WO2010118782A1 - Methods and tools for predicting the efficiency of anthracyclines in cancer - Google Patents

Abstract

The present invention relates to a gene set, kit and method to predict the efficiency of anthracyclines-based treatment of breast cancer.

Description

Methods and tools for predicting the efficiency of anthracyclines in cancer

Field of the invention

[0001] The present invention is related to methods and tools (gene set and kit or device) to predict the efficiency of an anthracycline (s) -based regimen in mammal subjects affected by an hyperproliferative disorder (cancer) , especially human patients affected by this disorder, especially human patients affected by Breast cancer (BC) .

Background of the invention [0002] Breast cancer (BC) is the most common cancer in women in Western countries.

[0003] Breast cancer is an heterogenous disease that can be subdivided into subgroups depending on markers . [0004] For instance, oestrogen receptor (ER) positive status is associated with a better outcome and may predict for a response to hormone treatments. [0005] Her2 (ERBB2, neu) overexpression is associated to a worse outcome, but Her2-specific treatments, such as administration of monoclonal antibodies (trastuzumab or herceptin) , may be beneficial for those patients .

[0006] Anthracyclines-based regimens are among the most active chemotherapies in Breast Cancer. However, their clinical use is associated with rare but severe toxicities, such as long-term hematological disorders (myelodisplastic syndrome and leukemia) and cardiac heart failure. Also the efficacy of anthracyclines appears to be restricted to a subset of the Breast Cancer patient population. Therefore, the identification of molecular markers that could predict a response of breast tumors to anthracyclines-based chemotherapy remains a priority.

[0007] Cell lines studies have suggested that cells with high amounts of Topoisomerase II alpha (TOP2A) might be more sensitive to anthracyclines. Several groups have investigated this hypothesis during the last decade and controversial results regarding TOP2A amplification/expression and response to anthracyclines in Breast Cancer patients have been reported. [0008] Several studies, mostly retrospective, hypothesized that the amplification/overexpression of TOP2A might influence a response to an anthracyclines therapy, often with contradictory results. [0009] It is known that a TOP2A overexpression or amplification were not predictive of response to neoadjuvant anthracyclines-based chemotherapy (Petit et al . , 2004, Eur J Cancer; 40:205-11) .

[0010] TOP2A quantification at the gene level and at the protein level are not correlated (Di Leo et al . , 2008, Eur. J. of Cancer, 44, 2791-2798) .

[0011] Some studies conclude that the prediction is restricted to the measure of the protein expression of TOP2A, while for other the best results are achieved by the monitoring of DNA amplification. For instance, Bartlett et al (J. Clin. Oncol, 2008, 31, 5027-5035) concludes that TOP2A gene amplification measures by FISH is not predictive of response to anthracyclin treatments. Knoop et al . (2005), J. Clin Oncol, 23, 7483-7490 conclude that both amplification and deletion of TOP2A is predictive for a response to anthracyclines .

[0012] These contrasting results reported in the literature regarding the predictive value of TOP2A may be explained by different reasons: the chemotherapy regimen used (monotherapy versus polychemotherapy, anthracyclines- based or including other drugs, such as taxanes) given in the neo-adjuvant or adjuvant setting, the diversity of treated patients included in these different studies, the heterogeneity in the assessment of a patient response (clinical, radiological and/or pathological response) and the different methods and levels of TOP2A evaluation. [0013] For instance, FISH results appear to be linked to a high variability between different test centres and up to 31% of discordance between local and central labs has been reported for FISH measurement of Her2 and/or TOP2A (Di Leo A et al, Cancer Res., 69 suppl (abstr 705) 2009) .

Aims of the invention

[0014] The present invention aims to provide new detection methods and tools that do not present the drawbacks of the state of the art.

[0015] The present invention aims to provide such methods and tools that improve the prediction of a response to antracyclines (-based) regimens in breast cancer patients .

Summary of the invention [0016] The present invention relates to a gene set representing TOP2A index.

[0017] Advantageously, the TOP2A index is measured by a mRNA quantification of these selected genes. [0018] Advantageously, the gene set of the invention (representing TOP2A index) comprises (or consist of) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 or all the genes mentioned in Table 1. [0019] The present invention relates also to a gene set (representing TOP2A index) consisting of TOP2A (Gene ID: 7153) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13,

14, 15, 16, 17, 18, 19, 20, 21, 22 or all the genes mentioned in Table 1. [0020] The present invention further relates to a gene set (representing TOP2A index) consisting of RARA (Gene ID: 5914) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,

12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0021] The present invention also relates to a gene set (representing TOP2A index) consisting of THRA (Gene ID: 7067) and further comprising 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0022] The present invention also relates to a gene set (representing TOP2A index) consisting of CDC6 (Gene ID: 990) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,

15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0023] The present invention also relates to a gene set (representing TOP2A index) consisting of GSDMl (Gene ID: 284110) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,

13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0024] The present invention also relates to a gene set (representing TOP2A index) consisting of PSMD3 (Gene

ID: 5709) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13,

14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0025] The present invention also relates to a gene set (representing TOP2A index) consisting of CSF3 (Gene ID: 1440) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0026] The present invention also relates to a gene set (representing TOP2A index) consisting of MED24 (Gene ID: 9862) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0027] The present invention also relates to a gene set (representing TOP2A index) consisting of SNORD124 (Gene ID: 100113390) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0028] The present invention also relates to a gene set (representing TOP2A index) consisting of NRlDl (Gene ID: 9572) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0029] The present invention also relates to a gene set (representing TOP2A index) consisting of TRNASTOP-UCA (Gene ID: 100126534) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0030] The present invention also relates to a gene set (representing TOP2A index) consisting of MSL-I (Gene ID: 339287) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0031] The present invention also relates to a gene set (representing TOP2A index) consisting of CASC3 (Gene ID: 22794) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0032] The present invention also relates to a gene set (representing TOP2A index) consisting of RAPGEFLl (Gene ID: 51195) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0033] The present invention also relates to a gene set (representing TOP2A index) consisting of WIPF2 (Gene ID: 147179) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,

13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0034] The present invention also relates to a gene set (representing TOP2A index) consisting of LOC100131821 (Gene ID: 100131821) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0035] The present invention also relates to a gene set (representing TOP2A index) consisting of GJD3 (Gene ID: 125111) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13,

14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0036] The present invention also relates to a gene set (representing TOP2A index) consisting of LOC390791 (Gene ID: 390791) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0037] The present invention also relates to a gene set (representing TOP2A index) consisting of LOC728207 (Gene ID: 728207) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1.

[0038] The present invention also relates to a gene set (representing TOP2A index) consisting of IGFBP4 (Gene ID: 3487) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0039] The present invention also relates to a gene set (representing TOP2A index) consisting of TNS4 (Gene ID: 84951) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0040] The present invention also relates to a gene set (representing TOP2A index) consisting of CCR7 (Gene ID: 1236) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0041] The present invention also relates to a gene set (representing TOP2A index) consisting of SMARCEl (Gene ID: 6605) and of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 or all the genes of Table 1. [0042] Preferably, the present invention relates to a gene set (representing TOP2A index) comprising (or consisting of) two or three genes selected from the group consisting of CDC6 (Gene ID: 990), THRA (Gene ID: 9572), RARA (Gene ID: 5914) and TOP2A (Gene ID: 7153) . [0043] Preferably, the present invention relates to a gene set (representing TOP2A index) comprising (or consisting of) THRA (Gene ID: 9572), CDC6 (Gene ID: 990), RARA (Gene ID: 5914) and TOP2A (Gene ID: 7153) . [0044] Preferably the present invention relates to a gene set (representing TOP2A index) comprising (or consisting of) THRA (Gene ID: 9572), CDC6 (Gene ID: 990), RARA (Gene ID: 5914) and TOP2A (Gene ID: 7153) and further comprising (consisting of) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, or all the genes of Table 1. [0045] The present invention relates to a gene set (representing TOP2A index) that may further comprise 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 or all the genes of Table 5 and/or Table 6. [0046] Alternatively, the present invention relates to a gene set comprising (or consistingof) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 or all the genes of Table 5. [0047] Alternatively, the present invention relates to a gene set comprising (or consisting of) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 or all the genes of Table 6. [0048] The present invention relates to a gene set

(representing TOP2A index) that further comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or all the genes of Table 10.

[0049] The present invention relates to a gene set

(representing TOP2A index) that further comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or all the genes of Table 11.

[0050] The present invention relates to a gene set

(representing TOP2A index) that further comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or all the genes of Table 12.

[0051] The present invention relates to a gene set (representing TOP2A index) that further comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or all the genes of Table 13.

[0052] The present invention relates to a gene set

(representing TOP2A index) that further comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or all the genes of Table 14. [0053] The present invention relates to a gene set

(representing TOP2A index) that further comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or all the genes of Table 15.

[0054] The present invention relates to a gene set

(representing TOP2A index) that further comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or all the genes of Table 16.

[0055] The present invention relates to a gene set

(representing TOP2A index) that further comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or all the genes of Table 17.

[0056] The present invention relates to a gene set (representing TOP2A index) that further comprises 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or all the genes of Table 18. [0057] The present invention further relates to a diagnostic kit or device comprising capture probes

(nucleotides sequences or proteinic probes such as antibodies, nanobodies or hypervariable portions thereof) possibly fixed upon a solid support, (such as a multiwell plate or glass side) able to detect specifically, the expression of complementary target genes (being complementary by hybridization to the capture probes) (or their corresponding proteins encoded by these genes) of the set of the invention (representing TOP2A index) , and/or the gene set of Tables 5 and/or 6 and possibly other means used for real time PCR analysis. [0058] Preferably, the kit or device according to the invention comprises means for real time PCR, preferably means for qRT-PCR.

[0059] Advantageously, the kit or device of the invention is a computerized system comprising - a bio-assay module configured for detecting a gene expression from a tumor sample (preferably a breast tumor sample) based upon the gene set or kit according of the invention and - a processor module configured to calculate expression of these genes or protein synthesis of these genes and to generate a risk assessment for the tumor (preferably a breast tumor) sample.

[0060] Another aspect of the invention concerns a method for a prediction (prognosis or prognostic) of cancer in mammal subject, which comprises the step of measuring gene expression in a (breast) tumor sample obtained from the said mammal subject, by putting into contact nucleotide sequences (or proteins) obtained from this tumor sample with the gene set, the kit or the device according to the invention and possibly generating a risk assessment for the said tumor sample by designating the outcome of a regimen comprising administration to the said mammal subject, of a sufficient amount of one or more Anthracycline compound (s) and by selecting the adequate compound to be administrated to this patient, preferably the sufficient amount of one or more anthracycline compound (s) .

[0061] Preferably in this method, the patient is a human patient, suffering from a cancer selected from the group consisting of ovary cancer, breast cancer, lung cancer, cancer of the womb, cancer of the bladder, cancer of the stomach, sarcomas, thyroid cancer, leukaemia, Hodgkin's lymphoma and multiple myeloma (preferably a breast cancer, more preferably a ER- breast cancer still more preferably a ER- Her2+ breast cancer) .

[0062] Preferably, the gene set, kit and device of the invention predicts the response of regimen consisting essentially into the administration to the tested patient of one or more Anthracycline compound (s), possibly combined with a simultaneous or a separated administration of another anti-tumoral therapeutic treatment or compound. [0063] Another aspect of the invention is related to an Anthracycline compound or a mixture of Anthracycline compounds (regimen) for use in the treatment (and/or the prevention) of cancer patients determined (as measured) by the gene set, kit or method of the invention. [0064] Advantageously, Anthracycline compound or a mixture of Anthracycline compounds (regimen) is for use in the treatment (and/or prevention) of a cancer selected from the group consisting of ovary cancer, breast cancer, lung cancer, cancer of the womb, cancer of the bladder, cancer of the stomach, sarcomas, thyroid cancer, leukemias, Hodgkin's lymphoma and multiple myeloma, having (increased) index (as measured) according to Table 5 and/or a worse prediction (prognosis) assessed by the (prediction or prognostic) method of the invention.

[0065] Alternatively, Anthracycline compound or a mixture of Anthracycline compounds (regimen) is for use in the treatment (and/or prevention) of a cancer selected from the group consisting of ovary cancer, breast cancer, lung cancer, cancer of the womb, cancer of the bladder, cancer of the stomach, sarcomas, thyroid cancer, leukemias, Hodgkin's lymphoma and multiple myeloma, having (increased) index (as measured) according to Table 6 and/or a worse prediction (prognosis) assessed by the prediction (prognostic) method of the invention.

