WO2017096458A1 - Immune gene signature predictive of anthracycline benefit - Google Patents

Abstract

There is described herein a method of predicting a benefit of anthracycline therapy for a subject with breast cancer, the method comprising: a) providing a sample of a breast cancer tumour of the subject; b) determining the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1, SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; c) comparing said expression levels to a reference expression level of the group of genes from control samples from a population; and d) determining the benefit of anthracycline therapy for the subject; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to an immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with a lesser benefit of anthracycline therapy

Description

IMMUNE GENE SIGNATURE PREDICTIVE OF ANTHRACYCLINE BENEFIT

CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of priority of U.S. Provisional Patent Application Nos. 62/263,831 and 62/264,029, both filed December 7, 2015 and incorporated herein by reference in their entirety.

FIELD OF INVENTION

[0002] The present disclosure relates generally to prognosing or classifying a subject with breast cancer. More particularly, the present disclosure relates to methods and devices directed to predicting anthracycline treatment benefit in a subject with breast cancer.

BACKGROUND

[0003] Breast cancer is the leading cause of cancer death for women. Besides surgery and radiation, treatment options for breast cancer patients include chemotherapy, endocrine therapy and trastuzumab. Anthracyclines are commonly used as they non- specifically target rapidly-proliferating cells. Consequently, patients treated with anthracyclines may recover from breast cancer, but suffer severe side-effects such as cardiac toxicity and leukemia. Conversely, patients may not respond to anthracyclines at all or develop drug resistance after prolonged use, which includes cross-resistance to structurally unrelated anti-cancer drugs.¹ It has been identified that TOP2A and CEP17 (duplication of centromeric region on chromosome 17), a surrogate marker of chromosomal instability,^2"4 as well as CIN4 (chromosomal instability)⁵ are predictive biomarkers of clinical anthracycline sensitivity; yet it remains to be established whether these biomarkers can be targeted with drugs in clinic. Therefore, it is not only important to spare non-responders from unnecessary side-effects by discovering novel biomarkers, but also to identify new therapeutic approaches to improve patient survival. [0004] The contribution of immune cells is well appreciated in cancer development, progression and therapy resistance. Tumour-infiltrating lymphocytes (TIL) are detected in both stromal and intratumoral compartments of breast cancer;⁶ their numbers vary and are greater in estrogen receptor negative (ER^") and HER2⁺ subtypes rather than ER⁺ cancers.⁷ Furthermore, CD4⁺ and CD8⁺ T cells outnumber other lymphocytes as well as myeloid cells.⁸ However, only recently has TIL's translational potential as cancer-associated prognostic and predictive markers emerged. In particular, high densities of TILs correlate with improved clinical outcome in triple negative breast cancers (TNBC); ^9;10 which was followed by the finding of a TIL-gene signature as a prognostic biomarker in TNBC.¹¹ [0005] However, whether TILs can also predict specific chemotherapy benefit in breast cancer patients remains largely unknown. In neoadjuvant setting, lymphocytes were predictive of response to chemotherapy but only in lymphocyte-predominant breast cancers (LPBC) defined as having >50-60 lymphocytic infiltration.¹² This is an arbitrary cut-off point that the researchers used to demonstrate the principle, rather than a biological subtype of breast cancer. In adjuvant setting, intratumoral, but not stromal CD8⁺ T cells were shown to be predictive of anthracycline therapy but only in ER^" breast cancers.¹³ Lastly, in HER2 amplified cancers, some studies have reported that patients whose breast cancers have LPBC phenotype benefit from Herceptin¹⁴'¹⁵, whereas other studies have shown the opposite effect.¹⁶ Therefore, breast cancers still need to be evaluated specifically for lymphocyte populations and profiled for their functional orientation, type and effector function in order to determine their predictive biomarker potential.

SUMMARY OF INVENTION

[0006] In an aspect, there is provided a method of predicting a benefit of anthracycline therapy for a subject with breast cancer, the method comprising: a) providing a sample of a breast cancer tumour of the subject; b) determining the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; c) comparing said expression levels to a reference expression level of the group of genes from control samples from a population; and d) determining the benefit of anthracycline therapy for the subject; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to an immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with a lesser benefit of anthracycline therapy.

[0007] In an aspect, there is provided a computer-implemented method of predicting benefit of anthracycline therapy for a subject with breast cancer, the method comprising: a) receiving, at at least one processor, data reflecting the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; b) comparing, at the at least one processor, said expression levels to a reference expression level of the group of genes from control samples from a population; c) outputting, at the at least one processor, an immune score; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to the immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with lesser benefit of anthracycline therapy. [0008] In an aspect, there is provided a computer program product for use in conjunction with a general-purpose computer having a processor and a memory connected to the processor, the computer program product comprising a computer readable storage medium having a computer mechanism encoded thereon, wherein the computer program mechanism may be loaded into the memory of the computer and cause the computer to carry out the method described herein.

[0009] In an aspect, there is provided a computer readable medium having stored thereon a data structure for storing the computer program product described herein.

[0010] In an aspect, there is provided a device for predicting benefit of anthracycline therapy for a subject with breast cancer, the device comprising: at least one processor; and electronic memory in communication with the at one processor, the electronic memory storing processor-executable code that, when executed at the at least one processor, causes the at least one processor to: a) receive data reflecting the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; b) compare said expression levels to a reference expression level of the group of genes from control samples from a population, said reference expression level being stored in the memory; c) output an immune score; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to the immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with a lesser benefit of anthracycline therapy. [0011] In an aspect, there is provided a method of treating a subject with breast cancer, comprising: a) determining the immune score of the subject according to the method described herein; and b) selecting a treatment based on said immune score, and preferably treating the subject according to the treatment.

[0012] In an aspect, there is provided a composition comprising a plurality of isolated nucleic acid sequences, wherein each isolated nucleic acid sequence hybridizes to: (a) the mRNA of a group of genes comprising at least one of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; and/or (b) a nucleic acid complementary to a).

[0013] In an aspect, there is provided an array comprising one or more polynucleotide probes complementary and/or hybridizable to an expression product of each gene of a group genes comprising at least three of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3.

[0014] In an aspect, there is provided a kit comprising reagents for detecting mRNA from a sample of a breast cancer tumour of at least three genes selected from the group comprising: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3. [0015] Other aspects and features of the present disclosure will become apparent to those ordinarily skilled in the art upon review of the following description of specific embodiments in conjunction with the accompanying figures.

BRIEF DESCRIPTION OF FIGURES [0016] Embodiments of the present disclosure will now be described, by way of example only, with reference to the attached Figures.

[0017] FIG. 1 is a modified Reporting Recommendations for Tumour Marker Prognostic Studies (REMARK) diagram of the BR9601 trial. The trial recruited 374 pre- and post-menopausal women with completely excised, histologically confirmed breast cancer and a clear indication for adjuvant chemotherapy. Patients were randomized to receive either eight cycles of CMF (i.v. cyclophosphamide 750mg/m2, methotrexate 50mg/m2 and 5-fluorouracil 600mg/m2) every 21 days, or E-CMF (four cycles of epirubicin 100mg/m2 every 21 days followed by four cycles of CMF regimen). Profile shows the total number of samples available for the gene expression analyses.

[0018] FIG. 2 shows an overview of individual gene prediction values. P-values were obtained from the treatment-by-marker interaction calculations. The genes are listed in the following order: FOXP3, CD4, IFNgamma, GZMB, CD3 epsilon, CD45RO, Tbet, CXCL9, CD48, IRF1 , CD8 alpha, VCAM1 , CXCL12, LCK, CXCR3, JAK2, CCL5, CXCL10, ICAM1 , CXCL13, SELL, IL-12, CCR5, CD68, MS4A1 , STAT1 , CCR1 , CCL22, CD19, CCR3, STAT3, CCR4, CXCR2, IL8, CXCR4, MADCAM1 , PRF1 , and CX3CL1. For six genes on the top, FOXP3, CD4, IFNy, GZMB, CD3e, CD45RO,, when their gene expression is low, patients benefit from E-CMF over conventional CMF therapy, whereas when their gene expression is high, there is no difference in patient survival. For two genes on the bottom, PRF1 , CX3CL1 , it is their high expression that predicts for benefit of E-CMF over CMF, whereas there is no difference in patient survival when their gene expression is low.

[0019] FIG. 3 shows directionality of expression of each gene within the 9-immune gene signature. At a patient level, the expression of each gene is categorized as being "low" if it is below the population median, or "high" if it is above the population median. After training for the predictive biomarker of anthracycline resistance, developed 9-gene signature contains patients whose tumours have various expression values (z-scores) of these genes.

[0020] FIG. 4 shows a novel 9-immune gene signature is predictive of benefit from E- CMF (dark line) over CMF (light line). Kaplan-Meier analyses showing A) distant recurrence- free survival (DRFS) and B) breast cancer-specific overall survival (OS).

