WO2022003146A1

WO2022003146A1 - Methods for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment

Info

Publication number: WO2022003146A1
Application number: PCT/EP2021/068290
Authority: WO
Inventors: Oriol CASANOVAS CASANOVAS; Jorge SENSERRICH; Luis PALOMERO; Lidia MOSERLE; Roser PONS; María de Mar Martínez Lozano; Gabriela JIMÉNEZ VALERIO
Original assignee: Fundació Institut D'investigació Biomèdica De Bellvitge (Idibell); Institut Català D'oncologia (Ico)
Priority date: 2020-07-03
Filing date: 2021-07-02
Publication date: 2022-01-06

Abstract

The present invention relates to methods useful for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer, or for predicting the response to an antiangiogenic therapy, wherein the methods comprise the determination of the expression level of ALDH1A3, as well as methods for providing personalized medicine to said patients. These methods can be used to identify those patients who are at a high risk of local invasion and/or metastasis. The identification of this subgroup of patients may guide the selection of therapies, improving financial and health outcomes. The invention relates as well to the use of reagents for carrying out the predictive methods.

Description

METHODS FOR PREDICTING THE RISK OF LOCAL INVASION AND/OR METASTASIS INDUCED BY AN ANTIANGIOGENIC TREATMENT

FIELD OF THE INVENTION

The invention relates to the field of prognosis and, more in particular, to methods for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient suffering from cancer, as well as methods for providing personalized medicine to said patients. The invention relates as well to the use of reagents for carrying out the predictive methods.

BACKGROUND ART

Antiangiogenic therapies have been typically identified as promising therapies for treatment of several types of cancers sustained by angiogenesis to progress. Since their introduction in the treatment of oncologic diseases, several patients have benefited from these drugs. Nevertheless, they have not been the expected panacea. Clinical studies indeed have shown that antiangiogenic treatments exert promising results in terms of prolonged progression-free survival (PFS) but eventually the efficacy decreases thus resulting in less clear effects on overall survival (OS). Several studies in mouse models have tried to explain the discrepancy between promising short-term effects and the long term limitations of antiangiogenic therapies and allowed finding out some possible mechanisms involved in failure. Interestingly, several preclinical studies have demonstrated that antiangiogenic treatments are able to induce or exacerbate the invasive and metastatic behavior of different tumor models. Among human tumors, renal cell carcinoma (RCC) is commonly treated with antiangiogenics but gain in patient survival is controversial. Furthermore, when evaluated, the impact on metastatic progression is unclear. Doubts also exist in respect of the safety of antiangiogenics administered as neoadjuvant therapy. Renal tumors grow surrounded by a fibrous peritumoral capsule and capsular invasion in localized RCC has been suggested as a prognostic factor of recurrence and of cancer-related death. Intriguingly, a study in a small subset of RCC patients treated with the antiangiogenic tyrosine kinase inhibitor axitinib as neoadjuvant therapy has described a fibrous reaction in the tumor-normal kidney interface and an invasive proliferative pattern in most treated tumors. However, there is still a need for biomarkers with prognostic and predictive potential in order to define when antiangiogenic therapies can increase tumor invasiveness and unexpectedly modify the progression of neoplasia toward more aggressive behavior.

SUMMARY OF THE INVENTION

The present invention relates to the finding that the expression level of the ALDH1A3 gene is a useful marker in the prognosis of cancer patients, particularly of patients who suffer from clear cell renal cell carcinoma (ccRCC). This method can be used to determine those patients who are at a high risk of developing local invasion and/or metastasis induced by an antiangiogenic treatment. The method also allows for the identification of patients that can exhibit a better response to anti-angiogenic agents, improving financial and health outcomes

Thus, in a first aspect, the invention relates to a method for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer, wherein the method comprises

(i) determining the expression level of ALDH1A3 in a sample from said patient, and

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1 A3 that is increased with respect to the reference value is indicative of an increased risk of local invasion and/or metastasis in response to the antiangiogenic treatment.

In a different aspect, the invention relates to a method for predicting the response to an antiangiogenic therapy of a patient who suffers from cancer, wherein the method comprises

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1 A3 that is increased with respect to the reference value is indicative that the patient is at an increased risk of developing local invasion and/or metastasis in response to the antiangiogenic treatment. In another aspect, the invention relates to a method for selecting a patient suffering from cancer for treatment with an antiangiogenic therapy, wherein the method comprises

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1A3 that is decreased with respect to the reference value is indicative that the patient is selected for treatment with an antiangiogenic therapy.

In yet a different aspect, the invention relates to a method for selecting a therapy for a patient suffering from cancer, wherein the method comprises

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1A3 that is decreased with respect to the reference value is indicative that an antiangiogenic therapy is selected for the patient.

In another aspect, the invention relates to an antiangiogenic compound for use in the treatment of a patient suffering from cancer, wherein the patient is selected using a method that comprises

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1A3 that is decreased with respect to the reference value is indicative that the patient will benefit from treatment with an antiangiogenic therapy.

In an additional aspect, the invention relates to the use of ALDH1A3 for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer.

In a final aspect, the invention relates to the use of reagents specific for the determination of the expression level of ALDH1A3, for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer. BRIEF DESCRIPTION OF THE FIGURES

Figure 1. Establishment and histological characterization of Ren-PDOX.

A) Representation of protocol for the establishment of Ren-PDOXs. Ren-PDOXs were obtained by implantation of a total of 56 ccRCC human specimens on left mouse kidney. Representative images of Ren-PDOX tumor and lung metastases (mets, arrows) are shown. Total number of established and metastatic Ren- PDOXs is reported. Graph describes the proportion of metastatic Ren-PDOXs generated by tumor from patients with negative (pNO and/or pMO) or positive (pN1 and/or pM1 and/or metachronous) metastatic status. *, one Ren-PDOX has not been evaluated for lung metastasis.

B) Overall mass architecture of paired human specimens and Ren-PDOX tumors observed by HE staining (20X). TNM staging and Fuhrman grade (F) of implanted tumors are reported. For Ren-PDOXs, lung micrometastases in HE and vimentin stained sections are shown (40X).

C) Human to mouse stroma substitution. CD34/CD31 staining for respectively human (h) and mouse (m) blood vessels in original human specimens and paired Ren13 and 28-PDOX are shown (20X). Ren-PDOX OT referred to tumor generated from first implantation of human specimen in mice. Inserts show human and mouse kidney as positive/negative controls of staining (20X).

D) Frequency histogram of SNV and/or indels found in Ren-PDOX by high- throughput sequencing in Ren13, Ren28, Ren36, Ren86 (targeted-NGS analysis by Haloplex) and Ren50-PDOX (WES). Only non-synonymous/stop gain SNVs and frameshift indels novel or with MAF<0.01 are included.

E) Genetic comparison of original human specimen (patient #50) and its derived PDOX (Ren50-PDOX) by WES. Venn diagram shows number and percentage of variants called in human specimen (left), in its PDOX (right) and overlapped variants (center). Figure 2. Robust antitumor effects of anti-VEGF/R therapies in Ren-PDOX.

A-E) Effects of DC and Beva treatments on Ren 13, Ren86, Ren28 and Ren50- PDOX. Evaluation of tumor progression (A) in control (black) compared to DC (dark gray, *) and Beva (light gray, #) treatments from 5-11 animals per tumor and treatment group by Mann Whitney test. Quantification of tumor weight (B), vessel number (C) and tumor necrosis (D) in control and treated tumors from 4- 5 samples/tumor/treatment group by Mann Whitney test. Evaluation of overall survival (E) comparing control and treated mice from 4-13 mice/tumor/treatment group by Mantel-Cox test.

Figure 3. Effects of anti-VEGF/R therapies on local invasion.

A) Capsular invasion was defined as i) absent if tumor cells were separated from normal renal parenchyma by well-defined layers of FN or ii) present if strands of tumor cells were directly in contact with normal renal parenchyma

B-C) Effects of DC and Beva treatment on Ren 13, Ren86, Ren28 and Ren50-PDOX on capsular invasion (B) and tumor front of invasion (C). B) Incidence of Capsular invasion (stripes) in each model and treatment. 5-8 animals per model and treatment group were analyzed by Chi-square test. C) Representative images of Vimentin staining (4X) and quantification of fold-invasion (D invasion) of treated vs control tumors. 6-12 animals per model and treatment group were analyzed by Mann-Whitney test.

Figure 4. Effects of anti-VEGF/R therapies on metastasis.

A-C) Evaluation of lung micrometastases using HE and vimentin-stained lung sections (A) upon antiangiogenic therapies in Ren13BM and Ren28. Incidence (B) (pie chart, *p<0.05 by Chi-square) and proportion of mice with 0, 1, 2 or > 3 metastases (mets) per lungs (C) (*p<0.05 by Mann-Whitney test) are shown.

D) Anti-VEGF/R treatments exacerbate the invasive capacity quantified as capsular invasion in the tumor kidney interface in some Ren-PDOXs. Moreover, the evaluation of lung metastasis confirms the pro-malignant effects of antiangiogenics in some Ren-PDOXs, recapitulating the inter-patient variable response.

Figure 5. Segregation of tumor and stromal expression profile to define tumor- specific molecular hallmarks of pro-invasive behavior.

A) Comparison of tumor-derived (human, left) and stroma-derived (mouse, right) differentially expressed genes in a pro-invasive (Ren13) and non-pro-invasive (Ren50) Ren-PDOX tumor models, obtained from species-specific RNAseq analysis. Venn diagrams illustrate gene expression differentials and overlap between Ren13 and Ren50 tumor types. B) Tumor specific (human) enriched gene sets in Ren13 and Ren50 Ren-PDOX tumors, after GSEA analysis from RNAseq data. Enrichment plots of top Ren13- and Ren50-enriched pathways are depicted (top, in dark grey and light grey boxes respectively). Gene sets most significantly enriched in Ren13 or Ren50 are indicated in the table (bottom left), including their normalized enrichment score (NES), p-value and false discovery rate (FDR q-value). Heatmaps illustrate expression levels of the genes responsible for the enrichment of KRAS signaling up and down in Ren13 and Ren50, respectively (bottom right).

C) Venn diagram comparing differentially expressed genes in the KRAS signature among Stage (l/llvslll/IV), Tumor Size (T1/2vsT3/4) and Metastasis (M0vsM1) from the TCGA KIRC primary tumors cohort. Only significant (FDR-adjusted p- value <0.05) age- and gender-corrected correlated genes are included. The fold change for the overlapping 19 genes is listed separating upregulated (dark grey) and downregulated (light grey) in more advanced/aggressive tumors.

