FIELD OF THE INVENTION

The present invention provides a method for diagnosing, identifying or monitoring proliferative disorders due to e.g cancer preferably breast cancer in a subject by measuring the change of gene expression in a sample e.g. a blood sample.

The present invention also encompasses oligonucleotide probes and primers corresponding to genes differentially expressed compared to the expression pattern in a normal cell. The use of the method and the oligonucleotides is also an aspect of the invention together with a kit comprising said oligonucleotides.

BACKGROUND OF THE INVENTION

Carcinomas or solid tumors are composed of neoplastic epithelial cells, which form the heart of the tumor, as well as a variety of mesenchymal cell types and extracellular matrix components that comprise the tumor stroma, often termed the tumor microenvironment (TME) ¹ . Historically, it was believed that leukocytic infiltrates, in and around developing neoplasms, were representative of the host's attempt to eradicate neoplastic cells. However, it is now clear that infiltrating immune cells regulate and exert complex, paradoxical roles that are both pro- and anti-tumor during each stage of cancer development ² .

Recent advances have revealed that tumor-host interactions extend well beyond the local TME (i.e. interactions between the neoplastic cells and the nearby stroma) and that cancer development largely depends on the ability of malignant cells to hijack and exploit the normal physiological processes of the host ² ' ³ . Although there is evidence that pronounced abnormalities in a host's systemic immune system caused by aging, stress, immuno-suppressive medications, autoimmune disease or chronic inflammation are associated with increased cancer risk ⁴ , we lack a precise understanding of how a tumor influences circulating immune cells and vice versa.

Blood pervades the entire body and is in a constant state of renewal ⁵ . It is the vehicle by which immune cells circulate between central and peripheral lymphoid organs, and migrate to and from tumor sites. Such a crucial role for blood cells as enhancers of tumor physiology raises questions about how they convey their tumor-promoting effects and whether, if understood, it is reflected in the gene expression of circulating cells. Advances in blood RNA processing and transcriptomics allow whole genome gene expression profiling of circulating blood cells ⁶ . The inventors work within the unique Norwegian Women and Cancer study (NOW AC) ⁷ proposed to use transcriptional profiling of cells preferably blood cells to investigate the genesis of cancer preferably breast cancer (BC) ⁷ when compared to population-based controls.

The successful treatment of cancer depends in part on early detection and diagnosis. It is however, well established knowledge that a diagnostic method will be dependent on the prevalence of the disease in the population from which the cases (persons diagnosed with the disease) originate.

Unlike gene expression profiles from breast cancer that has been reported previously ⁴⁹ ' ⁵⁰ which were determined using cases and control group based on what is called a "workout" which means that the tests are based on control groups with positive findings (abnormal mammograms) and the prevalence of breast cancer in the populations involved will be about 50 %, while the prevalence using samples from the control group of the present invention wherein population-based controls from the NOW AC cohort (healthy controls matched by time of follow-up and birth year within the cohort) were studied will be less than 0,5 %o.

It is believed that the method according to the present invention will be a more reliable diagnostic method, as the gene expression profile has been built up as a result of genes differentially expressed in samples from cases compared to samples from healthy people from the population wherein the test is to be used. It is therefore an objection of the present invention to provide an in vitro method for diagnosing, identifying or monitoring proliferative disorder in a subject, preferably a human being.

The differences in the control groups as decribed above between WO 2011/086174 A2 and Aarae et al. and the present invention may explain the differences between the identified gene profiles.

It should also be noted that our list is based on a study run as an epidemiological study and not with the potentials for quality controls as in a clinical setting. This is an important aspect for improving the test. The inventors have disclosed how the presence of a proliferative disorder are reflected in the gene expression of the host ^' s cells, preferably circulating blood cells, and further how to utilize this in diagnosing, identifying or monitoring proliferative disorder e.g. cancer, preferably breast cancer. It has further been shown that specific functional "groups" of genes are specifically expressed. The method of the present invention may be able to detect the expression pattern reflecting a proliferative disorder long before the onset of symptoms appear.

The primary benefit of breast cancer screening is early detection, which typically results in simpler treatment, a lower chance of recurrence and a greater chance of survival. For younger woman, sensitivity of mammography screening can be as low as 58% and false negatives results cause a delay in diagnosis for women who think they are safe of breast cancer. The cumulative risk of false positive reading after 10 mammographic screens ranges 21% to 49 % for all women. In population offered mammography screening, the risk of unnecessary surgeries was increased by 31% compared to women not using mammography.

More specific technologies complementing existing screening will clearly be beneficial and we believe that the present invention providing a robust new diagnostic method may be such a complement. The method may however, further be used separately for diagnosing, identifying or monitoring proliferative disorder without being part of a screening program.

SUMMARY OF THE INVENTION

The present invention encompasses in first aspect, an in vitro method for diagnosing, identifying or monitoring proliferative disorder in a subject, which method comprises the following step: a) measuring the level of gene expression in a subset of genes set forth in Table 1 or 2 in a sample from said subject; b) comparing the level of gene expression of the subset of genes in the sample from said subject with the level of gene expression of the subset of genes in a standard gene expression pattern extracted from healthy subjects ; c) wherein change of gene expression of each and one gene of said subset in said sample as compared to a standard gene expression pattern being indicative for a proliferative disorder.

A second aspect of the present invention comprises a set of oligonucleotide probes, wherein said set is selected from the oligonucleotides of Table 1 or 2 or

oligonucleotides derived from a sequence set forth in Table 1 or 2 or any combination thereof, or a oligonucleotide with a complementary sequence, or a functional equivalent thereof.

In a further aspect the invention comprises a kit for in vitro diagnosing, identifying or monitoring proliferative disorder comprising:

a collection of oligonucleotide probes and primers capable of detecting the level of expression of at least about 50, preferably at least about 30, more preferably at least about 20, most preferably at least 10 and even more preferably at least 2 forth in Table 1 or 2 or any combination thereof.

The present invention also provides, in one aspect use of the method as described, or the set of oligonucleotide probes, or the kit for diagnosing, identifying or monitoring proliferative disorder in a subject.

Preferred embodiments are set forth in the dependent claims and in the detailed description of the invention.

DESCRIPTION OF THE FIGURES

Figure 1. Gene expression changes in circulating blood cells of breast cancer patients compared to controls.

A. Venn diagram depicting the overlap between genes differentially expressed in circulating blood cells of breast cancer patients compared to controls in the primary (CCl) and the secondary (CC2) dataset. Differential expression was assessed at an FDR< 0.005 in the paired linear analyses. B. Expression fold changes for the 345 overlapping genes differentially expressed in CCl and CC2. Log fold-changes (logFC) in CCl are plotted on the x-axis against the log FCs for the same genes in CC2 on the y- axis. Light grey, genes downregulated in the breast cancer group in both data sets; dark grey genes upregulated in the breast cancer group group in both data sets. . C. Ordering of blood samples according to the sum of expression over those 345 overlapping genes. Heat map colors represent mean-centered fold change expression in log-space D. Accuracy significances of predictors in the validation dataset (CC3). Vertical line represents the accuracy significance of a predictor based on the expression of the 345 overlapping genes. The stippled line represents the distribution of accuracy significances that can be obtained from 100,000 predictors built using 345 random genes present in CC3 (N = 8529). The dotted line represents the distribution of accuracy significances that can be obtained from 100,000 predictors built using 50 genes among the 341 expressed in CC3 of the 345 overlapping genes. Plane line represents the distribution of accuracy significances that can be obtained from 100,000 predictors built using 50 random genes present in the dataset (N = 8529).

