or sub population. As an example, the number of individuals of a cohort or sub-population does not exceed the threshold of 3 times the number of individuals of another cohort, more particularly the threshold of 2.5 times the number of individuals of another cohort.

PCT/EP2012/052232 40 When the statistical analysis carried out uses a mathematical function, such as in the case of a mROC method, for example, the number of individuals required per cohort may optionally be of the order of 20 to 40 individuals per reference cohort. In the case of a machine learning analysis method, such as a KNN, WKNN, RF or NN method, it is 5 preferable to have at least 30 individuals per cohort, preferably at least 70 individuals, still more particularly at least 100 individuals. In the examples that follow, the total number of individuals included in the set of cohorts is more than 120. The individuals who make up the reference cohorts are individuals who have received an 10 anti-HCV treatment and for whom the status of responder or non-responder has been determined after application of that treatment, in particular by measuring the HCV load of these individuals at the end of treatment and if this load has become undetectable, 6 months after treatment. 15 In order to determine the status of responder or of non-responder of an individual, and consequently of attributing that individual to a reference cohort, the skilled person can employ any means that is judged appropriate. The VERSANT® HCV-RNA 3.0 (bDNA) ASSAY HCV RNA quantification test from Siemens Healthcare Diagnostics (quantification limit = 615 - 7 690 000 IU/mL) is an example of means that 20 can be used to measure the viral load and to determine whether that load has become undetectable at the end of the treatment and remains so 6 months after treatment (responder individuals) or whether said load is still detectable at the end of treatment (non responder individuals). 25 Although the number of samples taken from a given individual should of course be limited, in particular in the case of hepatic biopsy puncture, several samples can be collected from the same individual. In this case, the results of measuring the various samples of the same individual are considered as their resultant mean; it is not assumed that they could be equivalent to the measurement values obtained from distinct individuals. 30 The comparison of the values of the measurements in each of said cohorts may be carried out using any means known to the skilled person. It is generally carried out by statistical treatment and interpretation of those values. This multivariate statistical comparison can be PCT/EP2012/052232 41 used to construct a multivariate classification model which infers a value for the status of responder or non-responder from the combination of these values. Once said multivariate classification model has been constructed, it can be used to analyse 5 the values of measurements obtained for said subject, and above all be re-used for the analysis of the values of measurements from other subjects. Thus, said multivariate classification model can be established independently of measurements made for said subject or said subjects and may be constructed in advance. 10 Should it be necessary, rather than constitute the cohorts and combine the data from the individuals who make them up, in order to construct examples of multivariate classification models in accordance with the invention, the skilled person may use subjects who are described in the Examples section below as individuals of the cohorts and may, in the context of individual cohort data, use the data which are presented for these subjects in the Examples 15 below, more particularly in Tables 12 to 15 below. Preferably, said multivariate classification model is a particularly discriminating system. Advantageously, said multivariate classification model has a particular area under the ROC curve (or AUC) and/or LOOCV error value. 20 The acronym "AUC" denotes the Area Under the Curve, and ROC denotes the Receiver Operating Characteristic. The acronym "LOOCV" denotes Leave-One-Out-Cross Validation, see Hastie, Tibishirani and Friedman, 2009. 25 The characteristic of AUC is that it can be applied in particular to multivariate classification models which are defined by a mathematical function such as, for example, the models using a mROC classification method. Multivariate artificial intelligence or machine learning models cannot properly be said to 30 be defined by a mathematical function. Nevertheless, since they involve a decision threshold, they can be understood by means of a ROC curve, and thus by an AUC calculation. This is the case, for example, with models using a RF (random forest) method. In fact, in the case of the RF method, a ROC curve may be calculated from predictions of OOB (out-of-bag) samples.

PCT/EP2012/052232 42 In contrast, those of the multivariate artificial intelligence or machine learning models which could not be characterized by an AUC value, in common with all other multivariate artificial intelligence or machine learning models, can be characterized by the value for the 5 "classification error" parameter which is associated with them, such as the value of the LOOCV error, for example. Said particular value for the AUC may in particular be at least 0.84, or at least 0.85 preferably, with a 95% confidence interval of at most ± 11%, more particularly of less 10 than ± 10.5%, still more particularly of less than ± 9.5%, in particular of less than ±8.5%); see for example, combination Nos. 1 to 30 in Tables 7 and 26 below. Said particular AUC value may be at least 0.86, or at least 0.87, or at least 0.88, or at least 0.89, or at least 0.90 (preferably, with a 95% confidence interval of at most ± 11%, more 15 particularly of less than ± 10.5%, still more particularly of less than ± 9.5%, in particular of less than ± 8.5%). This may in particular be the case when one of combination Nos. I to 30 presented in Table 2 below is combined with one or more other biological factors and/or one or more virological factors and/or one or more clinical factors, for example with at least one virological factor and/or at least one other biological factor; see, for example, 20 combination No. 29 combined with two other factors (in fact the viral load before treatment and the concentration of gamma glutamyl transpeptidase) in Table 26 below. Advantageously, said particular LOOCV error value is at most 17%, or at most 16%, or at most 15%, or at most 14% (see for example, combination Nos. I to 30 in Table 3 below). 25 Said particular LOOCV error value may be at most 13%, or at most 12%, or at most 11%, or at most 10%, or at most 9%, or at most 8%, or at most 7 %, or at most 6%, or at most 5%, or at most 4%, or at most 3%, or at most 2%, or at most 1%. This may in particular be the case when one of combination Nos. 1 to 30 presented in Table 2 below is combined 30 with one or more other biological factors and/or one or more virological factors and/or one or more clinical factors, for example with at least one virological factor and/or at least one other biological factor.

PCT/EP2012/052232 43 The diagnostic performances of a biomarker are generally characterized in accordance with at least one of the following two indices: - the sensitivity (Se), which represents its capacity to detect the population termed "pathological" constituted by individuals termed "cases" (in fact, patients who have 5 a status of non-responders); - the specificity (Sp or Spe), which represents its capacity to detect the population termed "healthy", constituted by patients termed "controls" (in fact, patients who have a status of responders). 10 When a biomarker generates continuous values (for example concentration values), different positions of the Prediction Threshold (or PT) may be defined in order to assign a sample to the positive class (positive test: y = 1). The comparison of the concentration of the biomarker with the PT value means that the subject can be classified into the cohort to which it has the highest probability of belonging. 15 As an example, if a cohort of individuals which have a status of responders and a cohort of individuals which have a status of non-responders are considered, and if a subject or patient p is considered for whom the status is to be determined and for whom the value of the combination of measurements is V (V being equal to Z in the case of mROC models), 20 the decision rule is as follows: - when the mean value for the combination of measurements in the cohort of "responder" individuals is less than that of the cohort of "non-responder" individuals: - if V > PT: the test is positive, a status of non-responder is assigned to said patient p, 25 - if V < PT: the test is negative, a status of responder is assigned to said patient p, or - when the mean value of the combination of measurements in the cohort of "responder" individuals is higher than that of the cohort of "non-responder" individuals: 30 - if V > PT: the test is negative, a status of responder is assigned to said patient p, - if V < PT: the test is positive, a status of non-responder is assigned to said patient p.

PCT/EP2012/052232 44 Since the combination of biomarkers of the invention is effectively discriminate, the distributions, which are assumed to be Gaussian, of the combination of biomarkers in each population of interest are clearly differentiated. Thus, the optimal threshold value which will 5 provide this combination of biomarkers with the best diagnostic performances can be defined. In fact, for a given threshold PT, the following values may be calculated: - the number of true positives: TP; 10 - the number of false negatives: FN; - the number of false positives: FP; - the number of true negatives: TN. The calculations of the parameters of sensitivity (Se) and specificity (Sp) are deduced from 15 the following formulae: Se = TP / (TP + FN); Sp = TN /(TN + FP). The sensitivity can thus be considered to be the probability that the test is positive, 20 knowing that the status of the subject is a status of non-responder; and the specificity can be considered to be the probability that the test is negative, knowing that the status of the subject is a status of responder. An ROC curve can be used to visualize the predictive power of the biomarker (or, for the 25 multivariate approach, the predictive power of the combination of biomarkers integrated into the model) for different values of PT (Swets 1988). Each point of the curve represents the sensitivity versus (1-specificity) for a specific PT value. For example, if the concentrations of the biomarker of interest vary from 0 to 35, different 30 PT values may be successively positioned at 0.5; 1; 1.5; ... ; 35. Thus, for each PT value, the test samples are classified, the sensitivity and the specificity are calculated and the resulting points are recorded on a graph.

PCT/EP2012/052232 45 The closer the ROC curve comes to the first diagonal (straight line linking the lower left hand corner to the upper right hand corner), the worse is the discriminating performance of the model. A test with a high discriminating power will occupy the upper left hand portion of the graph. A less discriminating test will be close to the first diagonal of the graph. The area 5 under the ROC curve (AUC) is a good indicator of diagnostic performance. This varies from 0.5 (non-discriminating biomarker) to 1 (completely discriminating biomarker). A value of 0.76 is indicative of a discriminating biomarker. An ROC curve can be approximated by two principal techniques: parametric and non 10 parametric (Shapiro 1999). In the first case, the data are assumed to follow a specific statistical distribution (for example Gaussian) which is then adjusted to the observed data to produce a smoothed ROC curve. Non-parametric approaches consider the estimation of Se and (1-Sp) from observed data. The resulting empirical ROC curve is not a smoothed mathematical function but a step function curve. 15 The choice of threshold or optimal threshold, denoted 6 (delta), depends on the priorities of the user in terms of sensitivity and specificity. In the case where equal weights are attributed to sensitivity and specificity, this latter can be defined as the threshold maximizing the Youden's index (J = Se + Sp - 1). 20 Advantageously, the means of the invention can be used to obtain: - a sensitivity [Se = TP / (TP + FN)] of at least 77% (or more), and/or - a specificity [Sp = TN / (TN + FP)] of at least 80% (or more). 25 In accordance with the invention, the sensitivity may be at least 77%, or at least 78%, or at least 79%, or at least 80%, or at least 81 %, or at least 82%, or at least 83%, or at least 84% (see, for example, combination Nos. 1 to 30 in Tables 3 and 23 below). More particularly, the sensitivity may be at least 80%, at least 81%, or at least 82%, or at 30 least 83%, or at least 84% (see, for example, combination Nos. 1 to 18 in Table 3; see also combination No. 29 combined with other factors (in fact another biological factor and a virological factor) in Table 23 below).

PCT/EP2012/052232 46 Alternatively or in a complementary manner, the specificity may be at least 79%, at least 80%, or at least 81%, or at least 82%, or at least 83%, or at least 84%, or at least 85%, or at least 86%, or at least 87%, or at least 88%, or at least 89%, or at least 90% (see, for example, combination Nos. 1 to 30 in Tables 3 and 23 below). 5 More particularly, the specificity may be at least 84%, or at least 85%, or at least 86%, or at least 87%, or at least 88%, or at least 89%, or at least 90% (see, for example, combination Nos. 1 to 24 in Table 3). 10 All combinations of these sensitivity thresholds and these specificity thresholds are explicitly included in the content of the application (see, for example, combination Nos. 1 to 30 in Tables 3 and 23 below). More particularly, all combinations comprising at least the combination of a sensitivity 15 threshold and a specificity threshold are explicitly included in the content of the application. Alternatively or in a complementary manner to these characteristics of sensitivity and/or specificity, the negative predictive values (NPV) reached or which might be reached by the 20 means of the invention are particularly high. The NPV is equal to TN / (TN+FN), with TN = true negatives and FN = false negatives, and thus represents the probability that the test subject is a responder to anti-HCV treatment, knowing that the test of the invention is negative. 25 In accordance with the invention, the NPV may be at least 80%, or at least 81%, or at least 82%, or at least 83%, or at least 84%, or at least 85%, or at least 86%, or at least 87% (see, for example, combination Nos. 1 to 30 in Tables 3 and 23 below). 30 More particularly, the NPV may be at least at least 84%, or at least 85%, or at least 86%, or at least 87% (see, for example, combination Nos. 2 to 6, 14 to 28 in Table 3; see also combination No. 29 combined with other factors (in fact another biological factor and a virological factor) in Table 23 below).

PCT/EP2012/052232 47 All combinations of NPV thresholds and/or sensitivity thresholds and/or specificity thresholds are explicitly included in the content of the application. More particularly, all combinations comprising at least the combination of a sensitivity 5 threshold and a NPV threshold are explicitly included in the content of the application. Alternatively or in a complementary manner to these characteristics of sensitivity and/or specificity and/or NPV, the positive predictive values (PPV) obtained or which might be obtained by the means of the invention are particularly high. 10 The PPV is equal to TP / (TP + FP) with TP = true positives and FP = false positives, and thus represents the probability that the test subject is a non-responder, knowing that the test of the invention is positive. 15 In accordance with the invention, the PPV may be at least 72%, or at least 73%, or at least 74%, or at least 7 5 %, or at least 76%, or at least 7 7 %, or at least 78%, au at least 79%, or at least 80%, or at least 81%, or at least 82%, or at least 83%, or at least 84%, or at least 85%, or at least 86%, or at least 87%, or at least 88%, or at least 89% (see, for example, combination Nos. 1 to 30 in Tables 3 and 23 below). 20 More particularly, the PPV may be at least 78%, or at least 79%, or at least 80%, or at least 81%, or at least 82%, or at least 83%, or at least 84 %, or at least 85 %, or at least 86%, or at least 87%, or at least 88%, or at least 89% (see, for example, combination Nos. 1 to 24 in Table 3 below). 25 All combinations of PPV and/or NPV thresholds and/or sensitivity thresholds and/or specificity thresholds are explicitly included in the content of the application. More particularly, all combinations comprising at least the combination of a sensitivity 30 threshold and a PPV threshold are explicitly included in the content of the application. More particularly, all combinations comprising at least one of said NPV thresholds and/or at least one of said sensitivity thresholds, more particularly at least one of said NPV thresholds and one of said sensitivity thresholds, more particularly at least one of said PCT/EP2012/052232 48 NPV thresholds and one of said sensitivity thresholds and one of said specificity thresholds are included in the application. The predictive combinations of the invention comprise combinations of levels of gene 5 expression selected as indicated above. As will be indicated in more detail below, and as illustrated in the examples below (see Example 2d) below), it may, however, be possible to elect to involve one or more factors in these combinations other than the levels of expression of these genes, in order to 10 combine this or these other factors and the levels of expression of the selected genes into one decision rule. This or these other factors are preferably selected so as to construct a classification model the predictive power of which is further improved with respect to the model which does not 15 comprise this or these other factors. In addition to the level of expression of said selected genes, it is thus possible to assay or measure one or more other factors, such as one or more clinical factors and/or one or more virological factors and/or one or more biological factors other than the level of expression of 20 said selected genes (see for example, Tables 23 to 26 below, which present therein an example for combination No. 29). The value(s) of this (these) other factors may then be taken into account in order to construct the multivariate classification model and may thus result in still further improved 25 classification performances, more particularly in augmented sensitivity and/or specificity and/or NPV and/or PPV characteristics. As an example, if the values presented for combination No. 29 in Tables 3 and 7 on the one hand, and in Tables 23 and 26 on the other hand (see below) are compared, it will be 30 seen that the values for sensitivity, specificity, NPV, PPV and AUC increase (without, however, reducing the specificity) when the combination of the levels of transcription of said selected genes are also combined with other factors, in particular another biological factor (in fact, ALP) and a virological factor (in fact, VLbeforeTTT).

PCT/EP2012/052232 49 Advantageously, when one or more other factors are combined with a combination of genes selected from said list of twenty-eight genes of the invention, at least one of the AUC characteristics (if appropriate, the LOOCV error), sensitivity, specificity, NPV and PPV, is improved thereby. 5 As indicated above, and as illustrated below, the means of the invention involve measuring the level of expression of: - at least two genes from among HERC5, IL8 and STAT2; and 10 - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT 1, STMN2 and USP 18. Advantageously, the total number of genes selected thereby is 3, 4, 5 or 6, more 15 particularly 3, 4, or 5, for example 4 or 5. The choice of genes is made as a function of the demands or wishes for the performance to be obtained, for example as a function of the sensitivity and/or specificity and/or NPV and/or PPV which is to be obtained or anticipated. Clearly, the lower the number of 20 selected genes, the simpler the means of the invention are to implement. All possible choices of genes are explicitly included in the application. In a manner similar to that indicated above for the sensitivity thresholds, the specificity 25 thresholds, the NPV thresholds, the PPV thresholds and the total number of selected genes, all combinations of genes selected from each of the lists of genes and/or the total numbers of genes selected and/or sensitivity thresholds and/or specificity thresholds and/or NPV thresholds and/or PPV thresholds are explicitly included in the content of the application. 30 Thirty examples of combinations of genes in accordance with the invention are presented in Table 2 below.

PCT/EP2012/052232 50 Tables 3 to 7 below illustrate the performances of the combination of the levels of transcription of four or five genes (in fact, the value of Ct which was measured for the RNA transcripts of that gene and which has been normalized using the 2-Nct method). 5 Tables 23 to 26 below illustrate the performances of the combination of the levels of transcription of five genes selected in accordance with the invention (combination No. 29) when these levels are also combined with other factors (in fact, one other biological factor such as the concentration of alkaline phosphatase, and a virological factor such as the HCV load before treatment). 10 As an example, said genes selected from said list of twenty-eight genes of the invention are: - HERC5, IF144, IL8 and MDK (combination No. 1), or - HERC5, IF144, IL8, MDK and OASI (combination No. 2), or 15 - G1P2, IL8, OCLN, STAT2 and USP18 (combination No. 3), or - CXCL6, IFIT4, IL8, STAT 1 and STAT2 (combination No. 4), or - CXCL6, IL8, MX1, PLSCR1 and STAT2 (combination No. 5), or - CXCL6, IL8, MX1, STATI and STAT2 (combination No. 6), or - HERC5, IF144, IL8, MDK and STMN2 (combination No. 7), or 20 - HERC5, IL8, PLSCR1 and STMN2 (combination No. 8), or - HERC5, IF135, IFITI, IL8 and MX1 (combination No. 9), or - HERC5, IF144, IL8, OASI and RSAD2 (combination No. 10), or - HERC5, IF144, IL8, ITGA2 and MDK (combination No. 11), or - HERC5, IFIT1, IL8 and MX1 (combination No. 12), or 25 - HERC5, IL8, MDK, OAS3 and RSAD2 (combination No. 13), or - CCL21, G1P2, IL8, MDK and STAT2 (combination No. 14), or - G1P2, IFITM1, IL8, OCLN and STAT2 (combination No. 15), or - G1P2IL8, OAS 1, OCLN and STAT2 (combination No. 16), or - CLDN1, IL8, OAS2, OAS3 and STAT2 (combination No. 17), or 30 - CXCL6, IFITM1, IL8, MX1 and STAT2 (combination No. 18), or - CXCL6, IFIT1, IL8 and STAT2 (combination No. 19), or - FOXO1, G1P2, IL8, MDK and STAT2 (combination No. 20), or - CXCL6, G1P2, IL8, MDK and STAT2 (combination No. 21), or PCT/EP2012/052232 51 - CXCL6, IL8, OAS2, STAT 1 and STAT2 (combination No. 22), or - FOXO1, IFI27, IFITM1, IL8 and STAT2 (combination No. 23), or - HERC5, IFI35, IFI44, IL8 and OAS2 (combination No. 24), or - IL8, CCL21, G1P3, HERC5 and RSAD2 (combination No. 25), or 5 - IL8, G1P3, HERC5, OAS3 and RSAD2 (combination No. 26), or - IL8, ITGA2, G1P3, HERC5 and RSAD2 (combination No. 27), or - IL8, STMN2, G1P3, HERC5 and RSAD2 (combination No. 28), or - IL8, CLDN1, G1P3, HERC5 and RSAD2 (combination No. 29), or - IL8, G1P3, HERC5, LGALS3BP and RSAD2 (combination No. 30). 10 In one embodiment of the invention, said at least two genes selected from HERC5, IL8 and STAT2 comprise IL8, more particularly IL8 and HERC5 or IL8 and STAT2. They are in particular combination Nos. 1 to 30 listed above. 15 In accordance with an alternative or complementary embodiment of the invention, the number of genes selected from CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IFI27, IFI35, IFI44, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MIX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STATI, STMN2 and USP18, is 1, 2, 3 or 4, more particularly 1, 2 or 3, for example 2 or 3. 20 In accordance with an alternative or complementary embodiment of the invention, the number of genes selected from CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IFI27, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STATI, STMN2 and USP18, is at least two, and these 25 at least two genes comprise at least two genes from among: CXCL6, IFI44, MDK, MIXI and RSAD2. In accordance with an alternative or complementary embodiment of the invention, said at least one gene selected from CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IFI27, 30 IFI35, IFI44, IFIT1, IFIT4, IFITM1, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP 18, is: CXCL6, IFI44, MDK, MX or RSAD2. This is particularly the case with combination Nos. 1, 2, 4, 5, 6, 7, 9, 10, 11, 13, 14, 18, 19, 20, 21, 22, 24, 25, 26, 27, 28, 29 and 30 listed above.

PCT/EP2012/052232 52 In accordance with an alternative or complementary embodiment of the invention, the number of genes selected from CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, 5 OAS3, OCLN, PLSCR1, RSAD2, STATI, STMN2 and USP18, is at least two, and these at least two genes comprise IF144 and MDK. This is particularly the case with combination Nos. 1, 2, 7 and 11 listed above. In accordance with an alternative or complementary embodiment of the invention, said 10 genes selected in step i) do not comprise CLDN1, G1P3, IF127, IFITMI, ITGA2, OCLN and PLSCR1. This is particularly the case with combination Nos. 1, 2, 4, 6, 7, 9, 10, 12, 13, 14, 19, 20, 21, 22 and 24 listed above. The genes CLDN1, G1P3, IF127, IFITMI, ITGA2, OCLN and PLSCR are those genes from said list of twenty-eight genes of the invention which code for a protein the localization of which is strictly membrane in 15 nature. In accordance with an alternative or complementary embodiment of the invention, said genes selected in accordance with the invention do not comprise any genes which code for a protein the localization of which is strictly membrane in nature. 20 The levels of expression of such genes cannot be measured in a sample of biological fluid from said subject (such as a sample of intracorporal fluid such as blood, serum, plasma or a urine sample). 25 In accordance with an alternative or complementary embodiment of the invention: - the number of genes selected from CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STATIC, STMN2 and USP18 is at least two, and these at least two genes comprise at least two genes from among CXCL6, IF144, 30 MDK, MX1 and RSAD2, more particularly from among IF144, MDK, MX1 and RSAD2, and - said genes selected in step i) do not comprise CLDN1, G1P3, IF127, IFITM1, ITGA2, OCLN or PLSCR1, and/or said genes selected in accordance with the invention do not PCT/EP2012/052232 53 comprise any genes which code for a protein the localization of which is strictly membrane in nature. This is particularly the case with combination Nos. 1, 2, 5, 11, 18, 25, 26, 27, 28, 29 and 5 30 listed above.