[0066] Preferably, Anthracycline compound or a mixture of Anthracycline compounds (regimen) is for use in the treatment (and/or prevention) of (breast) cancer, preferably ER- breast cancer, more preferably a ER- Her2- breast cancer having (increased) index (as measured) according to the prediction (prognostic) method of the invention . [0067] Alternatively, Anthracycline compound or a mixture of Anthracycline compounds (regimen) is for use in the treatment (and/or prevention) of (an ER- Her2+ breast) cancer having (increased) index (as measured) according to the prediction (prognostic) method of the invention. [0068] Another aspect of the invention is related to a chemotherapeutic (anthracycline) compound (or regimen) for use in the treatment (and/or the prevention) of breast cancer having a specific (increased) index (as measured) according to Table 10 and/or Table 11, possibly combined with a specific (increased) TOP2A index (as measured according to Table 1), being preferably THRA (Gene ID: 9572), CDC6 (Gene ID: 990), RARA (Gene ID: 5914) and TOP2A (Gene ID: 7153) . [0069] Another aspect of the invention is related to a chemotherapeutic (anthracycline) compound (or regimen) for use in the treatment (and/or the prevention) of breast cancer having a specific (reduced) index (as measured) according to Table 12 and/or Table 13 and/or Table 16, possibly combined with a specific (increased) TOP2A index (as measured according to Table 1) , being preferably THRA (Gene ID: 9572), CDC6 (Gene ID: 990), RARA (Gene ID: 5914) and TOP2A (Gene ID: 7153) .

Table 1: genes composing TOP2A index

Symbol GenelD , Cyto

RARA 5914 17q21

CDC6 , 990 17q21.3

TΉRA ^" 7067 ^{" " "} 17q11.2 ^"

GSDM1 284110 17q12

PSMD3 5709 Ϊ7q12-q21 1

CSF3^{~~ ~} 1440 17q11.2-q12

MED24 9862 17q21.1

SNORD124 100113390 17q21.1

NRΪD1 9572 17q11.2

TRNASTOP-UCA 100126534 17q21.1

MSL-1 339287 17q21.1

CASC3 22794 17q1^~1-q21 3

^" RAPGEFL1 51195 ' Ϊ7q2Ϊ.1-q21.2

WIPF2 147179 j 17q21.2

LOC100131821 ^"10013Ϊ82T ^{" ~ ~}17q21.2

GJD3 125111 17q21.2

LOC390791 390791 17q21.2

LOC728207 728207 17q21 2

IGFBP4 3487 17qf2-q21 1

TNS4 84951 ^" ■ ^"I7q21.2

CCR7 ^{" "} 1236 , 17q12-q21.2

TOP2A 7153 17q21-17q22

SMARCE 1 ^" 6605 17q21.2

Note: RARA stands for retinoic acid receptor, alpha; CDC6 for cell division cycle 6 homolog (S. cerevisiae) ; THRA for thyroid hormone receptor, alpha (erythroblastic leukemia viral (v-erb-a) oncogene homolog, avian) ; GSDMl for gasdermin 1; PSMD3 for proteasome (prosome, macropain) 26S subunit, non-ATPase, 3; CSF3 for colony stimulating factor 3 (granulocyte) ; MED24 for mediator complex subunit 24; SNORD124 for small nucleolar RNA, C/D box 124; NRlDl for nuclear receptor subfamily 1, group D, member 1; TRNASTOP- UCA for transfer RNA opal suppressor (anticodon UCA) ; MSL-I for male-specific lethal-1 homolog; CASC3 for cancer susceptibility candidate 3; RAPGEFLl for Rap guanine nucleotide exchange factor (GEF) -like 1; WIPF2 for WAS/WASL interacting protein family, member 2; LOC100131821 for hypothetical protein LOC100131821; GJD3 for gap junction protein, delta 3, 31.9kDa; LOC390791 for similar to peptidylprolyl isomerase A isoform 1; LOC728207 for similar to 60S ribosomal protein L23a; IGFBP4 for insulin- like growth factor binding protein 4; TNS4 for tensin 4; CCR7 for chemokine (C-C motif) receptor 7; TOP2A for DNA topoisomerase II and SMARCEl for SWI/SNF related, matrix associated, actin dependent regulator of chromatin, subfamily e, member 1.

[0070] A last aspect of the invention is a method of treatment of patients comprising the steps of measuring gene expression (using one or several the gene sets of the Tables 1, 5-18) in a (breast) tumor sample obtained from the said mammal subject, by putting into contact nucleotide sequences with the gene set, the kit or the device according to the invention; - possibly generating a risk assessment for the said tumor sample by designating the outcome of a regimen comprising administration to the said mammal subject, of a sufficient amount of one or more Anthracycline compound (s) and - selecting the adequate compound to be administrated to this patient, preferably the sufficient amount of one or more anthracycline compound ( s ) .

Detailed description of the invention

Patient's description

[0071] One hundred and forty-nine patients were included in the prospective TOP trial. Out of these patients, 89 received 4 cycles of anthracyclines every 3 weeks, 59 patients received 6 cycles of anthracyclines administered every 2 weeks (dose-dense scheme) and one was not treated according to the protocol due to ineligibility .

10072] One hundred and thirty-two patients completed treatment as per protocol, 16 discontinued (2 due to disease progression, 2 due to consent withdraw, 3 due to adverse experience and 9 for another reason) .

[0073] A pathological complete response (pCR) rate of 15% was obtained, 14% and 16% for patients treated with the 3 -weekly and dose-dense scheme respectively. The patient and tumour baseline characteristics are illustrated in Table 2.

Table 2 : Patient and tumor baseline characteristics

Patients (n=149)

Patients registered 149

Case Report Forms 149 collected

Median age (range) 47 (27-68)

Age £ 50 91

Age > 50 58

Missing 0

T size (at baseline)

£ 2 cm 21

> 2 and £ 5 cm 105

> 5 cm 5 T4 17 Missing 1*

N status (at baseline)

NO 74

Nl 68

N2 3

N3 3

Missing 1* Histological type ductal 139** lobular 2 other 8 Missing 1*

Histological grade***

Gl 2

G2 28

G3 110

Gx (unknown) 9

Missing 1*

Type of surgery

Mastectomy 49

Conservative 88

Other 1

* 1 patient was ineligible

** 1 patient had multifocal tumour and both tumors were ductal

***1 patient had multifocal tumour with different grade

[0074] The age of the patient, the size and grade of the tumor and the nodal status were not associated with pCR. Ki67 protein expression (a proliferation marker) was not significantly associated with pCR, both when considering it as a continuous and binary variable (Table 3) .

Table 3 Association between the clinical parameters, FISH results and response to treatment (pCR= pathological complete response)

Topoisomerase II α (TOP2A)

[0075] The prospective investigation of the predictive value of TOP2A for response to anthracyclines was the primary aim of this trial. Advantageously, all T0P2A evaluations were carried out in a blinded fashion: T0P2A gene, mRNA and protein evaluations were done independently .

[0076] Gene expression profiles could be obtained for 120 patients out of the 149 (81%) . The ER mRNA levels were used to double-check the ER-negativity reported by IHC.

[0077] TOP2A was represented by 3 different Affymetrix probe-sets: 201 291_s_at, 201 292_at and 237 469_at. Since the probe-set 201 291_s_at showed the greatest variance, the inventors used that probe- set for further analyses .

[0078] Ninety-one samples had results both at DNA and mRNA level and a statistically significant correlation was observed between these results, both when considering FISH results as a continuous (Spearman rho= 0.35, p= 0.001) or as a discrete variable (p= 4 10^"4 when considered as amplified/non-amplified, and p=0.001 when considered as deleted/normal/amplified) . [0079] FISH results were available for 113 out of the 149 patients. Out of these samples, 36 were HER2 amplified (32%) ; TOP2A was amplified in 12 patients (11% of the global population) and deleted in 16 patients (14% of the population) . Noteworthy, all TOP2A amplified and 85% of the TOP2A deleted samples were also HER2 amplified. [0080] TOP2A measures by IHC were available for 120 patients. The median percentage of positively stained cells was 15% (range 0-90%) . When considering TOP2A expression as a continuous variable, the inventors did not observe any correlation with FISH results (based on the 99 ER-negative samples available for this comparison) .

[0081] The inventors were able to observe a small but statistically significant correlation between protein and mRNA levels (Spearman rho= 0.23, p= 0.02, based on 96 samples) . However, when the inventors used the median as a cut point to define overexpression, they did not find any correlation between the protein and other measurements of TOP2A. [0082] The inventors then assessed whether the different measurements of T0P2A were associated with common clinico-pathological parameters such as age at diagnosis, tumor size, nodal status and histological grade. Except a significant association between TOP2A mRNA levels and histological grade, high grade tumors presenting higher levels of TOP2A mRNA, no other associations were observed (Table 4) .

Table 4: TOP2A gene, mRNA and protein levels with regard to the clinic-pathologic parameters

[0083] TOP2A amplification was found to be predictive of response (kappa= 0.39, p=3 10^"s) to anthracyclines . Indeed, as shown in Table 3, 55% of the patients with TOP2A amplified tumors responded to the anthracyclines therapy as opposed to 9% of the patients whose tumors had no TOP2A amplification. Also, a higher proportion of HER2-positive patients presented a pCR (23%) compared to HER2 -negative patients (10%) , however this difference did not reach statistical significance (kappa=0.16, p=0.06) .

[0084] In this study, TOP2 mRNA levels (considered as a continuous variable or as a binary variable with the median as cut point) and TOP2A protein levels were not associated with response to therapy. [0085] However, when carrying subgroup analyses according to the HER2 mRNA status (cfr methods) , the inventors observed a C-index of 0.84 [95% confidence interval (CI) : 0.70-0.97), p= 2.10^"7] for TOP2A mRNA, which means that a C-index of 0.84 can be interpreted that the probability that a patient who has a pathological complete response, has a higher TOP2A mRNA value than a patient who has no pathological complete response is of at least 84%.

The TOP2A amplicon [0086] In order to better represent the TOP2A amplicon and to investigate whether other genes located closely to TOP2A might also be important in defining response to anthracyclines, the inventors developed different indices based on the averaged sum of the expression values of at least 2 genes located in the TOP2A amplicon.

[0087] By using a TOP2A index of 4 mRNA (THRA (Gene ID: 9572), CDC6 (Gene ID: 990), RARA (Gene ID: 5914) and TOP2A(Gene ID: 7153)), the inventors observed a C-index of 0.62 [95% CI: 0.46-0.78, p=0.06] .

[0088] The inventors observed that the expression profile of the genes composing the TOP2A index are correlated, but that the inclusion of more than one gene improved the performance of a test based on this index. [0089] When they evaluated the predictive performance of this index in subgroups of patients defined by the FISH evaluation of HER2, they observed a highly significant C-index of 0.84 (95%CI : 0.71-0.97 , p=10-7) in the HER2+ subgroup, whereas it was not significant in the HER2- subgroup [0.47 (95%CI : 0.32-0.63 ) , p=0.37)] .

[0090] The ROC curves of the TOP2A gene expression and TOP2A index and their corresponding AUC, which can be interpreted as the C-index, are illustrated in Figure 1 for both patient cohorts. The investigation of the prognostic value of the T0P2A index in untreated breast cancer patients using publicly available gene expression data showed that this index was not prognostic in the ER-/HER2- and HER2+ patients, but that high values of this index were associated with worse prognosis in the ER+/HER2 patients

(Figure 2) . More precisely, this signature was associated with prognosis in the low proliferative ("luminal A") subgroup of ER+/HER2- patients but not in the high proliferative subgroup ("luminal B"), suggesting that this index is not simply a good quantification of proliferation.

[0091] The inventors further validated the gene index they developed using the publicly available data of the EORTC 10994/BIG00-01 clinical trial (Bonnefoi et al . , 2007, Lancet Oncology, 8, 1071-1078). [0092] The inventors evaluated the predictive performance of this index in a cohort of patients treated pre-operatively either by an anthracyclines-based regimen

(FEC) or by a taxane-based regimen (TET) . It yielded a statistically significant C-index in the anthracyclines -arm [C-index=0.67 (95%CI : 0.54-0.79) , p=0.006] , but not in the taxane-based treatment arm [C-index=0.47 (95%CI : 0.32-0.62) , p=0.35] .

[0093] When studying the TOP2A index according to subgroups defined by HER2, the C-index was only predictive of response in the HER2+ [C-index= 0.81 (95%CI : 0.63-1.00) , p=0.0005] but not HER2- patients [C-index=0.56 (95%CI:0.40- 0.72), p=0.24] of the anthracyclines-arm. Of note, the TOP2A mRNA levels alone were not significantly associated with response in this validation cohort.

[0094] Again the inventors observed that TOP2A gene index predicts a better response to anthracyclines in the whole ER- patients and especially in ER-/HER2+ patients) . Moreover, in this subsequent validation study, the TOP2A index clearly outperformed mRNA measurements of TOP2A amplification. (Figure 3) .

Predictive value of biology-driven gene expression signatures [0095] The inventors further investigated the predictive value of the gene expression modules that they developed previously in WO 2009/030770 and Desmedt et al. Clin Can Res 2008) . In the global population, high levels of the estrogen receptor module were statistically associated with response to anthracyclines, whereas high levels of the tumor invasion module were associated with the presence of residual disease.

[0096] When looking into the ER-/HER- and ER-/HER2+ subgroups, the inventors observed that high levels of the tumor invasion module were associated with the presence of residual disease [C-index=0.31 (95%CI: 0.15-0.47), p=0.01 and 0.35 (95%CI : 0.19-0.50) , p=0.02 respectively] in both subgroups .