Figure 6. Patient validation of tumor-specific biomarkers of pro-invasive behavior.

A) Patient characteristics of this unique series of 39 RCC patients (GSE29609) where microvascular invasion (MVI) and renal vein involvement (RV) are specifically annotated. B) Validation of candidate gene expression is shown for each variable and condition, *p<0.05 **p<0.01 by Mann-Whitney test.

C) Levels of MAP7 and ALDH1A3 could differentiate between non-pro-invasive and pro-invasive tumors. Low levels of MAP7 and high levels of ALDH1A3 are associated to more invasive tumors.

Figure 7. Validation of ALDH1A3 as marker in patients treated with antiangiogenics

A) Immunohistochemistry staining of ALDH1A3 in Ren13, Ren86, Ren28 and Ren50 RCC PDOX tumors.

B) Quantification of ALDH1A3 RNA expression levels by TaqMan in non-pro- invasive and pro-invasive RCC PDOX tumors.

C) Quantification of ALDH1A3 expression levels by immunohistochemistry in non- pro-invasive and pro-invasive RCC PDOX tumors.

D) Quantification of ALDH1A3 expression levels by immunohistochemistry in non- pro-invasive and pro-invasive tumor samples from RCC patients. Score: 0 = low expression, 1 = medium expression, 2 = high expression. E) Receiver operating characteristic (ROC) curve for ALDH1A3 for the sensitivity and specificity defines pro-invasive prediction in RCC tumors. Area under the curve (AUC) for ALDH1A3 is 0.80 with p=0.05.

F) Significant correlation between ALDH1A3 expression levels and pro-invasive signature in RCC tumors. Statistical analysis using Spearman’s t-test p = 0.0095.

Figure 8. Validation of MAP7 as marker in patients treated with antiangiogenics.

A) Immunohistochemistry staining of MAP7 in Ren13, Ren86, Ren28 and Ren50 RCC PDOX tumors.

B) Quantification of MAP7 RNA expression levels by TaqMan in non-pro-invasive and pro-invasive RCC PDOX tumors.

C) Quantification of MAP7 expression levels by immunohistochemistry in non-pro- invasive and pro-invasive RCC PDOX tumors.

D) Quantification of MAP7 expression levels by immunohistochemistry in non-pro- invasive and pro-invasive tumor samples from RCC patients.

E) Receiver operating characteristic (ROC) curve for MAP7 for the sensitivity and specificity does not determine pro-invasive prediction of MAP7 in RCC tumors. Area under the curve (AUC) for MAP7 is 0.58 with p=0.7.

F) No correlation between MAP7 expression levels and pro-invasive signature in RCC tumors. (Spearman’s t-test p = 0.94).

G) No correlation between ALDH1A3 expression levels and MAP7 expression in RCC tumors. (Spearman’s t-test p = 0.22).

Figure 9. Effects of sunitinib treatment on local invasion. Effects of sunitinib treatment on Ren13BM and Ren28 on capsular invasion and tumor front invasion. Bar graphs show incidence of capsular invasion (stripes) in control and sunitinib treated tumors. 3-9 animals per model and treatment group were analyzed by Chi-square test. Box plots represent fold-invasion (D invasion) of sunitinib-treated vs control tumors. 5-8 animals per model and treatment group were analyzed by Mann- Whitney test

Figure 10: Capsular invasion and invasive front of Ren-PDOXs..

Effects of sunitinib treatment on an independent series of five new Ren-PDOXs showing variable therapy-induced effects on capsular invasion (top, bar graphs) and effects on tumor front invasion (bottom, box plots). 5-10 animals per tumor were used. Box plots represent median, Q1/Q3 and max/min value whiskers analyzed by Mann-Whitney test *P < 0.05.

Figure 11. Association of pre-treatment ALDH1A3 protein levels by IHC on a series of Ren-PDOX models treated with sunitinib. (A) Representation of ALDH1A3 expression on pro-invasive and non-pro-invasive Ren-PDOX after sunitinib treatment (**P < 0.01 by Chi-Square test). (B) Correlation of pre-treatment ALDH1A3 protein levels by IHC with increased tumor invasion found after sunitinib treatment n = 7, Spearman’s non- parametric correlation, P = 0.0095.

Figure 12. (A) Patient characteristics of a series of 15 metastatic ccRCC patients specifically annotated for their type of progression after antiangiogenic therapy (Sunitinib, Pazopanib and Bevacizumab). 60% progressed in preexisting lesions (non- pro-aggressive) while 40% progressed with new lesions (pro-aggressive). (B) Representation of ALDH1A3 score (0 = absent, 1 = low, 2 = high expression) on pre treatment samples of pro-aggressive and non-pro-aggressive ccRCC patients described above. n = 15, Chi-square test for independence P = 0.015, and Chi-square test for trend P = 0.030 (*) (C) OC curve for the ALDH1 A3 score on 15 ccRCC patients. ROC variation method was used due to categorical predictor factor. AUC = 80.6 13.8, Sensitivity 83.3% [95%CI: 35.9-99.5], Specificity 33.3% [95%CI: 7.5-70.0], empirical P = 0.008 by Bootstrapping method (n = 1,000).

Figure 13: Patients with high ALDH13A levels present worst risk of progression (Statistical significance (p-value=<0.050) different cancer types: (A) Colon adenocarcinoma, (B) bladder Urothelial Carcinoma, (C) Esophageal Carcinoma, (D) kidney Renal Clear Cell Carcinoma, (E) low Grade Glioma, (F) Ovarian serous cystadenocarcinoma and (G) Stomach adenocarcinoma.

DETAILED DESCRIPTION OF THE INVENTION

As explained above, the present invention relates to the finding that the expression level of the ALDH1A3 gene is a useful marker in the prognosis of cancer patients, particularly of patients who suffer from clear cell renal cell carcinoma (ccRCC). This method can be used to determine those patients who are at a high risk of developing local invasion and/or metastasis induced by an antiangiogenic treatment. The method also allows for the identification of patients that can exhibit a better response to anti-angiogenic agents, improving financial and health outcomes.

Methods for predicting the risk of local invasion and/or metastasis

In a first aspect, the invention relates to a method for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer, wherein the method comprises

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1 A3 that is increased with respect to the reference value is indicative of an increased risk of developing local invasion and/or metastasis in response to the antiangiogenic treatment.

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1 A3 that is increased with respect to the reference value is indicative that the patient is at an increased risk of developing local invasion and/or metastasis in response to the antiangiogenic treatment.

In the context of the invention, ALDH1A3 (also known as ALDH6 or RALDH3) refers to a gene encoding for aldehyde dehydrogenase 1 family member A3. The human gene is shown in the Ensembl database under accession number ENSG00000184254. It will be understood that the method according to the present invention may comprise the determination of any naturally occurring polymorphic variant of the above gene.

The methods of the invention comprise comparing the expression level of the ALDH1A3 gene with a reference value. “Reference value”, as used herein, refers to a laboratory value used as a reference for values/data obtained by laboratory examinations of subjects or samples collected from subjects. The reference value or reference level can be an absolute value; a relative value; a value that has an upper and/or lower limit; a range of values; an average value; a median value, a mean value, or a value as compared to a particular control or baseline value. A reference value can be based on an individual sample value, such as for example, a value obtained from a sample from the subject being tested, but at an earlier point in time or from a non-cancerous tissue. The reference value can be based on a large number of samples, such as from population of subjects of the chronological age matched group, or based on a pool of samples including or excluding the sample to be tested. Various considerations are taken into account when determining the reference value of the marker. Among such considerations are the age, weight, sex, general physical condition of the patient and the like. For example, equal amounts of a group of at least 2, at least 10, at least 100 to preferably more than 1000 subjects, preferably classified according to the foregoing considerations, for example according to various age categories, are taken as the reference group. In another embodiment, the quantity of the biomarker in a sample from a tested subject may be determined directly relative to the reference value (e.g., in terms of increase or decrease, or fold-increase or fold-decrease). Advantageously, this may allow to compare the quantity of the biomarker in the sample from the subject with the reference value (in other words to measure the relative quantity of any one or more biomarkers in the sample from the subject vis-a-vis the reference value) without the need to first determine the respective absolute quantities of said biomarker.

Typically, reference values are the expression level of the gene being compared in a reference sample. In an embodiment, the “reference sample”, as used herein, means a sample obtained from a pool of healthy subjects who do not have a disease state or particular phenotype. For example, the reference sample may comprise samples from tissue of patients which do not suffer from cancer or which do not have a history of cancer. Thus, in an embodiment, the reference value is the mean level of expression of ALDH1A3 in a pool of samples from healthy tissue.

In another embodiment, the reference value for the expression level of ALDH1A3 is the mean level of expression of ALDH1A3 in a pool of samples from primary tumours, preferably obtained from subjects suffering from the same type of cancer as the patient object of the study. In a particular embodiment, the reference value is the expression levels of the gene of interest in a pool obtained from primary tumor tissue obtained from patients. This pool will include patients with good prognosis and patients with bad prognosis, that is, the pool will include patients who develop invasion and/or metastasis in response to treatment with anti-angiogenics and patients who do not develop invasion and/or metastasis in response to treatment with anti-angiogenics, and therefore, the expression levels would be an average value of the values found in the different types of patients. In another embodiment, the reference value is the expression levels of the gene of interest in primary tumor tissue obtained from a patient or patients identified as patients having a good prognosis, that is, patients who do not develop invasion and/or metastasis in response to treatment with anti-angiogenics. In another embodiment, the reference value is the expression levels of the gene of interest in a tumor tissue obtained from a patient or patients identified as patients having a bad prognosis, that is, patients who develop invasion and/or metastasis in response to treatment with anti-angiogenics.

The expression profile of the genes in the reference sample can preferably, be generated from a population of two or more individuals. The population, for example, can comprise 3, 4, 5, 10, 15, 20, 30, 40, 50 or more individuals. Furthermore, the expression profile of the genes in the reference sample and in the sample of the individual that is going to be diagnosed according to the methods of the present invention can be generated from the same individual, provided that the profiles to be assayed and the reference profile are generated from biological samples taken at different times and are compared to one another. For example, a sample of an individual can be obtained at the beginning of a study period. A reference biomarker profile from this sample can then be compared with the biomarker profiles generated from subsequent samples of the same individual. In a preferred embodiment, the reference sample is a pool of samples from several individuals and corresponds to portions of tissue that are far from the tumor area and which have preferably been obtained in the same biopsy but which do not have any anatomo-pathologic characteristic of tumor tissue.