PCT/EP2012/052232 54 Table 2: Examples of combinations of levels of expression of four or five genes No. of the combination Selected genes 1 HERC5 IF144 IL8 MDK 2 HERC5 IF144 IL8 MDK OASI 3 G1P2 IL8 OCLN STAT2 USP18 4 CXCL6 IFIT4 IL8 STAT1 STAT2 5 CXCL6 IL8 MX1 PLSCR1 STAT2 6 CXCL6 IL8 MX1 STATIC STAT2 HERC5 IF144 IL8 MDK STMN2 8 HERC5 IL8 PLSCR1 STMN2 9 HERC5 IF135 IFITI IL8 MX1 10 HERC5 IF144 IL8 OAS1 RSAD2 11 HERC5 IF144 IL8 ITGA2 MDK 12 HERC5 IFITI IL8 MX1 13 HERC5 IL8 MDK OAS3 RSAD2 14 CCL21 G1P2 IL8 MDK STAT2 PCT/EP2012/052232 55 Table 2 (continued to end): Examples of combinations of levels of expression of four or five genes No. of the combination Selected genes 15 G1P2 IFITMI IL8 OCLN STAT2 16 G1P2 IL8 OASI OCLN STAT2 17 CLDN1 IL8 OAS2 OAS3 STAT2 18 CXCL6 IFITMI IL8 MX1 STAT2 19 CXCL6 IFITI IL8 STAT2 20 FOXO1 G1P2 IL8 MDK STAT2 21 CXCL6 G1P2 IL8 MDK STAT2 22 CXCL6 IL8 OAS2 STAT1 STAT2 23 FOXO1 IF127 IFITMI IL8 STAT2 24 HERC5 IF135 IF144 IL8 OAS2 25 IL8 CCL21 G1P3 HERC5 RSAD2 26 IL8 G1P3 HERC5 OAS3 RSAD2 27 IL8 ITGA2 G1P3 HERC5 RSAD2 28 IL8 STMN2 G1P3 HERC5 RSAD2 29 IL8 CLDN1 G1P3 HERC5 RSAD2 30 IL8 G1P3 HERC5 LGALS3BP RSAD2 PCT/EP2012/052232 56 Table 3: Values for sensibility (Se), specificity (Spe), negative predictive value (NPV), positive predictive value (PPV) and LOOCV error which may be associated with the combinations of the levels of transcription of four or five genes selected in accordance with the invention (RNA transcripts, more particularly RNA from a sample of tissue or hepatic cells) No. of the combination Classification (see Table 2) model used Se Spe NPV PPV LOOCV Error 1 WKNN 84 86 80 89 15 2 WKNN 82 89 84 88 14 3 RF 82 86 87 80 16 4 RF 82 84 87 78 17 5 RF 82 84 87 78 17 6 RF 82 84 87 78 17 WKNN 80 87 81 86 16 8 WKNN 80 87 81 86 16 9 WKNN 80 86 80 86 17 10 WKNN 80 86 80 86 17 5 LOOCV error = Leave-One-Out Cross Validation (Hastie, Tibishirani and Friedman, 2009); ND = not determined PCT/EP2012/052232 57 Table 3 (continued): Values for sensibility (Se), specificity (Spe), negative predictive value (NPV), positive predictive value (PPV) and LOOCV error which may be associated with the combinations of the levels of transcription (RNA) of four or five genes selected in accordance with the invention (RNA transcripts, more particularly RNA from a sample of tissue or hepatic cells) No. of the combination Classification (see Table 2) model used Se Spe NPV PPV LOOCV Error 11 WKNN 80 86 80 86 17 12 WKNN 80 86 80 86 17 13 WKNN 80 86 80 86 17 14 RF 80 90 86 85 14 15 RF 80 87 86 81 16 16 RF 80 87 86 81 16 17 RF 80 86 86 80 17 18 RF 80 86 86 80 17 19 RF 77 90 85 85 15 20 RF 77 89 85 83 16 5 LOOCV error = Leave-One-Out Cross Validation (Hastie, Tibishirani and Friedman, 2009); ND = not determined PCT/EP2012/052232 58 Table 3 (continued to end): Values for sensibility (Se), specificity (Spe), negative predictive value (NPV), positive predictive value (PPV) and LOOCV error which may be associated with the combinations of the levels of transcription (RNA) of four or five genes selected in accordance with the invention (RNA transcripts, more particularly RNA from a sample of tissue or hepatic cells) No. of the combination Classification (see Table 2) model used Se Spe NPV PPV LOOCV Error 21 RF 77 87 85 81 17 22 RF 77 87 85 81 17 23 RF 77 87 85 81 17 24 RF 77 87 85 81 17 25 mROC 77 81 84 74 ND 26 mROC 77 81 84 74 ND 27 mROC 77 81 84 74 ND 28 mROC 77 81 84 74 ND 29 mROC 77 79 83 72 ND 30 mROC 77 80 83 72 ND 5 LOOCV error = Leave-One-Out Cross Validation (Hastie, Tibishirani and Friedman, 2009); ND = not determined PCT/EP2012/052232 59 Table 4: Examples of control parameters for classification models set up using the WKNN method No. of the combination (see Tables 2 and 3 above) Distance (D) Number of neighbours (k) Kernel (K) 1 1 5 inverse 2 2 4 triangular 7 1 6 triangular 8 1 7 cosine 9 1 14 tri-weighted 10 1 6 cosine 11 2 3 inverse 12 1 10 triangular 13 2 3 inverse PCT/EP2012/052232 60 Table 5: Examples of mROC models (Z function) combining the levels of transcription of four or five selected genes (RNA transcripts, more particularly RNA from a sample of tissue or hepatic cells), and example of PT threshold for these functions (in fact, the threshold maximizing the Youden's index 8) No. of Z function combining the levels of transcription (RNA) of the selected genes Name of PT combination function threshold (see Tables 2 and 3 (S) above) 25 Z = 0.218 x CCL21 t + 0.441 x G1P3 t + 0.98 x HERC5 t + 0.121 x IL8 - 0.482 x RSAD2 t Z25ARN -2.182 26 Z = 0.437 x G1P3 t + 0.99 x HERC5 t + 0.131 x IL8 + 0.151 x OAS3' - 0.522 x RSAD2 t Z26ARN -1.239 27 Z = 0.508 x G1P3 t + 1.036 x HERC5 t + 0.124 x IL8 + 0.084 x ITGA2 - 0.467 x RSAD2 t Z27ARN -1.444 28 Z = 0.511 x G1P3 t + 1.038 x HERC5 t + 0.127 x IL8 - 0.468 x RSAD2 t + 0.017 x STMN2 Z28ARN -1.472 29 Z = 0.134 x CLDNl t + 0.488 x G1P3 t + 0.966 x HERC5 t + 0.129 x IL8 - 0.487 x RSAD2 t Z29ARN -1.281 30 Z = 0.4 x G1P3 t + 1.028 x HERC5 t + 0.13 x IL8 + 0.11 x LGALS3BP t - 0.481 x RSAD2 t Z30ARN -1.594 PCT/EP2012/052232 61 Table 6: List of genes for which it is advised to normalize the measured values (in particular, the measured values for the levels of RNA transcription, more particularly when these RNAs are obtained from a sample of tissue or hepatic cells), for example by a Box-Cox normalisation, and example of values for the Box-Cox parameter (k) which can be used in the Z functions indicated in Table 5 above Genes for which it is Example of values for the advised to normalize the value Box-Cox parameter (k) which can be used for the Z for the level of functions of Table 5 above transcription (RNA) CCL21 0.02 GlP3 0.19 HERC5 0.19 RSAD2 -0.05 OAS3 0.13 CLDN1 0.68 LGALS3BP -0.06 5 PCT/EP2012/052232 62 Table 7: AUC for Z functions of Table 5 No. of combination Name of function AUC AUC, lower limit AUC, upper limit (see Table 2 above) (see Table 5 above) 25 Z25ARN 0.854 0.759 0.916 26 Z26ARN 0.853 0.758 0.915 27 Z27ARN 0.854 0.757 0.916 28 Z28ARN 0.855 0.759 0.917 29 Z29ARN 0.857 0.761 0.919 30 Z30ARN 0.856 0.761 0.917 PCT/EP2012/052232 63 In addition to the levels of expression of said genes selected from the list of seventeen genes of the invention, the means of the invention can also comprise a combination of one or more other factors, such as: - one or more clinical factors, such as: 5 o sex (female, F or male, M), o age at the date of sampling (Age), for example, age at the date of HBP, age at the date of hepatic cytopuncture, age at the date of sampling blood, serum, plasma or urine, o age of patient at the date of contamination, 10 o age of patient at the start of treatment, o body mass index (BMI), o insulin sensitivity index (HOMA), o diabetes, o alcohol consumption, 15 o degree of steatosis, o mode of contamination, o Metavir activity, o hepatic fibrosis score measured using the Metavir system (Metavir F score) or using the Ishak system; 20 and/or - one or more virological factors, such as: o viral genotype, more particularly genotype of the HCV or HCVs, o duration of infection, o viral load before treatment, more particularly HCV load before treatment 25 (VLbeforeTTT), o viral load, more particularly HCV load, measured for the patient at the date of start of treatment (viral load at DO), o viral load, more particularly HCV load, measured for the patient at the date of sampling (viral load at HBP, viral load at the date of hepatic 30 cytopuncture, viral load at the date of sampling blood, serum, plasma or urine); and/or - one or more biological factors other than the levels of expression of said selected genes, which may in particular be selected from the concentrations, contents or PCT/EP2012/052232 64 quantities of intracorporal proteins, concentrations, contents or quantities of intracorporal metabolites, concentrations, contents or quantities of elements occurring in blood, and assays representative of the quantity of circulating iron, such as: o concentration of haptoglobin (Hapto), 5 o concentration of apolipoprotein Al (ApoAl), o total quantity of bilirubin (BLT), o concentration of gamma glutamyl transpeptidase (GGT), o concentration of aspartate aminotransferase (AST), o concentration of alanine aminotransferase (ALT), 10 o platelet count (PLQ), o quantity of prothrombin (TP), o quantity of HDL cholesterol (Chol-HDL), o total quantity of cholesterol, o concentration of ferritin (Ferritin), 15 o level of glycaemia (glycaemia), o concentration of peptide C, o quantity of insulin (insulinaemia), o concentration of triglycerides (TG), o quantity of albumin, 20 o transferrin saturation (TSAT), o concentration of alkaline phosphatase (ALP). Tables 23 to 26 (combination No.29) below illustrate the performances of the combination of the levels of transcription of five genes (in fact, the Ct value which was measured for the RNA transcripts of this gene and which was normalized by the 2 -Act method), also 25 combined with one or more other biological factors and/or one or more virological factors and/or one or more clinical factors (in fact, with one other biological factor such as the concentration of alkaline phosphatase and a virological factor such as the HCV load before treatment). 30 This or these other factors may be assayed for a sample with a nature which differs from that used to assay the levels of expression of said selected genes. As an example, the biological sample for assaying the levels of expression of said genes selected from said list of twenty-eight genes of the invention may be a HBP or hepatic cytopuncture sample, and PCT/EP2012/052232 65 the biological sample for assaying the values of said other factors may be a sample of a biological fluid such as blood, plasma or serum or urine. Similarly, the nature of the assayed level of expression may be different; as an example, to assay the level of expression of said selected genes, it is possible to assay the levels of their transcription into 5 RNA, while for those of said other factors which are biological factors, the assayed level of expression will generally be a protein concentration. The measurement or assay of certain of these factors may sometimes be considered to be assaying the level of translation (measurement of protein concentration) of a gene other 10 than a selected gene of the invention (for example ALP; see Example 2d) below). The number of genes the level of expression of which is measured, and which are not genes selected from said list of twenty-eight genes of the invention (for example ALP) is preferably a maximum of 18, more particularly 14 or fewer, more particularly 11 or fewer, 15 more particularly 6 or fewer, more particularly 4 or 3 or 2 or 1 or 0, more particularly 3 or 2 or 1 or 0, in particular 2 or 1 or 0. Advantageously, this or these other factors are or comprise one or more biological factors, in particular the concentration of alkaline phosphatase (ALP). 20 Example 2d) above provides an illustration of combinations of this type. Alternatively or in a complementary manner, this or these factors may more particularly be or comprise one or more factors from among the following virological factors: 25 o viral load before treatment (VLbeforeTTT); and/or o genotype of the HCV or HCVs. Examples 2d) and 3b) below provide an illustration of such combinations. 30 Alternatively or in a complementary manner, this or these factors may more particularly be or comprise one or more clinical factors, in particular the hepatic fibrosis clinical factor score, which can be assayed using the Metavir system (Metavir F score) or using the Ishak system.

PCT/EP2012/052232 66 In one particular embodiment of the invention, in addition to measuring the levels of expression of genes selected from said list of twenty-eight genes of the invention, the means of the invention may further comprise measuring or assaying the following other factors: 5 - one or more clinical factors which is or comprises the hepatic fibrosis score (which can be assayed using the Metavir system (Metavir F score) or using the Ishak system); and/or - one or more virological factors which is or comprises the genotype of the HCV or HCVs and/or the HCV load before treatment; and/or - one or more biological factors other than the levels of expression of genes selected from 10 CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OASI, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STATI, STMN2 and USP18, which is or comprises the concentration of alkaline phosphatase (ALP). 15 Alternatively or in a complementary manner, in addition to measuring the levels of expression of genes selected from said list of twenty-eight genes of the invention, the means of the invention may further comprise: - determining the hepatic fibrosis score of said subject, more particularly determining whether the hepatic fibrosis score of said subject is a score which, in the Metavir 20 score system, is at most F1 or at least F2, more particularly at least F2; and/or - determining whether the HCV or HCVs which has infected said subject comprises an HCV of genotype 1, 4, 5 or 6, more particularly of genotype 1 or 4, more particularly of genotype 1. 25 These determinations may be made during step i), or be made independently of step i).

PCT/EP2012/052232 67 Table 23: Values for sensibility (Se), specificity (Spe), negative predictive value (NPV), positive predictive value (PPV) and LOOCV error which may be associated with a combination of the levels of transcription of five genes selected in accordance with the invention (RNA transcripts, more particular when these RNAs are obtained from a sample of tissue or hepatic cells), further combined with other biological 5 factors and/or clinical factors and/or virological factors Other biological factors and/or clinical factors and/or No. of the combination virological factors, combined Classification of selected genes Se Spe NPV PPV LOOCV error with the levels of expression model used (see Table 2) of the combination of selected genes viral load before treatment (VLbeforeT TT); 2 9 RNA(§) concent o akain mROC(*) 81 80 85 76 ND concentration of alkaline phosphatase (ALP) (*): the indicated values of Se, Spe, NPV, PPV and LOOCV error are those of the function of Table 24 below more particularly, RNA from a sample of tissue or hepatic cells 10 ND = not determined PCT/EP2012/052232 68 Table 24: Example of mROC model (Z function) for a combination of five genes selected in accordance with the invention (measurement of their levels of transcription into RNA), further combined with other factors (biological factors other than the levels of expression of selected genes in accordance with the invention and/or clinical factors and/or virological factors), and Example of PT threshold for this 5 function (in fact, threshold maximizing Youden's index 5), No. of the combination of Other factors Example of Z function Name of PT selected genes (mROC model) function threshold (see Table 2) (8 threshold) Z = 0.3 x CLDNl t + 0.634 x GIP3' + 1.154 x viral load before HERC5' + 0.114 x IL8 - 0.685 x RSAD2' + treatment (VLbeforeTTT); 0.036 x VLbeforeTT

T

t + 2.139 x PAL' Z29ARNsupp 4.687 concentration of alkaline phosphatase (ALP) more particularly, RNA from a sample of tissue or hepatic cells PCT/EP2012/052232 69 Table 25: Example of values for the Box-Cox parameter (k) which can be used in the Z function indicated in Table 23 above Genes for which it is advised to normalize the value for the level of Example of value of the parameter lambda transcription (RNA) or factors for which can be used for the Z functions of Table which it is advised to normalize the 24 above (*) value CLDN1 0.63 G1P3 0.21 HERC5 0.15 RSAD2 -0.02 VLbeforeTTT 0.2 ALP -0.26 (*):lambda, the parameter for Box-Cox transformations [BMQt = (BMQ'-1)/X] Table 26: AUC value for the function of Table 24 Name of function AUC AUC, lower AUC, upper limit limit Z29ARNsupp 0.904 0.828 0.948 (see Table 24) PCT/EP2012/052232 70 In accordance with a complementary aspect of the invention, the application relates to products or reagents for the detection and/or determination and/or measurement of said assays, more particularly for the detection and/or measurement of the levels of expression of said selected genes, and to manufactured articles, compositions, pharmaceutical 5 compositions, kits, tubes or solid supports comprising such reagents, as well as to computer systems (in particular, computer program product and computer device), which are specially adapted to carrying out a method of the invention. The application is in particular relative to a reagent which specifically detects a 10 transcription product (RNA) of one of said genes selected from said list of twenty-eight genes of the invention, or a translation product of one of said genes selected from said list of twenty-eight genes of the invention (protein, or post-translational form of this protein, such as a specific fragment of this protein). 15 The application is in particular relative to reagents which specifically detect each of the transcription products (RNA) of said genes selected from said list of twenty-eight genes of the invention, or each of the translation products of said genes selected from said list of twenty-eight genes of the invention (protein, or post-translational form of this protein, as a specific fragment of this protein). 20 Advantageously, a set of such reagents is formed, each of the reagents detecting said transcription products of said selected genes and/or each of the reagents detecting said translation products of said genes selected from said list of twenty-eight genes of the invention, i.e. a set of reagents which specifically detects at least one expression product 25 for each of these genes. Preferably, said reagents not only specifically detect a transcription or translation product, but can also quantify it. 30 In particular, the application pertains to a manufactured article comprising said reagents as a combination product (or combined form, or combined preparation), in particular for their simultaneous, separate or sequential use. This manufactured article may, for example, be in the form of a set of reagents, or a kit.

PCT/EP2012/052232 71 Clearly, the characteristics of combinations of selected genes described above and those illustrated below are applicable to the reagents of the invention mutatis mutandis. Said reagents may, for example, hybridize specifically to the RNA of said selected genes 5 and/or to the cDNA corresponding to these RNAs (under at least stringent hybridization conditions), or bind specifically to proteins encoded by said selected genes (or to specific fragments of these proteins), for example in an antigen-antibody type reaction. At least stringent hybridization conditions are known to the skilled person. The conditions 10 may, for example, be as follows: - for filter hybridization: in 5xSSC, 2% sodium dodecyl sulphate (SDS), 100 micrograms/mL single strand DNA at 55-65'C for 8 hours, and washing in 0.2xSSC and 0.2% SDS at 60-65'C for thirty minutes; - for a hybridization by PCR: the PCR conditions indicated in Example 1 below. 15 Said reagents of the invention may in particular be: - nucleic acids (DNA, RNA, mRNA, cDNA), including oligonucleotide aptamers, optionally tagged to allow them to be detected, in particular with fluorescent tags which are well known to the skilled person, or - protein ligands such as proteins, polypeptides or peptides, for example aptamers, 20 and/or antibodies or fragments of antibodies. The nucleic acids of the invention may, for example, be primers and/or probes (see SEQ ID NO: I to 56 in Table 10 below), in particular pairs of primers (see the pairs of primers indicated in Table 10 below). For each of said genes selected from said list of twenty-eight 25 genes of the invention, the skilled person can construct a pair of primers and/or a probe which specifically hybridizes to this gene. A manufactured article of the invention may thus comprise the number of primers and/or probes necessary for the detection of the RNA or cDNA of each of said selected genes. 30 The sequence of nucleic acids of the invention may, for example, be constituted by 9 to 40 nucleotides, more particularly 10 to 30 nucleotides, more particularly 14 to 29 nucleotides, more particularly 19 to 24 nucleotides.

PCT/EP2012/052232 72 The primer sequences of one pair may, for example, be the sequences of a fragment of the sequence of one of said selected genes and a fragment of its complementary sequence (see Table 1 indicating the accession numbers of the sequences for these genes). One and/or the other of these two primer sequences might not be strictly identical to the sequence of a 5 gene fragment or its complementary sequence; one and/or the other of these two primer sequences may: - be derived from one or more nucleotide substitutions and/or additions and/or deletions, more particularly one or more nucleotide substitutions, and/or have a sequence identity of at least 80%, or at least 85%, or at least 90%, or at least 95% 10 with the sequence for this fragment or its complementary sequence (identity calculated over the longest of the two aligned sequences - optimal alignment), - provided that the resulting pair of primers has conserved the capacity to specifically hybridize to one of said selected genes. 15 A primer pair of the invention advantageously has a delta Tm of approximately 1 C or less. In one embodiment of the invention, a primer pair of the invention targets an approximately 70 to 120 bp amplicon (i.e. the sense primer and the anti-sense primer hybridize at such positions on the target nucleic acid that the amplicon produced by elongation of these hybridized primers has a length of approximately 70 to 120 bp). 20 Examples of such primers and primer pairs are presented in Table 10 below (SEQ ID NO: 1 to 56, forming 28 primer pairs). The sequence for a probe of the invention may, for example, be: 25 - the sequence for a fragment of the sequence of one of said selected genes (see Table 1 indicating the accession numbers for sequences for these genes), said fragment hybridizing specifically to the sequence for that gene; - a sequence: o which derives from the sequence for such a fragment by one or more 30 nucleotide substitutions and/or additions and/or deletions, more particularly by one or more nucleotide substitutions, and/or a sequence which has a sequence identity of at least 80%, or at least 85%, or at least 90%, or at least 95% with the sequence for this fragment or its complementary PCT/EP2012/052232 73 sequence (identity calculated for the longest of the two aligned sequences optimal alignment), but o which has conserved the capacity to hybridize specifically to one of said selected genes; 5 and/or - a complementary sequence of such sequences. A probe of the invention may in particular be a probe for real time amplification, intended for use with a primer pair in accordance with the invention. Alternatively, detection by 10 real time PCR may use molecules known as intercalating (for example; SYB green) which have the ability of interposing themselves into double stranded structures. The ligands of the invention, which bind specifically to proteins encoded by the genes selected from said list of twenty-eight genes of the invention (or to specific fragments of these proteins) may, for example, be proteins, polypeptides or peptides, for example aptamers or antibodies or 15 antibody fragments. The skilled person can produce such a ligand for each of said selected genes. The antibodies may, for example, be produced by immunization of a non-human mammal 20 (such as a rabbit) with a protein encoded by said selected gene or with an antigenic fragment of such a protein, optionally associated or coupled with an immunization adjuvant (such as a Freund's adjuvant or KLH - keyhole limpet haemocyanin), for example by intraperitoneal or subcutaneous injection, and by collecting the antibodies obtained thereby in the serum of said mammal. 25 Monoclonal antibodies may be produced using a lymphocyte hybridization technique (hybridomas), for example using the technique by K6hler and Milstein 1975 (see also US 4 376 110), the human B cell hybridoma technique (Kosbor et al. 1983; Cole et al. 1983), or the technique for immortalizing lymphocytes with the aid of the Epstein-Barr virus 30 EBV- (Cole et al. 1985). Examples of such antibodies are IgG, IgM, IgE, IgA, IgD or any sub-class of these immunoglobulins.

PCT/EP2012/052232 74 Antibodies modified by genetic engineering may be produced, such as recombinant antibodies or chimeras, humanized by grafting one or more CDRs, (Complementary Determining Region). 5 The antibodies used in the invention may be fragments of antibodies or artificial derivatives of such fragments, provided that these fragments or derivatives have said specific binding property. Such fragments may, for example, be Fab, F(ab')2, Fv, Fab/c or scFv (single chain fragment variable) fragments. 10 Examples of antibodies are given in Table 27 below.

