Personalization in prostate cancer using the prostate cancer PDE4D7 knockdown score

Abstract

The present invention relates to a method of predicting a response of a prostate cancer subject to radiation therapy, the method comprising the steps of: determining or receiving a gene expression profile of each of three or more PDE4D7 knockdown differentially expressed genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2, wherein the gene expression profile is determined in a biological sample obtained from the prostate cancer subject; and determining a prediction of radiation therapy response based on the gene expression profiles of the three or more PDE4D7 knockdown genes, wherein the radiation therapy is a definitive radiation therapy or a salvage radiation therapy.

Description

[Technical field]

[0001] The present invention relates to a method for predicting the response of a subject with prostate cancer to radiation therapy, and a computer program for predicting the response of a subject with prostate cancer to radiation therapy.Furthermore, the present invention relates to a diagnostic kit, the use of said diagnostic kit, the use of said diagnostic kit in a method for predicting the response of a subject with prostate cancer to radiation therapy, the use of a respective gene expression profile of one or more PDE4D7 knockdown responsive genes in a method for predicting the response of a subject with prostate cancer to radiation therapy, and a corresponding computer program. [Background technology]

[0002] Cancer is a class of diseases in which a group of cells exhibit uncontrolled proliferation, invasion, and sometimes metastasis. These three malignant properties of cancer distinguish it from benign tumors that are self-limited and do not invade or metastasize. Prostate cancer (PCa) is the second most common non-cutaneous malignancy in men, with an estimated 1.3 million new cases diagnosed worldwide in 2018, resulting in 360,000 deaths (see Bray F. et al., "Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries," CA Cancer J Clin, Vol. 68, No. 6, pp. 394-424, 2018). In the United States, approximately 90% of new cases involve localized cancer, meaning that metastases have not yet formed (see Cancer Facts & Figures 2010, ACS, 2010).

[0003] Several definitive therapies are available for the treatment of primary localized prostate cancer, of which surgery (radical prostatectomy, RP) and radiation therapy (RT) are the most commonly used. RT is administered by external beam or by implantation of radioactive seeds in the prostate (brachytherapy), or a combination of both. RT is particularly preferred for patients who are not candidates for surgery or who are diagnosed with advanced stage tumors, either localized or regional. Definitive RT is offered to up to 50% of patients diagnosed with localized prostate cancer in the United States (see ACS, 2010, ibid.).

[0004] After treatment, prostate cancer antigen (PSA) levels are measured in the blood for disease monitoring. An increase in blood PSA level provides a biochemical surrogate measurement for cancer recurrence or progression. However, there is a large variability in reported biochemical progression-free survival (bPFS) (see Grimm P. et al., Comparative analysis of prostate-specific antigen free survival outcomes for patients with low, intermediate and high risk prostate cancer treatment by radical therapy. Results from the Prostate Cancer Results Study Group, BJU Int, Suppl. 1, pp. 22-29, 2012). For many patients, bPFS after definitive RT exceeds 90% at 5 years or even 10 years. Unfortunately, for patients with intermediate risk of recurrence and especially for the group of patients with higher risk of recurrence, bPFS can drop to approximately 40% at 5 years, depending on the type of RT used (see Grimm P. et al., supra, 2012).

[0005] Many patients with primary localized prostate cancer not treated with RT will undergo RP (ACS, 2010, ibid.). After RP, an average of 60% of patients in the highest risk group will experience biochemical recurrence after 5 and 10 years (Grimm P. et al., 2012, ibid.). In cases of biochemical progression after RP, one of the main challenges is the uncertainty of whether this is due to localized disease recurrence, one or more metastases, or even indolent disease that does not lead to clinical disease progression (see Dal Pra A. et al., "Contemporary role of postoperative radiotherapy for prostate cancer," Transl Androl Urol, Vol. 7, No. 3, pp. 399-413, 2018, and Herrera FG and Berthold DR, "Radiation therapy after radical prostatectomy: Implications for clinicians," Front Oncol, Vol. 6, No. 117, 2016). RT to eradicate remaining cancer cells in the prostate bed is one of the main treatment options to salvage survival after PSA increase after RP. The efficacy of salvage radiotherapy (SRT) has resulted in a 5-year bPFS in 18%-90% of patients depending on multiple factors (see Herrera FG and Berthold DR, 2016, ibid., and Pisansky TM et al., "Salvage radiation therapy dose response for biochemical failure of prostate cancer after prostatectomy - A multi-institutional observational study," Int J Radiat Oncol Biol Phys, Vol. 96, No. 5, pp. 1046-1053, 2016).

[0006] It is clear that for certain patient groups, definitive or salvage RT is ineffective. Their situation is further exacerbated by the serious side effects that RT can cause, such as intestinal inflammation and dysfunction, urinary incontinence, and erectile dysfunction (see Resnick MJ et al., "Long-term functional outcomes after treatment for localized prostate cancer," N Engl J Med, Vol. 368, No. 5, pp. 436-445, 2013, and Hegarty SE et al., "Radiation therapy after radical prostatectomy for prostate cancer: Evaluation of complications and influence of radiation timing on outcomes in a large, population-based cohort," PLoS One, Vol. 10, No. 2, 2015). Furthermore, the median cost of one course of RT, based on Medicare reimbursement, is $18,000 and varies widely up to approximately $40,000 (see Paravati AJ et al., "Variation in the cost of radiation therapy among medicare patients with cancer," J Oncol Pract, Vol. 11, No. 5, pp. 403-409, 2015). These figures do not include the substantial long-term costs of follow-up care after definitive and salvage RT.

[0007] Improved prediction of the efficacy of RT for each patient will lead to improved treatment choices and improved chances of survival, whether in the curative or salvage setting. This can be achieved by 1) optimizing RT for patients for whom RT is predicted to be effective (e.g., by dose escalation or altering the timing of initiation) and 2) directing patients for whom RT is predicted to be ineffective to alternative, potentially more effective, forms of treatment. This, in turn, reduces patient suffering by sparing ineffective treatments and reduces costs that would otherwise be spent on ineffective treatments.

[0008] Numerous studies have been performed on measurements for predicting response to definitive RT (see Hall WA et al., "Biomarkers of outcome in patients with localized prostate cancer treated with radiotherapy," Semin Radiat Oncol, Vol. 27, pp. 11-20, 2016, and Raymond E. et al., "An appraisal of analytical tools used in predicting clinical outcomes following radiation therapy treatment of men with prostate cancer: A systematic review," Radiat Oncol, Vol. 12, No. 1, p. 56, 2017) and definitive SRT (see Herrera FG and Berthold DR, ibid., 2016). Many of these measurements depend on the concentration of PSA, a blood-based biomarker. Metrics that have been investigated for predicting response before the initiation of RT (definitive and salvage) include absolute PSA concentration, absolute values ​​related to prostate volume, absolute increase over time, and doubling time. Other factors that are often considered are the Gleason score and clinical stage of the tumor. In the SRT setting, additional factors such as margin status, time to recurrence after RP, pre- / peri-surgical PSA levels, and clinicopathological parameters are pertinent.

[0009] Although these clinical variables have provided limited improvement in stratifying patients into various risk groups, better predictive tools are needed.

[0010] A wide range of potential biomarkers in tissues and body fluids are being explored, but validation is often limited and demonstrates general prognostic information, not (treatment specific) predictive value (see Hall WA et al., 2016, ibid.). A small number of gene expression panels are currently being validated by private organizations. One or several of these may show predictive value for RT in the future (see Dal Pra A. et al., 2018, ibid.). Summary of the Invention [Problem to be solved by the invention]

[0011] In conclusion, there remains a strong need for better prediction of response to RT for primary prostate cancer as well as for the post-surgical setting. [Means for solving the problem]

[0012] In a first aspect, the present invention relates to a method of predicting a response of a prostate cancer subject to radiotherapy, the method comprising the steps of: determining a gene expression level or receiving a result of determining a gene expression level of each of three or more genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2, wherein the gene expression levels are determined in a biological sample obtained from the prostate cancer subject; and determining a prediction of the radiotherapy response based on the gene expression levels of the three or more genes, wherein the prediction is a favorable or unfavorable response to radiotherapy, and wherein the radiotherapy is a definitive radiotherapy or a salvage radiotherapy.

[0013] In a second aspect, the present invention relates to a computer program comprising a plurality of instructions, the plurality of instructions being adapted to encode three or more of the following genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2, for example 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, The present invention relates to a computer program product that causes a computer to execute a method comprising the steps of: receiving data indicative of a gene expression profile for each of nine or all of the genes, wherein gene expression levels are determined in a biological sample obtained from the prostate cancer subject; and determining a prediction of the prostate cancer subject's response to treatment based on the gene expression profile(s) of the three or more genes, wherein the prediction is a favorable or unfavorable response to radiation therapy, and wherein the radiation therapy is curative radiation therapy or salvage radiation therapy.

[0014] In a third aspect, the present invention relates to a diagnostic kit comprising at least three sets of polymerase chain reaction primer pairs and optionally at least three probes for determining the expression level in a sample of three or more, such as 3, 4, 5, 6, 7, 8, 9 or all, genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2.

[0015] In a fourth aspect, the present invention relates to the use of a diagnostic kit according to the third aspect of the invention in a method for predicting the response of a prostate cancer subject to radiation therapy, preferably for use in a method according to the first aspect of the invention.