Once this reference value is established, the level of this marker expressed in tumor tissue from subjects can be compared with this reference value, and thus be assigned a level of “increased” or “decreased”, depending on whether the expression level of the marker is above, equal to or below the reference value. For example, an increase in expression level above the reference value of at least 1.1-fold, 1.5-fold, 5-fold, 10-fold, 20-fold, 30-fold, 40-fold, 50-fold, 60-fold, 70-fold, 80-fold, 90-fold, 100-fold or even more compared with the reference value is considered as “increased” expression level. Similarly, the expression of a gene is considered increased in a sample of the subject under study when the levels increase with respect to the reference sample by at least 5%, by at least 10%, by at least 15%, by at least 20%, by at least 25%, by at least 30%, by at least 35%, by at least 40%, by at least 45%, by at least 50%, by at least 55%, by at least 60%, by at least 65%, by at least 70%, by at least 75%, by at least 80%, by at least 85%, by at least 90%, by at least 95%, by at least 100%, by at least 110%, by at least 120%, by at least 130%, by at least 140%, by at least 150%, or more. On the other hand, a decrease in expression levels below the reference value of at least 0.9-fold, 0.75- fold, 0.2-fold, 0.1-fold, 0.05-fold, 0.025-fold, 0.02-fold, 0.01-fold, 0.005-fold or even less compared with reference value is considered as “decreased” expression level. Similarly, the expression of a gene is considered decreased when its levels decrease with respect to the reference sample by at least 5%, by at least 10%, by at least 15%, by at least 20%, by at least 25%, by at least 30%, by at least 35%, by at least 40%, by at least 45%, by at least 50%, by at least 55%, by at least 60%, by at least 65%, by at least 70%, by at least 75%, by at least 80%, by at least 85%, by at least 90%, by at least 95%, by at least 100% (i.e. , absent). The comparison of the expression levels of the gene of interest with the reference value allows determining whether the patient will show a good or poor prognosis.

In the methods of the invention, an expression level of ALDH1A3 that is increased with respect to the reference value is indicative of an increased risk of developing local invasion and/or metastasis in response to an antiangiogenic treatment. As used herein, "an increased risk of developing local invasion and/or metastasis" indicates that the subject is expected, i.e. predicted, to have, or is at high risk of having, local invasion and/or distant metastases within a set time period. The term "high" is a relative term and, in the context of this application, refers to the risk of the "high" expression group with respect to a clinical outcome (local invasion, distant metastases). A "high" risk can be considered as a risk higher than the average risk for a heterogeneous cancer patient population. The risk will also vary in function of the time period. The time period can be, for example, five years, ten years, fifteen years or even twenty years of initial diagnosis of cancer or after the prognosis was made. As used herein, "a decreased risk of developing local invasion and/or metastasis" indicates that the subject is expected (e.g. predicted) to have no, or is at low risk of having, local invasion and/or distant metastases within a set time period. The term "low" is a relative term and, in the context of this application, refers to the risk of the "low" expression group with respect to a clinical outcome (local invasion, distant metastases, etc.). A "low" risk can be considered as a risk lower than the average risk for a heterogeneous cancer patient population. The risk will also vary in function of the time period. The time period can be, for example, five years, ten years, fifteen years or even twenty years after initial diagnosis of cancer or after the prognosis is made.

In the context of the invention, the term “local invasion” refers to the direct extension and penetration by cancer cells into neighboring tissues. The proliferation of transformed cells and the progressive increase in tumor size eventually leads to a breach in the barriers between tissues, leading to tumor extension into adjacent tissue. Local invasion is typically the first stage in the process that leads to the development of secondary tumors or metastases. The term "metastasis" as used herein refers to the growth of a cancerous tumor in an organ or body part, which is not directly connected to the organ of the original cancerous tumor. Metastasis will be understood to include micrometastasis, which is the presence of an undetectable amount of cancerous cells in an organ or body part which is not directly connected to the organ of the original cancerous tumor. In a preferred embodiment, the metastasis is lung metastasis. The term "risk of local invasion and/or metastasis", as used herein, refers to a likelihood or probability assessment regarding the chances or the probability that a subject or individual may develop a similar or the same neoplastic disease at an adjacent tissue or at an anatomically distant location within a defined time interval, comparable to the one that the subject or individual has been treated for or diagnosed for.

The term “predicting the response to an anti-angiogenic treatment”, as used herein, relates to the prediction of a medical outcome following a therapeutic intervention using an anti-angiogenic treatment. The outcome after the treatment may be determined using any common end point for patient progression, such as, for example, a poor or good outcome (e.g., likelihood of long-term survival, overall survival, disease-specific survival, progression-free survival or disease-free survival). A negative prognosis, or poor outcome, includes a prediction of relapse, disease progression (e.g., tumor growth, local invasion and/or metastasis, or drug resistance), or mortality; whereas a positive prognosis, or good outcome, includes a prediction of disease remission, (e.g., disease- free status), amelioration (e.g., tumor regression), or stabilization. Any parameter which is widely accepted for determining the outcome of a patient can be used in the present invention including, without limitation: • overall survival rate, as used herewith, relates to the percentage of people in a study or treatment group who are alive for a certain period of time after they were diagnosed with or treated for a disease, such as cancer.

• disease-specific survival rate which is defined as the percentage of people in a study or treatment group who have not died from a specific disease in a defined period of time.

• disease-free survival (DFS), as used herewith, is understood as the length of time after treatment for a disease during which a subject survives with no sign of the disease.

• objective response which, as used in the present invention, describes the proportion of treated subjects in whom a complete or partial response is observed.

• tumor control which, as used in the present invention, relates to the proportion of treated subjects in whom complete response, partial response, minor response or stable disease ³ 6 months is observed.

• progression free survival which, as used herein, is defined as the time from start of treatment to the first measurement of cancer growth.

• time to progression (TTP), as used herein, relates to the time since a disease is treated until the disease starts to get worse.

• six-month progression free survival or “PFS6” rate which, as used herein, relates to the percentage of subjects who are free of progression in the first six months after the initiation of the therapy and

• median survival which, as used herein, relates to the time at which half of the subjects enrolled in the study are still alive.

In an embodiment, the term “predicting the response to an anti-angiogenic treatment”, as used herein, relates to the prediction of the risk of local invasion and/or metastasis following a therapeutic intervention using an anti-angiogenic treatment In a particular embodiment, the methods of the invention further comprise the determination of one or more clinical parameters which are also indicative of the prognosis of the cancer. Such indicators include the presence or levels of known cancer markers, or can be clinical or pathological indicators (for example, age, tumor size, tumor histology, differentiation grade, clinical stage, family history and the like). The person skilled in the art will understand that the determination of the prediction does not need to be correct for all subjects (i.e., for 100% of the subjects). Nevertheless, the term requires enabling the identification of a statistically significant part of the subjects (for example, a cohort in a cohort study). Whether a part is statistically significant can be determined in a simple manner by the person skilled in the art using various well known statistical evaluation tools, for example, the determination of confidence intervals, determination of p values, Student’s T test, Mann-Whitney test, etc. The preferred confidence intervals are at least 90%, at least 95%, at least 97%, at least 98% or at least 99%. The p values are preferably 0.1 , 0.05, 0.01 , 0.005 or 0.0001. More preferably, at least 60%, at least 70%, at least 80% or at least 90% of the subjects of a population can be suitably identified by the method of the present invention.

The term “patient” or “subject”, as used herein, refers to all animals classified as mammals and includes, but is not restricted to, domestic and farm animals, primates and humans, e.g., human beings, non-human primates, cows, horses, pigs, sheep, goats, dogs, cats, or rodents. Preferably, the patient is a male or female human of any age or race. In a particular embodiment, the subject has not been treated with antiangiogenics prior to the determination of the expression levels of the gene of interest, that is, the step of determining the expression level of ALDH1A3 in the sample from the patient is performed prior to initiating an antiangiogenic treatment. In another particular embodiment, the subject suffering from cancer whose response to treatment is to be determined by the methods of the invention is undergoing an anti-angiogenic treatment, wherein said anti-angiogenic treatment of cancer is based on at least one anti angiogenesis agent.

Anti-angiogenic agents and treatments according to the invention include, without limitation anti-VEGF agents, including monoclonal antibodies such as bevacizumab (Avastin, a recombinant humanized monoclonal lgG1 antibody that binds to and inhibits the biological activity of human VEGFA in in vitro and in vivo assay systems), antibody derivatives such as ranibizumab (Lucentis), or antibody fragments such as Fab IMC 1121 or F200 Fab or orally-available small molecules that inhibit the tyrosine kinases stimulated by VEGF such as lapatinib (Tykerb), sunitinib (Sutent), sorafenib (Nexavar), axitinib, and pazopanib; anti-fibroblast growth factor(anti-FGF) agents, such as suramin and its derivatives, pentosanpolysulfate, cediranib, pazopanib, or BIBF 1120); anti-EGF agents, such as cetuximab, gefitinib or erlotinib and anti-HGF agents, such as ARQ197, JNJ-38877605, PF-04217903, SGX523, NK4, or AMG102; and antiangiogenic polypeptides such as angiostatin, endostatin, anti-angiogenic anti-thrombin III or sFRP- 4. Further anti-angiogenic agents include Marimastat; AG3340; COL-3, BMS-275291, Thalidomide, Endostatin, SU5416, SU6668, EMD121974, 2-methoxyoestradiol, carboxiamidotriazole, CMIOI (GBS toxin), pentosanpolysulphate, angiopoietin 2 (Regeneron), herbimycin A, PNU145156E, 16K prolactin fragment, Linomide, thalidomide, pentoxifylline, genistein, TNP470, endostatin, paclitaxel, accutin, angiostatin, cidofovir, vincristine, bleomycin, AGM- 1470, platelet factor 4 or minocycline. Further anti-angiogenic agents include anti-angiogenic polypeptides, denoting polypeptides capable of inhibiting angiogenesis and including, without limitation, angiostatin, endostatin, anti-angiogenic anti-thrombin III, sFRP- 4 as described in W02007115376, an anti-VEGF antibody such as anibizumab, bevacizumab (avastin), Fab IMC 1121 and F200 Fab. Further anti-angiogenic agents include pegaptanib, sunitinib, pazopanib, sorafenib, vatalanib and aflibercept (VEGF-Trap). Further anti- angiogenic agents include VEGFR2 blocking antibodies, such as Ramucirumab (IMC- 1121 B) and DC101 (also known as anti-Flk-1 mAb).