Figure 2. Ordering of blood samples based on the enrichment scores of the 38 significant gene sets differentially expressed between breast cancer cases (in dark grey) and controls (in light grey) in the primary (CCl), secondary (CC2) and validation (CC3) case-control serie. Heat map colors represent mean-centered fold change enrichment score in log-space.

Figure 3. Gene set variation analysis of the antigen processing and presentation pathway and the natural killer cell gene set.

A. The antigen processing and presentation pathway from KEGG. Overlapping core genes driving the observed association of the Ag processing and presentation pathway with the presence of breast cancer within CCl and CC2 are colored according to their up- (dark grey) or down- (light grey) regulation in breast cancer patients.

B. Boxplot indicating the enrichment scores from gene set variation analysis for the natural killer (NK) cell gene set associated with the gene expression profiles from breast patients (dark grey) and controls (light grey) included in CCl, CC2 and CC3 (left). List of genes included in the NK gene set significantly associated to disease status in paired linear analysis with FDR<0.10 in at least two of the three datasets and their

corresponding median fold-changes over all datasets (right).

Figure 4. Accuracy significances of predictors in the validation dataset (CC3) that can be obtained from 100,000 predictors built using 10, 25 or 50 genes among the 341 expressed in CC3 of the 345 overlapping genes.

Figure 5. Ordering of blood samples (dark grey: cases, light grey: controls) based on the top quintile of most variable genes in the primary case-control serie (CCl). Heat map colors represent mean-centered fold-change expression in log-space. Figure 6. Accuracy significances obtained to predict breast cancer diagnosis in the publicly available dataset including gene expression profiles of peripheral blood mononuclear cells (PBMC) from BC patients, pre- and post-surgery samples, patients from benign breast diseases, controls, gastrointestinal and brain cancer

patients (GSE27562). Vertical lines represent the accuracy significances of a predictor based on the expression of the 49 genes present in the dataset among our 50-gene best predictor list to predict breast cancer case from controls (plain line) and to predict breast cancer or benign breast abnormalities cases from controls (dotted line). The distribution curves represent the respective distribution of accuracy significances that can be obtained from 100,000 predictors built using 49 random genes present in CC3 (N =8242).

Figure 7. Accuracy significances obtained to predict breast cancer diagnosis in the publicly available dataset including gene expression profiles of whole blood cells from BC patients and controls with suspect mammograms. Vertical lines represent the accuracy significances of a predictor based on the expression of the 32 genes present in the dataset among our 50-gene best predictor list (dotted line) and of the 63 genes present plain line). The dotted distribution curve represents the distribution of accuracy significances that can be obtained from 100,000 predictors built using 32 random genes present in CC3 (N =8242).

Figure 8. Covariate plot displaying the individual contributions of each gene in the antigen processing and presentation pathway to the overall test statistics for differential expression in the primary (CCl) and secondary (CC2) case-control series. Significantly contributing genes (p<0.1) are indicated by the dark line in the hierarchical tree. The grey and the black colors represent association with the control and the breast cancer cases groups, respectively.

DETAILED DESCRIPTION OF THE INVENTION

It is an object of the present invention to provide a method for diagnosing, identifying or monitoring proliferative disorders due to e.g cancer preferably breast cancer in a subject by measuring the change of gene expression in a sample e.g. cells from a blood sample. The present invention also encompasses oligonucleotide probes and primers corresponding to genes differentially expressed compared to the expression pattern in a normal cell. The use of the method and oligonucleotides is also an aspect of the invention together with a kit comprising said oligonucleotides.

Norway has around 436,338 women born between 1943 and 1957, of whom 148,536 (34.4%) participate in the Norwegian Women and Cancer study (NOW AC) ⁶ ' ⁸ ' ⁴⁰ .

This unique material formed the basis of the study resulting in the present invention.

Blood samples were collected at time of diagnosis. The inventors first selected 96 blood samples from breast cancer patients and 116 blood samples from healthy controls matched by time of follow-up and birth year within the NOW AC cohort. From each sample, RNA was extracted from whole blood, and one control sample of two with the highest RNA quality and quantity was then amplified and hybridised along with its corresponding case sample. Using Illumina microarrays, whole blood gene-expression profiles for 96 breast cancer cases and 96 matched controls were generated. Probes fluorescence intensities of scanned images were quantified, trimmed, averaged and normalized to yield the transcript abundance of a gene as an intensity value.

In this primary dataset (CC1), some 9,338 unique genes were regulated across the group of 55 pairs of case/control. The inventors investigated blood gene expression profiles from an additional 63 pairs of BC cases and controls processed in the same way than CC1. To test whether these data findings where replicable in an independent data set a secondary dataset (CC2) was tested. After preprocessing of this secondary dataset (CC2), some 7898 unique genes and 49 pairs of case-control were included in the analyses. 818 of the 7898 genes passing the quality controls were

differentially expressed in CC2, of which 345 were also differentially expressed in CC1. Remarkably, the directionality of differential expression between breast cancer cases and control of all 345 overlapping genes was conserved between the datasets. When patients were ranked according to the sum of expression over the 345 overlapping genes, the majority of blood samples from breast cancer cases were segregated from controls in both dataset.

These findings indicate that a basis of a reliable method for diagnosing, identifying and monitoring a proliferative disorder e.g cancer, preferably breast cancer, has been established. The overlapping 345 genes and alternative splice variants of some of said genes presented in Table 1 may serve as a source from where a subset of genes is chosen and the level of expression is measured and compared to a standard expression pattern. Also oligonucleotide probes and primers may be generated. A selection of preferred 50 genes is presented in Table 2. Each gene in Table 1 and 2 are denoted by international recognized gene symbols.

In order to validate the findings in CC1 and CC2 the validation dataset (CC3) consisted in 8529 unique genes expressed across 59 pairs of breast cancer cases and controls was investigated. Microarray data have been deposited at European Genome-phenome Archive (EGA; https://www.ebi.ac.uk/ega/).

In the 345-gene list, most differentially expressed transcription factors and genes that are part of the general transcription machinery (Table 4) are in blood cells from breast cancer patients compared to controls suggesting that down regulated steady state transcription rates are reduced.

Coordinate and sharp waves of expression of large sets of genes in response to a common stimulus imply that a broad regulatory mechanism is in place, possibly in response to the presence of breast cancer. Although transcription is an essential first step in the regulation of gene expression, the expression of key factors controlling immune cell differentiation and/or function are regulated at the translational level ¹⁵ . In our study, post-transcriptional regulons based on RNA binding proteins (RBPs) were identified as having essential roles in the control of blood cell gene expression changes associated with breast cancer (Table 4, Table 5). RBPs organize nascent RNA transcripts into groups in order to percolate them together down the chain of splicing, nuclear export, stability and translation so that proteins are efficiently produced to meet the needs of the cell ¹⁶ . The abundance of genes in the immune system that are alternatively spliced ¹⁷ , and the connections between splicing and disease in our study, imply that alternative splicing may be a crucial mechanism for regulating and fine-tuning the function of the circulating blood cells associated with the presence of breast cancer.

