WO2013149904A1

WO2013149904A1 - Marker gene based diagnosis, staging and prognosis of prostate cancer

Info

Publication number: WO2013149904A1
Application number: PCT/EP2013/056510
Authority: WO
Inventors: Kyrylo Litovkin; Aleyde Van Eynde; Mathieu Bollen; Monique Beullens; Olivier Gevaert
Original assignee: Katholieke Universiteit Leuven
Priority date: 2012-04-06
Filing date: 2013-03-27
Publication date: 2013-10-10
Also published as: WO2013041731A1; GB201206209D0

Abstract

This invention relates generally to a method of diagnosis and prognosis, in particular staging and/or typing and/or predicting outcome, for distinguishing between a benign prostate hyperplasia and a prostate cancer and between an hormone sensitive and an hormone refractory prostate cancer condition and specifically to identification of differentially methylated CpG islands in the regulatory regions surrounding the transcriptional start site of at least one marker gene of the present invention as a diagnostic and/or prognostic indicator of prostate cancer (PrCa) and for distinguishing androgen-refractory from androgen sensitive prostate cancer. The marker genes of the present invention comprise PITX2, and HOXD3. This invention relates more specifically to the detection of hypermethylation of PITX2, and HOXD3. This invention further relates to the prediction, prognosis or diagnosis of prostate cancer, including metastasis, more particularly in patients with prostate cancer. Marker genes have been identified of which promoter regions containing differentially methylated regions, compared to a reference sample, which are indicative for the prediction or prognosis of prostate cancer.

Description

Marker gene based diagnosis, staging and prognosis of prostate cancer

FIELD OF THE INVENTION

This invention relates generally to a method of diagnosis and prognosis, in particular staging and/or typing and/or predicting outcome, for distinguishing between a benign prostate hyperplasia and a prostate cancer and between an hormone sensitive and an hormone refractory prostate cancer condition and specifically to identification of differentially methylated CpG islands in the regulatory regions surrounding the transcriptional start site of at least one marker gene of the present invention as a diagnostic and/or prognostic indicator of prostate cancer (PrCa) and for distinguishing androgen-refractory from androgen sensitive prostate cancer.

The marker genes of the present invention comprise PITX2 and HOXD3. This invention relates more specifically to the detection of hypermethylation of PITX2 and HOXD3.

This invention further relates to the prediction, prognosis or diagnosis of prostate cancer, including metastasis, more particularly in patients with prostate cancer. Marker genes have been identified of which promoter regions containing differentially methylated regions, compared to a reference sample, are indicative for the prediction or prognosis of prostate cancer.

BACKGROUND OF THE INVENTION

Prostate cancer (PrCa), is the second most common malignancy in males worldwide after lung cancer, and the third leading cause of cancer death in men. Early detection greatly improves survival rates. If the malignant prostate tumor is still local, PrCa can be successfully treated by radiation therapy, surgery, hormone therapy and/or chemotherapy. Unfortunately, if the PrCa invades other parts of the body like bones, lymph nodes, rectum and bladder (metastatic PrCa), it becomes refractory to hormone therapy. For this advanced PrCa the prognosis is poor. Currently, PrCa is detected by an elevated level of Prostate- Specific Antigen (PSA) in the blood, along with a digital rectal exam. The PSA test is also used to monitor patients for the recurrence of PrCa following surgery or other treatments. However, although the PSA test has greatly improved the detection of PrCa, its usefulness is still controversial. A recent study by Concato et al. shows that PSA screening is not associated with lower mortality (Concato J, et al. (2006) Arch Intern Med. 166:38-43). Moreover, the serum PSA level is also elevated in non-cancerous prostate disorders such as benign prostate hyperplasia and infection. Initial tests for suspected prostate cancer is done by analysis of blood levels of proteins like PSA or for instance PSP94 protein. Positive tests are followed by a conformational diagnosis. The only test which can fully confirm the diagnosis of prostate cancer is a biopsy, the removal of small pieces of the prostate for microscopic examination. The present invention provides a novel diagnostic test of prostatic tissue or cells obtainable from prostatic tissue. A condition of benign prostatic hyperplasia (BPH), or benign prostatic hypertrophy is common as a man ages. It is thus very important to distinguish between a PrCa and a BPH. Moreover hormone-refractory prostate cancers are more aggressive and need specific treatments such as apoptosis and regression induction of the tumors and/or antimetastasis.

Besides or in addition to the presently used prostate cancer nomograms, there is thus a need in the art for additional prostate cancer screening or diagnosis methods, and more particularly for a biomarker(s) that can discriminate between benign and malignant tumors and between aggressive and indolent (slow-growing) cancers. The present invention fulfills these needs.

The latter type of cancer will remain localized in a person's lifetime and is unlikely to reduce life expectancy. In contrast, an aggressive cancer is more lethal, due to metastasis, and requires immediate intervention. Therefore, there is an unmet need for a reliable diagnostic assay and biomarker(s) to distinguish between these two types.

We have now found that the combined detection of hypermethylation of the regulatory region surrounding the transcription start site (TSS) of HOXD3 and PITX2 is predictive of biochemical recurrence, and/or indicative of the type and/or stage of a prostate cell proliferative disorder, and can thus be considered to allow discrimination between different types and/or stages of prostate cancer. Although PITX2 (Weiss et al, J. Urol; 181(4) 1678- 1685; 2009) and HOXD3 (Kron et al, Laboratory Investigation; 90. 1060-1067; 2010), are known prognostic markers of biochemical recurrence (BCR) in prostate cancer, we have now found that the combined detection of hypermethylation of both genes is an even better predictor for biochemical recurrence, compared to either gene alone. In a Multivariate Cox Proportional Hazard Model, neither PITX2 nor HOXD3 on their own added prognostic information to known clinicopathological parameters, however, the two-gene PITX2/HOXD3 methylation model unexpectedly added independent prognostic information to parameters such as pre -operative PSA, final Gleason score and surgical margin status. Finally, PITX2 was found to be an independent significant predictor for clinical failure in the context of preoperative PSA, pathological stage, final Gleason score, surgical margin status, lymph-node invasion, adjuvant radiotherapy (RT) and adjuvant hormonal therapy (ADT). Clinical failure is a harder endpoint for prostate cancer than biochemical recurrence.

SUMMARY OF THE INVENTION The present invention solves the problems of the related art by providing a diagnostic and prognostic assay that allows one to determine the predisposition to, or the incidence of prostate cancer and allows to distinguish between different types and/or stages of cancer, in particular between hormone-refractory and hormones-sensitive cancer, particularly in prostatic tissues or cells originating from prostatic tissues and to predict outcome, in particular of biochemical recurrence in prostate cancer. However, the test could also be used on body fluids.

In accordance with the purpose of the invention, as embodied and broadly described herein, the invention is broadly drawn to methods and assays for detecting a prostate proliferative disorder, in particular for identifying prostate tumor cells that have become refractory or resistant to hormone therapy, and thus allowing to identify the prostate cancer or/and to distinguish hormone sensitive from hormone refractory prostate cancers.

The present invention relates generally to the identification of the distinguishing difference between a hormone refractory prostate tissue cellular proliferative disorder and a hormone sensitive prostate tissue cellular proliferative disorder in a subject, preferably a human subject. The distinguishing difference relies on the identification of one or more hypermethylated CpG islands surrounding the transcription start site (TSS) of the human genes PITX2 and HOXD3, more in particular the hypermethylated CpG islands are found in regions upstream of the TSS or in the promoter region of said human genes. In particular embodiments said group of genes (PITX2, and HOXD3) are to be analysed for their methylation status, and are used to predict the incidence of and more particular the aggressiveness of prostate cancer. These markers: PITX2 and HOXD3, are thus particularly useful as prognostic markers, more specifically in the current invention hypermethylation of PITX2, and HOXD3 are indicative for a negative prognosis, or an indication for an aggressive tumor, more particular a prostate tumor. In a certain embodiment of this invention, this set of genes (PITX2, and HOXD3) can be used in a method of the invention to decide on the proper treatment or proper medicament of the patient. In a more particular embodiment, the method of the present invention wherein hypermethylation of PITX2 and HOXD3 is detected in a sample of prostatic tissue or in a biological sample that comprises prostatic cells or prostatic cell components from a human patient when comparing the methylation level of said genes in a reference sample, is used to decide on the proper treatment of said patient, in particular the methylation level of said genes is indicative for the decision about the initiation or continuation of a proper treatment, wherein in a more particular embodiment said proper treatment is selected from a prostatectomy, treatment with a methylation inhibitor, a gonadotropin-releasing hormone agonists, or treatment with a compound which reduces male hormones, radiotherapy, or treatment with neutraceuticals.

With regard to the prognostic markers of this invention, PITX2 and HOXD3 are hypermethylation markers, meaning that hypermethylation of PITX2 and HOXD3, when comparing the methylation status of a patient or a human being suspected to have prostate cancer, to the methylation level of said genes in a reference sample is indicative for prostate cancer or indicative for the predisposition to prostate cancer, more particularly for typing and/or staging tumors, in particular to identify an aggressive or high grade prostate cancer and for predicting outcome, in particular to predict biochemical recurrence (BCR) and/or clinical failure.

The prognostic methods that detect whether a prostate cancer in subjects, preferably human, comprises an androgen refractory cancer and/or an androgen sensitive cancer can be carried out by analysis of the methylation status of said genes in a sample of a subject.

Thus, in a first aspect the invention provides methods for detecting, and in particular for typing and/or staging and/or prediction of outcome; in a subject a prostate cell proliferative disorder, which methods comprise the steps of:

a) obtaining a biological sample from the subject; b) determining the methylation state of CpG island(s) upstream and/or downstream of the TSS and/or in the promoter region of the PITX2 and HOXD3 genes and wherein detection of hypermethylation of PITX2, and HOXD3 in these regions is indicative of the type and/or stage, the predisposition to, the incidence of prostate cancer, or to predict BCR. The latter in particular in case of a prostatectomy.

In a similar aspect, the invention provides methods for detecting in a subject an androgen refractory prostate cancer, which methods comprise the steps of:

a) obtaining a biological sample from the subject;

b) determining the methylation state of CpG island upstream and/or downstream of the TSS region and/or in the promoter region of PITX2, and HOXD3, in the subject's sample; and wherein detection of hypermethylation of said gene(s) is indicative of a predisposition to, or the incidence of, androgen sensitive prostate cancer. In a particular aspect, the present invention provides a method for typing and/or staging and/or predicting outcome of a prostate cell proliferative disorder in a human male subject, the method comprising:

a. analyzing the level of DNA methylation of the regulatory region surrounding the transcription start site (TSS) of the PITX2 and HOXD3 genes in a test sample of prostatic tissue or in a biological test sample that comprises prostatic cells or prostatic cell components from a human patient; and

b. comparing the methylation level of CpG dinucleotides in the regulatory region surrounding the transcription start sites of PITX2 and HOXD3 of step (a) in the test sample with said level in a reference sample;

wherein the methylation level of CpG dinucleotides in said regulatory region in the PITX2 and HOXD3 genes of said test sample is predictive for the outcome and/or indicative of the type and/or stage of said prostate cell proliferative disorder.

More in particular, step (b) of said method comprises analyzing the methylation level of CpG dinucleotides in the regulatory region surrounding the transcription start sites of PITX2 and HOXD3, using the following formula:

((% DNA methylation of PITX2)* (0.021 +/- 0.004)) + ((% DNA methylation of

HOXD3)*(0.003 +/-0.002)) Preferably, both the methods of the invention comprise a further step as follows: c) identifying methylation of region(s), wherein hypermethylation of PITX2, and HOXD3 is identified as being different when compared to the same region(s) of the gene or associated regulatory region in a subject having an androgen sensitive prostate cancer. Another aspect of the invention is that it provides methylation conditions of regulatory regions of the panel of genes (PITX2, and HOXD3), such as in the CpG islands surrounding the TSS of said panel of human genes, which can be used (a) to analyze the presence of cancer cells in prostate tissue and/or in prostatic secretions, for instance in seminal plasma and (b) to define patients that have a prostate cancer or alternatively patients that have a normal prostate, and (c) to define which patients with a prostate cancer have an androgen refractory prostate cancer or alternatively to define which patients with a prostate cancer have an hormone sensitive prostate cancer.

Such test provides an accurate means or tool to decide about the suitable treatment of the prostate cancer; in particular if the PITX2 and HOXD3 genes are hypermethylated the need for chemotherapy, surgery or radiation therapy is identified. The methods of present invention can also be used to predict effectiveness of such chemotherapies applicable on a prostate cancer.

Patients affected by a condition of hypermethylation of regulatory regions of the genes PITX2 and HOXD3 such as in the CpG islands surrounding said genes, and/or CpG islands upstream of the TSS or in the promoter region of said genes can for instance be treated by DNA methyltransferase (DNMT) inhibitors.

Diagnosis of hypermethylation of the CpG island in the regions surrounding the TSS or in the promoter of PITX2 and HOXD3 can thus be used as a decision tool for treatment of a patient affected with such hypermethylation with a therapeutically effective amount of an DNA methyltransferase (DNMT) inhibitor for treating the prostate cancer or for preventing that a androgen sensible prostate cancer evolves into an androgen refractory prostate cancer. MGI Pharma developed small molecule DNA methyltransferase (DNMT) inhibitors for the treatment of cancer. Short oligonucleotide DNA methylation inhibitors in the art are Decitabine 5-Aza-CdR, SI 10 AzapG, S53 GpAza, S54 GpAzapG, S55 AzapGpAzapG, S56 pGpAzapAzapG, S52R AzapsG, Zebularine and SI 12 HEGpAzapG. A specific DNMT inhibitor is for instance the compound called SI 10 of the company SuperGen which is a dinucleotide containing decitabine, S 110, which has superior activity due to increased stability because of less degradation by hydrolytic cleavage and deamination. This is a DNA demethylating agent with a similar activity as decitabine (5-aza-2'-deoxycytidine) or its derivatives. Decitabine is a potent DNA methylation inhibitor which is approved in the US for the treatment of myelodysplasia syndromes (Yoo DB, et al. Cancer Research. 67: 6400- 6408, No. 13, 1 Jul 2007). Another DNA methyltransferase (DNMT) inhibitor is MG 98 (HYB 101584) is described in US 6953783 and US 6506735. MG 98 is a second generation antisense oligonucleotide that selectively targets DNA methyltransferase 1 (DNMT1) mRNA. By inhibiting the production of DNMT, the methylation of DNA is reversed and leads to re- expression of the tumour suppression genes. MG 98 is created by MethylGene Inc. (Stewart D, et al. 11th NCI-EORTC-AACR symposium on new drugs in cancer therapy. : 148, 7 Nov 2000. ; Winquist E, et al. European Journal of Cancer. 38 (Suppl. 7): 141, Nov 2002. ; Stewart DJ, et al. Annals of Oncology. 14: 766-774, May 2003 and Ramchandani S, et al. Proceedings of the National Academy of Sciences of the United States of America. 94: 684- 689, Jan 1997. and Davis AJ, et al. 11th NCI-EORTC-AACR symposium on new drugs in cancer therapy. : 94, 7 Nov 2000. These compounds can be administered in a therapeutically efficient amount to patients that have been identified by the diagnostic method of present invention to be in need thereof. Thus, epigenetic loss of gene function due to hypermethylation can be rescued by the use of DNA demethylating agents and/or DNA methyltransferase inhibitors and/or HDAC inhibitors. Accordingly, the invention also provides for a method for predicting the likelihood of successful treatment of prostate proliferative disorder or prostate cancer, with a DNA demethylating agent and/or a DNA methyltransferase inhibitor and/or HDAC inhibitor comprising detecting a methylation change in the region surrounding the TSS or the promoter region of PITX2 and HOXD3 wherein detection of the methylation change is indicative of successful treatment to a higher degree than if the methylation modification is not detected.