[0016] In a fifth aspect, the present invention relates to a method comprising the steps of receiving a biological sample obtained from a subject with prostate cancer and using the diagnostic kit of claim 12 to determine a gene expression profile of each of three or more, such as 3, 4, 5, 6, 7, 8, 9 or all genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2 in the biological sample obtained from the subject with prostate cancer. [Brief description of the drawings]

[0017] [Figure 1] Figure 1 shows the Kaplan-Meier curve of PDE4D7_KD_model in a cohort of 185 patients (the training set used to develop PDE4D7_KD_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients of the PDE4D7_KD_model class analyzed. That is, the risk patients in all +20 months after surgery are shown. [Diagram 2]Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical_model in a cohort of 185 patients (the training set used to develop the PDE4D7_clinical_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients for the PDE4D7_clinical_model classes analyzed, i.e., risk patients in all +20 months after surgery are shown. [Diagram 3] Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical class_model in a cohort of 185 patients (the training set used to develop the PDE4D7_clinical class_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients for the PDE4D7_clinical class_model classes analyzed, i.e., risk patients in all +20 months after surgery are shown. [Figure 4]Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical class_model in a cohort of 381 patients (the training set used to develop the PDE4D7_clinical class_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients for the PDE4D7_clinical class_model classes analyzed, i.e., risk patients in all +20 months after surgery are shown. [Diagram 5] Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD_model in a cohort of 169 patients (the training set used to develop the PDE4D7_KD_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was metastasis-free survival (metastasis) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log-rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients for the PDE4D7_KD_model classes analyzed, i.e., risk patients in all +20 months after surgery are shown. [Figure 6]Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical_model in a cohort of 169 patients (the training set used to develop the PDE4D7_clinical_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was metastasis-free survival (metastasis) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log-rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_clinical_model classes analyzed, i.e., at risk patients in all +20 months after surgery. [Figure 7] Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical class_model in a cohort of 169 patients (the training set used to develop the PDE4D7_clinical class_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was metastasis-free survival (metastasis) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log-rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients for the PDE4D7_clinical class_model classes analyzed, i.e., risk patients in all +20 months after surgery are shown. [Figure 8]Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical class_model in a cohort of 379 patients (the training set used to develop the PDE4D7_clinical class_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was metastasis-free survival (metastasis) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log-rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients for the PDE4D7_clinical class_model classes analyzed, i.e., risk patients in all +20 months after surgery are shown. [Figure 9] Figure 1 shows the Kaplan-Meier curve of PDE4D7_KD_model in a cohort of 185 patients (the training set used to develop PDE4D7_KD_model) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients of the PDE4D7_KD_model class analyzed. That is, risk patients in all +20 months after surgery are shown. [Figure 10]Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical_model in a cohort of 185 patients (the training set used to develop the PDE4D7_clinical_model) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_clinical_model classes analyzed, i.e., patients at risk in all +20 months after surgery. [Figure 11] Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical class_model in a cohort of 185 patients (the training set used to develop the PDE4D7_clinical class_model) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_clinical class_model classes analyzed, i.e., at risk patients in all +20 months after surgery. [Figure 12]Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical class_model in a cohort of 381 patients (the training set used to develop the PDE4D7_clinical class_model) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients of the PDE4D7_clinical class_model classes analyzed, i.e., risk patients in all +20 months after surgery are shown. [Figure 13] Figure 1 shows the Kaplan-Meier curve of PDE4D7_Dose Gy class_model in a cohort of 47 patients (the training set used to develop the PDE4D7_Dose Gy class_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery in subjects who received high (>66Gy) or low (<=66Gy) radiation therapy doses. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients for the PDE4D7_dose Gy class_model classes analyzed. That is, risk patients in all +20 months after surgery are shown. [Figure 14]Figure 1 shows the Kaplan-Meier curves of the PDE4D7_dose Gy class_model in a cohort of 51 patients (the training set used to develop the PDE4D7_dose Gy class_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery in subjects who received high (>66Gy) or low (<=66Gy) radiation therapy doses. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients for the PDE4D7_dose Gy class_model classes analyzed. That is, risk patients in all +20 months after surgery are shown. [Figure 15] Figure 1 shows the Kaplan-Meier curve of PDE4D7_KD_model in a cohort of 151 patients (test set used to validate the PDE4D7_KD_model developed in the training set) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after starting salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence interval are shown in the figure. The included supplementary list shows the number of risk patients of the PDE4D7_KD_model class analyzed. That is, the risk patients in all +20 months after surgery are shown. [Figure 16]Figure 1 shows the Kaplan-Meier curve of PDE4D7_clinical_model in a cohort of 151 patients (test set used to validate the PDE4D7_clinical_model developed in the training set) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients of the PDE4D7_clinical_model class analyzed. That is, the risk patients in all +20 months after surgery are shown. [Figure 17] Figure 1 shows the Kaplan-Meier curve of PDE4D7_clinical class_model in a cohort of 151 patients (test set used to validate the PDE4D7_clinical class_model developed in the training set) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients of the PDE4D7_clinical class_model classes analyzed. That is, the risk patients in all +20 months after surgery are shown. [Figure 18]Figure 1 shows the Kaplan-Meier curve of PDE4D7_clinical class_model in a cohort of 136 patients (test set used to validate the PDE4D7_clinical class_model developed in the training set) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients of the PDE4D7_clinical class_model classes analyzed. That is, the risk patients in all +20 months after surgery are shown. [Figure 19] Figure 1 shows the Kaplan-Meier curve of PDE4D7_KD_model in a cohort of 145 patients (test set used to validate the PDE4D7_KD_model developed in the training set) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was metastasis-free survival (metastasis) after starting salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence interval are shown in the figure. The included supplementary list shows the number of risk patients of the PDE4D7_KD_model class analyzed. That is, the risk patients in all +20 months after surgery are shown. [Figure 20]Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical_model in a cohort of 145 patients (the test set used to validate the PDE4D7_clinical_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was metastasis-free survival (metastasis) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log-rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_clinical_model classes analyzed, i.e., the risk patients in all +20 months after surgery are shown. [Figure 21] Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical_model in a cohort of 145 patients (the test set used to validate the PDE4D7_clinical_model), all of whom underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was metastasis-free survival (metastasis) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log-rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_clinical_model classes analyzed, i.e., the risk patients in all +20 months after surgery are shown. [Figure 22]Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical_model in a cohort of 131 patients (the test set used to validate the PDE4D7_clinical_model) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was metastasis-free survival (metastasis) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_clinical_model classes analyzed, i.e., at risk patients in all +20 months after surgery. [Diagram 23] Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD_model in a cohort of 151 patients (the test set used to validate the PDE4D7_KD_model) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients of the PDE4D7_KD_model class analyzed, i.e., risk patients in all +20 months after surgery are shown. [Figure 24] Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical_model in a cohort of 151 patients (the test set used to validate the PDE4D7_clinical_model) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_clinical_model classes analyzed, i.e., at risk patients in all +20 months after surgery. [Diagram 25] Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical class_model in a cohort of 151 patients (the test set used to validate the PDE4D7_clinical class_model) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_clinical class_model classes analyzed, i.e., at risk patients in all +20 months after surgery. [Figure 26] Figure 1 shows the Kaplan-Meier curves of the PDE4D7_clinical class_model in a cohort of 136 patients (the test set used to validate the PDE4D7_clinical class_model) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was prostate cancer-specific survival (PCa Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_clinical class_model classes analyzed, i.e., at risk patients in all +20 months after surgery. [Figure 27] FIG. 1 shows an outline of the know down strategy used to identify an extensive set of PDE4D7 knockdown genes (113 genes). [Figure 28] FIG. 1 outlines the strategy used to select and validate the set of 28 genes. [Figure 29]FIG. 1 shows control (scrambled (top); untreated and scrambled (bottom)) PDE4D7 mRNA expression levels in LNCaP cells and PDE4D7 shRNA-transfected LNCaP cells. [Diagram 30] Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD 3.1_model in a cohort of 185 patients (model using 3 randomly selected genes from the PDE4D7 KD gene set, namely BRCA1, FOXA1, KLK2) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_KD_model classes analyzed, i.e., in all +20 months after surgery. [Diagram 31] Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD 3.2_model in a cohort of 185 patients (model using 3 randomly selected genes from the PDE4D7 KD gene set, namely MLH1, KLK3, and HOXB13), all of whom underwent SRT after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation therapy (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_KD_model classes analyzed, i.e., in all +20 months after surgery. [Diagram 32]Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD 3.3_model in a cohort of 185 patients (model using 3 randomly selected genes from the PDE4D7 KD gene set, namely ATM, FANCA, MYO6) in which all patients underwent SRT (salvage radiation treatment) after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation treatment (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_KD_model classes analyzed, i.e., in all +20 months after surgery. [Diagram 33] Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD 3.4_model in a cohort of 185 patients (model using 3 randomly selected genes from the PDE4D7 KD gene set, namely ATM, SLC45A3, and NQO1), all of whom underwent SRT after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation therapy (SRT) due to disease recurrence after surgery. Log-rank, HR, and confidence intervals are shown in the figure. The included supplementary list shows the number of risk patients for the PDE4D7_KD_model classes analyzed, i.e., risk patients in all +20 months after surgery. [Diagram 34]Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD 4.1_model in a cohort of 185 patients (model using 4 randomly selected genes from the PDE4D7 KD gene set, namely ETV1, BRCA1, ACPP, NRP1) in which all patients underwent SRT after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation therapy (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_KD_model classes analyzed, i.e., in all +20 months after surgery. [Diagram 35] Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD 4.2_model in a cohort of 185 patients (model using 4 randomly selected genes from the PDE4D7 KD gene set, namely PALB2, KLK3, AR, FANCA) in which all patients underwent SRT after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation therapy (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_KD_model classes analyzed, i.e., in all +20 months after surgery. [Diagram 36]Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD 4.3_model in a cohort of 185 patients (model using 4 randomly selected genes from the PDE4D7 KD gene set, namely CDH1, SPDEF, ATM, NAALADL2) in which all patients underwent SRT after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation therapy (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_KD_model classes analyzed, i.e., in all +20 months after surgery. [Figure 37] Figure 1 shows the Kaplan-Meier curves of PDE4D7_KD 4.4_model in a cohort of 185 patients (model using 4 randomly selected genes from the PDE4D7 KD gene set, namely EHF, KLK2, NQO1, MME) in which all patients underwent SRT after BCR (biochemical recurrence) after surgery. The clinical endpoint tested was overall survival (Death) after initiation of salvage radiation therapy (SRT) due to disease recurrence after surgery. Log rank, HR and confidence intervals are shown in the figure. The included supplementary list shows the number of patients at risk for the PDE4D7_KD_model classes analyzed, i.e., in all +20 months after surgery. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] Phosphodiesterases (PDEs) provide the only means to degrade the second messenger 3'-5'-cyclic AMP. As such, PDEs may play an important regulatory role. Thus, abnormal changes in PDE expression, activity, and subcellular location may all contribute to the molecular pathology underlying certain disease states. Indeed, it has recently been shown that prostate cancer patients have frequent mutations in PDE genes, which may elevate cAMP signaling and also predispose to prostate cancer. However, the various expression profiles in different cell types, coupled with the complex and diverse isoform variants within each PDE family, make it difficult to understand the link between abnormal changes in PDE expression and functionality during disease progression. Several studies have attempted to describe the total amount of PDEs in the prostate, all of which identified significant levels of PDE4 expression along with other PDEs.