[0097] Additionally, in the ER-/HER2+ subgroup, high values of the immune response module were associated with response to anthracyclines [C-index=0.75 (95%CI : 0.45-1.00) , p=0.05] . Also the inventors have found that high values of the immune response module were associated with increased response rates to other chemotherapies and anti-cancer treatments . Table 5: ER-/HER2- gene signature

Gb ace. # Symbol Coef Ugcluster Gb ace. # Symbol Coe Ugcluster Gb acc. # Symbol Coef Ugcluster

AU144449 -1 Hs.469287 NM_006419 CXCL13 1 Hs.100431 NM_000213 ITGB4 -1 Hs.370255

BE550153 -1 NM_002416 CXCL9 1 Hs.77367 AL049988 ITPR2 -1 Hs.512235

AL833150 1 Hs.327631 AF361370 CYB5R3 -1 Hs.517666 NM_002223 ITP R2 -1 Hs.512235

AI087792 -1 AF003114 CYR61 -1 Hs.8867 NM_006801 KDELR1 -1 Hs 515515

AU 92452 -1 HS.167793 NM_006182 DDR2 -1 Hs.275757 AB033097 KIAA1271 -1 Hs.567292

BG389789 1 AW571709 DDX11 -1 Hs 443960 BC002710 KLK1O -1 Hs.275464

AU155091 -1 NM_030881 DDX17 -1 Hs.528305 NM_006853 KLK11 -1 Hs.57771

R55784 -1 Hs.155736 U59321 DDX17 -1 Hs.528305 U90269 KRIT1 -1 Hs.531987

BF056892 -1 Hs.435132 NM_001356 DDX3X -1 Hs.380774 AJ278245 LANCL2 -1 Hs.224282

BF433269 -1 Hs.208690 NM_015362 DERP6 -1 Hs.417029 BE222220 LEMD1 -1 Hs.181245

AK024243 -1 Hs.557780 NM_012242 _|DKK1 -1 Hs.40499 NM_002305 LGALS 1 -1 Hs.445351

AW39255Ϊ 1 Hs.180559 A>Ϊ760Ϊ3^~ DNAJC12 ^" -1 Hs.260720 BF345728 LOC14772 -1 Hs.534560

BF969544 -1 Hs.523913 AL119957 DNAJC3 -1 Hs.59214 BM666010 LOC20016 -1

AA903473 1 Hs.245545 AI692645 DOCK5 -1 Hs.195403 BF028225 LOC20189 -1 Hs.205952

AF495383 ADAM9 -1 Hs.2442 AL049369 DSCR1 ^" -1 Hs.282326 AF519622 LOC28302 1 Hs.255729

S68290 AKRΪC1 -ϊ Hs.558319 NM_00Ϊ943 DSG2 ^{" ~}1 Hs.412597 AI285192 LOC28488 1

U05598 AKR1C2 -1 NM_004950 DSPG3 -1 Hs.435680 AV721528 LOC28616 -1

M33376 AKR1C2 -1 BC003143 DUSP6 -1 Hs 298654 BF970340 LOC33944 -1 Hs.103939

NMJ)OI 629 ALOX5AP -1 Hs.507658 AW269645 ECT2 1 Hs.518299 BF211019 LOC40094 -1 Hs.502948

BE670056 ANKRD10 -1 Hs.525163 AA534817 EDG3 -1 AK021807 LRP 11 -1 Hs 408355

AI097229 ANKRD22 1 Hs.217484 S81545 EDNRA -1 Hs.183713 AB017498 LRP5 -1 Hs.6347

NM_018685 ANLN -1 Hs.62180 NM_015507 EGFL6 1 Hs.12844 AF196468 LSM2 -1 Hs.103106

NM_006305 ANP32A -1 Hs 458747 U48722 EGFR -1 Hs.488293 AA112507 LSM4 -1 Hs.515255

NM 018153 AlvfTXRI -1 Hs.165859 U95089 EGFR -1 Hs.488293 BE566894 LTB4DH -1 Hs 555920

N48299^" APCDD1 -11 Hs .293274 NM_005228^~ EGFR -1 Hsr488293 NM_002343 LTF 1 Hs.565960

AY113699 ^" APH1A _t -1 .Hs.108408 AK000106^" EGFR ^* lHs.488293 W80468 MALAT1 1 Hs 187199

NM 017855 APIN -1 HS/J4381Ϊ N^~M_005228^~ EGFR Hs.488293 AI936566 MCM4 -1 Hs.460184

BG427393 APLP2 -1 Ηs .370247 BE878463 EGFR ^' -1 Hs.488293 BC007207 MGC1311' -1 Hs.239500

NM_00Ϊ647 APOD -1 Hs 522555 NM_001415 EIF2S3 Hs.539684 H09657 MGC3990I 1 Hs 496530

AI358867 APOE 1 Hs.515465 U88968 ENO1 Hs.517145 BF677486 MGC3990I 1 Hs.496530

N74607 AQP3 1 Hs.234642 BC001038 EP N3 Hs.165904 AK026366 MGC4862! -1 Hs.558780

AF052179 ARF1 -ϊ 'Hs 286221 BG479856 FAM60A Hs.505154 BE855713 MGC9913 -1 Hs 23133

AI343000 ARGBP2 -1 Hs 481342 NM_006329 FBLN5 - Hs.332708 NM_006533 MIA -1 Hs.279651

BC040474 ARHGEF1 -1 Hs.98594 NM 002006 FGF2 - Hs.284244 NM_002427 MMP13 -1 Hs.2936

AF019888 ARPC4 -1 .Hs.323342 NMJ502014 FKBP4 - Hs.524183 N74662 MRPL43 1 Hs 421848

NM_004318 ASPH -1 Hs.332422 BI087313 FL J10707 - Hs.288164 AB049959 MRPL51 -1 Hs 55847

AF306765 ASPH -1 Hs.332422 AK075525 FLJ14712 -^• Hs 50802 AK022122 MTBP -1 Hs.553528

D13119 ~ ATP5G2 -1 Hs.524464 NM_144682^" FLJ3Ϊ952 - I Hs.462833 NM_006097 MYL9 -1 Hs.504687

NMJJ01690 ATP6V1A -1 Hs.477155 AA507012 FLOT1 -^• Hs 179986 W74452 NDRG2 -1 Hs 525205

AB028869 BIRC5 -1 Hs 514527 AJ276395 FN1 -^• Hs 203717 BE741920 NDUFA11 -1 Hs 406062

NM_001200 BMP2 -1 Hs.73853 NM_016725 FOLR1 -^• Hs.73769 AI357376 NEDD4L -1 Hs.185677

L20471 BSG -1 Hs.501293 NM_005249 FOXG 1 B -^• Hs.525266 AA005023 NOD27 1 Hs 528836

AL049332 BTG3 -1 Hs.473420 BC004908 FSCN1 - I Hs 118400 AA694067 NOD3 1 Hs.128357

AW204712 ^" Cido^~rf128 1 Hs.385493 BE930017 ^" FUS I Hs ^"513522 AI868441 NPW -1 Hs.233533

AC007182^" C14orfl -1 Hs.15106 AF072872 FZD1 Hs .94234 NM_0Ϊ4581 OBP2B /// -1 Hs.449629

BC007010 C1S -1 Hs.458355 AI796169 GATA3 Hs.524134 NM_003999 ¹OSMR -1 Hs 120658

NM_014145 JC20orf30 -1 Hs.472024 AI796169 GATA3 Hs.524134 AI569974 OSR1 -1 Hs.123933

BF381837 C20or152 -1 Hs.472564 NM_002053 GBP1 Hs.62661 AK075503 P4HB -1 Hs.464336

AW001030 C9orf52 1 Hs.130086 AL136680 GBP3 Hs.534284 AI608902 PDCD1LG 1 Hs.521989

AF257659^~ CALL) -1 Hs 7753 BG271923 GBP5 Hs.513726 U67932 PDE7A -1 Hs.527119

NM_004058{CAPS -11 Hs .567229 NM_019067 GNL3L ^•Hs.522664 AB033831 PDGFC -1 Hs.148162

L07555 CD69 1 Hs.208854 AF064826 GPC4 Hs.58367 BG054550 ^fPDLIM5 -1 Hs.480311 Table 5 (Continued) ER-/HER2- gene signature

Gb ace # Symbol Coef Ugcluster Gb ace # Symbol Coef Ugcluster Gb ace # Symbol Coef Ugcluster