[0021] FIG. 5 is a heatmap illustrating preprocessing methods. The heatmap shows ranking of preprocessing methods based on their ability to maximize mRNA abundance differences between HER2+ and HER2- samples, while minimizing batch effects. For the 252 combinations of preprocessing methods assessed, two rankings were established using the two criteria, and subsequently aggregated using the rank product. The heatmap is then sorted on the rank product with the most effective preprocessing parameters listed at the top.

[0022] FIG. 6 is a heatmap of mRNA abundance levels scaled as z-scores. Immune- score genes are displayed as rows and patients as columns. Covariate bars for the patients were displayed below showing DRFS, HER2 status, age, grade, N stage and T stage. Patients were sorted based on the distant relapse events and the average mRNA abundance levels. Genes are displayed in rows in the following order: FOXP3, CD4, IFNgamma, GZMB, CD3 epsilon, CD45RO, Tbet, CXCL9, CD48, IRF1 , CD8 alpha, VCAM1 , CXCL12, LCK, CXCR3, JAK2, CCL5, CXCL10, ICAM1 , CXCL13, SELL, IL-12, CCR5, CD68, MS4A1 , STAT1 , CCR1 , CCL22, CD19, CCR3, STAT3, CCR4, CXCR2, IL8, CXCR4, MADCAM1 , PRF1 , and CX3CL1.

DETAILED DESCRIPTION

[0023] High densities of tumour-infiltrating lymphocytes (TILs) correlate with improved clinical outcome in breast cancer; whether TILs also predict anthracycline benefit in all, or only a particular subgroup, of breast cancer patients, remains largely unknown. Furthermore, since identification of TILs is generally based on H&E staining, it has not previously been possible to evaluate relative contribution of distinct T-cell types, and B cells, to patient outcome. [0024] Breast cancer samples were immunoprofiled for T and B lymphocytes and chemokines that are involved in recruiting these populations to the tumour site. A minimal gene signature was identified that would be a predictive biomarker of anthracycline benefit in breast cancer.

[0025] As used herein, "benefit" refers to drug effect and/or efficacy attributable to a drug treatment. For example, a drug benefit includes the impact of the drug on the likelihood of survival of a subject or patient, which can be expressed using overall survival and/or distant relapse-free survival. A "greater benefit" or "lesser benefit" may refer to the benefit of anthracycline therapy on the survival of patient compared to a lack of anthracycline therapy. [0026] Two hundred ninety 290 patient samples were immunoprofiled from the BR9601 adjuvant breast cancer trial (see FIG. 1 ) on the Nanostring platform to gain insight into the lymphocytic populations and develop a TIL gene signature that is predictive of anthracycline benefit. A immunoprofiling panel was used that included 38 TIL genes and chemokines that may be responsible for recruiting TILs to the tumour site.

[0027] The refinement of the 38-gene panel resulted in the generation of a novel 9- gene signature that includes cytotoxic T lymphocytes (CTL) and chemokine genes. In univariate Cox regression analysis, the 9-gene signature was a predictive biomarker of anthracycline benefit with respect to breast-cancer specific OS (HR: 0.371 , 95%CI 0.158- 0.868, p=0.022) and DRFS (HR: 0.395, 95%CI 0.172-0.907, p=0.028); this effect was no longer significant after adjustment for other prognostic factors (OS HR: 0.442, 95%CI 0.165- 1.182, p=0.104; DRFS HR: 0.561 , 95%CI 0.212-1.483, p=0.244).

[0028] Assessing the entire tumour is significant since TILs, tumour and stromal cells collectively engage in a complex interplay that contributes to disease development and progression. Importantly, it reveals that not only CTLs but also chemokines may be clinically relevant and should be validated as potential biomarkers of anthracycline benefit and as therapeutic targets.

[0029] As will become apparent, preferred features and characteristics of one aspect of the invention are applicable to any other aspect of the invention. It should be noted that, as used herein, the singular form "a", "an" and "the" include plural references unless the context clearly dictates otherwise.

Predictive Gene Signature

[0030] In one aspect, this disclosure provides a method of prognosing or classifying a subject or patient with breast cancer. In some aspect, the method predicts benefit of anthracycline therapy for a subject with breast cancer. The method involves determining mRNA abundance using a sample of a breast cancer tumour obtained from the subject.

[0031] As used herein, "anthracyclines" are a class of cell-cycle non-specific drugs used in cancer chemotherapy, and are derived from Streptomyces bacterium Streptomyces peucetius var. caesius. These compounds are used to treat many cancers, such as breast cancer. Examples of Anthracyclines include, but are not limited to, daunorubicin, doxorubicin, epirubicin, idarubicin, and valrubicin.

[0032] As used herein, "breast cancer" and "breast cancer tumour" refers to at least one or more breast cancer types having neither, one, or both of estrogen and progesterone receptors. . ER-positive denotes presence of estrogen receptors, and PR-positive denotes presence of progesterone receptors. Hormone-positive cancer denotes cancer where the cancer cells contain either estrogen or progesterone receptors, and hormone-negative cancer denotes cancer where the cancer cells do not contain either estrogen or progesterone receptors. A breast cancer may also be classified by the level of HER2/neu protein associated with the cancer tumor. HER2-positive denotes cancer with increased levels of HER2/neu and/or increased copies of the HER2/neu gene, and HER2-negative denotes cancer that does not have increased levels of HER2. Triple-negative denotes cancer that does not have estrogen or progesterone receptors and do not have increased levels of HER2. Triple-positive denotes cancer that are ER-positive, PR-positive, and have increased levels of HER2.

[0033] In an aspect, there is provided a method of predicting a benefit of anthracycline therapy for a subject with breast cancer, the method comprising: a) providing a sample of a breast cancer tumour of the subject; b) determining the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; c) comparing said expression levels to a reference expression level of the group of genes from control samples from a population; and d) determining the benefit of anthracycline therapy for the subject; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to an immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with a lesser benefit of anthracycline therapy.

[0034] The term "subject" as used herein refers to any member of the animal kingdom, preferably a human being and most preferably a human being that has breast cancer or that is suspected of having breast cancer. [0035] The term "sample" as used herein refers to any fluid, cell or tissue sample from a subject which can be assayed for biomarker expression products and/or a reference expression profile, e.g. peptides differentially present in a liquid biopsy.

[0036] The term "prognosis" as used herein refers to a clinical outcome group such as a worse survival group or a better survival group associated with a disease subtype which is reflected by a reference profile such as a biomarker reference expression profile or reflected by an expression level of the fifteen biomarkers disclosed herein. The prognosis provides an indication of disease progression and includes an indication of likelihood of death due to breast cancer. In one embodiment the clinical outcome class includes a better survival group and a worse survival group.

[0037] The term "prognosing or classifying" as used herein means predicting or identifying the clinical outcome group that a subject belongs to according to the subject's similarity to a reference profile or biomarker expression level associated with the prognosis. For example, prognosing or classifying comprises a method or process of determining whether an individual has a better or worse survival outcome, or grouping individuals into a better survival group or a worse survival group, or predicting whether or not an individual will respond to therapy.

[0038] The term "gene" as used herein means a polynucleotide which may include coding sequences, intervening sequences and regulatory elements controlling transcription and/or translation. Genes include normal alleles of the gene encoding polymorphisms, including silent alleles having no effect on the amino acid sequence of the gene's encoded polypeptide as well as alleles leading to amino acid sequence variants of the encoded polypeptide that do not substantially affect its function. These terms also may optionally include alleles having one or more mutations which affect the function of the encoded polypeptide's function.

[0039] The term "level of expression" or "expression level" as used herein refers to a measurable level of expression of the products of biomarkers, such as, without limitation, micro-RNA, or messenger RNA transcript expressed or of a specific exon or other portion of a transcript, the level of proteins, peptides or portions thereof expressed of the biomarkers, the number or presence of DNA polymorphisms of the biomarkers, the enzymatic or other activities of the biomarkers, and the level of specific metabolites. As used herein, an expression level of a group of genes refers to the expression level of the group as a whole.

[0040] The phrase "determining the expression" as used herein refers to determining or quantifying RNA or proteins or protein activities or protein-related metabolites expressed by the biomarkers. The term "RNA" includes mRNA transcripts, and/or specific spliced or other alternative variants of mRNA, including anti-sense products. In the case of "protein" or "peptides", it refers to proteins expressed by genes are measurable in a sample.

[0041] The term "expression profile" as used herein refers to a dataset representing the expression level(s) of one or more biomarkers. An expression profile may represent one subject, or alternatively a consolidated dataset of a cohort of subjects, for example to establish a reference expression profile as a control.

