Biomarker SPAG5

Abstract

There has been a recent shift in cancer therapy from one size fits all to a personalized and tailored treatment for individual patients to increase efficiency and avoid unnecessary toxicity. This invention relates to a method of determining the prognosis and suitable treatment of cancer in a subject by measuring the level of expression of SPAG5. In particular, it relates to a method where high expression of SPAG5 in tumour cells correlates with aggressive tumours.

Claims

1. A method for treating a subject having cancer, the method comprising: selecting a subject having cancer; detecting an elevated expression level of SPAG5 in a tumour tissue sample from said selected subject relative to a SPAG5 reference value; and administering, based on said detecting, to the selected subject a platinum chemotherapy agent, an anthracycline, HERCEPTIN (trastuzumab), or a combination thereof.

2. The method according to claim 1, wherein the selected subject is exhibiting a high expression level of SPAG5 protein or a high level of mRNA relative to a SPAG5 reference value.

3. The method according to claim 1, wherein the SPAG5 reference value is the level of SPAG5 expression in normal tissues or is standard histo-pathological criteria.

4. The method of claim 1, wherein the selected subject is a human.

5. The method of claim 1, wherein the selected subject has a cancer selected from breast cancer, ovarian cancer, gastric cancer, pancreatic cancer, prostate cancer, head and neck cancers, lung cancer or colorectal cancer.

6. A method for treating a subject having cancer, the method comprising: selecting a subject having cancer; detecting (i) an elevated expression level of SPAG5 in a tumour tissue sample from said subject relative to a SPAG5 reference value and (ii) an H-Score of greater than 10 in the tumour tissue sample from said subject; and administering, based on said detecting, to the selected subject a platinum chemotherapy agent, an anthracycline, HERCEPTIN (trastuzumab), or a combination thereof.

7. The method of claim 6, wherein the selected subject is a human.

8. The method of claim 6, wherein the selected subject has a cancer selected from breast cancer, ovarian cancer, gastric cancer, pancreatic cancer, prostate cancer, head and neck cancers, lung cancer or colorectal cancer.

Description

(1) There now follows by way of example only a detailed description of the present invention with reference to the accompanying drawings, in which;

(2) FIG. 1: Shows the results of an artificial neural network and pathway analysis which revealed that SPAG5 is one of the main hubs in KIF2C pathways and that SPAG5 expression is strongly correlated to expression of genes that are involved in mitotic cell cycle regulation especially anaphase phase and cell division, DNA repair and metastases.

(3) FIG. 2: Shows Kaplan Meier curves showing breast cancer specific survival (BCSS) and disease free survival (DFS) of patients with breast tumours that had low or high expression of SPAG5 mRNA. Left hand panels show survival of the whole cohort of 128 breast cancer patents. Right hand panels show survival of a sub-group with ER+ tumours.

(4) FIG. 3: Shows a multivariate logistic regression model for SPAG5 RNA expression.

(5) FIG. 4: Shows Dual immunoflurescent (IF) staining of SPAG5 (green) and proliferating cell nuclear antigen (PCNA, red) in breast cancer cell lines. Co-expression of SPAG5 and PCNA was expressed in 4 out of 5 of the breast cancer cell lines screened (MCF7, T47D, MDA468 and MDA231). In these breast cancer cells the IF staining strongly suggests that in proliferating cells (evidenced by PCNA staining) there are both co-expression of SPAG5 and PCNA as well as increased levels of expression of SPAG5 providing evidence for the importance of SPAG5 in cell proliferation. Furthermore there is evidence of localisation of SPAG5 within the centrosomes and spindles of dividing cells. In SkBr3 cells both SPAG5 and PCNA were expressed by the cells but there appeared to be no direct correlation between levels of SPAG5 and the proliferation of the cells.

(6) FIG. 5: Shows a correlation between SPAG5 mRNA expression and mitotic index.

(7) FIG. 6: Shows immunostained sections of normal tissue showing SPAG5 expression in normal, liver, kidney and breast tissue.

(8) FIG. 7: Shows immunostained sections of breast cancer tissue with negative and with positive SPAG5 expression.

(9) FIG. 8: Shows immunostained sections of ovarian cancer tissue with either negative or positive SPAG5 expression (left hand panels) and immunostained sections of pancreatic cancer tissue with either negative or positive SPAG5 expression (right hand panels).

(10) FIG. 9: Shows immunostained sections of gastric cancer tissue with either negative or positive SPAG5 expression (left hand panels) and immunostained sections of colorectal cancer tissue with either negative or positive SPAG5 expression (right hand panels).

(11) FIG. 10: Shows a multivariate logistic regression model for SPAG5 protein expression.

(12) FIG. 11: Shows clinical outcome of high risk ER Breast cancer patients who received either no adjuvant chemotherapy or received CMF (n=285).