The term “treatment”, as used herein comprises any type of therapy, which aims at terminating, preventing, ameliorating and/or reducing the susceptibility to a clinical condition as described herein. In a preferred embodiment, the term treatment relates to prophylactic treatment (i.e. a therapy to reduce the susceptibility of a clinical condition, a disorder or condition as defined herein). Thus, “treatment,” “treating,” and the like, as used herein, refer to obtaining a desired pharmacologic and/or physiologic effect, covering any treatment of a pathological condition or disorder in a mammal, including a human. The effect may be prophylactic in terms of completely or partially preventing a disorder or symptom thereof and/or may be therapeutic in terms of a partial or complete cure for a disorder and/or adverse effect attributable to the disorder. That is, “treatment” includes (1) preventing the disorder from occurring or recurring in a subject, (2) inhibiting the disorder, such as arresting its development, (3) stopping or terminating the disorder or at least symptoms associated therewith, so that the host no longer suffers from the disorder or its symptoms, such as causing regression of the disorder or its symptoms, for example, by restoring or repairing a lost, missing or defective function, or stimulating an inefficient process, or (4) relieving, alleviating, or ameliorating the disorder, or symptoms associated therewith, where ameliorating is used in a broad sense to refer to at least a reduction in the magnitude of a parameter, such as inflammation, pain, and/or immune deficiency.

The term “cancer” is referred to a disease characterized by uncontrolled cell division (or by an increase of survival or apoptosis resistance) and by the ability of said cells to invade other neighboring tissues (invasion) and spread to other areas of the body where the cells are not normally located (metastasis) through the lymphatic and blood vessels, circulate through the bloodstream, and then invade normal tissues elsewhere in the body. Depending on whether or not they can spread by invasion and metastasis, tumors are classified as being either benign or malignant: benign tumors are tumors that cannot spread by invasion or metastasis, i.e. , they only grow locally; whereas malignant tumors are tumors that are capable of spreading by invasion and metastasis. Biological processes known to be related to cancer include angiogenesis, immune cell infiltration, cell migration and metastasis. The term cancer includes, without limitation, lung cancer, sarcoma, malignant melanoma, pleural mesothelioma, bladder carcinoma, prostate cancer, pancreas carcinoma, gastric carcinoma, ovarian cancer, hepatoma, breast cancer, colorectal cancer, kidney cancer, esophageal cancer, suprarenal cancer, parotid gland cancer, head and neck carcinoma, cervix cancer, endometrial cancer, liver cancer, mesothelioma, multiple myeloma, leukaemia, and lymphoma.

In one embodiment, the cancer is colon cancer, bladder cancer, esophageal cancer, glioma, ovarian cancer or stomach cancer. In another embodiment, the colon cancer is colon adenocarcinoma, the bladder cancer is bladder urothelial carcinoma, the esophageal cancer is esophageal carcinoma, the glioma is low grade glioma, the ovarian cancer is ovarian serous cystadenocarcinoma or the stomach cancer is stomach adenocarcinoma.

In a particular embodiment of the invention, the cancer is renal cancer. In another particular embodiment, the cancer is renal cell carcinoma (RCC). In a still more particular embodiment, the cancer is clear cell renal cell carcinoma (ccRCC). In a particular embodiment, the cancer is not a brain tumor. In a particular embodiment, the cancer is not a breast cancer. In a particular embodiment, the cancer is not a cervical cancer. In a particular embodiment, the cancer is not a colorectal cancer. In a particular embodiment, the cancer is not an endometrial cancer. In a particular embodiment, the cancer is not a hepatocellular carcinoma. In a particular embodiment, the cancer is not a lung cancer. In a particular embodiment, the cancer is not uveal melanoma. In a particular embodiment, the cancer is not ovarian cancer. In a particular embodiment, the cancer is not a pancreatic tumor. In a particular embodiment, the cancer is not a pancreatic neuroendocrine tumor. In a particular embodiment, the cancer is not a prostate cancer. In a particular embodiment, the cancer is not a renal cancer.

As used herein, "sample" or "biological sample" means biological material isolated from a subject. The biological sample may contain any biological material suitable for determining the expression level of ALDH1A3. The sample can be isolated from any suitable biological tissue or fluid such as, for example, tumor tissue, blood, plasma, serum, sputum, bronchoalveolar lavage, urine or cerebral spinal fluid (CSF). In a preferred embodiment, the sample contains tumor cells. In a more preferred embodiment, the sample containing tumor cells is a tumor tissue sample. The tumor tissue sample is understood as the tissue sample originating from the primary tumor or from a distant metastasis. Said sample can be obtained by conventional methods, for example biopsy, using methods well known by the person skilled in related medical techniques. Alternatively, the tumor tissue sample may be a sample of a tumor which has been surgically resected. The methods for obtaining a biopsy sample include splitting a tumor into large pieces, or microdissection, or other cell separating methods known in the art. The tumor cells can additionally be obtained by means of cytology through aspiration with a small gauge needle. To simplify sample preservation and handling, samples can be fixed in formalin and soaked in paraffin or first frozen and then soaked in a tissue freezing medium such as OCT compound by means of immersion in a highly cryogenic medium which allows rapid freezing. Alternatively, the samples are biofluid samples. The terms "biological fluid" and "biofluid" are used interchangeably herein and refer to aqueous fluids of biological origin. The biofluid may be obtained from any location (such as blood, plasma, serum, urine, bile, cerebrospinal fluid, aqueous or vitreous humor, or any bodily secretion), an exudate (such as fluid obtained from an abscess or any other site of infection or inflammation), or fluid obtained from a joint (such as a normal joint or a joint affected by disease such as rheumatoid arthritis). In a particular embodiment, the sample is a urine sample.

The term "expression level" of a gene as used herein refers to the measurable quantity of gene product produced by the gene in a sample of the subject, wherein the gene product can be a transcriptional product or a translational product. As understood by the person skilled in the art, the gene expression level can be quantified by measuring the messenger RNA levels of said gene or of the protein encoded by said gene. Thus, the expression levels of the ALDH1A3 gene can be determined by measuring the levels of mRNA encoded by said gene, or by measuring the levels of the protein encoded by said gene, i.e. ALDH1A3 protein, or of variants thereof. Alternatively, in another particular embodiment, the activity level of ALDH1A3 is determined. In a particular embodiment, the methods of the invention do not comprise determining the expression levels of the c- Met gene or the expression levels of the PKCA gene. In another particular embodiment, the methods of the invention do not comprise determining the expression level of any gene other than ALDH1A3 as a biomarker for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer.

In order to measure the mRNA levels of the ALDH1A3 gene, the biological sample may be treated to physically, mechanically or chemically disrupt tissue or cell structure, to release intracellular components into an aqueous or organic solution to prepare nucleic acids for further analysis. The nucleic acids are extracted from the sample by procedures known to the skilled person and commercially available. RNA is then extracted from frozen or fresh samples by any of the methods typical in the art, for example, Sambrook, J., et al., 2001. Molecular cloning: A Laboratory Manual, 3^rd ed., Cold Spring Harbor Laboratory Press, N.Y., Vol. 1-3. Preferably, care is taken to avoid degradation of the RNA during the extraction process. The expression level can be determined using mRNA obtained from a formalin-fixed, paraffin-embedded tissue sample. mRNA may be isolated from an archival pathological sample or biopsy sample which is first deparaffinized. An exemplary deparaffinization method involves washing the paraffinized sample with an organic solvent, such as xylene. Deparaffinized samples can be rehydrated with an aqueous solution of a lower alcohol. Suitable lower alcohols, for example, include methanol, ethanol, propanols and butanols. Deparaffinized samples may be rehydrated with successive washes with lower alcoholic solutions of decreasing concentration, for example. Alternatively, the sample is simultaneously deparaffinized and rehydrated. The sample is then lysed and RNA is extracted from the sample. Samples can be also obtained from fresh tumor tissue such as a resected tumor. In a preferred embodiment samples can be obtained from fresh tumor tissue or from OCT embedded frozen tissue. In another preferred embodiment samples can be obtained by bronchoscopy and then paraffin-embedded. Determination of the levels of mRNA of the ALDH1A3 gene can be carried out by any method known in the art such as qPCR, northern blot, RNA dot blot, TaqMan, tag based methods such as serial analysis of gene expression (SAGE) including variants such as LongSAGE and SuperSAGE, microarrays. Determination of the levels of the above genes can also be carried out by Fluorescence In Situ Hybridization, including variants such as Flow-FISH, qFiSH and double fusion FISH (D-FISH) as described in W02010030818. The levels of the mRNA of the different genes can also be determined by nucleic acid sequence based amplification (NASBA) technology. In a preferred embodiment, the gene mRNA expression levels are often determined by reverse transcription polymerase chain reaction (RT-PCR). The detection can be carried out in individual samples or in tissue microarrays. Thus, in a particular embodiment, the mRNA expression levels of the ALDH1A3 gene are determined by quantitative PCR, preferably, Real-Time PCR. The detection can be carried out in individual samples or in tissue microarrays.

Alternatively, in another embodiment of the methods of the invention, the expression level of the ALDH1A3 gene is determined by measuring the expression of the polypeptide or of variants thereof. In a preferred embodiment the expression level of the protein or of variants thereof is determined by Western blot, ELISA or by immunohistochemistry. The expression levels of the protein encoded by the ALDH1A3 gene can be quantified by means of conventional methods, for example, using antibodies with a capacity to specifically bind to the proteins encoded by said genes (or to fragments thereof containing antigenic determinants) and subsequent quantification of the resulting antibody-antigen complexes. The antibodies to be employed in these assays can be, for example, polyclonal sera, hybridoma supernatants or monoclonal antibodies, antibody fragments, Fv, Fab, Fab’ and F(ab’)2, ScFv, diabodies, triabodies, tetrabodies and humanized antibodies. At the same time, the antibodies can be labeled or not. Illustrative, but non-exclusive examples of markers which can be used include radioactive isotopes, enzymes, fluorophores, chemiluminescent reagents, enzymatic substrates or cofactors, enzymatic inhibitors, particles, colorants, etc. There are a wide variety of well-known assays that can be used in the present invention, which use non-labeled antibodies (primary antibody) and labeled antibodies (secondary antibodies); among these techniques are included Western blot or Western transfer, ELISA (enzyme linked immunosorbent assay), RIA (radioimmunoassay), competitive EIA (enzymatic immunoassay), DAS-ELISA (double antibody sandwich ELISA), immunocytochemical and immunohistochemical techniques, techniques based on the use of biochips or protein microarrays including specific antibodies or assays based on colloidal precipitation in formats such as dipsticks. Other ways of detecting and quantifying the levels of the protein of interest include techniques of affinity chromatography, binding- ligand assays, etc. Alternatively, in another particular embodiment, the levels of the protein encoded by the ALDH1A3 gene or of the variants thereof are determined by Western blot. Western blot is based on the detection of proteins previously resolved by gel electrophoreses under denaturing conditions and immobilized on a membrane, generally nitrocellulose, by the incubation with an antibody specific and a developing system (e.g. chemoluminiscent).