PCT/EP2012/052232 75 Table 27: Examples of specific antibodies Encoding Antibody Example of Catalogue reference gene supplier of product HERC5 anti-HECT E3 ubiquitin ligase Abcam ab83853 antibody STAT2 Human anti-STAT2 MAb (Clone R&D Systems MAB1666 545117), mouse IgGI IFI35 anti-IFI35 MAb produced in Sigma WH0003430M1 mouse IF144 IF144 antibody Abcam ab68503 IFITI anti-IFITI (1-15) antibody Sigma SAB1100107 produced in rabbit IFIT4 IFIT3 antibody Abcam ab76818 Mouse anti-integrin alpha 2 ITGA2 biotinylated, affinity purified R&D Systems BAF1740 PAb, sheep IgG MX1 anti-Mx1 antibody Abcam ab57854 MAb = monoclonal antibody PAb = polyclonal antibody PCT/EP2012/052232 76 Table 27 (continued): Examples of specific antibodies Encoding Example of Catalogue geneAntiodysuppier reference of product OAS1 anti-OASI (Abl) antibody Sigma SAB2101667 produced in rabbit OAS3 anti-OAS3 (C-term) antibody Sigma SAB1300335 produced in rabbit OCLN anti-occludin antibody Abcam ab31721 PLSCR1 anti-PLSCR1 (6-20) antibody Sigma SABI 100402 produced in rabbit RSAD2 anti-viperin (anti-RSAD2) Abcam ab73864 antibody STAT1 human anti-STATI MAb (Clone R&D Systems MAB14901 655210), mouse IgG2B STMN2 anti-STMN2 MAb produced in Sigma WHOO11075M2 mouse USP18 anti-UBP43 antibody Abcam ab80815 MAb = monoclonal antibody PAb = polyclonal antibody PCT/EP2012/052232 77 Table 27 (continued to end): Examples of specific antibodies Encoding Antibody Example of rCatalogue gene Anioysupplier rfrneo product CCL21 anti-CCL21 antibody Abcam ab9851 CLDN1 anti-CLDN1 (MO1) MAb, clone Abnova H00009076-MO1 1C5-D9 FOXO1 human anti-FoxOl/FKHR PAb' R&D Systems AF5939 affinity purified, sheep IgG G1P2 anti-ISG15 antibody Abcam ab38639 G1P3 anti-G1P3 MAb, clone M2 Abnova H00002537-M02 IF127 anti-IF127 antibody Abcam ab41538 LGALS3BP mouse anti-LGALS3BP PAb Abnova H00003959-BO1 MaxPab (BOlI) OAS2 mouse anti-human OAS2 MAb R&D Systems MAB1925 (Clone HLS56/3), MAb = monoclonal antibody PAb = polyclonal antibody 78 PCT/EP2012/052232 Other examples of means for measuring the levels of transcription of selected genes are also presented in Table 16 below (immunoassay kits). Said reagents may also comprise a tag for their detection (for example a fluorophore). 5 Said reagents may be in the form of composition(s), pharmaceutical composition(s), for example in one or more tube(s) or in (a) well(s) of a nucleic acid amplification plate. Said reagents may be as a mixture or in distinct forms or physically separated from each 10 other. Said reagents may be fixed to a solid support, for example a support formed from a polymer, from plastic, in particular polystyrene, from glass or from silicon. 15 Said reagents may be directly or indirectly attached to said solid support, for example via a binding agent or capture agent which is attached to the solid support. This binding or capture agent may comprise a portion fixed to said solid support and a portion which comprises a ligand which binds specifically to one of said selected genes. Such a ligand may, for example, be an antibody, a monoclonal antibody, in particular a human antibody 20 such as a IgG, IgM or IgA, or a fragment of an antibody of this type which has conserved the binding specificity. Said solid support may, for example, be a plastic plate, in particular formed from polystyrene, comprising a plurality of analytical wells, such as a protein titre or microtitre 25 plate, for example an ELISA plate. Said solid support may also be formed by magnetic or non-magnetic microbeads, for microtitration, for example using the technique described by Luminex. 30 Said solid support may, for example, be a nucleic acid, protein or peptide chip, for example a plastic, glass or silicon chip. Said reagents do not have to be fixed to a solid support and may, for example, be contained in a solution such as a buffer, for example to store them until use. More 79 PCT/EP2012/052232 particularly, the reagents may be nucleic acids which are not bound to a solid support the nucleotide sequence of which is adapted to specific amplification (the case of primers or primer pairs) and/or to specific hybridization (in the case of probes) of the transcription product (RNA) of one of said genes selected from said list of twenty-eight genes of the 5 invention. In addition to reagents which detect the transcription or translation products of mammalian genes, more particularly human genes, and in particular genes selected from said list of twenty-eight genes of the invention, a manufactured article in accordance with the 10 application may optionally comprise other reagents, for example reagents that can be used to measure or determine one or more virological factors and/or one or more clinical factors. As an example, an article manufactured in accordance with the application may comprise reagents which specifically detect one or more hepatitis viruses, and/or its or their 15 genotype. In one embodiment, the application pertains to a manufactured article comprising reagents in a combined preparation for their simultaneous, separate or sequential use, said reagents being constituted by: 20 - reagents which specifically detect (preferably, which specifically detect and can be used for quantification) each of the transcription or translation products of 3 to 46 mammalian genes, more particularly 3 to 46 human genes, (for example, by specifically hybridizing to the RNA of these genes and/or to the cDNA obtained by reverse transcription of these RNA, or by specifically binding to proteins encoded by these genes), said 3 to 46 25 mammalian genes, or, if appropriate, said 3 to 46 human genes, comprising said genes selected from said list of twenty-eight genes of the invention, and - optionally, reagents which specifically detect (preferably which specifically detect and can be used for quantification) a hepatitis virus and/or the genotype of a hepatitis virus. 30 In this manufactured article, the number of mammalian genes, more particularly human genes the transcription or translation products of which may be detected is 3 to 36, more particularly 3 to 33, more particularly 3 to 28, more particularly 3 to 26, more particularly 3 to 25, more particularly 3 to 24, more particularly 3 to 23, more particularly 3 to 22, 80 PCT/EP2012/052232 more particularly 3 to 20, more particularly 3 to 19, more particularly 3 to 10, more particularly 3 to 9, more particularly 3 to 8, more particularly 3 to 7 (for example 4, 5, 6 or 7), more particularly 3 to 6 (for example 4, 5 or 6), more particularly 3 to 5 (for example, 3, 4 or 5). 5 The mammalian genes, more particularly the human genes, the transcription or translation products of which may be detected by the reagents contained in the manufactured article of the application comprise said genes selected from said list of twenty-eight genes of the invention, and optionally other genes, which are not the genes selected from said list of 10 twenty-eight genes of the invention, but for which the expression product, more particularly of translation, may be of interest, such as the genes listed here as "other biological factors" (for example, the gene coding for alkaline phosphatase or ALP). In the manufactured article of the application, the number of reagents which specifically 15 detect the expression product of mammalian genes (more particularly human genes) which are not genes selected from said list of twenty-eight genes of the invention (for example a reagent specifically detecting ALP) is preferably a maximum of 5, more particularly 4 or fewer, more particularly 3 or fewer, more particularly 2 or fewer, more particularly 2 or 1 or 0. 20 Said manufactured article may thus, for example, be: - one or more tubes, - a kit, in particular a kit comprising one or more tubes, - a solid support, for example, formed from plastic, polystyrene, glass, silicon or 25 polymer or comprising a magnetic material such as iron oxide, such as: o a plate formed from plastic comprising a plurality of analysis wells, such as - a nucleic acid amplification plate comprising wells for receiving a biological sample and a reaction mixture for nucleic acid amplification, 30 - a titration or microtitration plate, more particularly an ELISA plate, o magnetic microbeads (for example microbeads formed from iron oxide and coated with a polymer to which the proteins or polypeptides can adhere or be attached by chemical coupling); " a nucleic acid, protein, polypeptide or peptide chip.

81 PCT/EP2012/052232 Optionally, the manufactured article of the invention further comprises instructions (for example, an instruction sheet) for measuring the level of expression of said selected genes on a biological sample collected or obtained from said subject, more particularly to carry out 5 a method of the invention. Said manufactured article may further comprise one or more of the following elements: - an instrument for removing said sample, in particular: o a needle and/or a syringe, more particularly a needle and/or a syringe for 10 taking a sample of an intracorporal liquid such as blood, and/or o a needle adapted for hepatic cytopuncture, for example a needle with a diameter of 18 to 22G), and/or o a needle and/or a catheter and/or a biopsy gun adapted for HBP; - a computer program product or software product, in particular a computer program 15 product or statistical analysis software, for example a computer program product of the invention as described below; - RNA extraction reagents; - a reverse transcriptase; - a polymerase, for example a Taq polymerase; 20 - nucleotides (dNTP). In particular, the application pertains to said manufactured article or to said reagents for their use in a method for predicting whether a subject infected with one or more HCVs has a high probability of responding to an anti-HCV treatment which is to comprise administering 25 interferon and ribavirin (or their prodrugs) or whether, in contrast, that subject has a high probability of not responding to that anti-HCV treatment, more particularly to said manufactured article or to said reagents for their use in a predictive method of the invention. In particular, this use may comprise: 30 - taking a biological sample from said subject, in particular by inserting a needle or catheter into the body of said subject, and - using said reagents in said method on this biological sample, or on a sample comprising nucleic acids and/or proteins and/or polypeptides and/or peptides extracted or purified 82 PCT/EP2012/052232 from said biological sample, or on a sample comprising cDNAs which are susceptible of having been obtained by reverse transcription of said nucleic acids. This use may, for example, comprise: 5 - taking a biological sample of said subject, optionally transformed by: o extraction or purification of RNAs of said removed sample and optionally by reverse transcription of the extracted RNAs, or by o extraction or purification of its proteins from said sample, and - using said reagents of the invention on this optionally transformed biological 10 sample. Said biological sample may be taken by inserting a sampling instrument, in particular by inserting a needle or a catheter, into the body of said subject. 15 The sampling instrument is primarily inserted in order to remove intracorporal fluid from said subject (such as blood, for example) and/or a portion of hepatic tissue from said subject (for example by HBP) and/or hepatic cells from said subject (for example by hepatic cytopuncture). 20 This instrument may thus be inserted, for example: - into a vein, an artery or a blood vessel of said subject to remove blood from said subject; and/or - into the liver of said subject, in order to take a sample of hepatic parenchyma, i.e. to carry out a hepatic biopsy puncture (HBP), for example transjugularly or transparietally; and/or 25 - through the skin to the liver of said subject, so as to carry out a hepatic cytopuncture. The application pertains in particular to said manufactured article or to said reagents for their use in a method for the treatment of hepatopathy which comprises liver tissue damage, more particularly a hepatic fibrosis, more particularly an anti-HCV therapy 30 method, more particularly in an anti-HCV therapy method which comprises administering interferon and the administration of ribavirin (or their prodrugs), more particularly in an anti-HCV therapy method which comprises, as a first line treatment, the administration of interferon and the administration of ribavirin (or their prodrugs).

83 PCT/EP2012/052232 This use may in particular comprise using said reagents in a method of the invention in order to predict whether a subject infected with one or more HCVs has a strong probability of responding to an anti-HCV treatment which will comprise the administration of interferon and ribavirin or whether, in contrast, that subject has a high 5 probability of not responding to this anti-HCV treatment. If said subject is predicted to be a non-responder, the clinician may elect not to administer a treatment to the subject which comprises (more particularly which is essentially constituted by) administering interferon and administering ribavirin (or their prodrugs), more particularly 10 not to administer such a treatment as a first line treatment. In such a situation, the clinician may, for example, elect to administer an anti-HCV treatment which does not comprise (or which is not essentially constituted by) the administration of interferon and the administration of ribavirin (or their prodrugs) to the subject, more particularly to administer such a treatment to the subject as a first line treatment. The clinician may alternatively elect not to administer 15 anti-HCV treatment to the subject, at least as a first line treatment. If said subject is predicted to be a responder, the clinician may elect to administer an anti-HCV treatment, in particular a treatment which comprises (more particularly which is essentially constituted by) administering interferon and administering ribavirin (or their prodrugs), more particularly to administer, as a first line treatment, a treatment which comprises (more particularly which is 20 essentially constituted by) administering interferon and administering ribavirin (or their prodrugs). This use may, for example, comprise: - using said reagents of the invention on a biological sample which has been taken 25 from said subject, and which optionally has been transformed, for example: " by extraction and/or purification of the RNAs of said sample and, optionally, by reverse transcription of the extracted RNAs, or o by extraction and/or purification of proteins and/or polypeptides and/or peptides of said sample which has been taken, 30 in order to predict whether a subject infected with one or more HCVs has a strong probability of responding to an anti-HCV treatment which will comprise the administration of interferon and ribavirin or whether, in contrast, this subject has a high probability of not responding to this anti-HCV treatment, 84 PCT/EP2012/052232 - optionally, determining the HCV genotype infecting said patient and/or determining his hepatic fibrosis score (more particularly, determining whether this score is a score of at least F2 using the Metavir system). 5 If said subject is predicted to be a non-responder, the clinician may elect not to administer a treatment to the subject which comprises (more particularly which is essentially constituted by) administering interferon and administering ribavirin (or their prodrugs), more particularly of not administering such a treatment to him as a first line treatment. In such a situation, the clinician may, for example, elect to administer an anti-HCV treatment 10 which does not comprise (or which is not essentially constituted by) administering interferon and administering ribavirin (or their prodrugs), more particularly of administering such a treatment to the subject as a first line treatment. The clinician may alternatively elect not to administer anti-HCV treatment to the subject, at least as a first line treatment. If said subject is predicted to be a responder, the clinician may elect to administer 15 an anti-HCV treatment, in particular a treatment which comprises (more particularly which is essentially constituted by) administering interferon and administering ribavirin (or their prodrugs), more particularly of administering, as a first line treatment, a treatment which comprises (more particularly which is essentially constituted by) administering interferon and administering ribavirin (or their prodrugs). 20 Said treatment may, for example, be an anti-HCV treatment as described above and illustrated below. The application also pertains to a drug or combination of drugs for the treatment of a 25 hepatopathy comprising an attack of the tissue of the liver, more particularly a hepatic fibrosis (such as standard interferon or pegylated interferon, in a monotherapy or a polytherapy associating one or more other active principles, in particular ribavirin), in particular an anti-HCV treatment for its use in the treatment method of the invention. 30 The application also pertains to a computer program product to be stored in a memory of a processing unit or on a removable memory support for cooperation with a reader of said processing unit. The computer program product of the invention comprises instructions for carrying out a method of the invention, in particular for carrying out a statistical analysis adapted to carrying out a method of the invention (in particular adapted 85 PCT/EP2012/052232 to the multivariate statistical analysis of the measurements, and more particularly the levels of expression of said selected genes) and/or for the construction of a multivariate classification model adapted to carrying out a method in accordance with the invention. 5 The application also pertains to a computer unit, a computer device, or computer, comprising a processing unit with the following stored or recorded in its memory: - a computer program product of the invention, and, optionally, - assays, or measurement values, of the levels of expression (transcription and/or translation) of said selected genes. 10 The term "comprising", which is synonymous with "including" or "containing", is an open term and does not exclude the presence of one or more additional element(s), ingredient(s) or step(s) of the method which are not explicitly indicated, while the term "consisting" or "constituted" is a closed term which excludes the presence of any other additional element, 15 step or ingredient which is not explicitly disclosed. The term "essentially consisting" or "essentially constituted" is a partially open term which does not exclude the presence of one or more additional element(s), ingredient(s) or step(s) provided that this (these) additional element(s), ingredient(s) or step(s) do not materially affect the basic properties of the invention. 20 As a consequence, the term "comprising" (or "comprise(s)") includes the terms "consisting", "constituted" as well as the terms "essentially consisting" and "essentially constituted by". 25 With the aim of facilitating reading of the application, the description has been separated into various paragraphs, sections and embodiments. It should not be assumed that these separations disconnect the substance of one paragraph, section or embodiment from that of another paragraph, section or embodiment. On the contrary, the description encompasses all possible combinations of the various paragraphs, sections, phrases and embodiments which it 30 contains. The content of the bibliographic references cited in the application is specifically incorporated into the content of the application by reference.

86 PCT/EP2012/052232 The following examples are given purely by way of illustration. They do not in any way limit the invention. EXAMPLES 5 EXAMPLE 1: CONSTRUCTION OF CLASSIFICATION MODELS 1. Populations and patients, measurement of the level of gene expression, determination of response to treatment: The study was approved by the local Ethics Committee in accordance with the Helsinki 10 Declaration and all of the patients gave their informed written consent. Presentation of patients The patients were adult patients infected with the hepatitis C virus (HCV), monitored at 15 the H6pital Beaujon (Clichy, France). The clinical diagnosis of infection with the hepatitis C virus of the selected patients was established on the basis of the detection of antibodies directed against HCV proteins and the detection of circulating HCV RNA. 20 The serology of the HCV to be detected was carried out using the 3 rd generation Abbott test (AxSYM T M HCV Version 3.0 (Abbott) Technique MEIA; index > 1 = positive; index < 1 = negative) and the VERSANT@ HCV-RNA 3.0 (bDNA) ASSAY HCV RNA quantification test from Siemens Healthcare Diagnostics (quantification limit = 615 25 7 690 000 IU/mL). In order to establish a homogeneous cohort which was entirely representative of the exemplified pathology, patients susceptible of presenting chronic hepatic diseases of origins other than the hepatitis C virus (such as a chronic hepatic disease due to an 30 infection with hepatitis B virus) were excluded from the study. Other exclusion criteria were also applied, namely excessive alcohol consumption, haemochromatosis, auto immune hepatitis, Wilson's disease, c-I antitrypsin deficiency, primary sclerosing cholangitis, primary biliary cirrhosis or subsequent anti-HCV treatment.

87 PCT/EP2012/052232 One hundred and forty patients were thus selected. Table 8 below presents the clinical, biological and virological data of the patients who were 5 thus selected. These data were collected before the patient received an antiviral treatment, in this case during a hepatic biopsy puncture (HBP). In Table 8 below: IU = International Unit 10 NR patients = patients not responsive to treatment; R patients = patients responsive to treatment; RR patients= responder-relapser patients; see below for the definition of these three sub-populations or cohorts.

88 PCT/EP2012/052232 Table 8: Clinical, biological and virological data (set of selected patients) Clinical, biological and Patients NR patients R patients RR patients virological data n 140 51 68 21 Sex: male (%) / female (%) 89 (64) / 51 (36) 31 (61) / 20 (39) 43 (63) / 25 (37) 15 (71) / 6 (29) Age [mean ± standard deviation 45.8 ± 8.5 (27-72) 47.3 ± 8.6 (33-72) 44.7 ± 9.0 (27-65) 44.4 ± 4.9 (34-66) (range)] Source of infection [n(%)] blood transfusion 30(21) 11(22) 16(23) 3(14) intravenous administration 42(30) 17(33) 21(31) 4(19) of an unknown drug 68(49) 23(45) 31(46) 14(67) Alanine aminotransferase 106 ± 73 (18-459) 112 ± 81 (30-354) 102 ± 74 (20-459) 100 ± 36 (18 176) (ALT) IU/L [mean ± standard deviation (range)] 89 PCT/EP2012/052232 Table 8 (continued to end): Clinical, biological and virological data Clinical, biological and Patients NR patients R patients RR patients virological data HCV genotypes [n(%)] 1 76(54.3) 40(78.4) 28(41.2) 8(38.1) 2 13 (9.3) 10 (14.7) 3 (14.3) 3 19 (13.6) 3 (5.9) 12 (17.6) 4 (19.0) 4 31(22.1) 8 (15.7) 17 (25.0) 6(28.6) 5 1(0.7) 0 1(1.5) 0 Fibrosis score (Metavir F score) [n(%)] 0 1(0.7) 0 1(1.5) 0 1 45 (32.1) 15 (29.5) 26 (38.2) 4 (19.0) 2 53 (37.9) 18 (35.3) 29 (42.7) 6 (28.6) 3 18 (12.9) 9 (17.6) 3 (4.4) 6 (28.6) 4 22(15.7) 9(17.6) 9(13.2) 4(19.0) unknown 1 (0.7) 0 0 1 (4.8) 90 PCT/EP2012/052232 Sampling: A hepatic biopsy puncture (HBP) was carried out on each patient before any antiviral treatment was received. The HBPs were carried out in accordance with good clinical 5 practice. The biopsies were immediately stored at -80'C with a view to extracting total RNA, and treated with paraffin for the histological studies. A sample of serum was taken from each of the patients included in the study in a period of +/- 6 months from the date of the biopsy, but always before the patient received antiviral treatment. 10 Treatment of hepatic biopsy samples (for measurement of RNAs): The levels of expression of the genes (in fact, level of RNA transcription) were measured on each of the 140 biopsies (1 biopsy per patient). 15 The hepatic biopsies were ground in nitrogen using a ceramic pestle and mortar (100% manual grinding). The powder was recovered using a scalpel (Swann Morton 22, Reference 0208). 20 a) Extraction of RNAs The powder obtained was dissolved in 1 mL of RNAble@ Ref. GEXEXTO0, Laboratoires Eurobio, France, to which 100 gL of chloroform had been added. The mixture obtained was placed in ice or at 4'C for 5 minutes, then was centrifuged at 25 13 000 g for 15 minutes. The upper aqueous phase containing the RNAs was recovered into a fresh tube and 1 volume of isopropanol was added to it. 30 The tube was agitated by repeated inversion and was kept at 4'C overnight, then was centrifuged at 13 000 g for 15 minutes. The supernatant was eliminated and the pellet containing the RNAs was taken up in a volume of 70% ethanol (extemporaneously prepared) and centrifuged again.

