Technical Field

The present invention relates to the field fo Medicine, particularly to breast cancer, especifically to a a new method for predicting the response to therapy against HER2 in HER2+ breast cancer patients that are not receiving chemotherapy. The method has potential applications in the clinical management and monitoring of said HER2+ breast cancer patients.

Background Art

The HER2+ breast cancer, defined by IHC/FISH (standard definition) ¹ , accounts for -20% of all breast tumours. Initially established as a prognostic biomarker, its greatest value today is as a predictor of trastuzumab benefit as well as other agents that target the HER2 pathway. Introduction of trastuzumab therapy markedly improved the poor prognosis associated with HER2+ ² . Subsequent identification of resistance mechanisms and the incorporation of new drugs with a better or different blockade of HER2 have improved survival outcome in the metastatic setting ^3,4 . In early stages, incorporation of new anti- HER2 agents has provided discordant results. On one hand, locally advanced and large operable tumours showed dramatic increase in pathological complete rates (pCR) with the incorporation of lapatinib or pertuzumab to standard neoadjuvant trastuzumab and chemotherapy combination. With pCR validated as surrogate endpoint for disease-free survival (DFS) in patients with HER2+ disease ⁵ , pertuzumab has granted approval by the Eurpean Medicines Agency (EMA) and the Food and Drug Administration (FDA) for this population. On the other hand, the addition of lapatinib to standard adjuvant trastuzumab and chemotherapy combination, provided statistically non-significant absolute benefit in the range of 2% at 4.5-years in DFS ⁶ . Results from a second large study incorporating pertuzumab to trastuzumab in the same setting are awaited. However, a constraint to clinically relevant achievements in this population is the low - modest risk following the high efficacy of trastuzumab and chemotherapy. Indeed, a single-arm treatment study in patients with predominantly stage I HER2+ breast cancer (i.e. T1 and node-negative or N1 mic) exploring adjuvant low-intensity weekly paclitaxel for 12 weeks with 1 year of trastuzumab obtained a 3-year 98.7% DFS ⁷ .

New strategies are needed in early HER2+ breast cancer to optimize and de-escalate treatments. In the HER2-negative/HR+ disease, gene expression-based assays have been incorporated to personalize risk and, most important, to establish the benefits and needs of adjuvant chemotherapy. The lack of any predictive tool in the HER2+ landscape is a question addressed for years that is penalizing adjuvant studies.

Three previous neoadjuvant studies have shown that ERBB2 mRNA expression alone is associated with a higher likelihood of pCR following chemotherapy and anti-HER2 therapy in patients with HER2+ disease ^8"10 . In the NeoALTTO study ¹¹ , RNA sequencing of 254 baseline samples (of 455 patients included) was evaluated ⁸ . The NeoALTTO randomized 455 women with HER2+ early-stage breast cancer to trastuzumab, lapatinib, or the combination for 6 weeks followed by the addition of weekly paclitaxel for 12 weeks. After systemic treatment, patients underwent surgery ¹¹ . The results revealed that high ERBB2 mRNA expression was associated with pCR in all treatment arms ⁸ . In another

retrospective study from the NeoALTTO trial, HER2 protein expression-only was also found associated with a higher likelihood of pCR ¹² . In the second clinical trial, the

CALGB40601 , patients with stage II to III HER2+ breast cancer were randomly assigned to chemotherapy (i.e. paclitaxel) plus trastuzumab alone or with the addition of lapatinib for 16 weeks before surgery ⁹ . Retrospective analysis revealed that high expression of ERBB2 by mRNA were associated with pCR in the entire population ⁹ . Finally, in the Tryphaena open-label phase II study, patients with operable, locally advanced, or inflammatory HER2+ breast cancer were randomized 1 :1 :1 to receive 6 neoadjuvant cycles of 3 different multi-agent chemotherapy regimens in combination with trastuzumab and pertuzumab ¹⁰ . Of the different molecular biomarkers analyzed, HER2 levels (protein and mRNA) showed an association with pCR rates when data from all arms were pooled.

The previous associations between baseline ERBB2 mRNA or protein with pCR following anti-HER2 therapy needs special consideration. Indeed, the 3 clinical trials (i.e.

NeoALTTO, CALGB40601 and Tryphaena) included backbone chemotherapy in all their treatment arms. Thus, one cannot discriminate the predictive effect of ERBB2 expression over chemotherapy. In fact, a previous large study in the adjuvant setting observed a significant interaction between HER2-positivity (as defined standard criteria using IHC and/or FISH) and paclitaxel benefit ¹³ . In this study, 1 ,500 women with node-positive breast cancer who had been randomly assigned to receive doxorubicin (60, 75, or 90 mg per square meter of body-surface area) plus cyclophosphamide (600 mg per square meter) for four cycles, followed by four cycles of paclitaxel (175 mg per square meter) or observation. Tissue blocks from 1322 of these 1500 women were available ¹³ .

Immunohistochemical analyses of these tissue specimens for HER2 with the CB1 1 monoclonal antibody against HER2 or with a polyclonal-antibody assay kit and

fluorescence in situ hybridization for HER2 amplification were performed. The interaction between HER2 positivity and the addition of paclitaxel to the treatment was associated with a hazard ratio for recurrence of 0.59 (P=0.01 ) ¹³ . Patients with a HER2+ breast cancer benefited from paclitaxel, regardless of estrogen-receptor status, but paclitaxel did not benefit patients with HER2-negative, estrogen-receptor-positive cancers. Thus, one cannot exclude the possibility that high baseline levels of ERBB2 are also predictive of chemotherapy benefit, or predictive of a synergy effect between the two treatments (i.e. chemo and anti-HER2, single or double), something that NeoALTTO, CALGB40601 and Tryphaena cannot rule out because they did not include patients without chemotheray. Moreover, none of these studies have evaluated the predictive value of the changes in ERBB2 mRNA expression following 2 weeks of treatment. Given that the dual HER2 blockade improves the efficacy of single-agent HER2 therapy, a clinical question that arises is whether the dual blockade may eliminate the need for chemotherapy in a subset of patients. Exclusive dual HER2 blockade has shown high activity in a group of patients with metastatic and primary HER2+ breast cancer ^14"16 . In HER2+ metastatic breast cancer previously treated with trastuzumab, the addition of pertuzumab or lapatinib to trastuzumab achieves higher clinical benefit than either pertuzumab or lapatinib alone ¹⁶ . In primary HER2+ breast cancer, chemotherapy-free neoadjuvant trastuzumab-lapatinib or trastuzumab-pertuzumab combinations achieved pCR rates in the breast of 17-27% ^14,15 . Overall, results suggest that a subset of patients with HER2+ breast cancers is highly sensitive to dual anti HER2 blockade and could potentially be treated without cytotoxic therapy.

A major challenge today is to discover biomarkers that will identify the more sensitive patients to dual HER2 blockade without chemotherapy. To date, hormone receptor- positivity by immunohistochemistry (IHC) is the only molecular biomarker to predict response to dual HER2 blockade without chemotherapy. In the TBCRC006 trial, the pCR rate in estrogen receptor-positive disease was 21 % versus 36% in ER-negative disease following 12 weeks of treatment with lapatinib and trastuzumab (and endocrine therapy if the tumour was ER+) ¹⁵ . In the NeoSphere trial, the pCR rate in estrogen receptor (ER)- positive or progesterone receptor (PR)-positive disease was 5.9% versus 27.3% in ER- negative or PR-negative disease following 12 weeks of treatment with pertuzumab and trastuzumab (Group C) ¹⁴ . However, this biomarker is not enough to identify those patients that will gain the highest benefit from dual HER2 blockade without chemotherapy.

Currently, 30% of patients with HER2-positive (HER2+) breast cancer benefit substantially from dual HER2 blockade without chemotherapy. However, there is a need to identify these patients before and during treatment.

Nowadays, the combination of anti-HER2 doublets (either lapatinib+trastuzumab or pertuzumab+trastuzumab) with optimal multi-agent chemotherapy regimens are providing pCR rates in the range of ~60% ¹⁰ , and pertuzumab has been specifically approved by the FDA and and the EMA for patients with HER2+ early breast cancer with primary tumours > 2 cm or node-positive disease. On the other hand, patients with stage I HER2+ disease, weekly paclitaxel for 12 doses plus single anti-HER2 (i.e. trastuzumab) is considered an acceptable regimen ⁷ . This treatment strategy provides pCR rates ranging from 29% to 46% ¹¹ .

Nowadays, in order to select the more appropriate therapy for the treatment of breast cancer is the hormone receptor status test, a test that tells whether or not the breast cancer cells have receptors for the hormones estrogen and progesterone. A cancer is called estrogen-receptor-positive (or ER+) if more than 1 % of tumor cells express ER by IHC. This suggests that the cancer cells, like normal breast cells, may receive signals from estrogen that could promote their growth. The cancer is progesterone-receptor- positive (PR+) if more than 1 % of tumor cells express ER by IHC. Hormone receptor status test by IHC, however, fails in providing an accurate information of the receptor, in some particular cases of breast cases, which, unfortunately, can cause a physician to take a wrong decision in deciding the more appropriate therapeutic protocol.

In spite of the efforts made, there is the need of biological markers that provide accurate predictive information of the success of a particular therapy prior its administration to the patient diagnosed of breast cancer.

Summary of Invention

The inventors have found that the ERBB2 gene product expression, in particular mRNA levels, when they are quantified in a patient already diagnosed of HER2+ breast cancer, and before receiving any therapy, can provide useful information about the positive or negative response to the administration of anti-HER2 therapy in the absence of chemotherapy (see FIG. 5). From the data provided below, it is remarkable the fact that using ERBB2 as a biomarker, the information provided about the pathological complete response ("pCR") is substantially more accurate when compared with the protocol currently accepted by physicians, which is based on determining the hormone receptor status (see Table 4 below). It is remarkable that the information provided by the ERBB2 biomarker, according to the present invention, is for a population of patients with HER2+ disease that might be candidates to receive anti-HER2 therapy and avoid chemotherapy. This is of great importance because, as it has been pointed out above, ERBB2 can affect the chemotherapy effectiveness and previous studies have not discriminated the effect of ERBB2 biomarker with chemotherapy versus anti-HER2 therapy.

