WO2019068761A1 - Methods for identifying cancer patients at high risk of developing metastasis - Google Patents

Abstract

Methods for identifying cancer patients at high risk of developing metastasis. The present invention relates to the field of biomedicine and cancer. Specifically, it relates to an in vitro method of prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease based on the differential presence of exons in circulating cell free DNA (cfDNA) in cancer patients.

Description

Methods for identifying cancer patients at high risk of developing metastasis Field of the invention

The present invention relates to the field of biomedicine and cancer. Specifically, it relates to an in vitro method of prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease based on the differential presence of exons in circulating cell free DNA (cfDNA) in cancer patients.

Background of the Invention

In solid malignancies, metastases rather than primary tumors are responsible for most cancer deaths. To prevent these deaths, improved ways of diagnosis, prognosis and treatment of metastatic disease are needed.

Circulating cell-free DNA (cfDNA) has emerged as a non-invasive alternative to conventional serial tissue biopsying for the analysis of tumor molecular features (Heitzer E. et al. Clin Chem. 2015, 61 (1 ), 1 12-23). Increasing evidences support the potential clinical utility of this approach, known as liquid biopsy, in colorectal cancer (CRC), particularly at advanced stages (Toledo RA. et al. Oncotarget. 2016, 8(21 ), 35289-35300). Several studies have demonstrated that a combination of both elevated cfDNA levels and a heterogeneous pattern of hotspot mutational status (including KRAS, NRAS, BRAF and EGFR genes, among others) provide a strong predictor of clinical prognostic value (Bettegowda C, et al. Sci Transl Med. 2014, 6(224), 224ra24; Misale S et al. Cancer Discov. 2014, 4(1 1 ), 1269-80; Spindler KL, PLoS One. 2015, 10(4), e0108247; Siravegna G, et al. Nat Med. 2015, 21 (7), 827). Thus, cfDNA-based surveillance is able to anticipate disease progression months ahead of standard imaging follow- up (Misale S et al. Nature, 2012, 486(7404), 532-6; Reinert T. et al. Gut. 2016, 65(4), 625-34). In line with these results, cfDNA has been used in a very recent study to reflect tumor molecular dynamics in drug response of metastatic CRC patients, tracking the evolution of resistance mutations in KRAS pathway genes at different time points along treatment with anti-EGFR therapy (Toledo RA. et al. Oncotarget. 2016, 8(21 ), 35289-35300).

Circulating tumor DNA (ctDNA) fragments represent a minor proportion of the total cfDNA and, therefore, require extremely sensitive and specific detection techniques. Among them, next- generation sequencing (NGS) has gained an increasing interest in the last few years. Most publications focused on NGS for liquid biopsy report targeted sequencing approaches, aimed at analyzing panels of known genes potentially relevant for clinical management in tumor-derived DNA from different kinds of cancers, including CRC (Schwaederle M et al. Oncotarget. 2016, 7(9), 9707-17; Lebofsky R et al. Mol Oncol. 2015, 9(4), 783-90). NGS of cfDNA has been recently applied to CRC in search of serial changes of mutational profiles and tumor load fluctuations for early detection of recurrence (Kim ST et al. Oncotarget. 2015, 6(37), 40360-9; Zhou J et al. PLoS One. 2016, 1 1 (7), e0159708; Sakai K et al. PLoS One. 2015, 10(5), e0121891 ; Tie J et al. Oncol. 2015, 26(8), 1715-22). However, it must be noted that sometimes tumors shed an insufficient amount of DNA to be analyzed, especially, but not exclusively, at early stages of disease (Kim ST, Lee WS et al. Oncotarget. 2015, 6(37), 40360-9, Heitzer E et al. Int J Cancer. 2013, 133(2), 346-56). Clinical sensitivity seems to be significantly affected by the surgical excision of primary tumors, by mutational heterogeneity and by tumor burden, rendering the analysis of mutations inefficient in cancer patients with low levels of ctDNA (Rachiglio AM et al. Oncotarget. 2016, 7(41 ), 66595-66605). Whole-genome sequencing has also been used to search for chromosomal alterations, including copy number changes and amplifications of cancer driver genes in cfDNA of CRC patients (Heitzer E et al. Int J Cancer. 2013, 133(2), 346-56), Leary RJ et al. Sci Transl Med. 2012, 4(162), 162ra54, Mohan S. PLoS Genet. 2014, 10(3), e1004271 ). For example, MET amplification has been very recently detected by exome-sequencing in plasma of patients refractory to anti-EGFR therapy (Raghav K et al. Oncotarget. 2016, 7(34), 54627-54631 ).

The use of exome-sequencing to track specific mutations is impaired by several factors. One of the major hurdles is that the sensitivity of mutation detection is severely affected by the concentration of cfDNA in plasma, the background noise rate, the relative abundance of ctDNA and the capture efficiency (Klevebring D et al. PLoS One. 2014, 9(8), e104417). These approaches usually require sequencing at a high depth, which considerably increases costs, and even at a very high read depth, mutations present at extremely low levels could still be undistinguishable from the sequencing background (Calvez-Kelm FL et al. Oncotarget. 2016, 8(1 1 ), 18166-18176). Thus, the clinical validity of mutational studies by NGS in plasma of patients with low-shedding tumors could be limited. Moreover, the use of whole-genome and/or exome-sequencing for the assessment of panels of tumor mutations as clinical biomarkers may not be cost-effective (Sato KA et al. PLoS One. 2016, 1 1 (1 ), e0146275). Accordingly, despite significant advances having been made, there is still a need to find an easy, fast, non-invasive, cost-effective and affordable strategy for the identification of cancer patients at high risk of developing metastasis for use in the early and accurate differential diagnosis, classification and treatment of these patients, as well as the prediction of the progression, prognosis and monitoring of the cancer disease. Summary of the Invention

Tumor genotyping during disease follow-up is gaining an increasing interest for monitoring and clinical management in colorectal cancer (CRC). Growing evidences suggest that single biopsies are inefficient for molecular profiling, due to clonal evolution and intratumoral heterogeneity of primary tumors and/or metastasis. Circulating cell-free DNA (cfDNA) has emerged as an alternative source of genetic material that seems to be representative of the continuously changing tumor molecular features. Thus, the so-called 'liquid biopsy' also offers the additional advantages of accessibility and non-invasiveness for the patient. Novel strategies based on next-generation sequencing (NGS) of cfDNA are being currently developed, with most studies focusing on targeted deep sequencing panels of potential clinically actionable genes.

Instead of searching for known mutations by targeted deep sequencing, which is the most common strategy, but could be unnecessarily costly and have limited potential for minimally DNA-shedding tumors, the inventors have devised a method based on exome-sequencing performed at a relatively shallower depth to gain a more general overview of the circulating DNA in plasma, including both known and unknown cancer-related characteristics. The aim of this study has been to compare differential features in terms of cfDNA other than SNPs. This represents a new strategy that broadens the scope of NGS applications in liquid biopsy, reducing costs and making it more feasible for translation to the clinical scenarios. Moreover, the differential features detected with this strategy could probably come both from tumor and non-tumor derived cfDNA, thus differing from previous almost exclusively focused on cancer- related genetic alterations.