[0042] As used herein, the term "control" refers to a specific value or dataset that can be used to prognose or classify the value e.g expression level or reference expression profile obtained from the test sample associated with an outcome class. In one embodiment, a dataset may be obtained from samples from a group of subjects known to have cancer having different tumor states and/or healthy individuals. The expression data of the biomarkers in the dataset can be used to create a control value that is used in testing samples from new patients. In some embodiments, a cohort of subjects is used to obtain a control dataset. A control cohort patients may be a group of individuals with or without cancer. In an exemplified embodiment, the control cohort is the group of individuals with breast cancer, namely the BR9601 trial group.

[0043] In preferred embodiments, a reference expression level is obtained by taking the median expression level. As used herein, "median" is the value separating the higher half of a population from the lower half. In simple terms, it may be thought of as the "middle" value of a dataset, such as the control cohort dataset. For example, a subject is classified into a high immune score group where the subject has an immune score above the population median. On the other hand, a subject is classified into a low immune score group where the subject has an immune score below the population median.

[0044] As used herein, "overall survival" refers to the percentage of or length of time that people in a study or treatment group are still alive following from either the date of diagnosis or the start of treatment for a disease, such as cancer. In a clinical trial, measuring the overall survival is one way to see how well a new treatment works.

[0045] As used herein, "relapse-free survival" refers to, in the case of caner, the percentage of or length of time that people in a study or treatment group survive without any signs or symptoms of that cancer after primary treatment for that cancer. In a clinical trial, measuring the relapse-free survival is one way to see how well a new treatment works. It is defined as any disease recurrence (local, regional, or distant).

[0046] The term "good survival" or "better survival" as used herein refers to an increased chance of survival as compared to patients in the "poor survival" group. For example, the biomarkers of the application can prognose or classify patients into a "good survival group". These patients are at a lower risk of death after surgery.

[0047] The term "poor survival" or "worse survival" as used herein refers to an increased risk of death as compared to patients in the "good survival" group. For example, biomarkers or genes of the application can prognose or classify patients into a "poor survival group". These patients are at greater risk of death or adverse reaction from disease or surgery, treatment for the disease or other causes.

[0048] The term "differentially expressed" or "differential expression" as used herein refers to a difference in the level of expression of the biomarkers that can be assayed by measuring the level of expression of the products of the biomarkers, such as the difference in level of mRNA or a portion thereof expressed. In a preferred embodiment, the difference is statistically significant. The term "difference in the level of expression" refers to an increase or decrease in the measurable expression level of a given biomarker, for example as measured by the amount of mRNA as compared with the measurable expression level of a given biomarker in a control. [0049] In some embodiments, the group of genes is at least 4, 5, 6, 7, 8, or 9 of the genes.

[0050] In some embodiments, the method further comprises building a subject gene expression profile from the determined expression levels of the group of genes.

Anthracycline Therapy [0051] In some embodiments, a subject is treated with anthracycline if the subject is in the low immune score group. In some embodiments, Anthracycline therapy comprises administration of epirubicin. Anthracycline therapy may also comprise other anthracyclines. In other embodiments, the anthracycline therapy further comprises administration of a least one additional chemotherapeutic agent, such as a combination therapy of an anthracyclines (for example epirubicin) and least one additional chemotherapeutic agent.

[0052] As used herein, "chemotherapeutic agent" refers to cytotoxic agents that are known to be of use in chemotherapy for cancer. Chemotherapeutic agents are often combined into chemotherapy regimens for combination chemotherapy. For example, the Cyclophosphamide Methotrexate Fluorouracil (CMF) regimen is a commonly used regimen of breast cancer chemotherapy that combines three anti-cancer agents: cyclophosphamide, methotrexate, and 5-fluorouracil.

Data Processing and Signature Score

[0053] In some embodiments, the method further comprises a signature score comprising a weighted sum expression of each of the group of genes, optionally scaled for imRNA abundance. In preferred embodiments, the signature score may be calculated using equation (2) below:

For each gene g, an indicator function / was run to determine whether the expression level of that sample e_p is above or below the median population gene expression level m_g or the reference expression level. Scores are summed over all genes to calculate the signature score.

[0054] In some embodiments, the subject is classified into a high immune score group where the subject has an immune score above the population median, and wherein the subject is classified into a low immune score group where the subject has an immune score below the population median.

[0055] In some embodiments, in the low immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, or IRF1 was relatively low and STAT3 was relatively high, and in the high immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 was relatively high and STAT3 was relatively low, in comparison to a population cohort.

[0056] In some embodiments, determining the gene expression level comprises use of nanostring.

[0057] In some embodiments, determining mRNA abundance of the genes comprises use of quantitative PCR.

[0058] A person skilled in the art will appreciate that a number of methods can be used to detect or quantify the level of RNA products of the biomarkers within a sample, including arrays, such as microarrays, RT-PCR (including quantitative RT-PCR), nuclease protection assays and Northern blot analyses.

Systems and Devices

[0059] In an aspect, there is provided a computer-implemented method of predicting benefit of anthracycline therapy for a subject with breast cancer, the method comprising: a) receiving, at at least one processor, data reflecting the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL 3, IL8, IRF1 and STAT3; b) comparing, at the at least one processor, said expression levels to a reference expression level of the group of genes from control samples from a population; c) outputting, at the at least one processor, an immune score; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to the immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with a lesser benefit of anthracycline therapy.

[0060] As used herein, "processor" may be any type of processor, such as, for example, any type of general-purpose microprocessor or microcontroller (e.g., an Intel™ x86, PowerPC™, ARM™ processor, or the like), a digital signal processing (DSP) processor, an integrated circuit, a field programmable gate array (FPGA), or any combination thereof. [0061] As used herein "memory" may include a suitable combination of any type of computer memory that is located either internally or externally such as, for example, random-access memory (RAM), read-only memory (ROM), compact disc read-only memory (CDROM), electro-optical memory, magneto-optical memory, erasable programmable read- only memory (EPROM), and electrically-erasable programmable read-only memory (EEPROM), or the like. Portions of memory 102 may be organized using a conventional filesystem, controlled and administered by an operating system governing overall operation of a device.

[0062] As used herein, "computer readable storage medium" (also referred to as a machine-readable medium, a processor-readable medium, or a computer usable medium having a computer-readable program code embodied therein) is a medium capable of storing data in a format readable by a computer or machine. The machine-readable medium can be any suitable tangible, non-transitory medium, including magnetic, optical, or electrical storage medium including a diskette, compact disk read only memory (CD-ROM), memory device (volatile or non-volatile), or similar storage mechanism. The computer readable storage medium can contain various sets of instructions, code sequences, configuration information, or other data, which, when executed, cause a processor to perform steps in a method according to an embodiment of the disclosure. Those of ordinary skill in the art will appreciate that other instructions and operations necessary to implement the described implementations can also be stored on the computer readable storage medium. The instructions stored on the computer readable storage medium can be executed by a processor or other suitable processing device, and can interface with circuitry to perform the described tasks.

[0063] As used herein, "data structure" a particular way of organizing data in a computer so that it can be used efficiently. Data structures can implement one or more particular abstract data types (ADT), which specify the operations that can be performed on a data structure and the computational complexity of those operations. In comparison, a data structure is a concrete implementation of the specification provided by an ADT.

[0064] In some embodiments, the method further comprises building, at the at least one processor, a subject gene expression profile from the determined expression levels of the group of genes. [0065] In some embodiments, the immune score comprises the weighted sum expression of the group of genes.

[0066] In some embodiments, the subject is classified into a high immune score group where the subject has an immune score above the population median, and wherein the subject is classified into a low immune score group where the subject has an immune score below the population median.

[0067] In some embodiments, in the low immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, or IRF1 was relatively low and STAT3 was relatively high, and in the high immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL 3, IL8, IRF1 was relatively high and STAT3 was relatively low.

[0068] In some embodiments, the processor further outputs a recommendation to treat the subject with anthracycline if the subject has a relatively low immune score or is in the low immune score group. [0069] In an aspect, there is provided a computer program product for use in conjunction with a general-purpose computer having a processor and a memory connected to the processor, the computer program product comprising a computer readable storage medium having a computer mechanism encoded thereon, wherein the computer program mechanism may be loaded into the memory of the computer and cause the computer to carry out the method described herein.

[0070] In an aspect, there is provided a computer readable medium having stored thereon a data structure for storing the computer program product described herein.

[0071] In an aspect, there is provided a device for predicting benefit of anthracycline therapy for a subject with breast cancer, the device comprising: at least one processor; and electronic memory in communication with the at one processor, the electronic memory storing processor-executable code that, when executed at the at least one processor, causes the at least one processor to: a) receive data reflecting the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; b) compare said expression levels to a reference expression level of the group of genes from control samples from a population, the reference expression levels being stored in the memory; c) output an immune score; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to the immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with a lesser benefit of anthracycline therapy.

[0072] In some embodiments, the code further causes the at least one processor to build a subject gene expression profile from the determined expression levels of the group of genes.