Also provided is a kit for typing and/or staging and/or predicting outcome, detecting a predisposition to, or detecting the incidence of, prostate cancer in a sample comprising:

(a) means for detecting a methylation change in the region surrounding the TSS or the promoter region of PITX2, and HOXD3.

(b) means for processing a sample derived from the prostate. In particular embodiments the analysis of said genes is restricted to the region surrounding their TSS. In particular embodiments said region extends from 1.5 kb upstream to about 1.5 kb downstream from the transcription start site of said genes. In other particular embodiments, said region extends from 1.0 kb upstream to about 1.0 kb downstream from the transcription start site of said genes.

More in particular, the regulatory region surrounding the transcription start site of the gene PITX2 preferably corresponds to position about -16 to about +73 of the transcription start site. Furthermore, the regulatory region surrounding the transcription start site of the gene HOXD3 preferably corresponds to position about 909 to 823 upstream of the transcription start site.

In more particular embodiments of this invention the detection of hypermethylation in said region of PITX2, and HOXD3 indicates the presence of prostate cancer cells or is indicative of a predisposition to, or the incidence of, prostate cancer. In a particular embodiment of this invention, said hypermethylation is detected when comparing the methylation status of the DNA of a test sample to the methylation status of a control sample and/or a benign prostate hyperplasia sample.

In the context of the present invention, hypermethylation (and/or hypomethylation) of the (marker) genes of this invention has the meaning of differential methylation i.e. hypermethylation (increased) and/or hypomethylation (decreased) of said genes, when compared to the methylation status of said genes in a reference or control sample. Iso- methylation of the (marker) genes of this invention has the meaning of substantially the same methylation level of said genes, when compared to the methylation status of said genes in a reference or control sample, i.e. the said genes are not differentially methylated compared to the methylation status of said genes in a reference or control sample.

In particular embodiments of this invention, the control sample or reference sample is a sample from a healthy prostate. In other particular embodiments of this invention, the control sample or reference sample is a sample from a benign hyperplasia substrate.

In specific embodiments of this invention, the method of this invention comprises PCR analysis of polynucleotide materials of the cells derived from prostatic tissue. In other particular embodiments of this invention, the method of this invention comprises PCR analysis of polynucleotide materials of the cells derived from prostatic fluid.

An embodiment of the present invention is a method of diagnosing a disease state or cell proliferative disorder in the prostate of a subject, said method comprising: (a) analyzing the level of DNA methylation of the regulatory region surrounding the transcription start site (TSS) of PITX2, and HOXD3 or an homologous sequence of said genes in a biological sample isolated from said subject, and (b) comparing said DNA methylation with the DNA methylation in a control sample and/or a benign prostate hyperplasia sample; whereby increased CpG methylation in PITX2, and HOXD3 relative to the control sample or the benign prostate hyperplasia sample in the regulatory regions surrounding the transcriptional start site of said genes, in particular as defined hereinbefore, is an indication for prostate cancer and/or an indication of an hormone refractory prostate cancer, androgen-independent prostate cancer (AIPC) or androgen-independent metastatic prostate cancer.

Further specific embodiments of these previous methods of diagnosis, typing and/or staging can be :

The previous method further comprising a step of analyzing histone (de)acetylation of the gene(s) of this invention in said sample.

The previous method whereby the disease state or cell proliferative disorder is a cancer. The previous method to distinguish between a healthy prostate and a disordered or diseased prostate.

The previous method to distinguish between a benign prostate hyperplasia and a prostate cancer

The previous method to distinguish between an hormone sensitive prostate cancer and an hormone refractory prostate cancer.

The previous method to distinguish between an androgen sensitive prostate cancer or androgen dependent prostate cancer and androgen-independent prostate cancer (AIPC) The previous method to discover an androgen-independent metastatic prostate cancer in a prostate cell or prostate tissue.

The previous method to carry out a prostate cancer grading or prostate cancer staging. The previous method to decide on the proper treatment or proper medicament of the prostate disease state

The previous method to decide on the treatment with a pharmaceutically acceptable DNA methylation inhibitor

The previous method to decide on the treatment with a pharmaceutically acceptable HDAC inhibitor.

The previous method to decide on the treatment to decrease the activity of the EZH2 protein The previous method to decide on the treatment with a DNA demethylating agent and/or a DNA methyltransferase inhibitor and/or HDAC inhibitor.

The previous method to decide on a prophylactically effective amount of a nutraceutical to treat a subject with a prostate disease status.

Numbered statements of the invention are as follows:

1. A method for typing and/or staging and/or predicting outcome of a prostate cell proliferative disorder in a human male subject, the method comprising:

2. The method according to statement 1, wherein step (b) comprises analyzing the methylation level of CpG dinucleotides in the regulatory region surrounding the transcription start sites of PITX2 and HOXD3, using the following formula:

((% DNA methylation of PITX2)* (0.021 +/- 0.004)) + ((% DNA methylation of HOXD3)*(0.003 +/-0.002))

3. The method according to statements 1 or 2, wherein hypermethylation of CpG dinucleotides in the regulatory region surrounding the transcription start site of PITX2 and HOXD3 is predictive for the outcome and/or indicative of the type of a prostate cell proliferative disorder as a more aggressive type and/or of a further advanced stage. The method according to anyone of statements 1-3 wherein said prostate cell proliferative disorder of a more aggressive type is a biochemical recurrent type of disorder.

The method according to anyone of statements 1-4 wherein said prostate cell proliferative disorder of a more advanced stage is a high clinical stage disorder of pT stage III or IV.

The method of any of statements 1 to 5, for use in deciding on the proper treatment or proper medicament dependent on the type and/or stage of said prostate cell proliferative disorder.

The method of statement 6, wherein detection of hypermethylation of PITX2 and HOXD3, is indicative for the decision about the initiation or continuation of a treatment selected from a prostatectomy, a histone deacetylase inhibitors, a DNA methylation inhibitor, a gonadotropin-releasing hormone agonists, a neutraceutical, radiotherapy, or a compound in an effective amount to reduce male hormones.

The method of any of statements 1 to 7, wherein methylation is determined using PCR analysis, bisulfite genomic sequencing PCR analysis, Methylation-Specific PCR analysis or an equivalent amplification technique.

The method of any of statements 1 to 8, wherein methylation is determined in an assay comprising primers for assessing the presence of methylation in a regulatory region surrounding the TSS of said genes.

The method of any of statements 1 to 9, wherein at least one primer of the group consisting of methylated specific primers for PITX2 (SEQ ID N° 67 and 68), and at least one primer of the group consisting of methylated specific primers for HOXD3 (SEQ ID N° 61 and 62) and at least one primer of the group consisting of unmethylated specific primers for PITX2 (SEQ ID N° 69 and 70) and at least one primer of the group consisting of unmethylated specific primers for HOXD3 (SEQ ID N° 63 and 64) is used. The method of any of statements 1 to 10, wherein the reference sample is a sample from a healthy individual or from an individual having a typical benign hyperplasia prostate.

The method of any of statements 1 to 11, wherein said regulatory region surrounding the TTS comprises one or more CpG islands and extends about 1.5 kb upstream to about 1.5 kb downstream from said transcription start site of said gene(s). The method according to statement 12, wherein the regulatory region surrounding the transcription start site of the gene PITX2 corresponds to position about -16 to about +73 of the transcription start site. The method according to statement 12, wherein the regulatory region surrounding the transcription start site of the gene HOXD3 corresponds to position about 909 to 823 upstream of the transcription start site. The method of any of statements 1 to 14, wherein the test and/or reference sample is selected from the list comprising prostatic tissue, prostatic fluid, seminal fluid, ejaculate, blood, urine, prostate secretions, histological slides, and paraffin-embedded tissue. The method of any of statements 1 to 15, further comprising analysing the level of DNA methylation of control plasmids comprising the inserts represented by SEQ ID N° 98 and 99. A kit for typing and/or staging a prostate cell proliferative disorder in a human male subject, comprising at least one primer of the group consisting of methylated specific primers for PITX2 (SEQ ID N° 67 and 68), and at least one primer of the group consisting of methylated specific primers for HOXD3 (SEQ ID N° 61 and 62) and at least one primer of the group consisting of unmethylated specific primers for PITX2 (SEQ ID N° 69 and 70) and at least one primer of the group consisting of unmethylated specific primers for HOXD3 (SEQ ID N° 63 and 64) is used. The kit according to statement 17 further comprising control plasmids comprising the inserts represented by SEQ ID N° 98 and 99. Further scope of applicability of the present invention will become apparent from the detailed description given hereinafter. However, it should be understood that the detailed description and specific examples, while indicating preferred embodiments of the invention, are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description. It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed.

DESCRIPTION OF THE FIGURES Figure 1 : Figure 1A is a schematic representation of the strategy we have chosen for the analysis of methylation status of different potential prostate cancer biomarkers. For the fast estimation of promoter methylation a melting curve assay was applied to a set of model genotypes corresponding to non-cancerous DNA and tissues (human genomic DNA (HG DNA), cell lines PZ-HPV7, BPH1) and PCa cell lines (androgene-sensitive LNCaP and androgene-insensitive PC-3 and DU 145). Melting curve assay implies amplification of a part of the gene promoter with methylation independent primers after bisulphite conversion of DNA, followed by registration of melting profile of the resulting amplicons. Differences in melting profile help to discriminate between initially methylated and nonmethylated DNA templates. Results of the analysis of the APC gene are shown in Fig. IB. Figure 2: Results of the bisulphite sequencing of the PCR- fragments covering MAGEA2 CpG- island around the transcription site, obtained from whole blood human genomic DNA (HG DNA) and LNCaP cell line. Numbers from -2 to 5 and from 12 to 16 represent positions of CpG-dinucleotides relative to the transcription start site (TSS). The selected CpG-island is completely hypomethylated in cancer LNCaP cell line in comparison with whole blood human genomic DNA.

Figure 3: Graphs showing PCR cycles (X-axis) plotted against the fluorescence intensity of the PCR product accumulated in EvaGreen® reaction mixture (Y-axis) using 100% M and 100% U plasmid standards as a template tested with M and U primers for APC. The M and U reactions were 100%) specific since M primers did not cross-react with U standard and vice versa. No primer dimers were observed in "no template" control. Specificity of primers is confirmed by melting curve analysis: melting temperature of the PCR product generated from 100% M template is shifted to the right and corresponds to that of the completely methylated sequence, as indicated by the melting peak; melting temperature of the PCR product generated from 100% U template is shifted to the left and corresponds to that of the completely unmethylated sequence, as indicated by the melting peak. Figure 4: Graphs showing validation of primers for qMSP and qUSP using APC methylated (M) and unmethylated (U) primer sets to amplify serially diluted plasmid standards in EvaGreen® PCR mixture. Efficency 0.95 - 1.00 reflects a 2-fold amplification of DNA per cycle. The correlation coefficient (R^A2) shows linearity (0.998) over the range of DNA concentrations. Figure 5 : Procedure of validation of primer sets for the two-step Quantitative multiplex nested-MSP assay.

Figure 6: Schematic representation of the protocol for quantitative multiplex nested-MSP analysis. In Multiplex PCR step a mixture of gene-specific methylation-independent primer pairs is used to co-amplify 80-180 bp fragments of CpG islands covering regulatory elements of the selected genes. In the quantitative step two real time PCRs (qMSP and qUSP) are performed for each gene separately with primer sets specific for methylated (M) and unmethylated (U) sequences using the DNA template derived from Multiplex PCR step (diluted 1 :500 in sterile water). Serial dilutions 3x 10² - 3x 10⁶ of plasmid standards with cloned gene fragments corresponding to completely methylated (M) and completely unmethylated (U) sequences are used to generate separate standard curves and quantify M and U fragments. The percent methylation for each gene in the panel is calculated as %M = [M / (U + M)] 100.

Figure 7: Gray-scale representation of the levels of 16 genes promoter hypermethylation as determined by the invention methods employing quantitative multiplex nested-MSP on prostate cell lines, prostate tissues and HG DNA. Intensity of color correlates with the degree of methylation, also indicated by number (%). For the TDRD1 and MAGEA2 genes reverse methylation value is presented (100 - % of methylation).

Figure 8: Gray-scale representation of the levels of 16 genes promoter hypermethylation as determined by the invention methods employing quantitative multiplex nested-MSP on matched tumor/adjacent benign prostate tissue samples from 7 patients. Intensity of color correlates with the degree of methylation, also indicated by number (%). For the TDRDl and MAGEA2 genes the reverse methylation values are presented (100 - % of methylation).

Figure 9: Average tumor volume measured (A) in ml and (B) in % of the total prostate gland volume in patients with low (LM) and high (HM) methylation of tumor DNA. Low and high methylation levels are discriminated based on the median methylation value for each gene. Study group: PCa2 cohort (N=63). Genes: 1 - RARB, 2 - GSTP1, 3 - CCND2, 4 - PTGS2, 5 - APC, 6 - LGALS3, 7 - TDRDl, 8 - RASSFI, 9 - PITX2, 10 - HOXD3, 11 - CDH13. The differences are significant at * P < 0.05, ** P < 0.01 and *** P < 0.001.

Figure 10: Results of Kaplan-Meier analysis for biochemical progression-free survival probability within 16 years after radical prostatectomy in groups of patients with high (HM - above the cutoff methylation value) and low (LM - below the cutoff methylation value, indicated on the graph) degree of HOXD3 (A) and TDRDl (B) methylation in PCa tumors from the PCa2 cohort.