(13) FIG. 12: Shows clinical outcome of high risk ER breast cancer patients who received adjuvant anthracycline (n=256).

(14) FIG. 13: Shows A) clinical outcome of patients with high risk HER+ breast cancer who did not receive HERCEPTIN (trastuzumab, anti-HER2 monoclonal antibody) and either received CMF or no chemotherapy (n=285), B) clinical outcome of patients with high risk HER+ breast cancer who received anthracycline only (n=66), and C) clinical outcome of patients with high risk HER+ breast cancer who received both anthracycline and HERCEPTIN (trastuzumab, anti-HER2 monoclonal antibody) (n=108).

(15) FIG. 14: Shows multivariate analysis that confirms Top2A protein as an independent predictor for pCR.

(16) FIG. 15: Shows clinical outcome of a cohort of ovarian cancer patients having SPAG5 positive or SPAG5 negative tumours,

(17) FIG. 16: Shows Clinical outcome of mRNA SPAG5 in the gene expression discovery set (Nottingham series; n=128) and Clinical outcome of mRNA SPAG5 in the gene expression Confirmatory set (Uppsala series; n=249),

(18) FIG. 17: Shows Clinical outcome of Discovery set (n=128) and clinical outcome of the Nottingham set (n=1650) for each of the graphs Low expression and High expression refer to expression of SPAG5 protein.

(19) FIG. 18: Shows A comparison of the survival times for Estrogen Receptor (ER) positive breast cancer tumours for the test set (top row) and validation set (bottom row). For all breast cancer cases (left hand column); survival times for high risk Estrogen receptor (ER) positive tumours with Tamoxifen treatment (centre column) and survival times for low risk ER receptor positive tumours with no Tamoxifen treatment.

(20) FIG. 19: Shows A comparison of the survival times for Estrogen Receptor (ER) negative breast cancer tumours for the test set (top row) and validation set (bottom row). Survival time is shown for high risk ER expressing tumours when no chemotherapy was used (bottom left); high risk ER expressing tumours with CMF chemotherapy and high risk ER expressing tumours when Anthracycline chemotherapy was used.

(21) FIG. 20: Shows the survival times of patients with HER2 overexpressing breast cancer tumours with: no chemotherapy (left hand side), anthracycline only (centre) and anthracycline with HERCEPTIN (trastuzumab, anti-HER2 monoclonal antibody) (right hand side).

(22) FIG. 21: Shows survival times for ovarian, gastric, colon and rectal cancers for patients with tumours that have high SPAG 5 expression and patients that have low SPAG5 expression,

(23) FIG. 22: Shows clinico-pathological correlations of SPAG5 protein expression in both the Training and Test sets,

(24) FIG. 23: Shows A comparison between the clinico-pathological characteristics of the Training and Validation sets showing both sets were well balanced,

(25) FIG. 24: Shows a list of the top 100 genes that are most closely associated with SPAG 5 in breast cancer tumours that over express estrogen receptors (ER+);

(26) FIG. 25: Shows a list of the top 100 genes that are most closely associated with SPAG 5 in all breast cancers regardless of ER, PR and HER2 status of the tumours.

(27) FIG. 26: Shows a list of the top 100 genes that are most closely associated with SPAG 5 in all breast cancers.

(28) Expression of Sperm-associated antigen 5 (SPAG 5) in human cancer can be used as a prognostic, predictive and new therapeutic target for human cancers including breast, ovarian, gastric, and colorectal and pancreatic cancers.

(29) Patients and Methods:

(30) Study Patients

(31) A) Breast Cancer

(32) A retrospective study was carried out in a consecutive series of 2800 BC patients with primary operable invasive BCs who were diagnosed and treated in a single centre (Nottingham City Hospital) from 1990 to 2010 and constituted three cohorts:

(33) 1) Nottingham Tenovus Primary BC Treated from 1986 to 1998 (Nottingham/Tenovus Series):

(34) This is a consecutive series of 1650 patients with primary invasive breast carcinomas who were diagnosed between 1989 and 1999 and entered into the Nottingham Tenovus Primary Breast Carcinoma series. This is a well-characterized series of patients under the age of 71 years (median=55 years) with long-term follow-up data (Supplemental Table S1). All patients were treated uniformly in a single institution and have been investigated in a wide range of biomarker studies. Patient characteristics such as age, family history, and menopausal status, and pathology characteristics such histologic grade, histologic tumor type, tumor size, lymph node stage, Nottingham Prognostic Index (NPI), and vascular invasion were routinely assessed in this large cohort (Pinder et al., Histopathology 24:41-7, 1994). Patients received standard surgery (mastectomy or wide local excision) with radiotherapy. Prior to 1989, patients did not receive systemic adjuvant treatment (AT). After 1989, AT was scheduled based on prognostic and predictive factor status, including NPI, oestrogen receptor-a (ER-a) status, and menopausal status. Patients with NPI scores of <3.4 (low risk) did not receive AT. In pre-menopausal patients with NPI scores of 3.4 (high risk), classical Cyclophosphamide, Methotrexate, and 5-Flourouracil (CMF) chemotherapy was given; patients with ER-a positive tumours were also offered HT. Postmenopausal patients with NPI scores of 3.4 and ER-a positivity were offered HT, while ER- negative patients received classical CMF chemotherapy. Median follow up was 111 months (range 1 to 233 months).