Therefore, the invention means a great advance in accurately predicting, before starting the therapy, how a patient already diagnosed from HER2+ breast cancer could positively respond to anti-HER2 therapy without chemotherapy. This can be of great value for the physician in order to decide the best therapeutic strategy to successfuly overcome the disease. The first aspect of the invention refers to an in vitro method to determine the efficacy of an anti-HER2 therapy in the absence of chemotherapy in a patient with HER2+ breast cancer comprising the detection and/or quantification of a gene expression product of HER2 in an isolated test sample from the patient, before starting the anti-HER2 therapy. In addition to the above, the present inventors have also found that determining the ratio of ERBB2 product gene expression before starting an anti-HER2 therapy and after a time of starting the therapy, it can also helps to predict the efficacy of anti-HER2 treatment in the absence of chemotherapy in a patient already diagnosed with HER2+ breast cancer. As it is shown below, ERBB2 levels determined before and after 15 days of starting the anti-HER2 therapy in the absence of chemotherapy, predicts treatment efficacy compared with hormone receptor status (see Table 6 below). In addition to this, determining ERBB2 gene product expression levels between these two timepoints (i.e. before and after 15 days) during anti-HER2 therapy provides valuable information for deciding whether anti- HER2 treatment should be withdrawn.

Thus, a second aspect of the invention refers to an in vitro method to determine the efficacy of an anti-HER2 therapy in the absence of chemotherapy of a patient with HER2+ breast cancer comprising:

(a) the detection and/or quantification of a gene expression product of HER2 in an isolated biological sample of the patient:

(a.i) before starting the anti-HER2 therapy, and

(a.ii) after the initiation of the anti-HER2 therapy. The results provided herein open the door to further studies in HER2+ breast cancer evaluating the long-term survival outcomes of chemotherapy-free dual HER2 blockade after selecting patients based on ERBB2 mRNA expression levels. With the method of the second aspect of the invention, 64.9%-75.0% pCR rates were observed in the group of patients treated with dual HER2 blockade without chemotherapy with high baseline ERBB2 expression, or high ratio of ERBB2 expression between week 2 and baseline time-points, suggesting that chemotherapy could be avoided in a subset of patients, which represents around -25% (i.e. a quartile) of all HER2+ patients. These pCR rates are currently achieved with multi-agent chemotherapy in combination with dual HER2 blockade if no patient selection is taken into account.

A third aspect of the present invention refers to an in vitro method for deciding or recommending a patient with HER2+ breast cancer whether to initiate an alternative medical regime to an anti-HER2 therapy, that comprises:

(a) the detection and/or quantification of a gene expression product of HER2 in an isolated biological sample of the patient:

(a.i) before starting an anti-HER2 therapy in the absence of chemotherapy, and (a.ii) after the initiation of an anti-HER2 therapy in the absence of chemotherapy.

A fourth aspect of the present invention refers to the use of a gene expression product of HER2 as an in vitro marker for determining the efficacy of anti-HER2 therapy in the absence of chemotherapy in a patient with HER2+ breast cancer, or alternatively as an in vitro marker for deciding or recommending whether to initiate an alternative medical regime to anti-HER2 therapy in a patient with HER2+ breast cancer before receiving an anti-HER2 therapy without chemotherapy.

A fifth aspect of the present invention refers to the use of means for determining the presence or for quantifying the gene expression product of HER2 in the methods of the invention.

Brief Description of Drawings Fig. 1 PAMELA trial schema.

Fig. 2 shows the diagram that resumes the patient information of the PAMELA trial.

Fig. 3 describes the balanced accuracy analyses using variable number of genes

(measured at baseline) and different methods of classification and variable selection, dlda, diagonal linear discriminant analysis; Ida, linear discriminant analysis; qda, quadratic discriminant analysis; rf, random forests; rfe, recursive feature elimination; t.test,

Student ^' s t-test; welch. test, Welch ^' s t-test. Fig. 4 shows the cross-validation area under the curve (AUC) analyses using baseline samples-only after selection of 1 , 2, 3, 4, 5 and 10 genes.

Fig. 5 describes the association of ERBB2 expression with pCR in the entire dataset of baseline samples (n=151 ). A, cross-validation area under the curve (AUC) analysis; B; box-whisker plot of ERBB2 expression in patients that achieved a pCR versus those that did not (non-pCR).

Fig. 6 shows the AUC analysis of ERBB2 expression (measured at baseline, ratio week 2 / baseline or week 2) for predicting pCR in the entire dataset of paired samples (n=144).

Fig. 7 Balanced accuracy analyses in using variable number of genes (ratio of week 2 / baseline) and different methods of classification and variable selection, dlda, diagonal linear discriminant analysis; Ida, linear discriminant analysis; qda, quadratic discriminant analysis; rf, random forests; rfe, recursive feature elimination; t.test, Student ^' s t-test; welch. test, Welch ^' s t-test.

Fig. 8 shows the AUC analyses using week 2 / baseline ratio after selection of 1 , 2, 3, 4, 5 and 10 genes.

Detailed description of the invention

HER2 (HER2-positive) breast cancer is a breast cancer that tests positive for a protein called human epidermal growth factor receptor 2 (HER2). The techniques used by clinical practice to determine the expression of HER2 are well known by the expert in the field, for example by detecting the protein by immunohystochemistry or by detecting the number of copies by Fluorescence in situ Hybridization (FISH), SPoT-Light HER2 CISH test

(Subtraction Probe Technology Chromogenic In Situ Hybridization) or by Inform HER2 Dual ISH test (Inform Dual In Situ Hybridization).

The gene "HER2" ("ERBB2", v-erb-b2 avian erythroblastic leukemia viral oncogene homolog 2) (GenelD: 64) encodes a member of the epidermal growth factor (EGF) receptor family of receptor tyrosine kinases. Amplification and/or overexpression of this gene has been reported in numerous cancers, including breast and ovarian tumors. Gene synonyms are the following CD340; HER-2; HER-2/neu; HER2; MLN 19; NEU; NGL; and TKR1 .

SEQ ID NO: 1 (ERBB2) (NM_001005862.1 , date of 19-JAN-2014) corresponds to the complementary DNA (cDNA) that codifies for the mRNA of the Homo sapiens variant 2. Alternative splicing results in several additional transcript variants, some encoding different isoforms. Allelic variations at amino acid positions have been reported.

The HER2 protein ID is the following: "NP_001005862.1 " (SEQ ID NO: 2).

The target sequence of ERBB2 for the detection and/or quantification in a preferred embodiment is SEQ ID NO: 3

(ACAGACACGTTTGAGTCCATGCCCAATCCCGAGGGCCGGTATACATTCGGCGCCAG CTGTGTGACTGCCTGTCCCTACAACTACCTTTCTACGGACGTGG).

In the present invention, the detection and/or quantification of a gene expression product of HER2 has been performed in patients with HER2+ breast cancer before and during anti-HER2 therapies in the absence of chemotherapy. Therefore, in a preferred embodiment of the methods of the invention the patient in addition has not received any chemotherapy before the detection and/or quantification of the gene expression product of HER2.

In one embodiment of the first aspect of the invention, when the gene expression product of HER2 is overexpressed it is indicative of anti-HER2 efficacy in the absence of chemotherapy. The overexpression is in relation to a reference sample, the reference sample is a normal breast tissue of a healthy person.

In one embodiment of the first aspect of the invention, when the amount of gene expression product of HER2 is highly expressed (defined, for example, as the top 25% percentile, or a ERBB2 gene expression score of≥3.22), it is indicative of high anti-HER2 efficacy in the absence of chemotherapy.

In the present invention the term "gene expression product" refers to the messenger ribonucleic acid (messenger RNA or mRNA) or the protein.

In one embodiment, the gene expression product is mRNA. By "mRNA" it is encompassed both the whole mRNA sequence as well as fragments thereof. In another embodiment, the term "gene expression product" refers to HER2 protein. By "HER2 protein" it is encompassed both the whole HER2 protein of sequence SEQ ID NO: 2, as well as functional fragments thereof (such as immunological fragments thereof) or a protein with a sequence having a percentage of identity of at least 70, 71 , 72, 73, 74, 75, 76, 77, 78, 79, 80, 81 , 82, 83, 84, 85, 86, 87, 88, 89, 90, 91 , 92, 93, 94, 95, 96, 97, 98, 99, or 100 percent sequence identity, preferably 100% identity with SEQ ID NO: 2.

In the present invention the term "identity" refers to the percentage of residues that are identical in the two sequences when the sequences are optimally aligned. If, in the optimal alignment, a position in a first sequence is occupied by the same amino acid residue as the corresponding position in the second sequence, the sequences exhibit identity with respect to that position. The level of identity between two sequences (or "percent sequence identity") is measured as a ratio of the number of identical positions shared by the sequences with respect to the size of the sequences (i.e., percent sequence identity = [number of identical positions/total number of positions] x 100).