The inventors surprisingly detected that circulating blood cells in breast cancer patients are enriched in genes which are associated with, and may depend on, systemic immunosuppression, cell motility, metabolism and proliferation. These findings may further be determining when selecting the subset of genes and/or oligonucleotides or primers of the present invention.

In conclusion, the gene expression of circulating blood cells is markedly perturbed in a specific pattern by irregularities other places in the body as e.g the presence of cancer, preferably breast cancer. These detectable changes form a reliable specific pattern not seen in blood cells form healthy persons and will serve as an excellent platform for diagnostic, identification or monitoring purposes. As a consequence the proliferative disorder may be detected long before the onset of other symptoms appears.

The blood-based gene expression method according to the present invention produced surprisingly uniquely robust and reproducible results across microarray platforms and external datasets to detect proliferative disorder, e.g. cancer, preferably breast cancer from population-based controls.

Thus the invention comprises a method for diagnosing, identifying or monitoring proliferative disorders due to e.g. cancer, preferably breast cancer in a subject by measuring a change in gene expression in cells from a sample e.g. a blood sample.

Accordingly a first aspect of the present invention relates to an in vitro method for diagnosing, identifying or monitoring proliferative disorder in a subject, which method comprises the following step: a) measuring the level of gene expression in a subset of genes set forth in Table lor 2 in a sample from said subject; b) comparing the level of gene expression of the subset of genes in the sample from said subject with the level of gene expression of the subset of genes in a standard gene expression pattern extracted from healthy subjects; and c) change of gene expression of each and one gene of said subset in said sample as compared to a standard gene expression pattern being indicative for a proliferative disorder.

As used herein, "proliferative disorder", refers to a condition where

cells are produced in excessive quantities as e.g. in cancer.

As used here in "subject" refers to a mammal. The term "subject" therefore includes for example primates (e.g. humans) and e.g. animals as cows, sheep goats, horses, dogs, cats.

As used here in "sample" refers to a biological sample (e.g. blood sample or fluid sample) As used herein, "gene expression" refers to translation of information encoded in a gene into a gene product (e.g., RNA, protein). Expressed genes include genes that are transcribed into RNA (e.g., mRNA) that is subsequently translated into protein as well as genes that are transcribed into non-coding functional RNA that are not translated into protein (e.g., tRNA, rRNA ribozymes etc.)

As used herein "level of gene expression" or "expression level" refers to the level (e.g., amount) of one or more products (e.g. RNA, protein) encoded by a given gene in a sample or reference standard.

As used herein "subset of genes" refers to a combination of two or more genes, each of which display a change in the expression pattern relative to the standard gene expression pattern. The subset of genes may be selected from the list of Table lor 2 or any combinations thereof.

In one embodiment the subset of genes does not contain the gene: ABHDIO

In one embodiment the subset of genes does not contain the gene: AB13

In one embodiment the subset of genes does not contain the gene: ANXA1

In one embodiment the subset of genes does not contain the gene: AQP9

In one embodiment the subset of genes does not contain the gene: C14orf2

In one embodiment the subset of genes does not contain the gene: CLN5

In one embodiment the subset of genes does not contain the gene: ELA C2

In one embodiment the subset of genes does not contain the gene: EXOC6

In one embodiment the subset of genes does not contain the gene: EXCOSC10

In one embodiment the subset of genes does not contain the gene: GMFG

In one embodiment the subset of genes does not contain the gene: GPBAR1

In one embodiment the subset of genes does not contain the gene: GPR68

In one embodiment the subset of genes does not contain the gene: HIST1H2BK

In one embodiment the subset of genes does not contain the gene: HSPBAPl

In one embodiment the subset of genes does not contain the gene: KIF13B

In one embodiment the subset of genes does not contain the gene: RPL4

In one embodiment the subset of genes does not contain the gene: RPL15

In one embodiment the subset of genes does not contain the gene: RPS29

In one embodiment the subset of genes does not contain the gene: RPL7

In one embodiment the subset of genes does not contain the gene: H2AFX i one embodiment the subset of genes does not contain the gene: MAPREIi one embodiment the subset of genes does not contain the gene: NUP62

I one embodiment the subset of genes does not contain the gene: PGAM1

I one embodiment the subset of genes does not contain the gene: PLAGL2I one embodiment the subset of genes does not contain the gene: RELA

I one embodiment the subset of genes does not contain the gene: RNH1

I one embodiment the subset of genes does not contain the gene: RPL11

I one embodiment the subset of genes does not contain the gene: RPL15

I one embodiment the subset of genes does not contain the gene: RPL21

I one embodiment the subset of genes does not contain the gene: RPS3A

I one embodiment the subset of genes does not contain the gene: S100A8 I one embodiment the subset of genes does not contain the gene: TBC1D15I one embodiment the subset of genes does not contain the gene: THOC4

I one embodiment the subset of genes does not contain the gene: VPS52

I one embodiment the subset of genes does not contain the gene: YWHAB n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 2 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 3 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 20 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 33 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 58 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 96 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 165 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 176 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 181 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 182 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 209 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 214 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 215 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 220 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 228 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 229 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 251 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 273 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 280 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 283 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 292 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 293 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 294 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 296 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 297 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 300 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 311 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 336 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 346 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 347 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 353 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 394 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 406 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 407 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 409 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 410 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 412 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 413 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 414 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 419 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 420 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 421 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 422 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 423 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 424 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 425 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 434 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 479 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 487 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 522 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 532 n one embodiment the set of oligonucleotide probes does not contain SEQ ID NO 533 As used herein "healthy subject" refers to a subject from the NOW AC cohort representing population-based controls (healthy controls matched by time of follow-up and birth year within the cohort).

In one embodiment of this aspect the subset of genes may at least be 50, preferably at least 30, more preferably at least 20, most preferably at least 10 and even more preferably at least 2 or any combination thereof .

In one embodiment the expression from all 50 genes of Table 2 may be measured.

In a further embodiment the expression from all 345 genes of Table 1 may be measured.

In yet a further embodiment only one gene of Table 1 or 2 may be measured.

In a further embodiment the subset of genes may be associated with systemic immunosuppression, cell motility, metabolism and/or proliferation as set forth in Table 1 or 2 or any combination thereof. Also other functional groups of genes may constitute a subset of genes.

In yet a further embodiment the change of gene expression is measured to be at least 10%, at least 20%, at least 30 % when measured as intensity value of scanned images. Other detection methods may be but are not limited to e.g. PCR.

In one or more embodiments of the method of the present invention the difference in gene expression may be measured by using oligonucleotide probes selected from a group set forth in Table 1 or 2, said oligonucleotide probes may be derived from a sequence set forth in Table 1 or 2, or may further be complementary to the sequence of the oligonucletides set forth in Table 1 or 2, or be functional equivalent to the oligonucleotides of Table 1 or 2. The oligonucleotide may act as a detectable probe targeting the correspondent target sequence or may be used as a primer in the amplification step described in the Example section. The subset of genes, oligonucleotides probes or primers according to the present invention may be selected from Table 1 or 2 or any combination thereof.