(35) TABLE-US-00001 TABLE S1 Clinicopathological characteristics of whole cohort (N = 1650) Variable n* Cases (%) Menopausal status 1650 Pre-menopausal 612 (37.0) postmenopausal 1038 (63.0) Tumor Grade (NGS) 1650 G1 306 (18.5) G2 531 (32.2) G3 813 (49.3) Lymph node stage 1650 Negative 1056 (64.0) Positive (1-3 nodes) 486 (29.5) Positive (>3 nodes) 108 (6.5) Tumour size (cm) 1650 T1 a + b (1.0) 187 (11.0) T1 c (>1.0-2.0) 868 (53.0) T2 (>2.0-5) 5729 (35.0) T3 (>5) 16 (1.0) Tumour type 1650 IDC-NST 941 (57) Tubular 349 (21) ILC 160 (10) Medullary (typical/atypical) 41 (2.5) Others 159 (9.5) NPI subgroups 1650 Excellent PG (2.08-2.40) Low risk 207 (12.5) Good PG (2.42-3.40) 331 (20.1) Moderate I PG (3.42 to 4.4) High risk 488 (29.6) Moderate II PG (4.42 to 5.4) 395 (23.9) Poor PG (5.42 to 6.4) 170 (10.3) Very poor PG (6.5-6.8) 59 (3.6) Survival at 20 years 1650 Alive and well 1055 (64.0) Dead from disease 468 (28.4) Dead from other causes 127 (7.6) Adjuvant systemic therapy (AT) 1602 No AT 665 (42.0) Hormone therapy (ET) 642 (41.0) Chemotherapy 307 (20.0) Hormone + chemotherapy 46 (3.0) HAGE expression 1650 No staining (0) normal 1581 Strong (+++) positive 134 *Number of cases for which data were available. NPI; Nottingham prognostic index, PG; prognostic group

(36) 2) Anthracycline Adjuvant Series:

(37) 256 primary invasive BC diagnosed and managed between 1998 and 2007 and all patients were primarily treated with surgery followed by radiotherapy and anthracycline-based chemotherapy

(38) 3) HER2+/HERCEPTIN (Trastuzumab, Anti-HER2 Monoclonal Antibody) Adjuvant Series

(39) 140 primary BC HER2+ patients diagnosed and managed between 2005 and 2011 and all patients were primarily treated with surgery followed by radiotherapy and subsequent treatment with anthracycline-based and HERCEPTIN (trastuzumab, anti-HER2 monoclonal antibody) chemotherapy.

(40) 4) Anthracycline-Neoadjuvant Series

(41) To validate SPAG5 as a predictor factor for anthracycline chemotherapy, 260 patients with locally advanced primary breast cancer treated with anthracycline-based combination followed by surgery were included and the pathological complete response (PCR) was used to evaluate the response to chemotherapy. The complete pathological response was defined as an absence of invasive carcinoma in both primary site and axillary lymph nodes.

(42) FIGS. 16 to 20 and 22 to 25 relate to data collected from a discovery gene expression set (Nottingham series; n=128) and a confirmatory gene expression set (Uppsala set; n=249).

(43) These datasets were used to investigate the clinical outcome in terms of survival time for patients with tumours that had high or low expression of SPAG 5 and high or low expression of estrogen receptors (ER). This data is set out in FIGS. 16 to 20 and 21 to 25. The datasets were used to evaluate how the expression level of SPAG 5 combined with the expression or not of estrogen receptors can be used to predict the prognosis of patients and be used to evaluate which tumours are likely to be sensitive to particular chemotherapeutic agents with or without Tamoxifen.

(44) Overall, for all breast tumours, high expression of SPAG 5 correlated with poorer outcome (FIGS. 16 and 17).

(45) For breast tumours that express estrogen receptors (ER+) low SPAG 5 expression correlated with better prognosis in terms of life expectancy than for ER+tumours that had high SPAG 5 expression. About 60% of breast cancers express estrogen receptors. The effect of using Tamoxifen for high risk and low risk ER+ tumours are shown in FIG. 18.

(46) For breast tumours that do not express estrogen receptors (ER) SPAG 5 expression is a less-strong predictor of survival time when chemotherapy is not applied. When chemotherapy with CMF or Anthracycline is applied tumours with high expression of SPAG 5 responded better to treatment as shown in FIG. 19.