In particular, the present work has studied two groups of colorectal cancer patients, with disseminated (M) or localized disease (N), using whole-exome sequencing and a bioinformatics analysis pipeline. A set of 379 exons present at different levels in cfDNA from both groups of colorectal cancer patients was identified, giving rise to a completely new concept in this field, termed here as 'Differential Presence of Exons' (DPE). Moreover, DPE was subsequently used for M and N patients clustering and classification, allowing for the design of a predictive 'DPE algorithm' with encouraging preliminary results. Indeed, Figures 4 and 5 show that normalized quantification values of differentially present exons enabled to accurately distinguish and classify patients with metastasis and localized disease.

Moreover, the identified set of exons was successfully used to classify a small subset of patients that could not be initially classified attending to the selection criteria established in Table 2. Indeed, every member of the U group, namely those with locally advanced disease (pT4) or affected nodes (pN1 -N2) in the absence of distant metastasis, were correctly classified by the DPE algorithm as belonging to the M group.

Accordingly, in a first aspect, the invention relates to an in vitro method for identifying exons which are differentially present in patients having metastasis (DPE) with respect to patients having localized cancer disease, wherein said method comprises the following steps:

i. quantifying individual exons by whole-exome sequencing of the circulating cell-free DNA (cfDNA) by next-generation sequencing in a blood, plasma or serum sample obtained from a cancer patient or group of patients with localized disease (N group); ii. quantifying individual exons by whole-exome sequencing of the cfDNA by next- generation sequencing in a blood, plasma or serum sample obtained from a cancer patient or group of patients with metastasized disease (M group);

iii. comparing the quantification values in i) with the quantification values in ii) and identify as differentially present exons those which abundance differs in a statistically significant manner between both groups.

In a preferred embodiment of the invention, the exome capturing for whole-exome sequencing is conducted by a method comprising hybridization of cfDNA in said blood, plasma or serum sample with probes substantially complementary to substantially all the coding DNA sequences in the patient's species genome. In another preferred embodiment of the invention, the exon quantification is conducted by counting the number of sequence reads obtained by whole- exome sequencing that map to known exons when these are aligned to a reference genome sequence. In another preferred embodiment of the invention, the quantification values of DPE have been normalized, preferably data normalization has been conducted by the trimmed mean of M values (TMM) method.

The second embodiment of the present invention refers to an in vitro method of prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease, the method comprising:

i. quantifying individual exons identified as differentially present in metastatic patients (DPE) according to a method as defined above in a blood, plasma or serum sample obtained from said patient; and

ii. comparing the quantification values in the patient's sample obtained in i) with the quantification values in a reference sample, wherein said reference sample is isolated from a patient or group of patients suffering from localized cancer disease; wherein when the quantification values in the patient's sample are increased or decreased in comparison with those in the reference sample the patient has a high risk of suffering from metastasis. In a preferred embodiment of the invention, DPE are quantified by a method selected from the group consisting of next generation sequencing, quantitative PCR (qPCR), PCR- pyrosequencing, PCR-ELISA, DNA microarrays, branched DNA, dot-blot, Fluorescence In Situ Hybridization assay (FISH), and multiplex versions of said methods. In another preferred embodiment of the invention, DPE are quantified by whole-exome sequencing using next generation sequencing. In another preferred embodiment of the invention, said patient is a human patient. In another preferred embodiment of the invention, said cancer is one characterized by presence of circulating tumor DNA (ctDNA) in the blood. In another preferred embodiment of the invention said cancer is colorectal cancer, preferably adenocarcinoma. In another preferred embodiment of the invention said differentially present exons are selected from the group consisting of the 379 exons found to be differentially present in colorectal cancer defined in Table 1 .

The third embodiment of the invention refers to an in vitro method for selecting a treatment for a patient having localized cancer disease wherein said method comprises selecting a treatment according to the classification of said patient according to its prognosis or risk of metastasis by a method as defined above. In a preferred embodiment of the invention, the method further comprises storing the method results in a data carrier, preferably wherein said data carrier is a computer readable medium. The fourth embodiment of the invention refers to a computer implemented method, wherein the method is as defined above.

The fifth embodiment of the present invention refers to the use of a kit for the prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease on the basis of the differential presence of exons in circulating cell-free DNA (cfDNA) according to a method as defined above, said kit comprising:

a) a reagent for quantifying individual exons in a blood, plasma or serum sample;

b) optionally, instructions for the use of said reagent(s) in determining the levels of said exons, in a blood, plasma or serum sample. Brief Description of the Figures

Figure 1. (A) Box Plot of DNA concentration in plasma of colorectal cancer patients. Metastatic patients (M) showed a higher median concentration of cell-free DNA (cfDNA) in plasma than non-metastatic ones (N). The distribution of cfDNA concentration in unclassifiable patients (U) shares common characteristics with both groups. (B) Size distribution of a cfDNA library from a patient showing a nucleosomal laddering pattern with fragment sizes of 302, 472 and 641 bp, including adapter sequences.

Figure 2. Schematic workflow of the experimental procedure performed. Cell-free DNA (cfDNA) was isolated from plasma of colorectal cancer (CRC) patients. Exome capture was performed before sequencing and the resulting reads were subsequently aligned to the reference genome sequence (hg38). In this study, a pipeline for NGS data analysis was applied in cfDNA from CRC patients.

Figure 3. MA plots for selected Differentially Present Exons (DPE) (pv<0.005). The log ratio of fold-change (FC) is plotted on the y-axis, called M (from "minus") while in the x-axis the average of the normalized counts (Counts Per Million) are represented, called A values (from "average"). A total number of 379 exons were obtained with EdgeR combining two different methods: (A) Likelihood Ratio Tests (LRT) with 297 exons and (B) Quasi-Likelihood F-tests (QLF) with 366 exons. Overpresent exons in metastatic (M) and non-metastatic (N) groups are represented with A and■, respectively. Figure 4. Patient clusterization using normalized quantification values of Differentially Present Exons (DPE) for Ward's Method. Metastatic (M) and non-metastatic (N) patients are clearly separated in two groups, whereas unclassifiable patients (U) are located in between, sharing common traits with both groups. Recurrence-free patients after the two year follow-up period are marked with an asterisk (^*) and, remarkably, tend to group together. (!) High-risk patients treated by HIPEC: hyperthermic intraperitoneal chemotherapy; (#) Correctly predicted metastasis; (+) Exitus.

Figure 5. Bidimensional Principal Components Analysis (PCA) plot. Metastatic (M) and non- metastatic (N) patients are properly clustered and clearly separated. Unclassifiable patients (U) form another group between the limits of M and N, probably due to their intermediate characteristics. (!) High-risk patients subjected to prophylactic treatment (HIPEC, hyperthermic intraperitoneal chemotherapy); (#) Correctly predicted metastasis; (+) Exitus. Detailed Description

Definitions

The term "cancer patient" and "subject suffering from cancer" are used herein interchangeably. It may refer to those subjects diagnosed after a confirmatory test (e.g., biopsy and/or histology) and subjects suspected of having cancer. The term "subject suspected of having cancer" as used herein, refers to a subject that presents one or more signs or symptoms indicative of a cancer and is being screened for cancer. A subject suspected of having cancer encompasses for instance an individual who has received a preliminary diagnosis (e.g., an X-ray computed tomography scan showing a mass) but for whom a confirmatory test (e.g., biopsy and/or histology) has not been done or for whom the stage of cancer is not known. The term further includes individuals in remission.