[0073] In some embodiments, the code further causes the at least one processor to classify the subject, wherein the subject is classified into a high immune score group where the subject has an immune score above the population median, and wherein the subject is classified into a low immune score group where the subject has an immune score below the population median.

Treatment Selection and Compositions

[0074] In an aspect, there is provided a method of treating a subject with breast cancer, comprising: a) determining the immune score of the subject according to the method described herein; and b) selecting a treatment based on said immune score, and preferably treating the subject according to the treatment.

[0075] In some embodiments, anthracycline therapy is selected as the treatment where the immune score is relatively low.

[0076] In some embodiments, the selected treatment further comprises a checkpoint inhibitor therapy. [0077] As used herein, "checkpoint Inhibitor" and "checkpoint inhibitor therapy" refers to a type of drug that blocks certain proteins made by some types of immune system cells, such as T cells, and some cancer cells. These proteins help keep immune responses in check and can keep T cells from killing cancer cells. When these proteins are blocked, the "brakes" on the immune system are released and T cells are able to kill cancer cells better. Examples of checkpoint proteins found on T cells or cancer cells include PD-1/PD-L1 and CTLA-4/B7-1/B7-2. Some immune checkpoint inhibitors are used to treat cancer.

[0078] In an aspect, there is provided a composition comprising a plurality of isolated nucleic acid sequences, wherein each isolated nucleic acid sequence hybridizes to: (a) the mRNA of a group of genes comprising at least one of. GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; and/or (b) a nucleic acid complementary to a); wherein the composition is used to measure the expression levels of the group of genes.

[0079] In an aspect, there is provided an array comprising one or more polynucleotide probes complementary and/or hybridizable to an expression product of each gene of a group genes comprising at least three of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3.

[0080] In an aspect, there is provided a kit comprising reagents for detecting mRNA from a sample of a breast cancer tumour of at least three genes selected from the group comprising: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL 3, IL8, IRF1 and STAT3. [0081] Examples of primers include an oligonucleotide which is capable of acting as a point of initiation of polynucleotide synthesis along a complementary strand when placed under conditions in which synthesis of a primer extension product which is complementary to a polynucleotide is catalyzed. Such conditions include the presence of four different nucleotide triphosphates or nucleoside analogs and one or more agents for polymerization such as DNA polymerase and/or reverse transcriptase, in an appropriate buffer ("buffer" includes substituents which are cofactors, or which affect pH, ionic strength, etc.), and at a suitable temperature. A primer must be sufficiently long to prime the synthesis of extension products in the presence of an agent for polymerase. A typical primer contains at least about 5 nucleotides in length of a sequence substantially complementary to the target sequence, but somewhat longer primers are preferred. A primer will always contain a sequence substantially complementary to the target sequence, that is the specific sequence to be amplified, to which it can anneal.

[0082] The terms "complementary" or "complement thereof, as used herein, refer to sequences of polynucleotides which are capable of forming Watson & Crick base pairing with another specified polynucleotide throughout the entirety of the complementary region. This term is applied to pairs of polynucleotides based solely upon their sequences and does not refer to any specific conditions under which the two polynucleotides would actually bind

[0083] The term "probe" refers to a molecule which can detectably distinguish between target molecules differing in structure, such as allelic variants. Detection can be accomplished in a variety of different ways but preferably is based on detection of specific binding. Examples of such specific binding include antibody binding and nucleic acid probe hybridization.

[0084] The term "hybridize" or "hybridizable" refers to the sequence specific non- covalent binding interaction with a complementary nucleic acid. In a preferred embodiment, the hybridization is under high stringency conditions. Appropriate stringency conditions which promote hybridization are known to those skilled in the art, or can be found in Current Protocols in Molecular Biology, John Wiley & Sons, N.Y. (1989), 6.3.1 6.3.6. For example, 6.0 x sodium chloride/sodium citrate (SSC) at about 45°C, followed by a wash of 2.0 x SSC at 50°C may be employed. [0085] The polynucleotide compositions can be primers, can be cDNA, can be RNA, can be DNA complementary to target cDNA or a portion thereof, genomic DNA, unspliced RNA, spliced RNA, alternately spliced RNA, synthetic forms, and mixed polymers, both sense and antisense strands, and may be chemically or biochemically modified or may contain non-natural or derivatized nucleotide bases, as will be readily appreciated by those skilled in the art.

[0086] Where nucleic acid includes RNA, reference to the sequence shown should be construed as reference to the RNA equivalent, with U substituted for T.

[0087] The methods of nucleic acid isolation, amplification and analysis are routine for one skilled in the art and examples of protocols can be found, for example, in the Molecular Cloning: A Laboratory Manual (3-Volume Set) Ed. Joseph Sambrook, David W. Russel, and Joe Sambrook, Cold Spring Harbor Laboratory; 3rd edition (Jan. 15, 2001 ), ISBN: 0879695773. Particularly useful protocol source for methods used in PCR amplification is PCR (Basics: From Background to Bench) by M. J. McPherson, S. G. Moller, R. Beynon, C. Howe, Springer Verlag; 1 st edition (Oct. 15, 2000), ISBN: 0387916008. [0088] Examples of amplification techniques include strand displacement amplification, as disclosed in U.S. Pat. No. 5,744,311 ; transcription-free isothermal amplification, as disclosed in U.S. Pat. No. 6,033,881 ; repair chain reaction amplification, as disclosed in WO 90/01069; ligase chain reaction amplification, as disclosed in European Patent Appl. 320 308; gap filling ligase chain reaction amplification, as disclosed in U.S. Pat. No. 5,427,930; and RNA transcription-free amplification, as disclosed in U.S. Pat. No. 6,025, 134.

[0089] "Kit" refers to a combination of physical elements, e.g., probes, including without limitation specific primers, labeled nucleic acid probes, antibodies, protein-capture agent(s), reagent(s), instruction sheet(s) and other elements useful to practice the invention, in particular to identify the levels of particular RNA molecules in a sample. These physical elements can be arranged in any way suitable for carrying out the invention. For example, probes and/or primers can be provided in one or more containers or in an array or microarray device.

[0090] In one embodiment, levels of RNA encoded by a target gene can be determined in one analysis. A combination kit may therefore include primers capable of amplifying cDNA derived from RNA encoded by different target genes. The primers may be differentially labeled, for example using different fluorescent labels, so as to differentiate between RNA from different target genes.

[0091] Multiplex, such as duplex, real-time RT-PCR enables simultaneous quantification of 2 targets in the same reaction, which saves time, reduces costs, and conserves samples. These advantages of multiplex, real-time RT-PCR make the technique well-suited for high-throughput gene expression analysis. Multiplex qPCR assay in a realtime format facilitates quantitative measurements and minimizes the risk of false-negative results. It is essential that multiplex PCR is optimized so that amplicons of all samples are compared insub-plateau phase of PCR. Yun, Z., I. Lewensohn-Fuchs, P. Ljungman, L. Ringholm, J. Jonsson, and J. Albert. 2003. A real-time TaqMan PCR for routine quantitation of cytomegalovirus DNA in crude leukocyte lysates from stem cell transplant patients. J. Virol. Methods 1 10:73-79. [PubMed]. Yun, Z., I. Lewensohn-Fuchs, P. Ljungman, and A. Vahlne. 2000. Real-time monitoring of cytomegalovirus infections after stem cell transplantation using the TaqMan polymerase chain reaction assays. Transplantation 69: 1733-1736. [PubMed]. Simultaneous quantification of up to 2, 3, 4, 5, 6, 7, and 8 or more targets may be useful.

[0092] For example, the primers and probes contained within the kit may include those able to recognize any of genes of the gene signature described herein. [0093] A primer which "selectively hybridizes" to a target polynucleotide is a primer which is capable of hybridizing only, or mostly, with a single target polynucleotide in a mixture of polynucleotides consisting of RNA in a sample, or consisting of cDNA complementary to RNA within the sample.

[0094] A gene expression profile for breast cancer found in a sample at the RNA level of one or more genes comprising, but preferably not limited to, any of the genes described herein, can be identified or confirmed using many techniques, including but preferably not limited to PCR methods, as for example discussed further in the working examples herein, Northern analyses and the microarray technique, NanoString® and quantitative sequencing. This gene expression profile can be measured in a sample, using various techniques including e.g. microarray technology. In an embodiment of this method, fluorescently labeled cDNA probes may be generated through incorporation of fluorescent nucleotides by reverse transcription of RNA extracted from a sample. Labeled cDNA probes applied to the chip hybridize with specificity to each spot of DNA on the array. Quantitation of hybridization of each arrayed element allows for assessment of corresponding mRNA abundance. For example, with dual color fluorescence, separately labeled cDNA probes generated from two sources of RNA are hybridized pair wise to the array. The relative abundance of the transcripts from the two sources corresponding to each specified gene is thus determined simultaneously. Such methods have been shown to have the sensitivity required to detect rare transcripts, which are expressed at a few copies per cell, and to reproducibly detect at least approximately two-fold differences in the expression levels (Schena et al., Proc. Natl. Acad. Sci. USA 93(2): 106-149 (1996)). Microarray analysis can be performed by commercially available equipment, following manufacturer's protocols, such as by using the Affymetrix GenChip technology, or Incyte's microarray technology.