(47) Breast tumours that overexpress HER2 and also have over expression of SPAG 5 were shown to have a better prognosis if treated with anthracycline and HERCEPTIN (trastuzumab, anti-HER2 monoclonal antibody) than when treated with anthracycline only. Patients whose breast tumours overexpressed SPAG 5 and HER2 had a good survival rate with treatment of anthracycline only and also with anthracycline and HERCEPTIN (trastuzumab, anti-HER2 monoclonal antibody) as shown in FIG. 20.

(48) High expression of SPAG 5 correlated with better prognosis in patients with ovarian, gastric and colon cancer but not in patients with rectal cancer see FIG. 21.

(49) Tumours were graded as shown in FIGS. 22 and 23 which show that high expression of SPAG 5 at the gene level follows through to high SPAG 5 protein expression.

(50) Assessment of the expression levels genes linked to SPAG 5 in different types of breast cancer, see genes listed in FIGS. 24 to 26, can help to understand the pathways that SPAG 5 is involved in different types of cancer cells. Gene hubs, which are genes linked to a number of other genes in the SPAG 5 pathway, for example SPAG 5 itself, can be knocked out in model systems to test the effect on expression levels of the genes that are linked to SPAG 5. This can lead to better understanding of the pathways linked to SPAG 5.

(51) B) Ovarian Cancer

(52) Investigation of the expression of SPAG5 in ovarian cancer was carried out on a tissue microarray of 172 ovarian cancer cases. This cohort comprised of female patients with a median age of 61 years (range 33-87, mean 60 years). Of the 172 patients included in the study, 54 (34%) were dead before the end of the follow-up period. Histologically, most patients were found to have a serous cystadenocarcinoma (56%), followed by endometrioid (21%), clear cell carcinoma (13%), mucinous cystadenocarcinoma (8%) or other types (2%). Tumours displayed poor histologic differentiation in 73% of cases. The majority of tumours were classified as FIGO stage III (45%). Tissue was obtained from patients with primary ovarian cancer treated at Nottingham University Hospitals (NUH) between 2000 and 2007. Survival was calculated from the operation date until 1st of April 2009 when any remaining survivors were censored. During the study period, patients were treated with either single agent carboplatin [65 patients (41.4%)] or platinum-based combination chemotherapy [89 patients (56.7%); 79/89 patients received carboplatin and paclitaxel, 4/89 received Cyclophosphamide/Doxorubicin/Cisplatinum (CAP), 3/89 were treated in ICON-5 trial using paclitaxel, carboplatin plus gemcitabine or topotecan 29 and 3/89 were treated according to the SCOTROC trial protocol by carboplatin, docetaxel+/topotecan 30. Platinum resistance was defined as patients who had progression during first-line platinum chemotherapy or relapse within 6 months after treatment.

(53) The Reporting Recommendations for Tumor Marker Prognostic Studies (REMARK) criteria, recommended by McShane et al, J. Natl Cancer Inst 97:1180-4, 2005, were followed throughout this study. This work was approved by Nottingham Research Ethics Committee.

(54) Survival Data

(55) Survival data including survival time, disease-free survival (DFS), and development of loco-regional and distant metastases (DM) were maintained on a prospective basis. DFS was defined as the number of months from diagnosis to the occurrence of local recurrence, local LN relapse or DM relapse. BC specific survival (BCSS) was defined as the number of months from diagnosis to the occurrence of BC related-death. DM-free survival was defined as the number of months from diagnosis to the occurrence of DM relapse. Survival was censored if the patient was still alive, lost to follow-up, or died from other causes.

(56) Immunohistochemistry (IHC)

(57) Tumours from Nottingham/Tenovus series, HER2/HERCEPTIN (trastuzumab, anti-HER2 monoclonal antibody), Anthracycline-adjuvant and ovarian cancer series were arrayed in tissue microarrays (TMAs) constructed with 2 replicate 0.6 mm cores from the centre and periphery of the tumors for each patient. While full-face sections from pre-chemotherapy core biopsy of Anthracycline-Neoadjuvant series were used. The TMAs and full face sections were immunohistochemically profiled for SPAG5 and other biological antibodies (Table 5) as previously described (9-12). Immunohistochemical staining was performed using NOVOLINK Detection kit according to the manufacturer instructions (Leica Microsystems). Sections were incubated for 30 mins at room temperature with 1/50 Anti-SPAG5 rabbit polyclonal (HPA022479; Sigma). Anti-SPAG5 antibody produced in rabbit, a Prestige Antibody, is developed and validated by the Human Protein Atlas (HPA) project (www.proteinatlas.org) where the antibody is tested by immunohistochemistry against hundreds of normal and disease tissues.