Figure 11 : Results of Kaplan-Meier analysis for biochemical progression-free survival probability within 16 years after radical prostatectomy in groups of patients with high (HM - above the cutoff methylation value) and low (LM - below the cutoff methylation value, indicated on the graph) degree of PITX2 (A) and RASSFI (B) methylation in PCa tumors from the PCa2 cohort.

Figure 12: Results of Kaplan-Meier analysis for clinical failure (CF) probability within 16 years after radical prostatectomy in groups of patients with high (HM - above the cutoff methylation value) and low (LM - below the cutoff methylation value, indicated on the graph) degree of PITX2 methylation in PCa tumors from the PCa2 cohort.

DETAILED DESCRIPTION OF THE INVENTION As used herein, the singular forms "a", "an" and "the" include plural referents unless the context clearly dictates otherwise.

"Disease state" as used herein means any disease, disorder, condition, symptom, or indication. As used herein, the term "cell proliferative disorder" refers to conditions in which the unregulated and/or abnormal growth of cells can lead to the development of an unwanted condition or disease, which can be cancerous or non-cancerous. The detection of the cell proliferative disorder may be by way of routine examination, screening for a cell proliferative disorder or pre-stadia such cell proliferative disorder, monitoring and/or staging the state and/or progression of the cell proliferative disorder, assessing for recurrence following treatment, and monitoring the success of a treatment regimen. In a preferred embodiment, the cell proliferation disorder is cancer.

As used herein, the term "cancer" concerns malignant neoplasm, malignant tumor or invasive tumor and also can include solid neoplasm or solid tumors cancers. Cancers are classified by the type of cell that resembles the tumor and, therefore, the tissue presumed to be the origin of the tumor. Examples of general categories include: Carcinoma: Malignant tumors derived from epithelial cells. This group represents the most common cancers, including the common forms of breast, prostate, lung and colon cancer. Sarcoma: Malignant tumors derived from connective tissue, or mesenchymal cells. Lymphoma and leukemia: Malignancies derived from hematopoietic (blood-forming) cells Germ cell tumor: Tumors derived from totipotent cells. In adults most often found in the testicle and ovary; in fetuses, babies, and young children most often found on the body midline, particularly at the tip of the tailbone; in horses most often found at the poll (base of the skull). Blastic tumor: A tumor (usually malignant) which resembles an immature or embryonic tissue. In a preferred embodiment, the cancer is prostate cancer.

"Hormone refractory prostate cancer" and in particular "androgen-independent prostate cancer (AIPC)" has to be understood for the meaning of this invention as prostate cancer that has become refractory, that is, it no longer responds to hormone therapy. "Prostate cancer grading" or "typing" as used herein means describing how abnormal or aggressive the cancer cells appear. The grade/type helps to predict long-term results, response to treatment and survival. In the art there is for instance the Gleason scale that is the most common scale used for grading prostate cancer. This system assigns cancer cells a score from 1 to 10, by combining the two most common patterns of cells to give a total score (i.e., 3 + 4 = grade 7). Scores generally range between 4 and, most commonly, 6 or 7. These scores are broken down into three main levels: Low-grade (well differentiated): This type of slow- growing cancer has an appearance most like normal prostate cells and is the least dangerous. It has a Gleason score of 4 or less. Intermediate grade (moderately differentiated): This type is somewhere between the low- and high-grade cancers and the most common of the three. Depending on PSA level and tumor volume, it can act like a high- or low-grade cancer. It has Gleason score between 4 and 7. High-grade (poorly differentiated): This type of cancer has an appearance least like normal prostate cells. It is the most deadly since it is very aggressive and grows very fast - even into surrounding areas - such as lymph nodes and bones. These cancer cells also tend to be large, hard to treat, and reappear more frequently. They have a Gleason score between 8 and 10 (Antoinette S Perry et al Endocrine-Related Cancer 13 (2) 357-377).

"Prostate cancer staging " or "staging" as used herein concerns how much and where the cancer is located. The more cancer there is in the body, the more likely it is to spread and less likely that treatments will work. Therefore, the more advanced stages can affect long-term results and survival. According an older prostate cancer staging the prostate cancer is broken down into four primary stages for instance the four ABCD stages of staging to gauge the severity of prostate cancer to describe the detection and location of the cancer. Stage A: Cancer found when not suspected or due to a high PSA level, Stage B: Cancer found due to abnormal digital rectal exam and is held in the prostate, Stage C: Cancer that has spread to the tissues outside of the prostate, Stage D: Cancer that has spread to the lymph nodes or bone. A particular system in the art which replaced the ABCD staging system of prostate cancer to give an even more accurate description of the cancer is the TNM grading system. "T" describes the tumor and uses different numbers to explain how large it is; "N" stands for nodes and tells whether the cancer has spread to the lymph nodes; "M" means metastatic, and tells whether the cancer has spread throughout the body. There are various T Status stages : Stage Tl : Microscopic tumor confined to prostate and undetectable by a digital rectal exam (DRE) or ultrasound; Stage Tla: Tumor found in 5% or less of prostate tissue sample; Stage Tib: Tumor found in more than 5% of a prostate tissue sample; Stage Tic: Tumor is identified by needle biopsy as a follow-up to screening that detected elevated PSA results; Stage T2: Tumor confined to prostate and can be detected by DRE or ultrasound; Stage T2a: Tumor involves less than half of one lobe of the prostate, and can usually be discovered during DRE exam; Stage T2b: Tumor involves more than half of one lobe of the prostate, and can usually be felt during DRE exam; Stage T2c: Tumor involves both lobes of the prostate and is felt during a DRE exam; Stage T3 : Tumor has spread to surrounding tissues or to the seminal vesicles; Stage T3a: Tumor has spread to outside of the prostate on only one side; Stage T3b: Tumor has spread to outside of the prostate on both sides; Stage T3c: Tumor has spread to one or both of the seminal tubes; Stage T4: Tumor is still within the pelvic region but may have spread to organs near the prostate, such as the bladder; Stage T4a: Tumor has spread beyond the prostate to any or all of the bladder neck, the external sphincter, and/or the rectum and Stage T4b: Tumor has spread beyond the prostate and may affect the levator muscles (the muscles that help to raise and lower the organ) and/or the tumor may be attached to the pelvic wall and various N Status stages : Stage NO: Cancer cells have spread, but not yet to pelvic lymph nodes; Stage Nl : Cancer cells have spread to a single lymph node in the pelvic area and are 2 cm (approximately 3/4 of one inch) or less in size; Stage N2: Cancer cells have spread either to a single lymph node and are more than 2 cm but less than 5 cm (approximately 2 inches) in size, or the prostate cancer cells are found in more than one lymph node and are no larger than 5 cm in size; Stage N3: Cancer cells have spread to the lymph nodes and are larger than 5 cm in size and various M Status stages: Stage MO: Cancer cells have spread, but only regionally in the pelvic area & Stage Ml : Cancer cells have spread beyond the pelvic area to other parts of the body (Dr F. H. Schroder et al. The Prostate Volume 21, Issue S4 , Pages 129 - 138, 20 Jul 2006).

As used herein, the term "effective amount" refers to an amount of a compound, or a combination of compounds, of the present invention effective when administered alone or in combination as an anti-proliferative agent. For example, an effective amount refers to an amount of the compound present in a formulation or on a medical device given to a recipient patient or subject sufficient to elicit biological activity, for example, anti-proliferative activity, such as e.g., anti-cancer activity or anti-neoplastic activity. The combination of compounds optionally is a synergistic combination. Synergy, as described, for example, by Chou and Talalay, Adv. Enzyme Regul. vol. 22, pp. 27-55 (1984), occurs when the effect of the compounds when administered in combination is greater than the additive effect of the compounds when administered alone as a single agent. In general, a synergistic effect is most clearly demonstrated at sub-optimal concentrations of the compounds. Synergy can be in terms of lower cytotoxicity, or increased anti-proliferative effect, or some other beneficial effect of the combination compared with the individual components.

"A therapeutically effective amount" as used herein means the amount of a compound that, when administered to a mammal for treating a disease, is sufficient to effect such treatment for the disease. The "therapeutically effective amount" will vary depending on the compound, the disease and its severity and the age, weight, etc., of the mammal to be treated. A therapeutically effective amount of one or more of the compounds can be formulated with a pharmaceutically acceptable carrier for administration to a human or an animal. Accordingly, the compounds or the formulations can be administered, for example, via oral, parenteral, or topical routes, to provide an effective amount of the compound. In alternative embodiments, the compounds prepared in accordance with the present invention can be used to coat or impregnate a medical device.

The term "prophylactically effective amount" as used herein means an effective amount of a compound or compounds, of the present invention that is administered to prevent or reduce the risk of unwanted cellular proliferation. "Pharmacological effect" as used herein encompasses effects produced in the subject that achieve the intended purpose of a therapy. In one preferred embodiment, a pharmacological effect means that primary indications of the subject being treated are prevented, alleviated, or reduced. For example, a pharmacological effect would be one that results in the prevention, alleviation or reduction of primary indications in a treated subject. In another preferred embodiment, a pharmacological effect means that disorders or symptoms of the primary indications of the subject being treated are prevented, alleviated, or reduced. For example, a pharmacological effect would be one that results in the prevention or reduction of primary indications in a treated subject.

"Prostate biopsy" as used herein is a procedure in which small samples are removed from a man's prostate gland to be tested for the presence of cancer. It is typically performed when the scores from a PSA blood test rise to a level that is associated with the possible presence of prostate cancer.

A subject from which a biological sample can be obtained for analysis according to the invention is an animal such as a mammal, e.g. a dog, cat, horse, cow, pig, sheep, goat, primate, rat, or mouse. A preferred subject is a human being, particularly a patient suspected of having or at risk for developing a cell proliferative disorder such as a prostate cancer, or a patient with such a cell proliferative disorder such as a prostate cancer.

"Treating", includes any effect, e.g., lessening, reducing, modulating, or eliminating, that results in the improvement of the condition, disease, disorder, etc. "Treating" or "treatment" of a disease state includes: (1) preventing the disease state, i.e. causing the clinical symptoms of the disease state not to develop in a subject that may be exposed to or predisposed to the disease state, but does not yet experience or display symptoms of the disease state; (2) inhibiting the disease state, i.e., arresting the development of the disease state or its clinical symptoms; or (3) relieving the disease state, i.e., causing temporary or permanent regression of the disease state or its clinical symptoms.

"Prostatic cell component(s)" mean any part of or component of a prostatic cell or cell compartment derived from a prostatic cell, including exosomes and preferably comprising DNA of said prostatic cell.

By "homologous sequence" is meant, a nucleotide sequence that is shared by one or more polynucleotide sequences, such as genes, gene transcripts and/or non-coding polynucleotides. For example, a homologous sequence can be a nucleotide sequence that is shared by two or more genes encoding related but different proteins, such as different members of a gene family, different protein epitopes, different protein isoforms or completely divergent genes, such as a cytokine and its corresponding receptors. A homologous sequence can be a nucleotide sequence that is shared by two or more non-coding polynucleotides, such as noncoding DNA or RNA, regulatory sequences, introns, and sites of transcriptional control or regulation. Homologous sequences can also include conserved sequence regions shared by more than one polynucleotide sequence. Homology does not need to be perfect homology (e.g., 100%), as partially homologous sequences are also contemplated by the instant invention (e.g., 99%, 98%, 97%, 96%, 95%, 94%, 93%, 92%, 91%, 90%, 89%, 88%, 87%, 86%, 85%, 84%, 83%, 82%, 81%, 80% etc.).

With the "regulatory region surrounding the transcription start site (TSS)" is meant a regulatory region located upstream or 5' to the TSS and/or a regulatory region around the TSS and/or a regulatory region located downstream or 3' to the TSS of the concerned gene. The location of the concerned region can vary from 15Kbp upstream to 15 Kbp downstream of the TSS. Thus the region under investigation may correspond to all or part of the promoter region of the concerned gene. Alternatively, the region under investigation corresponds an exon and/or intron region and/or TSS region of the concerned gene. The region of the concerned gene is preferably between about 1500 bp upstream and about 1500 bp downstream from the TSS of the gene. In a particular embodiment of this invention, said region extends from -1500 bp and +1500 bp from the TSS of the gene. Whenever in the present invention, reference is made to the position of the regulatory region surrounding the transcription start site of a gene, said position is determined based on the information available in genome version hgl9 (Genome Reference Consortium).

The term "promoter" refers to the regulatory region located upstream, or 5' to the structural gene and/or TSS. Such a region extends typically between approximately 5 Kb, 500 bp or 150 to 300 bp upstream from the transcription start site of the concerned gene. For all the genes of this invention, we identified at least 1 CpG islands (genomic regions that contain a high frequency of CG dinucleotides) surrounding the transcriptional start site.

By "conserved sequence region" is meant, a nucleotide sequence of one or more regions in a polynucleotide does not vary significantly between generations or from one biological system, subject, or organism to another biological system, subject, or organism. The polynucleotide can include both coding and non-coding DNA and RNA.

By "control plasmid" is meant, a plasmid comprising a sequence insert, corresponding to methylated or unmethylated DNA sequence after bisulphite conversion, which is representative for the respective methylated or unmethylated state of its corresponding gene. Said insert is in particular selected from the list comprising SEQ ID N° 77-100, as depicted in table 13. Serial dilutions of plasmid standards with inserts corresponding to completely methylated (M) and completely unmethylated (U) sequences can be used to generate separate standard curves and quantify M and U fragments. The percent methylation for each gene in the panel can then be calculated as %M = [M / (U + M)] x 100.

The genes of this invention comprise all the genes, variants, alternative names of said genes that are known to a person skilled in the art. Some of the alternative names of the genes of this invention are summarized here below.

Description of the genes of this invention:

Accordingly, in a first aspect the invention provides a method for typing, staging, predicting outcome and/or identifying a prostate cell proliferative disorder in a human male subject, the method comprises:

- providing a sample of prostatic tissue and/or biological fluid of the prostate from a human patient susceptible to a prostate cancer and,

- analyzing the sample for the presence of methylation (on CpG dinucleotides) in a regulatory region surrounding the transcription start site or promoter region of certain genes of the panel of genes of this invention (PITX2, and HOXD3),

wherein the presence of hypermethylation in this region in PITX2, and HOXD3 relative to the methylation status of the corresponding region in the control sample or the benign prostate hyperplasia sample is indicative for typing, staging, predicting outcome and/or identifying of prostate cancer.