AW006735 CD8A 1 Hs 85258 NM_013308 GPR171 1 Hs 549152 AF116705 PDLIM5 -1 Hs 480311

BC003682 CDC42 1 Hs 467637 U87460 GPR37 1 Hs 406094 NMJJ00287 PEXB -1 Hs 567243

NM_004360 CDH1 -^• 1 Hs 461086 NM_005756 GPR64 -1 Hs 146978 NM 002624 PFDN5 -1 Hs 288856

D21254 CDH11 1 Hs 116471 NM_000826 GRIA2 -1 Hs 32763 NM_016134 PGCP -1 Hs 156178

BC015877 |CDH19 Hs 42771 NM_000846 GSTA 1 -1 Hs 446309 R81072 PHACTR2 -1 Hs 102471

AA922068 |CDK6 Hs 119882 BF569051 H19 -1 Hs 556040 AV721177 PICALM -1 Hs 163893

AW19270(f CDK6 Hs 119882 NM_017445 H2BFS -1 Hs 473961 NM_002645 PIK3C2A -1 Hs 175343

D00682 CFL1 Hs 170622 NM_007071 HHLA3 -1 Hs 142245 NM_003628 PKP4 -1 Hs 407580

AK025141 CHPT1 Hs 293077 AC007130 HIBADH -1 Hs 406758 K03226 PLAU -1 Hs 77274

AF279779 CHRM3 Hs 7138 NM_152696 HIPK1 -1 Hs 532363 Z54367 PLEC1 -1 Hs 434248

T15991 CHRM3 Hs 7138 BC002649 HIST1 H1C -1 Hs 7644 AW469573 PLEKHC1 -1 Hs 509343

AK056349 CHRM3 I Hs 559103 NM_003514 HIST1 H2A -1 Hs 134999 AI754404 PLOD2 -1 Hs 477866

NM_000740 CHRM3 I Hs 7138 NM_003523 HIST1 H2B -1 Hs 534369 AW665155 POLH -1 Hs 439153

NM_006984 CLDN10 I Hs 534377 NM_003522 HIST1 H2B -1 Hs 182137 BQ613856 POLR2J2 -1 Hs 533383

N74924 CLDN20 1 Hs 352244 BE271470 HIST1 H2B -1 Hs 352109 BC005903 POLR2L -1 Hs 441072

BE791251 CLDN3 Hs 25640 NM_003525 HIST1 H2B -1 Hs 553506 NM_000942 PPIB -1 Hs 434937

Y15916 COL1A1 Hs 172928 AI313324 HIST2H2A. -1 Hs 530461 AB041836 PQBP1 -1 Hs 534384

AA909035 ,COL4A2 Hs 508716 BC001629 HIST2H2A -1 Hs 530461 BC017833 PRIM2A -1 Hs 485640

M20776 'COL6A1 Hs 474053 U70544 HLA-DRB4 -1 Hs 534321 AL121975 PRIM2A -1 Hs 485640

AY029208 |COL6A2 Hs 420269 NM_002130 HMGCS1 -1 Hs 397729 NM_014086 PRO1073 1

NM_052889 COP1 Hs 348365 BF983406 HNRPH1 -1 Hs 202166 NMJ44707 PROM2 1 Hs 469313

AI621079 ¹COPA NM_020386 HRASLS 1 Hs 36761 AU 84802 PRPF4 -1 Hs 374973

NM_001863 ^OXBBI Hs 431668 BG612458 HSPCB -1 Hs 509736 NM_007244 PRR4 -1 Hs 408153

AF 142573 CRISPLD1 I HS 436542 D13889 ID1 -1 Hs 504609 BF576710 PTP4A1 -1 Hs 227777

BC008745 iCRTAP I Hs 517888 BC001872 IGHM 1 Hs 525648 AU147115 PTPRK -1 Hs 155919

NMJJ05212 |CSN3 I Hs 54415 AU160004 IMP-3 -1 Hs 432616 AF312393 PTRF -1 Hs 437191

BG3~30076 CTNNA1 1 Hs 445981 AV733308 ^" ΓTGAΘ -1 Hs 133397 NMJ)16277 RAB23 -1 Hs 555016

AY072911 CXADR Hs 473417 NM 033669 ITGB 1 -1 Hs 429052 AB029004 RAB6IP2 -1 Hs 400431

NMJJ01565 CXCL10 Hs 413924 BF305661 ITGB4 -1 Hs 370255 NM_006506 RASA2 -1 Hs 15999

AF493929 ^V|RGS5 ^Hs 24950 ^..121985 SLAMF7 1 Hs 517265 K03199 TP 53 -1 Hs 408312

AF498970 iRHOA Hs 247077 BG251467 SLC25A37 -1 Hs 122514 AW665624 TPARL -1 Hs 479766

BI668074 RHOB Hs 502876 NM_004955 SLC29A1 -1 Hs 25450 BG389015 'TPD52 -1 Hs 368433

BC025770 I RHOJ Hs 525389 BC039498 SLC39A6 -1 Hs ^"79136 NM_000366 TPM1 -1 Hs 133892

NM_002933 RNASE1 i Hs 78224 AU 37517 SLITRK6 -1 Hs 525105 AA602532 TPP1 -1 Hs 523454

NM_000980 ₄RPL18A // Hs 558383 NM_003064 SLPI -1 Hs 517070 AW235355 TPR -1 Hs 279640

NM_000978 RPL23 Hs 406300 AK026426 SMARCA1 -1 Hs 152292 NM_003294 TPSAB1 -1

NM_000991 RPL28 Hs 356371 AI989477 SOX4 -1 Hs 357901 AF206666 TPSAB1 -1

NMJJ01004 RPLP2 1 Hs 437594 BC002704 STAT1 1 Hs 470943 AF099143 TPSAB1 -1

NM_001020 RPS16 1 Hs 397609 U46768 STC1 -1 Hs 25590 AF206667 TPSAB 1 // -1 Hs 405479

NM_012250 RRAS2 Hs 502004 AL553320 STIP1 -1 Hs 337295 NM_003293 TPSAB 1 // -1 Hs 405479

AI692974 IRRMI Hs 558393 NM_007271 STK38 -1 Hs 409578 NM_024164 TPSB2 -1 Hs 405479

AK022166 (RSBN1 Hs 486285^{^} AF008937 STX16 -1 Hs 307913 BI857154 TRAM1 -1 Hs 491988

NMJJ06054 JRTN3 ^" Hs 473761 AL034418 SULF2 I -1 Hs 162016^~ AI884858 TUSC3 -1 Hs 426324

NM_006271 S100A1 Hs 515715 U46837 'SURB7 -1 Hs 286145 U42349 TUSC3 -1 Hs 426324

AW238654 |S100A8 Hs 416073 AF077053 ^JTAF9L -1 Hs 546635 NM_016267 ,VGLLI -1 Hs 496843

BC001766 S 100B Hs 422181 BC010946 TAGLN^" -1 Hs 503998 BE542323 VGLL1 -1 Hs 496843

NM_007281 SCRG1 Hs 7122 BG330520 TALDO1 // -1 Hs 356766 AF199015 VIL2 -1 Hs 487027

AI380298 SDC2 Hs 1501 NM_003220 TFAP2A -1 Hs 519880 BF663141 VIL2 -1 Hs 487027

AI380298 SDC2 Hs 1501 AF047002 THOC4 -1 Hs 534385 NM_016312 WBP11 -1 Hs 524281

AW294630 'SEC15L2 Hs 303454 U67195 T1MP3 -1 Hs 297324 AF274954 WDR1 -1 Hs 128548

AF346602 SEC61A1 Hs 518236 BC005176 TM7SF3 -1 Hs 438641 AM45745 WNK1 -1 Hs 504432

AL574210 ISERPINE1 Hs 414795 |NM_016551 TM7SF3 -1 Hs 438641 AF303378 YSG2 -1 Hs 10056

Z95126 SET /// LO Hs 436687 BF540749 TM7SF3 -1 Hs 438641 BC003623 YWHAZ -1 Hs 558651

AF 131749 SEZ6L2 -1 Hs 6314 AB004064 TMEFF2 1 Hs 144513 BG483802 ZBTB 10 -1 Hs 205742

NM 006275 ιSFRS6 .1 Hs 6891 BE568134 TNFRSF21 -1 Hs 443577 NM 016620 ZNF644 -1 Hs 173001

De novo identification of predictive gene expression signatures

[0098] Given the unique characteristics of the TOP trial, i.e. that patients were treated with anthracyclines monotherapy in the neo-adjuvant setting; the inventors aimed at developing de novo gene expression signatures that would predict the efficacy of anthracyclines both in ER- /HER2- and ER-/HER2+ tumors (Tables 5 and 6) . [0099] The ER-/HER2- gene signature was composed of 321 probe-sets corresponding to 294 unique genes (Table 6) . The signature included PLAU, which is the prototype from the tumor invasion gene expression module of WO 2009/030770 and of Table 13 of the present invention; EGFR whose over- expression has been correlated with poor survival (Bartlett et al . 2008) and with resistance to anthracyclines -based chemotherapy in triple-negative breast cancer, TP53 whose predictive value for efficacy of anthracyclines-based chemotherapy has been matter of controversy and many others .

[0100] The IAP analysis, which included 218 genes, revealed that cellular movement, cellular growth and proliferation, cell-to-cell signaling, cell death and cellular assembly were the most significant functional classes.

[0101] The ER-/HER2+ gene signature was composed of 261 probe-sets corresponding to 218 unique genes. Eleven genes were in common with the ER-/HER2+ signature and both signatures displayed a low but significant correlation (p=0.22, p=0.02) . As expected, this signature included TOP2A as well as other genes located on 17q21-q22. Of interest, IGFlR, GRB7, a gene located close to HER2 , and different metallothioneins were also part of this signature and their over-expression was associated with the presence of residual disease.

Table 6 ER-/HER2+ gene signature O Gbb aaccec ff # . SSyymmbbooll Coef'Ugcluster Gb ace # Symbol Coef Ugcluster Gb ace # Symbol Coef Ugcluster AW574798 JKLHL6 1 Hs 333181 M24669 IGHM 1 Hs 525648 NM_014258 SYCP2 Hs 202676 AW043921 ⁱ 1 Hs 547618 M87789 IGH@ /// K 1 Hs 567449 AA770014 DSCR8 Hs 192371 AF 103530 1 Hs 559330 L14454 IGHG3 1 Hs 510635 AB014341 IGLJ3 Hs 517453 X93006 ' 1 Hs 561078 AJ275439 IGHA1 /// 1 1 Hs 497723 AB001733 IGU3 Hs 517453 AF043583 Hs 449599 U80139 IGHM 1 Hs 525648 X57812 IGL@ /// IGL Hs 449585 AJ239383 Hs 551925 AJ243643 IGHA 1 //M 1 Hs 497723 AA680302 IGL@ /// IGL Hs 449585 X84340 AJ275408 IGHA1 /// I 1 Hs 497723 AV698647 IGL@ /// IGL Hs 449585 AL359605 Hs 283851 U80164 IGH@ /// K 1 Hs 567449 D84143 IGLC2 Hs 567242 AF043584 Hs 449599 BF002659 IGHM 1 Hs 525648 D87021 IGLC2 Hs 567242 AI743780 BC001872 IGHM 1 Hs 525648 AF103591 IGLC2 Hs 567242 AI739132 Hs 152812 NM_01479 KIAA0125 1 Hs 395486 AJ249377 IGLC2 Hs 567242 AI923633 _t Hs 96886 S55~735 IGHA1 III I 1 Hs 558339 NM_017424 CECR1 Hs 170310 BF476080 1 Hs 418040 AJ275355 IGHG1 /// 1 1 Hs 497723 AF043586 IGLC1 Hs 555877 AK025909 ^~ Hs 288741 X17115 IGHM 1 Hs 525648 M87790 IGLC2 Hs 567242 BF108778 ⁱ Hs 28360_ AJ275397 |IGHG1 /// I 1 ,Hs 497723 D01059 JGLC2 Hs 567242

AI004009 I -1 |Hs 130526 AB035175 IGHA1 III \ 1 Hs 525648 U96394 IGLC2 Hs 449585

L06102 VHs 547404 U92706 IGHA1 /// 1 1 Hs 510635 L21961 IGLC2 Hs 567242

NM_052889 COP1 1 Hs 348365 AJ225092 IGHA1 /// I 1 Hs 510635 D84140 IGLC2 Hs 567242

AL833150 -1 Hs 327631 BG340548 IGH@ /// K 1 Hs 558339 H53689 IGLV3-25

AK097976 KLHL6 1 Hs 333181 NM_00302 SGNE1 -1 Hs 156540 AJ249377 IGLV2-14

M74303 1 Hs 560200 NM_02014 MEIS2 -1 Hs 510989 D87016 IGLC2 Hs 449567

L23516 1 Hs 383169 BC003610 MFGE8 -1 Hs 3745 AI761713 MMP11 Hs 143751

AF 103529 I Hs 560823 NM_00257 PCSK6 -1 Hs 498494 AI655697 DERL3 Hs 159971

L06101 I 1 Hs 64568 BC008777l lTGAL 1 Hs 174103_^ AL022324 LOC91353 Hs 546463

BG536224 11 Hs 559333 BC025741 ^"C16orf54 1 Hs 331095 AA398569 LOC91316 Hs 148656

M85256 1 Hs_554197 AI377875 EIF3S8 -1 Hs 192425 NM_005080 XBP1 Hs 437638

L34164 1 Hs 448957 AA~102581 USP31 -1 Hs 183817 AL008583 NPTXR Hs 91622^"

M87268 1¹Hs 448957 NM_00584 IGSF6 1 Hs 530902 NM_021822 APOBEC3G Hs 474853

BG482805 1 Hs 551722 NM_00119 TNFRSF17 1 Hs 2556 NM_025225 ADPN Hs 377087

AW404894 1 Hs 552522 AI401105 TXNDC11 1 Hs 313847 AK025665 ADPN Hs 377087

AW405975 1 Hs 449575 U90304 IRX5 -1 Hs 435730 AI631846 MGC16635 Hs 137007

L23518 1 Hs 560223 M 10943 MT1F -11 Hs 513626 NM_000878 IL2RB Hs 474787

BE620374 I -1 Hs 559452 BF217861 MT1E -1 Hs 534330 M20812 LOC339562 Hs 449972

AI225238 1 !Hs 445500 NM_00595 MT1X -1 Hs 374950 BG540628 IGKV1-5

BF244402 I -1 ' NMJJ2471 ELMO3 -1 Hs 377416 AJ408433 IGKC Hs 449621

BF244402 -1 NM_00174 CALB2 -1 Hs 106857 M63438 IGKC /// IGK Hs 449621

AI803010 -1 Hs 554052 NM_02430 FA2H_ 1 HSJ61329 BC005332 IGKC /// IGK Hs 449621

AW628735 1¹Hs 567989 NM_01485 KIAA0672 ^" 1 i Hs 499758 U4457 1 IGKC Hs 449621

AW504569 I|HS 551955 BC005926|EVr2B 1jHs 5509 AW408194 [IGKVI D-13 Hs 390427

^*~ r

AI660245 i !Hs 547f30 AB000221 JCCL18 1 |HS 143961 L14458 HGKC Hs 449621

AA424537 ^rC10orf18 _L -1 ^f Hs 4-32548 Y13710_jCCU8 _^ J ItI⁵L¹⁴³U⁶J BG548679 ' IGKC_ _ J Hs 449621

NM_000698 'ALOX5 11Hs 89499 NM_qO735 CASC3 ^~ f[Hs 350229^" AW575927 ΓGKC /// IGK Hs 449621

AU 18571 ¹C10orf74 -1¹Hs ^99833 AFO6Ϊ8I2IKRTI6 -1 Hs 432448 AW006735 ¹CDeA Hs 85258

U11058 KCNMA 1 1 |Hs 144795 AA001203 LOC33928 1 Hs 532786 BG485135 IGKC Hs 449621

AI917901 ACTA2 -1 'Hs 500483 AI702962 SMARCE1 1 Hs 463010 M85276 GNLY Hs 105806

BE888744 IFIT2 -1 Hs 437609 NM_00125 CDC6 1 Hs 405958 NM_006433 GNLY Hs 105806

AW043782 LDLRAD3 -1 Hs 205865 U77949 CDC6 1 Hs 405958 NM_001615 ACTG2 Hs 516105

AI755024 LOC38775 -1 Hs 32478 AA046439 SFRS 1 1 Hs 68714 AK022277 DTNB Hs 307720

NM_030754 lSAA1 /// S -1 ^! Hs 332053 AU159942 TOP2A 1 Hs 156346 BE465475 KBTBD9 Hs 130593 AL042088 TUB -1 Hs 231850 AL561834 TOP2A 1 Hs 156346 AI337069 RSAD2 Hs 17518 AI928342 ASRGL1 1 _jHs 535326 AK000271 SUPT4H1 1 Hs 439481 AW004016 ST6GAL2 Hs 98265 NM_025080 ASRGL1 1 'Hs 535326 AJ082827_ MPPE1 -1 Hs 514713 NM_001450 FHL2 Hs 443687 NM_000852¹GSTP1 1 IHs 523836 NM_02112 PMAIP 1 -11 Hs 96 X51887 LOC391427 U94592 UCP2 1 'Hs 80658 NM 02112 PMAIP 1 -1 Hs 96 AI928242 TFCP2L1 Hs 156471 AA725246 SPTBN2 1 Hs 26915^" NMJ)0343 ZNF91 1 Hs 558418 AF255647 DKFZP566N Hs 369471 AI246687 CTSC 1 Hs 128065 NM_00154 APOC1 1 Hs 110675 AU146532 PDK1 Hs 470633 AI246687 CTSC 1 Hs 128065 NM_01211 CBLC -1 Hs 466907 BF060783 DLX1 Hs 407015 NM_002421 MMP1 1 Hs 83169 NMJJ0277 KLK6 -1 Hs 79361 NM_000648 CCR2 Hs 511794 NM_019604 CRTAM 1 HsJ 59523 AF2~43527 KLK5 -1 Hs 50915 AA225165 LOC339903 Hs 146346 NM_006235 POU2AF1 1 ' Hs 2407 NM_00719 KLK8 -1 Hs 104570 R42166 CNTN4 Hs 298705 AF329841 'C1QTNF5 -1 |HS 157211 AL042588 PEG3 -1 Hs 201776 AJ240085 TRAT1 Hs 138701 Gb ace # Symbol Coef Ugcluster Gb ace # Symbol Coef Ugcluster Gb acc # Symbol Coef Ugcluster