[0095] The above listed aspects and/or embodiments may be combined in various combinations as appreciated by a person of skill in the art. The advantages of the present disclosure are further illustrated by the following examples. The examples and their particular details set forth herein are presented for illustration only and should not be construed as a limitation on the claims of the present invention.

EXAMPLES MATERIALS AND METHODS

Clinical trials and participants

[0096] The BR9601 trial (clinicaltrials.gov NCT00003012) recruited 374 pre- and postmenopausal women with completely excised, histologically confirmed breast cancer and a clear indication for adjuvant chemotherapy. Patients were randomized to receive either eight cycles of CMF (i.v. cyclophosphamide 750mg/m², methotrexate 50mg/m² and 5-fluorouracil 600mg/m²) every 21 days, or E-CMF (four cycles of epirubicin 100mg/m² every 21 days followed by four cycles of the above CMF regimen)¹⁷ (see FIG. 1 ). The primary outcomes of the BR9601 study were RFS and breast cancer-specific overall survival (OS), but distant relapse-free survival (DRFS) was also reported.¹⁷ [0097] The MA5 trial (National Cancer Institute of Canada Clinical Trials Group) included 710 premenopausal or perimenopausal women with axillary node-positive breast cancer who had undergone surgery. Patients were randomized to receive CMF (cyclophosphamide 100 mg/m² on days 1-14, methotrexate 40 mg/m² and fluorouracil 600 mg/m² on days 1-8) or CEF (cyclophosphamide 75 mg/m² on days 1 -14, epirubicin 60 mg/m² and fluorouracil 500 mg/m² on days 1-8).

[0098] Both protocols were approved by central and local ethics committees, where each patient provided written informed consent prior to randomization. For the current analysis, tissue blocks were retrieved and RNA was extracted. Two endpoints used in the study were DRFS (BR9601 n=99; MA5 n=X) and OS (BR9601 n=101 ; MA5 n=X). RNA extraction and NanoString gene expression analysis

[0099] Total RNA from formalin fixed paraffin embedded (FFPE) breast cancer tissue samples (2 x 10μΜ full sections) was extracted using the RecoverAII Total Nucleic Acid Isolation kit (Life Technologies, Burlington, Canada) following the manufacturer's protocol. Sample concentrations were determined with the NanoDrop ND-1000 spectrophotometer (ThermoScientific, Wilmington, USA). nCounter gene expression codeset were designed to include a panel of 38 immune-function genes and six housekeeping genes (see Table 1). The codesets were processed on nCounter according to manufacturer's instructions (NanoString Technologies, Seattle, USA).

Table 1 : All 38 immune-function genes and 6 housekeeping genes included on a

Nanostring codeset

Immune Housekeeping

CD3 epsilon GUSB

CD8 alpha PUM1

FOXP3 SF3A1

CD4 TBP

CD20 TRFC

CD68 TMED10

CD45RO

CD19

CD48

GZMB

LCK

PRF1

SELL

CCR1

CCR3

CCR5

CCR4

CCL5

CCL22

CXCL9

CXCL10

CX3CL1

CXCR3

CXCL13

IL8

CXCR2

CXCR4

CXCL12

Tbet

IFNgamma

IL-12

STAT1

IRF1

MADCAM1

VCAM1

ICAM1

JA 2

STAT3 NanoString Normalization

[00100] Raw mRNA abundance count data were pre-processed using the NanoStringNorm R package (v1.1.19)¹⁸ using normalization factors derived from the geometric mean of all genes, followed by quantile normalization (FIG. 5). No samples were discarded following QAQC. Six samples were run in duplicates, and their raw counts were averaged and subsequently treated as a single sample prior to normalization. Optimization of pre-processing methods is described in Supplementary Methods.

Derivation of 9-Gene Signature

[00101] In order to evaluate the performance of the current marker, a stratified five-fold cross validation approach was used. For each fold, stepwise backward feature selection was performed on these 38 genes using the Akaike Information Criterion (AlC) represented by equation (1 ) below:

AIC= 2K- 2 \n {L) _{(1 )} where k is the number of parameters and L is the maximum value of the likelihood function. The selected genes were then used to derive a multifeature signature score. For each gene, the patients were dichotomized into low/high expression groups based on the median expression values of the training patients within that fold. Each patient is then given +1 or -1 depending on whether they fall into the low or high expression group, respectively. Each gene iterated through, assigning +1 or -1 for the patients. Finally, the sum of scores for all genes was calculated using equation (2) below:

[00102] For each gene g, an indicator function / was run to determine whether the expression level of that sample e_p is above or below the median gene expression m₉. Scores are summed over all genes to calculate the 9-gene signature score. Statistical analyses [00103] Statistical analyses were performed using SPSS (v.23) software (IBM). Patients were first sorted based on the number of events for DRFS and OS, then dichotomized based on the population median into low or high expression group. Survival analyses were based on the Cox regression model and used to calculate hazard ratios (HR) and significance (p<0.1 ); HRs (with 95% CI) lower than 1 suggested an advantage of E- CMF treatment over CMF alone. Survival data was available for 10 years and visualized with the Kaplan-Meier plots. For multivariate survival analyses, age was used as binary variable (dichotomized at age >50), while nodal status, pathological grade, ER status and HER2 status were used as ordinal variables; tumor size was treated as a continuous variable. Treatment by marker interaction analysis was used to obtain hazard ratios and significance (p<0.1) between biomarker high and biomarker low patient groups, which were stratified based on the treatment type: E-CMF versus CMF. Spearman correlation test was used to obtain the strength (coefficient) and significance (p < 0.05) of the relationship between two genes, as well as the immune signature score and ER status. mRNA Abundance Data Processing

[00104] In order to identify an optimal normalization method in this current cohort, a combination of 252 preprocessing methods were evaluated. Six positive controls, 8 negative controls and 6 housekeeping genes (TRFC, TBP, GUSB, TMED10, SF3A1 , and PUM1 ) were used. Two criteria were used to determine the most optimal preprocessing parameters for this study as previously described [Sabine et a/., in preparation]. First, each of the 252 combinations of preprocessing methods was ranked based on their ability to maximize Euclidean distance of ERBB2 mRNA abundance between HER2-positive and HER2- negative samples. The process was repeated for 1 million random subsets of HER2-positive and HER2-negative samples for each of the preprocessing schemes. Preprocessing schemes were subsequently ranked based on inter-batch variation as measured by five replicates of a cell line control sample. A mixed effects linear model was used and residual estimates were extracted as an estimate of inter-batch variation (nlme v3.1 -1 17). Cumulative ranks based on these two criteria were estimated using RankProduct [19] resulting in selection of an optimal pre-processing scheme of normalization to the geometric mean derived from all genes for sample content followed by quantile normalisation.

RESULTS [00105] Patient and tumour characteristics from the BR9601 trial are shown in Table 2, whereas the modified REMARK diagram of the trial scheme is shown in FIG. 1. Tumour samples for Nanostring analysis were available for 144 (79%) of the 183 patients receiving E-CMF and 146 (76%) of the 191 patients receiving CMF, suggesting no bias by group in sample availability. The remaining samples were not included in the analysis due to block exhaustion or technical issues.

[00106] Gene-expression heatmap and associated clinical characteristics on a per- patient basis are included in FIG. 6. At the individual gene level, three genes were found to be prognostic: high expression of CXCL13 (HR: 0.485, 95%CI 0.32-0.735, p=0.001) and PRF1 (HR: 0.575, 95%CI 0.384-0.863, p=0.007), and low expression of IL8 (HR: 1.787, 95%CI 1.195-2.67, p=0.005) (see Table 3). The overall prediction values for each gene, based on its expression level, are shown in FIG. 2. In particular, six genes were found to be predictive of epirubicin benefit when their gene expression was low: FOXP3 (HR: 0.34, 95%CI 0.782-0.146, p=0.011 ), CD4 (HR: 0.35, 95%CI 0.792-0.153, p=0.012), IFNy (HR: 0.48, 95%CI 1.092-0.213, p=0.080), GZMB (HR: 0.49, 95%CI 1 .1 14-0.216, p=0.089), CD3e (HR: 0.49, 95%CI 1.107-0.220, 0.087) and CD45RO (HR: 0.50, 95%CI 1.125-0.223, p=0.094); in contrast, it was the high expression of PRF1 (HR: 2.14, 95%CI 4.950-0.927, p=0.075) and CX3CL1 (HR: 2.51 , 95%CI 5.682-1.1 10, p=0.027) that was predictive of E- CMF benefit (see Table 4).