Preferably, the method comprise the steps of:

a) obtaining a biological sample from the subject;

b) determining the methylation state of CpG island upstream and/or downstream of the TSS region and/or in the promoter region of the panel of genes of this invention (PITX2, and HOXD3); and

c) identifying hypermethylation of the region(s), wherein hypermethylation (on CpG dinucleotides) is identified as being different when compared to the same region(s) of the gene or associated regulatory region in a subject not having the prostate cellular proliferative disorder,

wherein detection of hypermethylation in PITX2, and HOXD3 is indicative of a predisposition to, predicting outcome of, or the incidence of, prostate cancer.

In a particular embodiment, the present invention provides a method for typing and/or staging and/or predicting outcome of a prostate cell proliferative disorder in a human male subject, the method comprising:

a. analyzing the level of DNA methylation of the regulatory region surrounding the transcription start site (TSS) of the PITX2 and HOXD3 genes in a test sample of prostatic tissue or in a biological test sample that comprises prostatic cells or prostatic cell components from a human patient; and b. comparing the methylation level of CpG dinucleotides in the regulatory region surrounding the transcription start sites of PITX2 and HOXD3 of step (a) in the test sample with said level in a reference sample;

More in particular step (b) of the above methods comprises analyzing the methylation level of CpG dinucleotides in the regulatory region surrounding the transcription start sites of PITX2 and HOXD3, using the following formula:

The sample for use in such methods can be any suitable sample such as prostatic tissue, prostatic fluid, seminal fluid, ejaculate, blood, urine, prostate secretions, histological slides, and paraffin-embedded tissue, and is preferably a tissue sample. Prostate biopsy is a procedure in which small samples are removed from a man's prostate gland to be tested for the presence of cancer. It is typically performed when the scores from a PSA blood test rise to a level that is associated with the possible presence of prostate cancer. A biopsy thus provides a specific example of a biological sample for use in present methods. Examination of the condition of the prostate may be performed transrectally, through the ureter or through the perineum. The most common procedure is transrectal, and may be done with tactile finger guidance, (Ghei, M; Pericleous S et al (2005 Sep). Ann R Coll Surg Engl 87 (5): 386-7.) or with ultrasound guidance. If cancer is suspected, a biopsy is offered. During a biopsy tissue samples from the prostate are obtained for instance via the rectum. A biopsy gun can be used to insert and remove special hollow-core needles (usually three to six on each side of the prostate) in less than a second.

Suitable samples for diagnostic, prognostic, or personalized medicinal uses can be obtained from surgical samples, such as biopsies or surgical resection. However, other suitable samples for use in the methods of present invention comprise fine needle aspirates, paraffin embedded tissues, frozen tumor tissue samples, fresh tumor tissue samples, fresh or frozen body fluid. Examples of body fluids include prostatic fluids, blood samples, serum, plasma, urine, ejaculate, wash or lavage fluid. In fact, any tissue or fluid containing cells or nucleic acid, preferably DNA, derived from cells of the prostate is a suitable reagent for use in the methods of present invention. Present methods preferably also include the step of obtaining the suitable sample. Cells may need to be lysed for release of the nucleic acid. The nucleic acid may need to be cleared of proteins or other contaminants, e.g. by treatment with enzymes. The nucleic acid may also need to be concentrated prior to further use in the method of the invention, in particular when the nucleic acid is derived from bodily fluids.

Thus in a particular aspect, the present invention provides for an in vitro method for distinguishing a hormone independent proliferative disorder or hormone refractory proliferative disorder from a hormone sensitive proliferative disorder in tissue and/or in at least one cell obtainable from tissue of the prostate from a subject. Such prognostic/diagnostic method comprises contacting a DNA of a tissue or a DNA of a biological fluid with a reagent which detects the methylation status of the promoter region of PITX2, and HOXD3, wherein hypermethylation of PITX2, and HOXD3, as compared to the methylation status of the promoter region or upstream of the promoter region of said gene(s) from said group from a normal cell or compared to the methylation status of promoter region or upstream of the promoter region of said gene(s) from said group from cells or of tissue of a prostate with steroidal hormone sensitive proliferative disorder, is indicative of said steroidal hormone refractory proliferative disorder.

The test is particularly suitable to distinguish between hormone refractory and homone sensitive and in particular for androgen sensitive and androgen-refractory prostate proliferative disorders and to distinguish between benign prostate hyperplasia and prostate cancer.

In one embodiment, the invention provides a method for distinguishing between androgen sensitive and androgen-refractory prostate cancer by contacting a cellular component of a prostate tissue sample or another sample with a reagent which detects the methylation status of PITX2, and HOXD3 in the promoter or upstream of the promoter region of said gene(s).

As aforementioned, methylation sensitive restriction endonuclease can be utilized to identify a hypermethylated promoter or upstream region of the genes of this invention, for example.

Other approaches for detecting methylated CpG dinucleotide motifs use chemical reagents. In particular chemical reagents that selectively modify the methylated or non-methylated form of CpG dinucleotide motifs can be used in the methods of present invention. Such chemical reagents include bisulphite ions. Sodium bisulphite converts unmethylated cytosine to uracil but methylated cytosines remain unconverted. Analysis of the nucleic acid sequence after bisulfite conversion indicates if the original nucleic acid was all or not methylated.

Multiple techniques for analysing the methylation status of CpG dinucleotide motifs in CpG islands are known in the art. They comprise without limitation sequencing, methylation- specific PCR (MS-PCR), McMS-PCR, MLPA, QAMA, MSRE-PCR, MethyLight, HeavyMethyl, ConLight-MSP, BS-MSP, COBRA, McCOBRA, MS-SNuPE, MS-SSCA, PyroMethA, MALDI-TOF, MassARRAY, ERMA, QBSUPT, MethylQuant, Quantitative PCR sequencing, oligonucleotide-based microarray systems, Pyrosequencing, and Meth- DOP-PCR. A review of techniques for the detection of the methylation state of a gene is given for instance in Oral Oncology, 2006, Vol. 42, 5-13 and references cited therein.

A preferred technique for the detection and/or quantification of methylated DNA is the Methylation Specific PCR (MSP) technique. This technique can be used in end-point format, wherein the presence of methylated DNA is for instance detected by electroforesis or by the use of dyes such as SYBR Green I or Ethidium Bromide that bind double-stranded DNA that accumulates during the amplification reaction. Alternatively, the method is based on the continuous optical monitoring of an amplification process and utilises fluorescently labeled reagents. Their incorporation in a product can be quantified as the reaction processes and is used to calculate the copy number of that gene or sequence region in the sample. The quantification of the amplification product may require the use of controls to avoid false negativity/positivity of the reaction. Particularly suitable for the quantification of the amplification product are reference genes (e.g. beta-actin) whose methylation status is known, and/or DNA standards (e.g. methylated or unmethylated standards).

Accumulation of an amplification product can be monitored through the incorporation of labeled reagents. Some techniques use labeled primers; others rely upon the use of labeled probes to monitor the amplification product. Real-time quantitative methylation specific PCR techniques comprise the use of Amplifluor primers and/or Molecular Beacon probes and/or Fret probes and/or Scorpion primers and/or Taqman probes and/or oligonucleotide blockers (eg. HeavyMethyl approach) and/or DzyNA primers. All these probes and primers have been described and their mode of action is well known in the art. In a preferred embodiment, the methods of the invention use unmethylated specific primers indicated by SEQ ID NO's 63, 64, 69 and 70 and/or methylated specific primers indicated by SEQ ID NO's 61, 62, 67 and 68.

Alternatively to PCR, other amplification methods such as NASBA, 3SR, TMA, LCR, selective amplification of target polynucleotide sequences (US Patent No. 6,410,276), consensus sequence primed polymerase chain reaction (US Patent No 4,437,975), arbitrarily primed polymerase chain reaction (WO 90/06995), invader technology, strand displacement technlology, and nick displacement amplification (WO 2004/067726) may be used to amplify the appropriate nucleic acid.

Primers may be designed in both the sense and antisense orientation to direct sequencing across the relevant region of the genes of this invention. Said primers can easily be designed by a person skilled in the art.

These amplification primers, amplification probes and sequencing primers form a further aspect of the invention.

This invention provides prognostic and/or diagnostic tools or means to determine a prostate cancer and to distinguish between androgen sensitivity and androgen independency of such prostate cancer. Methylation changes are not only ideal for screening purposes, but also interesting targets for monitoring staging or grading of the cancer. Methods for identifying a prostate cell proliferative disorder in a subject, can comprise the steps of: a) obtaining a biological sample from the subject; b) determining the methylation state of CpG island upstream and/or downstream of the TSS region and/or in the promoter of the genes of this invention; and c) identifying hypermethylation of the region(s) of PITX2, and HOXD3, wherein hypermethylation on CpG and/or non-CpG dinucleotides of said gene(s) is identified as being different when compared to the same region(s) of the gene(s) or associated regulatory region in a subject not having the prostate cellular proliferative disorder: or wherein detection of said hypermethylation is indicative for the stage/type or grade of the prostate cancer. This unexpected finding allows to diagnose for hormone-independent cancers by a simple assay that detects the hypermethylated CpG islands in the promoter region or upstream of the promoter region of the genes of this invention directly by for instance restriction endonuclease analysis to select the proper treatment for subjects with a prostate cancer, depending on the fact of the prostate cancer is hormone refractory or hormone sensitive or depending on the stage or grade of prostate cancer as can be indicated by the hypermethylation status. This is more reliable than detecting levels of mRNA or gene products of said genes. The diagnostic methods will also allow to indicate the proper treatment for hormone -refractory cancers or avoid that subjects with an hormone sensitive cancer will receive an inadequate treatment or assure that they can be treated differently. For instance patients by the diagnosis of present invention to have hypermethylation of a CpG island in the promoter region or upstream of the promoter region of PITX2, and HOXD3, can be subjected to an antimitotic drug therapy methods of treatment or the treatment can now adequately be directed to replacing the hypermethylated CpG islands (or non-CpG islands) with a non-methylated islands which for instance is possible by a treatment with a therapeutically sufficient dosage of a pharmaceutically acceptable DNA methylation inhibitor.

The findings of the present invention allow to diagnose prostatic cells or tissues for prostate cancer and to distinguish between a condition of benign prostate hyperplasia and prostate cancer.

The findings of present invention now specifically allow to diagnose for androgen- independent prostate cancer (AIPC) by a simple assay that detects the hypermethylated promoter or upstream region of the promoter directly of the genes of this invention (PITX2, and HOXD3) and to select the proper treatment for subjects with this AIPC or avoid that subjects with an androgen sensitive cancer will receive an inadequate treatment or allow that a such subject will be treated differently than subjects with androgen-sensitive prostate cancer.

Several documents are cited throughout the text of this specification. Each of the documents herein (including any manufacturer's specifications, instructions etc.) are hereby incorporated by reference; however, there is no admission that any document cited is indeed prior art of the present invention. EXAMPLES

Example 1: Materials and Methods Patients and sample collection

42 benign prostate hyperplasia (BPH) samples, 69 prostate cancer (PCa) samples (PCal cohort), 67 PCa samples (PCa2 cohort) and 16 PCa samples (PCa3 cohort), all paraffin embedded, were obtained from the UZ Leuven (Leuven, Belgium). Characteristics of the cohorts of patients analyzed are represented in Table 1. Furthermore, the PCal A cohort represents the high-risk patients of PCal . The PCa2A cohort represents a further cohort of high-risk patients, of which PCa2 represent the untreated ones.

PCa patients from PCal cohort were selected based on following criteria:

^■ Patients < 75 years of age, with histologically proven invasive adenocarcinoma of the prostate

^■ Patients with, according to the Partin tables, a risk of equal or more than 10% but less than 35% of lymph node metastasis (intermediate to high risk of locoregional or metastatic disease)

^■ No involvement of pelvic lymph nodes assessed by CT scan

^■ No evidence for bone metastasis

^■ WHO performance status > 2 ■ No previous pelvic irradiation or radical prostatectomy

^■ No previous hormonal therapy

^■ No other malignancy except adequately treated basal cell carcinoma of the skin or other malignancy from which the patient has been cured for at least 5 years

All patients provided written informed consent according to ICH/GCP, and national/local regulations.

Additionally BPH samples from 2 patients (analyzed together with 40 paraffin embedded BPH samples) and matched PCa/adjacent normal tissue samples from 7 patients were obtained from the University of Liege (Belgium). All tumor samples were obtained from macroscopically abnormal areas within radical prostatectomy specimens. All samples were directly obtained in the OR and snap frozen in liquid nitrogen vapors. Before snap freezing, a slice was cut for formalin fixation and further paraffin embedding. Sections were cut from these slices and the percentage of tumor was estimated (tumor samples contained at least 80% of tumor cells).

In addition, the prostate cell lines LNCaP, DU 145, PC-3, PZ-HPV-7, BPH1 (American Type Culture Collection, Rockville, MD, USA) and human genomic DNA (Clontech Laboratories, Inc., Mountain View, CA, USA) were used in the experiments. Table 1. A summary of the groups of patients used for the DNA methylation analysis.

Group Number of Diagnosis and tissue characteristics Remark

patients

BPH 42 Bening Prostatic Hyperplasia

PCal 75 (69 final) PCa with high-risk for nodal involvement based on - 3.5 year follow-up

partin table - DNA from tumor +

*clinical cT2-T3 OR surrounding benign

*biopsy Gleason score >= 6 OR tissue

*preoperative serum PSA >=10 μg/L,)

*with negative bone scan and CT-scan of the pelvis

Histologically proven invasive

adenocarcinoma

More than 10 % but less than 35% of LN

metastasis according Partin

No evidence of bone metastasis

No previous hormonal therapy

PCa2 67 The patients of PCa2A cohort that were not

neoadjuvant= treated

PCa3 16 High-risk^$ PCa - DNA from primary tumors and lymph node metastases from the same patients

PCalA 63 The high-risk^$ patients of PCal cohort

PCa2A 84 High-risk^$ PCa - 16 year follow -up

- DNA only from tumor

The criteria of High risk PCa patients are based on the EAU-NCCN guidelines, meaning cT3-cT4 or biopsy Gleason score 8-10 or PSA>20 ng/ml.

Nucleic acid extraction

Genomic DNA was extracted using the GenElute Mammalian Genomic DNA Purification Kit (Sigma-Aldrich, St. Louis, MO, USA) for cell lines and snap-forzen tissues, and the WaxFreeTM DNA kit (TrimGen, Sparks, MD, USA) for paraffin-embedded tissues following the manufacturer's protocol. The concentration of DNA was determined with the spectrophotometer NanoDrop ND-1000 (Thermo Fisher Scientific, Wilmington, DE, USA). Methylation analysis

Genomic DNA from all prostrate samples (500 ng) was bisulfite-converted using the EZ DNA methylation kit (Zymo Research Corp., Orange, CA, USA) according to the manufacturer's protocol. The final elution of bisulfite treated DNA was done in 25 ul elution buffer. Samples were stored at -80°C. The modified DNA was used as a template for quantitative multiplex nested-MSP.