BG112263 ASAM -1 Hs 504187 AF208967 PEG3 -1 Hs 201776 AW170015 PLCXD2 1 Hs 292419

NM_003622 PPFIBP1 -1 Hs 172445 NM_00504 KLK7 -1 Hs 151254 NM_018456 EAF2 1 Hs 477325

BC005961 PTHLH -1 IHs 89626 AL046017 FAM46C 1 Hs 356216 AF303889 ROPN1 -1 Hs 558504

M31157 PTHLH 1 Hs 89626 NM_00176 CD2 1 Hs 523500 AI340264 ROPN1 -1 Hs 558504

NM_002258]KLRB1 1 Hs 169824 AI743596 TSPAN2 -1 Hs 310458 NM_017578 ROPN1B -1 Hs 528203

NM_001038 SCNN1A 1 Hs 130989 AI082747 PALMD -1 Hs 483993 AI674183 EPHB1 -1 Hs 116092

NMJJ07360 KLRK1 1 Hs 387787 NM_00260 PDE4B 1 Hs 198072 NM_000685 AGTR1 -1 Hs 477887

NMJ501769 CD9 -1 Hs 114286 L20966 PDE4B 1 Hs 198072 AI754404 PLOD2 -1 Hs 477866

AY007436 RBP5 1 Hs 246046 AU 49963 DKFZp761 -1 Hs 132121 AI684991 CP 1 Hs 554736

NM_003979 GPRC5A Hs 194691 AA456955 ANKRD38 -1 Hs 283398 NMJJ00096 CP 1 Hs 554736

AK026776 LRRK2 Λ Hs 187636 NM_01240 PLA2G2D 1 Hs 189507 AI922198 HPS3 /// CP 1 Hs 477898

NM_173600 | MUC19 , 1 Hs 244017 AW24198C FCRL5 1 Hs 415950 NM_022443 MLF1 -1 Hs 85195

J00269 KRT6A /// -1 Hs 558758 AF343662 FCRL5 1 Hs 415950 AI911434 MLF1 -1 Hs 85195

U89281 HSD17B6 ₊ -1 Hs 524513 NM_00146 FMO5 1 Hs 303476 NMJJ14398 LAMP3 1 Hs 518448

NM_000785 CYP27B1 1 Hs 524528 AI659418 'RCSD1 1¹Hs 493867 NM_006548 IMP-2 -1 Hs 35354

BF510098 PPP1R1A 1 Hs 505662 AL121985 SLAMF7 1 Hs 517265 AI743792 ST6GAL1 1 Hs 207459

NM_001874 CPM KHs 48455Y AL121985 |SLAMF7 1 Hs 517265 AA702685 OSTalpha 1 Hs 567320

NM 004950 DSPG3 1 Hs 435680 AJ271869 ¹SLAMF? 1 Hs 517265 AU 26453 COX7B2 1 Hs 479656

NM_013300 HSU79274 -1 UHs 436618 AL514445 RGS4 -1 Hs 386726 AK026815 KIAA1102 -1 Hs 335163

NM_007197 FZD10 -1 Hs 31664 NM_00065 SELL 1 Hs 82848 AI494113 KIAA1102 -1 Hs 335163

M86849 GJB2 -1 Hs 524894 M83772 FMO3 1 Hs 445350 AI609256 SLC30A9 -1 Hs 479634

BF432648 TNFRSF1E 1 Hs 149168 AW003297 RALGPS2 1 Hs 496222 AB018289 KIAA0746 1 Hs 479384

AWΪ37148 POSTN -1 Hs 136348 NM_00346 CHlYi f Hs 201688 AA522514 KIAA0746 1 Hs 479384

AU 37517 SLΓΓRK6 -1 Hs 525105 BG339064 BTG2 1 Hs 519162 NM_002416 CXCL9 1 Hs 77367

R70320 SLITRK6 -1 Hs 525105 NMJD1777 LAX1 1 Hs 272794 AV733266 IGJ 1 Hs 381568

AI680986 SLITRK6 -1 Hs 525105 BC014852 IRF6 -1 Hs 355827 AF116705 PDLIM5 -1 Hs 480311

AF 193855 ZIC2 -1 _LHs 369063 NM_00614 IRF6 -1 Hs 355827 BF671400 PDLIM5 -1 Hs 480311

AI057619 UGCGL2 1 Hs 193226 M19154 TGFB2 -1 Hs 133379 AF065389 TSPAN5 1 Hs 118118

M 12959 TRAC 1 NM_00323TGFB2 -1 Hs 133379 AA059445 TSPAN5 1 Hs 118118

NM_004496 FOXA1 1 'HS 163484 NM_00250 NKX2-2 -1 Hs 516922 NM_001977 ENPEP -1 Hs 435765

N79004 SIX1 1 Hs 558398 NMJ30311 SPAG4 1 Hs 123159 NM_024090 ELOVL6 -1 Hs 412939

NM_000295 SERPINAI 1 Hs 525557 M35533 LBP 1 ) Hs 154078 BC001305 ELOVL6 -1 Hs 412939

L23519 1GHV1-69 1 Hs 449011 AW02438C RPS21 -1 Hs 190968 BG545288 FGB 1 Hs 300774

M24670 IGHV1-69 1 Hs 449011 NM_00195 EEF1A2 -1 Hs 433839 NMJJ05141 FGB 1 Hs 300774

NM_000509 FGG 1 Hs 546255 M32577 HLA-DQB 1 1 Hs 409934 NM_016027 LACTB2 -1 Hs 118554

NM_004362 CLGN 1 Hs 86368 NM_00352 HIST1H2Bι -1 Hs 546314 BC040474 ARHGEF10 -1 Hs 98594

S81545 EDNRA -1 Hs 183713 AA037483 HIST1H2Bι -1 Hs 546314 AF306765 ASPH 1 Hs 332422

NM_000909 NPY1R -1 Hs 519057 BC002842 HIST1 H2B -1 Hs 130853 NM_006823 PKIA 1 Hs 433700

NM_000163 GHR -1 Hs 125180 M 16276 HLA-DQB 1 -1 Hs 409934 BF245954 PKIA 1 Hs 433700

N71063 ADAMTS6 -1 Hs 482291 BE740761 HIST1H4H -1 Hs 421737 AL136588 DKFZp761D -1 Hs 492187

NM_017614 BHMT2 -1 Hs 114172 NM_00059 C4A /// C4 1 Hs 546241 NMJJ12082 ZFPM2 -1 Hs 431009

AW206234 FLJ42709 -1 Hs 171132 NM_00354 HIST1H4H -1 Hs 421737 BF979497 SQLE -1 Hs 71465

BC030966 KIAA0372 ^! -1 Hs 482868 M17955 "HLA-DQBI 1 Hs 409934 ALΪ37725 EPPK1 1 Hs 200412

NM_004772 |C5orf13 ^"ι -1 Hs 483067 NM_00322 TFAP2B 1 Hs 33102 AL137725 EPPK1 1 Hs 200412

NMJJ01046 SLC12A2 -1 Hs 162585 NM_00145 FOXC1 8883 NM_003289 TPM2 -1 Hs 300772

NM_014031 SLC27A6 -1 j Hs 49765 NM 00145 FOXF2 x - Hs 34 jf Hs 484423 AF109294 MTAP -1 Hs 193268

NM_016459 PACAP j 1 Hs~409563 NM_01225 HEY2 -1 Hs 144287 AU145658 MGC24103 -1

AF151024 PACAP 1 Hs 409563 AU 29628 SAMD3 1 Hs 558660 NMJJ03558 PIP5K1 B 1 Hs 534371

NM_033136iFGF1 -1 Hs 483635 AI633559 MAP3K4 -1 Hs 390428 AI949136 COL27A1 -1 Hs 494892

AB046841 PCDHB16 -1 Hs 147674 NM_00079 DDC -1 Hs 359698 AK021957 COL27A1 1 Hs 494892

AL832028 PCDHGA4 -1 NM_01519 COBL -1 Hs 99141 BC005939 PTGDS 1 Hs 558373

AF231124 SPOCK -1 Hs 124611 BF855173 SEPT7 1 Hs 191346 NM_000954 PTGDS 1 Hs 558373

D13720 ITK 1 Hs 558348 M30894 TRGC2 1 M61900 PTGDS 1 Hs 558373

N49841 ZNF300 -1 Hs 134885 M 13231 TRGC2 /// 1 Hs 534032 NM_018159 NUDT11 -1 Hs 200016

NM_014211 GABRP -1 Hs 26225 M27331 TRGC2 /// 1 Hs 534032 AI308863 CYBB 1 Hs 292356

AL577024 LOC22136 1 Hs 7921 M 16768 TRGC2 /// 1 Hs 534032 AI436587 P2RY8 1 Hs 111377

AB046819 CPNE5 1 |Hs 372129 BF792917 HOXA10 -1 Hs 110637 AI625739 ARHGEF9 -1 Hs 54697

NM_005084 PLA2G7 1 Hs 554780 AI949827 NFE2L3 1 Hs 404741 NM_014467 SRPX2 -1 Hs 306339

AI377755 HLA-DQA1 -1 Hs 555872 BE674103 CROT -1 ' Hs 125039 BF591534 TCEAL7 -1 Hs 21861

BE669692 HLA-DRB 1 -1 Hs 554754 NM_01228 KCND2 -1 Hs 21703 NM_017938 FAM70A 1 Hs 437563

NM_002122 HLA-DQA1 1 Hs 387679 NM_00240 MEST 1 Hs 270978 BE542323 VGLL1 -1 Hs 496843

K02403 C4A /// C4 1 Hs 546241 NM_01325 CLEC5A -1 Hs 446235 NM_016267 VGLL1 -1 Hs 496843

AI435670 SPDEF 11 Hs 485158 AF495383 ADAM9 -1 Hs 2442 U10691 MAGEA6 1 Hs 441113

AI583173 HLA-DQB 1 1 Hs 409934 NM_00622 PNOC 1 Hs 88218 BC000340 MAGEA3 1 Hs 417816

BG397856 HLA-DQA1 1 Hs 387679 NM_01447 ADAMDEC 1 Hs 521459 NM_005491 CXorte -1 Hs 20136

AL581873 HLA-DOA 1 Hs 351874 AF117949 L0XL2 -1 Hs 116479 [0102] The IAP analysis, which included 141 genes, revealed that cell-to-cell signaling and interaction, cellular movement, antigen presentation, cell death, as well as cellular growth and proliferation were the most significant functional classes.

[0103] Both signatures were positively correlated with the ER and immune response modules and negatively correlated with the tumor invasion and angiogenesis modules, although the level of the correlation differs according to the signature (see Table 7) .

Table 7 : Spearman rho correlations between the de novo signatures and the gene expression modules

Correlations Spearman's rho

ER-/HER2+ signature ER-/HER2- signature

ER-/HER2+ signature Correlation Coefficient 1 0,592219678

Sig. (2-tailed) 0,000001

N 118 118

ER-/HER2- signature Correlation Coefficient 0,592219678 1

Sig. (2-taιlecf) 0, 000001

N Ϊ18 118

ER module Correlation Coefficient 0,491642289 0,367889007

Sig. (2-taιled) 0,000001 4.15939E-05

N 118 118

HER2 module Correlation Coefficient 0,17831122 0,02552416

Sig (2-tailed) 0,053378276 0,783810651

N 118 118

Proliferation module Correlation Coefficient -0,138255563 -0,074315515

- - — sig. (2-tailed)^" 0.13542872Ϊ 0,423834615

N 118 118

Tumor invasion modu Correlation Coefficient -0,287295622 -0,378991231

Sig. (2-tailed) 0,001608367 2.31558E-05

N 118 118

Angiogenesis module Correlation Coefficient -0,483739258 -0,457137744

Sig. (2-tailed) 0,000001 0,000001

N 118 118

Immune response moi Correlation Coefficient 0,695601839 0,43849039

Sig (2-taιled) 0,000001 0,000001

N 118 118

Apoptosis module Correlation Coefficient 0,165083504 0,178501127

Sig. (2-tailed) 0,074025712 0,053120616

N 118 118

** Correlation is significant at the 0.01 level (2-tailed).

* Correlation is significant at the 0.05 level (2-tailed). [0104] Interestingly, when applying these signatures to the validation cohort, the ER-/HER2+ signature was predictive of pCR in the HER2+ patients treated in the anthracyclines arm [C- index: 0.75 (95%CI: 0.56-0.94), p=0.005] , but not in the taxane arm, similarly the ER- /HER2- signature was only predictive of pCR in the HER2- patients treated with A but not with T chemotherapy [C- index: 0.65 (95%CI: 0.50-0.81), p=0.03] . The detailed results are given in Table 8.