The terms "subject", or "individual"' are used herein interchangeably to refer to all the animals classified as mammals and includes but is not limited to domestic and farm animals, primates and humans, for example, human beings, non-human primates, cows, horses, pigs, sheep, goats, dogs, cats, or rodents. Preferably, the subject is a male or female human being of any age or race.

The term "therapeutically effective amount" as used herein refers to an amount that is effective, upon single or multiple dose administration to a subject (such as a human patient) in the prophylactic or therapeutic treatment of a disease, disorder or pathological condition.

The term "probe" as used herein refers to synthetic or biologically produced nucleic acids, between 10 and 285 base pairs in length which contain specific nucleotide sequences that allow specific and preferential hybridization under predetermined conditions to target nucleic acid sequences, and optionally contain a moiety for detection or for enhancing assay performance. A minimum of ten nucleotides is generally necessary in order to statistically obtain specificity and to form stable hybridization products, and a maximum of 285 nucleotides generally represents an upper limit for length in which reaction parameters can be easily adjusted to determine mismatched sequences and preferential hybridization. Probes may optionally contain certain constituents that contribute to their proper or optimal functioning under certain assay conditions. For example, probes may be modified to improve their resistance to nuclease degradation (e.g., by end capping), to carry detection ligands (e.g., fluorescein), to carry ligands for purification or enrichment purposes (e.g. biotin) or to facilitate their capture onto a solid support (e.g., poly- deoxyadenosine "tails"). The term "specific" as used herein in connection with a nucleotide sequence means that a nucleotide sequence will hybridize to/amplify a predetermined target sequence and will not substantially hybridize to/amplify a non-target sequence under the assay conditions, generally stringent conditions are used.

The term "hybridization" as used herein refers to a process by which, under predetermined reaction conditions, two partially or completely complementary strands of nucleic acid are allowed to come together in an antiparallel fashion to form a double-stranded nucleic acid with specific and stable hydrogen bonds, following explicit rules pertaining to which nucleic acid bases may pair with one another.

The term "substantial hybridization" means that the amount of hybridization observed will be such that one observing the results would consider the result positive with respect to hybridization data in positive and negative controls. Data which is considered "background noise" is not substantial hybridization.

The term "stringent hybridization conditions" means approximately 35°C to 65°C in a salt solution of approximately 0.9 molar NaCI. Stringency may also be governed by such reaction parameters as the concentration and type of ionic species present in the hybridization solution, the types and concentrations of denaturing agents present, and the temperature of hybridization. Generally as hybridization conditions become more stringent, longer probes are preferred if stable hybrids are to be formed. As a rule, the stringency of the conditions under which hybridization is to take place will dictate certain characteristics of the preferred probes to be employed.

Methods of the Invention

In a first aspect, the invention relates to an in vitro method for identifying exons which are differentially present in patients having metastasis (DPE) with respect to patients having localized cancer disease, wherein said method comprises the following steps: i. quantifying individual exons by whole-exome sequencing of the circulating cell-free DNA (cfDNA), preferably by next-generation sequencing, in a biological fluid sample (e.g. blood, plasma or serum sample) obtained from a cancer patient or group of patients with localized disease (N group);

ii. quantifying individual exons by whole-exome sequencing of the cfDNA, preferably by next-generation sequencing, in a biological fluid sample (e.g. blood, plasma or serum sample) obtained from a cancer patient or group of patients with metastasized disease (M group);

iii. comparing the quantification values in i) with the quantification values in ii) and identify as differentially present exons those which abundance differs in a statistically significant manner between both groups.

The term "DPE" as used herein refers to exons which are differentially present in patients having metastasis with respect to patients having localized cancer disease, i.e, exons which levels have been found to significantly differ between a patient group with localized disease with respect to a patient group with metastasis. A method for identifying DPE has been defined under the first aspect of the invention.

The term "exon" as used herein refers to any part of a gene that will encode a part of the final mature RNA produced by that gene after introns have been removed by RNA splicing. The term exon may refer to the DNA sequence within a gene and/or to the corresponding sequence in RNA transcripts. In RNA splicing, introns are removed and exons are covalently joined to one another as part of generating the mature messenger RNA. The entire set of exons for a species constitutes the exome.

The term "patient having localized cancer disease" as used herein refers to a patient having a primary solid tumor in the absence of distant metastasis, preferably in the absence of regional lymph node metastasis or distant metastasis. For instance, it would correspond to tumors in stages TO, T1 , T2, T3 or T4 of the TNM system classification, preferably to stages TO, T1 , T2 or T3, with no signs of regional lymph node or distant metastasis (American Joint Committee on Cancer, AJCC. Chicago, Illinois. AJCC Cancer Staging Manual, 7th edition, published by Springer- Verlag New York, www.cancerstaging.org).

The term "metastasis" as used herein refers to distant metastasis affecting organs other than the primary tumor site. Metastasis may be defined as the process by which cancer spreads or transfers from the primary site to other regions of the body with the development of a similar cancerous lesion at the new location (see for instance: Chambers AF et al., Nat Rev Cancer 2002; 2: 563-72), for instance in colorectal cancer metastasis in another organ (e.g., the liver) typically shows an enteroid adenocarcinoma pattern. A "metastatic" or "metastasizing" cell is typically one that loses adhesive contacts with neighboring cells and migrates via the bloodstream or lymph from the primary site of disease to invade neighboring body structures. The term "biological fluid sample" as used herein includes biological fluids, such as whole blood, serum, plasma, synovial fluid, cerebrospinal fluid, bronchial lavage, ascites fluid, bone marrow aspirate, pleural effusion and urine. Preferably, said biological fluid sample is blood, plasma or serum. These types of samples are routinely used in the clinical practice and a person skilled in the art will know how to identify the most appropriate means for their obtaining and preservation. Once a sample has been obtained, it may be used fresh, it may be frozen or preserved using appropriate means. Such biological samples can be taken around the time of diagnosis, before, during or after treatment (e.g. surgical resection). In a particular embodiment, the samples are obtained from patients diagnosed with cancer which have a resectable tumor (i.e., candidates for surgical resection).

The human genome contains approximately 3 billion base pairs (bp), 20.800 coding genes and over 200.000 exons, which represents about 1 -3% of the genome (Cunningham F et al. Nucleic Acids Res. 2015;43, D662-669; Clamp M, et al., Proc Natl Acad Sci USA. 2007; 104(49): 19428- 19433). On average, there are nine exons per gene, with an average exon size of 170 bp (Sakharkar MK et al., In Silico Biol. 2004, 4(4):387-393). The term "whole-exome sequencing" as used herein refers to the sequencing of substantially all the coding genes in a genome.

Focusing next generation sequencing methodologies to particular regions of interest (e.g. exons) requires enrichment of the desired target regions first. Accordingly, various exome capturing methods have been described and are well known in the art. These include polymerase chain reaction strategies, hybridization-based methods and the use of selective circularization probes which are described in further detail in Ballester et al (Expert Review of Molecular Diagnostics 2016, 16 (3), 1 -37) which is hereby incorporated by reference. Any of these exome-capturing methods can be used in the methods of the invention.