91 PCT/EP2012/052232 The pellet of RNA precipitate obtained was dried in the open air for approximately 1 hour then dissolved in 15 pL of water and stored at -80'C. b) Measurement of RNAs 5 The evaluation of the concentration of extracted RNAs was carried out by measuring the optical density using a spectrometer (Nanodrop), and was verified after a freeze/thaw cycle. The extracted RNAs were then diluted to obtain a 50 ng/pL solution. 10 Quality controls of the RNA were carried out by real time PCR (see below) by screening a ubiquitous expression control gene (known as endogenous), to verify that the RNA had not degraded (in fact, screening RPLPO). 15 Reverse Transcription or RT step: The reverse transcription was carried out on 200 ng of RNA in a reaction mixture produced in a volume of 20 pL, comprising the following reagents: 92 PCT/EP2012/052232 Table 9: Reagent and reference product Starting solution Volume SUPERSCRIPT II RNase H reverse 200 U/ptL 0.5 pL transcriptase, Invitrogen, ref: 18064014 SUPER SCRIPT 5X buffer - 4.0 pL Invitrogen, ref: 18064014 RNAsin 40 U/ptL 0.5 pL Promega, ref: N2111 DTT 100 mM 2.0 pL The 4 dNTPs 10 mM lgL GE Healthcare, ref: 28406552 Pd(N) primers 0.5 pg/tL 6.0 tL RANDOM HEXAMERS 50 (A260) units, 51 Perbio, ref: MB216601 RNA 50 ng/pL 4.0 pL

H

2 0 qs 20 tL The reverse transcription reactions were carried out at the following temperatures: - at 20'C for 10 minutes, then 5 - at 42'C for 30 minutes, and - at 99'C for 5 minutes. At this stage, the reaction mixtures were frozen or aliquoted or used directly for real time PCR amplification. 10 Quantitative real time PCR step (qPCR): The amplification was carried out using a Light Cycler@ 480 (Roche Diagnostics, Mannheim, Germany). The results were generated using Light Cycler@ Software 4.05/ 4.1. 15 Light Cycler@ technology can be used to continuously monitor the appearance of the amplification products due to emission of a quantity of fluorescence which is proportional 93 PCT/EP2012/052232 to the quantity of amplified product, which is itself dependent on the quantity of targets initially present in the sample to be analysed. Quantification (in relative values) of the gene expression was carried out using the method which is known by the name 2-Act

(

2 -ACt = 2 -(Cttarget - Ct reference); see Livak and Schmittgen 2001; Schmitten and Livak 2008), 5 utilizing the values for "Cycle Threshold", or Ct, determined by the quantitative real time PCR apparatus. The smaller the value of Ct, the higher the initial quantity of transcribed RNA. The reaction mixtures and the protocol used are described in the instruction leaflet in the 10 LIGHT CYCLER@ 480 SYBR GREEN I MASTER MIX kit (Roche Diagnostics, Mannheim, Germany; US 4 683 202; US 4 683 195; US 4 965 188; US 6 569 627). After the reverse transcription step, the reaction mixtures (cDNAs) were diluted to 1/40th (to verify the quality) or to 1/100th (for the target genes) before using them in qPCR. 15 For each gene, the qPCRs were carried out in a reaction volume of 10 pL on a 384 well plate: - 5 pL of reverse transcription reaction, diluted to 1/40th (or 1/100th); 20 - 4.8 pL of reaction mixture from the Light Cycler@ 480 SYBR Green I Master mix kit; - 0.1 pL of a 50 pM solution for each of the two primers, i.e. a final volume of 0.5 pM for each primer. The reaction mixtures were generally prepared for the 384 well plates. 25 The following primers were used: 94 PCT/EP2012/052232 Table 10: Examples of primers Symbol Sense primer SEQ ID NO: Antisense primer SEQ ID NO: HERC5 CCT GGG CCA CAC TGA GAG TAA A 1 GAG CCA CCA CAA GCG ACA AA 2 IL8 CAC CGG AAG GAA CCA TCT CAC TGT 3 TCC TTG GCA AAA CTG CAC CTT CA 4 STAT2 GAG TCA GGG TTT GAT TTG GGA CTT 5 CTG TCA CAC CTA GTG GCC CCT TA 6 CCL21 CTC CAT CCC AGC TAT CCT GTT CTT 7 TCT GCA CAT AGC TCT GCC TGA GA 8 CLDN1 TAT TTC TTC TTG CAG GTC TGG CTA 9 CAA ATT CGT ACC TGG CAT TGA CT 10 CXCL6 GTT TAC GCG TTA CGC TGA GAG TAA A 11 CGT TCT TCA GGG AGG CTA CCA 12 FOXO1 GTC AAG AGC GTG CCC TAC TTC A 13 TGA ACT TGC TGT GTA GGG ACA GAT TAT 14 G1P2 GAG GCA GCG AAC TCA TCT TTG CCA 15 CCG CCA GCA TCT TCA CCG TCA 16 G1P3 GAT GAG CTG GTC TGC GAT CCT GAA 17 CCA CCA GCC CCG AGG CTC T 18 95 PCT/EP2012/052232 Table 10 (continued): Examples of primers Symbol Sense primer SEQ ID NO: Antisense primer SEQ ID NO: IF127 TGG GTA CTC TGC AGT CAC TGG GA 19 CCG CAA TGG CAG ACC CAA T 20 IF135 AGG CCA GAC TCA AGA TGA GGC TGT 21 GGG CAC TGA AAA TGG GAC CTT GT 22 IF144 AAA ATC TTG GAC TTG CTC AAA ATT GTA 23 CTT TCA TCC AGT GAA TCT TCG CA 24 IFITI GAG CAA CCA TGA GTA CAA ATG GTG A 25 CGT CAT CAA TGG ATA ACT CCC ATG T 26 IFIT4 ATC AGC CTG GTC ACC AGC TTT 27 ATT CTT GGT GAC CTC ACT CAT GAC T 28 IFITMI TCA ACA CCC TCT TCT TGA ACT GGT 29 CAA CCA TCT TCC TGT CCC TAG ACT T 30 ITGA2 AGG TGC CTG CAG AAG AAT ATG GT 31 GAC AAC ATC AGA GGG CTC CTG TAT 32 LGALS3BP TGA CCC CTC CGA GGC TCT TC 33 ATG TCA CCA TCG TTC ACG CCT T 34 MDK GGG CAG CGA GAT GCA GCA C 35 CCA CTC AGC GCA CTC GCT CC 36 MX1 ACA CAC CGT GAC GGA TAT GGT C 37 GGC GGT TCT GTG GAG GTT AAA 38 96 PCT/EP2012/052232 Table 10 (continued to end): Examples of primers Symbol Sense primer SEQ ID NO: Antisense primer SEQ ID NO: OAS1 CTG AAG GAA AGG TGC TTC CGA G 39 GCC TGA GGA GCC ACC CTT TAC 40 OAS2 TGA AGC ACT GGT ACA AAG AGT GTG A 41 TGA GCA GCT CCA AGG CAT ACT T 42 OAS3 TCG ACA TCA TCT TGC GCT GC 43 CCA AAT GAG CCC CCT TTA CTG A 44 OCLN CCC AAT GTC GAG GAG TGG GTT A 45 AGT ATG CCA TGG GAC TGT CAA CTC T 46 PLSCR1 GAA TGC TTC TCA CCC GGA AAC A 47 AGC CAC TAT ATC CTG GAG GTC CTT G 48 RSAD2 AGC AAA GAG AGG ATT GCT TTT GC 49 AGG TAT TCT CCC CGG TCT TGA A 50 STAT1 AGC ATG AAA TCA AGA GCC TGG A 51 ACC ATT GGT CTC GTG TTC TCT GTT 52 STMN2 GCG GAG GAA AAG CTG ATC CTG A 53 TCC GCA GCA TGC CTC TCC TT 54 USP18 CCG TGG GAA ACA GGT CTT GAA 55 ATG GAG AAT CGC ATG AGG TGG 56 RPLPO GGC GAC CTG GAA GTC CAA CT 57 CCA TCAG CAC CAC AGC CTT C 58 97 PCT/EP2012/052232 The qPCRs were carried out using the following temperature conditions: - a step for initiating denaturing at 95'C for 10 minutes; - 50 cycles of: 5 - denaturing at 95'C for 15 seconds; - hybridization / elongation at 65'C for 30 seconds. Each target sample was amplified in duplicate. In order to overcome variations in the initial quantities of total RNA from one sample to another, at the same time a duplicate amplification was carried out of the RNAs of a gene used as an endogenous control, such 10 as a gene involved in cellular metabolic cascades, for example RPLPO (also known by the name 36B4; GENBANK accession number NM_001002) or TBP (GENBANK accession number NM_003194). In fact, the gene RPLP0 was used here as the endogenous control. The quality of RNA extraction from the 140 biopsies was evaluated on the basis of the value of Ct of the reference gene, RPLP0. The classification was carried out as follows: 15 - RPLPO Ct less than 22: very good RNA quality; - RPLPO Ct from 22 to 24: good RNA quality; - RPLPO Ct more than 24 and less than 26: average RNA quality; - RPLPO Ct of 26 or more: poor RNA quality. 20 In order to increase the reliability of the bio-statistical analyses, only the data from RNA extraction of very good and good quality (RPLPO Ct < 24) were retained; there were 128 biopsies [91.4% of the 140 samples] of which 107 had a status of responder or non responder strict; see Table 11 below. 25 The quantity of transcripts of a target gene was deduced from the Ct ("Cycle threshold") which corresponded to the number of PCR cycles necessary in order to obtain a significant fluorescence signal. The target samples were normalized on the basis of their RPLPO (or, if necessary, TBP) content, using the 2 -Act method. 30 This value for the normalized measurement in this case is in general denoted "BMK" (for biomarker).

98 PCT/EP2012/052232 The BMK values obtained for each of the 128 patients are presented in Tables 12 to 15 below. Treatment of serum samples (for the measurement of seric proteins): 5 The protein measurements were carried out using the kits indicated in Table 16 below, following the recommendations of the manufacturer.