Table 8: Results of the predictive ability of the de-novo gene signatures, presented as C- indices, in the validation cohort, according to treatment arm and HER2 mRNA status

[0105] The inventors further investigated the prognostic value of these signatures. Table 9 : Prognostic analysis of the ^xde novo' gene signatures

As illustrated in figure 2 and in Table 9, high values of the ER-/HER2+ signature were associated with better prognosis in the global population, in the ER+/HER2- and HER2+ patients, whereas high values of the ER-/HER2- signature were only associated with better prognosis in HER2+ patients.

[0106] The inventors then combined the present invention with the results they previously obtained (Desmedt et al . , 2008) .

[0107] The inventors coupled the gene set of the invention, with a gene set comprising several genes from Table 10 (immune module 1) . gene.symbol EntrezGene.ID gene.symbol EntrezGene.ID gene.symbol EntrezGene.ID

ALPI 248 HLA-A 3105 NR3C1 2908

ANPEP 290 HLA-DRBl 3123 NSMAF 8439

ARHGDIB 397 HLA-DRB5 3127 PAK2 5062

BAG4 9530 ICAMl 3383 PDK2 5164

BAX 581 ICOSLG 23308 PIK3C2G 5288

BBS9 27241 IKBKB 3551 PLCBl 23236

BID 637 ILlORA 3587 PPP1R13B 23368

BIRC3 330 IL12B 3593 PPP3CA 5530

BLVRA 644 IL12RB2 3595 PRFl 5551

C17orf46 124783 IL13 3596 PRKARlB 5575

CASPlO 843 IL15 3600 PRKDC 5591

CASP6 839 ILIA 3552 PTEN 5728

CASP8 841 IL2RA 3559 PTENPl 11191

CASP9 842 IL3 3562 PTPRC 5788

CD28 940 IL4R 3566 PVRLl 5818

CD33 945 IRAK2 3656 RAFl 5894

CD4 920 ITGA4 3676 RELA 5970

CD40 958 ITGAM 3684 RHEB 6009

CD44 960 ITGAX 3687 RPS6KB1 6198

CD5 921 ITK 3702 SPTANl 6709

CD7 924 JAKl 3716 STAT3 6774

CD80 941 JAK3 3718 STAT5A 6776

CD86 942 JUNB 3726 TANK 10010

CFLAR 8837 LMNA 4000 TAPl 6890

CR2 1380 LMNBl 4001 TAP2 6891

CRADD 8738 LTA 4049 TGFBl 7040

CSNKlD 1453 MADD 8567 TNF 7124

CUTLl 1523 MAF 4094 TNFRSFlOA 8797

CYCS 54205 MAP2K3 5606 TNFRSF13B 23495

DAXX 1616 MAP3K14 9020 TNFRSFlB 7133

EIF4A1 1973 MAP3K7IP1 10454 TNFRSF25 8718

EIF4E 1977 MAP4K2 5871 TNFSF13B 10673

ELKl 2002 MAPKl 5594 TOLLIP 54472

FAFl 11124 MAPK8 5599 TRA@ 6955

FAS 355 MYD88 4615 TRAFl 7185

FKBPlA 2280 NCF2 4688 TRAF3 7187

GRB2 2885 NFKBl 4790

[0108] The inventors measured that increased expression of TOP2A index coupled with increased expression of genes of Table 10 increased the prognosis for chemotherapy (anthracyclines) .

[0109] The inventors coupled the gene set of the invention, with a gene set comprising several genes from Table 11 (immune module 2) . gene.symbol EntrezGene.ID gene.symbol EntrezGene.ID gene.symbol EntrezGene.ID

ACP5 54 FU20035 55601 MX2 4600

ADAMDECl 27299 GLRX 2745 NMI 9111

APOCl 341 GPR171 29909 P2RX5 5026

ARHGAP15 55843 GPR18 2841 PIM2 11040

BIRC3 330 GZMK 3003 PIP3-E 26034

BST2 684 HCLSl 3059 PLA2G7 7941

BTN3A2 11118 HCP5 10866 PLAC8 51316

CCL5 6352 HERC5 51191 PSCDBP 9595

CCL8 6355 HERC6 55008 PSMEl 5720

CCRL2 9034 IFI30 10437 PTPN7 5778

CD2 914 IFI44L 10964 RAB8A 4218

CD3G 917 IFI6 2537 RASGRPl 10125

CD40LG 959 IFIT3 3437 REC8L1 9985

CD48 962 IFIT5 24138 RFX5 5993

CD69 969 IFITMl 8519 RSAD2 91543

CECRl 51816 IGSF6 10261 RTP4 64108

CLEC4A 50856 IL18 3606 SECTMl 6398

CTSC 1075 INDO 3620 SH2D1A 4068

CXCLlO 3627 IRFl 3659 SNXlO 29887

CXCLIl 6373 IRF8 3394 SP140 11262

CXCL9 4283 ISG15 9636 SP0CK2 9806

DDAH2 23564 ITGB2 3689 STATl 6772

DDX58 23586 KLRC3 3823 STAT4 6775

DNAL4 10126 KLRKl 22914 TAPl 6890

EBI2 1880 LAG3 3902 TFEC 22797

EBI3 10148 LAMP3 27074 TRAF3 7187

ECGFl 1890 LAPTM5 7805 TRGV9 6983

EFNAl 1942 LGP2 79132 TRIM22 10346

ETV7 51513 LILRA4 23547 UBD 10537

FASLG 356 LILRBl 10859 VAVl 7409

FGL2 10875 MGC29506 51237 ZC3HAV1 56829

FU11286 55337 MXl 4599

:oiio] The inventors measured that increased expression of TOP2A index coupled with increased expression of genes of Table 11 increased the prognosis for chemotherapy (anthracyclines) .

[0111] The inventors coupled the gene set of the invention, with a gene set comprising several genes from Table 12 (stroma module 1) . gene.symbol EntrezGene.ID gene.symbol EntrezGene.ID gene.symbol EntrezGene.ID

FGD6 55785 LRPlB 53353 VIT 5212

PLAC9 219348 TIMP4 7079 HOP 84525

CAB39L 81617 STXBP6 29091 GPX3 2878

FGD6 55785 WNTIl 7481 RRM2 6241

LONRF3 79836 PLAC9 219348 GPX3 2878

CGI-38 51673 MICAL2 9645 MYOC 4653

STXBP6 29091 PKD1L2 114780 CLEC3B 7123

FHLl 2273 SDCl 6382 GRP 2922

STXBP6 29091 FHLl 2273 GJB2 2706

LEPR 3953 FHLl 2273 AADAC 13

CA4 762 F2RL2 2151 MATN3 4148

TNMD 64102 AKR1C2 1646 PPAPDClA 196051

POSTN 10631 LEFl 51176 LOC646324 646324

LOC58489 58489 ADAM12 8038 COLlOAl 1300

LOC284825 284825 ADHlC 126 COLlOAl 1300

[0112] The inventors measured that increased expression of TOP2A index coupled with reduced expression of genes of Table 12 increased the prognosis for chemotherapy (anthracyc lines) .

[0113] The inventors coupled the gene set of the invention, with a gene set comprising several genes from Table 13. Table 13 : stroma module 2 gene.symbol EntrezGene.ID gene.symbol EntrezGene.ID gene.symbol EntrezGene.ID

PLAU 5328 BICD2 23299 EPYC 1833

BMPl 649 TNFRSF12A 51330 ANKRD46 157567

MMP14 4323 VDR 7421 CPNEl 8904

THYl 7070 SNAI2 6591 BCL3 602

COL5A2 1290 EPB41L2 2037 GLBl 2720

ADAM 12 8038 FKBP14 55033 UBL5 59286

ANGPTL2 23452 NBLl 4681 ULKl 8408

MFAP2 4237 CAPl 10487 N0L8 55035

SERPINHl 871 ATP6V1B2 526 TGFB2 7042

COL6A1 1291 EPHB4 2050 PDGFB 5155

ISLR 3671 TRAM2 9697 BASPl 10409

PDLIM7 9260 DDR2 4921 SDS 10993

PARVA 55742 GFPT2 9945 RPS27A 6233

OLFML2B 25903 NIDI 4811 ENCl 8507

TAGLN 6876 OFDl 8481 ACAN 176

CTSA 5476 CADMl 23705 ZNF518 9849

PDGFRB 5159 STABl 23166 GPR89A 51463

MXRA8 54587 TPST2 8459 RPL18 6141

OSMR 9180 PPP1R15A 23645 MEF2A 4205

COL3A1 1281 PDLIM3 27295 DNASElLl 1774

GREMl 26585 ATPIFl 93974 MYOlB 4430

FAP 2191 TRIM33 51592 JPH2 57158

DBNl 1627 MMP3 4314 [0114] The inventors measured that increased expression of TOP2A index coupled with reduced expression of genes of Table 13 increased the prognosis for chemotherapy (anthracyclines) . [0115] The inventors coupled the gene set of the invention, with a gene set comprising several genes from Table 14 (ESR-) .

[0116] The inventors measured that increased expression of TOP2A index coupled with increased expression of genes of Table 14 increased the prognosis for chemotherapy (anthracyclines) .

Table 14: ESR- Module (Gene ID means Entrez Gene ID)