In a particular embodiment, optionally in combination with one or more of the embodiments or features described above or below, exome capture is based on a hybridization method comprising the use of sequences substantially complementary to the regions of interest (i.e. capture probes). The expression "substantially complementary" to the target sequence as used herein means that is at least about 90%, preferably at least about 95%, 96%, 97%, 98%, or 99% complementary to a target polynucleotide sequence. In one embodiment, the capture probe comprises a sequence that is 100% complementary to a target polynucleotide sequence. Capture-based methods for enrichment of coding genes may be solid-phase methods, typically using oligonucleotide microarrays or in-solution methods. The latest is an approach that can be scaled up for a large number of targets and is preferred. Commercially available in-solution hybridization options for whole-exome capture and library preparation are offered by lllumina Inc (Nextera and TruSeq; San Diego, CA); Agilent Technologies (SureSelect; Santa Clara, CA); and Roche NimbleGen (SeqCap EZ; Madison, Wl). These technologies differ in the choice of target regions, the length of the baits, baits density and type of molecule used for capture (DNA or RNA). Protocols for whole exome-enrichment using the three technologies (TruSeq, SureSelect and NimbleGen) are available from Cold Spring Harbor Laboratories (Chen et al. "Whole-exome enrichment with the agilent SureSelect Human All Exon Platform". Cold Spring Harb Protoc. 2015 (7), 626-633; Chen et al. "Whole-Exome Enrichment with the Roche NimbleGen SeqCap EZ Exome Library SR Platform" Cold Spring Harbor Protocols 2015 (7), 634-641 ; Chen et al. "Whole-exome enrichment with the lllumina TruSeq Exome enrichment platform" Cold Spring Harb Protoc. 2015(7), 642-648).

Several studies have compared various in-solution hybridization methods and despite the differences among these methods, all major commercial methods referred above can capture more than 90% of the targeted regions (Bodi K, et al. Biomol Tech:JBT. 2013, 24(2):73-86; Clark MJ, et al Nat Biotechnol. 201 1 , 29(10):908-914). In a particular embodiment, exome capturing for whole-exome sequencing is conducted by a method comprising hybridization of cfDNA in said blood, plasma or serum sample with probes specifically hybridizing to substantially all the coding DNA sequences in the subject's species genome. The expression "substantially all coding sequences" refers to at least about 90%, preferably about 95%, 96%, 97%, 98%, 99% or 100% of the coding sequences available in public databases.

These probes may be modified with different ligands for target enrichment, for instance capture probes may be biotinylated. In a particular embodiment, these probes hybridize to the same sequences that the probes comprised in the SeqCap EZ probe pool provided by Roche NimbleGen to perform sequence capture. The SeqCap EZ probe pool enriches for ~44 Mb of the human exonic regions. The SeqCap system uses 55- to 105-base DNA biotinylated probes to capture known coding DNA sequences (CDS) from the NCBI Consensus CDS Database, RefSeq, and Sanger miRBase, for further details see "Whole-Exome Enrichment with the Roche NimbleGen SeqCap EZ Exome Library SR Platform" Chen et al., Cold Spring Harbor Protocols 2015. In a preferred embodiment, exome capturing for whole-exome sequencing is conducted by a method comprising hybridization of cfDNA in said blood, plasma or serum sample with probes comprising or consisting of the probes in the SeqCap EZ probe pool, preferably, wherein these probes are biotinylated. Diverse next-generation sequencing methods have been described and are well known to a person skilled in the art. These include for instance sequencing by synthesis with cyclic reversible termination approaches (e.g., Illumina, SEQLL, Qiagen), sequencing by synthesis with single-nucleotide addition approaches (e.g., Roche-454, Thermo Fisher-Ion Torrent), sequencing by ligation (e.g., Thermo Fisher SOLiD and BGI-Complete Genomics), real-time long-read sequencing (e.g., Pacific Biosciences, Oxford Nanopore Technologies), synthetic long-read sequencing (e.g., Illumina, 10X Genomics, iGenomeX), see for instance Goodwin S, et al., Nat Rev Genet. 2016, 17(6):333-51 ). In a particular embodiment, optionally in combination with any of the features or embodiments described above or below, whole-exome sequencing is conducted by paired-end sequencing.

The depth of sequencing coverage when performing whole-exome sequencing for use in a method of the invention will typically be below the one used in applications aiming at identifying single nucleotide polymorphisms (SNPs), which have generally a read depth of about 100x. In a particular embodiment, optionally in combination with any of the features or embodiments described above or below, whole-exome sequencing is characterized by an average read depth of 40-80x per sample. The term "quantifying" or "determining the levels ", as used herein, refers to ascertaining the absolute or relative amount or concentration of the exons in the sample. Techniques to assay levels of individual exons from test samples are well known to the skilled technician, and the invention is not limited by the means by which the components are assessed. In the methods of the invention exon quantification is typically conducted by counting the number of sequence reads that map to known exons when these are aligned to a reference genome sequence. The reference genome sequence will be of the same species as the subject from which a cfDNA sample has been obtained and an exome library has been prepared. In a particular embodiment said subject is a human subject and said reference genome sequence is hg38.

In some embodiments of the method of the invention, may be classified as DPE those exons which are differentially over-present in the N group, those which are differentially over-present in the M group or those which are differentially over-present in either the N group or the M group; preferably, are classified as DPE those exons which are differentially over-present in either the N group or the M group. In a preferred embodiment, in step iii) are classified as DPE those exons having a differential expression between the N and M group of more than about 1.2, 1.3, 1.4, 1.5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15 or 20 fold levels deviation (i.e., increase or decrease). Methods for determining that the difference in the exon levels between the two groups is statistically significant are well known in the art and the skilled person will know the most appropriate method. A statistically significant difference between the target exon levels can be established by a person skilled in the art by means of using different statistical tools; illustrative, non-limiting examples of said statistical tools include determining confidence intervals, determining the p-value, the Chi-Square test discriminating functions, etc. Preferred confidence intervals are at least 90%, at least 95%, at least 97%, at least 98%, at least 99%. The p-values are, preferably less than 0.1 , less than 0.05, less than 0.01 , less than 0.005 or less than 0.0001.

As above mentioned, exon quantification in the method of the invention is typically conducted by counting the number of sequence reads that map to known exons when these are aligned to a reference genome sequence. Analysis of differential presence as described herein may be performed using edgeR software package for the differential expression of RNA-seq data as described herein. RNA-seq data is typically summarized by counting the number of sequence reads that map to genomic features of interest. Negative binomial models are used in the edgeR software to capture the quadratic mean-variance relationship that can be observed in the RNA-seq data. Empirical Bayes methods are used to allow exon-specific variation estimates. In a particular embodiment of the methods of the invention, differentially expressed exons are identified using either or both of the Likelihood Ratio Tests (LRT) Quasi-Likelihood F-tests (QLF) as statistical methods.

In a particular embodiment, optionally in combination with one or more of the features or embodiments described above or below, quantification values of DPE have been normalized, preferably data normalization has been conducted by the trimmed mean of M values (TMM) method described in Robinson MD et al. (Bioinformatics. 2010, 26(1 ), 139-40) and used by the EdgeR software.

In a second aspect, the present invention provides an in vitro method of prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease, the method comprising:

i. quantifying individual exons identified as differentially present in metastatic patients

(DPE) according to a method as defined in any of claims 1 to 7 in a blood, plasma or serum sample obtained from said patient; and ii. comparing the quantification values in the patient's sample obtained in i) with the quantification values in a reference sample, wherein said reference sample is isolated from a patient or group of patients suffering from localized cancer disease; wherein when the quantification values in the patient's sample are increased or decreased in comparison with those in the reference sample the patient has a high risk of suffering from metastasis.