In a further embodiment said oligonucleotide probes is a set of probes of at least about 50, preferably at least about 30, more preferably at least about 20, most preferably at least about 10 and most preferably at least about 2 oligonucleotide probes. As used herein "oligonucleotide" refers to a nucleic acid molecule and may be DNA, RNA or PNA (peptid nucleic acid) or hybrids thereof or modified forms thereof, e.g chemically modified by e.g methylation. An oligonucleotide probe according to the invention have the ability to bind and probe complementary sequences in the target gene or gene product.

As used herein "derived" refers to oligonucleotides derived from the genes corresponding to the sequences set forth in Table 1 or 2. As the Tables provides Illumina identifiers and internationally recognized HUGO gene symbols, the probe or the primer sequences may be derived from anywhere on the gene to allow specific binding to that gene or its transcript.

As used herein "complementary sequences" refers to sequences with consecutive complementary bases, making it possible for the sequences to bind to one another through their complementarity.

As used herein "functional equivalent" refers to a oligonucleotide which is capable of identifying and binding to the transcript (or DNA) of the same gene as the oligonucleotides set forth in Table 1 or 2 or sequences derived from Table 1 or 2.

In a further embodiment the oligonucleotide may be a probe or a primer which hybridizes under stringent conditions, preferably under high stringency conditions. In further embodiments the oligonucleotide probe may be immobilized on one or more solid supports, which may be but are not limited to e.g a membrane, a plate or a biochip.

As used herein "hybridize under high stringency condition" refers to a specific association of two complementary nucleotide sequences (e.g., DNA, RNA or a combination thereof) in a duplex under stringent conditions as a result of hydrogen bonding between complementary base pairs. Under high stringency conditions do not permit hybridization of two nucleic acid molecules that are not complementary (two nucleic acids molecules that have less than 70% sequence complementarities).

In yet another embodiment said oligonucleotides or primers may preferably be at least about 20, 50, 100 or 200 nucelotides in length. In further embodiments of the present invention the levels of gene expression in said subset of genes may be detected in said sample by determining the levels of RNA molecules encoded by said genes. The levels may but are not restricted to be determined by the use of microarray technique. Also other detection methods as e.g. PCR may be used.

As used herein "RNA" refers to total RNA in a cell.

In another embodiment the sample wherein the gene expression is measured may be circulating blood cells e.g from whole blood, peripheral blood mononuclear cells or other subsets of blood cells. Also cells from body fluids or other tissues may be a sample source as the inventors surprisingly detected that the expression profile of differentially expressed genes found in blood cells also were seen in circulating mononuclear cells including monocytes, T-cells and natural killers cells. In one embodiments these cells may isolated and used as a sample source of the present invention.

In further embodiments the proliferative disorder may be breast cancer and the subject may be but is not restricted to a human being.

In one embodiment the detection of differentially expressed genes may be obtained by translating the subset of genes from Table 1 or 2 to PCR technology and identify the genes differentially expressed compared to control genes.

In another embodiment the detection may be to identify every pair of genes

in the subset of genes from Table 1 or 2 and identify the differentially expressed genes.

I yet another embodiment the pair of genes in the subset of genes of Table 1 or 2 may be investigated and rules created as e.g. if gene 1 > gene 2 etc. then it ^' s a positive case.

In order to decide if a sample from a subject reflects the specific expression pattern characteristic of a proliferative disorder as e.g. cancer, preferably breast cancer, a standard gene expression pattern may be provided, one embodiment of the present invention comprises an in vitro method for preparing a standard gene expression pattern reflecting proliferative disorder in a subject, which method comprises the following step: a) measuring the level of gene expression in a sample from said subject, b) measuring level of gene expression in a control sample from a healthy subject; c) comparing level of gene expression of the sample from said subject with the level of gene expression in a control sample from the healthy subject to produce a characteristic standard gene expression pattern from genes reflecting proliferative disorder as set forth in Table 1 or 2.

According to a second aspect the present invention relates to a set of oligonucleotide probes, wherein said set may be selected from the oligonucleotides of Table 1 or 2 or oligonucleotides derived from a sequence set forth in Table 1 or 2 or a oligonucleotide with a complementary sequence, or a functional equivalent oligonucleotide.

In a further embodiment the set of oligonucleotide probes may comprise at least about 50, preferably at least about 30, more preferably at least about 20, most preferably at least about 10 and most preferably at least about 2 oligonucleotide probes.

The oligonucleotide may act as a detectable probe targeting the correspondent target sequence or may be used as a primer in the amplification step described in the Example section. The oligonucleotide probes or primers according to the present invention may be selected from a group of Table 1 or 2 or any combination thereof.

In one or more embodiments a set of oligonucleotide probes may hybridize under high or medium stringency conditions with the corresponding subset of genes of Table 1 or 2 Further the oligonucleotide probes may be immobilized on a solid support, which may be but are not limited to e.g a membrane, a plate or a biochip.

The use of the oligonucleotides in a product, e.g. a kit form further aspects of the present invention.

Accordingly a further aspect of the present invention relates to a kit for in vitro diagnosing, identifying or monitoring a proliferative disorder comprising:

a collection of oligonucleotide probes and/or primers capable of detecting the level of expression of at least about 50, preferably at least about 30, more preferably at least about 20, most preferably at least 10 and even more preferably at least 2 genes selected from the group set forth in Table 1 or 2. The collection of oligonucleotides provided in the kit of the present invention may be selected from a group or any combination thereof according to Table 1 or 2. In another embodiment the said probes may specifically hybridize to RNA transcripts of said genes, further the the kit may comprise a set of oligonucleotide probes as defined above.

In further embodiments the oligonucleotide probes are preferably immobilized on one or more a solid supports wherein the solid support may be e.g a membrane, a plate, or a biochip.

The present invention also provides, in one aspect use of the method as described, or the set of oligonucleotide probes, or the kit for diagnosing, identifying or monitoring proliferative disorder in a subject.

Having now fully described the present invention in some detail by way of illustration and example for purpose of clarity of understanding, it will be obvious to one of ordinary skill in the art that same can be performed by modifying or changing the invention by with a wide and equivalent range of conditions and other parameters thereof, and that such modifications or changes are intended to be encompassed within the scope of the appended claims.

EXAMPLES Example 1

Blood-wide transcriptional signal of breast cancer and its diagnostic potential

NOW AC is a prospective cohort that follows a large representative sample of the Norwegian female population (-34% of all women born between 1943-1957) and biobanks blood samples both prior to (n=50,000), and at the time of, breast cancer diagnosis (n=385 cases with age-matched controls) ⁸ . We selected from this cohort 96 blood samples from breast cancer patients and 116 blood samples from controls matched by time of follow-up and birth year. From each case sample, total RNA was extracted from whole blood, and a matched control sample was selected that exhibited the highest quality and quantity of RNA. Both case and control were amplified and hybridized simultaneously. to ablate technical effects using Illumina microarrays. In total, we generated whole blood gene-expression profiles for 96 breast cancer cases and 96 matched controls. Samples received more than 4 days after collection (N=6), with low RNA quality (RIN<7, N=24), outliers (N=14), misdiagnosis (N=3) and unmatched samples (N=35) were excluded from our analyses. Probes intensities of scanned images were quantified, trimmed, normalized and averaged to yield the transcript abundance of each gene as an intensity value. In this first case-control dataset (CC 1), these quality control steps identified 9,338 unique genes of sufficient quality across a group of 55 pairs of case/control (Table 3). Disease status was associated with substantial differences in blood gene expression profiles across the case/control pairs included in CC1 (P = 6 x 10 ^"8 , global test. This blood-wide signal of breast cancer diagnosis is illustrated by the grouping of samples according to disease status in an unsupervised clustering based on the top quintile of most variable genes (Figure 5). We identified 3479 genes showing significant differences in expression (False Discovery Rate (FDR) < 0.005; paired linear analysis) between the breast cancer and controls groups with a relatively low median absolute value of fold-change equal to 1.13 and ranging from 1.03 to 2.03. This indicates that gene expression changes associated with breast cancer are of relatively low amplitude but ubiquitous in circulating blood cells.