[0019] The currently identified PDE isoforms were previously analyzed for expression in 19 prostate cancer cell lines and xenografts using sequence information [Henderson, 2014]. In such studies, PDE3B, PDE4B, PDE4D, PDE7A, PDE8A, PDE8B and PDE9A isoforms were identified as being abundantly expressed at the mRNA level in cancer prostate cells [Henderson, 2014], whereas PDE1, PDE3A, PDE5A, PDE10A and PDE11A mRNA levels were lower (unpublished data), highlighting the complexity of cyclic nucleotide signaling in the prostate epithelium. Importantly, by separating prostate cancer cell samples into androgen-sensitive and androgen-insensitive castration-resistant prostate cancer (CRPC) cell phenotypes, we found that the expression of PDE4D isoforms was downregulated in CRPC samples. In particular, PDE4D7, the most abundant PDE4 isoform in many androgen-sensitive samples, was found to be downregulated to a significant degree in CRPC cell models, suggesting a scenario in which PDE4D7 downregulation may directly contribute to altered cAMP signaling driven by disease progression.Furthermore, these observations suggest that PDE4D7, when measured, may be informative for prostate cancer disease progression, where low levels of expression may be associated with a more aggressive phenotype.

[0020] Based on the correlation of PDE4D7 expression with the pathological features of the disease, we specifically aimed to identify a prognostic association between the expression of PDE4D7 in patients' prostate tissues obtained either by biopsy or surgery and clinically useful information on the outcome of individual patients. Clinically meaningful endpoints or surrogate endpoints that significantly correlate with the development of metastasis, cancer-specific mortality or overall mortality are usually evaluated as prognostic cancer biomarkers. The most appropriate rationale for using surrogate endpoints relates to situations where either data on established clinical endpoints are unavailable or the number of events in the data cohort is too small for statistical data analysis. To develop a PDE4D7 prognostic biomarker, we evaluated either BCR (biochemical recurrence) progression-free survival or initiation of second-line treatment after surgery as surrogate endpoints for metastasis and prostate cancer death. These specific endpoints were used to identify the number of events associated in the clinical cohort (e.g., for BCR, >30%), which is particularly relevant for multivariate data analysis.

[0021] In the evaluation, standard methods of multivariate analysis, such as Cox regression and Kaplan-Meier survival analysis, were selected to investigate the additive and independent value of the continuous and / or categorical "PDE4D7 score" compared with established prognostic clinical variables, such as PSA and Gleason score [Alves de Inda, 2018]. Logistic regression was then used to build risk models combining the "PDE4D7 score" with either pre- or post-operative clinical predictors of progression after surgery. The resulting models were then tested with Kaplan-Meier survival and ROC curve analysis in multiple independent patient cohorts to predict progression-free survival after treatment [Alves de Inda, 2018].

[0022] Using such a strategy, we started to test the prognostic value of the PDE4D7 score in biopsies from retrospectively collected resected prostate tissue in a consecutively managed patient cohort from a single surgical center in the post-surgical setting [Alves de Inda, 2018]. The patient population consisted of approximately 500 individuals with long-term follow-up of both pathological and biological outcomes. These clinical data were available for all patients and were collected during a median follow-up of 120 months after treatment. The "PDE4D7 score" was determined as described above and tested in both univariate and multivariate analyses, using post-surgical covariates available to adjust for the multivariate setting (i.e., pathological Gleason score, pT stage, resection margin status, seminal vesicle invasion status, and lymph node invasion status). In this example, biochemical progression-free survival after the primary intervention was set as the clinical endpoint to be evaluated. Univariate analysis of these clinical samples [Alves de Inda, 2018] showed an inverse correlation between PDE4D7 expression (in terms of the "PDE4D7 score") and biological recurrence after surgery (HR per unit change = 0.53; 95% CI 0.41-0.67; p<0.0001), strongly supporting our previous data [Boettcher 2015; Boettcher, 2016]. In a multivariate analysis using such clinical variables, the "PDE4D7 score" remained an independently effective means of predicting clinical outcome (HR = 0.56 / unit change; 95% CI 0.43-0.73; p<0.0001). Moreover, in multivariate analysis, very similar results were obtained when the "PDE4D7 score" was evaluated using the validated and clinically used risk model CAPRA-S (HR=0.54 95%CI 0.42-0.69; p<0.0001). The CAPRA-S score is based on preoperative PSA and pathological parameters determined at the time of surgery and was developed and validated in US and other populations to provide clinicians with information aimed at helping predict disease recurrence, including BCR, systemic progression, and PCSM.

[0023] Interestingly, when assessing the hazard ratio (HR) compared to the continuous "PDE4D7 score", we revealed a linear increase in risk with decreasing "PDE4D7 score" for score values ​​between 2 and 5. However, for PDE4D7 scores below 2, the risk of progression after surgery increases sharply [Alves de Inda, 2018]. This is also evident in the Kaplan-Meier survival curves, where patients falling into the lowest "PDE4D7 score" category have the highest risk of disease recurrence. Logistic regression analysis was used to combine the CAPRA-S score with the continuous "PDE4D7 score". When this model was tested using ROC curve analysis, a notable improvement of AUC of 4-6% was noticed compared to CAPRA-S alone for both 2- and 5-year prediction of progression after treatment for BCR. Thus, we evaluated a Cox regression combination model of combined CAPRA-S and "PDE4D7 score" in Kaplan-Meier survival analysis and compared it to the CAPRA-S score category alone. With this in mind, the added value in risk prediction when using a model of the combined “PDE4D7 and CAPRA-S” score was confirmed compared to using the clinical metric of the CAPRA-S score alone [Alves de Inda, 2018].

[0024] Following a diagnosis of prostate cancer, an accurate risk assessment must be performed before stratification to a defined first-line treatment. With this in mind, we decided to see if the prognostic use of the “PDE4D7 score” could be translated in a pre-surgical setting testing tumor tissue obtained from diagnostic needle biopsy samples [van Strijp 2018]. Here, needle biopsies were performed on 168 patients from a single diagnostic clinical center who underwent surgery as a first-line treatment. The minimum follow-up period for each patient was 60 months after this intervention. Clinical covariates used to adjust the “PDE4D7 score” in the multivariable analysis were age at surgery, pre-surgical PSA, PSA density, biopsy Gleason score, percentage of tumor-positive biopsy cores, percentage of tumor in biopsy and clinical cT stage. Here, we evaluated the utility of the “PDE4D7 score” and the combination of the “PDE4D7 and CAPRA” scores in comparison to the pre-surgical CAPRA score in a Cox regression analysis for biochemical recurrence [van Strijp 2018].

[0025] When evaluating this patient cohort, the "PDE4D7 score" was found to be inversely associated with BCR in multivariate analysis, adjusting for clinical variables (HR=0.43; 95%CI 0.29-0.63; p<0.0001) as well as clinical CAPRA score (HR=0.53; 95%CI 0.38-0.74; p=0.0001) [van Strijp 2018]. Kaplan-Meier analysis showed that, as before, in the post-surgery setting, the "PDE4D7 score" category was significantly associated with BCR progression-free survival (log-rank p<0.0001) and second-line treatment-free survival (log-rank p=0.01). A combined logistic regression model developed in the previous cohort was then used [van Strijp 2018]. This consisted of a combined "CAPRA and PDE4D7" score, which showed that patients in the highest "CAPRA and PDE4D7" combined score category were virtually at no risk of biochemical progression after surgery or transition to any second-line treatment. This logistic regression model was also evaluated using a receiver operating characteristic (ROC) curve analysis to predict BCR 5 years after surgery. This revealed a 5% increase in AUC over the CAPRA score alone (AUC = 0.82 vs. 0.77, respectively; p = 0.004). Decision curve analysis of the combined "CAPRA and PDE4D7" score model to determine whether to perform an intervention (e.g., surgery) based on an individual patient's risk threshold of experiencing disease progression after surgery confirmed the superior net benefit of using this combined score compared to either score alone across all decision thresholds [van Strijp 2018].

[0026] The efficacy of both definitive RT and SRT for localized prostate cancer is limited, especially for patients at high risk of recurrence, leading to disease progression and ultimately patient death. Prediction of treatment outcome is very complicated, as many factors contribute to treatment efficacy and disease recurrence. Important factors may not yet be identified, and the influence of other factors cannot be precisely determined. Several clinicopathological measurements are currently being investigated and applied in clinical practice to improve response prediction and treatment selection, and have brought about some improvements. Nevertheless, better prediction of response to definitive RT and SRT is still strongly needed to increase the success rate of these treatments.

[0027] Herein, a new molecule was identified by using PDE4D7 knockdown strategy, whose expression was shown to be significantly associated with mortality after definitive RT and SRT, and therefore is expected to improve the prediction of the efficacy of these treatments.Using LNCaP prostate cancer cell line, stable PDE4D7 knockdown line was generated and analyzed.Using this method, genes that are differentially expressed in knockdown cell line were tested and validated for predictive value in human patient cohorts to predict response to radiation therapy. It has been found that the genes ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11, and PALB2, when used alone or in combination, can predict a favorable or poor response of a prostate cancer subject to radiation therapy.

[0028] Thus, in a first aspect, the present invention provides a method of predicting a response of a prostate cancer subject to radiation therapy, said method comprising: determining a gene expression level or receiving a result of determining a gene expression level of each of three or more genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11, and PALB2, wherein the gene expression levels are determined in a biological sample obtained from the subject with prostate cancer; determining a prediction of radiation therapy response based on the gene expression levels of said three or more genes, said prediction being a favorable or unfavorable response to radiation therapy.