The term "reference value", as used herein, relates to a predetermined criteria used as a reference for evaluating the values or data obtained from the samples collected from a subject. The reference value or reference level can be an absolute value, a relative value, a value that has an upper or a 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 or 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.

The reference value according to the method of the invention can be obtained from one or more subjects having localized cancer disease, from subjects suffering from metastasis, from subjects suffering from cancer at early stage, such as non-symptomatic (preclinical stage) or from the same subject that was diagnosed as having cancer but at an earlier time point.

In a particular embodiment, said reference sample is isolated from a patient or group of patients suffering from localized cancer disease, and when the quantification values of said set of individual exons identified as differentially present in metastatic patients (DPE) are significantly different from those in the reference sample the patient has a high risk of suffering from metastasis.

In another particular embodiment, said reference sample is isolated from a patient or group of patients suffering from metastasis; and when the quantification values of said set of individual exons identified as differentially present in metastatic patients (DPE) in the patient's sample are significantly different from those in the reference sample, the patient has a low risk of suffering from metastasis.

In an additional particular embodiment, the method of the invention is a method of monitoring cancer progression and the reference value is obtained from the same subject that was diagnosed as having cancer but at an earlier time point. In the methods of the invention, the level of presence (i.e. abundance) of an exon is considered "decreased" when the level of said exon in a sample is lower than its reference value. Preferably, the level is considered to be lower than its reference value when it is at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 100%, at least 1 10%, at least 120%, at least 130%, at least 140%, at least 150%, or more lower than its reference value; more preferably, the level is considered to be lower than its reference value when it is at least 20% lower than its reference value.

Likewise, in the context of the methods of the invention, the level of presence (i.e. abundance) of an exon is considered "increased" when the level of said exon in a sample is higher than its reference value. Preferably, the level is considered to be higher than its reference value when it is at least 10%, at least 15%, at least 20%, at least 25%, at least 30%, at least 35%, at least 40%, at least 45%, at least 50%, at least 55%, at least 60%, at least 65%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, at least 100%, at least 1 10%, at least 120%, at least 130%, at least 140%, at least 150%, or more higher than its reference value; more preferably the level is considered to be higher than its reference value when it is at least 20% higher than its reference value.

Alternatively or in addition, subjects having more than about 1 .1 ,1 .2, 1.3, 1 .4, 1 .5, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15 or 20 fold levels deviation (i.e., increase or decrease) than the reference value as described herein, preferably having more than about 1 .2 folds levels deviation than the reference value, may be identified as being differentially present.

Metastasis prediction in the method of the invention, as it is understood by a person skilled in the art, does not claim to be correct in 100% of the analyzed samples. However, it requires that a statistically significant amount of the analyzed samples are classified correctly. The amount that is statistically significant can be established by a person skilled in the art by means of using different statistical tools; illustrative, non-limiting examples of said statistical tools include determining confidence intervals, determining the p-value, the Chi-Square test discriminating functions, etc. Preferred confidence intervals are at least 90%, at least 95%, at least 97%, at least 98%, at least 99%. The p-values are, preferably less than 0.1 , less than 0.05, less than 0.01 , less than 0.005 or less than 0.0001. The teachings of the present invention preferably allow correctly diagnosing in at least 60%, in at least 70%, in at least 80%, or in at least 90% of the subjects of a determining group or population analyzed. There are different types of solid tumors which are typically named according to the type of cells that form them. Examples of solid tumors are lung cancer, sarcoma, malignant melanoma, 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, mesothelioma and lymphoma.

The differentially present exons detected using the method of the invention can probably come from tumor and non-tumor derived cfDNA. In a preferred embodiment, the localized cancer disease is one characterized by presence of circulating tumor DNA (ctDNA) in the blood, in particular where the amount of ctDNA arising from the primary solid tumor is such that exons can be quantified from a blood, plasma or serum sample by molecular methods known in the art such as next-generation sequencing. This method being particularly useful for the determination of metastasis risk in those cancer types where one of the mechanisms involved in metastasis mediation is the dissemination of acellular material in the plasma (e.g. exosomes, microparticles, lysosomes). For instance, illustrative non limiting examples of cancer types where presence of circulating tumor DNA has been described include colorectal cancer, breast cancer, lung cancer, ovarian cancer and pancreatic cancer (Nature Reviews Cancer | AOP, published online 12 May 201 1 ; Genomics Proteomics Bioinformatics 15 (2017) 59-72).

In a preferred embodiment, said solid tumor is a carcinoma. The term "carcinoma" as used herein refers to a malignant neoplasm of epithelial origin or cancer of the internal or external lining of the body. Carcinomas are divided into two major subtypes: adenocarcinoma, which develops in an organ or gland, and squamous cell carcinoma, which originates in the squamous epithelium. Carcinoma can be classified according to the site of origin or cell type and include inter alia: breast carcinoma; ovarian carcinoma; cervix carcinoma; gastric carcinoma; non-small cell lung carcinoma; small cell lung carcinoma; pancreatic carcinoma; prostate carcinoma; colon carcinoma; liver carcinoma; renal carcinoma; bladder carcinoma; prostate carcinoma; head and neck carcinoma; squamous cell carcinoma; epidermoid carcinoma; choriocarcinoma; seminoma; embryonal cell carcinoma; and mesothelioma. In a preferred embodiment, said carcinoma is colon adenocarcinoma. Preferably, said cancer is colorectal cancer.

In a preferred embodiment, said cancer is colorectal cancer and said differentially present exons are selected from the group consisting of the 379 exons found to be differentially present in colorectal cancer defined in Table 1 below. Table 1. DPE between CRC patients with metastatic and localized disease (EnsembI Exon ID; www.ensembl.org/); Version: EnsembI GRCh38.p2 (dec 2014).