Data analysis

The data analysis in the examples was performed using R (http://cran.r-project.org), an open-source-interpreted computer language for statistical computation and graphics, and tools from the Bioconductor project (http://www.bioconductor.org), adapted to our needs. To identify single gene differentially expressed between cases and controls, we conducted paired gene-wise linear analysis with application of empirical Bayes methods implemented in the software package Limma for the R computing environment. False discovery rate ⁴¹ (FDR) was calculated to adjust for multiple testing. We trained naive Bayes' classifiers to predict outcome using the ranked log-odds that the gene is differentially expressed in the paired linear analysis and used internal leave-one-out cross-validation. We investigated the distribution of accuracy significances that can be obtained from 100,000 predictors built using a defined number of genes among the selected gene list, and compared it with predictors using the same number of genes randomly chosen within the dataset. Functional annotations were curated from the GeneCards encyclopedia ⁴² (www.genecards.org). We applied functional clustering via the Database for Annotation, Visualization, and Integrated Discovery (DAVID) (ref) (http : // davi d . ab cc . ncif erf . gov/ ^' ) . Gene set variation analysis package for the R computing environment implements a non-parametric unsupervised method for assessing gene set enrichment in each sample. We used several compendia and ontologies including the Gene Ontology ⁴³ (GO), the KEGG ⁴⁴ and Biocarta (http://www.biocarta.com/) pathway databases, and curated gene signatures databases obtained from the literature, all available from the Molecular Signature Database, MSigDB . Gene set specific to immune cell subsets were generated using CD markers of no more than 3 immune cell subtypes ²¹ and transcripts specifically overexpressed in each differentiated immune cell subtype ²² . The global test {Goeman, 2004 24 /id} was used to test the influence of groups of genes on pairs of breast cancer cases and controls. Genes or cluster of genes driving the observed association of the gene set with the presence of BREAST CANCER were defined as core genes (multiplicity corrected P-value < 0.1).

Example 2

To test whether these findings were replicable in an independent data set (CC2), we investigated blood gene expression profiles from an additional 49 pairs of breast cancer cases and controls subjected to the same data processing as CC1 (Table 3). 418 of the 7898 genes passing quality controls were differentially expressed in CC2 with a FDRO.005, of which 345 were also differentially expressed in CC1 (P = 3 x 10 ^"60 , hypergeometric test; Figure 1 A, Table 4). Remarkably, the directionality of differential expression between breast cancer cases and controls of all 345 overlapping genes was conserved between datasets (Figure IB). When patients were ranked according to the sum of expression over the 345 overlapping genes, the majority of blood samples from breast cancer cases were segregated from controls in both datasets (Figure 1C).

Example 3

Using both CC1 and CC2 to select genes differentially expressed in blood samples from breast cancer patients compared to controls, we next asked whether we could accurately classify a third independent dataset encompassing (CC3). Data were subjected to the same processing as CC1 and CC2 (Table 3) and included the expressions of 8529 unique genes across 59 new case-control pairs from NOW AC. Of note, amplified RNA from the blood samples in CC3 was hybridized using a different version of the Illumina array system that includes 12 samples per array with about 40% less probes per signal. That can explain, at least partly, the higher FDR associated with the 345-gene list (Table 4). We built a predictor including all 341 expressed genes in CC3 of the 345- gene list and accurately predicted disease status in this validation dataset (P = 8.7 ^χ 10 ^{" 5} ; fisher test; Figure ID). We investigated the distribution of accuracy significances that can be obtained from 100,000 predictors built using 50 genes among the 341 expressed genes, and compared it with predictors using 50 random genes present in CC3 (N = 8529, Figure ID). The "best" 50-gene predictor chosen among the 341 significant genes (Table 4) had an accuracy of 72.9% to predict the presence of breast cancer (sensitivity = 83.1% and specificity = 62.7%; P = 3.0 χ 10 ^"9 ; fisher test). In all three datasets, we investigated whether RNA quality quantified by the RIN value and the use of menopausal hormone therapies or other specific medications could explain the misclassification of controls (i.e. false positives) and cases (i.e. false negatives). In CC2 only, true and false positives had a significantly lower RNA quality defined by the RIN value compared to false and true negatives, respectively (P = 1 ^χ 10 ^"3 and 0.02; student test). This indicates that gene expression changes in blood cells of breast cancer patients compared to controls can be confounded by RNA degradation although all samples were selected with a good RNA quality (RIN>7). Also, a significant proportion of controls misclassified as cases in CC3 (36.4%) were currently using of either a selective serotonin reuptake inhibitors (ATC N06AB) or a selective beta-blocking agent (ATC C07AB). Both drugs were previously found to inhibit the expression of T-cell and adaptive immunity-related genes ^9-11 that may explain the lower specificity in CC3 of the best 50-gene predictor identified. Notably, perturbagen signatures from the connectivity map associated with histone deacetylase, Hsp90, tyrosine kinase and immune response inhibitors were significantly enriched in our 50-gene predictor (Table 9, Table 5). Overall, this indicates that our 50-gene predictor contains cytostatic signals including the specific suppression of tumor immunity and that medications influencing transcription involved in those processes can be confounder of the blood-based signal associated with the presence of breast cancer. Considering breast cancer samples included in all 3 datasets, we did not observe tumor receptor specific or stage specific differences between true and false positives.

Example 4

To further validate the results, we investigated whether we were able to predict breast cancer diagnosis in two external datasets deposited in NCBI's Gene Expression Omnibus ¹² including gene expression profiles of peripheral blood mononuclear cells (PBMC) from breast cancer patients, pre- and post-surgery samples, patients from benign breast diseases, controls, gastrointestinal and brain cancer patients ¹³ (GSE27562) and gene expression profiles of whole blood cells from breast cancer patients and controls with suspect mammograms ¹⁴ (GSE164430). In the PBMC dataset, our 50-gene predictor was able to accurately predict the presence of breast cancer compared to controls from PBMC gene expression profiles (91.5% accuracy, Figure 6). This indicates that our diagnostic profile for breast cancer identified from whole blood cells is clearly found in circulating mononuclear cells including monocytes, T-cells, B-cells, and natural killer (NK) cells. Of note, the naive bayes classifier based on the top 3 factors published in LaBreche et al. (N = 77 genes) achieved a significantly lower accuracy (79.3%; P = 0.02 McNemar χ test) in their own dataset. All PBMC samples from other cancer types were predicted as normal, which indicates that our predictor is specific for breast cancer. Since our predictor was not trained to differentiate malignant breast cancer from benign breast diseases, we obtained significantly lower accuracy when we included those samples (63.4% accuracy; Figure 6). The expressions of only 33 genes of our 50-gene predictor were available in the second dataset (GSE164430) although we were able to significantly predict breast cancer diagnosis compared to women with suspect screening mammograms (P = 0.008; fisher test, Figure 7). We were not granted access of the Breast Cancer Test gene list derived from this study. In conclusion, our blood-based gene expression analysis produced uniquely robust and reproducible results across microarray platforms and external datasets to specifically detect breast cancer from population-based controls, warranting further analysis of the processes found deregulated by the presence of a breast cancer in circulating blood cells including PBMC.