PCT/EP2012/052232 99 Table 11: Clinical, biological and virological data Clinical, biological and Patients NR patients R patients RR patients virological data n 128 44 63 21 Sex: male (%) / female (%) 82 (64) / 46 (36) 25 (57) / 19 (43) 42 (67) / (21 (33) 15 (71) / 6 (29) Age [mean ± standard deviation 47.0 ± 8.7 (27-73) 46.1 ± 8.9 (27-66) 48.1 ± 8.9 (35-73) 47.4 ± 7.8 (35-67) (range)] Source of infection [n(%)] blood transfusion 28 (22) 10 (23) 15 (24) 3 (14) intravenous administration 37 (29) 13 (30) 20 (32) 4 (19) of an unknown drug 63 (49) 21(48) 28 (44) 14(67) Alanine aminotransferase 112 ± 82 (18-459) 114 + 80 (30-354) 119 ± 92 (30- 459) 88 ± 37 (18-176) (ALT) IU/L [mean ± standard deviation (range)] PCT/EP2012/052232 100 Table 11 (continued to end): Clinical, biological and virological data Clinical, biological and Patients NR patients R patients RR patients virological data HCV genotypes [n(%)] 1 70(55) 37(84) 25(40) 8(38) 2 12(9) 0(0) 9(14) 3(14) 3 19(15) 3 (7) 12(19) 4(19) 4 26(20) 4(9) 16(25) 6(29) 5 1(1) 0(0) 1(2) 0(0) Fibrosis score (Metavir F score) [n(%)] 0 0(0) 0(0) 0(0) 0(0) 1 41(32) 14(32) 23(37) 4(19) 2 49(38) 15 (34) 28(44) 6(29) 3 17(13) 8(18) 3(5) 6(29) 4 20(16) 7(16) 9(14) 4(19) unknown 1 (1) 0 (0) 0 (0) 1(5) PCT/EP2012/052232 101 Table 12: Patients' BMK values for the genes HERC5, IL8, STAT2, CXCL6, GIP2, IF135 and IF144 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R HERC5 IL8 STAT2 CXCL6 G1P2 IF135 IF144 or RR) 45 R 0.588 1.641 1.072 0.092 0.959 0.415 0.056 50 R 0.144 0.444 0.730 0.027 1.586 0.174 0.018 55 R 0.235 1.395 0.307 0.053 1.275 0.082 0.034 59 R 0.269 0.329 0.454 0.000 0.877 0.240 0.039 62 R 0.280 0.058 0.771 0.080 0.943 0.318 0.074 63 R 1.129 0.199 0.904 0.000 12.295 0.523 0.291 65 R 0.258 0.168 0.448 0.013 1.613 0.209 0.055 66 R 0.374 0.420 0.580 0.119 0.308 0.103 0.018 71 R 0.255 0.228 0.304 0.000 0.564 0.087 0.015 72 R 0.097 0.064 0.290 0.052 0.176 0.074 0.010 73 R 1.409 0.191 0.691 0.027 1.064 0.181 0.180 76 R 0.293 1.257 1.091 0.051 0.653 0.163 0.067 86 R 0.259 0.451 0.564 0.102 14.123 0.244 0.086 88 R 0.660 0.297 0.745 0.072 2.042 0.172 0.092 90 R 0.342 0.089 0.516 0.045 1.912 0.207 0.046 91 R 0.149 0.040 0.270 0.096 0.664 0.130 0.042 92 R 0.025 0.087 0.127 0.027 0.049 0.046 0.006 125 R 0.248 0.000 0.126 0.000 1.257 0.099 0.047 222 R 0.631 0.330 0.318 0.023 5.187 0.253 0.181 227 R 0.145 0.260 0.210 0.000 0.119 0.062 0.036 306 R 0.380 0.372 0.976 0.186 3.568 0.559 0.130 344 R 0.429 0.923 1.683 0.330 0.646 0.114 0.181 346 R 0.503 0.182 2.267 0.029 0.601 0.293 0.146 366 R 0.194 0.179 0.279 0.020 0.159 0.081 0.029 501 R 0.151 0.601 0.443 0.070 1.424 0.200 0.067 502 R 0.158 2.676 0.189 0.557 0.465 0.074 0.016 503 R 0.154 0.109 0.230 0.011 0.755 0.131 0.026 504 R 0.150 2.289 0.274 0.142 0.563 0.162 0.034 506 R 0.042 10.483 0.347 0.343 0.678 0.115 0.043 508 R 0.114 0.446 0.545 0.000 0.143 0.084 0.040 513 R 0.105 0.538 0.332 0.000 0.280 0.086 0.038 523 R 0.028 2.329 0.212 0.176 0.134 0.070 0.009 5 PCT/EP2012/052232 102 Table 12 (continued): Patients' BMK values for the genes HERC5, IL8, STAT2, CXCL6, GIP2, IF135 and IF144 (Ct normalized in accordance with the 2-ACt method) Status Patient (NR, R HERC5 IL8 STAT2 CXCL6 G1P2 IF135 IF144 or RR) 528 R 0.253 4.547 0.883 1.189 1.261 0.263 0.056 529 R 0.188 0.354 0.350 0.053 1.210 0.127 0.037 530 R 0.218 0.000 1.240 0.000 0.144 0.130 0.042 532 R 0.173 0.387 0.561 0.032 0.245 0.073 0.019 535 R 0.804 0.328 0.609 0.026 15.348 0.446 0.138 536 R 0.212 0.119 0.448 0.180 0.313 0.125 0.035 537 R 0.052 0.058 0.182 0.027 0.080 0.047 0.010 538 R 0.422 0.041 0.413 0.000 5.152 0.375 0.068 546 R 0.408 0.000 0.735 0.062 0.236 0.164 0.076 556 R 0.409 0.063 0.221 0.000 6.869 0.254 0.170 560 R 0.203 0.530 0.270 0.082 3.249 0.157 0.058 565 R 0.538 1.079 0.079 0.000 4.857 0.301 0.124 567 R 0.238 1.297 0.051 0.356 2.497 0.262 0.059 568 R 0.034 0.127 0.005 0.000 0.055 0.038 0.003 569 R 0.077 0.559 0.018 0.000 0.092 0.202 0.031 570 R 0.113 0.371 0.868 0.000 0.118 0.207 0.060 571 R 0.325 0.804 0.747 0.036 0.609 0.245 0.111 572 R 0.169 0.064 0.243 0.005 2.129 0.125 0.050 575 R 1.653 0.892 1.952 0.226 15.348 1.061 0.350 577 R 0.215 0.262 0.841 0.084 0.224 0.159 0.057 581 R 0.084 2.549 0.309 0.129 1.210 0.126 0.041 583 R 0.231 0.230 0.551 0.367 1.297 0.432 0.058 585 R 0.396 0.818 0.660 0.112 3.411 0.331 0.094 598 R 0.050 0.660 0.201 0.033 0.200 0.073 0.023 601 R 0.431 0.459 1.133 0.077 1.053 0.249 0.199 604 R 0.200 0.084 0.473 0.082 1.061 0.221 0.079 605 R 0.847 0.178 0.622 0.000 5.315 0.228 0.149 613 R 0.083 0.111 0.419 0.000 0.108 0.161 0.028 614 R 0.132 0.218 0.541 0.078 0.135 0.098 0.040 629 R 0.070 0.152 0.644 0.000 0.074 0.113 0.026 639 R 0.121 0.399 0.895 0.000 0.109 0.119 0.042 6 NR 0.247 2.648 0.624 0.000 0.685 0.113 0.046 5 PCT/EP2012/052232 103 Table 12 (continued): Patients' BMK values for the genes HERC5, IL8, STAT2, CXCL6, GIP2, IF135 and IF144 (Ct normalized in accordance with the 2-ACt method) Status Patient (NR, R HERC5 IL8 STAT2 CXCL6 G1P2 IF135 IF144 or RR) 46 NR 0.563 0.911 1.693 0.115 8.969 0.329 0.177 58 NR 0.607 0.616 0.787 0.152 13.881 0.490 0.149 75 NR 1.035 0.908 2.158 0.082 8.664 0.601 0.319 80 NR 0.570 0.745 0.493 0.096 0.963 0.146 0.057 83 NR 2.549 0.231 2.063 0.154 6.169 0.798 0.218 145 NR 1.283 0.755 0.838 0.196 7.490 0.423 0.183 167 NR 0.653 0.923 0.502 0.112 5.046 0.405 0.075 308 NR 0.886 0.853 0.440 0.115 6.658 0.437 0.154 509 NR 0.719 1.157 0.637 0.218 6.126 0.264 0.102 516 NR 0.358 0.844 0.559 0.049 3.160 0.232 0.098 521 NR 1.670 3.732 3.745 0.170 4.362 1.292 0.785 524 NR 0.236 0.949 0.578 0.025 0.286 0.120 0.028 526 NR 0.418 2.370 0.940 0.169 5.260 0.441 0.127 527 NR 0.868 0.657 1.057 0.060 9.318 0.495 0.192 534 NR 0.667 1.828 0.853 0.194 2.129 0.340 0.114 549 NR 0.639 0.979 1.759 0.080 2.297 0.434 0.290 574 NR 1.210 5.756 2.540 0.139 4.959 0.877 0.735 582 NR 0.476 2.949 0.455 0.409 4.675 0.230 0.081 596 NR 0.374 0.883 0.639 0.080 0.914 0.230 0.066 602 NR 0.047 0.936 0.180 0.059 0.091 0.086 0.012 618 NR 0.547 1.035 1.206 0.294 4.757 0.486 0.273 619 NR 0.276 0.301 1.218 0.064 1.542 0.284 0.237 636 NR 2.395 0.270 3.063 0.081 18.765 1.010 0.898 645 NR 0.026 0.000 0.124 0.000 0.069 0.058 0.008 646 NR 1.113 0.518 1.866 0.087 35.383 1.177 0.221 647 NR 0.859 0.192 1.283 0.110 20.393 0.883 0.216 649 NR 1.083 2.742 1.161 0.167 3.918 0.288 0.110 650 NR 1.000 5.816 2.354 0.446 12.381 0.990 0.294 651 NR 0.705 0.940 1.094 0.150 4.773 0.394 0.093 657 NR 0.356 4.840 0.787 2.412 2.949 0.322 0.135 658 NR 0.236 0.557 0.392 0.000 0.322 0.081 0.048 659 NR 0.547 4.423 0.536 0.237 0.690 0.150 0.067 5 PCT/EP2012/052232 104 Table 12 (continued to end): Patients' BMK values for the genes HERC5, IL8, STAT2, CXCL6, GIP2, IF135 and IF144 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R HERC5 IL8 STAT2 CXCL6 G1P2 IF135 IF144 or RR) 660 NR 0.874 1.113 0.856 0.308 6.612 0.284 0.138 662 NR 1.050 0.046 0.426 0.051 5.483 0.179 0.066 664 NR 0.420 0.953 1.157 0.139 2.704 0.316 0.066 665 NR 0.176 0.949 0.495 0.061 0.576 0.118 0.024 666 NR 0.578 2.799 1.275 0.000 5.796 0.358 0.100 67 NR 0.133 0.104 0.369 0.036 0.314 0.125 0.021 563 NR 0.208 0.330 0.085 0.063 0.235 0.221 0.030 573 NR 0.940 0.323 0.990 0.064 6.892 0.543 0.207 599 NR 0.826 0.262 0.221 0.005 1.248 0.115 0.050 641 NR 0.408 0.657 0.462 0.150 2.454 0.204 0.048 642 NR 0.516 13.408 1.993 1.682 6.635 0.807 0.285 49 RR 0.521 0.284 1.853 0.000 1.905 0.262 0.177 56 RR 0.468 0.363 0.607 0.000 3.249 0.225 0.115 60 RR 0.690 0.145 0.841 0.000 7.945 0.362 0.126 87 RR 1.072 1.608 1.310 0.117 11.432 0.570 0.183 505 RR 0.476 0.410 0.325 0.000 4.084 0.263 0.138 514 RR 0.200 1.053 1.193 0.066 0.525 0.224 0.055 531 RR 0.039 0.053 0.101 0.004 0.103 0.036 0.010 533 RR 0.135 3.084 0.405 0.189 1.145 0.132 0.038 543 RR 0.486 0.597 1.157 0.059 1.548 0.375 0.204 554 RR 0.033 1.072 0.324 0.125 0.148 0.096 0.011 557 RR 0.355 0.700 0.304 0.061 3.317 0.248 0.073 558 RR 0.045 0.983 0.030 0.097 0.289 0.067 0.023 559 RR 0.076 0.536 0.113 0.069 0.871 0.101 0.015 562 RR 0.459 1.380 0.216 0.140 3.668 0.305 0.122 576 RR 0.493 1.064 0.956 0.127 5.187 0.332 0.198 579 RR 0.370 0.483 0.490 0.072 0.147 0.065 0.059 588 RR 0.025 0.255 0.384 0.041 0.053 0.059 0.016 589 RR 0.697 0.315 0.750 0.033 7.387 0.324 0.159 591 RR 0.412 0.113 0.262 0.048 2.378 0.179 0.071 592 RR 0.959 7.781 1.149 1.094 6.000 0.355 0.196 643 RR 0.083 15.835 0.597 1.197 0.390 0.228 0.026 5 PCT/EP2012/052232 105 Table 13: Patients' BMK values for the genes IFITI, IFIT4, ITGA2, MDK, MX1, OAS1 and OAS3 (Ct normalized in accordance with the 2- method) Status Patient (NR, R IFITI IFIT4 ITGA2 MDK MX1 OAS1 OAS3 or RR) 45 R 0.020 0.753 0.408 0.037 0.254 1.014 1.879 50 R 0.011 0.111 0.166 0.132 0.681 0.826 1.479 55 R 0.034 0.305 0.275 0.054 1.035 0.883 0.911 59 R 0.017 0.426 0.000 0.090 0.771 1.380 0.856 62 R 0.020 0.607 0.160 0.043 0.986 1.444 1.821 63 R 0.261 4.347 0.063 0.106 4.042 3.506 5.676 65 R 0.023 0.256 0.000 0.105 0.582 0.908 0.979 66 R 0.011 0.179 0.066 0.054 0.361 0.378 0.465 71 R 0.013 0.193 0.000 0.021 0.455 0.437 0.760 72 R 0.008 0.092 0.115 0.001 0.245 0.162 0.391 73 R 0.048 0.885 0.000 0.055 0.841 1.364 0.568 76 R 0.024 0.391 0.376 0.045 0.717 1.161 1.283 86 R 0.045 0.491 0.070 0.060 1.474 1.053 2.007 88 R 0.046 0.620 0.000 0.024 1.165 0.976 1.227 90 R 0.033 0.372 0.030 0.115 1.500 0.853 1.057 91 R 0.017 0.211 0.082 0.051 0.796 0.651 0.518 92 R 0.003 0.064 0.045 0.001 0.052 0.187 0.084 125 R 0.026 0.345 0.198 0.058 0.465 0.963 0.102 222 R 0.089 2.250 0.074 0.063 1.840 1.537 1.137 227 R 0.011 0.259 0.071 0.002 0.218 0.301 0.120 306 R 0.032 1.886 0.553 0.125 3.732 3.972 2.204 344 R 0.047 0.858 0.022 0.045 1.766 1.378 1.388 346 R 0.037 0.695 0.422 0.032 1.352 2.519 0.965 366 R 0.014 0.137 0.273 0.012 0.297 0.325 0.547 501 R 0.027 0.358 0.094 0.046 1.017 0.826 0.083 502 R 0.007 0.139 0.693 0.056 0.476 0.299 0.019 503 R 0.018 0.257 0.000 0.014 0.690 0.936 0.059 504 R 0.012 0.376 0.000 0.028 0.301 0.747 0.067 506 R 0.017 0.289 0.702 0.057 0.415 0.525 0.173 508 R 0.006 0.349 0.080 0.032 0.163 0.588 0.143 513 R 0.012 0.298 0.425 0.034 0.409 0.553 0.113 523 R 0.003 0.079 0.214 0.045 0.052 0.140 0.062 5 PCT/EP2012/052232 106 Table 13 (continued): Patients' BMK values for the genes IFITI, IFIT4, ITGA2, MDK, MX1, OAS1 and OAS3 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R IFITI IFIT4 ITGA2 MDK MX1 OAS1 OAS3 or RR) 528 R 0.018 0.432 0.285 0.188 2.092 1.647 0.787 529 R 0.014 0.207 0.136 0.084 0.920 0.669 0.331 530 R 0.008 0.281 0.216 0.023 0.498 0.847 0.200 532 R 0.010 0.152 0.121 0.207 0.288 0.179 0.329 535 R 0.075 1.161 0.000 0.057 2.969 1.892 2.770 536 R 0.012 0.221 0.087 0.010 0.443 0.351 0.406 537 R 0.004 0.072 0.140 0.014 0.196 0.154 0.131 538 R 0.039 0.536 0.072 0.108 2.462 1.705 1.210 546 R 0.012 0.372 0.644 0.039 0.355 1.061 0.667 556 R 0.073 1.189 0.093 0.011 2.121 2.676 2.266 560 R 0.049 0.370 0.172 0.028 1.257 0.930 1.149 565 R 0.135 1.625 0.071 0.049 3.042 1.879 2.042 567 R 0.051 0.412 0.399 0.038 1.834 0.963 1.306 568 R 0.004 0.054 0.000 0.005 0.146 0.146 0.184 569 R 0.021 0.216 0.000 0.008 0.566 0.633 0.683 570 R 0.012 0.441 0.000 0.046 0.785 0.889 0.707 571 R 0.018 0.509 0.000 0.050 12.000 0.883 1.275 572 R 0.028 0.206 0.080 0.009 1.253 0.674 0.626 575 R 0.123 3.837 0.543 0.525 7.945 6.431 9.849 577 R 0.020 0.525 0.000 0.024 12.951 1.117 1.181 581 R 0.030 0.329 0.367 0.034 1.959 0.683 1.165 583 R 0.015 0.737 0.118 0.056 1.495 1.542 1.575 585 R 0.038 0.886 0.217 0.063 2.445 2.445 1.474 598 R 0.010 0.103 0.056 0.009 0.214 0.203 0.365 601 R 0.053 1.376 0.507 0.125 3.482 0.631 2.979 604 R 0.021 0.481 0.173 0.157 1.376 1.244 1.711 605 R 0.078 0.953 0.109 0.173 2.313 1.586 2.313 613 R 0.011 2.313 0.753 0.017 0.177 0.620 0.446 614 R 0.004 0.211 0.525 0.036 0.233 0.371 0.444 629 R 0.006 0.225 0.000 0.012 0.356 0.420 0.449 639 R 0.007 0.292 0.541 0.031 0.305 0.486 0.367 6 NR 0.010 0.171 17.569 0.688 0.457 0.228 0.702 5 PCT/EP2012/052232 107 Table 13 (continued): Patients' BMK values for the genes IFITi, IFIT4, ITGA2, MDK, MX1, OAS1 and OAS3 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R IFITI IFIT4 ITGA2 MDK MX1 OAS1 OAS3 or RR) 46 NR 0.080 1.283 0.091 0.308 2.173 3.543 3.811 58 NR 0.100 1.231 0.175 0.631 2.838 2.313 2.657 75 NR 0.198 3.306 0.459 1.248 7.062 3.959 6.612 80 NR 0.022 0.331 0.152 0.030 0.829 0.559 0.969 83 NR 0.235 2.732 0.237 0.476 6.498 3.411 7.037 145 NR 0.079 1.292 0.270 0.255 3.340 2.395 0.563 167 NR 0.100 1.214 0.231 0.521 0.463 2.204 2.549 308 NR 0.044 1.253 0.000 0.446 1.711 2.305 1.693 509 NR 0.061 0.908 0.444 0.172 3.021 1.324 0.651 516 NR 0.063 1.072 0.105 0.043 1.079 1.664 0.261 521 NR 0.349 7.648 1.193 1.676 8.140 6.233 2.114 524 NR 0.009 0.270 0.142 0.028 0.294 0.444 0.167 526 NR 0.063 0.997 0.282 0.189 2.630 2.305 1.035 527 NR 0.066 0.956 0.582 0.276 5.598 2.732 1.586 534 NR 0.083 1.046 0.300 0.140 2.173 1.608 2.395 549 NR 0.064 1.414 0.323 0.156 3.506 4.084 3.306 574 NR 0.163 3.193 0.435 0.191 10.629 6.277 6.916 582 NR 0.038 0.624 0.547 0.034 2.898 1.057 2.282 596 NR 0.035 0.551 0.262 0.069 1.682 1.017 1.390 602 NR 0.005 0.110 0.169 0.012 0.151 0.154 0.240 618 NR 0.072 1.796 0.000 0.152 2.949 4.213 3.238 619 NR 0.033 1.602 0.557 0.075 3.193 1.474 2.949 636 NR 0.256 6.364 0.277 0.712 15.242 9.580 13.690 645 NR 0.002 0.134 0.179 0.032 0.052 0.141 0.268 646 NR 0.091 4.611 0.000 0.314 1.564 5.389 10.666 647 NR 0.111 3.084 0.000 0.053 3.317 3.771 7.362 649 NR 0.048 0.993 0.104 0.053 0.448 1.474 1.366 650 NR 0.154 2.505 0.179 0.233 3.864 5.028 6.364 651 NR 0.031 0.859 0.218 0.253 1.925 2.488 2.114 657 NR 0.075 0.990 0.620 0.782 0.979 1.376 2.158 658 NR 0.021 0.379 0.271 0.018 0.410 0.518 0.308 659 NR 0.025 0.467 0.251 0.024 1.039 0.898 1.068 5 PCT/EP2012/052232 108 Table 13 (continued to end): Patients' BMK values for the genes IFITi, IFIT4, ITGA2, MDK, MX1, OAS1 and OAS3 (Ct normalized in accordance with the 2 -ACt method) Status Patient (NR, R IFITI IFIT4 ITGA2 MDK MX1 OAS1 OAS3 or RR) 660 NR 0.077 0.760 0.206 0.037 1.061 1.979 1.688 662 NR 0.019 0.284 0.072 0.047 1.803 0.790 1.110 664 NR 0.071 0.681 0.013 0.046 1.495 1.165 1.723 665 NR 0.024 0.182 0.000 0.026 0.258 0.366 0.514 666 NR 0.058 0.676 0.000 0.221 2.780 6.233 2.878 67 NR 0.007 0.144 0.188 0.017 0.264 0.374 0.545 563 NR 0.012 0.534 0.232 0.011 0.269 0.568 0.908 573 NR 0.123 1.747 0.105 0.047 3.850 3.375 4.362 599 NR 0.020 0.271 0.000 0.010 0.793 0.463 0.657 641 NR 0.027 0.459 0.138 0.028 0.717 0.914 1.110 642 NR 0.089 3.824 1.352 0.688 1.939 6.342 6.169 49 RR 0.045 0.678 0.202 0.076 0.859 2.471 1.952 56 RR 0.057 0.889 0.067 0.349 1.853 1.591 2.014 60 RR 0.031 0.642 0.000 0.192 1.853 1.790 2.227 87 RR 0.153 1.564 0.384 0.077 7.362 2.387 3.904 505 RR 0.049 1.404 0.000 0.037 1.338 1.641 0.622 514 RR 0.016 0.451 0.123 0.021 1.333 1.320 0.279 531 RR 0.004 0.094 0.209 0.034 0.214 0.150 0.115 533 RR 0.024 0.289 0.155 0.041 1.050 0.662 1.000 543 RR 0.031 0.940 0.179 0.075 1.747 1.532 1.409 554 RR 0.005 0.176 0.091 0.014 0.328 0.366 0.325 557 RR 0.053 0.605 0.216 0.001 1.952 0.946 0.787 558 RR 0.009 0.102 0.186 0.006 0.305 0.270 0.315 559 RR 0.015 0.191 0.091 0.007 1.091 0.349 0.572 562 RR 0.050 0.911 0.367 0.075 2.049 1.424 2.266 576 RR 0.073 1.521 0.435 0.069 0.245 3.117 2.630 579 RR 0.008 0.247 0.000 0.006 0.202 0.465 0.331 588 RR 0.006 0.105 0.246 0.008 0.158 0.249 0.219 589 RR 0.054 1.434 0.205 0.054 3.605 1.959 3.117 591 RR 0.035 0.432 0.171 0.035 1.569 0.818 0.747 592 RR 0.089 1.270 1.925 0.138 3.758 0.549 2.639 643 RR 0.011 0.261 0.755 0.054 0.299 0.859 1.137 5 PCT/EP2012/052232 109 Table 14: Patients' BMK values for the genes OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18 (Ct normalized in accordance with the 2- method) Status Patient (NR, R OCLN PLSCR1 RSAD2 STAT1 STMN2 USP18 or RR) 45 R 0.072 0.584 0.597 0.541 0.000 0.821 50 R 0.016 0.228 0.194 0.137 0.127 0.732 55 R 0.015 0.242 0.221 0.114 0.253 0.584 59 R 0.023 0.361 0.283 0.080 0.046 0.835 62 R 0.031 0.662 0.568 0.370 0.103 1.569 63 R 0.028 0.877 5.637 0.705 0.174 3.031 65 R 0.022 0.313 0.480 0.135 0.164 0.485 66 R 0.035 0.346 0.100 0.148 0.061 0.434 71 R 0.013 0.224 0.322 0.071 0.000 0.599 72 R 0.016 0.075 0.054 0.041 0.000 0.218 73 R 0.048 0.483 2.493 0.064 0.007 0.467 76 R 0.031 0.476 0.340 0.363 0.143 0.629 86 R 0.027 0.432 0.681 0.289 0.119 0.883 88 R 0.045 0.278 0.557 0.309 0.125 0.753 90 R 0.019 0.320 0.376 0.232 0.081 0.352 91 R 0.018 0.304 0.248 0.170 0.000 0.376 92 R 0.008 0.097 0.052 0.054 0.038 0.075 125 R 0.012 0.173 0.448 0.129 0.108 0.354 222 R 0.019 0.158 1.235 0.483 0.000 0.856 227 R 0.020 0.119 0.099 0.077 0.000 0.295 306 R 0.030 0.973 1.925 0.536 0.115 2.144 344 R 0.042 0.493 2.785 0.335 0.000 1.150 346 R 0.043 0.605 1.691 0.212 0.000 0.520 366 R 0.021 0.196 0.420 0.099 0.000 0.625 501 R 0.028 0.297 0.490 0.297 0.336 0.495 502 R 0.015 0.195 0.099 0.139 1.261 0.254 503 R 0.017 0.188 0.241 0.085 0.054 0.251 504 R 0.018 0.170 0.237 0.128 0.307 0.574 506 R 0.018 0.207 0.186 0.285 2.858 0.398 508 R 0.000 0.129 0.176 0.357 0.205 0.228 513 R 0.008 0.137 0.220 0.137 0.057 0.485 523 R 0.010 0.070 0.053 0.137 1.083 0.158 5 PCT/EP2012/052232 110 Table 14 (continued): Patients' BMK values for the genes OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R OCLN PLSCR1 RSAD2 STAT1 STMN2 USP18 or RR) 528 R 0.020 0.118 0.354 0.209 2.878 0.536 529 R 0.016 0.218 0.356 0.138 0.070 0.448 530 R 0.050 0.354 0.185 0.124 0.266 0.382 532 R 0.022 0.117 0.112 0.068 0.000 0.288 535 R 0.021 0.119 2.454 0.195 0.000 1.682 536 R 0.018 0.154 0.182 0.134 0.292 0.497 537 R 0.017 0.104 0.065 0.051 0.066 0.117 538 R 0.028 0.267 1.181 0.254 0.063 1.189 546 R 0.057 0.379 0.260 0.177 0.000 0.631 556 R 0.011 0.250 1.729 0.260 0.076 0.956 560 R 0.018 0.401 0.480 0.254 0.140 0.395 565 R 0.013 0.486 2.144 0.350 0.226 1.223 567 R 0.021 0.387 0.740 0.275 0.312 0.609 568 R 0.005 0.064 0.056 0.028 0.146 0.080 569 R 0.027 0.202 0.147 0.134 0.000 1.161 570 R 0.036 0.395 0.202 0.283 0.707 0.963 571 R 0.034 0.547 0.497 0.620 0.147 0.597 572 R 0.014 0.240 0.252 0.089 0.060 0.576 575 R 0.074 1.619 5.011 0.847 0.223 3.904 577 R 0.021 0.405 0.304 0.181 0.000 1.310 581 R 0.016 0.296 0.295 0.177 1.778 0.574 583 R 0.022 0.493 0.402 0.247 0.363 0.898 585 R 0.016 0.605 1.292 0.432 0.067 1.449 598 R 0.009 0.131 0.090 0.084 0.108 0.207 601 R 0.019 0.167 0.790 0.511 0.000 1.815 604 R 0.020 0.475 0.609 0.156 0.163 1.050 605 R 0.022 0.485 1.218 0.287 0.074 1.035 613 R 0.035 0.264 0.152 0.120 0.000 0.379 614 R 0.026 0.305 0.067 0.202 0.193 0.262 629 R 0.035 0.150 0.122 0.092 0.000 0.760 639 R 0.025 0.257 0.202 0.126 0.000 0.667 6 NR 0.020 0.755 0.266 0.151 87.730 0.224 5 PCT/EP2012/052232 111 Table 14 (continued): Patients' BMK values for the genes OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R OCLN PLSCR1 RSAD2 STAT1 STMN2 USP18 or RR) 46 NR 0.063 0.690 2.395 0.367 0.620 1.613 58 NR 0.041 0.503 1.439 0.283 0.041 1.753 75 NR 0.079 1.197 2.959 0.943 0.298 3.434 80 NR 0.027 0.283 0.309 0.147 0.179 0.441 83 NR 0.088 1.315 4.362 0.700 0.330 2.918 145 NR 0.047 0.653 1.469 0.578 0.315 1.253 167 NR 0.027 0.559 0.129 0.467 0.580 1.821 308 NR 0.049 0.069 1.057 0.209 0.000 2.362 509 NR 0.025 0.428 1.564 0.580 0.110 0.973 516 NR 0.031 0.282 0.892 0.416 0.189 0.917 521 NR 0.077 0.973 6.681 2.258 0.880 7.808 524 NR 0.027 0.230 0.168 0.183 0.362 0.281 526 NR 0.028 0.311 0.973 0.301 0.131 2.007 527 NR 0.036 0.532 2.121 0.425 0.369 1.919 534 NR 0.024 0.236 1.338 0.234 0.106 1.240 549 NR 0.041 0.446 1.202 0.707 0.000 2.667 574 NR 0.063 1.847 6.042 1.619 0.509 2.723 582 NR 0.033 0.620 1.050 0.272 2.523 0.551 596 NR 0.020 0.382 0.324 0.352 0.164 0.541 602 NR 0.014 0.174 0.080 0.124 0.022 0.180 618 NR 0.022 0.039 2.305 0.648 0.700 2.092 619 NR 0.010 0.441 1.010 0.460 0.000 1.193 636 NR 0.082 1.847 12.773 2.196 0.363 6.105 645 NR 0.000 0.105 0.163 0.022 0.000 0.126 646 NR 0.063 2.151 3.945 0.435 0.000 7.111 647 NR 0.045 1.925 3.543 0.662 0.170 3.000 649 NR 0.041 0.631 0.559 0.354 0.105 0.993 650 NR 0.080 1.266 3.732 0.824 0.547 4.098 651 NR 0.055 0.545 0.868 0.206 0.000 1.717 657 NR 0.046 0.541 0.732 0.418 1.676 1.244 658 NR 0.019 0.309 0.140 0.259 0.173 0.538 659 NR 0.023 0.422 0.676 0.248 0.123 0.438 PCT/EP2012/052232 112 Table 14 (continued to end): Patients' BMK values for the genes OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18 (Ct normalized in accordance with the 2 -ACt method) Status Patient (NR, R OCLN PLSCR1 RSAD2 STAT1 STMN2 USP18 or RR) 660 NR 0.037 0.603 1.385 0.291 0.123 0.785 662 NR 0.023 0.398 1.189 0.115 0.091 0.507 664 NR 0.046 0.449 0.674 0.315 0.710 1.117 665 NR 0.015 0.153 0.141 0.122 0.156 0.349 666 NR 0.040 0.605 1.210 0.310 0.262 1.301 67 NR 0.017 0.272 0.091 0.078 0.221 0.349 563 NR 0.019 0.199 0.098 0.139 0.000 0.215 573 NR 0.034 0.785 3.458 0.898 0.139 2.173 599 NR 0.008 0.198 0.219 0.088 0.023 0.409 641 NR 0.026 0.448 0.519 0.180 0.031 0.829 642 NR 0.116 1.580 0.862 1.094 1.919 2.532 49 RR 0.048 0.707 0.593 0.296 0.336 1.014 56 RR 0.018 0.273 1.283 0.345 0.212 1.046 60 RR 0.037 0.540 1.283 0.205 0.000 1.516 87 RR 0.039 0.624 3.294 0.697 0.105 1.741 505 RR 0.012 0.241 1.153 0.299 0.386 1.218 514 RR 0.060 0.422 0.460 0.300 0.073 0.727 531 RR 0.001 0.050 0.149 0.027 0.011 0.119 533 RR 0.020 0.183 0.530 0.226 1.569 0.676 543 RR 0.064 0.151 0.582 0.674 0.247 1.039 554 RR 0.024 0.115 0.091 0.160 0.179 0.384 557 RR 0.016 0.226 0.930 0.240 0.340 0.807 558 RR 0.009 0.096 0.084 0.067 0.351 0.194 559 RR 0.008 0.189 0.306 0.100 0.128 0.182 562 RR 0.037 0.518 1.177 0.365 0.904 1.061 576 RR 0.019 0.763 2.121 0.527 0.702 1.177 579 RR 0.018 0.137 0.163 0.127 0.188 0.235 588 RR 0.019 0.094 0.038 0.059 0.138 0.253 589 RR 0.023 0.609 1.735 0.412 0.284 1.064 591 RR 0.021 0.372 0.584 0.248 0.140 0.584 592 RR 0.041 0.973 1.747 0.868 4.141 1.297 643 RR 0.039 0.380 0.052 0.190 1.790 0.597 5 PCT/EP2012/052232 113 Table 15: Patients' BMK values for the genes CCL21, CLDN1, FOXO1, G1P2, G1P3, IF127, LGALS3BP and OAS2 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R CCL21 CLDN1 FOXO1 G1P2 G1P3 IF127 IFITMI LGALS3BP OAS2 or RR) 45 R 0.033 3.732 0.402 0.959 0.020 0.120 0.162 0.063 0.152 50 R 0.013 1.409 0.083 1.586 0.097 0.844 0.189 0.087 0.202 55 R 0.006 0.856 0.026 1.275 0.060 0.949 0.115 0.046 0.224 59 R 0.006 1.840 0.071 0.877 0.089 0.705 0.195 0.045 0.109 62 R 0.017 3.042 0.101 0.943 0.141 1.653 0.232 0.113 0.271 63 R 0.025 2.235 0.064 12.295 0.401 2.479 0.714 0.115 0.595 65 R 0.012 1.664 0.092 1.613 0.061 1.235 0.150 0.048 0.776 66 R 0.013 1.400 0.154 0.308 0.012 0.115 0.079 0.028 0.076 71 R 0.002 0.607 0.082 0.564 0.038 0.331 0.065 0.014 0.086 72 R 0.005 0.406 0.050 0.176 0.006 0.124 0.050 0.015 0.094 73 R 0.004 2.230 0.174 1.064 0.038 0.171 0.109 0.060 0.230 76 R 0.012 2.403 0.166 0.653 0.041 0.498 0.148 0.064 0.153 86 R 0.023 1.361 0.124 14.123 0.329 2.063 0.378 0.103 0.328 88 R 0.010 1.828 0.102 2.042 0.124 2.549 0.163 0.034 0.201 90 R 0.012 1.357 0.067 1.912 0.150 0.702 0.182 0.048 0.280 91 R 0.008 1.261 0.056 0.664 0.038 0.495 0.096 0.023 0.152 92 R 0.005 0.224 0.026 0.049 0.003 0.043 0.024 0.008 0.019 125 R 0.005 0.801 0.028 1.257 0.077 2.166 0.128 0.012 0.088 222 R 0.016 0.774 0.052 5.187 0.091 1.641 0.313 0.021 0.582 227 R 0.003 0.801 0.060 0.119 0.005 0.042 0.052 0.014 0.030 PCT/EP2012/052232 114 Table 15 (continued): Patients' BMK values for the genes CCL21, CLDN1, FOXO1, G1P2, G1P3, IF127, LGALS3BP and OAS2 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R CCL21 CLDN1 FOXO1 G1P2 G1P3 IF127 IFITMI LGALS3BP OAS2 or RR) 306 R 0.018 2.612 0.128 3.568 0.137 2.021 0.835 0.273 0.516 344 R 0.067 2.500 0.489 0.646 0.099 0.892 0.196 0.037 0.211 346 R 0.004 2.088 0.342 0.601 0.340 1.521 0.718 0.133 0.324 366 R 0.009 2.031 0.161 0.159 0.014 0.042 0.088 0.015 0.049 501 R 0.016 2.007 0.034 1.424 0.112 0.801 0.113 0.032 0.290 502 R 0.009 0.953 0.065 0.465 0.020 0.056 0.055 0.022 0.104 503 R 0.004 1.790 0.390 0.755 0.067 0.160 0.133 0.024 0.128 504 R 0.012 0.796 0.078 0.563 0.037 0.232 0.237 0.044 0.091 506 R 0.019 0.815 0.053 0.678 0.028 0.847 0.084 0.016 0.085 508 R 0.013 0.892 0.070 0.143 0.004 0.046 0.072 0.028 0.073 513 R 0.006 0.223 0.118 0.280 0.045 0.265 0.139 0.035 0.081 523 R 0.018 0.901 0.049 0.134 0.005 0.221 0.039 0.028 0.034 528 R 0.022 1.945 0.140 1.261 0.099 1.129 0.304 0.082 0.807 529 R 0.012 1.444 0.048 1.210 0.130 1.153 0.137 0.031 0.149 530 R 0.009 2.266 0.295 0.144 0.015 0.073 0.168 0.052 0.053 532 R 0.010 2.049 0.340 0.245 0.074 0.161 0.066 0.029 0.020 535 R 0.025 0.002 0.060 15.348 0.324 3.732 0.431 0.053 0.394 536 R 0.010 1.784 0.105 0.313 0.025 0.588 0.014 0.015 0.066 537 R 0.007 0.946 0.066 0.080 0.006 0.102 0.050 0.012 0.029 538 R 0.014 1.966 0.064 5.152 0.237 2.282 0.403 0.068 0.452 546 R 0.012 1.670 0.253 0.236 0.030 0.309 0.134 0.056 0.108 556 R 0.008 0.859 0.017 6.869 0.071 2.437 0.252 0.022 6.084 PCT/EP2012/052232 115 Table 15 (continued): Patients' BMK values for the genes CCL21, CLDN1, FOXO1, G1P2, G1P3, IF127, LGALS3BP and OAS2 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R CCL21 CLDN1 FOXO1 G1P2 G1P3 IF127 IFITMI LGALS3BP OAS2 or RR) 560 R 0.008 1.647 0.071 3.249 0.093 1.753 0.166 0.038 0.220 565 R 0.006 0.829 0.134 4.857 0.065 2.056 0.220 0.022 0.503 567 R 0.006 1.235 0.050 2.497 0.101 1.050 0.108 0.068 0.134 568 R 0.002 0.740 0.018 0.055 0.005 0.087 0.035 0.007 0.023 569 R 0.003 2.313 0.086 0.092 0.016 0.167 0.177 0.047 0.072 570 R 0.006 2.028 0.493 0.118 0.023 0.186 0.221 0.028 0.085 571 R 0.026 2.378 0.205 0.609 0.033 0.063 0.116 0.057 0.155 572 R 0.005 1.161 0.038 2.129 0.115 2.021 0.148 0.020 0.125 575 R 0.036 3.706 0.202 15.348 0.821 9.254 1.414 0.360 1.157 577 R 0.007 2.035 0.052 0.224 0.022 0.047 0.116 0.093 0.046 581 R 0.006 1.357 0.082 1.210 0.072 0.332 0.085 0.036 0.138 583 R 0.015 1.117 0.045 1.297 0.106 1.641 0.386 0.146 0.113 585 R 0.017 1.575 0.064 3.411 0.155 0.886 0.403 0.121 0.228 598 R 0.008 0.732 0.017 0.200 0.024 0.486 0.053 0.009 0.040 601 R 0.005 1.275 0.056 1.053 0.253 1.759 0.362 0.069 0.157 604 R 0.011 1.310 0.063 1.061 0.123 3.470 0.248 0.124 0.003 605 R 0.009 1.153 0.064 5.315 0.258 2.338 0.311 0.048 0.023 613 R 0.007 3.000 0.087 0.108 0.011 0.053 0.113 0.035 0.049 614 R 0.011 1.705 0.249 0.135 0.005 0.084 0.056 0.057 0.041 629 R 0.003 2.437 0.142 0.074 0.009 0.588 0.065 0.059 0.036 639 R 0.005 1.553 0.033 0.109 0.005 0.043 0.086 0.021 0.021 6 NR 0.004 0.237 0.355 0.685 0.041 0.388 0.083 0.044 0.092 PCT/EP2012/052232 116 Table 15 (continued): Patients' BMK values for the genes CCL21, CLDN1, FOXO1, G1P2, G1P3, IF127, LGALS3BP and OAS2 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R CCL21 CLDN1 FOXO1 G1P2 G1P3 IF127 IFITMI LGALS3BP OAS2 or RR) 46 NR 0.021 4.112 0.295 8.969 0.209 3.329 0.511 0.164 0.850 58 NR 0.016 1.966 0.072 13.881 0.476 5.464 0.467 0.149 0.512 75 NR 0.033 4.170 0.215 8.664 0.607 1.993 0.877 0.265 1.035 80 NR 0.018 2.329 0.144 0.963 0.031 0.288 0.131 0.032 0.088 83 NR 0.073 5.776 0.220 6.169 0.730 7.701 0.892 0.164 1.266 145 NR 0.023 2.959 0.186 7.490 0.360 2.770 0.345 0.193 0.601 167 NR 0.018 1.500 0.098 5.046 0.434 1.765 0.283 0.099 0.750 308 NR 0.022 2.305 0.144 6.658 0.545 6.821 0.613 0.251 0.485 509 NR 0.025 1.464 0.060 6.126 0.198 1.459 0.312 0.045 0.521 516 NR 0.017 1.591 0.309 3.160 0.199 1.007 0.222 0.065 0.266 521 NR 0.052 4.317 1.253 4.362 1.315 4.377 1.315 0.576 0.774 524 NR 0.009 2.523 0.094 0.286 0.017 0.108 0.072 0.042 0.068 526 NR 0.028 1.741 0.084 5.260 0.293 3.106 0.413 0.130 0.459 527 NR 0.045 1.636 0.238 9.318 0.432 3.811 0.468 0.457 0.904 534 NR 0.017 2.742 0.058 2.129 0.259 0.990 0.402 0.048 0.491 549 NR 0.023 2.878 0.314 2.297 0.245 1.490 0.415 0.142 0.370 574 NR 0.031 3.986 0.512 4.959 0.305 3.945 1.042 0.241 1.343 582 NR 0.012 2.549 0.077 4.675 0.120 0.470 0.143 0.037 0.277 596 NR 0.017 1.945 0.077 0.914 0.079 0.622 0.189 0.059 0.221 602 NR 0.010 1.479 0.055 0.091 0.005 0.087 0.047 0.019 0.040 PCT/EP2012/052232 117 Table 15 (continued): Patients' BMK values for the genes CCL21, CLDN1, FOXO1, G1P2, G1P3, IF127, LGALS3BP and OAS2 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R CCL21 CLDN1 FOXO1 G1P2 G1P3 IF127 IFITMI LGALS3BP OAS2 or RR) 618 NR 0.025 3.399 0.137 4.757 0.475 4.070 0.603 0.221 0.601 619 NR 0.011 1.253 0.080 1.542 0.387 1.444 0.432 0.151 0.266 636 NR 0.048 6.169 0.089 18.765 0.969 6.498 1.659 0.459 1.729 645 NR 0.004 0.009 0.642 0.069 0.009 0.081 0.071 0.017 0.011 646 NR 0.018 2.305 0.284 35.383 0.463 11.551 1.210 0.529 1.526 647 NR 0.028 3.021 0.287 20.393 0.297 5.979 0.936 0.097 1.564 649 NR 0.034 3.543 0.286 3.918 0.109 1.157 0.334 0.051 0.332 650 NR 0.022 3.272 0.254 12.381 0.465 2.181 0.889 0.241 0.821 651 NR 0.009 3.238 0.392 4.773 0.416 6.298 0.398 0.146 0.306 657 NR 0.013 2.362 0.114 2.949 0.202 2.258 0.236 0.075 0.315 658 NR 0.009 1.361 0.116 0.322 0.016 0.279 0.125 0.019 0.101 659 NR 0.006 1.847 0.127 0.690 0.026 0.094 0.085 0.016 0.175 660 NR 0.021 2.979 0.150 6.612 0.399 2.523 0.334 0.074 0.286 662 NR 0.022 1.206 0.059 5.483 0.087 1.619 0.180 0.049 0.053 664 NR 0.029 3.106 0.116 2.704 0.184 0.917 0.371 0.063 0.002 665 NR 0.014 0.298 0.040 0.576 0.067 0.505 0.059 0.042 0.057 666 NR 0.028 3.387 0.159 5.796 0.351 3.127 0.326 0.095 0.774 67 NR 0.006 1.404 0.088 0.314 0.011 0.146 0.051 0.034 0.069 563 NR 0.005 1.608 0.050 0.235 0.010 0.070 0.128 0.038 0.058 573 NR 0.028 2.809 0.446 6.892 0.403 2.576 0.488 0.134 0.660 599 NR 0.005 0.505 0.050 1.248 0.065 1.516 0.109 0.013 0.079 641 NR 0.012 1.636 0.174 2.454 0.148 1.828 0.187 0.035 0.259 PCT/EP2012/052232 118 Table 15 (continued to end): Patients' BMK values for the genes CCL21, CLDN1, FOXO1, G1P2, G1P3, IF127, LGALS3BP and OAS2 (Ct normalized in accordance with the 2 -Act method) Status Patient (NR, R CCL21 CLDN1 FOXO1 G1P2 G1P3 IF127 IFITMI LGALS3BP OAS2 or RR) 642 NR 0.046 2.949 0.200 6.635 0.382 2.990 0.917 0.429 0.877 49 RR 0.018 3.986 0.084 1.905 0.261 0.549 0.480 0.128 0.467 56 RR 0.011 1.087 0.092 3.249 0.270 1.741 0.264 0.062 0.252 60 RR 0.015 2.462 0.128 7.945 0.230 4.332 0.435 0.059 0.201 87 RR 0.021 3.605 0.123 11.432 0.361 3.193 0.521 0.104 1.057 505 RR 0.006 0.844 0.010 4.084 0.188 3.042 0.226 0.082 0.357 514 RR 0.011 2.488 0.352 0.525 0.031 0.200 0.184 0.082 0.194 531 RR 0.002 0.099 0.091 0.103 0.013 0.091 0.049 0.006 0.025 533 RR 0.008 1.548 0.059 1.145 0.073 0.702 0.072 0.046 0.092 543 RR 0.040 1.173 0.131 1.548 0.098 2.085 0.570 0.112 0.139 554 RR 0.005 1.772 0.100 0.148 0.009 0.094 0.035 0.029 0.062 557 RR 0.007 1.597 0.065 3.317 0.090 1.419 0.199 0.057 0.213 558 RR 0.002 0.730 0.070 0.289 0.019 0.331 0.032 0.015 0.048 559 RR 0.005 0.730 0.082 0.871 0.047 0.467 0.054 0.028 0.025 562 RR 0.007 2.828 0.102 3.668 0.242 1.676 0.266 0.112 0.271 576 RR 0.010 1.625 0.113 5.187 0.272 1.625 0.590 0.136 0.164 579 RR 0.004 1.206 0.053 0.147 0.006 0.034 0.060 0.019 0.021 588 RR 0.002 2.063 0.044 0.053 0.004 0.018 0.033 0.127 0.028 589 RR 0.012 2.694 0.048 7.387 0.280 2.305 0.387 0.094 0.435 591 RR 0.013 1.693 0.137 2.378 0.125 1.347 0.161 0.140 0.246 592 RR 0.020 2.297 0.191 6.000 0.204 1.010 0.468 0.158 0.553 643 RR 0.014 2.063 0.189 0.390 0.015 0.088 0.115 0.058 0.081 PCT/EP2012/052232 119 Table 16: Kits for protein measurements Tags IL8 LGALS3BP MDK CCL21 QUANTIKINE HUMAN HUMAN 90K/MAb-2 BP HUMAN MIDKINE EIA kit CCL21/6Ckine CXCL8/IL-8 ELISA ELISA IMMUNOASSAY IMMUNOASSAY Supplier R&D Systems Abnova Abnova R&D Systems Reference D8000C KA0140 KA0028 V.02 D6COO ELISA type Sandwich Sandwich Sandwich Sandwich Serum, plasma, serum, cell culture Serum, plasma, tissue, Types of samples serum, plasma cell culture medium supernatant cell culture medium Test volume 50 L 20pL 25pL 100pL PCT/EP2012/052232 120 Table 16 (continued): Kits for protein measurements Solid phase mAb anti-IL8 mAb anti-LGALS3BP pAb anti MDK mAb anti-CCL21 mAb-HRP anti Conjugate pAb-HRP anti-IL8 LGALS3BP pAb-biotin anti-MDK pAb-HRP anti-CCL21 Sensitivity 3.5 pg/mL 0.92 ng/mL 0.33 ng/mL 9.9 pg/niL Detection range 31.2-2000 pg/mL 12.5 to 200 ng/mL 2 - 10 ng/mL 91-371pg/mL PCT/EP2012/052232 121 Table 16 (continued to end): Kits for protein assays Human recombinant IL8, , no cross reaction with ANG, AR, CNTF, b ECGF, EGF, Epo, acidic FGF, basic FGF, FGF-4, FGF-5, FGF-6, GCSF, Specificity GM-CSF, Human LGALS3BP Human MIDKINE Human CCL21 GROa, GROb, GROg, sgpl30, HBEGF, HGF, 1-309, IFN-g , IGF-I, IGF-II, IL-la , IL-lb, IL-Ira, IL-I sRI pAb = polyclonal antibody mAb = monoclonal antibody PCT/EP2012/052232 122 Administration of antiviral treatment and analysis of patient's response: After HBP and removing serum, each patient received an antiviral treatment which is currently considered to be the standard treatment for hepatitis C, namely a treatment based 5 on a combination of two antiviral agents, namely alpha interferon and ribavirin. In the context of the test described here, all of the patients received the following treatment: - either: 10 o pegylated alpha-2b interferon (PEG-INTRON@; Schering Plough Corporation; Kenilworth, NJ; U.S.A.) in a dose of 1.5 g/kg/week, and o ribavirin (REBETOL@; Schering Plough Corporation; Kenilworth, NJ; U.S.A.) in a dose of: - 800 to 1200 mg/kg/day for those patients who had been infected 15 with at least one genotype 1 and/or 4 and/or 5 of HCV, or in a dose of - 800 mg/kg/day for those patients who had been infected with at least one genotype 2 and/or 3 of HCV, - or: 20 o pegylated alpha-2a interferon (PEGASYS@; Roche Corporation; F. Hoffmann-La Roche Ltd.; Basel, Switzerland) in a dose of 180 g/kg/week, and o ribavirin (COPEGUS@; Roche Corporation; F. Hoffmann-La Roche Ltd.; Basel, Switzerland) in a dose of 1000 to 1200 mg/kg/day. 25 The treatment was administered for 24 weeks for those patients who had been infected with at least one genotype 2 and/or 3 of HCV, and for 48 weeks for those patients who had been infected with at least one genotype 1 and/or 4 and/or 5 of HCV. 30 The viral load was measured in week 24, at the end of treatment and 6 months after treatment had ended by quantification of the HCV RNAs present in the serum from each patient, with the aid of the VERSANT@ HCV RNA 3.0 (bDNA) ASSAY HCV RNA quantification test from Siemens Healthcare Diagnostics (quantification limit = 615 7 690 000 IU/mL).