ENSE00000434238 ENSE0000071 1736 ENSE00000756079

ENSE00000653232 ENSE00000718690 ENSE00000765456

ENSE00000653391 ENSE00000722587 ENSE00000774903

ENSE00000656382 ENSE00000737375 ENSE00000776449

ENSE00000657505 ENSE00000740608 ENSE00000777978

ENSE00000686566 ENSE00000743794 ENSE00000778155

ENSE00000686774 ENSE00000743816 ENSE00000783622

ENSE00000689282 ENSE00000745482 ENSE00000783996

ENSE00000689806 ENSE00000748078 ENSE00000790723

ENSE00000698040 ENSE00000751409 ENSE00000796406

ENSE00000796933 ENSE00000958441 ENSE00001 139210

ENSE00000800497 ENSE00000962639 ENSE00001 148562

ENSE00000802167 ENSE00000970330 ENSE00001 150459

ENSE00000806199 ENSE00000970745 ENSE00001 151941

ENSE00000836542 ENSE00000973444 ENSE00001 154392

ENSE00000836822 ENSE00000983683 ENSE00001 154874

ENSE00000837388 ENSE00000985419 ENSE00001 172744

ENSE00000845244 ENSE00000988136 ENSE00001 173836

ENSE00000847773 ENSE00000991439 ENSEOOOO 1 182049

ENSE00000859183 ENSE00001005678 ENSE00001 186180

ENSE00000879785 ENSE00001006075 ENSEOOOO 1 186768

ENSE00000880163 ENSE00001009766 ENSEOOOO 1 187499

ENSE00000881769 ENSE00001018420 ENSE00001 196195

ENSE00000887360 ENSE00001019109 ENSE00001201096

ENSE00000893357 ENSE00001026920 ENSE00001210364

ENSE00000900718 ENSE00001030161 ENSE00001216121

ENSE0000090661 1 ENSE00001034760 ENSE00001218189

ENSE00000906613 ENSE00001037219 ENSE00001218212

ENSE00000906663 ENSE00001042850 ENSE00001222140

ENSE00000915853 ENSE00001046392 ENSEOOOO 1225273

ENSE00000919350 ENSE00001054410 ENSEOOOO 1229237

ENSE00000923654 ENSE00001067780 ENSE00001244123

ENSE00000924669 ENSE00001068195 ENSEOOOO 1245555 ENSE00000926037 ENSE00001072073 ENSEOOOO 1252408

ENSE00000928244 ENSEOOOO 1085947 ENSEOOOO 1255660

ENSE00000928742 ENSEOOOO 1086752 ENSE00001258014

ENSE00000931905 ENSEOOOO 1086848 ENSEOOOO 1259847

ENSE00000932329 ENSEOOOO 1087864 ENSEOOOO 1262755

ENSE00000932683 ENSE00001093991 ENSEOOOO 1267874

ENSE00000938453 ENSEOOOO 1094477 ENSE00001269619

ENSE00000942189 ENSEOOOO 1094840 ENSE00001272710

ENSE00000943999 ENSE00001099123 ENSE00001281957

ENSE00000948540 ENSEOOOO 1 122020 ENSEOOOO 1283934

ENSE00000952678 ENSEOOOO 1 122449 ENSEOOOO 1288764

ENSE00000957948 ENSE00001 138641 ENSEOOOO 1292358

ENSEOOOO 1294554 ENSEOOOO 1477424 ENSEOOOO 1654437

ENSE00001317652 ENSEOOOO 1482260 ENSEOOOO 1655745

ENSEOOOO 1324426 ENSEOOOO 1487204 ENSE00001659171

ENSEOOOO 1327286 ENSE00001500109 ENSEOOOO 1660809

ENSEOOOO 1327628 ENSEOOOO 1500245 ENSEOOOO 1677669

ENSE00001330154 ENSEOOOO 1505294 ENSEOOOO 1678238

ENSE00001331282 ENSE00001507564 ENSEOOOO 1679508

ENSE00001331970 ENSE00001532501 ENSE00001680713

ENSE00001335315 ENSEOOOO 1536025 ENSE00001681577

ENSEOOOO 1338847 ENSE00001544012 ENSE00001690721

ENSEOOOO 1348898 ENSEOOOO 1544497 ENSEOOOO 1692722

ENSEOOOO 1363598 ENSEOOOO 1544499 ENSE00001695592

ENSEOOOO 1365279 ENSE00001552512 ENSEOOOO 1697223

ENSE00001367514 ENSEOOOO 1555479 ENSEOOOO 1698002

ENSE00001370613 ENSEOOOO 1559349 ENSE00001700106

ENSEOOOO 1374565 ENSE00001577541 ENSEOOOO 1703944

ENSEOOOO 1377582 ENSE00001580614 ENSEOOOO 1708743

ENSE00001378512 ENSE00001587001 ENSEOOOO 1722665

ENSE00001385581 ENSE00001590717 ENSE00001722751

ENSE00001387195 ENSE00001604217 ENSEOOOO 1724086

ENSE00001399393 ENSE00001607375 ENSEOOOO 1725060

ENSE00001400048 ENSE00001610688 ENSEOOOO 1726502

ENSE0000141 1632 ENSE0000161 1930 ENSE00001729071 ENSE00001425520 ENSE00001612229 ENSE00001729707

ENSE00001433450 ENSE00001612795 ENSEOOOO 1736622

ENSE00001442068 ENSE00001615936 ENSE00001738815

ENSE00001447946 ENSE00001619008 ENSEOOOO 1740253

ENSE00001448750 ENSE00001620608 ENSEOOOO 1743273

ENSE00001451570 ENSEOOOO 1620623 ENSEOOOO 1745825

ENSEOOOO 1453252 ENSE00001621025 ENSE00001746619

ENSE00001453907 ENSEOOOO 1623204 ENSEOOOO 1759285

ENSEOOOO 1454928 ENSEOOOO 1629777 ENSEOOOO 1762320

ENSEOOOO 1462984 ENSE00001633936 ENSEOOOO 1767060

ENSEOOOO 1465369 ENSEOOOO 1652593 ENSE00001773796

ENSEOOOO 1465380 ENSEOOOO 1652658 ENSE000017761 12

ENSEOOOO 1779364 ENSE00002229442 ENSE00002518209

ENSE00001779813 ENSE00002233678 ENSE00002518335

ENSEOOOO 1784545 ENSE00002243315 ENSE00002519157

ENSEOOOO 1788092 ENSE00002260934 ENSE00002521 166

ENSE00001799334 ENSE00002265455 ENSE00002521763

ENSE00001801381 ENSE00002268656 ENSE00002534362

ENSEOOOO 1804294 ENSE00002276159 ENSE00002551446

ENSE00001806781 ENSE00002278414 ENSE00002563835

ENSE00001839365 ENSE00002316371 ENSE00002593366

ENSE00001841354 ENSE00002318685 ENSE00002605818

ENSE00001887130 ENSE00002322170 ENSE00002606846

ENSEOOOO 1905643 ENSE00002328787 ENSE00002685281

ENSE00001930056 ENSE00002333390 ENSE00002694730

ENSE00001979195 ENSE00002345671 ENSE00002707913

ENSE00002031472 ENSE00002371 184 ENSE00002716524

ENSE00002037226 ENSE00002375408 ENSE00002749781

ENSE00002039916 ENSE00002385145 ENSE00002872785

ENSE00002052516 ENSE00002393310 ENSE00002899954

ENSE00002054040 ENSE00002431283 ENSE00002946189

ENSE00002054327 ENSE00002451 123 ENSE00003160461

ENSE00002067388 ENSE00002457525 ENSE00003378215

ENSE00002071708 ENSE00002459604 ENSE00003407676

ENSE00002100538 ENSE00002460354 ENSE00003456898 ENSE00002113883 ENSE00002472627 ENSE00003467016

ENSE00002116906 ENSE00002474653 ENSE00003484499

ENSE00002117591 ENSE00002478241 ENSE00003489833

ENSE00002129095 ENSE00002479642 ENSE00003515415

ENSE00002159094 ENSE00002488634 ENSE00003518162

ENSE00002177860 ENSE00002489974 ENSE00003522662

ENSE00002195738 ENSE00002497133 ENSE00003524777

ENSE00002206839 ENSE00002497158 ENSE00003531526

ENSE00002208070 ENSE00002500983 ENSE00003567951

ENSE00002216003 ENSE00002505263 ENSE00003568223

ENSE00002218429 ENSE00002511764 ENSE00003574811

ENSE00002222662 ENSE00002518041 ENSE00003576904

ENSE00003580867 ENSE00003688051 ENSE00003729869

ENSE00003580877 ENSE00003702436 ENSE00003731831

ENSE00003584693 ENSE00003715388 ENSE00003733828

ENSE00003589370 ENSE00003716981 ENSE00003734079

ENSE00003593814 ENSE00003717538 ENSE00003749342

ENSE00003648889 ENSE00003719100 ENSE00003750954

ENSE00003657877 ENSE00003721716 ENSE00003770543

ENSE00003665630 ENSE00003721972 ENSE00003786302

ENSE00003671238 ENSE00003723030 ENSE00003786618

ENSE00003681685 ENSE00003724449 ENSE00003786655

ENSE00003685102 ENSE00003728302

ENSE00003685588 ENSE00003729471

It is further noted that the accuracy of the method of the invention can be increased by determining the presence and/or the quantification of other biomarkers which have been described to be associated to metastasis (see for instance, Martini G, Troiani T, Cardone C,et al., World J Gastroenterol. 2017 23(26):4675-4688) and/or clinical signs or symptoms with reported prognostic/predictive value, such as morphological features of the tumor, histological subtypes, radiological traits of the imaging tests (e.g. size, shape, volume, radiological texture, morphological details or other features in a CT scan, X-Ray or SUV or alternative ways to analyze nuclear tracer levels in a PET imaging, etc); clinical characteristics of the patients (e.g. age, sex, race, respiratory function levels, performance status). The potential additional biomarkers to be associated to the present invention can be found in the tumor specimen itself or other cells, or other bodily sample, such as body fluids (e.g. blood or plasma), feces or exhaled breath obtained from the same patient.

In a particular embodiment, optionally in combination with one or more of the features or embodiments described above or below, the methods of the invention further comprise determining the presence and/or the quantification of other biomarkers, clinical signs and/or symptoms, and/or clinical characteristics of the patients predictive of metastasis.