[0029] In an alternative embodiment, the present invention provides a method of predicting a prostate cancer subject's response to radiation therapy, the method comprising: determining a gene expression level or receiving a result of determining a gene expression level of each of three or more genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11, and PALB2, wherein the gene expression levels are determined in a biological sample obtained from the subject with prostate cancer; determining a prediction of radiation therapy response based on gene expression levels of the three or more genes, the prediction being a favorable or unfavorable response to radiation therapy, the three or more genes comprising: BRCA1, FOXA1, and KLK2; MLH1, KLK3, and HOXB13; ATM, FANCA, and MYO6; ATM, SlC45A3, and NQO1, ETV1, BRCA1, ACPP, and NRP1; PALB2, KLK3, AR, and FANCA; CDH1, SPDEF, ATM, and NAALADL2; or comprising EHF, KLK2, NQO1, and MME; The present invention relates to a method comprising the steps of:

[0030] The currently identified PDE isoforms were analyzed for expression in 19 prostate cancer cell lines and xenografts using sequence information (see Henderson DJ et al., "The cAMP phosphodiesterase-4D7 (PDE4D7) is downregulated in androgen-independent prostate cancer cells and mediates proliferation by compartmentalizing cAMP at the plasma membrane of VCaP prostate cancer cells," Br J Cancer, Vol. 110, No. 5, pp. 1278-1287, 2014). In such studies, PDE3B, PDE4B, PDE4D, PDE7A, PDE8A, PDE8B and PDE9A isoforms were identified as being highly expressed at the mRNA level in cancer prostate cells (see Henderson DJ et al., 2014, supra), whereas PDE1, PDE3A, PDE5A, PDE10A and PDE11A mRNA levels were lower (unpublished data), highlighting the complexity of cyclic nucleotide signaling in the prostate epithelium.Importantly, by separating prostate cancer cell samples into androgen-sensitive and androgen-insensitive castration-resistant prostate cancer (CRPC) cellular phenotypes, it was found that expression of the PDE4D isoform was downregulated in CRPC samples. In particular, PDE4D7, the most abundant PDE4 isoform in many androgen-sensitive samples, was found to be downregulated to a significant degree in CRPC cell models, suggesting a scenario in which PDE4D7 downregulation may directly contribute to altered cAMP signaling driven by disease progression.Furthermore, these observations suggest that PDE4D7, when measured, may be informative for prostate cancer disease progression, where low levels of expression may be associated with a more aggressive phenotype.

[0031] Based on the correlation of PDE4D7 expression with the pathological features of the disease, we specifically aimed to identify a prognostic association between the expression of PDE4D7 in patients' prostate tissues, either obtained by biopsy or surgery, and clinically useful information regarding the outcome of individual patients. Surrogate endpoints that are clinically meaningful or significantly correlated with the development of metastasis, cancer-specific mortality, or overall mortality are usually evaluated as prognostic cancer biomarkers. The most appropriate rationale for using surrogate endpoints relates to situations where either data on established clinical endpoints are unavailable or the number of events in the data cohort is too small for statistical data analysis. To develop a PDE4D7 prognostic biomarker, we evaluated either BCR (biochemical recurrence) progression-free survival, or the initiation of second-line treatment after surgery, as surrogate endpoints for metastasis and prostate cancer death. These specific endpoints were used to identify the number of associated events (e.g., for BCR, >30%) in the clinical cohort, which is particularly relevant for multivariate data analysis.

[0032] In the evaluation, standard methods of multivariate analysis, such as Cox regression and Kaplan-Meier survival analysis, were selected to explore the additive and independent value of the continuous and / or categorical "KD score" compared with established prognostic clinical variables, such as PSA and Gleason score (see Alves de Inda M. et al., "Validation of Cyclic Adenosine Monophosphate Phosphodiesterase-4D7 for its Independent Contribution to Risk Stratification in a Prostate Cancer Patient Cohort with Longitudinal Biological Outcomes", Eur Urol Focus, Vol. 4, No. 3, pp. 376-384, 2018). Logistic regression was then used to build risk models combining the "KD score" with either pre- or post-operative clinical predictors of postoperative progression. The resulting models were then tested with Kaplan-Meier survival and ROC curve analyses in multiple independent patient cohorts to predict progression-free survival after treatment (see Alves de Inda M. et al., 2018, supra).

[0033] Using such a strategy, we started to test the prognostic value of the KD score in biopsies from retrospectively collected resected prostate tissue in a consecutively managed patient cohort from a single surgical center in the post-surgical setting (see Alves de Inda M. et al., 2018, ibid.). The patient population consisted of about 500 individuals with long-term follow-up of both pathological and biological outcomes. These clinical data were available for all patients and were collected during a median follow-up of 120 months after treatment. The "KD score" was determined as described above and was tested in both univariate and multivariate analyses using post-surgical covariates available to adjust for the multivariate setting (i.e., pathological Gleason score, pT stage, resection margin status, seminal vesicle invasion status, and lymph node invasion status). In this example, biochemical progression-free survival after the primary intervention was set as the clinical endpoint to be evaluated.

[0034] The present invention is based on the idea that, since PDE4D7 biomarker has been proven to be a good predictor of radiotherapy response, if the markers that are differentially expressed upon PDE4D7 knockdown can be identified, it can help to enable better prediction of overall RT response.To achieve this, in the above strategy, a group of 28 genes (ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2) that can be used individually or together to predict the response of prostate cancer subjects to radiotherapy, as described in the examples.

[0035] The term "ACPP" refers to the acid phosphatase 3 gene, also known as ACP3 (Ensembl: ENSG00000014257; HGNC: 125). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_001099.5, which encodes the coding sequence CCDS3073, i.e., the nucleotide sequence as defined in SEQ ID NO: 1, which corresponds to the sequence of the coding sequence of the ACPP transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 2, which corresponds to the protein sequence as defined in NCBI protein accession reference sequence NP_001090.2, which encodes the ACPP polypeptide.

[0036] The term "ACPP" also refers to a nucleotide sequence that exhibits high homology to ACPP, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:1, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:2. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:2, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:1.

[0037] The term "AR" refers to the androgen receptor gene (Ensembl: ENSG00000169083; HGNC: 644).For example, the exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_000044.6, which encodes the coding sequence CCDS87754, i.e., the nucleotide sequence as defined in SEQ ID NO: 3, which corresponds to the sequence of the coding sequence of the AR transcript shown above, and encodes the corresponding amino acid sequence, for example, as defined in SEQ ID NO: 4, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_000035.2, which encodes the AR polypeptide.

[0038] The term "AR" also refers to a nucleotide sequence that exhibits high homology to AR, e.g., a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:3, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:4. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:4, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:3.

[0039] The term "CDH1" refers to the Cadherin 1 gene (Ensembl: ENSG00000039068; HGNC: 1748). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_004360.5, which encodes the coding sequence CCDS82005, i.e., the nucleotide sequence as defined in SEQ ID NO: 5, which corresponds to the sequence of the coding sequence of the CDH1 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 6, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_004351.1, which encodes the CDH1 polypeptide.

[0040] The term "CDH1" also refers to a nucleotide sequence that exhibits high homology to CDH1, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:5, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:6. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:6, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:5.

[0041] The term "EHF" refers to the ETS homology factor gene (Ensembl: ENSG00000135373; HGNC: 3246). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_012153.6, which encodes the coding sequence CCDS55752, i.e., the nucleotide sequence as defined in SEQ ID NO: 7, which corresponds to the sequence of the coding sequence of the EHF transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 8, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_036285.2, which encodes the EHF polypeptide.

[0042] The term "EHF" also refers to a nucleotide sequence that exhibits high homology to EHF, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:7, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:8. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:8, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:7.

[0043] The term "ETV1" refers to the ETS variant 1 gene (Ensembl: ENSG00000006468; HGNC: 3490). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_004956.5, which encodes the coding sequence CCDS55085, i.e., the nucleotide sequence as defined in SEQ ID NO: 9, which corresponds to the sequence of the coding sequence of the ETV1 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 10, which corresponds to the protein sequence as defined in NCBI protein accession reference sequence NP_004947.2, which encodes the ETV1 polypeptide.

[0044] The term "ETV1" also refers to a nucleotide sequence that exhibits high homology to ETV1, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:9, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:10. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:10, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:9.

[0045] The term "FOLH1" refers to the folate hydrolase 1 gene (Ensembl: ENSG00000086205; HGNC: 3788). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_004476.3, which encodes the coding sequence CCDS31493, i.e., the nucleotide sequence as defined in SEQ ID NO: 11, which corresponds to the sequence of the coding sequence of the FOLH1 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 12, which corresponds to the protein sequence as defined in NCBI protein accession reference sequence NP_004467.1, which encodes the FOLH1 polypeptide.

[0046] The term "FOLH1" also refers to a nucleotide sequence that exhibits high homology to FOLH1, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:11, or an nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:12. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:12, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:11.

[0047] The term "FOXA1" refers to the Forkhead Box A1 gene (Ensembl: ENSG00000129514; HGNC: 5021). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_004496.5, which encodes the coding sequence CCDS9665, i.e., the nucleotide sequence as defined in SEQ ID NO: 13, which corresponds to the sequence of the coding sequence of the FOXA1 transcript shown above, and encodes the corresponding amino acid sequence, for example, as defined in SEQ ID NO: 14, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_004487.2, which encodes the FOXA1 polypeptide.

[0048] The term "FOXA1" also refers to a nucleotide sequence that exhibits high homology to FOXA1, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:13, or an nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:14. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:14, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:13.

[0049] The term "HOXB13" refers to the homeobox B13 gene (Ensembl: ENSG00000159184; HGNC: 5112). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_006361.6, which encodes the coding sequence CCDS11536, i.e., the nucleotide sequence as defined in SEQ ID NO: 15, which corresponds to the sequence of the coding sequence of the HOXB13 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 16, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_006352.2, which encodes the HOXB13 polypeptide.

[0050] The term "HOXB13" also refers to a nucleotide sequence that exhibits high homology to HOXB13, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:15, or at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:16. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:16, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:15.

[0051] The term "KLK2" refers to the kallikrein-related peptidase 2 gene (Ensembl: ENSG00000167751; HGNC: 6363). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_005551.5, which encodes the coding sequence CCDS12808, i.e., the nucleotide sequence as defined in SEQ ID NO: 17, which corresponds to the sequence of the coding sequence of the KLK2 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 18, which corresponds to the protein sequence as defined in NCBI protein accession reference sequence NP_005542.1, which encodes the KLK2 polypeptide.

[0052] The term "KLK2" also refers to a nucleotide sequence that exhibits high homology to KLK2, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:17, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:18. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:18, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:17.

[0053] The term "KLK3" refers to the kallikrein-related peptidase 3 gene (Ensembl: ENSG00000142515; HGNC: 6364). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_001648.2, which encodes the coding sequence CCDS12807, i.e., the nucleotide sequence as defined in SEQ ID NO: 19, which corresponds to the sequence of the coding sequence of the KLK3 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 20, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_001639.1, which encodes the KLK3 polypeptide.

[0054] The term "KLK3" also refers to a nucleotide sequence that exhibits high homology to KLK3, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:19, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:20. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:20, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:19.