PCT/EP2012/052232 123 Each patient was classified as a function of their response to treatment as measured by the test for assaying the seric viral load of HCV. 5 A patient was considered to be: - a patient who was a responder to treatment (patient classified as R), when the viral load of HCV had become undetectable in the patient's blood at the end of treatment and it remained undetectable for 6 months after treatment had been stopped; - a patient who was a non-responder to treatment (patient classified as NR), when 10 the viral load of HCV remained detectable in the patient's blood at the end of treatment; - a patient who was a responder-relapser (patient classified as RR), when the viral load of HCV had become undetectable in the patient's blood at the end of treatment, but which became detectable again 6 months after treatment had been stopped. 15 The viral load of HCV was considered to be undetectable in the patient's blood when the measurement of the HCV RNAs in the serum of patient gave a value of less than 12 International Units (IU) per mL of serum, as measured with the aid of the VERSANT@ HCV RNA 3.0 (bDNA) ASSAY kit from Siemens Healthcare Diagnostics as indicated 20 above. Three sub-populations, or cohorts, were thus formed (sub-population of R patients, sub population of NR patients and sub-population of RR patients). 25 2. Comparison of measurement values for the sub-populations NR and R in order to set up a multivariate classification model The measurement values obtained in § 1 above for the sub-populations "responders" (R) and "non-responders" (NR) were compared in order to construct a multivariate 30 classification model which, starting from the combination of these values, classifies the test patient among the patients who have a high probability of responding to anti-HCV treatment (class R) or among the class of patients who have a high probability of not responding to anti-HCV treatment (NR class).

PCT/EP2012/052232 124 The measurement values obtained at § 1 above for the "responders-relapsers" (RR) sub population were also compared with measurement values obtained for the R and NR sub populations. It was observed that the RR sub-population was very distinct from that of R; RR patients are primarily classified as R. 5 A classification model may, for example, be obtained by following a multivariate statistical analysis method or a multivariate mathematical analysis method. mROC models: 10 A suitable multivariate mathematical analysis method is the mROC method (multivariate Receiver Operating Characteristic method). By using the measurement values obtained in § 1 above for the R and NR sub-populations, 15 mROC models were constructed as described in Kramar et al. 1999 and Kramar et al. 2001. To this end, the mROC version 1.0 software, available commercially from the designers (Andrew Kramar, Antoine Fortune, David Farragi and Benjamin Reiser), was used. 20 Andrew Kramar and Antoine Fortune may be contacted at or via the Unit6 de Biostatistique du Centre R6gional de Lutte contre le Cancer (CRLC) [Biostatistics Unit, Regional Cancer Fighting Centre] Val d'Aurelle - Paul Lamarque (208, rue des Apothicaires; Parc Eurom6decine; 34298 Montpellier Cedex 5; France). 25 David Faraggi and Benjamin Reiser may be contacted at or via the Department of Statistics, University of Haifa (Mount Carmel; Haifa 31905; Israel). Starting from the input measurement data, the mROC method generates a decision rule in the form of a linear function [Z = f(BMK 1 , BMK 2 , BMK 3 ,... )] of the type Z = a.BMK 1 + 30 p.BMK 2 + y.BMK 3 ... , where BMK 1 , BMK 2 , BMK 3 ... are the measurement values for the levels of expression of each of the selected genes, and PCT/EP2012/052232 125 the user identifies the reference or threshold value (6) which provides this combination with the best performance. This function and this threshold constitute a multivariate classification model. 5 The function f calculated by the mROC method was then applied to the measurement values of the level of expression of the genes BMKI, BMK 2 , BMK 3 ... measured for a test subject p. The value Z calculated for a test subject p was then compared with the threshold 6. 10 For example, when the mean value of the combination of the levels of expression of said selected genes in the cohort of "R" individuals is lower than that of the cohort of individuals "NR": - if Z > 6, the test is positive: the subject p is declared to be a NR patient (the subject is 15 predicted to be a non-responder to treatment); and - if Z < 6, the test is negative: the subject p is declared to be a R patient (the subject is predicted to be a responder to treatment). Conversely, when the mean value of the combination of the levels of expression of said 20 selected genes in the cohort of "R" individuals is higher than that of the cohort of "NR" individuals: - if Z > 6, the test is negative: the subject p is declared to be a R patient (the subject is predicted to be a responder to treatment); and - if Z < 6, the test is positive: the subject p is declared to be a NR patient (the subject is 25 predicted to be a non-responder to treatment). WKNN models: A suitable multivariate statistical analysis method is the WKNN (Weighted k Nearest 30 Neighbours) method. WKNN models were constructed as described by Hechenbichler and Schliep, 2004 using the measurement values obtained in § 1 above for the sub-populations R and NR.

PCT/EP2012/052232 126 In outline, a WKNN method attributes each new case (y,x) to the class I of maximum weight in a neighbourhood of k neighbours in accordance with the formula: I = max, K(D(x,x))I(y(I) r 5 where r represents the index of the clinical classes of interest (in fact, sub-population R or sub-population NR), and is equal to 0 or 1. In order to construct the WKNN models, R software (WKNN library), which is freely available from http://www.r-project.org/, was used. The following control parameters were 10 used: (see, for example, combination No. 1) - Kernel (K): inverse; - Parameter of Minkowski distance (D): 1; - Number of neighbours (k): 5; 15 or (see, for example, combination No. 2) - Kernel (K): triangular; - Parameter of Minkowski distance (D): 2; - Number of neighbours (k): 4; or (see, for example, combination No. 7) 20 - Kernel (K): triangular; - Parameter of Minkowski distance (D): 1; - Number of neighbours (k): 6; or (see, for example, combination No. 8) - Kernel (K): cosine; 25 - Parameter of Minkowski distance (D): 1; - Number of neighbours (k): 7; or (see, for example, combination No. 9) - Kernel (K): triweight; - Parameter of Minkowski distance (D): 1; 30 - Number of neighbours (k): 14; or (see, for example, combination No. 10) - Kernel (K): cosine; PCT/EP2012/052232 127 - Parameter of Minkowski distance (D): 1; - Number of neighbours (k): 6; or (see, for example, combination Nos. 11 and 13) - Kernel (K): inverse; 5 - Parameter of Minkowski distance (D): 2; - Number of neighbours (k): 3; or (see, for example, combination No. 12) - Kernel (K): triangular; - Parameter of Minkowski distance (D): 1; 10 - Number of neighbours (k): 10; The WKNN models constructed in this manner were then used to determine the status, R or NR, of the subjects by inputting the measurement values for these subjects into the WKNN models constructed in this manner. 15 The measurement values for the levels of expression of the selected genes of a test subject p were compared with those of these neighbours (k). The WKNN model calculates the weight which has to be attributed to the "R sub-population" class and that which has to be attributed to the " NR sub-population " for this subject p. The subject p is then classified 20 by the WKNN model into the major class, for example into the "NR sub-population" class if the weight of the "R sub-population" and "NR sub-population" classes calculated by the WKNN method are 0.3 and 0.7 respectively. The LOOCV ("Leave-One-Out-Cross-Validation") error is as defined by Hastie, 25 Tibishirani and Friedman, 2009. Random Forest or RF models: Random Forest or RF models were constructed using the measurement values obtained in 30 § 1 above for the R and NR sub-populations as described in Breiman in 2001, Liaw and Wiener in 2002. To this end, R software, which is freely available from http://www.r-project.org/, was used.

PCT/EP2012/052232 128 The following parameters were used: - NumberOfTrees = 500; - NumberOfDescriptors = sqrt(D). 5 The digital data listed in the output file from R could be used to evaluate the signatures by calculating the following parameters: calculation of the True Positive (TP), False Positive (FP), True Negative (TN) and False Negative (FN) values, see below. 10 The data extracted from the output file for the RF models constructed thereby had the following form: "OOB estimate of error rate: Confusion matrix: NR R Classification error 15 NR TP FN NR classification, error rate R FP TN R classification, error rate ROC score (out-of-bag data): ROC score for predicted samples" OB is the acronym for Out-Of-Bag, and represents an evaluation of the error. 20 These output data directly indicate the values for the parameters TP (number of NR patients who have been classified as NR), FP (number of R patients who have been classified as NR), TN (number of R patients who have been classified as R) and FN (number of NR patients who have been classified as R). 25 The formulae below can be used to calculate the values for sensitivity (Se), specificity (Spe), positive predictive value (PPV), and negative predictive value (NPV): Se = TP/ (TP + FN); Sp = TN / (TN + FP); 30 PPV =TP / (TP + FP); NPV= TN / (TN + FN). The output data also directly indicate the error rate and the ROC score of the constructed model.

PCT/EP2012/052232 129 The RF models constructed in this manner were then used to determine the hepatic fibrosis score of test subjects. The measurement values of the levels of expression of the genes of these test subjects were input into a RF model, which generated output data as presented 5 above and classified the test subject into the "R sub-population" or "NR sub-population" class. The LOOCV error was as defined by Hastie, Tibishirani and Friedman, 2009. 10 Neural network models Another appropriate method for multivariate statistical analysis is a neural network method. In brief, a neural network comprises an orientated weighted graph the nodes of which symbolize neurons. The network is constructed from sub-population measurement 15 values (in this case R versus NR) and is then used to determine to which class (in this case R or NR) a new element (in this case a test patient p) belongs. Neural network models were constructed as described by Intrator and Intrator 1993, Riedmiller and Braun 1993, Riedmiller 1994, Anastasiadis et al. 2005 using the 20 measurement values obtained in § 1 above for the R and NR sub-populations; see http://cran.r-project.org/web/packages/neuralnet/index.html. To this end, R software which is freely available from http://www.r-project.org/, was used (version 1.3 of Neuralnet, written by Stefan Fritsch and Frauke Guenther, following the 25 work by Marc Suling). The following computation options were used: "NumberOfHiddenNodes = 1 and 2 WeightDecayFactor = 0.001 30 CrossValidate = True CrossValidationFolds = 5 MaxNumberIterations = 2000 MaxNumberWeights= 2000".

PCT/EP2012/052232 130 For each of the combinations, the confusion matrix was extracted in the following format: "Cross-validation results (5-fold): Nodes Decay ROC Score Best 1 1 5 2 2 Contingency Table (best CV model): Predicted Actual R NR R TN FP 10 NR FN TP In this example, it will be observed that the best model is model 2, indicated by "***" in the "ScoreBest" column. 15 These output data directly indicate the values for the parameters TP (number of NR patients who have been classified as NR), FP (number of R patients who have been classified as NR), TN (number of R patients who have been classified as R) and FN (number of NR patients who have been classified as R). 20 The following parameters were evaluated: the sensitivity (Se), the specificity (Spe), the positive predictive value (PPV) and the negative predictive value (NPV) (see formulae for Se, Spe, PPV and NPV above). The ROC score was extracted directly from the output file on the line identified by "***" 25 which corresponded to the best model. The error was calculated by the following formula: Classerr = (FP + FN) / (FP + TP + FN + TN). The neural network models constructed thereby were then used to determine whether a test subject had a high probability of responding or, in contrast, of not responding to anti 30 HCV treatment. The measurement values for the levels of expression of the genes of these test subjects were entered into a neural network model which generated output data as presented above and classified the test subject into the "R sub-population" or "NR sub population" class.