Gene IDl Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol

2099 ESR1 3295 HSD17B4 2800 GOLGA 1 51103 NDUFAF1

23158 TBC1D9 11013 TMSL8 8326 FZD9 11042 NA

2625 GATA3 51604 PIGT 6376 CX3CL1 10040 TOM 1L1

771 CA12 6663 SOX10 8399 PLA2G10 1117 CHI3L2

3169 FOXA 1 85377 MICALL1 5327 PLAT 112398 EGLN2

4602 MYB 58495 OVOL2 22885 ABLIM3 9258 MFHAS1

7802 DNALI1 1116 CHI3L1 11094 C9on7 374 AREG

18 ABAT 11001 SLC27A2 5321 PLA2G4A 2982 GUCY1A3

7494 XB P 1 25841 ABTB2 57348 TTYH 1 688 KLF5

57758 SCUBE2 64080 RBKS 6787 NEK4 1960 EGR3

2066 ERBB4 375035 SFT2D2 123872 LRRC50 ^" , 7993 UBXD6

9 NAT1 10479 SLC9A6 10421 CD2BP2 25823 TPSG1

10551 .AGR2 5002 SLC22A18 5971 RELB 4485 MST1

987 J LRBA 8645 KCNK5 6833 ABCC8 23528 ZNF281

56521 ' DNAJC12 79885 HDAC11 11122 PTPRT 1672 DEFB1

2203J FBP1 11254 SLC6A14 23650 TRIM29 28960 DCPS

^"51466| EVL 122616 _;C14orf79 79629 OCE L1 5268 SERPINB5

51442-VGLL1 79650 C16or157 8722 CTSF 934 CD24

^{" "} 57496 MKL2 23321 TRIM2 57110 HRASLS 55450 CAMK2N1

7031 TFF1 23327 NEDD4L 6697 SPR 6261 RYRΪ

1153t CIRBP 22977IAKR7A3 2919^' CXCL 1 2627 GATA6

^"26227^"[PHGDH 8581 ,^'LΫ6D 27250 PDCD4 57180 ACTR3B

1555 CYP2B6 8842J PROM1^" 23245 ASTN2 _; 4036 LRP2

6648_| ^"SOD2 4953 ODC1 10265 IRX5 29116 MYLIP

55638^'[NA 55544- RBM38 2824^"^GPMeB 57211 GPR126

221061|CΪ0orf38 55663 i^"ZNF446^" 10644 IGF2BP2 4435' CITE D1

7033;TFF3 27124 PIB5PA 7436 VLDLR 54913 RPP25

53335' BCL11A 6715 SRD5A1 25825 BACE2 9982 FGFBP 1

79818 ZNF552 51809 GALNT7 10827 C5orf3 11170 FAM107A

57613; KIAA1467 89927 C16orf45 4828 NMB 3294 HSD17B2

8416 ANXA9 1827 DSCR1 6720 SREBF1 6583 SLC22A4

582: BBS1 51706 CYB5R1 10477 UBE2E3 79170 ATAD4

54463 j NA ^{' "} 33831 ICAM 1 ^"3066; HDAC2 79745 CLIP4

55733 HHAT 5806_j PTX3 , 55224 ETNK2 2813_[GP2

^" 2674 GFRA 1 9501 J RPH3AL ^" 875! CBS ^"6723' SRM

4478J MSN ; 3613|IMPA2 3872: KRT17 136θ[cPB1

51097¹ SCCPDH 7568^!ZNF20 753 C18orfϊ ^l 5016 OVGP 1

^"54502 [NA 6280 S100A9 136 ADORA2B 5271 SERPINB8

26018; LRIG1 22929 SE PHS 1 2013 EMP2 347902 AMIGO2

55793| FAM63A^{~ '} 81563 C1orf21 ^" 1917 EEF1A2 79719 NA

3868 KRT16 1389 CREBL2 3576 IL8 55258 NA

54961 SSH3 1410 CRYAB 419 ART3 8563 THOC5

60481 ELOVL5 Tθ884:MRPS3(J^{" " ""}55650^' PIGV^" 83464 APH1 B

36671 IRS 1 55614;C20or123 23107. MRPS27 23532; PRAME

83439 JTCF7L1 1824 DSC2 25818 KLK5 6834 SURF 1

^" 10950 | BTG3 ^{' ~} 78~5Ϊ^' MALL 8309 ACOX2^* 60119 RLN2

3572_| IL6ST 2743| GLRB 1047. CLGN 214 ALCAM

^" 4783ΪNFIL3^{" ' ' "} 427TASAHI 10002 NR2E3 55333 SΫNJ2BP

511^~61 . C3orf18^~ 524Ϊ| PGR 60487 TRMT11 10525 HYOU 1

2296| FOXC1 51364^"'.ZMYNDIO 10656 KH DRBS 3 2232 FDXR

6664 SOX11 6926 TBX3 ^{" " "} 55240 STEAP3 274 BIN1

5613 PRKX 5193 PEX12 3315 HSPB1 10307 APBB3

8543 LMO4 8531 CSDA 10273 STUB1 8986 RPS6KA4

55686; MREG 23 ABCF1 2171 FABP5 56938 ARNTL2

8100 IFT88 7545 ZIC1 55184 C20orf12 9510 ADAMTS 1

^' 2617 GARS 819 CAMLG 5783 PTPN 13 2770 GNAH

3945; LDHB 2947 GSTM3 1877 E4F1 4350 MPG

8382i^rNME5 5825 ABCD3 11098 PRSS23 863 CBFA2T3

10614 HEXJM1 5860 QDPR ^" 10202 DH RS 2 2891 GRIA2 Table 14 : ESR- Module ( Ctd . )

Gene ID Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol

9633 MTL5 59342 SCPEP1 80223 RAB11 FIP1 10309 UNG2

2568 GABRP 51806 CALML5 79627 OGFRL1 7037 TFRC

23324 MAN2B2 79603 LASS4 6948 TCN2 3574 IL7

55765 C1orfl06 21 ABCA3 3097 HIVEP2 55293 UEVLD

5104 SERPINA5 54847 SIDT1 8985 PL0D3 27165 GLS2

5174 PDZK1 8537 BCAS 1 3892 KRT86 55188 RIC8B

56674 7MEM9B 10874 NMU 10575 CCT4 11202 KLK8

1054 CEBPG 54149 C21orf91 51004 COQ6 51181 DCXR

9120 SLC16A6 9929 JOS D1 4071 TM4SF1 827 CAPN6

79641 ROGDI 5317 PKP1 1718 DHCR24 390 RND3

23303 KIF13B 7388 UQCRH 1381 CRABP1 54438 GFOD1

2173 FABP7 64764 CREB3L2 9368 SLC9A3R1 10079 ATP9A

23171 GPD1 L 10127 ZNF263 92104 TTC30A 4285 MIPEP

9674 KIAA0040 80347 COASY 9518 GDF15 8324 FZD7

27134 TJP3 126353 C19orf21 6364 CCL20 9052 GPRC5A

79921 TCEA L4 50865 HEBP1 3306 HSPA2 9508 ADAMTS3

54898 ELOVL2 54812 AFTPH 79605 PGBD5 10519 CIB1

1345 COX6C 64087 MCCC2 23336 DMN 7138 TNNT1

5937^*RBMS1 8884 SLC5A6 1356 CP 51735 RAPGEF6

400451 NA 5269 SERPINB6 54619 CCNJ 54970 TTC12

3898 LAD1 4321 MMP12 9200 PTPLA 2591 GALNT3

2530 FUT8 8190 MIA 51302 CYP39A1 2348 F0LR1

51306 C5orf5 6769ISTAC 5191 PEX7 2954 GSTZ1

25837 RAB26 51368[TEX264 706^TSPO 23318 ZCCHC11

10982 MAPRE2 23541 SEC14L2 7159 TP53BP2 10267 RAMP1

1632 DCI 9185^!REPS2 55218 EXDL2 25984 KRT23

7905 REEP5 185 AGTR1 79669 C3orf52 6496 SIX3

1101 CHAD 7368 UGT8 10140 TOB1 786 CACNG1

323 APBB2 399665 FAM102A 11226 GALNT6 22976 PAXIP1

28958 CCDC56 12 SERPINA3 6652 SORD 283232 TMEM80

1476 CSTB 55975 KLHL7 3418 IDH2 629 CFB

9435 CHST2 25864 ABHD14A 10200 MPHOSPH6 7286 TUFT1

7371 UCK2 4851 NOTCH1 7345 UCHL1 5562 P RKAA 1

2737 GLI3 9091 PIGQ 6564 SLC15A1 9851 KIAA0753

8685 MARCO 1299 COL9A3 54458 PRR13 79622 C16orf33

55316 RSAD1 83988 NCALD 60686IC14orf93 23266 LPHN2

6271 S100A1 2920 CXCL2 8792 TNFRSF11A 29104 N6AMT1

55859 BEX1 8870 IER3 54894 RNF43 1783 DYNC1 LI2

3595 IL12RB2 55245 C20orf44 5737 PTGFR 8987 NA

5100] PCDH8 6666 SOX12 1501 CTNND2 79852 ABHD9

2861 GPR37 80279 CDK5RAP3 7764 ZNF217 57586 SYT13

26278,SACS 1644 DDC 8405 SPOP 8785 MATN4

55506 H2AFY2 5441 POLR2L 1847 DUSP5 10331 B3GNT3

64215 DNAJC1 9022 CLIC3 4488 MSX2 5357 PLS 1

3096 HIVEP1 7769 ZNF226 7163 TPD52 54880 BCOR

23059 CLUA P 1 27239 GPR162 25790 CCDC19 55790 NA

79602 ADIPOR2 26504 CNNM4 5803 PTP RZ1 4139 MARK1

^* 56683^" 021OFfSg 3400 ID4 ^~23635^rSSBP2 81539 SLC38A1

22943, DKK1 1733, DIO1 6548 SLC9A1 10810 WASF3

6277 S100A6 ^" 25915 C3ori60 8187 ZNF239 926 CD8B

65983 GRAMD3 1525 CXADR 2588 GALNS 50805 IRX4

4255 MGMT 1475 CSTA 54903 MKS1 58513 EPS15L1

10406 WFDC2 2155 F7 55163 PNPO 6304 SATB 1

3760 KCNJ3 4188 MDFI 55101 NA 79446 WDR25

23552 CCRK 3622 ING2 4682 NUBP1 23366 NA

9722 NOS 1AP 25980 C20orf4 3779 KCNM B 1 55699 IARS2

23613 PRKCBP1 8310 ACOX3 64849 SLC13A3

202 AIM1 54820 NDE 1 4691 NCL

51207 DUSP13 5816 PVALB 64428 NARFL L[0u.1.17. j] The inventors coupled the gene set of the iinnvveennttion, with a gene set comprising several genes from Table 15.

Table 15: Erbb2 module (Gene ID means Entrez Gene ID)

Gene ID Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol

2064 ERBB2 147179 WIPF2 1573 CYP2J2 404093 CUEDC1

93210 PERLD1 55040 EPN3 26154 ABCA12 3675 ITGA3

5709 PSMD3 5245 PHB 3081 HGD 55129. TMEM16K

5409 PNMT 9635 CLCA2 8804 CREG1 24147 FJX1

55876 GSDML 3227 HOXC11 9914 ATP2C2 1048 CEACAM5

22794 CASC3 29095 ORM DL2 5129 PCTK3 9572 NR1D1

3927 LASP1 5909 RAP 1GAP 54793 KCTD9 51375 SNX7

[0118] The inventors coupled the gene set of the invention, with a gene set comprising several genes from Table 16.

[0119] The inventors measured that increased expression of TOP2A index coupled with decreased expression of genes of Table 16 increased the prognosis for chemotherapy (anthracyclines) .

Table 16 : AURKA proliferation module

Gene ID Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol

6790 AURKA 2331 FMOD 51204 CCDC44 3028 HSD17B10

11065 UBE2C 51512 GTSE1 54845 RBM35A 26519 T1MM10

9133 CCNB2 6424 SFRP4 283 ANG 25960 GPR124

1058 CENPA 55353 LAPTM4B 79652 C16orf30 10252 SPRY1

332 BIRC5 8404 SPARCL1 56944 OLFML3 6199 RPS6KB2

11004 KIF2C 990 CDC6 3297 HSF1 9824 ARHGAP11A

10112 KIF20A 7043 TGFB3 27235 COQ2 55630 SLC39A4

991 CDC20 11047 ADRM 1 2487 FRZB 7049 TGFBR3

2305 FOXM1 58190 CTDSP1 3251 HPRT1 8607 RUVBL1

^" 891 CCNB1 ^{* "} 79838 TMC5 5119 PCOLN3 2581 GALC

22974 TPX2 84823 LMNB2 ^" 6839 SUV39H1 862 RUNX1T1

9088 PKMYT1 ^"83989 C5or121 27303 RBMS3 8458 TTF2

54478 FAM64A 1793 DOCK1 10468 FST 9775 EIF4A3

4751 NEK2 9358 ITGB L1 26289 AK5 3181 HNRPA2B1

24137 KIF4A 8836 GGH 55038 CDCA4 26039 SS18L1

23397 NCAPH 57088 PLSCR4 7283 TUBG1 10580 SORBS 1

9319 ^"TKIP 13 6642 SNX1 23212 RRS1 7056 THBD

4085 MAD2L1 4969 OGN 65094 JMJD4 8322 FZD4

9156 EXO1 90627 STARD13 55379 LRRC59 1003 CDH5

10615 S PAG5 11260 XPOT 10956 NA 2152 F3

7083 TK1 22827 NA 51022 GLRX2 55068 NA

6491 STlL 9793 CKAP5 54915 YTHDF1 64785 GINS3

6241 RRM2 ' 2791 GNG11 54861 SNRK 79042 TSEN34

55839^CENPN ^" 55247 "NEIL3 79000 C1orf135 8805 TRIM24

7298 TYMS 10234 LRRC17 79776 ZFHX4 1478 CSTF2

641 BLM 9353 SLIT2 79971 GPR177 1746 DLX2

4171 MCM2 1841 DTYMK 7718 ZNF165 57125 PLXDC1

1164 CKS2 9631 NUP155 201254 STRA13 22998 NA

79682 M LFI IP 5424 P0LD1 1848 DUS P6 79915 C17orf41

10129 FRY 6631 SNRPC 9037 SEMA5A 7026 NR2F2

^{" ~}51659 GINS2 10186 LHFP 5433 POLR2D 7474 WNT5A

10212 DDX39 452I i NUDTI 29087¹THYNI 55857 C20orf19

3925 STMN1 3479 IGF1 ^" 79864 C11orf63 114625 E RMAP^"

7980rSHCBP1 4172IMCM3 358 AQP1 8857 FCGBP

3014 H2AFX 2205I FCERTA^"" 136341SNRPD3 26872 STEAP 1

10535 RNAS E H2A ^" 55732^^112 2621 GAS6 7226 TRPM2

5984 RFC4^{" "} 9077 DIRAS3 562>0 WDR45L 29844 TFPT

55970 GNG Ϊ2 5557. PRIM1 5187 PER1 4719 NDUFS1

1033 CDKN3 54963 UCKL1 2098 ESD 4013 LOH11CR2A

55388 MCM 10 54512 EXOSC4 81887 LAS1 L 3396 ICT1

55257 C20orf20 79901 CYBRD1 1811 SLC26A3 397 ARHGDIB

1163 CKS1 B 10161 P2RY5 54535 CCHCR1 10436 EMG1

8914 TIMELESS 29097 CNIH4 55526 DHTKD1 51582 AZIN1

54821 NA 6513 SLC2A1 57161 PELI2 10598 AHSA1

23371^' TENC1 51123¹ZNFVOe 2353 FOS 333 APLP1

8544 PIR 857 CAV1 ^" 51279 C1 RL 51142 CHCHD2

83Ϊ7 CDC7 51110¹ LACTK- 60436 TGIF2 27123 DKK2 ^"

55020¹ NA ^" 58500 ZNF250 4649 MYO9A 594 BCKDHB

^" 23460 ABCA6 11081 KERA 53820 DSCR6 6286 S100P

64321^" SOX17 7064 THOP1 3638 INSIG1^' 3954 LETM 1

7098 TLR3 55799 CACNA2D3 11171 STRAP 51087 YBX2

6338 SCNN1 B 49855 ZNF291 10992 SF3B2 10953 TOMM34

3692 ITGB4BP 54606 DDX56 6832 SUPV3L1 8317 CDC7

10253 SPRY2 7164 TPD52U 55922 NKRF 51110 LACTB2

2669 GEM 80775 TMEM 177 10557 RPP38 60436 TGIF2

79679 VTCN 1 667 DST 3216 HOXB6 27123 DKK2

79618 HMBOX1 2781 GNAZ 54785 C17orf59

8772 FADD 23464,GCAT 1933ΕEF1B2

9986 RCE 1 79763 ISOC2 8161 , COIL [0120] The inventors coupled the gene set of the invention, with a gene set comprising several genes from Table 17 (VEGF) .