A number of mathematical algorithms for rapidly obtaining the optimal alignment and calculating identity between two or more sequences are known and incorporated into a number of available software programs. Examples of such programs include the MATCH- BOX, MULTAIN, GCG, FASTA, and ROBUST programs for amino acid sequence analysis, among others. Preferred software analysis programs include the ALIGN, CLUSTAL W, and BLAST programs (e.g., BLAST 2.1 , BL2SEQ, and later versions thereof). For amino acid sequence analysis, a weight matrix, such as the BLOSUM matrixes (e.g., the BLOSUM45, BLOSUM50, BLOSUM62, and BLOSUM80 matrixes), Gonnet matrixes, or PAM matrixes (e.g., the PAM30, PAM70, PAM120, PAM160, PAM250, and PAM350 matrixes), are used in determining identity. The BLAST programs provide analysis of at least two amino acid sequences, either by aligning a selected sequence against multiple sequences in a database (e.g., GenSeq), or, with BL2SEQ, between two selected sequences. BLAST programs are preferably modified by low complexity filtering programs such as the DUST or SEG programs, which are preferably integrated into the BLAST program operations. If gap existence costs (or gap scores) are used, the gap existence cost preferably is set between about -5 and -15. Similar gap parameters can be used with other programs as appropriate. The BLAST programs and principles underlying them are further described in, e.g., Altschul et al., "Basic local alignment search tool", 1990, J. Mol. Biol, v. 215, pages 403-410. For multiple sequence analysis, the CLUSTAL W program can be used. The CLUSTAL W program desirably is run using "dynamic" (versus "fast") settings. Amino acid sequences are evaluated using a variable set of BLOSUM matrixes depending on the level of identity between the sequences. The CLUSTAL W program and underlying principles of operation are further described in, e.g., Higgins et al., "CLUSTAL V: improved software for multiple sequence alignment", 1992, CABIOS, 8(2), pages 189-191 .

In an embodiment of the present invention, optionally in combination with any of the embodiments provided above or below, the gene expression product is mRNA

(messenger RNA) (in a preferred embodiment is SEQ ID NO: 1 ). In another embodiment, the sequence detected and/or quantified is SEQ ID NO: 3.

In a preferred embodiment of the present invention the product of expression of HER2 is quantified. In a more preferred embodiment the mRNA of HER2 is quantified. In a more preferred embodiment SEQ ID NO: 1 is quantified.

In a preferred embodiment of the present invention the the product of expression of HER2 quantified by an amplification technique. In a more preferred embodiment of the present invention he mRNA of HER2 is quantified using specific primers and/or probes.

The expert in the field knows that adding additional steps to detection techniques quantification can be achieved.

Detection and/ or quantification can be performed by any method known to the skilled person, provided that said method permits the detection or quantification of mRNA in a biological sample. Included among the examples of these procedures are PCR, quantitative real-time PCR (QPCR), multiplex PCR, NASBA, LCR, RT-PCR, RNA sequencing, array hybridization or "Northern" transfer, or combinations of these. In a preferred embodiment, the determination of the mRNA is performed by the nCounter platform (Nanostring Technologies). In most procedures, the use of primers and/or probes are required to detect and/or quantify the mRNA of interest. A skilled person would get easily and directly the sequence of the primers and or probes that can be used from the sequence of the mRNA of HER2.

In most methods of detection and quantification of RNA mentioned above, before performing this procedure it is necessary to convert the RNA to complementary DNA (cDNA). This conversion is accomplished by known techniques by skilled in the art, such as reverse transcription, among others.

In one embodiment of any of the methods provided by the present invention, the detection and/or quantification of the gene expression product of HER2 after the initiation of an anti- HER2 therapy is performed at a day from the 5 ^th to the 20 ^th day (5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14, 15, 16, 17, 18, 19 or 20) after the initiation of the anti-HER2 therapy. In another embodiment, the detection and/or quantification of the gene expression product of HER2 after the initiation of an anti-HER2 therapy is performed at a day from the 5 ^th to the 19 ^th day, more preferably from the 10 ^th to the 16 ^th day. In another embodiment, the detection and/or quantification of the gene expression product of HER2 after the initiation of an anti- HER2 therapy is performed at day 14 ^th after the initiation of the anti-HER2 therapy.

In a preferred embodiment of the methods of the invention the gene expression product of HER2 is mRNA and it is quantified by at least a pair of primers and/or probes. In a preferred embodiment of the present invention the probe detects SEQ ID NO: 3, in a particular embodiment two probes detect SEQ ID NO: 3.

In the present invention the pathological complete response (pCR) is the absence of invasive neoplastic cells at microscopic examination of the primary tumour at surgery after a treatment HER2+ breast cancer by, preferably a dual HER2 blokade, more preferably with lapatinib and trastuzumab, has been completed.

In the present invention the term "after initiation of an anti-HER2 therapy" means that the subject has already received the treatment or that is receiving said treatment (ongoing treatment).

The anti-HER2 therapy in the present invention is not given in combination with chemotherapy. Therefore the anti-HER2 therapy is given to the patient in the absence of chemotherapy (without chemotherapy).

Known anti-HER2 therapies (treatment) include trastuzumab (Herceptin®), lapatinib (Tykerb®), neratinib (HKI-272), pertuzumab (Perjeta®) and ado-trastuzumab emtansine (Kadcyla®). In one embodiment of the in vitro methods provided by the present invention, the anti-HER2 therapy is selected from the list consisting of: trastuzumab, lapatinib, neratinib, pertuzumab and/or ado-trastuzumab emtansine, or any combinations thereof. Preferably is trastuzumab and lapatinib.

Therefore, in the case the patient receives trastuzumab and lapatinib, the method determines that said medical regime is effective when the gene expression of HER2, preferably by quantifying and/or detecting the mRNA, after the initiation of said therapy is descreased in comparison to the basal expression (before receiving said therapy). Thus, the treatment outocome of said patient is good. On the contrary, when said comparison shows that there is not a decrease in gene expression, then said medical regime is less effective or ineffective. Thus the treatment outcome of said patient is bad. In that case, the method of the present invention is useful for deciding or recommending to change said medical regime and in particular to initiate another treatment, and therefore is useful for determining the best therapeutic regime for a given patient with HER2+ breast cancer. Chemotherapy (cytotoxic therapy) that could be used as said medical regime would be paclitaxel, docetaxel, carboplatin, doxorubicin, epirubicin, nab-paclitaxel, vinorelbine, capecitabine and eribulin.

In the present invention the term "efficacy" is related to the pCR of the HER2+ breast cancer, therefore the absence of invasive neoplastic cells at microscopic examination of the primary tumour at surgery is indicative that the treatment has been effective.

In a preferred embodiment of the invention the efficacy is pCR. The efficacy can also be observed as any decrease in tumor size wherein imaging techniques are used.

The term "biological sample" includes, without being limited thereto, biological tissues and/or fluids from an individual, obtained by any method designed for that purpose known to persons skilled in the art. The biological sample comprises the product of expression of the gene that codifies for HER2.

In an embodiment of the in vitro methods provided by the present invention the sample is a breast tissue, blood, serum or plasma. In a preferred embodiment is a biopsy sample from breast cancer tissue. In the present invention, the biological sample is fresh, frozen, fixed or fixed and embedded in paraffin. In a preferred embodiment, the sample is a breast cancer tissue fixed and embedded in paraffin. The biological sample can be collected by any means known by the expert in the field, for example by needle biopsy of the breast.

In the present invention the terms "patient", "subject" and "individual" are used

interchangeably.

In the present invention the patient is a mammal, such as a mouse, rat, guinea pig, rabbit, dog, cat, bovine, horse, goat, sheep, primate or human, preferably is a human, more preferably is a woman. The patient can be of any age, gender or race.

In another preferred embodiment of the first, second, and third in vitro methods of the present invention the patient is a woman. In the present invention, the patient has not received any previous cancer therapy (nor chemotherapy) before the initiation of the anti-HER2 therapy. In another preferred embodiment of the in vitro methods of the present invention the anti- HER2 therapy is combined with endocrine therapy in hormone receptor-positive (HR+) patients.

The patient can be also a hormone receptor-negative (HR-) patient.

Endocrine therapy known by the expert in the field is for example: selective estrogen- receptor response modulators (SERMs) (for example tamoxifen or toremifene), aromatase inhibitors (for example anastrozole, exemestane, letrozole), estrogen-receptor

downregulators (ERDs) (for example fulvestrant) and luteinizing hormone-releasing hormone agents (LHRHs) (for example goserelin, leuprolide and Triptorelin). In a preferred embodiment of the methods and uses of the present invention the endocrine therapy is selected form list consisting of: a selective estrogen-receptor response modulator, an aromatase inhibitor, an estrogen-receptor downregulators (ERDs) and/or a luteinizing hormone-releasing hormone agent, or any combination thereof. In a more preferred embodiment the endocrine therapy is selected form the list consisting of:

tamoxifen, toremifene, anastrozole, exemestane, letrozole, fulvestrant, goserelin, leuprolide and/or Triptorelin, or any combinations thereof. In a more preferred

embodiment, is letrozole or tamoxifen. Thus a preferred embodiment of the methods of the invention refers to a method wherein the anti-HER2 therapy is trastuzumab and lapatinib. In a more preferred embodiment the patient in addition has not received chemotherapy. In a more preferred embodiment the gene expression product of HER2 is mRNA, in a preferred embodiment is SEQ ID NO: 1. In a preferred embodiment the patient is a HR- patient. In another preferred embodiment the patient is a HR+ patient, and the HER2-therapy is combined with letrozole or tamoxifen. In relation to the methods of the second and third aspect of the invention, in addition in a preferred embodiment the detection and/or quantification of the mRNA after the initiation of an anti-HER2 therapy in absence of chemotherapy is performed at day 14 after the initiation of the anti-HER2 therapy.

Thus a preferred embodiment of the methods of the invention refers to a method wherein the anti-HER2 therapy is trastuzumab and lapatinib; the patient is a HR- patient; the gene expression product of HER2 is mRNA, and in a preferred embodiment is SEQ ID NO: 1. In relation to the methods of the second and third aspect of the invention, in addition in a preferred embodiment the detection and/or quantification of the mRNA after the initiation of an anti-HER2 therapy in absence of chemotherapy is performed at day 14 after the initiation of the anti-HER2 therapy. In a preferred embodiment in addition the patient has not received chemotherapy previously to the HER2 detection/quantification.

In a preferred embodiment of the methods of the invention refers to a method wherein the anti-HER2 therapy is trastuzumab and lapatinib; the patient is a HR+ patient; the HER2- therapy is combined with letrozole or tamoxifen; the gene expression product of HER2 is mRNA, and in a preferred embodiment is SEQ ID NO: 1 . In relation to the methods of the second and third aspect of the invention, in addition in a preferred embodiment the detection and/or quantification of the mRNA after the initiation of an anti-HER2 therapy in absence of chemotherapy is performed at day 14 after the initiation of the anti-HER2 therapy. In a preferred embodiment in addition the patient has not received chemotherapy previously to the HER2 detection/quantification.