The term "biomarker" as used herein refers to markers of disease which are typically substances found in a bodily sample which generally can be easily measured. Typically, the measured amount correlates to an underlying disease pathophysiology, making it useful for diagnosing, predicting and/or measuring the progress of a disease or the effects of a treatment. The term biomarker encompasses biophysical and biochemical determinations, including genetic and serological markers.

The method of prognosis or for predicting the risk of suffering from metastasis in a cancer patient of the invention may be used for the classification or selection of patients as belonging to a particular group of risk (i.e. those patients with an increased risk of suffering from metastasis). Furthermore, a treatment may be selected or personalized according to the risk group into which the cancer patient has been classified. For instance, patients in the high risk group will be treated with the best possible treatment (i.e, surgical, adjuvant and/or neoadjuvant treatment) and treatment regimen; whereas a less aggressive treatment or no treatment will be selected for those patients with lower risk of suffering from metastasis. Accordingly, in another aspect, the invention relates to an in vitro method for selecting the treatment for a patient having localized cancer disease, wherein said method comprises determining the prognosis or predicting the risk of suffering from metastasis in said cancer patient by a method as defined in previous aspects of the invention. In a related aspect, the invention refers to a method for the treatment of a patient having localized cancer disease comprising administering to said patient a therapeutically effective amount of a treatment (e.g. a drug or drug combination) wherein said treatment is selected according to the classification of said cancer patient according to its risk of suffering from metastasis, wherein said risk has been determined by the method of prognosis or for predicting the risk of suffering from metastasis according to the invention. A person skilled in the art will know how to select the most appropriate treatment according to the type of cancer and the risk of metastasis. This treatment may be a neoadjuvant treatment administered prior to the surgical removal of the tumor and/or an adjuvant treatment after the surgical intervention.

In a further aspect, the invention refers to an in vitro method of monitoring disease progression or response to a treatment in a patient having localized cancer disease, wherein said method comprises determining the prognosis or the risk of suffering from metastasis in said cancer patient by a method according to the invention.

The term "monitoring" as used herein refers to determining the evolution of the disease and/or the efficacy of a therapy, for example determining whether there has been a change in the levels of a set of DPE which is indicative of worst prognosis or increased risk of metastasis. One of the goals of the method of monitoring of the invention is to early detect an increased risk of metastasis. Preferred embodiments and features of the invention are as described under previous aspects.

The methods of the present invention might be implemented by a computer. Therefore, a further aspect of the invention refers to a computer implemented method, wherein the method is any of the methods disclosed herein or any combination thereof.

It is noted that any computer program capable of implementing any of the methods of the present invention or used to implement any of these methods or any combination thereof, also forms part of the present invention. It is also noted that any device or apparatus comprising means for carrying out the steps of any of the methods of the present invention or any combination thereof, or carrying a computer program capable of, or for implementing any of the methods of the present invention or any combination thereof, is included as forming part of the present specification. The methods of the invention may also comprise the storing of the method results in a data carrier, preferably wherein said data carrier is a computer readable medium. The present invention further relates to a computer-readable storage medium having stored thereon a computer program of the invention or the results of any of the methods of the invention. As used herein, "a computer readable medium" can be any apparatus that may include, store, communicate, propagate, or transport the results of the determination of the method of the invention. The medium can be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system (or apparatus or device) or a propagation medium.

Kit and use of a kit in the methods of the invention

The invention further relates to a kit as defined herein below and to the use of said kit for the prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease on the basis of the differential presence of exons in circulating cell-free DNA (cfDNA) according to a method of the invention, said kit comprising:

a) a reagent for quantifying individual exons in a biological fluid sample (e.g. blood, plasma or serum sample);

b) optionally, instructions for the use of said reagent(s) in determining the levels of said exons, in said biological fluid sample.

Methods for nucleic acids quantification are well known in the art and have been described herein above. This may include next generation sequencing, quantitative PCR (qPCR), PCR- pyrosequencing, PCR-ELISA, DNA microarrays, branched DNA, dot-blot, Fluorescence In Situ Hybridization assay (FISH), and multiplex versions of said methods.

The kit of the invention will therefore include reagents according to the selected method for exon quantification.

In a particular embodiment, said kit comprises reagents suitable for performing a real-time or qPCR reaction, which typically contain a DNA polymerase, such as Taq DNA polymerase (e.g., hot- start Taq DNA polymerase), buffer, magnesium, dNTPs, and optionally other agents (e.g., stabilizing agents such as gelatin and bovine serum albumin). In addition, real-time PCR reaction mixtures also contain reagents for real time detection and quantification of amplification products.

In another particular embodiment, said kit comprises reagents suitable for whole-exome sequencing of cfDNA by next generation sequencing, for instance it will comprise reagents suitable for whole-exome capturing and/or for next generation sequencing. Such kit may comprise for instance reagents typically used in DNA extraction protocols (e.g. wash and/or dilution buffers, proteinase, etc.) and/or reagents usually used in the capturing of nucleic acid sequences (e.g. buffers, magnesium, probes, etc.).

In a particular embodiment, optionally in combination with one or more of the embodiments or features described above or below, said patient has colorectal cancer and said kit comprises reagents suitable for the quantification of individual exons selected from the group consisting of the 379 exons found to be differentially present in colorectal cancer defined in Table 1 .

ITEMS OF THE INVENTION

An in vitro method for identifying exons which are differentially present in patients having metastasis (DPE) with respect to patients having localized cancer disease, wherein said method comprises the following steps:

i. quantifying individual exons by whole-exome sequencing of the circulating cell- free DNA (cfDNA) by next-generation sequencing in a blood, plasma or serum sample obtained from a cancer patient or group of patients with localized disease (N group);

ii. quantifying individual exons by whole-exome sequencing of the cfDNA by next- generation sequencing in a blood, plasma or serum sample obtained from a cancer patient or group of patients with metastasized disease (M group);

iii. comparing the quantification values in i) with the quantification values in ii) and identify as differentially present exons those which abundance differs in a statistically significant manner between both groups.