Example 5

Pathway and gene set variation analyses

Pathway enrichment analysis showed that our 345-gene list differentially expressed in blood with the presence of breast cancer patients was enriched with a FDR < 0.10 for gene ontology categories related to apoptosis, RNA binding, spliceosome and RNA splicing, protein synthesis, RNA metabolism, transcriptional regulation, cell cycle, metabolism and signal transduction (Table 10). Expert functional annotation curated from GeneCards (Table 4) revealed additional grouping of genes involved in immune processes, cell growth/proliferation, cytoskeletal regulation, signal transduction, protein and cell metabolism. To further investigate how breast cancer affects gene expression in blood, we performed gene set variation analysis in CC1 and CC2 datasets and validate the results in CC3. The collection of gene sets used in the analysis consisted of release 3.0 of the C2 (curated gene sets) and C5 (Gene ontology gene sets) sub-collections of the Molecular Signatures Database (http://wvvw.broad.mit.edu/gsea/msigdb/) of size ranging from 10 to 500 genes. We found 58 gene sets overlapping between the top 200 gene sets deregulated in CC1 and CC2 (FDR < 2 ^χ 10 ^"4 ; linear analysis). Although we previously identified a confounding factor with the current use of specific drugs by controls in CC3, forty- five of the 58 significant gene sets overlapping in CC1 and CC2 were validated in CC3 (FDR < 0.15, Table 6) and showed remarkably comparable enrichment scores according to disease status in all three datasets (Figure 2). Gene set variation analysis revealed similar processes seen after functional clustering of our 345- gene list including apoptosis, metabolism and transcription pathways, but also identified additional gene signatures notably involved in antigen processing and presentation and Myc target genes (Figure 2, Table 6).

Example 6

Circulating blood cells in breast cancer patients expressed changes in genes related to immune functions

The antigen processing and presentation pathway and NK-cell mediated immunity

The antigen processing and presentation (APP) pathway downregulated in blood cells of BC patients (Figure 2, Table 6) was the most direct evidence that our signature is based specifically on the transcriptome of circulating immune effector cells. Mechanisms that regulate APP alter the form and the quantity of the epitopes that are presented by the major histocompatibility complex (MHC) molecules for immune recognition and can dictate tumor immunogenicity ¹⁹ ' ²⁰ .

We investigated the overlapping core genes driving the observed association of the APP pathway with the presence of breast cancer within CCl and CC2 (Figure 3 A, Figure 8). All core genes that part of the MHC class II pathway were downregulated in blood samples from breast cancer patients compared to controls including the interferon gamma-inducible protein 30 (IFI30) and cathepsin S (CTSS) involved in the endocytic generation of MHC class II-restricted epitopes as well as CD74 involved in the formation and transport of MHC class II protein and CD4 a co-receptor that assists the T cell receptor (TCR) (Figure 3 A). A recent study integrating measurements of copy number and gene expression in breast cancer tumor tissue identified trans-acting deletion hotspots localized to both TCR loci. Given that genomic copy number loss at the TCR loci drived a trans-acting immune response module, we asked whether cognate mRNAs modulated by both TCR loci in tumor-infiltrating lymphocytes (N= 114, Table 7) were differentially expressed in circulating blood cells of breast cancer patients included in CCl and CC2. The gene set was significantly enriched when comparing the gene expression profiles of circulating blood cells from breast cancer patients compared to controls (P = 1.4 ^χ 10 ^"8 and 1.5 ^χ 10 ^"5 in CCl and CC2, respectively; global test). Ranking the samples according to the sum of expression over 42 genes defined as core genes in at least one of the dataset, the majority of blood samples from breast cancer cases were segregated from controls in both datasets (Figure 8). This indicates that part of the signal from tumor-infiltrating lymphocytes with rearranged TCR loci was also found in circulating blood cells of breast cancer patients compared to controls. Within the MHC class I pathway, PSME3 of the immune-proteasome was defined as a core gene in the APP pathway downregulated in breast cancer patients (Figure 3A). In addition, three genes coding for the proteasome (PSMB2, PSMB10, PSMD1) within our 345-gene list (Table 5) were downregulated in blood cells of breast cancer patients compared to controls. These findings support that alterations in epitope processing occur in blood cells of breast cancer patients owing at least partly to changes in the proteasomal machinery. Also, core genes involved in peptide-loading onto MHC class I molecules (TAPBP, CALR) were downregulated in breast cancer patients (Figure 3 A). Finally, several heat shock proteins within the APP pathway were also defined as core genes (Figure 3 A) including the downregulation in breast cancer blood samples of HSP70-1 and -2 genes in breast cancer blood samples (HSPA1A and HSPA1B) that encode the major heat-inducible 70kDa heat shock protein (HSP70). In contrast, we observed upregulation in breast cancer blood samples of HSPA5 with anti-inflammatory properties.

When investigating the expression of CD markers of no more than 3 immune cell subtype ²¹ and transcripts specifically overexpressed in each differentiated immune cell subtypes ²² , genes specific to NK cells (Table 8) were consistently downregulated within the set and in samples from breast cancer patients compared to controls (Figure 3B). Consistent with this finding, the NK cell-mediated cytotoxicity pathway from KEGG was significantly downregulated in blood samples from breast cancer patients compared to controls (mean FDR = 0.08 across all three datasets; gene set variation analysis). Overall, this indicates that breast cancer evasion of anti-tumoe immunity is associated with the downregulation of the APP pathway and NK cell-mediated immunity not only in murine models or in the TME of good prognosis solid tumors ²³ but also in the transcriptome of circulating blood cells. Remarkably, one epidemiological study has linked low peripheral blood NK cell cytotoxic activity with increased cancer risk ²⁴ . Changes in expression of several cytoskeleton-regulating genes among our 345-gene list (Table 4) may imply further deregulation of the APP pathway as well as the disruption of the overall immune response against the damaged cells. For example, integrin-linked kinase (ILK), signaling modules of GTPase-mediated actin polymerization (ABR, ABB, ARHGAPl), several actin-related genes (ACTB, ACTGJ, ARPC5L, PFN1), several genes involved in microtubule-based motor proteins dynein and kinesin (DYNLRB1, DYNC1H1, K1F13B), and two engulfment and cell motility genes {ELMO I, ELMO 2) were downregulated in blood cells of breast cancer patients (Table 6). A deregulated cellular cytoskeleton could disrupt the vacuolar transport of MHC-peptide complex in blood cells of breast cancer patients, as well as deregulate the cellular programs that lead to activation, proliferation, differentiation, secretion, cell-cell interaction and survival of immune cells, and phagocytosis ²⁵ . During the activation of a resting lymphocyte, large metabolic demands are also placed on the cell as it initiates proliferation and cytokine production ²⁶ ' ²⁷ . However, it has been shown that antigen stimulation signaling is required to control the ability of resting cells to take up and utilize nutrients at levels sufficient to maintain viability ²⁸ . Consistent with the defect in the APP pathway in blood cells from breast cancer patients, we observed a lowered cell metabolism with an overall downregulation of glycolysis and glucose metabolism (Table 10, Figure 2). In accordance with our results, altered plasma levels of enzymes involved in glucose, lipid, and amino acid metabolism were observed during tumor development in mice ²⁹ indicating the presence of a tumor triggers systemic metabolic dysregulation. When extrinsic growth factors are limiting like it is suggested in blood cells of breast cancer patients, cells can activate pathways such as autophagy that promotes the degradation of intracellular constituents to provide a source of ATP. In accordance with this, two genes promoting autogphagy (ATG12, VPM1) were upregulated in blood cells of breast cancer patients (Table 5).