Thus a preferred embodiment of the third aspect of the invention the anti-HER2 therapy is trastuzumab and lapatinib; the patient is a HR- patient; the gene expression product of HER2 is mRNA, and in a more preferred embodiment is SEQ ID NO: 1 ; the detection and/or quantification of the mRNA after the initiation of an anti-HER2 therapy in absence of chemotherapy is performed at day 14 after the initiation of the anti-HER2 therapy; and the alternative regime is chemotherapy. In a preferred embodiment in addition the patient has not received chemotherapy previously to the HER2 detection/quantification.

In a preferred embodiment of the third aspect of the invention the anti-HER2 therapy is trastuzumab and lapatinib; the patient is a HR+ patient; the HER2-therapy is combined with letrozole or tamoxifen; the gene expression product of HER2 is mRNA, in a more preferred embodiment is SEQ ID NO: 1 ; the detection and/or quantification of the mRNA after the initiation of an anti-HER2 therapy in absence of chemotherapy is performed at day 14 after the initiation of the anti-HER2 therapy; and the alternative regime is chemotherapy. In a preferred embodiment in addition the patient has not received chemotherapy previously to the HER2 detection/quantification.

In a preferred embodiment of the of the in vitro methods of the present invention, the method also comprises imaging the subject for breast cancer, for example by ultrasound. The imaging can be performed in any order in the method of the invention, therefore, before detecting and/or measuring the gene expression product of HER2. The reduction of cancer size is associated with efficacy of the HER2-therapy.

The present invention also refers to a method for determining efficacy and treatment of HER2+ breast cancer in a subject diagnosed with the disease, said method comprising the steps of:

a) obtaining a first sample comprising a breast cancer tissue biopsy sample from the subject before the beginning of an anti-HER2 therapy;

b) contacting the first sample with a first reagent, preferably a probe, that binds to the mRNA of HER2;

c) measuring an amount of mRNA of HER2 that is bound to the first reagent in the first sample;

d) comparing the amount of mRNA of HER2 bound to the first reagent in step c) with the mRNA of HER2 obtained from a second sample comprising a breast cancer tissue biopsy sample from the subject after the initiation of an anti-HER2 therapy;

e) determining the treatment outcome for the subject and treating the subject, wherein:

(i) if the amount of mRNA of HER2 bound to the first reagent in step c) is higher than the one of mRNA of HER2 value on the second sample, the anti-HER2 treatment is more or highly effective; and

(ii) if the amount of mRNA bound to the first reagent in step c) is lower than the one of the mRNA of HER2 value on the second sample, the anti-HER2 treatment is less effective or ineffective, and the treatment is selected from the group consisting of: breast cancer removal, follow-up, chemotherapy, radiotherapy, and combinations thereof.

In an embodiment of the fourth aspect of the present invention, the gene expression product of HER2 is mRNA. More preferably wherein the gene expression product is SEQ ID NO: 1 .

A preferred embodiment of the fourth aspect of the present invention is referred to the use wherein the gene expression product is the protein, in a preferred embodiment is SEQ ID NO: 2. A preferred embodiment of the fourth aspect of the present invention is referred to the use wherein the anti-HER2 therapy is selected from the group consisting of: trastuzumab, lapatinib, neratinib, pertuzumab, ado-trastuzumab emtansine, or a combination thereof, preferably is trastuzumab and lapatinib. In preferred embodiment of the fourth aspect of the present invention the patient is a woman, preferably is a hormone receptor-positive (HR+) patient. The patient is a HR+ patient or a receptor-negative (HR-) patient. Wherein the patient is a HR+ patient, the anti- HER2 therapy can be combined with endocrine therapy. In a more preferred embodiment the endocrine therapy is selected form the list consisting of: tamoxifen, toremifene, anastrozole, exemestane, letrozole, fulvestrant, goserelin, leuprolide and/or Triptorelin, or any combinations thereof. In a more preferred embodiment, is letrozole or tamoxifen.

In a preferred embodiment of the fourth aspect of the present invention the anti-HER2 therapy is trastuzumab and lapatinib; the patient is a HR- patient; the gene expression product of HER2 is mRNA, and in a preferred embodiment is SEQ ID NO: 1. In a preferred embodiment in addition the patient has not received chemotherapy previously to the HER2 detection/quantification. In a preferred embodiment of the fourth aspect of the present invention the anti-HER2 therapy is trastuzumab and lapatinib; the patient is a HR+ patient; the HER2-therapy is combined with letrozole or tamoxifen; the gene expression product of HER2 is mRNA, and in a preferred embodiment is SEQ ID NO: 1. In a preferred embodiment in addition the patient has not received chemotherapy previously to the HER2 detection/quantification.

In a preferred embodiment of the fifth aspect of the present invention the gene expression product is mRNA, preferably SEQ ID NO: 1 ; or protein, preferably SEQ ID NO: 2.

In a preferred embodiment of the fifth aspect of the present invention the means form part of a kit.

Another aspect of the present invention is referred to a kit that comprises the specific means to detect the presence or absence of or quantify a gene expression product of HER2 , preferably its mRNA, for use in the methods of the present invention. In a particular embodiment the kit comprises specific primers and/or probes, antibodies, or combinations thereof. In a particular embodiment the kit comprises specific primers and/or probes for detecting and/or quantifing SEQ ID NO: 1 , in a more particular embodiment for detecting and/or quantifing SEQ ID NO: 3. Throughout the description and claims the word "comprise" and variations of the word, are not intended to exclude other technical features, additives, components, or steps.

Furthermore, the word "comprise" encompasses the case of "consisting of". Additional objects, advantages and features of the invention will become apparent to those skilled in the art upon examination of the description or may be learned by practice of the invention. The following examples and drawings are provided by way of illustration, and they are not intended to be limiting of the present invention. Reference signs related to drawings and placed in parentheses in a claim, are solely for attempting to increase the intelligibility of the claim, and shall not be construed as limiting the scope of the claim. Furthermore, the present invention covers all possible combinations of particular and preferred embodiments described herein.

Examples Example 1 : ERBB2 as a predictor of response of dual HER2 blockade in the absence of cytotoxic therapy

Material and methods: Study design and patients:

PAMELA (NCT01973660) is a nonrandomised, multicentre, prospective, open-label phase 2 study in women with HER2+ breast cancer (Figure 1 ). All eligible patients had centrally confirmed HER2, centrally performed estrogen receptor and progesterone receptor by immunohistochemistry, stage l-IIIA breast cancer with primary tumours larger than 1 cm in diameter, were aged 18 years or older, and had not received any previous cancer therapy. Tumours had to be HER2 immunohistochemistry 3+ or 2+ and positive for chromogenic in-situ hybridisation. Of note, HER2, ER, and PR testing were done under IS015189 accreditation.

Other main inclusion criteria were: baseline Eastern Cooperative Oncology Group

(ECOG) performance status of 0-2, baseline left ventricular ejection fraction (LVEF) of 50% or more, as measured by echocardiography or multiple gated acquisition (MUGA). Key exclusion criteria were multicentric tumours, inoperable stage III disease, stage IV disease, bilateral breast cancer, other malignancies, inadequate bone marrow or renal function, impaired liver function, impaired cardiac function, uncontrolled hypertension, pregnancy, and refusal to use contraception.

The study was undertaken in accordance with Good Clinical Practice guidelines and the World Medical Association Declaration of Helsinki. All patients provided written informed consent. Approvals for the study protocol were obtained from independent ethics committees.

Procedures:

Lapatinib was given orally at a daily dose of 1000 mg. Trastuzumab was given IV every 3 weeks at a loading dose of 8 mg/kg, followed by 6 mg/kg. Patients with HR+ received letrozole (2.5 mg daily) or tamoxifen (20 mg daily) according to menopausal status. The total duration of treatment was 18 weeks. At week 2, a core-needle biopsy was mandatory. At week 6, an early response evaluation by ultrasound was mandatory. Any increase in tumour size during the study or at week 6 was considered a treatment failure, and the patient would be categorized as not sensitive for the primary endpoint (i.e. pCR with dual blockade). These patients were treated with trastuzumab and weekly paclitaxel 80 mg/m2 for 12 doses and lapatinib 750 mg orally. Surgery was performed between 1 and 3 weeks after the last dose of dual HER2 blockade, or 2 and 3 weeks after the last dose of paclitaxel. Standard adjuvant chemotherapy was administered according to the physician ^' s discretion. Gene expression analysis:

A section of the formalin-fixed paraffin-embedded (FFPE) breast tissue was first examined with haematoxylin and eosin staining to confirm presence of invasive tumour cells (≥10%) and determine the minimum tumour surface area (>4 mm2). Patients could not be recruited unless the minimum tissue requirement for gene expression analysis was met. For samples at day-15, those without invasive tumour cells were also profiled. For RNA purification (Roche® High Pure FFPET RNA isolation kit), >1 -5 10 μηη FFPE slides were used for each tumour specimen, and macrodissection was performed, when needed, to avoid normal contamination. A minimum of ~100 ng of total RNA was used to measure the expression of the 555 breast cancer-selected genes and 5 housekeeping genes (ACTB, MRPL19, PSMC4, RPLP0, and SF3A1 ) using the nCounter platform (Nanostring

Technologies, Seattle, WA, US). Data were log base 2 transformed. The geometric mean of the 5 housekeeping genes was obtained for each sample, and was used as a normalization factor for each gene in each sample. The design of the 560-CodeSet, including the target sequences, can be found in table 8.

Statistical analysis:

The primary endpoint was pCR in the breast, which is defined as the absence of invasive neoplastic cells at microscopic examination of the primary tumour at surgery. Remaining in situ lesions were allowed.