In an alternative, the activity levels of the protein encoded by the ALDH1A3 gene are determined. The term "activity level" of a protein, more particularly of an enzyme, as used herein refers to a measure of the enzyme activity, particularly measured as moles of substrate converted per unit of time. Assays to determine the activity level of an enzyme are known by the skilled person and include, without limitation, initial rate assays, progress curve assays, transient kinetics assays and relaxation assays. Continuous assays of enzymatic activity include, without limitation, spectrophotometric, fluorometric, calorimetric, chemiluminiscent, light scattering and microscale thermopheresis assays. Discontinuous assays of enzymatic activity include, without limitation, radiometric and chromatographic assays. As the skilled person understands, factors that may influence enzymatic activity comprise salt concentration, temperature, pH, and substrate concentration.

All the terms and embodiments described elsewhere herein are equally applicable to these aspects of the invention.

Methods of personalized medicine with an antiangiogenic therapy

The authors of the present invention have found that the levels of ALDH1A3 determined in cancer patients show a statistically significant correlation with the risk that the patients show local invasion and/or metastasis after treatment with an anti-angiogenic therapy.

Thus, in a further aspect, the invention relates to a method for selecting a patient suffering from cancer for treatment with an antiangiogenic therapy, wherein the method comprises (i) determining the expression level of ALDH1A3 in a sample from said patient, and

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1A3 that is decreased with respect to the reference value is indicative that the patient is not at risk of local invasion and/or metastasis in response to an antiangiogenic therapy.

It is considered that the patient will benefit from treatment with an antiangiogenic therapy if said patient is not at risk of local invasion and/or metastasis in response to an antiangiogenic therapy.

In a related aspect, the invention relates to a method for treating a patient suffering from cancer, wherein the method comprises

(i) determining the expression level of ALDH1A3 in a sample from said patient,

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1A3 that is decreased with respect to the reference value is indicative that the patient is not at risk of local invasion and/or metastasis in response to an antiangiogenic therapy; and (iii) administering treatment based on an antiangiogenic compound if the patient is not at risk of local invasion and/or metastasis in response to the administration of said antiangiogenic therapy.

In another related aspect, the invention relates to the use of an antiangiogenic compound for the manufacture of a medicament for the treatment of a patient suffering from cancer, wherein the patient is selected using a method that comprises

The terms “patient”, “cancer”, “antiangiogenics”, “treatment”, “reference value”, “sample”, “predicting the risk of local invasion and/or metastasis” and “predicting the response to an antiangiogenic therapy”, among others, have been defined above and are equally applicable to these methods of personalized medicine according to the present invention. All the terms and embodiments described elsewhere herein are equally applicable to these aspects of the invention.

In the context of the invention, when a treatment based on antiangiogenics is administered to a patient, said antiangiogenic treatment is based on at least one antiangiogenic compound, i.e. , it may be based on a single antiangiogenic agent or on a combination of antiangiogenic compounds. In a particular embodiment, the antiangiogenic treatment is administered as a primary treatment, or administered as an adjuvant therapy or neoadjuvant therapy accompanying another treatment. The term "neoadjuvant therapy", as used herein, refers to any type of treatment of cancer given prior to administration of the primary treatment, i.e., prior to surgical resection of the primary tumor, in a patient affected with a cancer. The most common reason for neoadjuvant therapy is to reduce the size of the tumor so as to facilitate a more effective surgery. Other neoadjuvant therapies comprise radiotherapy and therapy, preferably systemic therapy, such as hormone therapy, chemotherapy, immunotherapy and monoclonal antibody therapy. The term "adjuvant therapy", as used herein, refers to any type of treatment of cancer given as additional treatment, usually concomitant with or after the primary treatment, i.e. , after surgical resection of the primary tumor, in a patient affected with a cancer that is at risk of metastasizing and/or likely to recur. The aim of such an adjuvant treatment is to improve the prognosis. Other adjuvant therapies comprise radiotherapy and therapy, preferably systemic therapy, such as hormone therapy, chemotherapy, immunotherapy and monoclonal antibody therapy.

Systemic treatments of cancer include but are not limited to chemotherapy, hormone treatment, immunotherapy, or a combination thereof. Additionally, radiotherapy and/or surgery can be used. The choice of treatment generally depends on the type of primary cancer, the size, the location of the metastasis, the age, the general health of the patient and the types of treatments used previously. The term “surgery” or “surgical treatment”, as used herein, means any therapeutic procedure that involves methodical action of the hand or of the hand with an instrument, on the body of a human or other mammal, to produce a cure or remedy. As used herein the term "chemotherapy", "chemotherapeutic drug" refers broadly to the use of a chemical drug or a combination thereof for the treatment of cancer, tumors or malign neoplasia, including both cytotoxic and cytostatic drugs. Examples of chemotherapy agents which may be in accordance to the present invention include: alkylating agents (for example mechlorethamine, chlorambucil, cyclophosphamide, ifosfamide, streptozocin, carmustine, lomustine, melphalan, busulfan, dacarbazine, temozolomide, thiotepa or altretamine); platinum drugs (for example cisplatin, carboplatin or oxaliplatin); antimetabolite drugs (for example 5- fluorouracil, capecitabine, 6-mercaptopurine, methotrexate, gemcitabine, cytarabine, fludarabine or pemetrexed); anti-tumor antibiotics (for example daunorubicin, doxorubicin, epirubicin, idarubicin, actinomycin-D, bleomycin, mitomycin-C or mitoxantrone); mitotic inhibitors (for example paclitaxel, docetaxel, ixabepilone, vinblastine, vincristine, vinorelbine, vindesine or estramustine); and topoisomerase inhibitors (for example etoposide, teniposide, topotecan, irinotecan, diflomotecan or elomotecan). The term "radiotherapy" is a term commonly used in the art to refer to multiple types of radiation therapy including internal and external radiation therapies or radioimmunotherapy, and the use of various types of radiations including X-rays, gamma rays, alpha particles, beta particles, photons, electrons, neutrons, radioisotopes, and other forms of ionizing radiations. The antiangiogenic compound or compounds for use according to the invention may be formulated into a pharmaceutical composition. This pharmaceutical composition may be in any dosage form suitable for administration to a subject, illustratively including solid, semi-solid and liquid dosage forms such as tablets, capsules, powders, granules, suppositories, pills, solutions, suspensions, ointments, lotions, creams, gels, pastes, sprays and aerosols. Liposomes and emulsions are well-known types of pharmaceutical formulations that can be used to deliver a pharmaceutical agent, particularly a hydrophobic pharmaceutical agent. The pharmaceutical compositions generally include a pharmaceutically acceptable carrier such as an excipient, diluent and/or vehicle. Delayed release formulations of compositions and delayed release systems, such as semipermeable matrices of solid hydrophobic polymers can be used. The term "pharmaceutically acceptable carrier" refers to a carrier which is suitable for use in a subject without undue toxicity or irritation to the subject and which is compatible with other ingredients included in a pharmaceutical composition. Pharmaceutically acceptable carriers, methods for making pharmaceutical compositions and various dosage forms, as well as modes of administration are well-known in the art, for example as detailed in Pharmaceutical Dosage Forms: Tablets, eds. H. A. Lieberman et al., New York: Marcel Dekker, Inc., 1989; and in L. V. Allen, Jr. et al., Ansel's Pharmaceutical Dosage Forms and Drug Delivery Systems, 8th Ed., Philadelphia, Pa.: Lippincott, Williams and Wilkins, 2004; A. R. Gennaro, Remington: The Science and Practice of Pharmacy, Lippincott Wiliams and Wlkins, 21st ed., 2005, particularly chapter 89; and J. G. Hardman et al., Goodman and Gilman's The Pharmacological Basis of Therapeutics, McGraw-Hill Professional, 10th ed., 2001. The dosage of the antiangiogenic compound or compounds for use according to the method of the invention will vary based on factors such as, but not limited to, the route of administration; the age, health, sex, and weight of the subject to whom the composition is to be administered; the nature and extent of the subject's symptoms, if any, and the effect desired. Dosage may be adjusted depending on whether treatment is to be acute or continuing. One of skill in the art can determine a pharmaceutically effective amount in view of these and other considerations typical in medical practice. Detailed information concerning customary ingredients, equipment and processes for preparing dosage forms is found in Pharmaceutical Dosage Forms: Tablets, eds. H. A. Lieberman et al., New York: Marcel Dekker, Inc., 1989; and in L. V. Allen, Jr. et al. , Ansel's Pharmaceutical Dosage Forms and Drug Delivery Systems, 8th Ed., Philadelphia, Pa.: Lippincott, Williams and Wlkins, 2004; A. R. Gennaro, Remington: The Science and Practice of Pharmacy, Lippincott Williams and Wilkins, 21st ed., 2005, particularly chapter 89; and J. G. Hardman et al., Goodman and Gilman's The Pharmacological Basis of Therapeutics, McGraw-Hill Professional, 10th ed., 2001.

Uses of the invention

In a final aspect, the invention relates to the use of reagents specific for the determination of the expression level of ALDH1A3, for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer. In a particular embodiment, the reagents are selected from the group of a set probes which specifically hybridize to the mRNA of said gene and a set of primer pairs which are capable of specifically amplifying the mRNA of said gene, or wherein the reagents are a set of antibodies which specifically bind to the polypeptide encoded by said gene. In another particular embodiment, the reagents are reagents to detect and quantify the activity of the protein encoded by the ALDH1A3 gene.

The reagents specific for the determination of the expression level of ALDH1A3 can be provided as part of a kit. In particular embodiments, the reagents adequate for the determination of the expression levels of the ALDH1A3 gene comprise at least 10%, at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 90% or at least 100% of the total amount of reagents adequate for the determination of the expression levels of genes forming the kit, including housekeeping genes.

In the context of the present invention, “kit” is understood as a product containing the different reagents necessary for carrying out the methods of the invention packed so as to allow their transport and storage. Materials suitable for packing the components of the kit include crystal, plastic (polyethylene, polypropylene, polycarbonate and the like), bottles, vials, paper, envelopes and the like. Additionally, the kits of the invention can contain instructions for the simultaneous, sequential or separate use of the different components which are in the kit. Said instructions can be in the form of printed material or in the form of an electronic support capable of storing instructions such that they can be read by a subject, such as electronic storage media (magnetic disks, tapes and the like), optical media (CD-ROM, DVD) and the like. Additionally or alternatively, the media can contain Internet addresses that provide said instructions.