[0055] The term "MAOA" refers to the monoamine oxidase A gene (Ensembl: ENSG00000189221; HGNC: 6833). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_000240.4, which encodes the coding sequence CCDS14260, i.e., the nucleotide sequence as defined in SEQ ID NO: 21, which corresponds to the sequence of the coding sequence of the MAOA transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 22, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_000231.1, which encodes the MAOA polypeptide.

[0056] The term "MAOA" also refers to a nucleotide sequence that exhibits high homology to MAOA, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:21, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:22. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:22, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:21.

[0057] The term "MLH1" refers to the mutL homolog 1 gene (Ensembl: ENSG00000076242; HGNC: 7127). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI Reference Sequence NM_000249.4, which encodes the coding sequence CCDS54562, i.e., the nucleotide sequence as defined in SEQ ID NO: 23, which corresponds to the sequence of the coding sequence of the MLH1 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 24, which corresponds to the protein sequence defined in NCBI Protein Accession Reference Sequence NP_000240.1, which encodes the MLH1 polypeptide.

[0058] The term "MLH1" also refers to a nucleotide sequence that exhibits high homology to MLH1, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:23, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:24. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:24, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:23.

[0059] The term "MME" refers to the membrane metalloendopeptidase gene (Ensembl: ENSG00000196549; HGNC: 7154). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_007289.4, which encodes the coding sequence CCDS87157, i.e., the nucleotide sequence as defined in SEQ ID NO: 25, which corresponds to the sequence of the coding sequence of the MME transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 26, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_001341573.1, which encodes the MME polypeptide.

[0060] The term "MME" also refers to a nucleotide sequence that exhibits high homology to MME, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:25, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:26. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:26, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:25.

[0061] The term "MYO6" refers to the myosin VI gene (Ensembl: ENSG00000196586; HGNC: 7605). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_004999.4, which encodes the coding sequence CCDS34487, i.e., the nucleotide sequence as defined in SEQ ID NO: 27, which corresponds to the sequence of the coding sequence of the MYO6 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 28, which corresponds to the protein sequence as defined in NCBI protein accession reference sequence NP_004990.3, which encodes the MYO6 polypeptide.

[0062] The term "MYO6" also refers to a nucleotide sequence that exhibits high homology to MYO6, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:27, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:28. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:28, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:27.

[0063] The term "NAALADL2" refers to the N-acetylated alpha-linked acidic dipeptidase-like 2 gene (Ensembl: ENSG00000177694; HGNC: 23219). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_207015.3, which encodes the coding sequence CCDS46960, i.e., the nucleotide sequence as defined in SEQ ID NO: 29, which corresponds to the sequence of the coding sequence of the NAALADL2 transcript shown above, and encodes the corresponding amino acid sequence, for example, as defined in SEQ ID NO: 30, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_996898.2, which encodes the NAALADL2 polypeptide.

[0064] The term "NAALADL2" also refers to a nucleotide sequence that exhibits high homology to NAALADL2, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO: 29, or at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO: 30. This includes an amino acid sequence, or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:30, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:29.

[0065] The term "NKX3-1" refers to the NK3 homeobox 1 gene (Ensembl: ENSG00000167034; HGNC: 7838). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_006167.4, which encodes the coding sequence CCDS6042, i.e., the nucleotide sequence as defined in SEQ ID NO: 31, which corresponds to the sequence of the coding sequence of the NKX3-1 transcript shown above, and encodes the corresponding amino acid sequence, for example, as defined in SEQ ID NO: 32, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_006158.2, which encodes the NKX3-1 polypeptide.

[0066] The term "NKX3-1" also refers to a nucleotide sequence that exhibits high homology to NKX3-1, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO: 31, or at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO: 32. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:32, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:31.

[0067] The term "NQO1" refers to the NAD(P)H quinone dehydrogenase 1 gene (Ensembl: ENSG00000181019; HGNC: 2874). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_000903.3, which encodes the coding sequence CCDS67067, i.e., the nucleotide sequence as defined in SEQ ID NO: 33, which corresponds to the sequence of the coding sequence of the NQO1 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 34, which corresponds to the protein sequence as defined in NCBI protein accession reference sequence NP_001273066.1, which encodes the NQO1 polypeptide.

[0068] The term "NQO1" also refers to a nucleotide sequence that exhibits high homology to NQO1, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:33, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:34. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:34, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:33.

[0069] The term "NRP1" refers to the Neuropilin 1 gene (Ensembl: ENSG00000099250; HGNC: 8004). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI Reference Sequence NM_003873.7, which encodes the coding sequence CCDS7177, i.e., the nucleotide sequence as defined in SEQ ID NO: 35, which corresponds to the sequence of the coding sequence of the NRP1 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 36, which corresponds to the protein sequence as defined in NCBI Protein Accession Reference Sequence NP_003864.5, which encodes the NRP1 polypeptide.

[0070] The term "NRP1" also refers to a nucleotide sequence that exhibits high homology to NRP1, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:35, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:36. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:36, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:35.

[0071] The term "SLC45A3" refers to the solute carrier family 45 member 3 gene (Ensembl: ENSG00000158715; HGNC: 8642). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_033102.3, which encodes the coding sequence CCDS1458, i.e., the nucleotide sequence as defined in SEQ ID NO: 37, which corresponds to the sequence of the coding sequence of the SLC45A3 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 38, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_149093.1, which encodes the SLC45A3 polypeptide.

[0072] The term "SLC45A3" also refers to a nucleotide sequence that exhibits high homology to SLC45A3, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:37, or at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:38. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:38, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:37.

[0073] The term "SPDEF" refers to the SAM pointed domain containing ETS transcription factor gene (Ensembl: ENSG00000124664; HGNC: 17257). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_012391.3, which encodes the coding sequence CCDS4794, i.e., the nucleotide sequence as defined in SEQ ID NO: 39, which corresponds to the sequence of the coding sequence of the SPDEF transcript shown above, and encodes the corresponding amino acid sequence, for example, as defined in SEQ ID NO: 40, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_036523.1, which encodes the SPDEF polypeptide.

[0074] The term "SPDEF" also refers to a nucleotide sequence that exhibits high homology to SPDEF, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:39, or an nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:40. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:40, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:39.

[0075] The term "ATM" refers to the ATM serine / threonine kinase gene (Ensembl: ENSG00000149311; HGNC: 795). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_000051.4, which encodes the coding sequence CCDS86245, i.e., the nucleotide sequence as defined in SEQ ID NO: 41, which corresponds to the sequence of the coding sequence of the ATM transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 42, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_000042.3, which encodes the ATM polypeptide.

[0076] The term "ATM" also refers to a nucleotide sequence that exhibits high homology to ATM, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:41, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:42. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:42, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:41.

[0077] The term "ATR" refers to the ATR serine / threonine kinase gene (Ensembl: ENSG00000175054; HGNC: 882). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_001184.4, which encodes the coding sequence CCDS3124, i.e., the nucleotide sequence as defined in SEQ ID NO: 43, which corresponds to the sequence of the coding sequence of the ATR transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 44, which corresponds to the protein sequence as defined in NCBI protein accession reference sequence NP_001175.2, which encodes the ATR polypeptide.

[0078] The term "ATR" also refers to a nucleotide sequence exhibiting high homology to ATR, e.g., a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:43, or an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:44. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:44, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:43.

[0079] The term "BRCA1" refers to the DNA repair-related gene BRCA1 (Ensembl: ENSG00000012048; HGNC: 1100). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_007294.4, which encodes the coding sequence CCDS11454, i.e., the nucleotide sequence as defined in SEQ ID NO: 45, which corresponds to the sequence of the coding sequence of the BRCA1 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 46, which corresponds to the protein sequence as defined in NCBI protein accession reference sequence NP_009229.2, which encodes the BRCA1 polypeptide.

[0080] The term "BRCA1" also refers to a nucleotide sequence that exhibits high homology to BRCA1, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:45, or an nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:46. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:46, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:45.

[0081] The term "BRCA2" refers to the DNA repair-related gene BRCA2 (Ensembl: ENSG00000139618; HGNC: 1101). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_000059.4, which encodes the coding sequence CCDS9344, i.e., the nucleotide sequence as defined in SEQ ID NO: 47, which corresponds to the sequence of the coding sequence of the BRCA2 transcript shown above, and encodes the corresponding amino acid sequence, for example, as defined in SEQ ID NO: 48, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_000050.3, which encodes the BRCA2 polypeptide.

[0082] The term "BRCA2" also refers to a nucleotide sequence that exhibits high homology to BRCA2, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:47, or an nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:48. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:48, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:47.

[0083] The term "CDK12" refers to the cyclin-dependent kinase 12 gene (Ensembl: ENSG00000167258; HGNC: 24224). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_016507.4, which encodes the coding sequence CCDS11337, i.e., the nucleotide sequence as defined in SEQ ID NO: 49, which corresponds to the sequence of the coding sequence of the CDK12 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 50, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_057591.2, which encodes the CDK12 polypeptide.

[0084] The term "CDK12" also refers to a nucleotide sequence that exhibits high homology to CDK12, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:49, or an nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:50. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:50, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:49.

[0085] The term "FANCA" refers to the FA complementation group A gene (Ensembl: ENSG00000187741; HGNC: 3582). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence as defined in NCBI reference sequence NM_000135.4, which encodes the coding sequence CCDS32515, i.e., the nucleotide sequence as defined in SEQ ID NO: 51, which corresponds to the sequence of the coding sequence of the FANCA transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 52, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_000126.2, which encodes the FANCA polypeptide.

[0086] The term "FANCA" also refers to a nucleotide sequence that exhibits high homology to FANCA, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:51, or an nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:52. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:52, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:51.

[0087] The term "MRE11" refers to the MRE11 homolog, a double-stranded break repair nuclease gene (Ensembl: ENSG00000020922; HGNC: 7230). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_005591.4, which encodes the coding sequence CCDS8298, i.e., the nucleotide sequence as defined in SEQ ID NO: 53, which corresponds to the sequence of the coding sequence of the MRE11 transcript shown above, and encodes the corresponding amino acid sequence, e.g., as defined in SEQ ID NO: 54, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_005581.2, which encodes the MRE11 polypeptide.

[0088] The term "MRE11" also refers to a nucleotide sequence that exhibits high homology to MRE11, for example, a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:53, or an nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:54. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:54, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:53.