PCT/EP2012/052232 131 3. Examples of classification models obtained: The inventors have thus identified the genes for which the levels of expression constitute biomarkers which, when taken in combination, are pertinent to determining the status of 5 "responder" (R) or "non-responder" (NR) of a subject. These genes are the following twenty-eight genes: HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXOl, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, 10 STAT1, STMN2 and USP18. The majority of these genes code for proteins with a localization which is not transmembrane in nature, or at least are not exclusively so. The majority of these genes thus code for proteins which are susceptible of being detected in a biological fluid of the 15 subject such as the blood, serum or plasma. This is in fact the case with the twenty-eight genes listed above, with the exception of the following seven genes which code for strictly membrane proteins: CLDN1, G1P3, IF127, IFITMI, ITGA2, OCLN and PLSCR1. The inventors have also identified that the most pertinent combinations comprise: 20 - at least two genes from among HERC5, IL8 and STAT2; and - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFITI, IFIT4, IFITMI, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18. 25 By way of illustration, examples of appropriate combinations of biomarkers in particular comprise combinations of four or five biomarkers (combinations of the levels of expression of four or five genes) presented in Table 2 above, in the description section. Examples of classification models which may be used with these combinations of 30 biomarkers are presented in Tables 3 to 5 above (combination of the levels of transcription of four or five genes selected in accordance with the invention), see also the examples below.

PCT/EP2012/052232 132 The predictive combinations of the invention are combinations of the levels of gene expression selected as indicated above. However, it may be elected to involve one or more factors in these combinations other 5 than the levels of expression of these genes, in order to combine this or these other factors and the levels of expression of the selected genes into one decision rule. This or these other factors are preferably selected so as to construct a classification model the predictive power of which is further improved compared with the model which did not 10 comprise this or these other factors (see example below). This or these other factors may, for example, be clinical, biological, or virological factors, for example: - one or more clinical factors, such as sex (feminine F or masculine M), age at the date 15 of sampling (for example, age at the date of HBP, age at the date of hepatic cytopuncture, age at the date of sampling blood, serum, plasma or urine), age at the date of contamination, age at the treatment start date, body mass index (BMI), insulin sensitivity index (HOMA), diabetes, alcohol consumption, degree of steatosis, mode of contamination, Metavir activity, or hepatic fibrosis score measured using the 20 Metavir system (Metavir F score) or using the Ishak system, and/or - one or more biological factors other than the levels of expression of said selected genes, such as concentration of haptoglobin (Hapto), concentration of apolipoprotein Al (ApoAl), total quantity of bilirubin (BLT), concentration of gamma glutamyl 25 transpeptidase (GGT), concentration of aspartate aminotransferase (AST), concentration of alanine aminotransferase (ALT), platelet count (PLQ), quantity of prothrombin (TP), quantity of HDL cholesterol (Chol-HDL), total quantity of cholesterol, concentration of ferritin (Ferritin), level of glycaemia (glycaemia), concentration of peptide C, quantity of insulin (insulinaemia), concentration of 30 triglycerides (TG), quantity of albumin, transferrin saturation (TSAT), or concentration of alkaline phosphatase (ALP); and/or - one or more virological factors, such as viral genotype, duration of infection, viral load before treatment (VLbeforeTTT), viral load assayed for the patient at the PCT/EP2012/052232 133 treatment start date (viral load at DO), viral load assayed for the patient at the date of sampling (viral load at HBP, viral load at the date of hepatic cytopuncture, or viral load at the date of sampling blood, serum, plasma or urine). 5 EXAMPLE 2: RNA from hepatic biopsy puncture (HBP) / applications of constructed models to test patients 2a) Example of application of the combination of the levels of expression (RNA) of the genes HERC5, IF144, IL8 and MDK (combination No.1 in Table 2 above): 10 Using the WKNN method (see Example 1 above), the LOOCV error associated with the combination of the levels of transcription (RNA) of the genes HERC5, IF144, IL8 and MDK (combination No.1 in Table 2 above) is 15 (see Table 3 above). 15 The best performances for this combination using the WKNN method (computed over the population of 107 responders and non-responders presented in Table 11 above) are as follows: sensitivity (Se) = 84%; specificity (Sp) = 86% (see Table 3 above). 20 The model parameters used for the WKNN method were as follows: Kernel (K): inverse Parameter of Minkowski distance (D): 1 Number of neighbours (k): 5 25 An example of a prediction over 20 subjects (human patients) is given in Table 17 below, which presents the measurement values for the levels of expression of the selected genes (BMK values obtained by the 2 -Act method; see Example 1 above). 30 One or more clinical, biological and virological factors may be combined with these biomarkers (levels of expression of four genes), and give rise to a decision rule the predictive power of which is even better than that of the rule presented above. The following PCT/EP2012/052232 134 Table 18 presents examples of such clinical, biological and virological factors, as well as their values for the test subjects of Table 17. ND = not determined VL = viral load 5 PCT/EP2012/052232 135 Table 17: Example of application of a classification model based on the combination of the levels of expression of the genes HERC5, IF144, IL8 and MDK (combination No.1 of Table 2 above) 5 WKNN model Status, No. of (kernel = inverse; parameter of Minkowski distance = 1; k = 5) R or NR, test as WVKNN determined subject HERC5 IF144 IL8 MDK afdeteried prediction afe treatment 521 1.670 0.785 3.732 1.676 NR NR 509 0.719 0.102 1.157 0.172 NR NR 308 0.886 0.154 0.853 0.446 NR NR 167 0.653 0.075 0.923 0.521 NR NR 145 1.283 0.183 0.755 0.255 NR NR 80 0.570 0.057 0.745 0.030 NR NR 75 1.035 0.319 0.908 1.248 NR NR 58 0.607 0.149 0.616 0.631 NR NR 46 0.563 0.177 0.911 0.308 NR NR 6 0.247 0.046 2.648 0.688 NR NR 91 0.149 0.042 0.040 0.051 R R 90 0.342 0.046 0.089 0.115 R R 86 0.259 0.086 0.451 0.060 R R 76 0.293 0.067 1.257 0.045 R R 72 0.097 0.010 0.064 0.001 R R 71 0.255 0.015 0.228 0.021 R R 65 0.258 0.055 0.168 0.105 R R 62 0.280 0.074 0.058 0.043 R R 59 0.269 0.039 0.329 0.090 R R 50 0.144 0.018 0.444 0.132 R R PCT/EP2012/052232 136 Table 18: Examples of clinical, virological, biological data for test subjects of Table 17 Age Steatosis Ishak Metavir Age at No. of Sex at BMI HOMA Diabetes Alcohol (score 0 to Mode of Metavir fibrosis fibrosis start of test subject HBP (kg/m2) (g/day) 3) contamination activity score score treat ment 521 M 58.4 23.8 ND No 30 0 Transfusion 1 4 3 58.9 509 F 48.2 20.8 1.4 No ND 0 Transfusion 1 4 3 48.4 308 M 34.9 27.7 ND No 0 0 Endemic area (Egypt) 1 1 1 35.4 167 F 48.6 37.8 2.3 No ND 0 Transfusion 2 2 1 48.8 145 M 40.8 28.4 2.3 No 10 0 Toxicomania 1 2 1 42.2 80 F 36.6 31.6 ND No 0 2 Transfusion 1 1 2 37.2 75 F 53.0 27.9 ND No 0 0 ND 2 2 2 53.3 58 M 48.1 25.2 2.1 No 0 0 Toxicomania 1 3 2 48.2 46 M 54.4 26.9 2.4 No 5 1 Unknown 2 2 1 55.1 6 M 59.0 19.6 1.2 No 30 0 Toxicomania 1 3 2 59.3 91 F 47.8 21.4 ND No 0 0 ND 1 2 1 48.0 90 F 58.7 25.2 ND No 0 2 Transfusion 1 3 2 59.0 86 M 50.1 28.4 ND Yes 40 1 Unknown 1 1 1 50.3 76 F 53.5 22.8 ND No 0 0 Unknown 2 3 2 53.6 72 M 37.2 29.0 2.2 No 0 2 Endemic area (Egypt) 1 3 2 37.3 71 M 43.3 20.2 0.6 No 0 0 Unknown 1 2 1 43.3 65 M 35.6 29.6 3.9 No 0 0 Endemic area (Egypt) 1 2 1 35.7 62 M 61.6 27.4 8.7 ND 50 1 Nosocomial 1 3 2 61.7 59 F 50.9 23.7 1.5 No 0 0 Transfusion 1 2 1 51.3 50 M 48.1 28.4 3.4 No 0 1 Endemic area (Egypt) 2 2 2 48.7 PCT/EP2012/052232 137 Table 18 (continued): Examples of clinical, virological, biological data for test subjects of Table 17 No. of test Duration subett Genotype nfetion VL HBP (copies/mL. 10E3) VLbeforeTTT (copies/mL. 10E3) 521 1 34.8 14654 14432 509 1 49.1 8779 8779 308 4 ND 423 423 167 4 ND 12616 12616 145 1 25.8 26631 18304 80 1 38.1 7606 12932 75 1 ND 3276 2347 58 1 30.9 5079 5079 46 1 ND 7903 9103 6 1 33.7 10450 10450 91 1 ND 3902 3902 90 1 ND 515 515 86 2 ND 57 57 76 1 ND 19227 19927 72 4 ND 1155 1155 71 4 ND 6935 6935 65 4 ND 1135 450 62 2 ND 14 185 59 5 26.8 524 1120 50 4 ND 445 445 PCT/EP2012/052232 138 Table 18 (continued): Examples of clinical, virological, biological data for test subjects of Table 17 No. of alpha2macroglobulin Haptoglobin Apo Al BLT GGT AST ALT PLQ TP (%) Chol-HDL Test subject (g/L) (g/L) (g/L) (Imol/L) (U/L) (U/L) (U/L) (x 10E3/mm 3 ) (mmole/L) 521 ND ND ND 11 127 83 166 189 100 1.62 509 4.24 0.39 1.89 13 43 154 243 121 86 1.88 308 ND ND ND 16 246 30 36 214 100 ND 167 2.9 0.54 1.64 15 378 163 144 183 78 1.28 145 2.1 0.48 1.56 11 133 36 92 225 ND 1.25 80 ND ND ND 18 132 44 67 311 92 ND 75 ND ND ND 12 135 64 50 233 100 2.15 58 3.16 0.63 1.79 4 162 62 93 222 91 1.43 46 4.77 0.21 1.48 14 174 91 125 216 100 0.89 6 2.97 0.3 1.82 13 78 50 69 200 99 1.63 91 ND ND ND 9 39 64 79 353 99 ND 90 ND ND ND 11 21 61 119 359 100 ND 86 ND ND ND 13 287 53 105 160 100 ND 76 ND ND ND 14 81 90 157 209 96 ND 72 1.5 0.96 1.44 15 23 31 71 249 98 0.76 71 1.89 0.47 1.92 7 22 37 58 177 94 1.63 65 4.07 1.15 1.17 8 46 61 152 210 92 0.72 62 4.37 0.94 1.37 12 46 34 47 257 87 1 59 2.3 0.97 2.04 20 18 49 80 239 100 2.13 50 5.21 1.09 1.6 14 160 93 167 161 89 0.81 PCT/EP2012/052232 139 Table 18 (continued to end): Examples of clinical, virological and biological data for test subjects of Table 17 No of test Ferritin Glycaemia Peptide C Insulin TG Albumin TSAT (%) Total chol ALP subject (p/)(mmole/L) (ng/mL) (ptUI/mL) (mmole/L) (g/L) (mmole/L) (U/L) 521 514 6.2 ND ND 1.76 47 32 4.62 62 509 179 4.3 2.8 7.5 0.56 48 32 4.02 56 308 455 5.1 ND ND 1.13 48 46 4.81 73 167 702 5.22 2.3 9.82 0.62 43 70 4.58 114 145 260 5.7 2.87 8.99 1.2 40 43 5.1 48 80 52 4.2 ND ND 0.97 47 30 5.16 100 75 271 5.4 ND ND 0.69 44 40 4.72 149 58 693 5.4 2.35 8.79 0.93 46 33 4.93 51 46 ND 4.9 3.03 11.18 1.5 43 36 4.22 71 6 150 5.4 1.88 4.87 0.66 45 35 4.5 40 91 206 4.7 ND ND 1 43 36 4.5 72 90 137 4.2 ND ND 0.88 48 42 5.4 52 86 837 6.6 ND ND 0.69 48 46 5.82 77 76 156 4.2 ND ND 1.12 47 36 4.96 58 72 397 5 2.34 9.74 3.57 47 27 4.72 70 71 101 4.5 0.81 3.13 0.36 47 15 4.96 65 65 147 5.2 3.31 17.07 1.09 46 43 3.82 49 62 822 6.3 5.21 30.99 0.9 45 45 5 52 59 30 5 2.18 6.65 0.71 46 27 5.46 77 50 182 5.2 3.15 14.74 1.59 47 36 3.57 84 PCT/EP2012/052232 140 2b) Example of application of the combination of the levels of expression (RNA) of the genes HERC5, IF144, IL8, MDK and OAS1 (combination No. 2 in Table 2 above): 5 Using the WKNN method (see Example 1 above), the LOOCV error associated with the combination of the levels of transcription (RNA) of the genes HERC5, IF144, IL8, MDK and OASI (combination No.2 in Table 2 above) is 14 (see Table 3 above). The best performances for this combination using the WKNN method (computed over the 10 population of 107 responders and non-responders presented in Table 11 above) are as follows: sensitivity (Se) = 82%; specificity (Sp) = 89% (see Table 3 above). The model parameters used for the WKNN method were as follows: 15 Kernel (K): triangular Parameter of Minkowski distance (D): 2 Number of neighbours (k): 4 An example of a prediction over 20 subjects (same human patients as in Example 2a) 20 above) is given in Table 19 below, which presents the measurement values for the levels of expression of the selected genes (BMK values obtained by the 2-act method ; see Example 1 above). One or more clinical, biological and virological factors may be combined with these 25 markers indicated above (levels of expression of five genes), and lead to a decision rule the predictive power of which is even better than that of the rule presented above. Table 18 above presents examples of such clinical, biological and virological factors, as well as their values for the test subjects of Table 19. 30 ND = not determined PCT/EP2012/052232 141 Table 19: Example of application of a classification model based on the combination of the levels of expression of the genes HERC5, IF144, IL8, MDK and OAS1 (combination No.2 of Table 2 above) WKNN model Status, No. of (kernel = triangular; parameter of Minkowski distance 2; k = 4) R or NR, test as test determined subject HERC5 IF144 IL8 MDK OAS1 WKNN after prediction treatment 521 1.670 0.785 3.732 1.676 6.233 NR NR 509 0.719 0.102 1.157 0.172 1.324 NR NR 308 0.886 0.154 0.853 0.446 2.305 NR NR 167 0.653 0.075 0.923 0.521 2.204 NR NR 145 1.283 0.183 0.755 0.255 2.395 NR NR 80 0.570 0.057 0.745 0.030 0.559 NR NR 75 1.035 0.319 0.908 1.248 3.959 NR NR 58 0.607 0.149 0.616 0.631 2.313 NR NR 46 0.563 0.177 0.911 0.308 3.543 NR NR 6 0.247 0.046 2.648 0.688 0.228 NR NR 91 0.149 0.042 0.040 0.051 0.651 R R 90 0.342 0.046 0.089 0.115 0.853 R R 86 0.259 0.086 0.451 0.060 1.053 R R 76 0.293 0.067 1.257 0.045 1.161 R R 72 0.097 0.010 0.064 0.001 0.162 R R 71 0.255 0.015 0.228 0.021 0.437 R R 65 0.258 0.055 0.168 0.105 0.908 R R 62 0.280 0.074 0.058 0.043 1.444 R R 59 0.269 0.039 0.329 0.090 1.380 R R 50 0.144 0.018 0.444 0.132 0.826 R R 5 PCT/EP2012/052232 142 2b) Example of application of the combination of the levels of expression (RNA) of the genes IL8, CLDN1, G1P3, HERC5 and RSAD2 (combination No.29 in Table 2 above): 5 The AUC relating to the combination of the levels of expression of the genes IL8, CLDN1, G1P3, HERC5 and RSAD2 (combination No.29 in Table 2 above) computed for the population of 107 responders and non-responders presented in Table 11 above is 0.857 (see Table 7 above). 10 Using the mROC method, the maximizing threshold of the Youden's index (6) for this combination is -1.281 (see Table 5 above). For this choice of threshold, the performances of the combination are as follows: Sensitivity (Se) = 77%; specificity (Spe) = 79% (see Table 3 above). 15 The following rule is an example of a decision rule: Z = 0.134 x CLDNl t + 0.488 x G1P3 t + 0.966 x HERC5 t + 0.129 x IL8 - 0.487 x RSAD2 t (function Z29ARN; see Table 5 above), where: - CLDN1, G1P3, HERC5, IL8 and RSAD2 are the measurement values for the biomarkers 20 BMK, i.e. the measurement values for the levels of expression of the indicated genes (Ct normalized in accordance with the 2 -Act method), and - the exponent t (carried here by CLDN1, G1P3, HERC5 and RSAD2) indicates that the value to be applied in the decision rule is the Box-Cox transformation (Box and Cox, 1964) of the measurement value of the level of expression of the gene under consideration 25 in order to normalize it using the following formula:

BMK

t = (BMK-1)/X. In the example of the decision rule indicated above, the parameters X are 0.68 for CLDN1, 0.19 for G1P3, 0.19 for HERC5 and -0.05 for RSAD2 (see Table above). 30 If Z > -1.281: the diagnostic test is positive (mROC prediction = 1), the subject is declared "NR" (subject predicted to be a non-responder to treatment).

PCT/EP2012/052232 143 If Z < -1.281: the test is negative (mROC prediction = 0), the subject is declared "R" (subject predicted to be a responder to treatment). An example of a prediction over 20 subjects is given in Table 20 below, which presents 5 the measurement values for the levels of expression of the selected genes (BMK values obtained by the 2 -Act method; see Example 1 above). One or more clinical, biological and virological factors may be combined with these markers indicated above (levels of expression of five genes), and lead to a decision rule 10 the predictive power of which is even better than that of the rule presented above. Table 21 above presents examples of such clinical, biological and virological factors, as well as their values for the test patients of Table 20. ND = not determined VL = viral load 15 PCT/EP2012/052232 144 Table 20: Example of the application of a classification model based on the combination of the levels of expression of the genes IL8, CLDN1, G1P3, HERC5 and RSAD2 (combination No. 29 in Table 2 above) mROC model (function Z29ARN; 6 = -1.28 1) Status, No. of test subject R or NR, as determined after treatment CLDN1 GIP3 HERC5 IL8 RSAD2 Z mROC prediction 46 4.112 0.209 0.563 0.911 2.395 -1.168 NR NR 50 1.409 0.097 0.144 0.444 0.194 -1.546 R R 55 0.856 0.060 0.235 1.395 0.221 -1.362 R R 58 1.966 0.476 0.607 0.616 1.439 -0.779 NR NR 59 1.840 0.089 0.269 0.329 0.283 -1.291 R R 62 3.042 0.141 0.280 0.058 0.568 -1.380 R R 65 1.664 0.061 0.258 0.168 0.480 -1.745 R R 71 0.607 0.038 0.255 0.228 0.322 -1.809 R R 72 0.406 0.006 0.097 0.064 0.054 -1.970 R R 75 4.170 0.607 1.035 0.908 2.959 -0.273 NR NR 76 2.403 0.041 0.293 1.257 0.340 -1.365 R R 80 2.329 0.031 0.570 0.745 0.309 -0.920 NR NR 83 5.776 0.730 2.549 0.231 4.362 0.630 NR NR 86 1.361 0.329 0.259 0.451 0.681 -1.348 R R 91 1.261 0.038 0.149 0.040 0.248 -1.988 R R 145 2.959 0.360 1.283 0.755 1.469 -0.080 NR NR 167 1.500 0.434 0.653 0.923 0.129 0.461 NR NR 308 2.305 0.545 0.886 0.853 1.057 -0.162 NR NR 521 4.317 1.315 1.670 3.732 6.681 0.592 NR NR 526 1.741 0.293 0.418 2.370 0.973 -0.902 NR NR PCT/EP2012/052232 145 Table 21: Examples of clinical, virological and biological data for test subjects of Table 20 No. of Age at BMI Alcohol Steatosis Mode of Metavir Ishak Metavir Age at Testet Sex HBP (kg/m 2 ) HOMA Diabetes (g/day) (score 0 to 3) contamination activity fibrosis fibrosis treat ent 46 M 54.4 26.9 2.4 No 5 1 Unknown 2 2 1 55.1 50 M 48.1 28.4 3.4 No 0 1 Endemic area (Egypt) 2 2 2 48.7 55 F 48.2 22.2 3.5 No 60 2 Unknown 1 3 2 49.8 58 M 48.1 25.2 2.1 No 0 0 Toxicomania 1 3 2 48.2 59 F 50.9 23.7 1.5 No 0 0 Transfusion 1 2 1 51.3 62 M 61.6 27.4 8.7 ND 50 1 Nosocomial 1 3 2 61.7 65 M 35.6 29.6 3.9 No 0 0 Endemic area (Egypt) 1 2 1 35.7 71 M 43.3 20.2 0.6 No 0 0 Unknown 1 2 1 43.3 72 M 37.2 29.0 2.2 No 0 2 Endemic area (Egypt) 1 3 2 37.3 75 F 53.0 27.9 ND No 0 0 ND 2 2 2 53.3 76 F 53.5 22.8 ND No 0 0 Unknown 2 3 2 53.6 80 F 36.6 31.6 ND No 0 2 Transfusion 1 1 2 37.2 83 F 51.4 25.7 ND No ND 2 NC 1 2 1 52.0 86 M 50.1 28.4 ND Yes 40 1 Unknown 1 1 1 50.3 91 F 47.8 21.4 ND No 0 0 ND 1 2 1 48.0 145 M 40.8 28.4 2.3 No 10 0 Toxicomania 1 2 1 42.2 167 F 48.6 37.8 2.3 No ND 0 Transfusion 2 2 1 48.8 308 M 34.9 27.7 ND No 0 0 Endemic area (Egypt) 1 1 1 35.4 521 M 58.4 23.8 ND No 30 0 Transfusion 1 4 3 58.9 526 F 73.0 24.8 ND No 0 1 Transfusion 2 6 4 73.2 PCT/EP2012/052232 146 Table 21 (continued): Examples of clinical, virological and biological data for test subjects of Table 20 No. of Duration test Genotype of VL HBP (copies/mL. 10E3) CVbeforeTTT (copies/mL .10E3) subject infection 46 1 ND 7903 9103 50 4 ND 445 445 55 1 ND 150 299 58 1 30.9 5079 5079 59 5 26.8 524 1120 62 2 ND 14 185 65 4 ND 1135 450 71 4 ND 6935 6935 72 4 ND 1155 1155 75 1 ND 3276 2347 76 1 ND 19227 19927 80 1 38.1 7606 12932 83 1 ND 1579 3928 86 2 ND 57 57 91 1 ND 3902 3902 145 1 25.8 26631 18304 167 4 ND 12616 12616 308 4 ND 423 423 521 1 34.8 14654 14432 526 1 38.4 778 778 PCT/EP2012/052232 147 Table 21 (continued): Examples of clinical, virological and biological data for test subjects of Table 20 No. of alpha2macroglobulin Haptoglobin Apo Al BLT GGT AST ALT PLQ Chol-HDL subset (g/L) (g/L) (g/L) (smol/L) (U/L) (U/L) (U/L) (x 1OE3/mm3) TP (%) (mmole/L) 46 4.77 0.21 1.48 14 174 91 125 216 100 0.89 50 5.21 1.09 1.6 14 160 93 167 161 89 0.81 55 4.4 0.38 1.4 13 118 210 314 156 95 1.05 58 3.16 0.63 1.79 4 162 62 93 222 91 1.43 59 2.3 0.97 2.04 20 18 49 80 239 100 2.13 62 4.37 0.94 1.37 12 46 34 47 257 87 1 65 4.07 1.15 1.17 8 46 61 152 210 92 0.72 71 1.89 0.47 1.92 7 22 37 58 177 94 1.63 72 1.5 0.96 1.44 15 23 31 71 249 98 0.76 75 ND ND ND 12 135 64 50 233 100 2.15 76 ND ND ND 14 81 90 157 209 96 ND 80 ND ND ND 18 132 44 67 311 92 ND 83 ND ND ND 13 82 78 95 232 101 ND 86 ND ND ND 13 287 53 105 160 100 ND 91 ND ND ND 9 39 64 79 353 99 ND 145 2.1 0.48 1.56 11 133 36 92 225 ND 1.25 167 2.9 0.54 1.64 15 378 163 144 183 78 1.28 308 ND ND ND 16 246 30 36 214 100 ND 521 ND ND ND 11 127 83 166 189 100 1.62 526 ND ND ND 17 172 128 210 187 77 ND PCT/EP2012/052232 148 Table 21 (continued to end): Examples of clinical, virological and biological data for test subjects of Table 20 No. of Ferritin Glycaemia Peptide C Insulin TG Albumin Total chol ALP subject (pg/L) (mmole/L) (ng/mL) (pUI/mL) (mmole/L) (g/L) TSAT (mmole/L) (U/L) 46 ND 4.9 3.03 11.18 1.5 43 36 4.22 71 50 182 5.2 3.15 14.74 1.59 47 36 3.57 84 55 943 5.6 4.04 14.04 0.68 46 43 4.88 55 58 693 5.4 2.35 8.79 0.93 46 33 4.93 51 59 30 5 2.18 6.65 0.71 46 27 5.46 77 62 822 6.3 5.21 30.99 0.9 45 45 5 52 65 147 5.2 3.31 17.07 1.09 46 43 3.82 49 71 101 4.5 0.81 3.13 0.36 47 15 4.96 65 72 397 5 2.34 9.74 3.57 47 27 4.72 70 75 271 5.4 ND ND 0.69 44 40 4.72 149 76 156 4.2 ND ND 1.12 47 36 4.96 58 80 52 4.2 ND ND 0.97 47 30 5.16 100 83 71 5.6 ND ND 0.76 44 37 3.68 97 86 837 6.6 ND ND 0.69 48 46 5.82 77 91 206 4.7 ND ND 1 43 36 4.5 72 145 260 5.7 2.87 8.99 1.2 40 43 5.1 48 167 702 5.22 2.3 9.82 0.62 43 70 4.58 114 308 455 5.1 ND ND 1.13 48 46 4.81 73 521 514 6.2 ND ND 1.76 47 32 4.62 62 526 320 4.4 ND ND 0.7 41 35 4.33 108 PCT/EP2012/052232 149 2d) Combination of the levels of expression (RNA) of the genes IL8, CLDN1, G1P3, HERC5 and RSAD2 (combination No.29 in Table No.2 above), further combined with clinical factors and/or other biological factors and/or virological factors 5 One or more clinical factors and/or one or more biological factors and/or one or more virological factors may be combined with the levels of expression of genes selected in accordance with the invention (in fact, levels of RNA transcription assayed in a HBP sample), and thus lead to a decision rule the predictive power of which is even better than that of just the combination of said levels of expression. 10 For example, the combination: - of the levels of expression (RNA) of the genes IL8, CLDN1, G1P3, HERC5 and RSAD2 (combination No. 29 in Table 2 above; see example 2c above), and - of the value for the following other biological factor: 15 o concentration of alkaline phosphatase or ALP (in IU/mL), and - of the value for the following virological factor: o viral load before the start of treatment or VLbeforeTTT (copies/mL. 103), leads to a decision rule the area under the ROC curve (AUC) of which, computed for those of the 107 responders and non-responders of Table 11 above for which the biological and 20 virological data were available (i.e. in total, 97 responders and non-responders; see Table 22 below) is 0.904 (see Table 26 above), while it is 0.857 (see Table 7 above) when the combination of the levels of expression of the genes IL8, CLDN1, G1P3, HERC5 and RSAD2 is used alone, without being combined with the clinical and virological factors indicated above (ALP and VLbeforeTTT). 25 Using the mROC method (see Example 1), the maximizing threshold of the Youden's index for this combination is 4.687 (see Table 24 above). For this choice of threshold, the performances of the combination are as follows: 30 Sensitivity (Se) = 810%; specificity (Spe) = 80% (see Table 23 above). The following rule is an example of a decision rule: Z = 0.3 x CLDNl t + 0.634 x GIP3 t + 1.154 x HERC5 t + 0.114 x IL8 - 0.685 x RSAD2 t + 0.036 x VLbeforeTTT t + 2.139 x PAL t 150 (function Z29ARNsupp; see Table 24 above), where: - IL8, CLDN1, G1P3, HERC5 and RSAD2 are the BMK measurement values for the biomarkers, i.e. the measurement values for the levels of expression of the indicated genes (Ct normalized in accordance with the 2 -ACt method), 5 - VLbeforeTTT and ALP are the values for the virological factor and the biological factor indicated above, and - the exponent t (carried here by CLDN1, G1P3, HERC5, RSAD2, VLbeforeTTT and ALP) indicates that the value to be applied in the decision rule is the Box-Cox transformation (Box and Cox, 1964) of the measurement value of the level of expression of 10 the gene under consideration in order to normalize it using the following formula: BMK= (BMK- 1)/X. In the example of the decision rule indicated above, the parameters k are 0.63 for CLDN1, 0.21 for G1P3, 0.15 for HERC5, -0.02 for RSAD2, 0.2 for VLbeforeTTT and -0.26 for 15 ALP (see Table 25 above). If Z > -4.687: the diagnostic test is positive (mROC prediction = 1), the subject is declared "NR". If Z < -4.687: the test is negative (mROC prediction = 0), the subject is declared "R". 20 The reference to any prior art in this specification is not, and should not be taken as, an acknowledgement or any form of suggestion that such art forms part of the common general knowledge in Australia. Further, the reference to any prior art in this specification is not, and should not be taken as, an acknowledgement or any form of suggestion that 25 such art would be understood, ascertained or regarded as relevant by the skilled person in Australia.