The expression “reagent which allows determining the expression level of a gene” means a compound or set of compounds that allows determining the expression level of a gene both by means of the determination of the level of mRNA or by means of the determination of the level of protein. Thus, reagents of the first type include probes capable of specifically hybridizing with the mRNAs encoded by said genes. Reagents of the second type include compounds that bind specifically with the proteins encoded by the marker genes and preferably include antibodies, although they can be specific aptamers.

In a particular embodiment of the kit of the invention, the reagents of the kit are nucleic acids which are capable of specifically detecting the mRNA level of the genes mentioned above and/or the level of proteins encoded by one or more of the genes mentioned above. Nucleic acids capable of specifically hybridizing with the genes mentioned above can be one or more pairs of primer oligonucleotides for the specific amplification of fragments of the mRNAs (or of their corresponding cDNAs) of said genes. In a preferred embodiment, the first component of the kit of the invention comprises a probe which can specifically hybridize to the gene mentioned above. The term “specifically hybridizing”, as used herein, refers to conditions which allow hybridizing of two polynucleotides under high stringent conditions or moderately stringent conditions. "Stringency" of hybridization reactions is readily determinable by one of ordinary skill in the art, and generally is an empirical calculation dependent upon probe length, washing temperature, and salt concentration. In general, longer probes require higher temperatures for proper annealing, while shorter probes need lower temperatures. Hybridization generally depends on the ability of denatured DNA to reanneal when complementary strands are present in an environment below their melting temperature. The higher the degree of desired homology between the probe and hybridizable sequence, the higher the relative temperature which can be used. As a result, it follows that higher relative temperatures would tend to make the reaction conditions more stringent, while lower temperatures less so. For additional details and explanation of stringency of hybridization reactions, see Ausubel et al. , Current Protocols in Molecular Biology, Wiley Interscience Publishers, (1995). The reagents to determine the expression levels of housekeeping controls are probes which hybridize specifically with genes which are expressed constitutively in the sample which is analyzed. The expression level controls are designed to control the physiological state and the metabolic activity of the cell. The examination of the covariance of the expression level control with the expression level of the target nucleic acid indicates if the variations in the expression levels are due to changes in the expression levels or are due to changes in the overall transcriptional rate in the cell or in its general metabolic activity. Thus, in the case of cells which have deficiencies in a certain metabolite essential for cell viability, the observation of a decrease both in the expression levels of the target gene as in the expression levels of the control is expected. On the other hand, if an increase in the expression of the expression of the target gene and of the control gene is observed, it probably due to an increase of the metabolic activity of the cell and not to a differential increase in the expression of the target gene. Probes suitable for use as expression controls correspond to genes expressed constitutively, such as genes encoding proteins which exert essential cell functions such as b-2-microglobulin, ubiquitin, ribosomal protein 18S, cyclophilin A, transferrin receptor, actin, GAPDH, tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein (YWHAZ) and beta-actin.

In the event that the expression levels of the gene according to the present invention is determined by measuring the levels of the polypeptide encoded by said gene, the kits according to the present invention comprise reagents which are capable of specifically binding to said polypeptides. For this purpose, the arrays of antibodies such as those described by De Wildt et al. (2000) Nat. Biotechnol. 18:989-994; Lueking et al. (1999) Anal. Biochem. 270:103-111 ; Ge et al. (2000) Nucleic Acids Res. 28, e3, l-VII; MacBeath and Schreiber (2000) Science 289:1760-1763; WO 01/40803 and WO 99/51773A1 are useful. The antibodies of the array include any immunological agent capable of binding to a ligand with high affinity, including IgG, IgM, IgA, IgD and IgE, as well as molecules similar to antibodies which have an antigen binding site, such as Fab', Fab, F(ab')2, single domain antibodies or DABS, Fv, scFv and the like. The techniques for preparing said antibodies are very well-known for the person skilled in the art and include the methods described by Ausubel et al. (Current Protocols in Molecular Biology, eds. Ausubel et al, John Wiley & Sons (1992)). The antibodies of the array can be applied at high speed, for example, using commercially available robotic systems (for example, those produced by Genetic Microsystems or Biorobotics). The substrate of the array can be nitrocellulose, plastic, crystal or can be of a porous material as for example, acrylamide, agarose or another polymer. In another embodiment, it is possible to use cells producing the specific antibodies for detecting the proteins of the invention by means of their culture in array filters. After the induction of the expression of the antibodies, the latter are immobilized in the filter in the position of the array where the producing cell was located. An array of antibodies can be put into contact with a labeled target and the binding level of the target to the immobilized antibodies can be determined. If the target is not labeled, a sandwich type assay can be used in which a second labeled antibody specific for the polypeptide which binds to the polypeptide which is immobilized in the support is used. The quantification of the amount of polypeptide present in the sample in each point of the array can be stored in a database as an expression profile. The array of antibodies can be produced in duplicate and can be used to compare the binding profiles of two different samples.

It should be noted that, as used in the specification and in the appended claims, the singular forms “a”, “an”, “the”, include their plural referents unless the context clearly indicates otherwise. Similarly, the term "comprises" also includes as a particular embodiment an embodiment in which no additional elements or components are present, that is, includes as an embodiment the term "consists of".

EXAMPLES

The following invention is hereby described by way of the following examples, which are to be construed as merely illustrative and not limitative of the scope of the invention.

Example 1. Materials and Methods

Generation of Ren-PDOX from human specimens ccRCC patient derived orthoxenograft models (Ren-PDOX/Ren) were generated by orthotopic implantation in mouse kidney of fresh human specimens of ccRCC obtained at surgical resection or biopsy from the Bellvitge or Vail d’Hebron Hospitals (Barcelona, Spain) under local ethics committee’s approved protocols (CEIC approvals ref. PR322/11 and PR[AG]240/2013). Patients enrolled provided an informed consent to participate to the study. 56 ccRCC specimens were collected at surgery and immediately implanted in mice left kidney. Ren-PDOX were then maintained and perpetuated in vivo, as previously described. For Ren13, we generated Ren13-PDOX and the paired metastatic variant Ren13BM-PDOX implanting in kidney of mice respectively piece of the human primary ccRCC and metachronous piece of brain metastasis from the same patient.

Treatment schedule

The following antiangiogenic regimens were used: i) anti-mouse VEGFR2 blocking antibody (DC101) intraperitoneally (i.p.) administered, 1 mg/dose/mouse, twice a week. DC101 was collected from a hybridoma culture (ATCC, Manassas, USA); ii) anti-human VEGF monoclonal antibody (Bevacizumab, Avastin, 25mg/ml, Roche Pharma AG, Grenzach-Wyhlen, Germany), i.p., 5 mg/kg/dose, twice a week. The administration started when tumor was palpable. To study tumor local invasion, the duration of treatment was calculated based on the median of control lifespan. To study metastatization, therapies were administered until survival of each mouse.

Immunohistochemistry

Three-micrometer thick of paraffin-embedded tumor or lung tissue sections were rehydrated and processed by standard procedure for HE staining and immunohistochemistry (IHC). The following primary antibodies were used: mouse anti- vimentin (clone V9; Thermo Fisher Scientific, Inc., Cambridge, UK); mouse anti-CD34 (clone QBEND-10, ab8536, Abeam, Cambrige, UK); rabbit anti-CD31 (ab28364, Abeam); rabbit anti-fibronectin (ab2413, Abeam). The EnVision system of labeled polymer-HRP anti-mouse or anti-rabbit IgG (DakoCytomation, Agilent Technologies, Santa Clara, USA) was used as a secondary antibody.

Evaluation of tumor local invasiveness

Evaluation of tumor local invasion was done by ImageJ software tracing the tumor-kidney interface in 4X images of vimentin stained sections taken through all invasive front and evaluating widest extension of tumor protrusion into kidney parenchyma for each image (depth). Then average of depth was calculated for each tumor. Due to the positive correlation found between depth and tumor weight in untreated tumors, we calculated the invasion as depth (pm) normalized for tumor weight (g). To compare different experiments, fold-invasion (D invasion) was calculated as invasion normalized to the average of invasion of control group of each experiment.

High-throughput DNA sequencing

For high-throughput sequencing, DNA was extract from snap frozen Ren 13, Ren28, Ren38, Ren50 and Ren86-PDOX tumor samples using the DNeasy kit (Qiagen, Hilden, Germany) according to manufacturer’s instructions. Next generation sequencing (NGS) was performed at Cancer genomics core Lab at VHIO (Barcelona, Spain). Whole-exome sequencing (WES) was performed in Hiseq 2500 sequencer (lllumina, Inc., San Diego, USA). Somatic variants were called when supported by at least 7 reads, representing at least 7.5% of the total reads. Targeted-NGS was performed in MiSeq sequencer (lllumina, Inc., San Diego, USA) using a custom Haloplex panel (Agilent Technologies, Santa Clara, USA) of 397 genes. Variants were retained with a threshold of depth >200 and variant allele frequency >10%. SNVs and indels characterization was performed using the following public databases: Genome Browser, COSMIC, VARSOME and ICGC Data Portal. Only non-synonymous/stop gain SNVs and frameshift indel novels or with reported allele frequency <1% (MAF < 0.01) were included in Fig.3A.

Massive RNA sequencing

The RNA extraction of tumor specimens was performed using the RNeasy Mini Kit (Qiagen, Hilden, Germany), according to the manufacturer’s recommendations. Samples were sequenced at Centro Nacional de Analisis Genomico (CNAG-CRG, Barcelona, Spain). A modified TruSeq™ Stranded Total RNA kit protocol (lllumina Inc.) was used to prepare the RNA-Seq libraries from samples. Each library was sequenced using TruSeq SBS Kit v3-HS (lllumina), in paired-end mode with a read length of 2x76bp. Image analysis, base calling and base quality scoring of the run were processed by integrated primary analysis software — Real Time Analysis (RTA 1.13.48) and followed by generation of FASTQ sequence files by CASAVA 1.8.

RNA-Seq reads were aligned to the human (GRCh38/hg38) and mouse (GRCm38/mm10) reference genomes using STAR (version 2.5.1b) and GSNAP (version 2015-06-23), respectively, with ENCODE parameters for long RNA. Genes were quantified using RSEM (version 1.2.28) and read counts were used as input for DESeq2 (version 1.10.1). The cut-off for considering a gene significantly up-sampled or down- sampled was FDR<5%. Subsequently, Gene Set Enrichment Analysis (GSEA) was used to detect coordinated expression within samples using default parameters.