Entrez Gene lD Gene Symbol Entrez Gene lD Gene Symbol Entrez Gene lD Gene Symbol

7422 VEGFA 6166 RPL36AL 22809 ATF5

911 CD1C 9450 LY86 23417 MLYCD

4005 LMO2 22900 CARD8 23592 LEMD3

4222 MEOX1 1776 DNASE1 L3 51621 KLF13

29927 SEC61A1 1119 CHKA

[0121] The inventors coupled the gene set of the invention, with a gene set comprising several genes from Table 18 (CASP3) .

Gene ID Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol Gene ID Gene Symbol

836 CASP3 7738 ZNF 184 8237 USP11 25978 CHMP2B 10393 iANAPCIO 3728 JUP 402 IARL2 6301 SARS

55361 NA 5977 ;DPF2

METHODS Patients Study population

[0122] The neoadjuvant prospective "TOP" (Trial of Principle) trial was conducted at different European hospitals and coordinated by the Institut Jules Bordet. This study is registered on the clinical trials site of the US National Cancer Institute website http://clinicaltrials.qov/ct2/show/NCT00162812?term=NCT0016 2812&rank=l . One hundred and forty-nine patients have been included in this trial. Anthracyclines (Epirubicin) monotherapy (100 mg/m2) was given as neo-adjuvant chemotherapy: every 3 weeks x 4 cycles for early BC or every 2 weeks x 6 cycles with Neulasta® 6mg on day 2 for patients with inflammatory or locally advanced breast cancer. At completion of chemotherapy, every patient underwent surgery with axillary node sampling. After surgery, adjuvant docetaxel (100 mg/m2 x 4 cycles) and loco-regional irradiation were administered using standard criteria. This study was primarily designed for the identification of biomarkers of response to anthracyclines (epirubicin) .

[0123] All patients underwent pretreatment core biopsies of the primary breast tumor before starting neoadjuvant chemotherapy using a 14G needle. Two biopsies were embedded in OCT (Sakura) , frozen in liquid nitrogen within 5 minutes and transferred to a -80⁰C freezer. Two biopsies were fixed in formalin and embedded in paraffin. Both fixed and frozen samples were retrieved and stored at the Institut Jules Bordet in Brussels, where the TOP2A evaluations were carried out. Pathologic response was determined by microscopic examination of the excised tumor and nodes after completion of chemotherapy. Pathological complete response (pCR) was defined by the absence of residual invasive breast carcinoma (macro and microscopic) in the breast and in the axillary nodes. Persistence of in- situ carcinoma without invasive component was considered pCR.

[0124] This analysis was performed after the clinical data until surgery had become available for all the patients. The clinical data was collected, monitored and validated by the BrEAST data centre, Institut Jules Bordet. This study has been approved by the local ethics committees and all patients have given written informed consent prior to study entry.

Validation cohort

[0125] This validation cohort includes patients reported by Bonnefoi et al . These patients are a subgroup of the patients included in the prospective phase III intergroup trial of neoadjuvant chemotherapy (European Organisation for Research and Treatment of Cancer [EORTC] 10994/Breast International Group [BIG] 00-01, registered as NCT00017095) .

[0126] Patients were randomly assigned to a non- taxane regimen of six cycles of 500 mg/m² fluorouracil, 100 mg/m² anthracyclines (epirubicin) , and 500 mg/m² cyclophosphamide (FEC) treatment, or to three cycles of 100 mg/m² docetaxel followed by three cycles of 90 mg/m² anthracyclines (epirubicin) plus 70 mg/m² docetaxel (TET) . This substudy was restricted to ER-negative tumours . The definition used in this study for pCR was slightly different than the one used in our study. It was defined as disappearance of the invasive component of the primary tumour after treatment, with at most a few scattered tumour cells detected by the pathologist in the resection specimen, but did not consider the disappearance of the invasive component in the axillary lymph nodes. Microarray data were deposited in the Gene Expression Omnibus database under accession number GSE6861.

Fluorescent Hybridization In Situ (FISH)

[0127] FISH assays were done with the Abbott Multicolor TOP2A Spectrum orange, HER2 Spectrum green and CEP17 Spectrum aqua probe. Briefly, the sections were deparaffined and incubated in pre- treatment buffer at 8O⁰C for 30 min. Enzymatic digestion was carried out with pepsin (10-20 min at 37°C) and the slides were dehydrated in graded dilutions of ethanol . The probe (10 μl) was applied to the slides under coverslips. The slides were co- denatured on a hot plate (73⁰C for 5 min) , followed by overnight hybridization at 37⁰C. After stringency washing

(2X SSC/0.3% Nonidet P-40 at 73°C for 2 min) , the slides were counterstained with 10 μl of 0.2 μM 4 , 6-diamino-2- phenylindole (DAPI) in antifade solution (Vectashield, Vector Laboratories, Inc., Burlingame, CA) . FISH was evaluated using an Olympus BX51 epifluorescence microscope. The invasive part of the tumor was circled on the slides with a diamond by superposition with haematoxylin- stained sections previously analyzed by the pathologist (DL) . Signals from at least 60 non-overlapping nuclei with intact morphology were evaluated to determine the mean number of signals/cell (ratio between mean number of TOP2A or HER2 signals and the mean

Immunohistochemistry (IHC)

[0128] TOP2A protein expression was evaluated by IHC. Briefly, the sections were dewaxed, rehydrated and incubated for 30 min in 0.5% hydrogen peroxide (H₂O₂) in methanol. After pretreatment (0.1% trypsine in 0.1% CaC12 pH 7.8 for 10 min at 37⁰C), non-specific staining was blocked by incubating in 10% normal serum for 1 h at 4⁰C and performing all steps in buffer PBS/0.1% BSA/1% Tween- 20. After incubation with the primary antibody (1 μg/ml overnight at 4⁰C, clone KiSl, Boehringer-Mannheim) , sections were incubated at room temperature with a secondary biotinylated anti-mouse antibody for 30 min and Streptavidin-HRP (Zymed) (1/20) for 10 min. 3'3- diaminobenzidine was used as a chromogen. Sections were counterstained with Mayer's haematoxylin. Negative controls consisted in serial sections incubated with buffer alone instead of primary antibody. Tonsil samples were used as positive control.

[0129] Ki-67 evaluation was carried out routinely by IHC using the monoclonal mouse antibody MIB-I (1/50, Dako, Carpinteria, CA) . In brief, the sections were dewaxed and rehydrated. Then a microwave (two times 10 min at 650 W) antigen retrieval method in citrate buffer pH 6.0 was implemented before using the Ventana Nexes automated immunostainer with standard Nexes reagents (Ventana Medical Systems, Tucson, AZ) . A cutoff of 25% of positively stained cells was used (Durbecq et al. 2004) .

Gene Expression Profiling

[0130] One 5-μm tissue section (usually after 10 30- μm sections) of each biopsy were hematoxylin and eosin stained to monitor the tumor cell percentage of the tissue. Only specimens with more than 30% of tumor cells were included in further analysis. Isolation of RNA was performed using the Trizol method (Invitrogen) according to the manufacturer's instructions and purified using RNeasy mini-columns (Qiagen, Valencia, CA) . The quality of the RNA obtained from each tumor sample was assessed based on the RNA profile generated by the bioanalyzer (Agilent Inc) . RNA amplification, hybridization and image scanning were done according to standard Affymetrix protocols. The inventors have used the Affymetrix Human Genome U133-2.0 plus GeneChip .

Statistical analysis

[0131] Correlations between continuous variables were assessed using the non-parametric Spearman coefficients. Correlations between binary variables were reported using the kappa statistics. Other correlations between categorical variables were performed using the chi- square test. Correlations between continuous and categorical variables were performed using the Mann-Withney U (for binary variables) or Kruskal-Wallis test. [0132] Both in the study population and validation cohort, HER2+ and HER2- patients were identified using the bimodality of the HER2 mRNA expression as previously described. [0133] The TOP2A index and other prognostic indexes were computed for each sample as module score = ∑w,x₍ /∑l ^w, I where x- is the expression of a gene included in the index and w^ is its coefficient. This index is a combination of table 1 genes.

[0134] More precisely, the inventors used an index made of the preferred genes :TOP2A, THRA, CDC6 and RARA.

When different probe- sets were available for one gene, the inventors considered the one with the greatest variance

(THRA: 31637_s_at, CDC6 : 203 967_at, RARA: 203 749_s_at and

TOP2A: 201 291_s_at) .

[0135] The signatures predictive of pathological response were identified through stability-based feature ranking as scoring function.

[0136] Once the signature was identified, a signature score was computed for each patient using the following formula :

where s is the signature score, x- is the expression of a gene in the signature, w- is either +1 or -1 depending on the sign of the association with pathological response. The signature scores were scaled such that quantiles 2.5% and 97.5% are equaled to -1 and +1 respectively. This scaling is robust to outliers and ensured that the scores lay approximately in [-1,+I] .

[0137] The area under the curve (AUC) was used to assess the prediction performance of a signature score. AUC was estimated through the concordance index, its confidence interval and significance being estimated assuming asymptotic normality. The corresponding p-values were onesided.

[0138] The inventors computed the prognostic value of a signature score through a meta-analytical framework. [0139] The corresponding gene module scores were computed in order to identify the breast cancer molecular subtype in this two dimensional space using a mixture of three Gaussians. The three subtypes are denoted by ER- /HER- (basal) , HER2+ (ERBB2 -enriched) and ER+/HER2- (luminal) .

Functional analysis

[0140] Functional analysis of the gene signatures was performed using Ingenuity Pathways Analysis (IPA) tools version 3.0. Affymetrix probe sets of each cluster were used as input and IPA then calculated a significance value for enrichment of the functional classes. Only significant functions are shown.

Claims

1. A gene set comprising or consisting of 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 16, 17, 18, 19, 20, 21, 22 or all the genes of Table 1.

2. The gene set of claim 1 comprising or consisting of THRA (Gene ID: 9572), CDC6 (Gene ID: 990), RARA (Gene ID: 5914) and TOP2A (Gene ID: 7153) . 3. The gene set of claim 2 further comprising 1, 2,

3, 4, 5, 6, 7, 8, 9, 10, 15, 16, 17, 18 or all the (other) genes of Table 1.

4. The gene set according to any of the preceding claims, further comprising 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 or all the genes of Table 5 and/or Table 6.

5. The gene set according to any of the preceding claims, further comprising 1, 2, 3, 4, 5, 6, 7,

8, 9, 10 or all the genes of Table 10 and/or of Table 11. 6. The gene set according to any of the preceding claims further comprising 1, 2, 3, 4, 5,

6, 7, 8,

9, 10 or all the genes of Table 12 and/or of Table 13.

7. A diagnostic kit or device comprising fixed upon a solid support, capture probes detecting specifically the expression of complementary target genes of the gene set according to any of the preceding claims and possibly other means for real time PCR analysis.

8. A diagnostic kit or device comprising fixed upon a solid support, capture probes detecting specifically the expression of complementary target genes of a gene set comprising or consisting of 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45; 50, 55, 60, 65, 70, 75 or all the genes of Table 5.

9. A diagnostic kit or device comprising fixed upon a solid support, captures probes detecting specifically the expression of complementary target genes of a gene set comprising or consisting of 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 or all the genes of Table 6.

10. The kit or device according to any of the preceding claims 7 to 9 being a computerized system comprising - a bio-assay module configured for detecting a gene expression or protein synthesis from a tumor sample (a breast tumor sample) based upon the gene set, kit or according to any of the preceding claims, a processor module configured to calculate expression of the said genes expression or the said protein synthesis and to generate a risk assessment for the tumor, preferably breast tumor sample.

11. A method for prediction (or prognosis or prognostic) of cancer in mammal subject, which comprises the step of measuring gene expression in a (breast) tumor sample obtained from the said mammal subject, by putting into contact nucleotide sequences obtained from this tumor sample with the gene set, the kit or the device according to any of the preceding claims and possibly generating a risk assessment for the said tumor sample by designating the outcome of a regimen comprising administration to the said mammal subject, of one or more Anthracycline compound (s) .

12. The method of claim 11 wherein the patient is a human patient, preferably suffering from a breast cancer preferably an ER- breast cancer (more preferably a ER- Her2+ breast cancer) .

13. An Anthracycline compound for use in the treatment and/or the prevention of cancer having an increased risk assessment as measured by the method of claim 11 or 12 and/or by the gene set, kit or device according to claims 1 to 10.

14. An Anthracycline compound or a mixture of Anthracycline compounds for use in the treatment and/or the prevention of a (breast) cancer having an increased risk assessment as measured by the method of claim 11 or 12 and/or by the gene set, kit or device according to claims 1 to 10, preferably in ER- patients, more preferably in ER- Her2- breast cancer patients.

15. An Anthracycline compound or a mixture of Anthracycline compounds for use in the treatment and/or the prevention of breast cancer having an increased risk assessment as measured by the method of claim 11 or 12 and/or by the gene set, kit or device according to claims 1 to 10 in ER- Her2+ patients.