If cells begin with lower energic levels and less biomass, they have to grow more to reach a size sufficient to enter a replicative division ²⁸ ' °. Protein synthesis has been defined as a key determinant of cell growth and proliferation ³³ and is regulated by either the rates of translation initiation and elongation or ribosome biogenesis" ^2"34 . Consistent with this, three components of the eukaryotic initiation factor 4 complex (EIF4A1, EIF4A3, EIF4H) were significantly downregulated in breast cancer patients compared to controls. Furthermore, several genes involved in ribosomal biogenesis (GAR1, SURF6, RRS1) were downregulated while several small (RPS3A, RPS29) and large (RPL4, RPL5, RPL7, RPLll, RPL15, RPL21, RPL41) ribosomal proteins (RPs) were upregulated in blood cells from breast cancer patients compared to controls. The accumulation of RPs can occur due to defects in ribosome assembly caused by an imbalance among RPs or caused by a defect in the assembly process ³⁵ . Recent findings have demonstrated that components of the translational apparatus are multifunctional and that several individual RPs play a role in regulating cell growth, transformation and death" ⁵ . Such an accumulation of any of several RPs can interface with the p53 system, leading to cell-cycle arrest or to apoptosis. For example, both RPL5 and RPLll upregulated in blood cells from breast cancer patients have been found necessary for the accumulation of p53 and the consequent Gl arrest ³⁶ . Also, RPL1 1 can bind and sequester c-myc, itself a positive promoter of ribosome synthesis, and particularly of RPLl l synthesis ³⁶ ' ³ '. Consistent with this, we found a decrease in gene set expression of the targets downregulated by Myc ^j (Figure 2) in blood cells from breast cancer patients. The downregulated Myc targets defined as core genes in all three datasets are involved in global gene regulatory networks with specific influence on cell growth and proliferation {ILK, ARPC4, PP2R4, ERBB2, CEBPA). Since some Myc target genes are regulators of cell growth while others function in cell differentiation and proliferation pathway, Myc is apparently poised at the interface of these processes in circulating blood cells from breast cancer patients compared to controls. The corresponding gene set including targets up regulated by Myc ³⁸ was upregulated but not significant in blood samples of breast cancer patients compared to controls (mean FDR = 0,34 across all three datasets; gene set variation analysis). During immune cell proliferation, a fine regulation of the cell cycle is required to maintain immune cell homeostasis ³⁹ . In our 345-gene list, differentially expressed genes in the cell cycle regulation (TERF2, CKAP5, CUL4B, MCM3, HBP1, NUDC, CTCF, TUBB, USP9X, H2AFX) in circulating blood cells of breast cancer patients were mostly involved in the arrest at mitosis checkpoint. In GSVA, the gene set associated with the loss of Nip from mitotic centrosomes required at the onset of mitosis was downregulated in the blood samples from breast cancer patients compared to controls (Figure 2, Table 6). Further, the integrin signaling pathway was downregulated in blood samples from breast cancer patients compared to controls. That indicates that integrin-mediated intracellular signals including cellular shape, mobility, and progression through the cell cycle were downregulated by the presence of breast cancer.

Together, these findings show that tumor development is associated with, and may depend on, systemic immunosuppression including the impairment of Ag processing/presentation pathway, and the metabolism, growth, motility, and proliferation of immune cells with a remarkable suppression of NK cell-mediated immunity.

The work leading up to the present invention has received funding from the European Research Council under the European Communit's Seventh Framework Programme (FPT/2007-2013)/ERC grant agreement no. 232997.

The human biological material applied in the study presented in present application has been approved by Regional Committees for Medical and Health Research Ethics in Norway; P REK NORD 27/2004 and P REK NORD 146/2006, and is in accordance with the "Biobankloven" (lov av 21 February 2003 no.12 om biobanker).

References

1 Bissell, M. J. & Radisky, D. Putting tumours in context. Nature reviews. Cancer

1, 46-54, doi: 10.1038/35094059 (2001).

2 de Visser, K. E., Eichten, A. & Coussens, L. M. Paradoxical roles of the immune system during cancer development. Nature reviews. Cancer 6, 24-37, doi: 10.1038/nrcl782 (2006).

3 Hanahan, D. & Weinberg, R. A. Hallmarks of cancer: the next generation. Cell

144, 646-674, doi: 10.1016/j cell.2011.02.013 (2011).

4 Grivennikov, S. I, Greten, F. R. & Karin, M. Immunity, inflammation, and cancer. Cell 140, 883-899, doi: 10.1016/j .cell.2010.01.025 (2010).

5 Ogawa, M. Differentiation and proliferation of hematopoietic stem cells. Blood

81, 2844-2853 (1993).

6 Dumeaux, V. et al. Gene expression analyses in breast cancer epidemiology: the Norwegian Women and Cancer postgenome cohort study. Breast cancer research : BCR 10, R13, doi : 10.1186/bcr 1859 (2008).

7 Lund, E. & Dumeaux, V. Systems epidemiology in cancer. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology 17, 2954-2957, doi: 10.1158/1055-9965.EPI-08-0519 (2008).

8 Lund, E. et al. Cohort profile: The Norwegian Women and Cancer Study— NOWAC— Kvinner og kreft. International journal of epidemiology 37, 36-41, doi: 10.1093/ije/dyml37 (2008).

9 Taler, M. et al. Immunomodulatory effect of selective serotonin reuptake inhibitors (SSRIs) on human T lymphocyte function and gene expression. European neuropsychopharmacology : the journal of the European College of Neuropsychopharmacology 17, 774-780, doi: 10.1016/j. euroneuro.2007.03.010 (2007).

10 Sacre, S., Medghalchi, M., Gregory, B., Brennan, F. & Williams, R. Fluoxetine and citalopram exhibit potent antiinflammatory activity in human and murine models of rheumatoid arthritis and inhibit toll-like receptors. Arthritis and rheumatism 62, 683-693, doi: 10.1002/art.27304 (2010). Gasser, R. Myocardial ischemia and the immune system: some thoughts and changing views J Clin Basic Cardiol 14, 7 (2011).