(58) Pre-treatment of sections was performed with citrate buffer (pH 6.0) antigen for 20 minutes. To validate the use of TMAs for immuno-phenotyping, full-face sections of 40 cases were stained and the protein expression levels of the different antibodies were compared. The concordance between TMAs and full-face sections was excellent (kappa=0.8). Positive and negative (omission of the primary antibody and IgG-matched serum) controls were included in each run.

(59) Evaluation of Immunohistochemical Staining

(60) The tumour cores were evaluated by three pathologists of co-authors blinded to the clinico-pathological characteristics of patients in two different settings. There was excellent intra and inter-observer agreements (k>0.8; Cohen's and multi-rater tests, respectively). Whole field inspection of the core was scored and intensities of cytoplasmic staining were grouped as follows: 0=no staining, 1=weak staining, 2=moderate staining, 3=strong staining. The percentage of each category was estimated. Positive SPAG5 (SPAG5+) expression was defined as the presence of granular cytoplasmic staining in >10% of malignant cells. Not all cores within the TMA were suitable for IHC analysis as some cores were missing or lacked tumour.

(61) Statistical Analysis

(62) Data analysis was performed using SPSS (SPSS, version 17 Chicago, Ill.). Where appropriate, Pearson's Chi-square, Fisher's exact, Student's t and ANOVAs one way tests were used. Cumulative survival probabilities were estimated using the Kaplan-Meier method and differences between survival rates were tested for significance using the log-rank test. Multivariate analysis for survival was performed using the Cox hazard model. The proportional hazards assumption was tested using standard log-log plots. Hazard ratios (HR) and 95% confidence intervals (95% CI) were estimated for each variable. All tests were two-sided with a 95% CI and a p value of <0.05 considered significant. For multiple comparisons, p values were adjusted according to Holm-Bonferroni correction method (13).

(63) Results

(64) i) 15% and 5% of BC showed gain and amplification of SPAG5 locus, respectively, at chromosome 17q11.2. SPAG5 mRNA expression levels displayed a statistically significant correlation with its copy number (p=0.01), ii) 30% and 20% of ovarian and breast cancer showed SPAG5 protein overexpression, iii) In breast cancer, SPAG5 overexpression at both mRNA and protein levels showed a statistically significant association with ER, PR, triple negative phenotype, high grade tumour, high mitotic index, high ki67, high KIF2C, basal like phenotype and epithelial mesenchymal transition (EMT) phenotype, p53 and absence of DNA repair genes (BRCA1, ATM and XRCC1); p values<0.0001, iii) SPAG5 mRNA overexpression was statistically associated with poor clinical outcome (p<0.0001). In ER BC who treated with S+RT followed by anthracycline based adjuvant chemotherapy, SPAG5 negative BC had 7-10 times risk death, progression and DM (p values<0.00001). Moreover, in locally advanced BC who received anthracycline-bases neoadjuvant chemotherapy, SPAG5 overexpression BC achieved 38% pCR vs., 7% of SPAG5-negative BC (p<0.00001). After controlling to other validated predictors for pCR, SPAG5 remained as a powerful independent predictor (HR; 2.4, CI 95%; 1.5-3.9; p=0.00001). vi) SPAG5 negative OVC were resistant to platinum chemotherapy (p=0.004) and SPAG5 negative expression had 3 fold increase of risk of death (p=0.008) and progression (p=0.001) and independently associated with poor survival [HR 2.6, p=0.008].

(65) Neural network (NNT) and pathways analysis of a breast cancer (BC) gene expression array data from 128 patients showed that that SPAG5 was among top 10 ranked genes out of 48,000 transcripts, that accurately predicted worse clinical outcome, differentiated between low and high grade cancers and distinguished between breast cancer with high mitotic index and breast cancer with low mitotic index. These results were based on a 10-fold external cross-validation analysis with an average classification accuracy of >99.999%. SPAG5 was identified as a major hub in the KIF2C pathway and its expression is strongly related to expression of genes that are involved in mitotic cell cycle regulation (FIG. 1).

(66) Dual immunoflorescent staining (IF) in breast cancer cell lines (FIG. 4) showed co-expression of SPAG5 and proliferating cell nuclear antigen (PCNA) in 4 out of 5 of the breast cancer cell lines screened (MCF7, T47D, MDA468 and MDA231). In these breast cancer cells the IF staining strongly suggested that in proliferating cells (evidenced by PCNA staining) there is both co-expression of SPAG5 and PCNA as well as increased levels of expression of SPAG5 providing evidence for the importance of SPAG5 in cell proliferation. Furthermore there is evidence of localisation of SPAG5 within the centrosomes and spindles of dividing cells.

(67) The molecular and clinicopathological functions of SPAG5 expression and its effect on management of both breast and ovarian cancers (OVC) have been investigated.