[0089] The term "PALB2" refers to the partner and localizer gene of BRCA2 (Ensembl: ENSG00000083093; HGNC: 26144). For example, an exemplary splice variant of the above gene is defined in the nucleotide sequence defined in NCBI reference sequence NM_024675.4, which encodes the coding sequence CCDS32406, i.e., the nucleotide sequence as defined in SEQ ID NO: 55, which corresponds to the sequence of the coding sequence of the PALB2 transcript shown above, and encodes the corresponding amino acid sequence, for example, as defined in SEQ ID NO: 56, which corresponds to the protein sequence defined in NCBI protein accession reference sequence NP_078951.2, which encodes the PALB2 polypeptide.

[0090] The term "PALB2" also refers to a nucleotide sequence that exhibits high homology to PALB2, for example a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:55, or an nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:56. or a nucleic acid sequence encoding an amino acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:56, or an amino acid sequence encoded by a nucleic acid sequence that is at least 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identical to the sequence as set forth in SEQ ID NO:55.

[0091] The term "biological sample" or "sample obtained from a subject" refers to any biological material obtained from a subject, e.g., a prostate cancer patient, via suitable methods known to those of skill in the art. The term "prostate cancer subject" refers to a person who has or is suspected of having prostate cancer.

[0092] The biological samples used are obtained in a clinically acceptable manner, for example in a manner that preserves nucleic acids (particularly RNA) or proteins.

[0093] The biological sample(s) include body tissues and / or body fluids, such as, but not limited to, blood, sweat, saliva, and urine. Furthermore, the biological sample contains cell extracts or cell populations containing epithelial cells, such as cancerous epithelial cells or epithelial cells from tissue suspected to be cancer. The biological sample contains cell populations from glandular tissues, for example, a sample from the prostate of a male subject. In addition, if necessary, cells are purified from the obtained body tissues and body fluids and then used as the biological sample. In some realizations, the sample is a tissue sample, a urine sample, a urine sediment sample, a blood sample, a saliva sample, a semen sample, a sample containing circulating tumor cells, extracellular vesicles, a sample containing exosomes secreted by the prostate, or a cell line or a cancer cell line. In a particular realization, a biopsy or resection sample is obtained and / or used. Such samples include cells or cell lysates.

[0094] Thus, in one embodiment, the biological sample obtained from the subject is a biopsy. In a further preferred embodiment, the method comprises the step of providing or obtaining a biopsy. In a preferred embodiment, the biopsy is a prostate biopsy.

[0095] It is also conceivable that the contents of the biological sample are subjected to an enrichment step, for example by contacting the sample with a ligand specific for a particular cell type, for example the cell membrane or organelles of prostate cells, functionalized, for example, with magnetic particles. The material enriched by the magnetic particles is then used for the detection and analysis steps described herein above or below.

[0096] Furthermore, cells, e.g., tumor cells, can also be enriched via a filtration process of a liquid or liquid sample, e.g., blood, urine, etc. Such a filtration process can also be combined with an enrichment step based on ligand-specific interactions as described herein above.

[0097] The term "prostate cancer" refers to cancer of the prostate gland in the male reproductive system, which occurs when cells in the prostate mutate and begin to grow uncontrollably. Typically, prostate cancer is associated with elevated levels of prostate-specific antigen (PSA). In one embodiment of the present invention, the term "prostate cancer" refers to cancers that exhibit a PSA level above 3.0. In another embodiment, the term refers to cancers that exhibit a PSA level above 2.0. The term "PSA level" refers to the concentration of PSA in the blood in ng / ml.

[0098] The term "non-progressive prostate cancer status" means that the individual's sample does not exhibit parameter values ​​indicative of "biochemical recurrence," "clinical recurrence," "metastasis," "castration-resistant disease," and / or "prostate cancer or disease-specific mortality."

[0099] The term "advanced prostate cancer status" means that an individual's sample exhibits parameter values ​​indicative of "biochemical recurrence," "clinical recurrence," "metastasis," "castration-resistant disease," and / or "prostate cancer or disease-specific mortality."

[0100] The term "biochemical recurrence" generally refers to the recurrence of an increase in the biological value of PSA, which indicates the presence of prostate cancer cells in the sample. However, other markers that can be used to detect the presence or that raise suspicion of such a presence can also be used.

[0101] The term "clinical recurrence" refers to the presence of clinical signs indicative of the presence of tumor cells, as measured, for example, using in vivo imaging.

[0102] The term "metastasis" refers to the presence of metastatic disease in organs other than the prostate.

[0103] The term "castration-resistant disease" refers to the presence of hormone-insensitive prostate cancer, i.e., cancer in the prostate gland no longer responds to androgen deprivation therapy (ADT).

[0104] The term "prostate cancer-specific or disease-specific mortality" refers to the death of a patient from prostate cancer in the patient.

[0105] As used herein, the term "PDE4D7 KD gene" is used interchangeably with "KD gene" or "knockdown gene" and refers to one or more genes selected from ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11, and PALB2.

[0106] As used herein, the term "response of a prostate cancer subject to radiation therapy" refers to a situation in which radiation therapy improves or does not improve the disease state of a subject with prostate cancer. As used herein, "does not improve" may refer to a situation in which radiation therapy has no significant effect or in which radiation therapy worsens the subject's condition. The worsening or improvement of a patient's condition may refer to tumor size or mass, cancer-free survival or survival time. Thus, a favorable response to radiation therapy is a reduction in tumor size or mass, an increase in tumor-free survival time, or an overall increase in survival time of the subject. Thus, an unfavorable response to radiation therapy is an increase in tumor size or mass, a decrease in tumor-free survival time, or an overall decrease in survival time of the subject.

[0107] The method is based on the expression levels of three or more target genes selected from ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2. It is understood that the method may be performed on input related to the expression levels of the three or more genes or the step of determining the expression levels may be part of the method.

[0108] It is further envisioned that the method is executed by a processor. Accordingly, in one embodiment, the present invention relates to a computer-implemented method of predicting a response of a prostate cancer subject to radiation therapy, the computer-implemented method comprising the steps of receiving a determination of the gene expression level of each of three or more genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2, wherein the gene expression levels are determined in a biological sample obtained from the prostate cancer subject; and determining a prediction of the radiation therapy response based on the gene expression levels of the three or more genes, wherein the prediction is a favorable or unfavorable response to radiation therapy, and wherein the radiation therapy is a definitive radiation therapy or a salvage radiation therapy.

[0109] Preferably, the step of determining the radiotherapy response prediction comprises combining the gene expression profiles of two or more, for example 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27 or all, of the PDE4D7 KD genes with a regression function derived from a population of prostate cancer subjects.

[0110] Thus, in one embodiment, said three or more genes include six or more, preferably nine or more, and most preferably all of said genes.

[0111] Cox proportional hazards regression allows the analysis of the impact of multiple risk factors on a tested event, such as survival, in real time. In it, risk factors are binary or discrete variables, such as risk scores or clinical stages, or continuous variables, such as biomarker measurements or gene expression values. The probability of an end point (e.g. death or disease recurrence) is called the hazard. In the regression analysis, next to information about whether the tested end point was reached or not reached, for example by subjects in a patient cohort (e.g. whether the patient died or not), the time to the end point is also taken into account. The hazard is expressed as H(t)=H 0 (t) exp(w 1 ·V 1 +w 2 ·V 2 +w 3 ·V 3 + ), where V 1 , V 2 , V 3 are the predictor variables, and H 0 (t) is the baseline hazard, while H(t) is the hazard at any time t. The hazard rate (i.e., the risk of achieving the event) is Ln[H(t) / H 0 (t)] = w 1 ·V 1 +w 2 ·V 2 +w 3 ·V 3+ , where the coefficients or weights w 1 , w 2 , w 3 ··· is estimated by Cox regression analysis and can be interpreted similarly to logistic regression analysis.

[0112] In one particular realization, the prediction of radiation therapy response is determined as follows:

number

[0113] Weight Wn Exemplary values ​​of are shown in Table 1 below. However, it should be understood that other values ​​may be used within the margin of error to determine the optimal constant based on this sample. Thus, it is envisioned that the weights are selected to be within the range of the constant listed in Table 1 below ± the standard error listed in the table. For example, the w value used listed for ETV1 is 0.1782, with a standard error of 0.2464, which means that any value between -0.0682 (0.1782-0.2464) and 0.4246 (0.1782+0.2464) is used. The same is true for the other 27 genes listed below.

[0114] [Table 1]

[0115] The radiotherapy response prediction may also be classified or categorized into one of at least two risk groups based on the value of the radiotherapy response prediction. For example, there are two risk groups, three risk groups, four risk groups, or more than four predefined risk groups. Each risk group covers a distinct range (non-overlapping) of values ​​of the radiotherapy response prediction. For example, the risk groups represent the probability of occurrence of a particular clinical event, such as 0 to <0.1, 0.1 to <0.25, 0.25 to <0.5, or 0.5 to 1.0.

[0116] It is further preferred that the step of determining said radiotherapy response prediction is further based on one or more clinical parameters obtained from said prostate cancer subject.

[0117] As mentioned above, various measurements based on clinical parameters have been investigated, and the prediction of radiotherapy response can be further improved by further basing the prediction on such clinical parameter(s).

[0118] In one embodiment, determining the prediction of radiation therapy response comprises combining gene expression levels of three or more, for example 3, 4, 5, 6, 7, 8, 9 or all of the genes with a regression function derived from a population of prostate cancer subjects.

[0119] In one embodiment, determining the prediction of radiotherapy response is further based on one or more clinical parameters obtained from the prostate cancer subject.

[0120] In one embodiment, the clinical parameters include one or more of: (i) prostate-specific antigen (PSA) level; (ii) pathological Gleason score (pGS); iii) clinical stage of the tumor; iv) pathological Gleason grade group (pGGG); v) pathological stage; vi) one or more pathological variables, such as surgical margin status, lymph node involvement, extraprostatic growth and / or seminal vesicle invasion; vii) CAPRA-S; and viii) another clinical risk score.

[0121] In one embodiment, determining the prediction of radiation therapy response comprises combining the gene expression levels of the three or more KD genes and one or more clinical parameters obtained from the prostate cancer subject with a regression function derived from a population of prostate cancer subjects.

[0122] It is further preferred to combine said gene expression profile of one or more PDE4D7 KD genes and one or more clinical parameters obtained from said prostate cancer subjects with a regression function derived from a population of prostate cancer subjects.