Patient Data

Clinical validation of gene expression with clinical variables was performed from two series of ccRCC patients: TCGA-KIRC and GSE29609. (1) In TCGA study, 528 KIRC patients were analyzed by age- and gender-adjusted linear regressions comparing normalized (z-score) gene expression to clinical variables as Tumor Stage (1/11=318,

11 I/I V=205), Tumor Size (T) (1/2=338, 3/4=190), Lymph nodes Invasion (N) (0=254, 1=238) and Metastasis (M) (0=416, 1=78); to obtain their fold change (computed using foldchange method at gtools R library), gene expression beta regression coefficients and nominal and FDR-adjusted p values. (2) In the GSE29609 study, 39 ccRCC patients were evaluated and normalized Iog10 ratio Cy5/Cy3 gene expression values were analyzed in patients with and without Renal Vein Involvement (RV; 0=23, 1=16) or Microvascular Invasion (MVI; 0=16, 1=23) by Unpaired Fisher’s exact test.

Statistical analysis and data representation

Results are presented as box plots (min to max values and median). Depth and tumor weight correlation was assessed by Spearman test (2-tailed). Statistical comparison between control and treated mice was done by Mann-Whitney T test (2-tailed) or Chi- square test. Differences were considered statistically significant at p<0.05.

Example 2. Ren-PDOX models reproduce architectural features and metastatic potential of their original patient’s tumor

In order to evaluate the impact of antiangiogenic drugs in clinically relevant models, we generated Ren-PDOXs by direct implantation of human tumor specimens from patient into mouse kidney (Fig. 1A). We collected and implanted a total of 56 ccRCC and we obtained 27 growing PDOX corresponding to 48.2% of engraftment (data not shown). During the in vivo passage, Ren-PDOX maintained the same structural features of original human specimen, distinctive of each tumor (Fig. 1 B). As expected, by CD31 and CD34 IHC staining as species-specific markers of murine and human blood vessels respectively, we found that human stroma components were lost in PDOXs and substituted by murine stroma from the first orthotopic in vivo growth (Fig. 1C). This was confirmed by species-specific TaqMan analysis of murine and human VEGF-A and VEGFR2 where only human VEGF-A ligand and only mouse VEGFR2 were detected.

Tumor harvested from metastatic patients (including positive regional lymph node and/or synchronous or metachronous metastasis) exhibited significantly higher engraftment rate compared to patients that didn’t present metastasis (pNO and pMO), generating PDOX respectively in 88.9% and 28.1% of implanted tumors (p<0.0001 by Chi-square test). Moreover, we found that 38.5% of growing Ren-PDOX developed lung metastases visible as macrometastases at sacrifice and/or micrometastases detectable by HE and vimentin staining of lung sections (Fig. 1A and 1 B). Interestingly, tumors from metastatic patients showed high metastatic capacity in mice and indeed 80% of metastatic Ren- PDOX derived from metastatic patients (Fig. 1A, graph). Altogether, this suggests that more aggressive and dysplastic tumors maintain the characteristics of malignancy acquired in patients and are more prone to seed and to adapt to different microenvironment for growing.

Studies from genomic consortiums such as TCGA have demonstrated that ccRCC tumors are characterized by prototypical mutations in some common genes such as VHL, PBRM1, BAP1 , SETD2 and TSC1 among others. Thus, we wondered whether Ren-PDOXs molecularly reproduced RCC typical genetic alterations. To this aim, we performed targeted-Next Generation Sequencing (NGS) analysis in DNA extracted from Ren13, Ren28, Ren38, Ren50, Ren86-PDOX tumors using a panel of 397 genes that included all the most common mutated genes in RCC. Collectively, Ren-PDOXs tumors presented 25 different alterations (15 already reported, 10 novel), distributed over 14 genes of Haloplex panel. We found potentially pathogenic truncating or frameshift mutation in VHL and PBRM1 genes that control processes frequently altered in ccRCC. In particular, three out of five Ren-PDOXs had point alterations in VHL, all already reported in public database, while alterations identified for PBRM1 gene were all novel. Other potentially pathogenic mutations were found in ATM, BAP1, MLL3, mTOR, SETD2 and TSC1 genes (Fig. 1 D), fully concordant with previously published TCGA mutation profile for clear cell kidney cancer.

Furthermore, in order to determine whether individual PDOX were reproducing the genomic alterations of their original patient’s tumor, we performed a Whole Exome Sequencing (WES) analysis to compare the genomic profile of Ren50-PDOX and its original human tumor specimen. The total number of variants detected was comparable between original human specimen and Ren50-PDOX sample (399 and 355 respectively). Only 9/355 variants were exclusive of PDOX tumor, therefore 97.5% of Ren-PDOX variants were already present in original human tumor (Fig. 1 E). Overall, the 408 variants identified by WES were homogeneously distributed over the human karyotype. Taken together, these results demonstrate that Ren-PDOXs share genomic characteristics with human tumor reproducing human RCC heterogeneity and complexity, key features for their use as relevant predictive models of human cancer.

Example 3. Ren-PDOXs respond to antiangiogenic treatments with varying degrees of efficacy in a patient-specific manner

In order to test antiangiogenic efficacy in our newly developed PDOX models, DC and Beva were administered to mice implanted with four different Ren-PDOX tumors: Ren13, Ren28, Ren50 and Ren86. In all tested Ren-PDOXs, short-term antiangiogenic therapies slowed down tumor growth with tumor weight significantly lower in treated compared to control condition (Fig. 2A-B). Indeed, DC and Beva treatments were effective in diminishing vessel number and increasing tumor necrosis (Fig. 2C-D) as the causal mechanism of their antitumor effects. Nevertheless, differential effects were observed regarding OS in these treatments: while DC and Beva treatments extended survival in Ren28 and Ren50 tumors, this effect was not observed in Ren13 and Ren86 PDOX models (Fig. 2E). These results demonstrate differential responses to antiangiogenics in different PDOX models that recapitulates the differential responses observed in kidney cancer patients treated with antiangiogenics.

Since pro-invasive effects have been described to be one of the consequences of tumor adaptation to antiangiogenics, we aimed at evaluating tumor invasion at the tumor-kidney interface. Indeed, Ren-PDOX masses grew bound to kidney establishing a tumor-kidney interface at the IF. In HE sections we observed that the interface presented a fibrous capsule. Screening by IHC staining, we found that the fibrous capsule, included renal capsule, was composed by fibronectin (FN) fibers. Moreover, we could identify different grade of capsular invasion (Cl). Thus, we distinguished tumors whose Cl was i) absent if tumor cells were separated from normal renal parenchyma by well-defined layers of FN or ii) present if strands of tumor cells were directly in contact with normal renal parenchyma (Fig. 3A). In particular, untreated Ren13 and Ren86 were characterized by low basal Cl (around 30% of tumors) whereas 60% of Ren28 and Ren50 tumors presented Cl (Fig. 3B, control).

Since it has been recently suggested that antiangiogenic therapy could modify RCC tumor-normal kidney parenchymal interface, we wondered if treatment could exert comparable effects on invasive front of Ren-PDOX tumors. When the four PDOXs were treated with DC and Beva, both treatments exerted differential effects on tumor invasive capacity in each PDOX model. In Ren13 and Ren86, DC and Beva treatment produced significantly more Cl than controls (Fig. 3B) resulting in >2-fold higher local invasion (Fig. 3C). In contrast, in Ren28 and Ren50 the percentage of tumor masses showing Cl was similar in control and treated animals (Fig. 3B), therefore VEGF/R blockade did not modify tumor invasive behavior of this other subset of tumors (Fig. 3C). Therefore, our findings show that VEGF-targeted therapy impaired tumor growth in all PDOX tested, but only increased the invasive capacity of a specific subset of ccRCC tumors, indicating a prominent patient-specificity of this pro-invasive response.

Example 4. Antiangiogenics promote patient-specific variable responses in systemic dissemination and metastasis We then wondered whether contrasting effects on local invasion could have different consequences on development of metastasis in Ren-PDOXs. The use of spontaneous metastatic models to assess change in metastatic dissemination is challenging. Thus, we compared the effects of VEGF/R inhibition on two independent Ren-PDOXs showing high basal metastatic capacity but opposite invasive behavior upon treatments (Ren13BM and Ren28) (Fig. 4A). Interestingly, we found that in Ren13BM the pro- invasive effects of VEGFR-targeted therapies on local invasion (data not shown) was associated with augmentation of lung metastasis and indeed incidence of metastasis as well as number of metastatic foci significantly increased in DC-treated compared to control mice (Fig. 4B-C, left). We also observed higher metastatic dissemination in Beva- treated mice, although not significantly, probably due to the limited sample size (Fig. 4B- C, left). On the other hand, in Ren28, antiangiogenic treatments did not modify incidence of metastases in treated mice compared to controls, nor number of metastatic foci per mouse (Fig. 4B-C, right), in accordance with unmodified local invasion. Taken together, our observations suggest that inhibition of VEGF signaling pathway could have different inter-tumor impact therefore resulting in contrasting local and subsequently systemic effects (Fig. 4D).

Example 5. Tumor cell-specific expression profile defines molecular hallmarks of pro- invasive behavior

In order to get a mechanistic insight into the molecular basis of tumor predisposition to acquire an invasive behavior upon antiangiogenic therapies, we performed a novel RNA sequencing (RNAseq) technique that allows for discrimination of tumor and stromal expression profiles. Briefly, whole-genome RNAseq of a pro-invasive PDOX model after antiangiogenics (Ren13) and a non-pro-invasive PDOX (Ren50) in basal untreated condition was performed where human and murine transcripts can be unequivocally distinguished, allowing respectively the analysis of tumoral and stromal contribution to the differential response to antiangiogenics (Fig. 5A). Results from tumor-derived (human origin) comparative expression showed that most genes were present in both tumor types (85.1%, Fig. 5A). However, a fraction of the transcripts were expressed exclusively either in Ren13 (6.7%) or in Ren50 tumors (8.2%, Fig. 5A). As expected, stromal (murine origin) gene expression overlap between Ren 13 and Ren50 was even higher (94.7%, Fig. 5A), even though a small proportion of stromal genes were found specific for each tumor type (1.6% and 3.7% for Ren13 and Ren50, respectively, Fig. 5A). These results demonstrate the intrinsic genetic differences from these two Ren- PDOX tumors mainly due to tumoral contribution.