Table 2: Patient and tumour characteristics: BR9601 trial

Distant relapse- free survival no

1 14 (53.5) 99 (46.5) 213 relapse

relapse 44 (40.7) 64 (59.3) 108

Table 3: Individual gene prognostic values

HR 95% CI lower 95% CI upper p-value

CXCL13 0.485 0.32 0.735 0.001

IL8 1.787 1 .195 2.67 0.005

PRF1 0.575 0.384 0.863 0.007

JAK2 0.672 0.451 1.001 0.051

CCR5 0.685 0.46 1 .021 0.063

CD8 alpha 0.707 0.475 1 .054 0.089

CCR4 1.387 0.933 2.063 0.106

STAT1 0.737 0.495 1.096 0.132

STAT3 0.755 0.507 1.126 0.168

IFNgamma 0.761 0.51 1 1.132 0.178

Tbet 0.767 0.516 1.141 0.190

CCL5 0.774 0.521 1.15 0.205

MS4A1 0.8 0.538 1 .189 0.270

CD19 0.803 0.54 1 .196 0.280 CCR1 0.81 1 0.546 1.205 0.300

LCK 0.824 0.554 1 .223 0.336

CD3 epsilon 0.833 0.561 1.237 0.365

CCL22 1 .171 0.789 1 .737 0.433

IL-12 1 .154 0.778 1.71 1 0.477

CCR3 1 .135 0.765 1.683 0.530

CXCL9 0.886 0.597 1.316 0.549

CD45RO 0.891 0.601 1.322 0.567

CXCR3 0.892 0.601 1.324 0.571

MADCAM1 1.121 0.755 1.663 0.572

CD68 0.916 0.617 1.36 0.665

CD48 0.918 0.619 1.362 0.671

GZMB 0.925 0.624 1.372 0.698

VCAM1 0.934 0.63 1 .386 0.735

IRF1 1 .06 0.715 1.571 0.773

CXCR2 0.947 0.639 1.405 0.788

FOXP3 0.948 0.638 1.407 0.790

ICAM1 1.05 0.708 1.557 0.810

CXCL12 0.958 0.646 1.421 0.831 CXCL10 1.036 0.699 1 .537 0.859

CX3CL1 1.035 0.698 1.535 0.865

SELL 0.977 0.658 1.449 0.906

CD4 0.983 0.663 1.458 0.932

CXCR4 0.998 0.673 1 .48 0.992

Table 4: Individual gene predictive hazard ratios from treatment by marker interaction analysis

HR 95% CI lower 95% CI upper p-value

FOXP3 0~34 0.782 0.146 0.01 1

CD4 0.35 0.792 0.153 0.012

CX3CL1 2.51 5.682 1.1 10 0.027

PRF1 2.14 4.950 0.927 0.075

IFNgamma 0.48 1.092 0.213 0.080

CD3 epsilon 0.49 1 .107 0.220 0.087

GZMB 0.49 1.1 14 0.216 0.089

CD45RO 0.50 1 .125 0.223 0.094

Tbet 0.51 1.174 0.220 0.1 13

CXCL9 0.53 1.203 0.234 0.129

MADCAM1 1.74 3.876 0.781 0.176 CD48 0.57 1.285 0.255 0.176

CXCR4 1.73 3.861 0.778 0.179

IRF1 0.59 1 .325 0.263 0.201

CD8 alpha 0.60 1.348 0.267 0.215

IL8 1 .63 3.717 0.715 0.245

VCAM1 0.65 1.451 0.289 0.292

CXCR2 1.50 3.322 0.675 0.320

CCR4 1.47 3.289 0.661 0.342

CXCL12 0.69 1.524 0.308 0.355

LCK 0.71 1.582 0.318 0.401

CXCR3 0.72 1.605 0.324 0.424

JAK2 0.73 1.631 0.325 0.441

STAT3 1.36 3.067 0.604 0.458

CCL5 0.77 1.706 0.344 0.514

CCR3 1 .24 2.762 0.559 0.593

CXCL10 0.82 1.835 0.363 0.623

ICAM1 0.83 1.835 0.372 0.641

CXCL13 0.83 1 .919 0.359 0.662

SELL 0.84 1.873 0.379 0.674

CD19 1 .18 2.653 0.530 0.679 CCL22 1.17 2.61 1 0.528 0.695

CCR1 1.16 2.591 0.521 0.715

STAT1 1.12 2.506 0.501 0.782

IL-12 0.90 1 .992 0.404 0.790

CCR5 0.91 2.049 0.408 0.828

CD68 0.96 2.141 0.432 0.924

MS4A1 1.04 2.320 0.461 0.934

[00107] In order to determine the minimal number of genes predictive of epirubicin benefit, the BR9601 trial was used as a training cohort for signature development. The resulting multi-gene signature included the following 9 genes: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3. A 9-immune gene signature score was calculated and each patient was sorted into either low immune-signature score group (below the population median) or high immune-signature score group (above the population median). FIG. 3 shows the directionality of gene expression for each gene in the signature. Briefly, in the low score group, the expression of all genes except STAT3 was low; similarly, in the high immune-signature score group, the expression of all genes, except STAT3, was high. Previous reports suggested that high percentages of TILs correlate with increased survival only in TNBC patients treated with E-CMF ¹⁸. It was observed that approximately 67% of patients in this group had ER negative breast cancer (see Table 5).

Table 5: Correlation between immune signature score and ER status

[00108] Since the immune signature included genes that are involved in cell killing and trafficking, as well as chemoattractants responsible for recruiting immune cells to the homing site, it was assessed whether these genes correlate with CD4 (Th phenotype), CD8 (CTL phenotype) and CD3 (total T-cell phenotype) gene expression (see Table 6) GZMB and PRF1 , which encode for lymphocyte effector molecules involved in cellular killing, correlated with CD8 and CD3 gene expression but not CD4. Similarly, IRF1 , a transcription factor involved in interferon response signaling correlated with CD8 and CD3 expression, but not CD4. Lastly SELL, a gene encoding a homing receptor responsible for lymphocyte trafficking, correlated with both T-cell co-receptors as well as with CD3. None of the chemokine (IL8, CDXCL10, CXCL13, CCL22) or receptor signaling (STAT3) genes correlated with T-cell markers, suggesting that they might be expressed by tumour cells or other stromal cells in the tumour microenvironment (see Table 6).

Table 6: Correlation of immune signature genes with T cell markers

CD8 CD4 CD3

GZMB 0.400** 0.168 0.458**

PRF1 0.755** 0.184** 0.668**

IRFl 0.430** 0.276** 0.387**

SELL 0.571 0.391** 0.717**

STAT3 -0.071** - , . 0.077 -0.143*

CXCL13 0.161** 0.108 0.287**

CCL22 ^■'^{' ■}0.006 ^' 0.071** -0.011

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

*. Correlation is significant at the 0.05 level (2- tailed).

CD3:CD4 correlation = 0.392** CD3:CD8 correlation = 0.725**

[00109] Next, the differential effects of the immune signature on DRFS and breast cancer-specific overall survival (OS) were analyzed between patients in the BR9601 trial receiving an anthracycline (E-CMF) and those given CMF alone by assessing hazard ratios and treatment by marker interactions. Patients whose tumours had low immune signature score had an increased DRFS (HR: 0.428, 95%CI 0.234-0.782) and OS (HR: 0.391 , 95%CI 0.206-0.741 ) when treated with E-CMF compared with patients treated with CMF alone (FIG. 4). Conversely, there was no differential benefit of E-CMF vs CMF in patients with high immune signature expression (FIG. 4) for DRFS (HR 1 .078, 95%CI 0.609-1.907) and OS (HR: 1.034, 95%CI 0.588-1.816). In a multivariate analysis, after adjustment for HER2 status, nodal status, age, grade, size and ER status, the treatment by marker interaction showed no statistical difference for DRFS (HR: 0.561 , 95%CI 0.212-1.483, p=0.244) and OS (HR: 0.442, 95%CI 0.165-1.182, p=0.104) (see Table 7).