(68) SPAG5 may be over-expressed in breast, ovarian, gastric, pancreatic and colorectal cancer tissues. Tumours that over express SPAG5 (SPAG5+ tumours) are associated with more aggressive highly proliferating breast and ovarian cancer. Moreover, it is possible from the amount of SPAG5 expression to identify patients who should be clinically treated in particular ways and could be a target for new therapeutics.

(69) If a tumour over expresses SPAG5 it indicates an increased probability that the cancer is aggressive. Tumours that overexpress SPAG5 were also shown to have a greater chance of response to chemotherapy with anthracycline adjuvant therapy, adjuvant therapy with HERCEPTIN (trastuzumab, anti-HER2 monoclonal antibody) and anthracycline and neoadjuvant therapy with anthracycline. SPAG5 expression may be determined by staining tissue samples with labelled SPAG5-specific antibodies (FIG. 6). The presence of strong granular cytoplasmic expression in >10% of the malignant cells indicates overexpression of SPAG5 (SPAG5+). The normal breast tissues showed weak SPAG5 expression (SPAG5-). Normal spleen, smooth muscle, colon, prostate and bladder tissues also usually show weak SPAG5 expression.

(70) By using high resolution array comparative genomic hybridization (aCGH) using data from the Chin et al database (Chin et al, Genome Biol 207; 8:R215) it was shown that between 5% and 15% of breast cancers showed amplification and gain of SPAG5 locus, respectively, at 17q11.2. SPAG5 mRNA expression displayed a significant correlation with its copy number (p<0.0001).

(71) Clinicopathological Features of SPAG5 mRNA Expression

(72) SPAG5 mRNA high expression (mRNA SPAG5+) was significantly associated with aggressive clinico-pathological features including: ER, high grade, high proliferation index, triple receptor negative (TNT) phenotypes, over-expression of HER2 and p53 mutation (Table 1). In multivariate regression analysis mRNA SPAG5 level was only associated with ER and high mitoses (FIG. 5). Moreover, mRNA SPAG5+ patients have 2-3-time increased risk of deaths and recurrence in the long term, compared to mRNA SPAG5-cases (FIG. 2).

(73) TABLE-US-00002 TABLE 1 Association between SPAG5 mRNA expression and other clinico-pathologic variables SPAG5 mRNA Expression Low High (N = 73) (N = 55) X.sup.2 Variable n (%) n (%) Adjusted p value A) Pathological Parameters Tumor Size 0.02 T1 a + b 8 (13) 2 (4) (1.0) T1 c (>1.0- 40 (64) 25 (516) 2.0) T2 (>2.0-5) 14 (23) 22 (22) T3 (>5) 0 (0) 0 (0) Lymph node stage Negative 56 (77) 30 (55) 0.028 Positive (1-3 15 (20) 21 (38) nodes) Positive (>3 2 3) 4 (7) nodes) Grade** 0.000000009 G1 30 (41) 2 (4) G2 31 (43) 22 (40) G3 12 (16) 31 (56) Mitotic 0.0000000004 Index M1 (low; 41 (66) 4 (8) mitoses < 10) M2 14 (23) 16 (33) (medium; mitoses 10- 18) M3 (high; 7 (11) 29 (59) mitosis > 18) Pleomorphism 0.001* 1 (small- 5 (8) 1 (2) regular uniform) 2 (Moderate 41 (661) 18 (37) variation) 3 (Marked 16 (26) 30 (611) variation) B) Hormonal receptors ER 0.000002 (Negative) 10 (14) 29 (53) (Positive) 63 (86) 26 (47) PR 0.001* (Negative) 16 (28) 28 (61) (Positive) 41(72) 18 (39) Cell cycle/apoptosis regulators p53 0.0000005* (Negative) 56 (98) 29 (59) (Positive) 1 (2) 20 (41) EGFR 0.015* (Negative) 47 (87) 30 (67) (Positive) 7 (13) 15 (33) vimentin 0.007* (Negative) 38 (97) 30 (77) (Positive) 1 (3) 9 (23) Bcl2 0.000002 (Negative) 16 (26) 35 (71) (Positive) 46 (74) 14 (29) Ki67 0.00000002 (Negative) 33 (85) 8 (21) (Positive) 6 (15) 30 (79) p16 0.0004* (Negative) 41 (100) 27 (73) (Positive) 0 (0) 10 (27) Aggressive phenotype Her2 0.006* overexpression (No) 58 (97) 36 (80) (Yes) 2 (3) 9 (20) Triple 0.0002* negative (No) 58 (95) 32 (68) (Yes) 3 (5) 15 (32) Basal like 0.02 (No) 57 (93) 37 (79) (Yes) 4 (7) 10 (21) **grade as defined by NGS; BRCA1: BC 1, early onset; HER2: human epidermal growth factor receptor 2; ER: oestrogen receptor; PR: progesterone receptor; CK: cytokeratin; Basal-like: ER-, HER2 and positive expression of either CK5/6, CK14 or EGFR; Triple negative: ER-/PR-/HER2-

(74) Clinicopathological feature of SPAG5 protein expression 20% of breast cancer showed SPAG5 protein overexpression. SPAG5 overexpression showed a statistically significant association with ER, PR, triple negative phenotype, high grade tumour, high mitotic index, high ki67, high KIF2C, basal like phenotype and epithelial mesenchymal transition (EMT) phenotype, p53 and absence of DNA repair genes (BRCA1, ATM and XRCC1); p values<0.0001 (Table 2). In multivariate regression analysis model, ER, mitotic index and p16 mutation were are only markers independently associated with SPAG5 overexpression (FIG. 10).