To identify a signature of molecular pathways involved in divergent invasive behavior upon antiangiogenic treatment, gene set enrichment analysis (GSEA) was performed from RNAseq expression data of the tumor cell component (human origin) (Fig. 5B). GSEA analysis identified UV response down, KRAS signaling up, coagulation, angiogenesis and Wnt^-catenin signaling as the most enriched pathways in pro-invasive tumors (Fig. 5B). On the other hand, non-pro-invasive tumors were enriched in gene sets involving E2F targets, G2/M checkpoint, estrogen response late, epithelial mesenchymal transition (EMT), KRAS signaling down, reactive oxygen species (ROS) pathway and IL2/STAT5 signaling (Fig. 5B). Remarkably, pro-invasive tumors showed increased KRAS signaling in comparison to non-pro-invasive (more expression of genes up- regulated by KRAS and less of genes down-regulated by KRAS), suggesting that this pathway may be an important axis implicated in the tumor predisposition to acquire a more malignant phenotype after antiangiogenic therapy.

In order to filter the KRAS signature genes to obtain a smaller more specific geneset, an extensive series of clear cell kidney cancer patients from TCGA was used. The possible association between candidate normalized (z-score) gene expression and clinical variables in primary TCGA-KIRC tumor samples was studied by age- and gender- adjusted linear regressions. Guided by both Tumor Stage (S), Tumor Size (T) and Metastasis (M), only 19 genes were found differentially expressed (FDR-adjusted p- value <0.05) in the more advanced/aggressive subset of all these clinical characteristics (Fig. 5C). Validation of filtered candidates using a unique series of 39 RCC patients where microvascular invasion and renal vein involvement were specifically evaluated. Candidate signature genes differential expression analysis was performed in the GSE29609 patient series as a validation dataset, evaluating renal vein involvement and microvascular invasion (Fig. 6A). As shown in Fig. 6B, only two of the 19 selected candidate genes, namely ALDH1A3 and MAP7, were robustly differentially expressed in tumors with microvascular invasion or with renal vein involvement. Specifically, ALDH1A3 was strongly upregulated while MAP7 was robustly down-regulated in invasive tumors to the microvasculature or further implicating the renal vein (Fig. 6C). Therefore, these data suggest that high ALDH1A3 and low MAP7 could be used to define the tumor predisposition to acquire aggressive/invasive capacity in RCC patients.

Example 6. Validation of ALDH1A3 as marker in patients treated with antiangiogenics The expression of both candidate biomarkers was validated in PDOX preclinical models and patient samples. Results showed that ALDH13A was over-expressed in tumors with pro-invasive characteristics from preclinical PDOX models (Fig. 7A-C) and also in patient samples (Fig.7D). Moreover, the receiver operating characteristic (ROC) curve was statistically significant (p=0.05) which defines ALDH1A3 as a good predictor of pro- invasive signature in RCC tumors (Fig 7E). Additionally, there is a significant correlation between ALDH1 A3 expression levels and pro-invasive signature in RCC tumors (P Value < 0.0095) (Fig. 7F). Therefore, these results demonstrate that ALDH1A3 up-regulation can predict early tumor malignization in response to anti-angiogenic therapies. In contrast, the candidate gene MAP7 seems to not be so involved in the process of predicting the malignant behavior of a tumor in response to antiangiogenic therapies, as neither association nor correlation to invasion was found (Fig 8A-G).

Example 7. ALDH13A Expression in mRCC PDOX model after sunitinib treatment

In order to test whether these pro-invasive effects were also observed when a clinically- approved antiangiogenic was administered to our Ren-PDOX, we treated a pro-invasive (Ren13BM) and a non-pro-invasive (Ren28) models with sunitinib and evaluated invasion. As shown in Figure 9, sunitinib treatment increased both Cl (bar graphs) and tumor local invasion (box plots) in Ren13BM but not in Ren28. Furthermore, sunitinib pro-invasive analysis was extended to a new set of 5 PDOXs that also showed patient- specific differences in capsular and local tumor invasion. In this set, 3 out of 5 PDOXs show enhanced tumor invasion (Figure 10). Thus, this suggests that sunitinib also produces increased invasive capacity of a specific subset of ccRCC tumors, indicating a prominent patient-specificity of this pro-invasive response.

Example 8. Validation of ALDH1A3 as a tumor-specific biomarker of pro-invasive behavior after sunitinib treatment

Results confirmed that ALDH1A3 was significantly associated with pro-invasive phenotype in these samples (Fig. 11 A) and its protein levels significantly correlated to invasion (Fig. 11 B).

Example 9. Validation of ALDH1A3 in human samples

Series of patients for prediction of response were gathered in a clinical study approved by local ethics committee (ref. HUVH-PR(AG)240/2013). 15 patients were included and gender, treatment regimen (Sunitinib, Pazopanib and Bevacizumab) and type of progression information were gathered for this study, together with tumor specimen from either nephrectomy or tumorectomy prior to treatment. Clinical data and samples were handled following ethical procedures under local ethics committee approval. In order to test the predictive clinical value of ALDH1A3, we evaluated a series of 15 patients clinically treated with sunitinib whose pro-aggressive response after therapy was fully annotated. As shown in Fig 12A, patient’s response to therapy was categorized as disease progression with new metastasis/ overt local infiltration, or disease progression with no new lesions/ local progression. Tumor specimen analysis before treatment in this series confirmed that ALDH1A3 can significantly discriminate pro-aggressive response in patients, showing not only association but also a linear trend of ALDH1A3 levels to pro-aggressive response (Fig 12B). Furthermore, the predictive power of ALDH1A3 was evaluated using a ROC curve study with very significant results and defining its initial sensitivity and specificity (Fig 12C). Overall, our data clearly defines ALDH1A3 as a possible predictive factor of pro-aggressiveness and could be used to predetermine tumor predisposition to acquire aggressive/invasive capacity in RCC patients treated with antiangiogenic therapies.

Example 10: Correlation between ALDH13A expression levels and overall survival of cancer patients in different types of tumors.

Clinical validation of gene expression with clinical variable was performed from seven series of patients with Colon adenocarcinoma, Bladder Urothelial Carcinoma, Esophageal Carcinoma, Kidney Renal Clear Cell Carcinoma, Low Grade Glioma, ovarian serous cystadenocarcinoma and stomach adenocarcinoma (Atlas, 2013- 2021). In TCGA studies, patients were analyzed by comparing normalized (z-score) gene expression to the clinical variable overall survival to obtain their Kaplan-Meier survival plot p values.

Patients with high ALDH13A levels present worst risk of progression (statistical significance(p-value=<0.050) in Colon adenocarcinoma, Bladder Urothelial Carcinoma, Esophageal Carcinoma, Kidney Renal Clear Cell Carcinoma, Low Grade Glioma, ovarian serous cystadenocarcinoma and stomach adenocarcinoma.

Claims

1. A method for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer, wherein the method comprises

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1A3 that is increased with respect to the reference value is indicative of an increased risk of local invasion and/or metastasis in response to the antiangiogenic treatment.

2. A method for predicting the response to an antiangiogenic therapy of a patient who suffers from cancer, wherein the method comprises (i) determining the expression level of ALDH1A3 in a sample from said patient, and

(ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1A3 that is increased with respect to the reference value is indicative that the patient is at an increased risk of developing local invasion and/or metastasis in response to the antiangiogenic treatment.

3. A method for selecting a patient suffering from cancer for treatment with an antiangiogenic therapy, wherein the method comprises

4. A method for selecting a therapy for a patient suffering from cancer, wherein the method comprises

(i) determining the expression level of ALDH1A3 in a sample from said patient, and (ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1A3 that is decreased with respect to the reference value is indicative that an antiangiogenic therapy is selected for the patient.

5. An antiangiogenic compound for use in the treatment of a patient suffering from cancer, wherein the patient is selected using a method that comprises (i) determining the expression level of ALDH1A3 in a sample from said patient, and (ii) comparing the level obtained in (i) to a reference value, wherein an expression level of ALDH1A3 that is decreased with respect to the reference value is indicative that the patient will benefit from treatment with an antiangiogenic therapy.

6. The method according to any one of claims 1 to 4, or the antiangiogenic compound for use according to claim 5, wherein the patient suffers from colon cancer, bladder cancer, esophageal cancer, glioma, ovarian cancer or stomach cancer.

7. The method according to claim 6 wherein the colon cancer is colon adenocarcinoma, the bladder cancer is bladder urothelial carcinoma, the esophageal cancer is esophageal carcinoma, the glioma is low grade glioma, the ovarian cancer is ovarian serous cystadenocarcinoma or the stomach cancer is stomach adenocarcinoma.

8. The method according to any one of claims 1 to 4, or the antiangiogenic compound for use according to claim 5, wherein the patient suffers from renal cancer.

9. The method according to claim 8, or the antiangiogenic compound for the use according to claim 8, wherein the renal cancer is renal cell carcinoma (RCC), preferably clear cell renal cell carcinoma (ccRCC).

10. The method according to any one of claims 1-4 or 6-9, or the antiangiogenic compound for the use according to any one of claims 5-9, wherein the step of determining the expression level of ALDH1A3 in the sample from the patient is performed prior to initiating an antiangiogenic treatment.

11. The method according to any one of claims 1-4 or 6-10, or the antiangiogenic compound for the use according to any one of claims 5-10, wherein the antiangiogenic treatment is administered as a primary treatment, or administered as an adjuvant or neoadjuvant accompanying another treatment, preferably a surgical treatment.

12. The method according to any one of claims 1-4 or 6-11, or the antiangiogenic compound for the use according to any one of claims 5-11 , wherein the reference value is the mean level of expression of ALDH1A3 in a pool of samples from healthy tissue, or wherein the reference value is the mean level of expression of ALDH1A3 in a pool of samples from primary tumors.

13. The method according to any one of claims 1-4, or 6-12, or the antiangiogenic compound for the use according to any one of claims 5-12, wherein the sample is a sample containing tumor cells, preferably wherein the sample containing tumor cells is tumor tissue.

14. The method according to any one of claims 1-4, or 6-13, or the antiangiogenic compound for the use according to any one of claims 5-13, wherein the determination of the expression levels of the gene is carried out by determining the level of the corresponding mRNAs or by determining the level of the polypeptide encoded by said gene.

15. Use of ALDH1A3 for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer.

16. Use of reagents specific for the determination of the expression level of ALDH1A3, for predicting the risk of local invasion and/or metastasis induced by an antiangiogenic treatment in a patient who suffers from cancer.

17. The use according to claim 16, wherein the reagents are selected from the group of a set probes which specifically hybridize to the mRNA of said gene and a set of primer pairs which are capable of specifically amplifying the mRNA of said gene, or wherein the reagents are a set of antibodies which specifically bind to the polypeptide encoded by said gene.