[0123] In one particular realization, the prediction of radiotherapy response is determined as follows (PDE4D7_clinical_model):

number

[0124] In further realization, the PDE4D7_clinical model can be further used for discrimination to identify more risk classes. In the figures using the PDE4D7_clinical score, the threshold value for dividing patients into two groups is shown in the figure legend. In the case of the PDE4D7_clinical_class model, patients are divided into three risk groups (low, medium, high) instead of just two groups. The cutoff value of each class is based on the PDE4D7_clinical score, which is calculated based on the Cox regression model: low risk (<0); medium risk (0-4); high risk (>4).

[0125] In one embodiment, the biological sample is obtained from the prostate cancer subject prior to the initiation of radiation therapy, preferably the biological sample is a prostate sample or a prostate cancer sample.

[0126] In one embodiment, the radiation therapy is definitive or salvage radiation therapy.

[0127] In one embodiment, a treatment is recommended or administered based on said prediction of radiation therapy response; If the radiotherapy response prediction is unfavorable, the recommended treatment is: (i) Radiation therapy provided earlier than standard; (ii) radiotherapy with increased radiation doses; (iii) adjunctive treatment, such as androgen deprivation therapy, preferably a combination of androgen deprivation therapy with a second-line antiandrogen, radiation, chemotherapy and / or immunotherapy; and (iv) Alternative non-radiotherapy treatments Contains one or more of the following: Depending on how negative the radiotherapy response prediction is, determines how much the recommended treatment deviates from standard forms of radiotherapy.

[0128] In one embodiment, a treatment is recommended based on said prediction of radiation therapy response; If the radiotherapy response prediction is favorable, the recommended treatment is: (v) definitive radiotherapy; (vi) salvage radiotherapy; (vi) salvage radiation therapy at tapered dose levels; and (vii) Surveillance standby Contains one or more of the following:

[0129] In a second aspect, the present invention provides a computer program comprising a plurality of instructions, the plurality of instructions, when run by a computer, comprising: receiving data indicative of a gene expression profile for each of three or more, such as 3, 4, 5, 6, 7, 8, 9 or all, genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11, and PALB2, wherein gene expression levels are determined in a biological sample obtained from the prostate cancer subject; determining a prediction of a prostate cancer subject's response to a treatment based on said gene expression profile(s) of said three or more genes, said prediction being a favorable or unfavorable response to radiation therapy, said radiation therapy being a definitive or salvage radiation therapy. In one embodiment, the computer program is used to perform the method of the first aspect of the invention.

[0130]

[0023] In a further aspect of the present invention there is provided an apparatus for predicting a prostate cancer subject's response to radiation therapy, comprising: an input adapted to receive data indicative of a gene expression profile for each of three or more, such as 3, 4, 5, 6, 7, 8, 9 or all, genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11, and PALB2, wherein the gene expression profile is determined in a biological sample obtained from the prostate cancer subject; and a processor adapted to determine a prediction of radiation therapy response based on said gene expression profile(s) of said three or more genes; and optionally a providing unit adapted to provide the radiotherapy response prediction or a treatment recommendation based on the radiotherapy response prediction to a healthcare caregiver or to the prostate cancer subject. An apparatus is presented comprising:

[0131] In a third aspect, the present invention provides a method for producing a pharmaceutical composition comprising the steps of: at least three sets of polymerase chain reaction primer pairs, and optionally at least three probes, for determining the expression level in a sample of three or more, such as 3, 4, 5, 6, 7, 8, 9 or all, of the genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2, Optionally, the diagnostic kit further comprises a computer program or device as described herein. The diagnostic kit may further comprise any one of a buffer, dNTPs, a magnesium salt solution, or a polymerase enzyme.

[0132] In a fourth aspect, the present invention relates to the use of a diagnostic kit according to the third aspect of the invention in a method for predicting the response of a prostate cancer subject to radiation therapy, preferably for use in a method according to the first aspect of the invention.

[0133] In a fifth aspect, the present invention provides a method for producing a pharmaceutical composition comprising the steps of: receiving a biological sample obtained from a subject with prostate cancer; using the diagnostic kit according to the third aspect of the invention to determine a gene expression profile for each of three or more, such as 3, 4, 5, 6, 7, 8, 9 or all, genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2 in said biological sample obtained from said prostate cancer subject; The present invention relates to a method comprising the steps of:

[0134] In a fifth aspect, the present invention relates to the use of a respective gene expression profile of three or more, such as 3, 4, 5, 6, 7, 8, 9 or all, genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2 in a method of predicting a prostate cancer subject's response to radiation therapy, comprising the steps of: determining a prediction of radiation therapy response based on the gene expression levels of the three or more genes, said prediction being a favorable or unfavorable response to radiation therapy, said radiation therapy being a definitive radiation therapy or a salvage radiation therapy; The present invention relates to use in a method comprising the steps of: Working Example EXAMPLES

[0135] LNCaP cells (wild type; wt) were cultured and transfected using various lentiviral vectors carrying short hairpin RNA (shRNA) constructs targeting the mRNA of PDE4D7. Several clones stably expressing the shRNA-PDE4D7 constructs were selected. Knockdown of PDE4D7 expression was confirmed using qPCR. RNA sequencing was then performed on LNCaP wt cells and three different shRNA-PDE4D7 expressing LNCaP clones. Quality control was performed on the sequencing reads using FastQC / MultiQC. A trimming step was then performed using SeqPurge to trim adapter sequences from the sequencing reads. rRNA sequences were then removed using Bowtie 2. The processed sequencing reads were then aligned using STAR, and the reads mapped to genes were quantified using featureCounts. From the resulting data, the number of transcripts per million was calculated for each gene, these values ​​were log2 transformed and normalized based on the reference genes, and finally, a z-score transformation was performed.

[0136] Differential gene expression between wild-type LNCaP and the three different knockdown clones was determined. Initially, 113 differentially expressed genes were identified with a confidence level of p<1E-20. For these genes, pathway analysis (DisGenNET) for enriched genes was performed using the terms prostate neoplasia, prostate malignancy, prostate carcinoma, neoplasia metastasis and tumor progression. Based on these criteria, 20 prostate neoplasia / metastasis-related genes and 8 DNA repair-related genes were selected, resulting in a group of 28 genes: ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAALADL2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2. A model was constructed using multivariate Cox regression analysis to give a PDE4D7 KD score (also referred to herein as "KD score" or knockdown score). A first cohort of 653 patient samples was obtained for downstream analysis after samples were used for quality control, and samples belonging to patients who received salvage radiotherapy with or without androgen deprivation treatment were selected from this first cohort of samples for training the model.The trained model was then validated on a second cohort of patients, in which patient samples from patients who received salvage radiotherapy with or without androgen deprivation treatment were used.These data are shown in the following figure: Training in patient cohorts: Figures 1 to 14 Figures 1 to 4: Overall survival endpoint Figures 5 to 8: Metastasis-free survival Figures 9-12: Prostate cancer-specific survival endpoints Figure 13, Figure 14: RT dose (low vs. high risk disease); endpoint overall survival Testing in a cohort of 151 patients: Figures 17 to 26 Figures 15 to 18: Overall survival endpoint Figures 19 to 22: Metastasis-free survival Figures 23-26: Prostate cancer-specific survival endpoint.

[0137] [Table 2] TIFF2024528065000006.tif25585TIFF2024528065000007.tif25583TIFF2024528065000008.tif12983

[0138] [Table 3] TIFF2024528065000010.tif234170

[0139] TPM gene expression values For each gene, a TPM (transcripts per million) expression value was calculated based on the following steps: 1. After mapping and aligning to the human genome, divide the number of reads from the raw RNAseq fastq data by the length of each gene in kilobases. This gives the RPK (reads per kilobase). 2. Sum up all the RPK values ​​in your sample and divide this number by 1,000,000. This gives you a "per million" scaling factor. 3. Divide the RPK value by the "per million" scaling factor. This gives the TPM expression value for each gene.

[0140] The TPM values ​​per gene were used to calculate the reference normalized gene expression for each gene.

[0141] Normalized gene expression values For Cox regression modeling, TPM (transcripts per million)-based expression values ​​from RNAseq data for all genes were log2-normalized by the following transformation:

number

number

[0142] The following reference genes were considered (Table 2):

[0143] [Table 4]

[0144] For these reference genes, the following four were selected: B2M, HPRT1, POLR2A, and PUM1, to calculate:

number

[0145] For multivariate analysis of genes of interest, the normalized log2(TPM) values ​​of the reference genes for each gene were used as input.

[0146] Cox regression analysis We then set out to test whether a combination of these 28 genes might offer more prognostic value. Using Cox regression, we modeled the expression levels of the 28 genes against prostate cancer-specific mortality after salvage RT after surgery with (PDE4D7_clinical_model) or without (PDE4D7_KD_model) the presence of the pathological Gleason grade group (pGGG) variable in a cohort of 571 prostate cancer patients. The two models were tested in ROC curve analysis (data not shown) as well as in Kaplan-Meier survival analysis.

[0147] The Cox regression function was derived as follows: PDE4D7_KD_model:

number

number

[0148] In the PDE4D7_clinical_class model, patients are divided into three risk groups (low, medium, high) instead of just two groups. The cutoff value for each class is based on the PDE4D7_clinical score, which is calculated based on the Cox regression model: low risk (<0); medium risk (0-4); high risk (>4).

[0149] Weight w 1 ~w 28 Details about the above are shown in Table 1 below.

[0150] ROC curve analysis The Cox regression model outlined above was then tested for its ability to predict prostate cancer-specific mortality (PCa Death) 5 years after initiation of salvage radiotherapy (SRT) due to disease recurrence after surgery. Model performance was compared with EAU-BCR risk groups (see Tilki D. et al., "External validation of the European Association of Urology Biochemical Recurrence Risk groups to predict metastasis and mortality after radical prostatectomy in a European cohort," Eur Urol, Vol. 75, No. 6, pp. 896-900, 2019) and pathological Gleason grade group (pGGG).

[0151] Kaplan-Meier survival analysis For Kaplan-Meier survival curve analysis, the Cox functions of the risk models (PDE4D7_KD_model and PDE4D7_clinical_model) were categorized into two subcohorts based on cutoff values. The thresholds for separating low- and high-risk groups were based on the mean output values ​​of PDE4D7_KD_model and PDE4D7_clinical_model, respectively, calculated by using the Cox regression model for each patient in the entire cohort.