RESULTS: The PAMELA clinical trial:

From October 2013 to December 2015, 151 patients were recruited across 19 sites in Spain. Of 151 recruited patients, 137 patients completed treatment as planned and 14 patients discontinued treatment (Figure 2). The baseline median tumour size by clinical breast examination was 2.4 cm, and most patients had negative axilla (64.9%) and were postmenopausal (59.6%) (Table 1 ). Among patients with HR+ disease (n=77), 52% and 48% received tamoxifen and letrozole respectively, accordingly with menopausal status. All patients who underwent surgery had a valid assessment of pathological response. Table 1. Patient demographics at baseline.

A pCR in the breast was noted in 46 of 151 women (30.5%, 95% CI 23.4-38.5). Consistent with previous findings, fewer pCRs were noted in tumours that were HR+ compared to those HR-negative (18.2% vs 43.2%; p=0.001 ). Among 14 patients who discontinued treatment, 6 had treatment failure (4.0% of all patients). Treatment failure occurred in HR+ (n=2) and HR-negative (n=4) disease. Five patients out of 6 with treatment failure received neoadjuvant paclitaxel, lapatinib and trastuzumab as per protocol and none achieved a pCR.

Among the different clinical-pathological variables evaluated (age, tumour size, tumour stage, menopausal status, nodal status and hormone receptor [HR] status), only HR status was found significantly associated with pCR (Table 2). Table 2. Logistic regression model analyses of treatment pathological response.

OR, odds ratio.

Prediction of pCR with gene expression from baseline samples:

Expression of 555 breast cancer-related genes and 5 house-keeping genes was performed successfully in all baseline samples (n=151 ) (see table 8). Cross validation analyses (10-fold, repeated 25 times) using 4 methods of variable selection (t-test,

Welch ^' s t-test, random forests [rf] and recursive feature elimination [rfe]), different number of selected genes (1 , 2, 3, 4, 5, 10, 12, 15, 20 and 30) and 3 classification methods (diagonal linear discriminant analysis [dlda], linear discriminant analysis [Ida] and quadratic discriminant analysis [qda]) were performed to select the best model. As shown in Figure 3 and Table 3, the best ^' balanced accuracy ^' , a measure of classification performance, was obtained with a single gene, which was ERBB2 in all the analyses performed.

Table 3. Genes selected during 10-fold cross-validation using different methods of classification and gene selection.

Variable

Class. Selection Balanced method Method N Gene Misclass. sensitivity specificity accuracy QDA t.test 1 ERBB2 24.4% 51.4% 86.2% 68.8%

QDA9 welch .test 1 ERBB2 24.4% 51.4% 86.2% 68.8%

DLDA14 welch .test 20 - 34.3% 71.8% 63.1 % 67.4%

DLDA13 welch .test 12 - 34.2% 70.3% 63.8% 67.1 %

LDA t.test 1 ERBB2 24.0% 42.7% 90.6% 66.7%

LDA9 welch .test 1 ERBB2 24.0% 42.7% 90.6% 66.7%

DLDA19 rfe 10 - 30.1 % 57.2% 75.5% 66.3%

DLDA12 welch .test 10 - 34.5% 68.0% 64.4% 66.2%

DLDA1 1 welch .test 5 - 33.1 % 64.3% 68.1 % 66.2%

DLDA5 t.test 10 - 34.7% 66.6% 64.7% 65.6%

DLDA9 welch .test 3 - 31.2% 56.7% 74.0% 65.4%

QDA10 welch .test 2 - 31.1 % 55.9% 74.6% 65.3%

"Class. Method": Classification method; N: "Number of genes selected"; "Misclass.":

Misclassification

Next, it was evaluated, using cross-validation analyses (10-fold, repeated 25 times) and the qda method, the prediction performance using Receiver Operating Characteristic (ROC) analysis (i.e. area under the ROC [auROC]curve) when 1 , 2, 3, 4, 5 and 10 genes were selected. As shown in Figure 4, 1 single gene, which was ERBB2 in all cases, showed the highest auROCs. Prediction of pCR with ERBB2 expression from baseline samples:

Overall, this data suggested that among the 555 breast cancer-related genes, ERBB2 was the most robust gene to predict response following dual HER2 blockade without chemotherapy. Then the ability of ERBB2 expression to predict pCR in the entire dataset of 151 patients with baseline tumour samples was explored. Firstly, it was estimated the performance of ERBB2 for predicting pCR (Figure 5). The results revealed an AUC of 0.804.

Secondly, it was evaluated the expression of ERBB2 in patients that achieved a pCR versus those that did not (non-pCR) (Figure 5). The results revealed that the median expression of ERBB2 in the pCR group was 3.24, and the median expression of ERBB2 in the non-pCR group was 1 .83. The difference was 1.42, which is equivalent to a 2.68-fold difference. Thirdly, were explored the pCR rates according to ERBB2 expression. Using tertiles (cutoffs of ERBB2 score of 2.93 and 1.61 ), the pCR rate in the highest,

intermediate and lowest tertiles were 58.8%, 24% and 8%, respectively. Using quartiles (cutoffs of ERBB2 score of 3.21 , 2.45 and 0.97), the pCR rate in the highest, intermediate (the 2 intermediate quartiles combined into 1 group) and lowest quartiles were 64.9%, 23.7% and 10.5%, respectively.

Ability of ERBB2 at baseline to predict pCR compared to HR status:

5

HR status was the only molecular predictor to date to predict pCR following dual HER2 blockade in the absence of cytotoxic therapy. Here the ability of ERBB2 to predict pCR beyond HR status was evaluated. In a bivariate logistic regression model that includes HR status and ERBB2 expression (Table 4), it was observed that ERBB2 remains significantly0 associated with pCR whereas HR status loses its statistical significance. This results suggest that ERBB2 provides more predictive information than HR status.

Table 4. Association of ERBB2 baseline and HR with pCR.

Univariate Bivariate

Breast Lower Upper , Lower Upper

Signatures N _pCR ^ OR _{9g % 9g %} p-valne OR _{9g % 9g %} p-valne

ERBB2 baseline 151 NA 2.62 1.8 3.9 O.001 1.6 3.7 O.001

HR-negative 74 43.2% 3.42 1.6 7.2 0.001 0.7 3.9 0.224 5

Prediction of pCR with ERBB2 expression at baseline, week 2 and ratio between week 2 and baseline:

A total of 144 paired samples were available in PAMELA from the 151 patients recruited.0 This represents 95% of all available samples. Thus, this paired dataset allowed to

compare the predictive ability of ERBB2 expression measured at baseline, at week 2 and the ratio of ERBB2 expression between the 2 time-points. To compare performances, the AUCs between the three biomarkers were compared (Figure 6). The results revealed that the ratio of ERBB2 expression between week 2 and baseline time-points was the best5 predictor of pCR (Figure 6) with an AUC of 0.878.

Secondly, the pCR rates according to the ratio of ERBB2 expression between week 2 and baseline time-points were explored. Using tertiles (cutoffs of ERBB2 ratio score of -3.04 and -0.35), the pCR rate in the lowest, intermediate and highest tertiles were 64.6%, 25%0 and 2%, respectively. Using quartiles (cutoffs of ERBB2 ratio score of -3.88, -1.37 and 0.009), the pCR rate in the lowest, intermediate (the 2 intermediate quartiles combined into 1 group) and highest quartiles were 75%, 22.2% and 2.7%, respectively. Overall, this data suggested that the best predictor of pCR was the ratio of ERBB2 expression between week 2 and baseline time-points. However, it was unclear if this can be improved by the addition of genes. Thus, using the 555 breast cancer-related genes, we evaluated the best ratio of gene expression to predict pCR. To do so, was calculated the ratio of expression between week 2 and baseline time-points (i.e. week 2 / baseline) for each gene. Similar to the previous analysis with baseline samples-only, we performed cross validation analyses (10-fold, repeated 25 times) using 4 methods of variable selection (t-test, Welch ^' s t-test, random forests [rf] and recursive feature elimination [rfe]), different number of selected genes (1 , 2, 3, 4, 5, 10, 12, 15, 20 and 30) and 3

classification methods (diagonal linear discriminant analysis [dlda], linear discriminant analysis [Ida] and quadratic discriminant analysis [qda]). As shown in Figure 7 and Table 5, similar ^' balanced accuracies ^' were obtained with different number of genes. Of note, when using different methods of classification and variable selection, ERBB2 was found the top gene associated with pCR. These results suggested that not much prediction performance is to be gained by the addition of new genes beyond ERBB2.

Table 5. Genes selected during 10-fold cross-validation using different methods of classification and gene selection.

"Class. Method": Classification method; "Variable Sel. Method": Variable Selection Method; " N. genes Sel.": Number of genes selected; "M.": misclassification; "B. acc": Balanced accuracy

Furthermore, it was evaluated, using cross-validation analyses (10-fold, repeated 25 times) and the qda method, the prediction performance using Receiver Operating Characteristic (ROC) analysis (i.e. area under the ROC [auROC]curve) when 1 , 2, 3, 4, 5 and 10 genes were selected. As shown in Figure 8, 1 single gene, which was ERBB2 in all cases, showed one of the the highest AUC. Indeed, a 2-gene model, which included ERBB2 and GRB7 (i.e. k=2), although they showed a numerically higher AUC, it did not significantly improve the AUC compared to ERBB2-alone. Overall, this data suggested that among the 555 breast cancer-related genes, the ratio of ERBB2 expression between week 2 and baseline time-points was the most robust to predict response following dual HER2 blockade without chemotherapy.

Ability of ERBB2 ratio to predict pCR compared to HR status:

HR status was the only molecular predictor to date to predict pCR following dual HER2 0 blockade in the absence of cytotoxic therapy. Here the ability of ERBB2 ratio to predict pCR beyond HR status was evaluated. In a bivariate logistic regression model that includes HR status and ERBB2 ratio (Table 6), it was observed that ERBB2 ratio remains significantly associated with pCR whereas HR status loses its statistical significance. This results suggested that ERBB2 ratio provides more predictive information than HR status. 5

Table 6. Association of ERBB2 ratio and HR with pCR.