PCT/EP2012/052232 151 Table 22: Clinical, biological and virological data Clinical, biological and virological data Patients NR patients R patients n 97 42 55 Sex: male (%) / female (%) 60 (62) / 37 (38) 25 (60) / 17 (40) 35 (64) / 20 (36) Age [mean ± standard deviation (range)] 53.83140 ± 9.03924 55.43874 8,54830 52.60397 ± 9,28617 (31 - 71) (31 -79) (41 -79) Source of infection [n(%)] Blood transfusion 22 (22.68) 9 (21.43) 13 (23.64) Intravenous administration of a drug 28 (28.87) 13 (30.95) 15 (27.27) unknown 37 (48.45) 75 (47.62) 69 (49.09) Alanine aminotransferase (ALT) UI/L 120.8247 ± 89,29702 117 ± 80,94744 123.7455 ± 95,81661 [mean ± standard deviation (range)] (30 - 459) (30 - 354) (30 - 459) PCT/EP2012/052232 152 Table 22 (continued to end): Clinical, biological and virological data Clinical, biological and virological data Patients NR patients R patients HCV genotypes [n(%)] 1 60 (61.86) 36 (85.71) 24 (43.64) 2 8 (8.25) 0 (0) 8 (14.55) 3 11(11.34) 2(4.76) 9(16.36) 4 17 (17.53) 4 (9.52) 13 (23.64) 5 1(1.03) 0(0) 1(1.82) Fibrosis score (score F Metavir) [n(%)] 1 33 (34.02) 12 (28.57) 21 (38.18) 2 38(39.18) 15 (35.71) 23(41.82) 3 11(11.34) 8(19.05) 3 (5.45) 4 15 (15.46) 7 (16.67) 8 (14.55) PCT/EP2012/052232 153 REFERENCES Anastasiadis et al. 2005; New globally convergent training scheme based on the resilient propagation algorithm. Neurocomputing 64: 253-270. 5 Asselah et al. 2008; "Liver gene expression signature to predict response to pegylated interferon plus ribavirin combination therapy in patients with chronic hepatitis C"; Gut 57: 516-524. Box and Cox 1964; An analysis of transformations. Journal of the Royal Statistical Society, Series B 26: 211-243. 10 Breiman 2001; Random Forests. Machine Learning 45: 5-32. Chambers 2008; Software for data analysis: programming with R. Springer, New York, ISBN 978-0-387-75935-7. Chen et al. 2005; Gastroenterology 128: 1437-1444. Chen et al. 2010; Gastroenterology 138: 1123-1133. 15 Cole et al. 1983; Proc. Natl. Acad. Sci. USA 80: 2026-2030. Cole et al. 1985; Monoclonal Antibodies And Cancer Therapy, Alan R. Liss, Inc., pp. 77 96. Falissard 2005; Comprendre et utiliser les statistiques dans les sciences de la vie [Understanding and using statistics in the life sciences], Masson. 20 Hastie, Tibishirani and Friedman, 2009; "The Elements of Statistical Learning: Data Mining, Inference and Prediction", 2 "d Edition, Springer. Hidetsugu Saito et al. 2010; "On-treatment predictions of success in peg interferon/ribavirin treatment using a novel formula"; World J. Gastroenterol. 16(1): 89 97. 25 Intrator and Intrator 1993; Using Neural Nets for Interpretation of Nonlinear Models. Proceedings of the Statistical Computing Section, San Francisco: American Statistical Society (eds), pages 244-249. K6hler and Milstein 1975; Nature 256: 495-497. Kosbor et al. 1983; Immunology Today 4: 72. 30 Kramar et al. 1999; Criteres ROC g6n6ralis6s pour l'6valuation de plusieurs marqueurs tumoraux [Generalized ROC criteria for the evaluation of a number of tumour markers]. Revue d'Epidemiologie and Sant6 Publique 47:376-383. Kramar et al. 2001; mROC: a computer program for combining tumour markers in predicting disease states. Computer methods and programs in biomedicine 66: 199-207.

PCT/EP2012/052232 154 Liaw and Wiener 2002; Classification and regression by Random Forest. R. News 2.3: 18-22. Livak and Schmittgen 2001; Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta C(T)) Method. Methods 25: 402-408. 5 Reiser and Faraggi 1997; Confidence intervals for the generalized ROC criterion. Biometrics 53: 644-652. Riedmiller 1994; Rprop - Description and Implementation Details. Technical Report. University of Karlsruhe. Riedmiller and Braun 1993; A direct adaptive method for faster backpropagation 10 learning: the RPROP algorithm. Proceedings of the IEEE International Conference on Neural Networks (ICNN), San Francisco, pages 586-591. Schmitten and Livak 2008; Analyzing real-time PCR data by the comparative Ct method. Nature Protocols 3(6): 1101-1108. Shapiro 1999; The interpretation of diagnostic tests. Statistical Methods in Medical 15 Research, 8: 113-134. Su and Liu 1993; Linear combinations of multiple diagnostic markers. Journal of the American Statistical Association 88: 1350-1355. Swets 1988; Measuring the accuracy of diagnostic systems. Science 240, 1285-1293. Theodoridis and Koutroumbos 2009; Pattern Recognition. Academic Press, Elsevier. 20 US 4 376 110 (in the name of Hybritech Inc.).

Claims

2. The method according to claim 1, characterized in that it is carried out before said anti-HCV treatment has commenced. 156
3. The method according to claim 1 or claim 2, in which the total number of said other clinical, virological and biological factors, the value of which is measured or determined in step i), is 0 to 4.
4. The method according to any one of claims 1 to 3, in which the comparison of step ii) is carried out by combining the measurement values obtained for said subject in step i) into a multivariate classification model which compares those values with their values, or the distribution of their values, in reference cohorts which have been pre-established as a function of their status of responder or non-responder, in order to classify said subject into that of those reference cohorts to which it has the highest probability of belonging.
5. The method according to any one of claims 1 to 4, in which the comparison of step ii) is made by combining measurement values obtained for said subject in step i) into a pre-constructed multivariate classification model as follows: a) for a population of individuals who are of the same species as said subject, and who are infected with one or more HCVs, determining for each of those individuals whether or not that individual responds to an anti-HCV treatment which comprises the administration of interferon and of ribavirin, and classifying those individuals into distinct sub-populations as a function of whether they are responders or whether they are non-responders to that treatment, thus constituting reference cohorts established as a function of the response or non-response of those individuals to anti-HCV treatment; b) in at least one sample which has previously been obtained from each of said individuals, the nature of which is identical to that of the sample from said subject, making the same measurements as those carried out for said subject in said step i); c) making an inter-cohort comparison of the values for the measurements obtained in step b), or of the distribution of those values, in order to construct a multivariate classification model which infers a status of responder to anti HCV treatment or a status of non-responder to that treatment, starting from the combination of said values for the measurements obtained in step b). 157
6. The method according to any one of claims 1 to 5, in which the comparison of step ii) is made by combining said measurement values obtained in step i) into a mathematical function, in particular a linear or non-linear function, more particularly a linear function, in order to obtain an output value which is indicative of the status of responder or of non-responder of said subject.
7. The method according to any one of claims 1 to 5, in which the comparison of step ii) is made by combining said values obtained in step i) into a multivariate machine learning model, for example a multivariate non-parametric classification model, a multivariate heuristic model, or a multivariate probabilistic prediction model, in order to obtain an output value which is indicative of the status of responder or of non-responder of said subject.
8. The method according to any one of claims 1 to 7, in which the classification of said subject into that of said reference cohorts to which it has the highest probability of belonging is made with: - a sensitivity (Se) of at least 77%, or at least 78%, or at least 79%, or at least 80%, or at least 81%, or at least 82%, or at least 83%, or at least 84%; and/or with a - specificity (Sp) of at least 79%, at least 80%, or at least 81%, or at least 82%, or at least 83%, or at least 84%, or at least 85%, or at least 86%, or at least 87%, or at least 88%, or at least 89%, or at least 90%; and/or with - a negative predictive value (NPV) of at least 80%, or at least 81%, or at least 82%, or at least 83%, or at least 84%, or at least 85%, or at least 86%, or at least 87%; and/or with - a positive predictive value (PPV) at least 72%, or at least 73%, or at least 74%, or at least 75%, or at least 76%, or at least 77%, or at least 78%, au at least 79%, or at least 80%, or at least 81%, or at least 82%, or at least 83%, or at least 84%, or at least 85%, or at least 86%, or at least 87%, or at least 88%, or at least 89%.
9. The method according to any one of claims 1 to 8, in which said multivariate classification model has: - an area under the ROC curve (AUC) of at least 0.84, at least 0.85, at least 0.86, or at least 0.87, or at least 0.88, or at least 0.89, more particularly at least 0.90, and/or 158 - a LOOCV error of at most 17%, at most 16%, at most 15%, at most 14%, at most 13%, or at most 12%.
10. The method according to any one of claims 1 to 9 in which, in step i), said genes selected from: - at least two genes from among HERC5, IL8 and STAT2, and - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFIT1, IFIT4, IFITM1, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18, are: - HERC5, IF144, IL8 and MDK (combination No. 1), or - HERC5, IF144, IL8, MDK and OAS 1 (combination No. 2), or - G1P2, IL8, OCLN, STAT2 and USP18 (combination No. 3), or - CXCL6, IFIT4, IL8, STAT1 and STAT2 (combination No. 4), or - CXCL6, IL8, MX1, PLSCR1 and STAT2 (combination No. 5), or - CXCL6, IL8, MX1, STAT1 and STAT2 (combination No. 6), or - HERC5, IF144, IL8, MDK and STMN2 (combination No. 7), or - HERC5, IL8, PLSCR1 and STMN2 (combination No. 8), or - HERC5, IF135, IFIT1, IL8 and MX1 (combination No. 9), or - HERC5, IF144, IL8, OAS 1 and RSAD2 (combination No. 10), or - HERC5, IF144, IL8, ITGA2, MDK (combination No. 11), or - HERC5, IFIT1, IL8 and MX1 (combination No. 12), or - HERC5, IL8, MDK, OAS3 and RSAD2 (combination No. 13), or - CCL21, G1P2, IL8, MDK and STAT2 (combination No. 14), or - G1P2, IFITM1, IL8, OCLN and STAT2 (combination No. 15), or - G1P2, IL8, OAS 1, OCLN and STAT2 (combination No. 16), or - CLDN1, IL8, OAS2, OAS3 and STAT2 (combination No. 17), or - CXCL6, IFITM1, IL8, MX1 and STAT2 (combination No. 18), or - CXCL6, IFIT1, IL8 and STAT2 (combination No. 19), or - FOXO1, G1P2, IL8, MDK and STAT2 (combination No. 20), or - CXCL6, G1P2, IL8, MDK and STAT2 (combination No. 21), or - CXCL6, IL8, OAS2, STAT1 and STAT2 (combination No. 22), or - FOXO1, IF127, IFITM1, IL8 and STAT2 (combination No. 23), or 159 - HERC5, IF135, IF144, IL8 and OAS2 (combination No. 24), or - IL8, CCL21, G1P3, HERC5 and RSAD2 (combination No. 25), or - IL8, G1P3, HERC5, OAS3 and RSAD2 (combination No. 26), or - IL8, ITGA2, G1P3, HERC5 and RSAD2 (combination No. 27), or - IL8, STMN2, G1P3, HERC5 and RSAD2 (combination No. 28), or - IL8, CLDN1, G1P3, HERC5 and RSAD2 (combination No. 29), or - IL8, G1P3, HERC5, LGALS3BP and RSAD2 (combination No. 30).
11. The method according to any one of claims 1 to 10 in which said at least two selected genes from among HERC5, IL8 and STAT2 comprise IL8.
12. The method according to any one of claims 1 to 11, in which said at least one gene selected from CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFIT1, IFIT4, IFITM1, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18, is CXCL6, IF144, MDK, MX1 or RSAD2.
13. The method according to any one of claims 1 to 12, in which said genes selected in step i) do not comprise CLDN1, G1P3, IF127, IFITM1, ITGA2, OCLN or PLSCR1.
14. The method according to any one of claims 1 to 13 in which, in step i), the total number of said genes selected from: - at least two genes from among HERC5, IL8 and STAT2, and - at least one gene from among CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFIT1, IFIT4, IFITM1, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18, is 3, 4 or 5.
15. The method according to any one of claims 1 to 14, in which: - said clinical factor or factors are selected from the following factors: sex, age at the date of sampling, age of patient at the date of contamination, age of patient at the treatment start date, body mass index, insulin sensitivity index, diabetes, 160 alcohol consumption, degree of steatosis, mode of contamination, Metavir activity, or hepatic fibrosis score measured using the Metavir system (Metavir F score) or using the Ishak system; and/or - said virological factor or factors are selected from the following factors: viral genotype, duration of infection, viral load before treatment, viral load assayed for the patient at the treatment start date, viral load assayed for the patient at the date of sampling; and/or - said biological factor or factors other than the levels of expression of genes selected from HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, GlP2, G1P3, IF127, IF135, IF144, IFIT1, IFIT4, IFITM1, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18 are selected from the following factors: concentration of haptoglobin, concentration of apolipoprotein Al, total quantity of bilirubin, concentration of gamma glutamyl transpeptidase, concentration of aspartate aminotransferase, concentration of alanine aminotransferase, platelet count, quantity of prothrombin, quantity of HDL cholesterol, total quantity of cholesterol, concentration of ferritin, level of glycaemia, concentration of peptide C, quantity of insulin, concentration of triglycerides, quantity of albumin, transferrin saturation, and concentration of alkaline phosphatase.
16. The method according to any one of claims 1 to 15, in which: - said clinical factor or factors comprise the hepatic fibrosis score measured using the Metavir system (Metavir F score) or using the Ishak system; and/or - said virological factor or factors comprise the viral genotype and/or the viral load before treatment; and/or - said biological factor or factors other than the levels of expression of genes selected from HERC5, IL8, STAT2, CCL21, CLDN1, CXCL6, FOXO1, G1P2, G1P3, IF127, IF135, IF144, IFIT1, IFIT4, IFITM1, ITGA2, LGALS3BP, MDK, MX1, OAS1, OAS2, OAS3, OCLN, PLSCR1, RSAD2, STAT1, STMN2 and USP18 comprise the concentration of alkaline phosphatase.
17. The method according to any one of claims 1 to 16, which comprises: - determining whether the hepatic fibrosis score of said subject is a score which, in the Metavir score system, is at least Fl, more particularly at least F2; and/or 161 - determining whether the HCV or HCVs with which said subject is infected comprises an HCV of genotype 1, 4, 5 or 6, more particularly of genotype 1 or 4.
18. The method according to any one of claims 1 to 17, in which said sample which has been obtained in advance from of said subject is: - a sample of intracorporal biological fluid which has been taken from said subject, such as a sample of blood, serum or plasma, or a sample of urine from said subject, or - a sample containing proteins and/or polypeptides and/or peptides extracted or purified from said biological sample.
19. A manufactured article when used in the method of any one of claims 1-18, which comprises reagents in a preparation for their simultaneous, separate or sequential use, said reagents being constituted by: - reagents which specifically detect each of the transcription or translation products of 3 to 46 human genes, said 3 to 46 human genes comprising said selected genes according to any one of claims 1 to 18, and optionally other human genes, and - optionally, reagents which specifically detect one or more hepatitis viruses and/or hepatitis virus genotype or genotypes.
20. A manufactured article when used in the method of any one of claims 1-18, which comprises reagents in a preparation for their simultaneous, separate or sequential use in the method according to any one of claims 1 to 18, said reagents comprising reagents which specifically detect each of the transcription or translation products of selected genes according to any one of claims 1 to 18.
21. A manufactured article when used in the method of any one of claims 1-18, which comprises reagents in a preparation for their simultaneous, separate or sequential use in an anti-HCV therapy method, said reagents comprising reagents which specifically detect each of the transcription or translation products of selected genes according to any one of claims 1 to 18. 162
22. The manufactured article according to any one of claims 19 to 21, in which said reagents are nucleic acids which hybridize specifically to the RNA of said selected genes according to any one of claims 1 to 18, and/or to the cDNA obtained by reverse transcription of these RNA, or are proteins, polypeptides or peptides which specifically bind to the proteins encoded by said selected genes.
23. The manufactured article according to any one of claims 19 to 22, in which said reagents are amplification primers and/or nucleic acid probes, or are antibodies or fragments of antibodies or protein, polypeptide or peptide aptamers.
24. The manufactured article according to any one of claims 19 to 23, in which said reagents are contained in one or more tubes, or in the wells of a nucleic acid amplification plate for receiving a sample containing nucleic acids and a reaction mixture for nucleic acid amplification, or in the wells of a protein titration plate, more particularly a protein microtitration plate, for example an ELISA plate, or on microbeads or on a chip for the detection and quantification of nucleic acids, proteins, polypeptides or peptides.
25. A computer program product, for storage in a memory of a processing unit or on a removable memory support for cooperation with a reader of said processing unit, when used in the method of any one of claims 1-18, characterized in that it comprises instructions for carrying out a method according to any one of claims 1 to 18.
26. A computer device, when used in the method of any one of claims 1-18, which comprises a processing unit in the memory of which is stored a computer program product according to claim 22, and measurement values for the levels of transcription and/or translation of said selected genes as defined in any one of claims 1 to 18.
27. An in vitro method for predicting, prior to treatment, whether a subject infected with one or more HCVs has a high probability of being a responder to an anti-HCV treatment which will comprise the administration of interferon and the administration of ribavirin or a prodrug of ribavirin or whether, in contrast, that 163 subject has a high probability of not responding to that anti-HCV treatment, said method substantially as herein described with reference to the Examples and/or Drawings. 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015 2012215436 15 Jun 2015