Table 7: Univariate and multivariate analyses for DRFS and OS

95.0% CI

HR Lower Upper p-value

Univariate DRFS

Treatment 1.076 0.608 1.903 0.801

Immune signature 1.645 0.96 2.819 0.070

Treatment by Immune signature 0.395 0.172 0.907 0.028

Univariate OS

Treatment 1.060 0.605 1.854 0.839

Immune signature 1.411 0.825 2.416 0.209

Treatment by Immune signature 0.371 0.158 0.868 0.022

Multivariate DRFS

Age 0.963 0.592 1.565 0.879

Size 0.966 0.543 1.718 0.907

Grade 1.514 0.937 2.445 0.090

Node 2.569 1.738 3.797 <0.001

ER 0.656 0.386 1.116 0.120

HER2 0.701 0.415 1.184 0.184

Treatment 0.717 0.365 1.409 0.335

Immune signature 1.378 0.722 2.631 0.330

Treatment by immune signature 0.561 0.212 1.483 0.244

Multivariate OS

Age 0.832 0.513 1.351 0.457

Size 0.841 0.46 1.536 0.572

Grade 1.302 0.796 2.128 0.293

Node 2.373 1.616 3.485 <0.001

ER 0.495 0.289 0.850 0.011

HER2 0.627 0.372 1.056 0.079

Treatment 0.824 0.425 1.600 0.568

Immune signature 1.418 0.739 2.718 0.293

Treatment by immune signature 0.442 0.165 1.182 0.104

[00110] The prognostic role of TILs and immune genes in breast cancer has already been established, but not their predictive significance. In this study, a 9-immune gene signature as a biomarker of anthracycline sensitivity was identified. The study's approach involved examining 300 breast cancer samples from BR9601 randomized adjuvant clinical trial to derive a predictive gene signature. This signature is unique in two ways. First, an immune biomarker that predicts for therapy benefit in all breast cancer patients has been identified. Previously, total TILs ¹² and CD8⁺ T cells ³ were found to be predictive of chemotherapy benefit but only in TNBC. In addition, Perez ef al have identified an immune- enriched gene signature that predicts trastuzumab benefit in patients with HER2+ breast cancer.¹⁵ This study is the first to report an immune biomarker, whether it is cell- or gene- based, as a predictive biomarker of adjuvant therapy in both ER⁺ as well as ER^" breast cancers. Second, the immune biomarker contained genes that correlate to cytotoxic T lymphocytes (GZMB, PRF1 , IRF1 and SELL), as well as STAT3 and chemokines (IL8, CXCL10, CXCL13 and CCL22) that are likely to be expressed by other stromal or tumour cells. Therefore, the signature encompasses immune features from the entire tumour microenvironment, reflecting that immune, tumour and stromal cells engage in a complex interplay during development of drug resistance.

[00111] Previously it was reported that only TNBC patients benefit from chemotherapy, possibly since poorly differentiated and highly heterogeneous TNBC cells are expected to generate a variety of tumour-associated antigens and elicit a stronger immune response than other breast-cancer subtypes. In this study however, all patients survived well when immune biomarker was high, regardless of the chemotherapy type that they had received. This might be attributed to one known mechanisms of chemotherapy, namely lymphopenia that is followed by homeostatic expansion of lymphocytes. Specifically, anthracyclines and cyclophosphamide can cause transient depletion of lymphocytes including immunosuppressive T regulatory cells and exhausted T cells in the tumour site;^{19" 2} a subsequent homeostatic expansion and recruitment of tumour-antigen specific T cells would lead to a more effective antitumour immune response following chemotherapy treatment.²¹

[00112] Interestingly, it was the low immune biomarker score that predicted benefit from anthracyclines. It is important to emphasize here that "low" score refers to those patients that are below the population median, therefore not entirely devoid of immune features associated with the signature. It is presumed at least some of those patients had low immune infiltration. Priming the immune system with anthracyclines might have contributed to the immunogenic cell death and better overall outcome. Previously, doxorubicin treated MCF7 cells have been shown to upregulate Fas expression, which induces Fas-mediated apoptosis.²² Furthermore, anthracyclines specifically have been shown to induce translocation of calreticulin from ER to the cell membrane as well as a release of an endogenous ligand HMGB1 (high mobility group box 1 ), both of which act as danger signals eliciting an immune response.^23"25

[00113] On the other hand, patients in the low immune biomarker group that did not benefit from anthracycline treatment may benefit from the checkpoint inhibitors such as anti- CTLA4, anti-PD-1 anti-PD-L1. CTLA4, an inhibitory molecule expressed by T regulatory and activated T cells, competes with CD28 for interaction with the co-stimulatory ligands CD86/80 that are necessary for T-cell activation. PD1 is expressed on activated lymphocytes as well as exhausted lymphocytes ²⁶ and functions by binding to antigen- presenting cells and tumour cells, thereby reducing T-cell activation.²⁷ Therefore, the mechanism of action of these drugs would involve multiple avenues, ranging from blocking inhibitory molecules on tumour cells and T regulatory cells, to directly activating cytotoxic T cells.²⁷

[00114] Simon et al. describes an example procedure for validation of gene signature studies using archived samples.²⁸

[00115] All documents disclosed herein, including those in the following reference list, are incorporated by reference. Although preferred embodiments of the invention have been described herein, it will be understood by those skilled in the art that that the detailed description and the specific examples while indicating preferred embodiments of the invention are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.

REFERENCES

(1 ) Guo B, Villeneuve DJ, Hembruff SL et al. Cross-resistance studies of isogenic drug- resistant breast tumor cell lines support recent clinical evidence suggesting that sensitivity to paclitaxel may be strongly compromised by prior doxorubicin exposure. Breast Cancer Res Treat 2004;85:31-51.

(2) Munro AF, Twelves C, Thomas JS, Cameron DA, Bartlett JM. Chromosome instability and benefit from adjuvant anthracyclines in breast cancer. Br J Cancer 2012; 107:71- 74.

(3) Pritchard Kl, Munro A, O'Malley FP et al. Chromosome 17 centromere (CEP17) duplication as a predictor of anthracycline response: evidence from the NCIC Clinical Trials Group (NCIC CTG) MA.5 Trial. Breast Cancer Res Treat 2012; 131 :541-551.

(4) Bartlett JM, McConkey CC, Munro AF et al. Predicting Anthracycline Benefit: TOP2A and CEP17-Not Only but Also. J Clin Oncol 2015;33: 1680-1687.

(5) Spears M, Yousif F, Lyttle N et al. A four gene signature predicts benefit from anthracyclines: evidence from the BR9601 and MA.5 clinical trials. Oncotarget 2015;6:31693-31701.

(6) Salgado R, Denkert C, Demaria S et al. The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an International TILs Working Group 2014. Ann Oncol 2015;26:259-271.

(7) Liu S, Lachapelle J, Leung S, Gao D, Foulkes WD, Nielsen TO. CD8+ lymphocyte infiltration is an independent favorable prognostic indicator in basal-like breast cancer. Breast Cancer Res 2012; 1 :R48.

(8) Ruffell B, Au A, Rugo HS, Esserman LJ, Hwang ES, Coussens LM. Leukocyte composition of human breast cancer. Proc Natl Acad Sci U S A 2012; 109:2796-2801.

(9) Loi S, Sirtaine N, Piette F et al. Prognostic and predictive value of tumor-infiltrating lymphocytes in a phase III randomized adjuvant breast cancer trial in node-positive breast cancer comparing the addition of docetaxel to doxorubicin with doxorubicin- based chemotherapy: BIG 02-98. J Clin Oncol 2013;31 :860-867. (10) Adams S, Gray RJ, Demaria S et al. Prognostic value of tumor-infiltrating lymphocytes in triple-negative breast cancers from two phase III randomized adjuvant breast cancer trials: ECOG 2197 and ECOG 1 199. J Clin Oncol 2014;32:2959-2966.

(1 1 ) West NR, Milne K, Truong PT, Macpherson N, Nelson BH, Watson PH. Tumor- infiltrating lymphocytes predict response to anthracycline-based chemotherapy in estrogen receptor-negative breast cancer. Breast Cancer Res 20 1 ;13:R126.

(12) Denkert C, Loibl S, Noske A et al. Tumor-associated lymphocytes as an independent predictor of response to neoadjuvant chemotherapy in breast cancer. J Clin Oncol 2010;28: 105-1 13.

(13) AN HR, Provenzano E, Dawson SJ et al. Association between CD8+ T-cell infiltration and breast cancer survival in 12,439 patients. Ann Oncol 2014;25: 1536-1543.

(14) Loi S, Michiels S, Salgado R et al. Tumor infiltrating lymphocytes are prognostic in triple negative breast cancer and predictive for trastuzumab benefit in early breast cancer: results from the FinHER trial. Ann Oncol 2014;25: 1544-1550.

(15) Perez EA, Thompson EA, Ballman KV et al. Genomic analysis reveals that immune function genes are strongly linked to clinical outcome in the North Central Cancer Treatment Group n9831 Adjuvant Trastuzumab Trial. J Clin Oncol 2015;33:701 -708.

(16) Perez EA, Ballman KV, Tenner KS et al. Association of Stromal Tumor-Infiltrating Lymphocytes With Recurrence-Free Survival in the N9831 Adjuvant Trial in Patients With Early-Stage HER2-Positive Breast Cancer. JAMA Oncol 2015; 1 -9.