Univariate Bivariate

Breast Lower Upper Lower Upper

Signatures N OR p-value

pCR rate ^OR 95% 95% ^p - ^Value 95% 95%

ERBB2 ratio 151 NA 0.49 0.4 0.6 O.001 0.51 0.4 0.6 O.001

Ability of ERBB2 ratio to predict pCR compared to ERBB2 baseline 0 Here it was compared the ability of ERBB2 ratio to predict pCR compared to ERBB2

baseline in the 144 paired samples. In a bivariate logistic regression model that includes ERBB2 baseline and ratio (Table 7), it was observed that ERBB2 ratio remains significantly associated with pCR whereas ERBB2 baseline loses its statistical significance. This results suggest that ERBB2 ratio provides more predictive information 5 than ERBB2 baseline, which is concordant with the previous AUC results.

Table 7. Association of ERBB2 ratio and ERBB2 baseline with pCR.

Univariate Bivariate

Breast Lower Upper Lower Upper

Signatures N OR p-value OR p-value pCR rate 95% 95% 95% 95%

ERBB2 baseline 151 NA 2.56 1.7 3.8 O.001 1.39 0.9 2.2 0.149

ERBB2 ratio 151 TSFA 0.49 0.4 0.6 <0.001 0.54 0.4 0.7 «0.001

CONCLUSIONS: In this study, it has been shown that ERBB2 expression alone is the best predictor of pCR following dual HER2 blockade without chemotherapy. This biomarker can be evaluated either at baseline, at week 2 of treatment, or both. These results suggest that the predictive ability of baseline ERBB2 expression is similar to week 2 ERBB2 expression; however, combination of ERBB2 expression data coming from both time-points (i.e.

ERBB2 ratio) is the best predictor among the three. Thus, from a clinical perspective, ERBB2 expression could be used either at baseline-only (i.e. before starting therapy) or at both time-points (i.e. ERBB2 ratio) if a biopsy at week 2 is available. Either way, both predictors can identify -25% (top quartile) of patients with HER2+ disease that will achieve a pCR in 64.9-75% of the cases if treated with dual HER2 blockade without chemotherapy. Importantly, ERBB2 at baseline, or ERBB2 ratio, provide independent and more information compared to HR status, which is the only molecular predictor to date consistently found associated with pCR in HER2+ breast cancer following dual HER2 blockade without chemotherapy.

References:

1 . Wolff AC, Hammond MEH, Hicks DG, et al. Recommendations for Human

Epidermal Growth Factor Receptor 2 Testing in Breast Cancer: American Society of

Clinical Oncology/College of American Pathologists Clinical Practice Guideline Update. Journal of Clinical Oncology 2013; 31 (31 ): 3997-4013.

2. Slamon D, Eiermann W, Robert N, et al. Adjuvant Trastuzumab in HER2-Positive Breast Cancer. New England Journal of Medicine 201 1 ; 365(14): 1273-83.

3. Blackwell KL, Burstein HJ, Storniolo AM, et al. Overall Survival Benefit With Lapatinib in Combination With Trastuzumab for Patients With Human Epidermal Growth Factor Receptor 2-Positive Metastatic Breast Cancer: Final Results From the EGF104900 Study. Journal of Clinical Oncology 2012; 30(21 ): 2585-92.

4. Baselga J, Cortes J, Kim S-B, et al. Pertuzumab plus Trastuzumab plus

Docetaxel for Metastatic Breast Cancer. New England Journal of Medicine 2012; 366(2): 109-19.

5. Cortazar P, Zhang L, Untch M, et al. Pathological complete response and long- term clinical benefit in breast cancer: the CTNeoBC pooled analysis. The Lancet;

384(9938): 164-72. 6. Piccart-Gebhart M, Holmes E, Baselga J, et al. Adjuvant Lapatinib and

Trastuzumab for Early Human Epidermal Growth Factor Receptor 2-Positive Breast Cancer: Results From the Randomized Phase III Adjuvant Lapatinib and/or Trastuzumab Treatment Optimization Trial. Journal of Clinical Oncology 2016; 34(10): 1034-42.

7. Tolaney SM, Barry WT, Dang CT, et al. Adjuvant Paclitaxel and Trastuzumab for Node-Negative, HER2-Positive Breast Cancer. New England Journal of Medicine 2015; 372(2): 134-41 . 8. Fumagalli D, Venet D, Ignatiadis M, et al. Rna sequencing to predict response to neoadjuvant anti-her2 therapy: A secondary analysis of the neoaltto randomized clinical trial. JAMA Oncology 2016.

9. Carey LA, Berry DA, Cirrincione CT, et al. Molecular Heterogeneity and Response to Neoadjuvant Human Epidermal Growth Factor Receptor 2 Targeting in CALGB 40601 , a Randomized Phase III Trial of Paclitaxel Plus Trastuzumab With or Without Lapatinib. Journal of Clinical Oncology 2015.

10. Schneeweiss A, Chia S, Hickish T, et al. Pertuzumab plus trastuzumab in combination with standard neoadjuvant anthracycline-containing and anthracycline-free chemotherapy regimens in patients with HER2-positive early breast cancer: a randomized phase II cardiac safety study (TRYPHAENA). Annals of Oncology 2013; 24(9): 2278-84.

1 1 . Baselga J, Bradbury I, Eidtmann H, et al. Lapatinib with trastuzumab for HER2- positive early breast cancer (NeoALTTO): a randomised, open-label, multicentre, phase 3 trial. The Lancet 2012; 379(9816): 633-40.

12. Scaltriti M, Nuciforo P, Bradbury I, et al. High HER2 Expression Correlates with Response to the Combination of Lapatinib and Trastuzumab. Clinical Cancer Research 2015; 21 (3): 569-76.

13. Hayes DF, Thor AD, Dressier LG, et al. HER2 and Response to Paclitaxel in Node-Positive Breast Cancer. New England Journal of Medicine 2007; 357(15): 1496-506. 14. Gianni L, Pienkowski T, Im Y-H, et al. Efficacy and safety of neoadjuvant pertuzumab and trastuzumab in women with locally advanced, inflammatory, or early HER2-positive breast cancer (NeoSphere): a randomised multicentre, open-label, phase 2 trial. The Lancet Oncology 2012; 13(1 ): 25-32. 15. Rimawi MF, Mayer IA, Forero A, et al. Multicenter Phase II Study of Neoadjuvant Lapatinib and Trastuzumab With Hormonal Therapy and Without Chemotherapy in Patients With Human Epidermal Growth Factor Receptor 2-Overexpressing Breast Cancer: TBCRC 006. Journal of Clinical Oncology 2013; 31 (14): 1726-31 .

16. Cortes J, Fumoleau P, Bianchi GV, et al. Pertuzumab Monotherapy After Trastuzumab-Based Treatment and Subsequent Reintroduction of Trastuzumab: Activity and Tolerability in Patients With Advanced Human Epidermal Growth Factor Receptor 2- Positive Breast Cancer. Journal of Clinical Oncology 2012; 30(14): 1594-600.