Quantitative multiplex nested-MSP analysis

Quantitative multiplex nested MSP analysis was performed in two subsequent steps. In step 1, multiplex nested PCR was performed to co-amplify 12 genes, using external primer pairs independent of DNA methylation, i.e. containing no CpG sites, or no more than one CpG site close to 5' end, designed according to guidelines (MSP PCR, PCR11). All primers are listed in Table 2. PCR was performed in a volume of 25 ul containing reaction buffer (16.6 mM (NH₄)₂S0₄, 67.0 mM Tris pH 8.8, 6.7 mM MgCl₂x6H₂0, 10.0 mM β-mercapto-ethanol), 2,5 ul of dNTP Mix 2mM each (Fermentas GmbH, St. Leon-Rot, Germany), 2.5 ul of 10* 24 primer mix 2uM each primer (Sigma-Aldrich N.V. Bornem, Belgium), 0.5U IMMOLASE™ DNA polymerase (Bioline USA Inc., Boston, MA, USA), 3 ul of bisulfite-converted DNA template. Reactions were carried out in triplicate using the following conditions: 95°C for 10 min, then 30 cycles at 95°C for 30 s, 57°C for 30 s, 69°C for 30 s; and a final extension step at 69°C for 3 min. A negative control for the assay (water only) was included. The final PCR product from each triplicate was diluted 1 :500 in sterile distilled water. In step 2, separate quantification of methylated and unmethylated DNA fragments of each gene preamplified in step 1 was performed in two independent quantitative reactions (MSP and USP) containing a pair of internal primers, correspondingly, for methylated (M) or unmethylated (U) sequences, for each of 3 repeats separately, on a Rotor-Gene TM 6000 (Corbett Life Science Pty Ltd, Mortlake, NSW, Australia). Reactions were carried out in a volume of 15 ul in the same PCR mix with addition of 0.75 ul EvaGreen® dye (Biotium Inc, Hayward, CA, USA), 0.4 uM of M or U forward and reverse primers (listed in Table 2), 0.3U IMMOLASE™ DNA polymerase (Bioline, London, UK) and 2ul of diluted PCR product from multiplex nested PCR. Cycling conditions were as follows: 95°C for 10 min, then 30-35 cycles at 95°C for 20 s, 61°C for 15 s, 69°C for 15 s. Melting curve analysis of amplification products was performed at the end of each PCR reaction by increasing the temperature from 70°C to 95°C by 0.5°C every 10 s. Separate standard curves were generated for MSP and USP using 5 serial dilutions of plasmids (3x 10⁷-3^X 10³ copies per reaction in triplicate) containing a cloned fragment of each gene of interest in, correspondingly, a fully "methylated" (with Cs in CpG dinucleotides) or "unmethylated" (containing no Cs) versions, and the number of "methylated" (M) and "unmethylated" (U) molecules for each gene in every cancer sample was quantified. The percent methylation for each gene in the panel was calculated as %M = [M / (U + M)] 100.

Plasmid M and U clones were obtained by separate amplification of a promoter region of every gene with methylation independent primers using alternatively methylated PCa cell lines (M standard) and human genomic DNA from whole blood (U standard) under the PCR conditions listed above. Amplified fragments were cloned in DH5a™ competent cells (Invitrogen Ltd, Paisley, UK), using pGEM®-T Easy Vector System (Promega Corporation, Madison, WI, USA).

Table 2: Primers

CCND2-UF3, SEQ. ID NO 21 AAGTATGTGTTAGAGTATGTGTTAGGGTTG 80

ATT

CCND2-U 3, SEQ. ID NO 22 CAAACTTTCTCCCTAAAAACCAACTACA

PTGS2-NF, SEQ. ID NO 23 GGCGATTAGTTTAGAATTGGTTTT 144

PTGS2-NR, SEQ. ID NO 24 AAAATAATCCCCACTCTCCTATCTA

PTGS2-MF, SEQ. ID NO 25 ATTGGTTTTCGGAAGCGTTC 81

PTGS2-MR3, SEQ. ID NO 26 TCCACCGCCCCAAACG

PTGS2-UF, SEQ. ID NO 27 GTTTAGAATTGGTTTTTGGAAGTGTTT 91

PTGS2-UR3, SEQ. ID NO 28 AATTCCACCACCCCAAACA

BCL2-NF2, SEQ. ID NO 29 TTAGTTCGGTGTTATTTGTGGTTT 1 1 1

BCL2-NR2, SEQ. ID NO 30 ACGTCAAATACAACTAACTAAACATCTC

BCL2-MF, SEQ. ID NO 31 TTTTCGTTAGGTCGGCGAC 59

BCL2-MR, SEQ. ID NO 32 ACATCTCGACGAAATCGCG

BCL2-UF, SEQ. ID NO 33 TATTTGATTTTTTGTTAGGTTGGTGAT 67

BCL2-UR, SEQ. ID NO 34 ACTAAACATCTCAACAAAATCACAACA

RASSF1 -NF7S, SEQ. ID NO 35 GTCGTTTAGTTTGGATTTTGG 131

RASSFl -NR6a, SEQ. ID NO 36 CTCAAACTCCCCCGACATAA

RASSF1 -MF, SEQ. ID NO 37 GGTTCGTTTTGTGGTTTCGTTC 72

RASSF1 -MR, SEQ. ID NO 38 CCCGATTAAACCCGTACTTCG

RASSF1 -UF, SEQ. ID NO 39 GGGTTTGTTTTGTGGTTTTGTTT 78

RASSF1 -UR, SEQ. ID NO 40 CATAACCCAATTAAACCCATACTTCA

TDRD 1-NF2, SEQ. ID NO 41 GGAATACGTGGGTATATTGAGTTGT 139

TDRD 1-NR2, SEQ. ID NO 42 GACTACCGATACTAAAAACCCTACC

TDRD l-MF, SEQ. ID NO 43 GGTATATTGAGTTGTACGTGGACGC 57

TDRD l-MR, SEQ. ID NO 44 CCTCCTAACCTCAACGCACG

TDRD l-UF, SEQ. ID NO 45 GTGGGTATATTGAGTTGTATGTGGATGT 63

TDRD I UR, SEQ. ID NO 46 CACCCTCCTAACCTCAACACACA

LGALS3-NF, SEQ. ID NO 47 AATTTTTTATTTAGGTGATTTTGGAGA 151

LGALS3-NR, SEQ. ID NO 48 CAAAAACGACCAAAAAACTCC

LGALS3-MF, SEQ. ID NO 49 AGTTTAGGTTTCGGGAGCGTTAC 61

LGALS3-MR, SEQ. ID NO 50 ACTAAAAAACGCGACCTCCG

LGALS3-UF, SEQ. ID NO 51 GTTGTAGTTTAGGTTTTGGGAGTGTTAT 71

LGALS3-UR, SEQ. ID NO 52 CAAACACTAAAAAACACAACCTCCA

CDH13-NF1 , SEQ. ID NO 53 GAGGTTGAGTTTTAATAGTTTAAAGAAGT 104

CDH13 NR1 , SEQ. ID NO 54 CTCCCTCGTTTTACATAACAAATAC

CDH13 MF2, SEQ. ID NO 55 GATGTTATTTTCGCGGGGTTC 50

CDH13 MR, SEQ. ID NO 56 AAATACGAAATAAACACCTCGCG

CDH13 UF2, SEQ. ID NO 57 GGGATGTTATTTTTGTGGGGTTT 59

CDH13 UR, SEQ. ID NO 58 ACATAACAAATACAAAATAAACACCTCACA HOXD3-NF, SEQ. ID NO 59 TGCGAGTTAAAGGTTATTTTAAAGGT 87

HOXD3-N , SEQ. ID NO 60 CCGAAAAAACCTCACACAAAA

HOXD3-MF, SEQ. ID NO 61 TTAAAGGTTTATGGTTGCGCG 56

HOXD3-MR, SEQ. ID NO 62 ACACAAAACGTTCCCGACG

HOXD3-UF, SEQ. ID NO 63 AAAGGTTATTTTAAAGGTTTATGGTTGTGT 70

HOXD3-UR, SEQ. ID NO 64 CCTCACACAAAACATTCCCAACA

PITX2-NF, SEQ. ID NO 65 TTTTTGGTTTTAAGATGTTAGGTTAATA 89

PITX2-NR, SEQ. ID NO 66 CGCAACTCAACTCCAAACAC

PITX2-MF, SEQ. ID NO 67 GTTAATAGGGAAGCGCGGAGTC 59

PITX2-MR, SEQ. ID NO 68 AAACACCCAAACGAACGACG

PITX2-UF, SEQ. ID NO 69 ATGTTAGGTTAATAGGGAAGTGTGGAGTT 63

PITX2-UR, SEQ. ID NO 70 CCAAACACCCAAACAAACAACA

GSTP1 -1 1NF, SEQ. ID NO 71 GGCGGGATTATTTTTATAAGGTT

124

GSTP1 -1 1NR, SEQ. ID NO 72 CTAAAAACTCTAAACCCCATCC

GSTP-1 1MF, SEQ. ID NO 73 GGAGGTCGCGAGGTTTTC

52

GSTP1 -1 1MR, SEQ. ID NO 74 CTAATAACGAAAACTACGACGACGA

GSTP1 -1 1UF, SEQ. ID NO 75 GTTTGGAGGTTGTGAGGTTTTT

64

GSTP1 -1 1UR, SEQ. ID NO 76 CATACTCACTAATAACAAAAACTACAACAA

CA

The HOXD3 promoter region that is amplified by the external HOXD3 primer set, corresponds to 909 to 823 bp upstream of its transcription start site (TSS). The HOXD3 TSS is located at chromosome2: 177,028,805 based on hgl9 genome version. The amplicon that is obtained by the external primer pair of PITX2 gene, corresponds to position -16 to +73 of the TSS. The PITX2 TSS that is used in this study, is located at chromosome 4: 111,544,254 based on hgl9 genome version and is known as the second TSS, which gives rise to a short alternative splice variant NM_000325.5 (NCBI reference sequence). Statistical analysis

Using the CpG island methylation data, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of every DNA methylation marker were calculated based on a methylation treshold below which a sample was regarded as unmethylated. The latter was determined for each marker based on its degree of methylation in BPH tissues. The methylation threshold was set at \%CCND2, 2% for RARB, GSTP1, APC, PTGS2 and BCL2, 5% for TDRDI, 15% for LGALS3, PITX2 and CDH13, 20% for RASSF1 and HOXD3. To determine the relation between methylation and clinicopathologic characteristics, the second methylation threshold, or a cutoff value (CV), was introduced for all markers based on the median methylation value (MV), which was calculated for samples methylated > 1%. If for some gene the median value was lower than the first methylation threshold determined based on methylation of BPH, the latter was applied as a cutoff value. For the PCal cohort the following statistical analysis was performed: using the methylation value (MV) at each gene as a dichotomous independent variable with MV > CV for each gene designated as highly methylated (HM) and MV < CV designated as lowly methylated (LM), the relation between the high/low DNA methylation and a categorical clinicopathologic variable was assessed by means of a Fisher's Exact test. Categorical clinicopathologic variables included pT stage (I, II vs III, IV), Gleason score (4-7a vs 7b- 10) and lymph nodes (negative vs positive). For continuous clinicopathologic variables (PSA level, tumor volume, age of patients), the mean variable was described descriptively between the samples with and without methylation and compared between the two groups by means of a Wilcoxon-rank sum test. Graphs were generated using GraphPad Prism 3.0 software (GraphPad Software, Inc., La Jolla, CA, USA).

For the PCa2 cohort similar statistical methodology with some variations was applied. Namely, the univariable association between two continuous variables or a continuous and ordered categorical variable was analyzed by the Spearman correlation coefficient. The association between a continuous and a categorical variable was analyzed by means of the Mann- Whitney U test (for two categories) or the Kruskal-Wallis test (for more than two categories). The association between two categorical outcomes was analyzed by means of the chi-square test or the Fisher exact test (in case of low (<5) cell frequencies).

For studying the relationship between DNA methylation percentage and time-to-event outcomes (BCR and clinical failure) a Cox proportional hazard model was used. The presence of a nonlinear relationship between the methylation percentage and event risk was investigated by considering models with a quadratic effect and using restricted cubic splines. The fit of these models was compared to the fit of a model with linear trend using a likelihood ratio test. Based on the best fitting model proposals for categorizations of the DNA methylation percentage were derived. In case of a linear trend the cutoff value that leads to the best dichotomization (highest likelihood) was selected by comparing all possible dichotomisations. Further the predictive accuracy of competing models (e.g. with continuous as categorical versions of the DNA methylation percentage) by means of the concordance probability index (CPE; Gonen and Heller 2005) was calculated. A graphical presentation of the results is given by plotting the Kaplan Meier estimates in case of categorical predictors or the model-predicted incidence curves for a selection of predictor values in case of continuous predictors.

The association between the methylation in primary tumor and metastasis for every single gene in the PCa3 cohort was analyzed by means of the Pearson correlation coefficient. As a global measure of correlation for all genes, the canonical correlation coefficient was calculated.

All analyses have been performed using SAS software, version 9.2 of the SAS System for Windows. Copyright © 2002 SAS Institute Inc. SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc., Cary, NC, USA.

Example 2: Results

1. Screening of prostate cancer-related hyper- and hypomethylated genes

The average number of genes methylated with functional significance in PCa is estimated as several hundred (1). Kim et al (2) identified 812 out of 1171 unique gene promoters methylated only in prostate cancer cell line LNCaP in comparison with the benign cell line PrEC. We started from the hypothesis, however, that all these methylation changes are the consequence of a general failure of the DNA methylation machinery in the onset of tumorigenesis and during cancer progression, and targeting of a small number of differentially methylated sites is quite enough to develop a methylation assay with substantial diagnostic and prognostic significance. Therefore our aim was to select from a pool of the described in the literature methylation markers of PCa a limited number of genes showing altered methylation status between normal and tumor tissue, and in a perspective - between different stages of PCa. For this purpose we screened a selection of genes reported to be frequently differentially methylated in PCa and benign prostate tissues and/or associated with PCa prognosis and outcome (2, 3 and other) using a set of 6 model genotypes. These genotypes comprised three PCa cell lines (androgene-sensitive LNCaP, androgene-refractory PC-3 and DU 145, corresponding to early and late stage cancer) and two benign prostate cell lines PZ- HPV 7 and BPHl, as well as a genomic DNA sample extracted from the whole blood, which corresponds to non-cancerous DNA with regular levels of methylation (Fig. 1A). For the Melting curve analysis methylation independent primers were designed to amplify promoter fragments of the candidate genes and analyze their methylation status by the melting curve assay to detect the presence of unmethylated and partially or fully methylated copies (Fig. IB). The analyzed genes are classified into two groups: hyper- and hypomethylated markers (Table 3).