(17) Poole CJ, Earl HM, Hiller L et al. Epirubicin and cyclophosphamide, methotrexate, and fluorouracil as adjuvant therapy for early breast cancer. N Engl J Med 2006;355: 1851-1862.

(18) AN AM, Provenzano E, Bartlett JM et al. Prognosis of early breast cancer by immunohistochemistry defined intrinsic sub-types in patients treated with adjuvant chemotherapy in the NEAT/BR9601 trial. Int J Cancer 20 3; 133: 1470-1478. (19) Zitvogel L, Galluzzi L, Smyth MJ, Kroemer G. Mechanism of action of conventional and targeted anticancer therapies: reinstating immunosurveillance. Immunity 2013;39:74-88.

(20) Ghiringhelli F, Larmonier N, Schmitt E et al. CD4+CD25+ regulatory T cells suppress tumor immunity but are sensitive to cyclophosphamide which allows immunotherapy of established tumors to be curative. Eur J Immunol 2004;34:336-344.

(21 ) Ercolini AM, Ladle BH, Manning EA et al. Recruitment of latent pools of high-avidity CD8(+) T cells to the antitumor immune response. J Exp Med 2005;201 :1591-1602.

(22) Crescenzi E, Pacifico F, Lavorgna A et al. NF-kappaB-dependent cytokine secretion controls Fas expression on chemotherapy-induced premature senescent tumor cells. Oncogene 201 ;30:2707-2717.

(23) Obeid M, Tesniere A, Ghiringhelli F et al. Calreticulin exposure dictates the immunogenicity of cancer cell death. Nat Med 2007; 13:54-61 .

(24) Apetoh L, Mignot G, Panaretakis T, Kroemer G, Zitvogel L. Immunogenicity of anthracyclines: moving towards more personalized medicine. Trends Mol Med 2008; 14:141-151.

(25) Apetoh L, Ghiringhelli F, Tesniere A et al. Toll-like receptor 4-dependent contribution of the immune system to anticancer chemotherapy and radiotherapy. Nat Med 2007; 13: 1050-1059.

(26) Freeman GJ, Sharpe AH. A new therapeutic strategy for malaria: targeting T cell exhaustion. Nat Immunol 2012; 13: 1 13-1 15.

(27) Mahoney KM, Rennert PD, Freeman GJ. Combination cancer immunotherapy and new immunomodulatory targets. Nat Rev Drug Discov 2015; 14:561 -584.

(28) Simon RM, Paik S, Hayes DF. Use of archived specimens in evaluation of prognostic and predictive biomarkers. J Natl Cancer Inst 2009; 101 :1446-1452.

Claims

WHAT IS CLAIMED IS:

1 . A method of predicting a benefit of anthracycline therapy for a subject with breast cancer, the method comprising: a) providing a sample of a breast cancer tumour of the subject; b) determining the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; c) comparing said expression levels to a reference expression level of the group of genes from control samples from a population; and d) determining the benefit of anthracycline therapy for the subject; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to an immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with a lesser benefit of anthracycline therapy.

2. The method of claim 1 , wherein the group of genes is at least 4, 5, 6, 7, 8, or 9 of the genes.

3. The method of any one of claims 1 -2, further comprising building a subject gene expression profile from the determined expression levels of the group of genes. 4. The method of any one of claims 1 -3, wherein the immune score comprises the weighted sum expression of the group of genes, optionally scaled for mRNA abundance.

5. The method of claim 4, wherein the subject is classified into a high immune score group where the subject has an immune score above the population median, and wherein the subject is classified into a low immune score group where the subject has an immune score below the population median.

6. The method of claim 5, wherein in the low immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, or IRF1 was relatively low and STAT3 was relatively high, and in the high immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 was relatively high and STAT3 was relatively low, in comparison to a population cohort.

7. The method according to any one of claims 1 -6, further comprising treating the subject with anthracycline if the subject has a relatively low immune score or is in the low immune score group.

8. The method according to any one of claims 1-7, wherein the anthracycline is epirubicin, daunorubicin, doxorubicin, idarubicin, or valrubicin.

9. The method according to any one of claims 1 to 8, wherein determining the gene expression level comprises use of NanoString®.

10. The method according to any one of claims 1 to 9, wherein determining mRNA abundance of the genes comprises use of quantitative PCR. 1 . A computer-implemented method of predicting benefit of anthracycline therapy for a subject with breast cancer, the method comprising: a) receiving, at at least one processor, data reflecting the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL 3, IL8, IRF1 and STAT3; b) comparing, at the at least one processor, said expression levels to a reference expression level of the group of genes from control samples from a population; c) outputting, at the at least one processor, an immune score; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to the immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with a lesser benefit of anthracycline therapy.

12. The method of claim 11 , wherein the group of genes is at least 4, 5, 6, 7, 8, or 9 of the genes.

13. The method of any one of claims 1 1 -12, further comprising building, at the at least one processor, a subject gene expression profile from the determined expression levels of the group of genes.

14. The method of any one of claims 1 1-13, wherein the immune score comprises the weighted sum expression of the group of genes.

15. The method of claim 14, wherein the subject is classified into a high immune score group where the subject has an immune score above the population median, and wherein the subject is classified into a low immune score group where the subject has an immune score below the population median.

16. The method of claim 15, wherein in the low immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, or IRF1 was relatively low and STAT3 was relatively high, and in the high immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 was relatively high and STAT3 was relatively low.

17. The method according to any one of claims 1 1 -16, wherein the processor further outputs a recommendation to treat the subject with anthracycline if the subject has a relatively low immune score or is in the low immune score group. 18. The method according to any one of claims 1 1 -17, wherein the anthracycline is epirubicin, daunorubicin, doxorubicin, idarubicin, or valrubicin.

19. A computer program product for use in conjunction with a general-purpose computer having a processor and a memory connected to the processor, the computer program product comprising a computer readable storage medium having a computer mechanism encoded thereon, wherein the computer program mechanism may be loaded into the memory of the computer and cause the computer to carry out the method defined in any one of claims 1 to 18.

20. A computer readable medium having stored thereon a data structure for storing the computer program product according to claim 19.

21 . A device for predicting benefit of anthracycline therapy for a subject with breast cancer, the device comprising: at least one processor; and electronic memory in communication with the at one processor, the electronic memory storing processor-executable code that, when executed at the at least one processor, causes the at least one processor to: a) receive data reflecting the level of expression in the sample for a group of genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10,

CXCL13, IL8, IRF1 and STAT3; b) compare said expression levels to a reference expression level of the group of genes from control samples from a population, said reference expression level being stored in the memory; c) output an immune score; wherein a statistically significant difference or similarity in the expression of the group of genes compared to the reference expression level correlates to the immune score; a relatively low immune score being associated with a greater benefit of anthracycline therapy, and a relatively high immune score being associated with a lesser benefit of anthracycline therapy.

22. The device of claim 21 , wherein the group of genes is at least 4, 5, 6, 7, 8, or 9 of the genes.

23. The device of any one of claims 21 -22, wherein the code further causes the at least one processor to build a subject gene expression profile from the determined expression levels of the group of genes.

24. The device of claims 21 -23, wherein the immune score comprises the weighted sum expression of the group of genes.

25. The device of claim 24, wherein the code further causes the at least one processor to classify the subject, wherein the subject is classified into a high immune score group where the subject has an immune score above the population median, and wherein the subject is classified into a low immune score group where the subject has an immune score below the population median.

26. The device of claim 25, wherein in the low immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, or IRF1 was relatively low and STAT3 was relatively high, and in the high immune score group, the expression of GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 was relatively high and STAT3 was relatively low. 27. A method of treating a subject with breast cancer, comprising: a) determining the immune score of the subject according to the method defined in any one of claims 1 to 18; and b) selecting a treatment based on said immune score, and preferably treating the subject according to the treatment. 28. The method of claim 27, wherein anthracycline therapy is selected as the treatment where the immune score is relatively low.

29. The method of claim 28, wherein the anthracycline therapy comprises epirubicin, daunorubicin, doxorubicin, idarubicin, or valrubicin.

30. The method of claim 28 of 29, wherein the selected treatment further comprises a checkpoint inhibitor therapy.

31. A composition comprising a plurality of isolated nucleic acid sequences, wherein each isolated nucleic acid sequence hybridizes to:

(a) the mRNA of a group of genes comprising at least one of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3; and/or (b) a nucleic acid complementary to a)

32. An array comprising one or more polynucleotide probes complementary and/or hybridizable to an expression product of each gene of a group genes comprising at least 3 of: GZMB, PRF1 , SELL, CCL22, CXCL10, CXCL13, IL8, IRF1 and STAT3.

33. A kit comprising reagents for detecting mRNA from a sample of a breast cancer tumour of at least 3 genes selected from the group comprising: GZMB, PRF1 , SELL, CCL22,

CXCL10, CXCL13, IL8, IRF1 and STAT3.