Table 8: nCounter™ CodeSet Design for the 560 genes

Gene Accession T. Region Gene Accession T. Region

A1CF NM_014576.2 1866-1965 BAG1 NM_004323.3 1491-1590

AARS NM_001605.2 836-935 BCL11A NM_018014.2 3781-3880

ABAT NM_000663.4 3336-3435 BCL2 NM_000633.2 1526-1625

ABCB1 NM_000927.3 3911-4010 BCL2A1 NM_001114735.1 1-100

ABCC3 NM_001144070.1 461-560 BDNF NM_001143805.1 436-535

ABCC8 NM_000352.3 481-580 BIRC5 NM_001012270.1 1096-1195

ACOT4 NM_152331.3 733-832 BLM NM_000057.2 2136-2235

ACTB NM_001101.2 1011-1110 BLVRA NM_000712.3 926-1025

ACTL8 NM_030812.1 1611-1710 BMI1 NM_005180.5 1146-1245

ACTR3B NM_001040135.1 276-375 BOP1 NM_015201.3 204-303

ADM NM_001124.1 1301-1400 BRAF NM_004333.3 566-665

ADRA2A NM_000681.2 2491-2590 BRCA1 NM_007294.2 631-730

ADRA2C NM_000683.3 1366-1465 BRCA2 NM_000059.3 116-215

AFF3 NM_001025108.1 4881-4980 BTG2 NM_006763.2 1701-1800

AGR2 NM_006408.2 1366-1465 BTG3 NM_001130914.1 876-975

AGR3 NM_176813.3 1-100 BUB1 NM_004336.2 101-200

AHCYL1 NM_006621.4 2436-2535 BYSL NM_004053.3 1081-1180

AKT1 NM_001014431.1 2008-2107 C11orf30 NM_020193.3 2226-2325

AKT3 NM_005465.3 2001-2100 C16orf45 NM_001142469.1 1621-1720

ALDH1A1 NM_000689.3 12-111 C1orf106 NM_001142569.1 2561-2660

ANGPTL4 NM_001039667.1 1137-1236 C1orf21 NM_030806.3 8911-9010

ANLN NM_018685.2 1901-2000 C4orf32 NM_152400.1 181-280

ANXA1 NM_000700.1 516-615 C8orf33 NM_023080.2 1366-1465

ANXA8L2 NM_001630.2 1351-1450 CA12 NM_001218.3 2446-2545

APC NM_000038.3 6851-6950 CABP7 NM_182527.2 2741-2840

APH1B NM_001145646.1 2931-3030 CAMK2N1 NM_018584.5 86-185

AR NM_000044.2 876-975 CAND1 NM_018448.3 2821-2920

ARAF NM_001654.1 1021-1120 CAPN13 NM_144575.2 1266-1365

AREG NM_001657.2 548-647 CAPN6 NM_014289.2 1131-1230

ASF1A NM_014034.2 351-450 CAV1 NM_001753.3 435-534

C12orf11 NM_018164.2 906-1005 CBX7 NM_175709.3 2556-2655

ATAD2 NM_014109.3 1036-1135 CCDC86 NM_024098.3 1461-1560

ATAD3A NM_001170535.1 501-600 CCNA2 NM_001237.2 1211-1310

ATM NM_000051.3 31-130 CCNB1 NM_031966.2 716-815

ATR NM_001184.2 566-665 CCND1 NM_053056.2 691-790

AURKA NM_003600.2 406-505 CCND2 NM_001759.2 5826-5925

AVEN NM_020371.2 441-540 CCND3 NM_001136017.2 1934-2033

AVL9 NM_015060.1 1346-1445 CCNE1 NM_001238.1 1636-1735

AXL NM_001699.4 1898-1997 CD19 NM_001178098.1 939-1038

AZGP1 NM_001185.2 124-223 CD24 NM_013230.2 96-195 Table 8: continuation

Gene Accession T. Region Gene Accession T. Region

CD3G NM_000073.2 405-504 CHEK1 NM_001 1 14121.1 2226-2325

CD4 NM_000616.3 836-935 CHEK2 NM_001005735.1 536-635

CD44 NM_000610.3 2461 -2560 CHPF NM_024536.5 2905-3004

CD68 NM_001040059.1 1571 -1670 CHST1 1 NM_018413.4 326-425

CD84 NM_001 184882.1 76-175 CHUK NM_001278.3 861 -960

CD86 NM_006889.3 147-246 CITED4 NM_133467.2 916-1015

CD8A NM_001768.5 1321 -1420 CKS1 B NM_001826.2 239-338

CDA NM_001785.2 323-422 CKS2 NM_001827.1 196-295

CDC123 NM_006023.1 496-595 CLDN3 NM_001306.3 607-706

CDC20 NM_001255.2 431 -530 CLDN4 NM_001305.3 1243-1342

CDC25B NM_004358.3 3006-3105 CLDN7 NM_001307.3 176-275

CDC25C NM_001790.2 1056-1 155 CLMN NM_024734.3 3336-3435

CDC45L NM_003504.3 1676-1775 C16orf61 NM_020188.3 532-631

CDC6 NM_001254.3 1301 -1400 COG8 NM_032382.4 1 151 -1250

CDCA5 NM_080668.3 321 -420 COX6C NM_004374.2 70-169

CDCA7 NM_031942.4 771 -870 COX7B NM_001866.2 4-103

CDCA7L NM_001 127370.1 71 -170 CRIM1 NM_016441.1 1521 -1620

CDCA8 NM_018101.2 1666-1765 CRYAB NM_001885.1 579-678

CDH1 NM_004360.2 1231 -1330 CTGF NM_001901.2 1 101 -1200

CDH3 NM_001793.4 3746-3845 CTNNB1 NM_001098209.1 181 1 -1910

CDK1 NM_001 170406.1 700-799 CTPS NM_001905.2 2571 -2670

CDK4 NM_000075.2 1056-1 155 CTSL1 NM_001912.4 1073-1 172

CDKN1A NM_000389.2 1976-2075 CTSL2 NM_001333.2 66-165

CDKN1 B NM_004064.2 366-465 CXCL1 NM_00151 1.1 446-545

CDKN2A NM_000077.3 976-1075 CXCL14 NM_004887.4 1 126-1225

CDKN2B NM_004936.3 1 176-1275 IL8 NM_000584.2 26-125

CDKN2C NM_001262.2 1296-1395 CXCR1 NM_000634.2 1951 -2050

CDKN2D NM_001800.3 871 -970 CXCR2 NM_001 168298.1 1 13-212

CDKN3 NM_001 130851.1 391 -490 CXXC5 NM_016463.7 1266-1365

CDT1 NM_030928.3 1437-1536 CYB5B NM_030579.2 481 -580

CDYL NM_001 143970.1 1591 -1690 CYBRD1 NM_001 127383.1 1216-1315

CEACAM6 NM_002483.4 1218-1317 CYCS NM_018947.4 1736-1835

CELSR1 NM_014246.1 10056-10155 CYR61 NM_001554.3 1391 -1490

CENPA NM_001042426.1 980-1079 DDB2 NM_000107.1 841 -940

CENPF NM_016343.3 5823-5922 DDIT4 NM_019058.2 86-185

CENPI NM_006733.2 661 -760 DDR1 NM_001954.4 1343-1442

CENPN NM_001 100624.1 1941 -2040 DEGS2 NM_206918.2 398-497

CEP55 NM_001 127182.1 559-658 DLGAP5 NM_001 146015.1 131 -230

CFLAR NM_001 127183.1 654-753 DNAJC12 NM_021800.2 621 -720 Table 8: continuation

Gene Accession T. Region Gene Accession T. Region

DNALI1 NM_003462.3 1786-1885 FLVCR2 NM_017791.2 1256-1355

DSP NM_001008844.1 6026-6125 FNBP1 NM_015033.2 1341-1440

E2F1 NM_005225.1 936-1035 FOXA1 NM_004496.2 2466-2565

ECE2 NM_001037324.2 1096-1195 FOXC1 NM_001453.1 1516-1615

EGFR NM_005228.3 2761-2860 FOXM1 NM_021953.2 3209-3308

EIF2S2 NM_003908.3 1611-1710 FZD6 NM_001164615.1 1231-1330

ELOVL5 NM_021814.3 2081-2180 FZD7 NM_003507.1 1891-1990

ELSPBP1 NM_022142.3 151-250 GABPB1 NM_002041.4 726-825

COX4NB NM_001142288.1 861-960 GAL NM_015973.3 445-544

EMP3 NM_001425.2 351-450 GALNT7 NM_017423.2 911-1010

EPCAM NM_002354.1 416-515 GARS NM_002047.2 1231-1330

EPN3 NM_017957.2 2533-2632 GATA3 NM_001002295.1 2836-2935

EPSTI1 NM_001002264.1 611-710 GGH NM_003878.2 693-792

ERBB2 NM_001005862.1 1256-1355 GINS2 NM_016095.2 991-1090

ERBB3 NM_001005915.1 421-520 GLRB NM_000824.3 1236-1335

ERBB4 NM_001042599.1 7301-7400 GNG11 NM_004126.3 431-530

ERCC1 NM_001166049.1 2856-2955 GOLT1A NM_198447.1 266-365

ESR1 NM_000125.2 2391-2490 GPR160 NM_014373.1 761-860

ESRP1 NM_001034915.2 1516-1615 GPR89A NM_001097612.1 1482-1581

EVI2A NM_001003927.1 246-345 GPSM2 NM_013296.3 1931-2030

EX01 NM_003686.3 2716-2815 GRB7 NM_001030002.1 971-1070

EZH2 NM_004456.3 191-290 GREM1 NM_013372.5 576-675

F11R NM_016946.4 2106-2205 GRHL1 NM_198182.1 941-1040

F3 NM_001993.3 1031-1130 GRHL2 NM_024915.3 3691-3790

FABP4 NM_001442.2 416-515 GSTM1 NM_000561.2 336-435

FABP5 NM_001444.1 101-200 GSTM3 NM_000849.3 1026-1125

FAM171A1 NM_001010924.1 2936-3035 GSTM4 NM_000850.4 61-160

FAM174B NM_207446.2 1076-1175 GSTP1 NM_000852.2 416-515

FAM198B NM_001031700.2 1631-1730 GTPBP4 NM_012341.2 81-180

KIAA1370 NM_019600.2 3266-3365 GUSB NM_000181.1 1351-1450

FANCA NM_000135.2 266-365 H19 NR_002196.1 1593-1692

FANK1 NM_145235.3 446-545 HEXIM1 NM_006460.2 2921-3020

FAP NM_004460.2 1491-1590 C8orf30A NM_016458.2 2226-2325

FBN1 NM_000138.3 6421-6520 HIF1A NM_001530.2 1986-2085

FBP1 NM_000507.3 591-690 HJURP NM_018410.3 1326-1425

FBXL6 NM_012162.1 548-647 HMGA1 NM_002131.3 92-191

FGFR1 NM_015850.2 1336-1435 HN1 NM_001002032.1 711-810

FGFR2 NM_000141.4 706-805 HRAS NM_001130442.1 397-496

FGFR4 NM_002011.3 1586-1685 HSPA14 NM_016299.2 1331-1430

FIGF NM_004469.2 581-680 HSPD1 NM_002156.4 924-1023 Table 8: continuation

Gene Accession T. Region Gene Accession T. Region

ID4 NM_001546.2 2049-2148 KRT14 NM_000526.4 524-623

IDH2 NM_002168.2 426-525 KRT16 NM_005557.3 1391 -1490