Table 3. Analysis of the methylation status of prospective PCa biomarkers in a set of model genotypes by melting curve assay.

Conclusion

We have selected 38 genes linked to PCa from the literature (Table 3) and analyzed their DNA methylation state in three PCa and two benign prostatic cell lines, and human whole blood by melting curve analysis. Based on the semi-quantitative data of the melting curve analysis we selected candidates for further analysis of their role in PCa development. The selected group contained four types of biomarkers: 1) genes frequently hypermethylated in PCa (diagnostic markers with possible prognostic significance); 2) metastatic markers: 3) markers associated with biochemical recurrence after prostatectomy; and 4) hypomethylation markers. From this group we selected 16 genes and developed a two-step quantitative multiplex nested-MSP as detailed below.

2. Development and validation of the two-step quantitative multiplex nested-MSP

Methylation-independent primers either from the screening step or designed separately were used to amplify a part of a CpG island in the promoter regions of the selected genes. PCR amplification was performed separately for each gene using genotypes that had shown differential methylation of a separate gene by melting curve analysis. Subsequently, the PCR fragments were subcloned in pGEM-T-easy plasmid vector and multiple clones were sequenced. The sequence information was used to validate the correct amplification of the gene and to determine the methylation status of all CG dinucleotides. The plasmids, containing the PCR fragment corresponding to the fully methylated and unmethylated DNA, were selected and labeled as plasmid M and U, respectively. After obtaining M and U standards for every gene, a final set of methylation independent primers was designed for amplification of PCR fragments around 100 base pairs (bp) in length whenever possible, but not exceeding 200 bp (listed in Table 1). Also, two sets of nested primers specific for either methylated or unmethylated bisulphite modified DNA sequence of each gene were designed (Table 1). To examine the specificity of the nested methylation and unmethylation specific primers, both plasmids M and U were used as a template for amplification with both primer sets (Fig. 3). Importantly, no methylated signal (0 % methylation) was detected with plasmid U, as well as in "no template" control, using methylation specific primers (qMPS reaction). Likewise, no unmethylated signal was detected with plasmid M, as well as in "no template" control, using unmethylated specific primers (qUSP reaction), indicating that the qMSP and qUSP reactions were 100% specific. To check the efficiency of both sets of primers, separate standard curves were generated for MSP and USP using 5 serial dilutions of plasmid standards M and U (3>< 10⁷-3x l0³ copies per reaction in triplicate) for each gene. Final sets of qMSP and qUSP amplifying the template with high efficiency and linearity were selected (Fig. 4). The whole scheme of the validation process is presented on figure 5. The scheme of the two-step quantitative multiplex nested- MSP is presented on figure 6. Briefly, a mixture of validated gene-specific multiplex primers was used to co-amplify 12 gene promoters independent of their methylation status. The amplification product was diluted 1 :500 with sterile distilled water and used as a template for the quantitative PCR. In the second PCR reaction, absolute quantification of methylated and unmethylated DNA fragments for each gene was performed separately by using validated M and U specific primers and separate standard curves generated using serially diluted M and U plasmid standards.

Conclusion. The advantage of the developed two-step quantitative multiplex nested-MSP assay is that it utilizes the same bisulphite-converted DNA template (which is often very limited in a volume and quantity) to preamplify the selected number of gene promoters of interest in one PCR tube. This makes the procedure independent of sampling and pipeting diversions and allows at the same time to obtain the sufficient amount of the DNA targets for MSP primers to reduce false priming errors. The second quantitative step enables (the researcher) to determine the lowest methylation levels and discriminate between functionally significant and insignificant or background methylation.

3. Determination of DNA methylation state of 16 genes in prostate cell lines and paired PCa/adjacent normal tissue samples

The developed two-step quantitative multiplex nested-MSP assay was used to determine the degree of methylation of the selected genes in prostate cell lines as well as in PCa samples and BPH genotypes. As it has been observed in a previous study (4), cancer lines usually exhibit higher levels of CpG island hypermethylation than primary cancers, which may be a result of repeated passages and adaptation to culture environment, as well as of contamination of tumor samples by adjacent normal cells. In our study PCa cell cultures also show more polar methylation values as compared to the PCa samples (Fig. 7 and 8).

LNCaP cell line corresponding to hormone-dependent (early stage) PCa surprisingly showed the highest methylation value for 11 out of 14 hypermethylated genes analyzed. However, two markers of PCa progression and biochemical recurrence CDH13 and HOXD3 were not significantly methylated in this line. PC-3 showed the highest value for 3 genes reported to be associated with biochemical recurrence: APC, PITX2 and HOXD3, DU 145 - for HOXD3 and CDH13. A combination of these markers has a greater prognostic value in comparison with that of a single marker.

Among non-malignant cells both BPH 1 and PZ-HPV7 showed methylation higher than 10 % for 5 different genes each (Fig. 7). In benign genotypes methylation higher than 20% was detected for RASSFI, PITX2, HOXD3, TDRD5, TBX20 and SOXl implying that a higher methylation cutoff value must be introduced for these markers to reveal their diagnostic and/or prognostic significance.

Methylation of all genes was detected to a much greater extent in tumor samples in comparison with histologically cancer-free adjacent tissues (Fig. 8).

The PCa samples 1 and 6 have a higher Gleason score 3+4 compared to other samples (GS 3+3). In PCa sample 1 and 6, 9 and 6 genes respectively, show methylation close to or above the 50%) level. PCa samples 5 and 7 show lower levels of methylation, which corresponds to the notion that the degree of methylation increases with tumor progression. HOXD3 shows methylation above median in a sample PCa 4 and PCa 6, while PITX2 has a higher methylation level in sample PCa 1.

Conclusions

1. The developed two-step quantitative multiplex nested-MSP assay effectively distinguished PCa cell lines from non-malignant cells, as well as PCa tumors from surrounding malignant tissues based on the quantification of the methylation values of 16 markers.

2. Cancer cell lines exhibited higher methylation values compared with the PCa samples. LNCaP cells showed the highest number of completely methylated genes (90-100%>): 11 out of 14 hypermethylated markers. 4. Determination of DNA methylation state of 12 genes in benign prostate hyperplasia and PCa samples

To determine sensitivity and specificity of the methylation markers, we applied the developed two-step quantitative multiplex nested-MSP assay for quantification the levels of methylation of 12 genes in four groups of patients: BPH (benign prostate hyperplasia, N=42), PCal (N=69), PCa2 (N=67), PCa3 (N=16). Based on the methylation levels in the BPH cohort (Table 4), we divide the analyzed genes into two groups: diagnostic and prognostic markers. Diagnostic markers, also indicated as PCa-specific markers, have no or very low levels of methylation in BPH samples not exceeding 1-2% of methylated gene copies. Thus, at setting the methylation cutoff value at 2% such markers attain 100% specificity, showing at the same time quite high sensitivity (Table 5). The group of PCa-specific markers includes RARB, GSTPl, CCND2, PTGS2, BCL2 (all 5 show 100% specificity at the 2% methylation cutoff value) and APC (98%> specificity). The second group of genes was moderately methylated in BPH samples, so the methylation cutoff value should be raised to 5-20% of methylated gene copies to increase specificity of these markers. This decreased their sensitivity, while specificity still never reached 100% (Table 5). However, we will utilize these markers for prognostic rather than diagnostic purposes. The group of prognostic markers includes LGALS3, TDRD1, RASSF1, PITX2, HOXD3, CDH13.

Table 4. Mean methylation values (%) of 12 genes in one BPH and two PCa cohorts. The Table represents the mean methylation values, standard deviations, the median, minimal and maximal methylation values, 25% and 75% percentiles (Ql and Q3). For the TDRD1 gene reverse methylation values (100 - % of methylation) are represented.

Table 5. Diagnostic information of DNA methylation markers as determined by the invention methods employing two-step Quantitative multiplex nested-MSP. Sensitivity, number of true positives which were identified as having cancer; specificity, number of true negatives which were identified as having no cancer; PPV, positive predictive value; NPV, negative predictive value. Methylation cutoff value signifies a threshold below which methylation of each gene is considered to be unspecific for PCa. Calculations of diagnostic parameters are performed using the methylation data from a BPH cohort (N=42) and the PCa cohorts: PCal (N=69) and PCa2 (N=67).

DNA of patients from the PCal cohort was extracted from the whole paraffin slide containing both tumors and non-tumor cells, while in the PCa2 cohort it was extracted exclusively from tumor sites. This explains why methylation values were lower in the PCal cohort compared with that in the PCa2 cohort for most of the genes (Table 4), as the tumor DNA in the PCal cohort was mixed with the DNA from surrounding non-malignant cells. This also explains why sensitivity of most of the markers was lower in the PCal cohort (Table 5). Methylation of markers could be detected, but due to presence of the unmethylated DNA from non- malignant cells the detected methylation levels could be lower than the methylation cutoff set for each separate marker (1-2% methylation for PCa-specific markers, 5-20% for prognostic markers.

In general, and considering the all said above, the two-step quantitative multiplex nested-MSP assay detected PCa cancer with very high efficiency. In the PCal cohort only 5 of 69 samples showed no significant methylation of PCa-specific genes. We attribute the cases with the insignificant methylation values detected in the PCal cohort to methylation-independent PCa types, having, however, quite low incidence. Still there is a number of cases in every cohort with methylation above the cutoff value, but not exceeding 10%> for most of the genes (5.79% and 10.77%) of cancer cases in the PCal and PCa2 cohorts correspondingly). We would prefer to call these cases as tumors with suppressed methylation; this phenomenon may have an impact on the prognosis of the tumor development based on methylation marker information and needs further more detailed investigation.

Conclusions

1. The two-step quantitative multiplex nested-MSP assay showed very high sensitivity for PCa at the 100% with specificity level: 92.75% in the PCal cohort (mixed DNA from malignant and non-malignant cells) and 100.00% in the PCa2 cohort (DNA predominantly from the tumor sites).

2. 7.25% of tumors in the PCal cohort may be attributed to methylation-independent cancer cases, as none of the markers (or no more than one marker) showed methylation above the cutoff value in such tumors. Methylation-independent tumors may be identified by the two-step quantitative multiplex nested-MSP assay and discriminated from methylation- associated tumors in case if PCa was detected by other diagnostic means.

3. Tumors with low (below 10%) methylation of all markers were detected in all two PCa cohorts analyzed; the incidence of such cases was 5.79% in PCal and 10.77% in PCa2. The two-step quantitative multiplex nested-MSP assay is able to detect such methylation- suppressed PCa tumors; however, no prognostic conclusion may probably be made for such cancer cases based on DNA methylation, as the methylation seems to be suppressed by some unknown mechanism. 5. Correlation of CpG island methylation alterations with clinicopathologic parameters.

In this example, we evaluated correlations between CpG island alterations of eleven genes (single loci, RARB, GSTP1, CCND2, PTGS2, APC, LGALS3, TDRD1, RASSF1, PITX2, HOXD3, CDH13,) and clinicopathologic parameters: preoperative serum PSA level, clinical tumor pT stage, extracapsular extension, seminal vesicle invasion, surgical margin status, Gleason score, age and tumor volume in the PCa2 cohort (Fig. 9, Table 8). Table 8. P-values indicating the level of statistical significance of association between methylation of 11 genes and the following clinicopathologic parameters: Gleason score, tumor pathological pT stage, extracapsular extension, seminal vesicle invasion, surgical margin status. P-values higher than 0.05 are considered non-significant (NS). Study group: PCa2

No correlation between DNA methylation and preoperative PSA levels was found, when those variables were regarded as continuous variables (analyzed by the Spearman correlation coefficient). Table 9. Significant associations between promoter methylation of 11 genes and clinicopathologic traits detected at least in one of the PCal and PCa2 cohorts. Clinicopathologic parameters: 1 - age, 2 - preoperative PSA, 3 - Gleason score, 4 - tumor pathological pT stage, 5 - extracapsular extension, 6 - seminal vesicle invasion, 7 - surgical margin status, 8 - tumor volume, 9 - lymph node invasion, 10 - biochemical recurrence, 11 - clinical failure.

DNA methylation was not associated with the age of patients (Spearman correlation, Table 9). Methylation of PITX2 and HOXD3 was highly associated with pathological tumor pT stage (pT as unordered categorical variable: pTl-2, 3a, 3b, 4; DNA methylation as a continuous variable, Kruskal-Wallis test, Table 8) and ecstracapsular extention (Mann-Whitney U test, Table 8). Besides, HOXD3 methylation was associated with seminal vesicle invasion (Mann- Whitney U test, Fisher exact test, Table 8) and lymph node invasion (Mann- Whitney U test, Table 9), although the number of patients with lymph node metastases in the PCa2 cohort was rather low (7 of 67). An association with tumor volume (measured as a percentage of a total gland volume) in the PCa2 cohort was detected for HOXD3 (Mann- Whitney U test, Fig. 9). All statistically significant association between clinicopathologic parameters and DNA methylation at least in one of the two PCa cohorts analyzed are indicated in Table 9. Conclusion

Some significant associations between CpG island methylation levels and clinicopathologic parameters are detected. Methylation of PITX2 and HOXD3 was highly associated with pathological tumor pT stage and extracapsular extention.

Example 3: Results on prognostic significance of DNA methylation markers.

To assess the prognostic value of DNA methylation markers, we analysed methylation frequencies in a PCa2 cohort of 69 patients with < 16 years of follow-up data. The primary clinical end point of this study was biochemical recurrence (BCR, also biochemical relapse or biochemical progression- free survival). Clinical failure was the second end point of this study. BCR was detected in 37 (55.22%) patients and for 11 (16.42%) patients clinical failure was observed. PCa samples were divided into high methylation (HM) and low methylation (LM) groups based on methylation cutoff value selected for each gene individually (Table 10). Table 10. Selection of the methylation cutoff value for the best dichotomisation (low methylation vs high methylation). Study group: PCa2. CF - clinical failure; BCR - biochemical recurrence.