(75) TABLE-US-00003 TABLE 2 Association between SPAG5 protein expression and other clinico- pathologic variables in BC SPAG5 protein Expression Low High X.sup.2 (N = 1125) (N = 259) Adjusted Variable n (%) n (%) p value A) Pathological Parameters Tumor Size 0.110 T1 a + b (1.0) 107 (11) 16 (7) T1 c (>1.0-2.0) 510 (53) 121 (51) T2 (>2.0-5) 337 (35) 95 (40) T3 (>5) 4 (1) 4 (2) Lymph node stage 0.188 Negative 612 (63) 136 (57) Positive (1-3 nodes) 293 (30) 86 (36) Positive (>3 nodes) 67 (7) 15 (6) Grade** 2 10.sup.22 G1 196 (20) 13 (5.5) G2 356 (37) 36 (15) G3 418 (43) 187 (79) Mitotic Index 1.5 10.sup.23 M1 (low; mitoses < 10) 401 (41) 32 (14) M2 (medium; 194 (20) 26 (11) mitoses 10-16) M3 (high:, mitosis > 18) 375 (39) 178 (75) Pleomorphism 4.3 10.sup.21* 1 (small-regular uniform) 27 (3) 2 (1) 2 (Moderate variation) 445 (46) 31 (13) 3 (Marked variation) 498 (51) 203 (86) B) Hormonal receptors ER (Negative) 176 (20) 136 (61) 1.6 10.sup.34 (Positive) 714 (80) 87 (39) PR (Negative) 343 (37) 156 (69) 1.0 10.sup.17* (Positive) 578 (63) 71 (31) AR (Negative) 296 (31) 137 (65) 2.8 10.sup.20* (Positive) 607 (69) 74 (35) JNK (Negative) 504 (52) 82 (59) 0.0000005* (Positive) 463 (48) 158 (41) Bcl2 (Negative) 299 (30) 132 (55) 5.1 10.sup.42 (positive) 709 (70) 109 (45) Tumour type IDC 525 (54) 173 (75) 1.6 10.sup.19 Tubular 218 (23) 24 (10) Medullary 11 (1) 20 (9) ILC 114 (12) 7 (3) others 98 (10) 7 (3) XRCC1 (Negative) 132 (14) 61 (26) 4.5 10.sup.6 (positive) 823 (86) 172 (74) HAGE (Negative) 1000 (93) 218 (86) 4.7 10.sup.4 (Positive) 79 (7) 36 (14) P53 (Negative) 756 (85) 140 (64) 9.9 10.sup.13 (Positive) 135 (15) 80 (36) KIF2C (Negative) 198 (25) 30 (16) 0.006 (Positive) 601 (75) 164 (84) MIB1 (low) 278 (49) 25 (23) 2.4 10.sup.7 (high) 284 (51) 91 (77) TOP2A (negative) 390 (47) 98 (47) 0.9 (High) 438 (53) 112 (53) Ck5/6 (negative) 820 (88) 157 (68) 1.1 10.sup.13 (positive) 112 (12) 74 (32) EGFR (negative) 709 (82) 156 (72) 0.001 (positive) 154 (18) 60 (23) BRCA1 (Loss) 124 (16) 60 (31) 4.1 10.sup.6 (positive) 649 (84) 137 (69) P21 (Negative) 432 (64) 147 (68) 2.2 10.sup.4 (positive) 363 (46) 68 (32) P16 (Negative) 717 (92) 136 (66) 4.7 10.sup.22 (Positive) 65 (8) 71 (34) MDM2 (negative) 562 (72) 190 (89) 3.3 10.sup.7 (positive) 215 (28) 23 (11) CK14 (negative) 819 (89) 192 (83) 0.01 (positive) 102 (11) 40 (17) Basal like (No) 850 (92) 142 (65) 4.3 10.sup.27 (Yes) 74 (8) 78 (35) Vimentine (Negative) 597 (94) 120 (70) 3.6 10.sup.18 (positive) 41 (6) 52 (30) SMA (Negative) 816 (89) 179 (79) 2.6 10.sup.5 (Positive) 98 (11) 48 (21) P63 (Negative) 906 (99) 219 (96) 0.001 (Positive) 11 (1) 10 (4) CK18 (negative) 64 (7) 53 (25) 1.4 10.sup.13 (positive) 807 (93) 160 (75) CK 19 (negative) 43 (5) 30 (13) 3.1 10.sup.6 (positive) 883 (95) 202 (87) P-Cadherin (negative) 415 (53) 53 (27) 3.3 10.sup.11 (positive) 366 (47) 145 (73) FIHT (negative) 148 (18) 61 (31) 7.6 10.sup.5 (positive) 677 (82) 139 (69) ATM (negative) 136 (40) 49 (55) 0.01 (Positive) 201 (60) 40 (45) Tubular formation 3.9 10.sup.7 T1 62 (6) 2 (1) T2 354 (36) 57 (24) T3 554 (57) 177 (75) T24 (Negative) 578 (72) 170 (23) 0.016* (Positive) 220 (28) 41 (19) *Statistically significant; **grade as defined by NGS; BRCA1: BC 1, early onset; HER2: human epidermal growth factor receptor 2; ER: oestrogen receptor; PR: progesterone receptor; CK; cytokeratin; Basal-like: ER-, HER2 and positive expression of either CK5/6, CK14 or EGFR; Triple negative: ER-/PR-/HER2-