[0152] In the two risk models developed (PDE4D7_KD_model; PDE4D7_clinical_model), the patient classes show an increased risk of experiencing the tested clinical endpoints: prostate cancer-specific death (PCa Death) after initiation of salvage radiotherapy (SRT) due to disease recurrence after surgery (Figures 9 to 12, Figures 23 to 26), death after initiation of salvage radiotherapy (SRT) due to disease recurrence after surgery (Figures 1 to 4, Figures 15 to 18), or development of metastasis or recurrence due to disease recurrence after surgery (Metastasis) (Figures 5 to 8, Figures 19 to 22). EXAMPLES

[0153] LNCaP clone FGC (ATCC® CRL1740™) cells were cultured in RPMI 1640 medium supplemented with fetal bovine serum to a final concentration of 10%. For lentiviral transduction, cells were cultured at 0.2 × 10 in 2 ml of growth medium per well. 6Cells were seeded in 6-well plates at a concentration of 1:10 ...

[0154] RNA was extracted from cells using the RNeazy kit (Qiagen). cDNA was synthesized using either oligo-dT or specific primers. qPCR was performed using PrimeTime Gene Expression Master Mix (IDT, Cat. 1055772).

[0155] Several PDE4D7-shRNA constructs were tested and evaluated. The construct represented by SEQ ID NO: 57 having the sequence: gatccgGAAATACCTGTGATTTGCTTTCTCAAGAGGAAAGCAAATCACAGGTATTTCttttttg was used in all experiments. EXAMPLES

[0156] Based on the PDE4D7_KD model using the expression levels of all 28 genes listed in Table 3, we reasoned that a subset of these genes is also likely to have predictive value. To further support this theory, three or four genes were randomly selected from the entire set of genes, as shown in Tables 5 to 12 below. Each of these models was used to analyze the data, and patient groups were analyzed as shown in Figures 30 to 37. As can be inferred from the log-rank values ​​of each plot, each of the three-gene model and the four-gene model was found to significantly stratify high and low risk patients. Therefore, it appears that any subset of three or more genes selected from the PDE4D7 KD genes (shown in Table 3) can be used to stratify patients into high and low risk groups.

[0157] Models used: [Table 5]

[0158] [Table 6]

[0159] [Table 7]

[0160] [Table 8]

[0161] [Table 9]

[0162] [Table 10]

[0163] [Table 11]

[0164] [Table 12]

[0165] Consideration The efficacy of both definitive RT and SRT for localized prostate cancer is limited, especially for patients at high risk of recurrence, leading to disease progression and ultimately patient death. Prediction of treatment outcome is very complicated, as many factors contribute to treatment efficacy and disease recurrence. Important factors may not yet be identified, and the influence of other factors cannot be precisely determined. Several clinicopathological measurements are currently being investigated and applied in clinical practice to improve response prediction and treatment selection, and have brought about some improvements. Nevertheless, better prediction of response to definitive RT and SRT is still strongly needed to increase the success rate of these treatments.

[0166] Molecules have been identified whose expression has been shown to be significantly associated with mortality after definitive RT and SRT and therefore is expected to improve the prediction of efficacy of these treatments. Improved prediction of efficacy of RT for each patient will lead to improved treatment selection and improved chances of survival, whether in the definitive or salvage setting. This can be achieved by 1) optimizing RT for patients in whom RT is predicted to be effective (e.g., by dose escalation or by changing the time of initiation) and 2) directing patients in whom RT is predicted to be ineffective to alternative, potentially more effective, forms of treatment. This, in turn, reduces patient suffering by sparing ineffective treatments and reduces costs that would otherwise be spent on ineffective treatments.

[0167] Those skilled in the art can understand and effect other variations to the disclosed implementations in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

[0168] In the claims, the words "comprise" and "comprise" do not exclude other elements or steps and the singular does not exclude a plurality.

[0169] One or more steps of the method illustrated in Fig. 1 are implemented in a computer program product running on a computer. The computer program product includes a non-transitory computer readable recording medium, such as a disk, a hard drive, etc., on which a control program is recorded (stored). Common forms of non-transitory computer readable media include, for example, a floppy disk, a flexible disk, a hard disk, a magnetic tape or other magnetic storage medium, a CD-ROM, a DVD or other optical medium, a RAM, a PROM, an EPROM, a FLASH-EPROM or other memory chip or cartridge, or other non-transitory medium that can be read and used by a computer.

[0170] Alternatively, one or more steps of the method may be implemented in a transitory medium such as a transmittable carrier wave in which the control program is embodied as a data signal using a transmission medium such as sound or light waves, such as those generated during radio wave and infrared data communications.

[0171] Exemplary methods are implemented on one or more general purpose computers, special purpose computers, programmed microprocessors or microcontrollers and peripheral integrated circuit elements, hardwired electronic or logic circuits such as ASICs or other integrated circuits, digital signal processors, discrete element circuits, programmable logic devices such as PLDs, PLAs, FPGAs, Graphical cards CPUs (GPUs) or PALs, etc. In general, any device capable of implementing a finite machine capable of sequentially performing the steps described herein can be used to implement one or more steps of the exemplary risk stratification method for treatment selection in patients with prostate cancer. As will be appreciated, the steps of the method may all be computer-implemented, although in some embodiments, one or more steps are performed at least partially manually.

[0172] A computer-implemented process may also be created by loading computer program instructions on a computer, other programmable data processing apparatus, or other device to cause a series of work steps to be executed on the computer, other programmable apparatus, or other device such that the instructions running on the computer or other programmable apparatus provide a process for performing the functions / operations specified herein.

[0173] Any reference signs in the claims should not be construed as limiting the scope. The attached sequence listing, entitled 2021PF00367_seq list_ST25, is hereby incorporated by reference in its entirety.

Claims

**Claim 1** A method for predicting the response of a prostate cancer subject to radiotherapy, the method comprising: determining the gene expression level of each of three or more genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAAALD2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11 and PALB2, or receiving the result of determining the gene expression level, wherein the gene expression level is determined in a biological sample obtained from the prostate cancer subject; determining a prediction of radiotherapy response based on the gene expression levels of the three or more genes, the prediction being a favorable response or an unfavorable response to radiotherapy, and the radiotherapy being salvage radiotherapy; A method having the above steps. **Claim 2** The method according to claim 1, wherein the three or more genes include six or more, preferably nine or more, and most preferably all of the genes. **Claim 3** The method according to claim 1 or 2, wherein the step of determining the prediction of radiotherapy response includes combining the gene expression levels of three or more, such as 3, 4, 5, 6, 7, 8, 9 or all of the genes with a regression function derived from a population of prostate cancer subjects. **Claim 4** The method according to any one of claims 1 to 3, wherein the step of determining the prediction of radiotherapy response is further based on one or more clinical parameters obtained from the prostate cancer subject. **Claim 5** The method according to claim 4, wherein the clinical parameters include (i) prostate-specific antigen (PSA) level, (ii) pathological Gleason score (pGS), (iii) clinical stage of the tumor, (iv) pathological Gleason grade group (pGGG), (v) pathological stage, (vi) one or more pathological variables, such as margin status, lymph node invasion, extraprostatic growth and / or seminal vesicle invasion, (vii) CAPRA-S, and (viii) one or more of another clinical risk score. **Claim 6** The step of determining the prediction of the radiation therapy response includes combining the gene expression levels of the three or more genes, and one or more clinical parameters obtained from the prostate cancer subject, with a regression function derived from a population of prostate cancer subjects, the method according to claim 4 or 5.

7. The biological sample is obtained from the prostate cancer subject before the start of the radiation therapy, and preferably, the biological sample is a prostate sample or a prostate cancer sample, the method according to any one of claims 1 to 6.

8. A treatment method is recommended based on the prediction of the radiation therapy response, When the prediction of the radiation therapy response is not favorable, the recommended treatment method is (i) Radiation therapy provided earlier than the standard, (ii) Radiation therapy with an increased radiation dose, (iii) Adjuvant therapy such as androgen blockade therapy, preferably, a combination of androgen blockade therapy and a second-choice anti-androgen agent, radiation therapy, chemotherapy, and / or immunotherapy, adjuvant therapy, and (iv) An alternative treatment other than radiation therapy The method according to any one of claims 1 to 7, including one or more of.

9. A treatment method is recommended based on the prediction of the radiation therapy response, When the prediction of the radiation therapy response is favorable, the recommended treatment method is (v) Salvage radiation therapy, (vi) Salvage radiation therapy at a tapered dose level, and (vii) Surveillance waiting The method according to any one of claims 1 to 8, including one or more of.

10. A computer program including a plurality of instructions, when the computer program is run by a computer, the plurality of instructions Receiving data indicating the gene expression profiles of three or more genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAAALD2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11, and PALB2, for example, 3, 4, 5, 6, 7, 8, 9 genes or all genes, the step being determined in a biological sample obtained from a prostate cancer subject, the step, Determining a prediction of the response of a prostate cancer subject to treatment based on the gene expression profiles of said three or more genes, wherein said prediction is a favorable response or an unfavorable response to radiation therapy, and wherein said radiation therapy is salvage radiation therapy; A computer program that causes the computer to execute a method having the steps above. Use of a kit in a method for predicting the response of a prostate cancer subject to radiation therapy, wherein said radiation therapy is salvage radiation therapy, and wherein said kit comprises: At least three sets of polymerase chain reaction primer pairs for determining the respective expression levels in a sample of three or more genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAAALD2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11, and PALB2, for example 3, 4, 5, 6, 7, 8, 9, or all of the genes, and optionally at least three probes Use of a kit comprising the above. Use of the kit according to claim 11, by the method according to any one of claims 1 to 8.

13. Use in a method for predicting the response of a prostate cancer subject to radiation therapy of the respective gene expression profiles of three or more genes selected from the group consisting of ACPP, AR, CDH1, EHF, ETV1, FOLH1, FOXA1, HOXB13, KLK2, KLK3, MAOA, MLH1, MME, MYO6, NAAALD2, NKX3-1, NQO1, NRP1, SLC45A3, SPDEF, ATM, ATR, BRCA1, BRCA2, CDK12, FANCA, MRE11, and PALB2, for example 3, 4, 5, 6, 7, 8, 9, or all of the genes, wherein said method comprises determining a prediction of radiation therapy response based on the gene expression levels of said three or more genes, wherein said prediction is a favorable response or an unfavorable response to radiation therapy, and wherein said radiation therapy is salvage radiation therapy. Use of a gene expression profile.

Images