Gene . . CHitkal factor ftietliflatiei . tta»§fatl# f ralue,Ccnc

. ^{■ '} . . cutoff wakie, %

ifWQ _. ^' _. CF 50.17 9.4 0.0015

_.§»§13Q ^' _. ^{■ ' '} _. BCR 23.55 3.9 0.0003

TNDML^' BCR 49.53 2.8 0.0058

HOXD3 ^' ,.^: BCR 25.05 5.2 0.0022

RASSFI : ^· · BCR 53.99 3.4 0.0014

Univariate Kaplan-Meier curves demonstrated that low-level methylation of TDRD1, and high high-level methylation RASSFI, PITX2 and HOXD3 predicted significantly earlier biochemical recurrence with the best methylation cutoff value at 69. 23% percentile, 71.88% percentile, 72.00%) percentile and 35.82%> percentile correspondingly (Fig. 10-11). High PITX2 methylation was also associated with the significant risk of clinical failure (CF, methylation cutoff value at 93.94% percentile, Fig. 12). Conclusion

High methylation levels of HOXD3 and PITX2 are significantly associated with the risk of BCR. A high methylation of PITX2 is also significantly associated with the risk of CF. Table 13: Nucleotide sequences of the inserts cloned into pGEM-T easy plasmid, to generate the control plasmids of the invention. Plasmid M and U clones were obtained by separate amplification of a promoter region of every gene with methylation independent primers using bisulphite converted genomic DNA from hypermethylated PCa cell lines (M standard) and human genomic DNA from whole blood (U standard). Amplified fragments were cloned in DH5a™ competent cells (Invitrogen Ltd, Paisley, UK), using pGEM®-T Easy Vector System (Promega Corporation, Madison, WI, USA) according to a manufacturer's protocol.

APC - M, SEQ. ID NO 81 ggaagcggagagagaagtagttgtgtaattcgttggatgcggattagggcgttttttattttcgtcgggagttcg tcgattggttgggtgtgggcgtacgtgatcgatatgtggttgtattggtgtagttcg

APC - U, SEQ. ID NO 82 ggaagtggagagagaagtagttgtgtaatttgttggatgtggattagggtgttttttatttttgttgggagtttgttg attggttgggtgtgggtgtatgtgattgatatgtggttgtattggtgtagttcg

CCND2 -M, SEQ. ID NO 83 ttttgtaaagatagttttgatttaagtatgtgttagagtacgtgttagggtcgatcgtgttggcggcgattttatcgt agtcggtttttagggagaaagtttggcgagtgaggcgcgaaatcggaggggttggtgaggatgtgggtgaa ggattgagtgtggaggttttatgttttcggggaaaggaaggggtggtggtgt

CCND2 -U, SEQ. ID NO 84 aaagatagttttgatttaagtatgtgttagagtatgtgttagggttgattgtgttggtggtgattttattgtagttggtt tttagggagaaagtttggtgagtgaggtgtgaaattggaggggttggtgaggatgtgggtgaaggattgagt gtggaggttttatgtttttggggaaaggaaggggtggtggtgt

PTGS2 - M, SEQ. ID NO 85 agataaattatagttatgtatattgaaggtagttattttattttataaaataagagttttttaaaaagttatgtatgtatg tgttgtatatagagtagatatatagtttattaagcgtcgttattaaaatataaaatatgttagttttttttaattttattcgt tttagtttgtttcgacgtgattttttcgattttttaaagacgtatagattagatacggcggcggcggcgggagagg ggattttttgcgttttcggattttagggtcgtttagatttttggagaggaagttaagtgtttttttgttttttttcggtattt tatttaaggcgattagtttagaattggttttcggaagcgttcgggtaaagattgcgaagaagaaaagatatttgg cggaaatttgtgcgtttggggcggtggaattcggggaggagagggagggattagataggagagtggggat tatttt

PTGS2 - U, SEQ. ID NO 86 agataaattatagttatgtatattgaaggtagttattttattttataaaataagagttttttaaaaagttatgtatgtatg tgttgtatatagagtagatatatagtttattaagtgttgttattaaaatataaaatatgttagttttttttaattttatttgttt tagtttgttttgatgtgatttttttgattttttaaagatgtatagattagatatggtggtggtggtgggagaggggatt ttttgtgtttttggattttagggttgtttagatttttggagaggaagttaagtgtttttttgtttttttttg ggtgattagtttagaattggtttttggaagtgtttgggtaaagattgtgaagaagaaaagatatttggtggaaattt gtgtgtttggggtggtggaatttggggaggagagggagggattagataggagagtggggattatttt

BCL2 - M, SEQ. ID NO 87 ttagttcggtgttatttgtggtttatttgatttttcgttaggtcggcgacgatttttttcgtcgttatcgtcgcgattttgt cgagatgtttagttagttgtatttgacgt

BCL2 - U, SEQ. ID NO 88 ttagttcggtgttatttgtggtttatttgattttttgttaggttggtgatgattttttttgttgttattgttgtgattttgttga gatgtttagttagttgtatttgacgt

RASSF1 - M, SEQ. ID NO 89 agtttttgtatttaggtttttattgcgcggttttttttagttttttttcgtcgtttagtttggattttgggggaggcgttgaa gtcggggttcgttttgtggtttcgttcggttcgcgtttgttagcgtttaaagttagcgaagtacgggtttaatcggg ttatgtcgggggagtttgagtttattgagtt

RASSF1 - U, SEQ. ID NO 90 agtttttgtatttaggtttttattgtgtggttttttttagtttttttttgttgtttagtttggattttgggggaggtgttgaagt tggggtttgttttgtggttttgtttggtttgtgtttgttagtgtttaaagttagtgaagtatgggtttaattgggttatgtc gggggagtttgagtttattgagtt

TDRD 1 - M, SEQ. ID NO 91 tgagtttgtaattagagtataagttgttttcggggaaggcggagggaatacgtgggtatattgagttgtacgtgg acgcggagtgcgtaggcgtgcgttgaggttaggagggcgtattggggattggaggcgagggaagtgtagg gcgtattttaggcggtagggtttttagtatcggtagtc

TDRD 1 - U, SEQ. ID NO 92 tgagtttgtaattagagtataagttgtttttggggaaggtggagggaatatgtgggtatattgagttgtatgtggat gtggagtgtgtaggtgtgtgttgaggttaggagggtgtattggggattggaggtgagggaagtgtagggtgt attttaggtggtagggtttttagtatcggtagtc

LGALS3 - M, SEQ. ID NO 93 aattttttatttaggtgattttggagagggcgggggatagacgcggtcgtagtttaggtttcgggagcgttacgg aatttaacggtggtagcggaggtcgcgttttttagtgttcgcgtttttttcgtcgggagttttttggtcgtttttgcgg cggcggttcggggtgtttggtcggcgcggggttcgtttagtttggttcggggagaggattggttgggtaggg gcgtcgtttcgtttcgggagaggcgggtcgggcggggttgggagtatttgaggttcg

LGALS3 - U, SEQ. ID NO 94 aattttttatttaggtgattttggagagggtgggggatagatgtggttgtagtttaggttttgggagtgttatggaat ttaatggtggtagtggaggttgtgttttttagtgtttgtgttttttttgttgggagttttttggttgtttttgtggtggtggt ttggggtgtttggttggtgtggggmgtttagtttggtttggggagaggattggttgggtaggggtgttgttttgtt ttgggagaggtgggttgggtggggttgggagtatttgaggttcg

CDH13 - M, SEQ. ID NO 95 gaggttgagttttaatagtttaaagaagtaaatgggatgttattttcgcggggttcgtttttcgcgaggtgtttattt cgtatttgttatgtaaaacgagggag

CDH13 - U, SEQ. ID NO 96 gaggttgagttttaatagtttaaagaagtaaatgggatgttatttttgtggggtttgttttttgtgaggtgtttattttgt atttgttatgtaaaacgagggag HOXD3 - M, SEQ. ID NO 97 ggtgtttattaaggggtgagttattgcggtgcgagttaaaggttattttaaaggtttatggttgcgcgttttagttttt agaagcgtcgggaacgttttgtgtgaggttttttcgggtgtagtttagtgttcgtaaattattgagtcgataaattg tatagggtagatgtaagagggggattttgtttttttttttttatttgacggggggtggttattcgttgagcggtgata gtggtgggggagtttggmgtggttttgggtgggggtggaaaggatgtttgttttttaggcgatttgtagtttgat tttgatttagaagtgggtagtttagggttaagagtgtggtgga

HOXD3 - U, SEQ. ID NO 98 ggtgtttattaaggggtgagttattgtggtgtgagttaaaggttattttaaaggtttatggttgtgtgttttagttttta gaagtgttgggaatgttttgtgtgaggtttttttgggtgtagtttagtgtttgtaaattattgagttgataaattgtata gggtagatgtaagagggggattttgtttttttttttttatttgatggggggtggttatttgttgagtggtgatagtggt gggggagtttggtttgtggttttgggtgggggtggaaaggatgtttgttttttaggtgatttgtagtttgattttgatt tagaagtgggtagtttagggttaagagtgtggtgga

PITX2 - M, SEQ. ID NO 99 cgtattgttgataggtgtaggtaggatagtttttttatcgcggttcggggcgttttgattggtgcggagttacgtta gtcgtattcggagaagggtttgggaggaggcggaggcggagagggttggggagggttgcggcggagtg acgtttcggtattaggaagttcgtttttggttttaagatgttaggttaatagggaagcgcggagtcgtagatttggt tcgtcgttcgtttgggtgtttggagttgagttgcggtaaggttcggtttttgttcgatcgttcgaggggtgtgcgtg tgcgcgttgcggagggtgcgtttagagggtcgcgtcgtggttgtagcggttgttgtcgtcgtaggggatttaat attatttatttgtttttg

PITX2 - U, SEQ. ID NO 100 cgtattgttgataggtgtaggtaggatagtttttttattgtggtttggggtgttttgattggtgtggagttatgttagtt gtatttggagaagggtttgggaggaggtggaggtggagagggttggggagggttgtggtggagtgatgtttt ggtattaggaagtttgtttttggttttaagatgttaggttaatagggaagtgtggagttgtagatttggtttgttgttt gtttgggtgtttggagttgagttgtggtaaggtttggtttttgtttgattgtttgaggggtgtgtgtgtgtgtgttgtg gagggtgtgtttagagggttgtgttgtggttgtagtggttgttgttgttgtaggggatttaatattatttatttgtttttg

Example 4: Evaluation of the prognostic value of PITX2 and HOXD3. MATERIALS AND METHODS

DNA methylation was examined within univariate and multivariate Cox proportional hazards regression models for biochemical recurrence (BCR)-free survival and Clinical failure (CF)- free survival. -values refer to the Wald test. DNA methylation is analyzed as a continuous variable. Using a multivariate Cox proportional hazard regression model comprising clinicopathological and methylation variables as indicated in the tables, the relative contribution of each variable to BCR or CF was assessed.

Using regularized linear regression modeling for creating multivariate models in combination with 8-fold cross validation, a two-gene model was built in cohort PCa2 or PCa2A for BCR-free survival and tested using Cox proportional hazard regression model [Tibshirani et al., 2010]. <0.05 was considered statistically significant. All analyses were done using SPSS (Chicago, IL), Matlab (version R2011b) and the survival packages in the R statistical software (survival R package v 2.36-10).

RESULTS

PITX2, and HOXD3, are univariate but not multivariate predictors for biochemical recurrence (BCR)

To study the relationship between DNA methylation and BCR, the correlation of DNA methylation level of PITX2, and HOXD3 as a continuous variable with BCR was investigated in multiple cohorts (Table A). The cohorts, consisting of prostatectomy samples from mainly high-risk patients are described in Material and Methods. Univariate analysis showed significant association of methylation of PITX2 in all cohorts analyzed with the hazard ratio (HR) between 1.035 and 1.058 ( -value <0.01) and of HOXD3 in three out of four cohorts with a HR between 1.022 and 1.046 ( -value <0.05).

To test whether PITX2 and HOXD3 DNA methylation added independent information to known prognostic clinicopathological parameters, multivariate cox proportional hazard model analysis was performed in the context of pre-operative PSA, pathological stage, final Gleason score, surgical margin status, lymph-node invasion, adjuvant radiotherapy (RT) and adjuvant hormonal therapy (ADT). In multivariate analysis, PITX2 and HOXD3 were not independent predictors of BCR in the context of these clinicopathological variables (Table B and C).

A two-gene model, consisting of PITX2 and HOXD3, is an independent significant predictor, both univariate and multivariate for BCR in the context of known prognostic clinical variables.

Based on the DNA methylation of TDRD1, RARB, GSTP1, APC, CCND2 PTGS2, RASSF1, LGALS3, CDH13, PITX2, and HOXD3 as a continuous variable, a regularized linear regression modeling was performed to create a multivariate model for BCR-free survival. Since only multivariate significant P- values were found for cohorts PCa2A and PCa2, a multivariate model in combination with 8-fold cross validation was built for these cohorts. The multivariate models on the cohort PCa2A and PCa2 are the two-gene models PITX2*0.022862+HOXD3*0.0014012 (two-gene model A) and PITX2*0.020677+ HOXD3*0.0043132 (two-gene model B), respectively (Table D). Cox proportional hazard regression analysis showed that the cross-validated two-gene model A had a P- value of 0.00046288 (Wald test) and HR of 4.3517 (95% CI 1.9-9.9) in the cohort PCa2A (Table D). In the validation set PCal, this model had a P-value of 0.0041718 and hazard ratio of 11.1273 (95% CI 2.1-57.84). The cross-validated two-gene model B had a P-value of 0.00039066 and a HR of 4.8467 (95% 2.0259-11.5953) in the cohort PCa2, and a P-value of 0.0044938 and a HR of 11.83 (95% CI 2.15-65.04) in the independent validation cohort PCal A. Taken together, this indicates that the two-gene models A and B are better predictors of BCR-free survival, as compared to the single genes PITX2 and HOXD3 (Compare Table A and E) in univariate analysis.