ID01 NM_002164.3 51 -150 KRT17 NM_000422.2 515-614

IFT74 NM_001099222.1 136-235 KRT18 NM_000224.2 841 -940

IGBP1 NM_001551.2 1486-1585 KRT19 NM_002276.4 97-196

IGF1 NM_000618.3 492-591 KRT23 NM_015515.3 1736-1835

IGF2R NM_000876.1 2606-2705 KRT5 NM_000424.2 131 -230

IGFBP2 NM_000597.2 676-775 KRT6A NM_005554.3 1 18-217

IKBKB NM_001556.1 1996-2095 KRT6B NM_005555.3 2096-2195

IKBKE NM_014002.2 2471 -2570 KRT6C NM_173086.4 1854-1953

IL1 B NM_000576.2 841 -940 KRT8 NM_002273.3 360-459

IL6 NM_000600.1 221 -320 KRTAP1 -1 NM_030967.2 565-664

IL6R NM_000565.2 994-1093 LAG 3 NM_002286.5 1736-1835

IL6ST NM_002184.2 2506-2605 LAMA3 NM_000227.3 4261 -4360

INHBA NM_002192.2 491 -590 LAMC2 NM_005562.2 4296-4395

INPP4B NM_001 101669.1 3056-3155 LEPRE1 NM_001 146289.1 861 -960

INSIG1 NM_005542.3 1 121 -1220 LHFP NM_005780.2 461 -560

IRX3 NM_024336.1 2103-2202 h.LOC389332 NR_024418.1 1606-1705

ITCH NM_031483.4 156-255 h.LOC400043 NR_026656.1 1056-1 155

ITGA6 NM_000210.1 3066-3165 h.LOC642077 XM_942735.1 262-361

ITGB1 NM_00221 1.3 356-455 h.LOC647456 XM_942813.1 121 -220

JUP NM_002230.2 1076-1 175 s.Cytokeratin-8 XM_937689.1 813-912

KCNJ15 NM_002243.3 2161 -2260 LRIG1 NM_015541.2 571 -670

KCTD1 NM_001 136205.1 1368-1467 LRP8 NM_001018054.1 2091 -2190

KDM4B NM_015015.2 121 -220 LRRC2 NM_024512.3 1 1 1 -210

KDR NM_002253.2 1421 -1520 LSR NM_015925.5 804-903

KIAA0040 NM_001 162893.1 2791 -2890 LTBP2 NM_000428.2 5985-6084

KIAA1324 NM_020775.2 1806-1905 MAD2L1 NM_002358.3 183-282

KIF13B NM_015254.3 1 16-215 MAGEA1 NM_004988.4 477-576

KIF20A NM_005733.2 301 -400 MAGOHB NM_018048.2 1523-1622

KIF23 NM_004856.4 2721 -2820 MAP2K1 NM_002755.2 971 -1070

KIF2C NM_006845.3 1941 -2040 MAP2K4 NM_003010.2 2831 -2930

KIF4A NM_012310.3 3232-3331 MAP7D3 NM_024597.2 806-905

KIFC1 NM_002263.3 1547-1646 MAPT NM_001 123066.2 5606-5705

KIT NM_000222.1 6-105 MCM2 NM_004526.2 2946-3045

KLF4 NM_004235.4 1981 -2080 MCM3 NM_002388.3 301 -400

KLHL7 NM_001031710.2 1681 -1780 MDM2 NM_001 145337.1 5871 -5970

KLHL9 NM_018847.1 3581 -3680 ME1 NM_002395.3 1406-1505

KPNA1 NM_002264.2 1421 -1520 MED21 NM_004264.3 124-223

KRAS NM_004985.3 1791 -1890 MELK NM_014791.2 366-465 Table 8: continuation

Table 8: continuation

Gene Accession T. Region Gene Accession T. Region

PIR NM_001018109.1 746-845 RBBP8 NM_002894.2 761-860

PITX1 NM_002653.4 1551-1650 RECK NM_021111.2 2136-2235

PLA1A NM_015900.2 1251-1350 RECQL NM_002907.3 1251-1350

PLOD1 NM_000302.2 966-1065 REEP6 NM_138393.1 387-486

PN01 NM_020143.2 716-815 RE LA NM_001145138.1 2356-2455

PNP NM_000270.2 1151-1250 RELB NM_006509.2 251-350

POLD1 NM_002691.2 2393-2492 RERG NM_032918.1 526-625

PPFIBP1 NM_003622.2 2586-2685 RFC4 NM_002916.3 956-1055

SAPS1 NM_014931.3 781-880 RGS22 NM_015668.3 2576-2675

PRAME NM_006115.3 1391-1490 RHBG NM_020407.2 661-760

PRC1 NM_003981.2 2046-2145 RINT1 NM_021930.4 1806-1905

PREP NM_002726.3 1451-1550 RNF103 NM_005667.2 2891-2990

PROM1 NM_001145847.1 601-700 RPLP0 NM_001002.3 251-350

PSMA7 NM_002792.2 640-739 RRAGD NM_021244.4 2281-2380

PSMC4 NM_006503.2 251-350 RRM2 NM_001034.1 1616-1715

PSMD14 NM_005805.4 701-800 RRP15 NM_016052.3 7076-7175

PSPH NM_004577.3 226-325 S100A11 NM_005620.1 474-573

PSPHL AJ001612.1 1-100 S100A14 NM_020672.1 461-560

PTDSS1 NM_014754.1 2376-2475 S100A8 NM_002964.3 116-215

PTEN NM_000314.3 1676-1775 S100A9 NM_002965.2 76-175

PTGER4 NM_000958.2 1381-1480 SCGB2A2 NM_002411.1 266-365

PTGS2 NM_000963.1 496-595 SCUBE2 NM_001170690.1 2291-2390

PTTG1 NM_004219.2 543-642 SEH1L NM_001013437.1 501-600

PUF60 NM_001136033.1 1686-1785 SEMA3C NM_006379.2 946-1045

PUM1 NM_001020658.1 641-740 SERPINA3 NM_001085.4 6-105

PVRL3 NM_015480.1 1111-1210 SETBP1 NM_001130110.1 1071-1170

PYROXD1 NM_024854.3 1049-1148 SF3A1 NM_001005409.1 236-335

RAB25 NM_020387.2 246-345 SFRP1 NM_003012.3 3321-3420

RAB35 NM_001167606.1 436-535 SH2B3 NM_005475.2 4286-4385

RACGAP1 NM_013277.3 10-109 SHC1 NM_001130040.1 1986-2085

RAD 17 NM_002873.1 26-125 SLC16A3 NM_001042422.1 390-489

RAD50 NM_005732.2 5398-5497 SLC25A19 NM_001126121.1 1086-1185

RAD51 NM_001164269.1 751-850 SLC39A6 NM_001099406.1 1041-1140

RAD51L1 NM_002877.4 91-190 SLC40A1 NM_014585.5 1666-1765

RAD51 C NM_002876.2 301-400 GPR172A NM_024531.3 941-1040

RAF1 NM_002880.2 1991-2090 SLC5A6 NM_021095.1 1456-1555

RAI2 NM_021785.3 1606-1705 SLC7A6 NM_001076785.1 2111-2210

RAN BP 1 NM_002882.2 381-480 SLC9A3 NM_004174.2 736-835

RARA NM_000964.2 116-215 SLC9A3R1 NM_004252.3 1811-1910

RB1 NM_000321.1 2111-2210 C4orf34 NM_174921.1 371-470 Table 8: continuation

Gene Accession T. Region Gene Accession T. Region

SMO NM_005631.3 1616-1715 TMEM208 NM_014187.3 141 -240

SNAI 1 NM_005985.2 64-163 TMEM25 NM_001 144034.1 1053-1 152

SNRPA1 NM_003090.2 120-219 TMEM45B NM_138788.3 2076-2175

SNRPD1 NM_006938.2 1205-1304 TNFRSF11A NM_003839.2 491 -590

SPAG5 NM_006461.3 51 1 -610 TNFSF1 1 NM_003701.2 491 -590

SPATA7 NM_001040428.2 1006-1 105 TOM1 L1 NM_005486.2 1431 -1530

SPDEF NM_012391.1 1336-1435 TOMM40 NM_001 128916.1 1585-1684

SPINT1 NM_001032367.1 1316-1415 TOP2A NM_001067.2 5377-5476

SPINT2 NM_001 166103.1 626-725 TOR1A NM_0001 13.2 626-725

SQLE NM_003129.3 251 -350 TP53 NM_000546.2 1331 -1430

SRC NM_005417.3 176-275 TP53BP2 NM_001031685.2 1541 -1640

ST18 NM_014682.2 1296-1395 TP63 NM_001 1 14978.1 1 176-1275

STAT1 NM_007315.2 206-305 TRIM29 NM_012101.3 2646-2745

STAT3 NM_003150.3 2061 -2160 TRIP13 NM_001 166260.1 951 -1050

STC2 NM_003714.2 2826-2925 TSHZ1 NM_005786.4 4466-4565

STK1 1 NM_000455.4 2061 -2160 TSPAN13 NM_014399.3 556-655

STK38L NM_015000.1 421 -520 TTK NM_001 166691.1 776-875

STMN1 NM_001 145454.1 81 1 -910 TUBA4A NM_006000.1 218-317

STRAP NM_007178.3 1536-1635 TUBB6 NM_032525.1 1396-1495

SUV39H2 NM_024670.3 2036-2135 TWIST1 NM_000474.3 36-135

TACC3 NM_006342.1 154-253 TWIST2 NM_057179.1 1266-1365

TAP1 NM_000593.5 2076-2175 TYMP NM_001953.3 720-819

TCEAL1 NM_001006639.1 471 -570 TYMS NM_001071.1 556-655

TCF7L1 NM_031283.1 2216-2315 UBE2C NM_007019.2 562-661

TFAM NM_003201.1 86-185 UBE2T NM_014176.3 596-695

TFF1 NM_003225.2 21 1 -310 UCHL1 NM_004181.3 451 -550

TFF3 NM_003226.2 582-681 UIMC1 NM_016290.3 996-1095

TFRC NM_001 128148.1 2041 -2140 USP10 NM_005153.2 1921 -2020

TGFBR2 NM_001024847.1 1761 -1860 VAMP 8 NM_003761.3 261 -360

TGFBR3 NM_003243.3 1951 -2050 VAV3 NM_001079874.1 353-452

THBS1 NM_003246.2 3466-3565 VEGFA NM_001025366.1 1326-1425

THY1 NM_006288.2 136-235 VIM NM_003380.2 695-794

TIMM17A NM_006335.2 86-185 WDR12 NM_018256.3 656-755

TIMM8A NM_001 145951.1 41 1 -510 WDR4 NM_018669.4 1636-1735

TK1 NM_003258.1 1216-1315 WIPF2 NM_133264.4 1801 -1900

TM7SF3 NM_016551.2 1316-1415 XBP1 NM_001079539.1 936-1035

TMCC2 NM_014858.2 2793-2892 YBX1 NM_004559.3 541 -640

TMEM125 NM_144626.1 956-1055 CSDA NM_001 145426.1 658-757

TMEM139 NM_153345.1 1416-1515 ZEB1 NM_001 128128.1 1451 -1550

TMEM158 NM_015444.2 1271 -1370 ZEB2 NM_014795.2 21 -120 Table 8: continuation

"T. region": target region; h. hypothetical protein; "s. Cytokeratin-8": similar to Keratin type II cytoskeletal 8 (Cytokeratin-8) (CK-8) (Keraton-8) (K8); "h. MGC18216": hypothetical protein MGC18216.