Barrett, T. et al. NCBI GEO: archive for functional genomics data sets~10 years on. Nucleic acids research 39, D1005-1010, doi: 10.1093/nar/gkql l84 (2011). LaBreche, H. G., Nevins, J. R. & Huang, E. Integrating factor analysis and a transgenic mouse model to reveal a peripheral blood predictor of breast tumors. BMC medical genomics 4, 61, doi: 10.1186/1755-8794-4-61 (2011).

Aaroe, J. et al. Gene expression profiling of peripheral blood cells for early detection of breast cancer. Breast cancer research : BCR 12, R7, doi: 10.1186/bcr2472 (2010).

Rebhan, M., Chalifa-Caspi, V., Prilusky, J. & Lancet, D. GeneCards: integrating information about genes, proteins and diseases. Trends in genetics : TIG 13, 163 (1997).

Anderson, P. Post-transcriptional regulons coordinate the initiation and resolution of inflammation. Nature reviews. Immunology 10, 24-35, doi: 10.1038/nri2685 (2010).

Keene, J. D. RNA regulons: coordination of post-transcriptional events. Nature reviews. Genetics 8, 533-543, doi: 10.1038/nrg2111 (2007).

Lynch, K. W. Consequences of regulated pre-mRNA splicing in the immune system. Nature reviews. Immunology 4, 931-940, doi: 10.1038/nril497 (2004). Garrido, C. et al. Alterations of HLA class I expression in human melanoma xenografts in immunodeficient mice occur frequently and are associated with higher tumorigenicity. Cancer immunology, immunotherapy : CII 59, 13-26, doi: 10.1007/s00262-009-0716-5 (2010).

Neefjes, J., Jongsma, M. L., Paul, P. & Bakke, O. Towards a systems understanding of MHC class I and MHC class II antigen presentation. Nature reviews. Immunology 11, 823-836, doi: 10.1038/nri3084 (2011).

Curtis, C. et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature 486, 346-352, doi: 10.1038/naturel0983 (2012).

Birnbaum, K. D. & Kussell, E. Measuring cell identity in noisy biological systems. Nucleic acids research 39, 9093-9107, doi: 10.1093/nar/gkr591 (2011). Watkins, N. A. et al. A HaemAtlas: characterizing gene expression in differentiated human blood cells. Blood 113, el-9, doi: 10.1182/blood-2008-06- 162958 (2009). Vivier, E., Ugolini, S., Blaise, D., Chabannon, C. & Brossay, L. Targeting natural killer cells and natural killer T cells in cancer. Nature reviews. Immunology 12, 239-252, doi: 10.1038/nri3174 (2012).

Imai, K., Matsuyama, S., Miyake, S., Suga, K. & Nakachi, K. Natural cytotoxic activity of peripheral-blood lymphocytes and cancer incidence: an 11-year follow-up study of a general population. Lancet 356, 1795-1799, doi: 10.1016/S0140-6736(00)03231-1 (2000).

Vicente-Manzanares, M. & Sanchez-Madrid, F. Role of the cytoskeleton during leukocyte responses. Nature reviews. Immunology 4, 110-122, doi: 10.1038/nril268 (2004).

Jones, R. G. & Thompson, C. B. Revving the engine: signal transduction fuels T cell activation. Immunity 27, 173-178, doi: 10.1016/j.immuni.2007.07.008 (2007).

Fox, C. J., Hammerman, P. S. & Thompson, C. B. Fuel feeds function: energy metabolism and the T-cell response. Nature reviews. Immunology 5, 844-852, doi: 10.1038/nril710 (2005).

Rathmell, J. C, Vander Heiden, M. G., Harris, M. H., Frauwirth, K. A. & Thompson, C. B. In the absence of extrinsic signals, nutrient utilization by lymphocytes is insufficient to maintain either cell size or viability. Molecular cell 6, 683-692 (2000).

Pitted, S. J. et al. Tumor microenvironment-derived proteins dominate the plasma proteome response during breast cancer induction and progression. Cancer research 71, 5090-5100, doi: 10.1158/0008-5472.CAN-l l-0568 (2011). Frauwirth, K. A. & Thompson, C. B. Activation and inhibition of lymphocytes by costimulation. The Journal of clinical investigation 109, 295-299, doi: 10.1172/JCI14941 (2002).

Pardee, A. B. Gl events and regulation of cell proliferation. Science 246, 603- 608 (1989).

Holland, E. C, Sonenberg, N., Pandolfi, P. P. & Thomas, G. Signaling control of mRNA translation in cancer pathogenesis. Oncogene 23, 3138-3144, doi: 10.1038/sj.onc.1207590 (2004).

Thomas, G. An encore for ribosome biogenesis in the control of cell proliferation. Nature cell biology 2, E71-72, doi: 10.1038/35010581 (2000). Warner, J. R., Vilardell, J. & Sohn, J. H. Economics of ribosome biosynthesis. Cold Spring Harbor symposia on quantitative biology 66, 567-574 (2001). Warner, J. R. & Mcintosh, K. B. How common are extraribosomal functions of ribosomal proteins? Molecular cell 34, 3-11, doi: 10.1016/j .molcel.2009.03.006 (2009).

Dai, M. S., Arnold, H., Sun, X. X., Sears, R. & Lu, H. Inhibition of c-Myc activity by ribosomal protein Ll l . The EMBO journal 26, 3332-3345, doi: 10.1038/sj . emboj .7601776 (2007).

van Riggelen, J., Yetil, A. & Felsher, D. W. MYC as a regulator of ribosome biogenesis and protein synthesis. Nature reviews. Cancer 10, 301-309, doi: 10.1038/nrc2819 (2010).

Zeller, K. I, Jegga, A. G., Aronow, B. J., O'Donnell, K. A. & Dang, C. V. An integrated database of genes responsive to the Myc oncogenic transcription factor: identification of direct genomic targets. Genome biology 4, R69, doi: 10.1186/gb-2003-4-10-r69 (2003).

Baek, K. H. et al. p53 deficiency and defective mitotic checkpoint in proliferating T lymphocytes increase chromosomal instability through aberrant exit from mitotic arrest. Journal of leukocyte biology 73, 850-861 (2003).

Lund, E. et al. External validity in a population-based national prospective study—the Norwegian Women and Cancer Study (NOW AC). Cancer causes & control : CCC 14, 1001-1008 (2003).

Smyth, G. K. Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Statistical applications in genetics and molecular biology 3, Article3, doi: 10.2202/1544-6115.1027 (2004).

Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B 57, 289-300 (1995).

Huang da, W., Sherman, B. T. & Lempicki, R. A. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nature protocols 4, 44-57, doi: 10.1038/nprot.2008.211 (2009).

Ashburner, M. et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nature genetics 25, 25-29, doi: 10.1038/75556 (2000).

Ogata, H. et al. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic acids research 27, 29-34 (1999).

Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences of the United States of America 102, 15545- 15550, doi: 10.1073/pnas.0506580102 (2005).

Goeman, J. J., van de Geer, S. A., de Kort, F. & van Houwelingen, H. C. A global test for groups of genes: testing association with a clinical outcome. Bioinformatics 20, 93-99 (2004).

WO 2011/086174 A2 Aarae, J. et al. Gene expression profiling of peripheral blood cells for early detection of breast cancer. Breast Cancer Research 12, no. 1, R7 (2010)