Finally, the two-gene model A and B is an independent predictor of BCR, as shown by multivariate analysis (Table E-F) in the PCa2A group and PCa2 groups, respectively. Thus, the two-gene PITX2+HOXD3 methylation model A added independent prognostic information to known prognostic clinicopathologic parameters like pre -operative PSA, final Gleason score, surgical margin status (HR 3.3, 95% CI 1.4-7.8; P-value 0.0065225, Table F). The two-gene PITX2+HOXD3 methylation model B added independent prognostic information to known prognostic clinicopathologic parameters like pre-patho logical T-stage and surgical margin status (HR 3.1, 95% CI 1.1-9.0, P-value 0.04047, Table F). In conclusion, a two-gene methylation model, consisting of both PITX2 and HOXD3, but not PITX2 or HOXD3 separately, is an independent significant predictor for BCR in the context of known prognostic clinical variables (compare Table B, C with F). Taken together, the data indicate that the two-gene models A and B are better predictors of BCR-free survival, as compared to the single genes PITX2 and HOXD3 in univariate analysis as well as multivariate analysis (Table A-F). PITX2 methylation is an independent significant predictor for clinical failure in the context of known prognostic clinical variables

Univariate analysis showed significant association of PITX2 with clinical failure (CF) in cohort PCa2 with a HR of 1.0328 (95% CI, 1.0023-1.0642) and -value of 0.034812 (Table G). On multivariate analysis, time to clinical failure was studied with the complete set of clinical variables and PITX2 methylation. PITX2 was an independent significant predictor for clinical failure (HR 1.04, 95% CI: 1-1.1) , -value <0.05) in the context of pre-operative PSA, pathological stage, final Gleason score, surgical margin status, lymph-node invasion, adjuvant radiotherapy (RT) and adjuvant hormonal therapy (ADT) (Table H). Pathological stage and Gleason score were also significant predictors for CF while pre-operative PSA was borderline significant and surgical margin status, lymph node invasion, adjuvant RT and ADT showed no impact (Table H). In conclusion, these data suggest that PITX2 methylation status provides prognostic information independent of the traditionally used clinicopathological parameters, including pathological T stage and Gleason score. Table A. Univariate Cox Proportional Hazard Model analyses on BCR-free survival for DNA methylation of PITX2 and HOXD3. HR, Hazard ratio; BCR, Biochemical recurrence; CI, confidence intervals

Gene name Cohort Univariate

HR (95%C1) -value

Wald test

PITX2 PCa2A 1.0354 (1.0152-1.056) 0.00052871

PCa2 1.0371 (1.0156-1.059) 0.00065982

PCalA 1.0579 (1.0171- 1.1003) 0.0050409

PCal 1.0568 (1.0171-1.1004) 0.0072969

HOXD3 PCa2A 1.0224 (1.0046-1.0406) 0.013483

PCa2 1.0253 (1.0065-1.0443) 0.0079786

PCalA 1.0463 (1.0023-1.0923) 0.039169 PCal 1.0445 (0.99894-1.0922) 0.055678 Table B. Multivariate Cox Proportional Hazard Model analyses on BCR-free survival for HOXD3 methylation (%), pre-operative PSA, pathological stage, final Gleason score, surgical margin status, lymph node invasion, adjuvant radiotherapy (RT) and adjuvant hormonal therapy (ADT) in cohort PCa2. HR, Hazard ratio; BCR, Biochemical recurrence; CI, confidence intervals

95% CI

HR P-value lower upper

HOXD3 1.01897566 0.09793689 0.99654132 1.04191504 pre.op.PSA 1.01556351 0.07081252 0.99869056 1.03272153 pathological.T.stage.grouped 3.26790425 0.00097063 1.61706962 6.60404354 final. Gleason. Score 1.26678087 0.46692734 0.66989661 2.39549467 surgical.margin.status 2.89802382 0.00958351 1.29562062 6.48225406 lymph.node nvasion 1.08148243 0.91026705 0.27694682 4.22320875

Adjuvant.RT 0.19607238 0.04318491 0.04041232 0.95130346

Adiuvant.ADT 0.33534439 0.33355438 0.03661946 3.07093152

Table C. Multivariate Cox Proportional Hazard Model analyses on BCR-free survival PITX2 methylation (%), pre-operative PSA, pathological stage, final Gleason score, surgical margin status, lymph node invasion, adjuvant radiotherapy (RT) and adjuvant hormonal therapy (ADT) in cohort PCa2. HR, Hazard ratio; BCR, Biochemical recurrence, CI, confidence

Variable HR P-value lower upper

PITX2 1.02481793 0.06483661 0.99849309 1.05183681 pre.op.PSA 1.01398479 0.11800699 0.99648118 1.03179586

pathological.T.stage.grouped 2.98215687 0.00300253 1.44914544 6.13689928 final. Gleason. Score 1.34621257 0.36281696 0.7096285 2.55385496

surgical.margin.status 2.45907664 0.0224899 1.13538824 5.32598251 lymph.node nvasion 1.12795178 0.86503976 0.28137722 4.52159984

Adjuvant.RT 0.23365931 0.07056508 0.04832766 1.12971888

Adiuvant.ADT 0.35383598 0.35083903 0.03989583 3.1381697

Table D. Univariate Cox Proportional Hazard Model analyses on BCR-free survival for two- gene models A (PITX2*0.022862 + HOXD3*0.0014012) and B (PITX2*0.020677 + HOXD3*0.0043132). HR, Hazard ratio; BCR, Biochemical recurrence; CI, confidence intervals

Gene name Cohort Univariate

HR (95%C1) -value

Wald test

Two-gene model A PCa2A 4.3517 (1.9105-9.9119) 0.00046288

PITX2*0.022862 + PCal 11.1273 (2.1405-57.8446) 0.0041718

HOXD3*0.0014012

Two-gene model B PCa2 4.8467 (2.0259-11.5953) 0.00039066

PITX2*0.020677 + PCalA 11.83 (2.1518-65.0368) 0.0044938

HOXD3*0.0043132

Table E. Multivariate Cox Proportional Hazard Model analyses on BCR-free survival for two- gene model A (PITX2*0.022862 + HOXD3*0.0014012), pre-operative PSA, clinical stage, final Gleason score, surgical margin status, lymph node invasion, adjuvant radiotherapy (RT) and adjuvant hormonal therapy (ADT) in cohort PCa2A. HR, Hazard ratio; BCR, Biochemical recurrence; CI, confidence intervals.

Variables HR -value lower Upper

Two gene model A 3.306545663 0.0065225 1.4 7.83

pre.op.PSA 1.018990843 0.0146227 1 1.03

Clinical stage 1.155295304 0.6829703 0.58 2.31

final.Gleason.Score 1.741843289 0.0313191 1.05 2.89

surgical.margin.status 2.583683001 0.0097143 1.26 5.31

lymph.node nvasion 0.554591461 0.3967029 0.14 2.17

Adjuvant.RT 0.190680152 0.0160051 0.05 0.73

Adiuvant.ADT 0.509708341 0.2708728 0.15 1.69

Table F. Multivariate Cox Proportional Hazard Model analyses on BCR-free survival for two- gene model B (PITX2*0.020677 + HOXD3*0.0043132), pre-operative PSA, pathological stage, final Gleason score, surgical margin status, lymph node invasion, adjuvant radiotherapy (RT) and adjuvant hormonal therapy (ADT) in cohort PCa2. HR, Hazard ratio; BCR, Biochemical recurrence, CI, confidence intervals.

95 %CI

Variables HR -value lower upper

Two gene model B 3.077472213 0.04047 1.05 9.02

pre.op.PSA 1.013027074 0.1431055 1 1.03

pathological.T.stage.grouped 3.256213335 0.00199 1.54 6.88

final. Gleason. Score 1.502266441 0.2330766 0.77 2.93

surgical.margin.status 2.328480421 0.03715 1.05 5.16

lymph.node nvasion 1.046749296 0.9488769 0.26 4.23

Adjuvant.RT 0.230679729 0.0731373 0.05 1.15

Adjuvant.ADT 0.401506084 0.4195107 0.04 3.68

Table G. Univariate Cox Proportional Hazard Model analyses on Clinical failure-free survival for DNA methylation of 11 genes in cohort PCa2. HR, Hazard ratio;, CI, confidence intervals.

Overall-

WaldTest Overall- 95% CI 95% CI

Gene name -Value HR lower Upper

GSTP1 0.71434 1.0056 0.97586 1.0363

TDRD1 0.98166 1.0002 0.9795 1.0214

PTGS2 0.80424 1.0048 0.96746 1.0436

APC 0.58334 1.0078 0.9802 1.0362

CCND2 0.36691 1.0199 0.97712 1.0647

HOXD3 0.095163 1.0257 0.99559 1.0566

PITX2 0.034812 1.0328 1.0023 1.0642

LGALS3 0.45699 1.0087 0.98596 1.0319

RASSF1A 0.63716 1.007 0.97836 1.0364

RARB 0.31482 1.0137 0.98719 1.0408

CDH13 0.084544 1.0419 0.99443 1.0916

Table H. Multivariate Cox Proportional Hazard Model analyses on CF-free survival DNA methylation of PITX2, pre-operatieve PSA, pathological stage, final Gleason score, surgical margin status, lymph node invasion, adjuvant radiotherapy and adjuvant hormonal therapy in cohort PCa2. HR, Hazard ratio; CF, clinical failure

95% CI 95% CI

Variable HR -value lower upper

PITX2 1.0401953 0.04948 1 1.1

pre.op.PSA 1.02821719 0.069263 1 1.1

pathological.T.stage.grouped 4.73715966 0.02729 1.2 19

final.Gleason.Score 13.4464008 0.00357 2.3 77

surgical.margin.status 1.35815199 0.695634 0.3 6.3

lymph.node nvasion 3.91968146 0.176361 0.5 28

Adjuvant.RT 0.08864632 0.114363 0 1.8

Adjuvant.ADT 4.24E-09 0.998585 0 Inf

Literature

1. Ibragimova I, Ibanez de Caceres I, Hoffman AM, Potapova A, Dulaimi E, Al-Saleem T, Hudes GR, Ochs MF, Cairns P. Global reactivation of epigenetically silenced genes in prostate cancer. Cancer Prev Res (Phila). 2010;3(9): 1084-92. 2. Kim JH, Dhanasekaran SM, Prensner JR, Cao X, Robinson D, Kalyana-Sundaram S, Huang C, Shankar S, Jing X, Iyer M, Hu M, Sam L, Grasso C, Maher CA, Palanisamy N, Mehra R, Kominsky HD, Siddiqui J, Yu J, Qin ZS, Chinnaiyan AM. Deep sequencing reveals distinct patterns of DNA methylation in prostate cancer. Genome Res. 2011;21(7): 1028-41.

3. Mahapatra S, Klee EW, Young CY, Sun Z, Jimenez RE, Klee GG, Tindall DJ, Donkena KV. Global methylation profiling for risk prediction of prostate cancer. Clin Cancer

Res. 2012 May 15;18(10):2882-95.

4. Chung W, Kwabi-Addo B, Ittmann M, Jelinek J, Shen L, Yu Y, Issa JP. Identification of novel tumor markers in prostate, colon and breast cancer by unbiased methylation profiling. PLoS One. 2008 Apr 30;3(4):e2079. 5. Ahmed H, Cappello F, Rodolico V, Vasta GR. Evidence of heavy methylation in the galectin 3 promoter in early stages of prostate adenocarcinoma: development and validation of a methylated marker for early diagnosis of prostate cancer. Transl Oncol. 2009;2(3): 146- 56.

6. Tibshirani R, Hastie T, Friedman J. Regularization Paths for Generalized Linear Models via Coordinate Descent. Journal of Statistical Software. 2010;33: 1-22.

Claims

2. The method according to claim 1, wherein step (b) comprises analyzing the methylation level of CpG dinucleotides in the regulatory region surrounding the transcription start sites of PITX2 and HOXD3, using the following formula:

3. The method according to claims 1 or 2, wherein hypermethylation of CpG dinucleotides in the regulatory region surrounding the transcription start site of PITX2 and HOXD3 is predictive for the outcome and/or indicative of the type of a prostate cell proliferative disorder as a more aggressive type and/or of a further advanced stage.

4. The method according to anyone of claims 1-3 wherein said prostate cell proliferative disorder of a more aggressive type is a biochemical recurrent type of disorder.

5. The method according to anyone of claims 1-4 wherein said prostate cell proliferative disorder of a more advanced stage is a high clinical stage disorder of pT stage III or IV.

6. The method of any of claims 1 to 5, for use in deciding on the proper treatment or proper medicament dependent on the type and/or stage of said prostate cell proliferative disorder.

7. The method of claim 6, wherein detection of hypermethylation of PITX2 and HOXD3, is indicative for the decision about the initiation or continuation of a treatment selected from a prostatectomy, a histone deacetylase inhibitors, a DNA methylation inhibitor, a gonadotropin-releasing hormone agonists, a neutraceutical, radiotherapy, or a compound in an effective amount to reduce male hormones.

8. The method of any of claims 1 to 7, wherein methylation is determined using PCR analysis, bisulfite genomic sequencing PCR analysis, Methylation-Specific PCR analysis or an equivalent amplification technique.

9. The method of any of claims 1 to 8, wherein methylation is determined in an assay comprising primers for assessing the presence of methylation in a regulatory region surrounding the TSS of said genes.

10. The method of any of claims 1 to 9, wherein at least one primer of the group consisting of methylated specific primers for PITX2 (SEQ ID N° 67 and 68), and at least one primer of the group consisting of methylated specific primers for HOXD3 (SEQ ID N° 61 and 62) and at least one primer of the group consisting of unmethylated specific primers for PITX2 (SEQ ID N° 69 and 70) and at least one primer of the group consisting of unmethylated specific primers for HOXD3 (SEQ ID N° 63 and 64) is used.

11. The method of any of claims 1 to 10, wherein the reference sample is a sample from a healthy individual or from an individual having a typical benign hyperplasia prostate.

12. The method of any of claims 1 to 11, wherein said regulatory region surrounding the TTS comprises one or more CpG islands and extends about 1.5 kb upstream to about 1.5 kb downstream from said transcription start site of said gene(s).

13. The method according to claim 12, wherein the regulatory region surrounding the transcription start site of the gene PITX2 corresponds to position about -16 to about +73 of the transcription start site.

14. The method according to claim 12, wherein the regulatory region surrounding the transcription start site of the gene HOXD3 corresponds to position about 909 to 823 upstream of the transcription start site.

15. The method of any of claims 1 to 14, wherein the test and/or reference sample is selected from the list comprising prostatic tissue, prostatic fluid, seminal fluid, ejaculate, blood, urine, prostate secretions, histological slides, and paraffin-embedded tissue.

16. The method of any of claims 1 to 15, further comprising analysing the level of DNA methylation of control plasmids comprising the inserts represented by SEQ ID N° 98 and 99.

17. A kit for typing and/or staging a prostate cell proliferative disorder in a human male subject, comprising at least one primer of the group consisting of methylated specific primers for PITX2 (SEQ ID N° 67 and 68), and at least one primer of the group consisting of methylated specific primers for HOXD3 (SEQ ID N° 61 and 62) and at least one primer of the group consisting of unmethylated specific primers for PITX2 (SEQ ID N° 69 and 70) and at least one primer of the group consisting of unmethylated specific primers for HOXD3 (SEQ ID N° 63 and 64) is used.

18. The kit according to claim 17 further comprising control plasmids comprising the inserts represented by SEQ ID N° 98 and 99.