(76) Clinical Outcome of SPAG5 Protein Expression

(77) In high risk Estrogen receptor negative (ER) breast cancer patients who did not received any adjuvant therapy or received ineffective CMF chemotherapy, tumours that overexpressed SPAG5 protein and tumours that did not overexpress SPAG5 had a similar risk of recurrence and the patients had a similar risk of death (FIG. 11). Receiving anthracycline chemotherapy had a positive impact on high risk ER breast cancer patients with patients whose tumours over expressed SPAG5 protein having 72-65% less chance of death, recurrence and metastases compared to patients whose tumours did not overexpress SPAG5 protein; p<0.0001 (FIG. 12). The positive impact of anthracycline on patients with SPAG5+ tumours has also been shown for tumours that are HER2 positive (HER2+). These patients either received anthracycline only (treated before 2006) or anthracycline plus HERCEPTIN (trastuzumab, anti-HER2 monoclonal antibody) (FIG. 13).

(78) Patients with locally advanced breast cancer tumours who received anthracycline-based neoadjuvant chemotherapy had 39% pathological complete response (pCR) for SPAG5+BC but only 6% pCR for SPAG5-negative BC (p<0.00001). After controlling to other validated predictors for pCR, SPAG5 remained as a powerful independent predictor (HR; 2.4, CI 95%; 1.5-3.9; p=0.00001).

(79) Ovarian Cancer

(80) In ovarian cancer, 46.5% of tumours showed SPAG5+ protein expression which was significantly associated with aggressive clinico-pathological features including higher stage; grade and clear cell histological subtype (Table 4). Moreover, SPAG5+ ovarian cancer tumours were more sensitive to platinum chemotherapy (p=0.002; Table 4) and had higher probability of pathological (p=015) and CA125 response (p=0.036). CA125 is a marker that is routinely used in practice to indicate the presence of ovarian cancer. In addition, SPAG5 negative expression had 3 fold increase of risk of death (p=0.00008) and progression and independently associated with poor survival [HR 2.6, p=0.001; FIG. 15].

(81) TABLE-US-00004 TABLE 4 Association between SPAG5 protein expression and other clinico- pathologic variables in ovarian cancer SPAG5 Protein Expression Low High X.sup.2 (N = 92) (N = 80) Adjusted Variable n (%) n (%) p value A) Pathological Parameters Cancer Stage 0.004 0 1 (13) 1 (1) I 17 (64) 36 (46) II 14 (23) 8 (10) III 47 (51) 29 (37) IV 13 (14) 5 (6) Optimal debunked 0.008 yes 51 (55) 61 (78) no 41 (45) 17 (22) Status of residual 0.015 tumour burden No residual 26 (28) 30 (38) microscopic 17 (18) 26 (33) <1 cm 8 (9) 3 (4) >1-2 cm 8 (9) 7 (9) >2 cm 33 (36) 13 (16) CA125 response 0.036 No response 5 (6) 3 (4) Complete response 65 (76) 64 (89) Partial response 10 (12) 3 (4) Steady response 0 (0) 2 (3) Progressive response 5 (6) 0 (0) Tumour type 0.02 Clear cell carcinoma 5 (7) 13 (20) Non clear carcinoma 68 (93) 52 (80) Platinum sensitivity sensitive 52 (60) 63 (82) 0.002 resistance 35 (40) 14 (18)

(82) In conclusion, SPAG5 is an important novel gene implicated in the survival of BC & OVC cells. Its protein expression is an independent predictor for Anthracycline/cisplatinum CT. SPAG5 may provide new avenues for the discovery of new predictive marker to guide therapeutic intervention.