The method according to item 1 , wherein exome capturing for whole-exome sequencing is conducted by a method comprising hybridization of cfDNA in said blood, plasma or serum sample with probes substantially complementary to substantially all the coding DNA sequences in the patient's species genome.

The method according to any of the precedent items, wherein whole-exome sequencing is characterized by an average read depth of 40-80x per sample.

The method according to any of the precedent items, wherein exon quantification is conducted by counting the number of sequence reads obtained by whole-exome sequencing that map to known exons when these are aligned to a reference genome sequence.

5. The method according to any of the preceding items, wherein in step iii) are classified as DPE those exons which quantification levels differ between the two groups in at least 1.2- fold. The method according to any of the preceding items, wherein in step iii) Likelihood Ratio Tests (LRT) and/or Quasi-Likelihood F-tests (QLF) are used as statistical methods for identifying differentially present exons. The method according to any of the preceding items, wherein quantification values of DPE have been normalized, preferably data normalization has been conducted by the trimmed mean of M values (TMM) method. An in vitro method of prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease, the method comprising:

i. quantifying individual exons identified as differentially present in metastatic patients (DPE) according to a method as defined in any of items 1 to 7 in a blood, plasma or serum sample obtained from said patient; and

ii. comparing the quantification values in the patient's sample obtained in i) with the quantification values in a reference sample, wherein said reference sample is isolated from a patient or group of patients suffering from localized cancer disease; wherein when the quantification values in the patient's sample are increased or decreased in comparison with those in the reference sample the patient has a high risk of suffering from metastasis. The method according to item 8, wherein DPE are quantified by a method selected from the group consisting of next generation sequencing, quantitative PCR (qPCR), PCR- pyrosequencing, PCR-ELISA, DNA microarrays, branched DNA, dot-blot, Fluorescence In Situ Hybridization assay (FISH), and multiplex versions of said methods. The method according to any of items 8 or 9, wherein DPE are quantified by whole-exome sequencing using next generation sequencing. The method according to any of the precedent items, wherein said patient is a human patient. The method according to any of the precedent items, wherein said cancer is one characterized by presence of circulating tumor DNA (ctDNA) in the blood. The method according to any of the precedent items, wherein said cancer is selected from the group consisting of colorectal cancer, breast cancer, lung cancer, ovarian cancer and pancreatic cancer. 14. The method according to item 13, wherein said cancer is colorectal cancer, preferably adenocarcinoma. 15. The method according to item 14, wherein said differentially present exons are selected from the group consisting of the 379 exons found to be differentially present in colorectal cancer defined in Table 1.

16. An in vitro method for selecting a treatment for a patient having localized cancer disease wherein said method comprises selecting a treatment according to the classification of said patient according to its prognosis or risk of metastasis by a method as defined in any of items 8 to 15.

17. An in vitro method of monitoring disease progression or response to a treatment in a patient having localized cancer disease, wherein said method comprises determining the prognosis or the risk of suffering from metastasis in said cancer patient by a method as defined in any of items 8 to 15.

18. A method according to any of the preceding items, wherein said method further comprises storing the method results in a data carrier, preferably wherein said data carrier is a computer readable medium.

19. A computer implemented method, wherein the method is as defined in any of items 1 to 18. 20. Use of a kit for the prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease on the basis of the differential presence of exons in circulating cell-free DNA (cfDNA) according to a method of any of items 8 to 15, said kit comprising:

a) a reagent for quantifying individual exons in a blood, plasma or serum sample;

b) optionally, instructions for the use of said reagent(s) in determining the levels of said exons, in a blood, plasma or serum sample.

21 . Use of a kit according to item 20, wherein said kit comprises reagents suitable for whole- exome capturing and/or for next generation sequencing.

22. Use of a kit according to any of items 20 or 21 , wherein said patient has colorectal cancer and said kit comprises reagents suitable for the quantification of individual exons selected from the group consisting of the 379 exons found to be differentially present in colorectal cancer defined in Table 1.

It is contemplated that any features described herein can optionally be combined with any of the embodiments of any method, kit, use of a kit, or computer program of the invention; and any embodiment discussed in this specification can be implemented with respect to any of these. It will be understood that particular embodiments described herein are shown by way of illustration and not as limitations of the invention. The principal features of this invention can be employed in various embodiments without departing from the scope of the invention. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, numerous equivalents to the specific procedures described herein. Such equivalents are considered to be within the scope of this invention and are covered by the claims. All publications and patent applications are herein incorporated by reference to the same extent as if each individual publication or patent application was specifically and individually indicated to be incorporated by reference.

The use of the word "a" or "an" may mean "one," but it is also consistent with the meaning of "one or more," "at least one," and "one or more than one". The use of the term "another" may also refer to one or more. The use of the term "or" in the claims is used to mean "and/or" unless explicitly indicated to refer to alternatives only or the alternatives are mutually exclusive.

As used in this specification and claim(s), the words "comprising" (and any form of comprising, such as "comprise" and "comprises"), "having" (and any form of having, such as "have" and "has"), "including" (and any form of including, such as "includes" and "include") or "containing" (and any form of containing, such as "contains" and "contain") are inclusive or open-ended and do not exclude additional, unrecited elements or method steps. The term "comprises" also encompasses and expressly discloses the terms "consists of" and "consists essentially of". As used herein, the phrase "consisting essentially of" limits the scope of a claim to the specified materials or steps and those that do not materially affect the basic and novel characteristic(s) of the claimed invention. As used herein, the phrase "consisting of" excludes any element, step, or ingredient not specified in the claim except for, e.g., impurities ordinarily associated with the element or limitation.

The term "or combinations thereof" as used herein refers to all permutations and combinations of the listed items preceding the term. For example, "A, B, C, or combinations thereof" is intended to include at least one of: A, B, C, AB, AC, BC, or ABC, and if order is important in a particular context, also BA, CA, CB, CBA, BCA, ACB, BAC, or CAB. Continuing with this example, expressly included are combinations that contain repeats of one or more item or term, such as BB, AAA, AB, BBC, AAABCCCC, CBBAAA, CABABB, and so forth. The skilled artisan will understand that typically there is no limit on the number of items or terms in any combination, unless otherwise apparent from the context.

As used herein, words of approximation such as, without limitation, "about", "around", "approximately" refers to a condition that when so modified is understood to not necessarily be absolute or perfect but would be considered close enough to those of ordinary skill in the art to warrant designating the condition as being present. The extent to which the description may vary will depend on how great a change can be instituted and still have one of ordinary skilled in the art recognize the modified feature as still having the required characteristics and capabilities of the unmodified feature. In general, but subject to the preceding discussion, a numerical value herein that is modified by a word of approximation such as "about" may vary from the stated value by ±1 , 2, 3, 4, 5, 6, 7, 8, 9, or 10%. Accordingly, the term "about" may mean the indicated value ± 5% of its value, preferably the indicated value ± 2% of its value, most preferably the term "about" means exactly the indicated value (± 0%). The following examples serve to illustrate the present invention and should not be construed as limiting the scope thereof.

EXAMPLES Materials and Methods Patients

Thirty patients with colorectal cancer were selected following the criteria shown in Table 2 from January to December 2014 in the Department of General Surgery at Fundacion Jimenez Diaz University Hospital according to a protocol approved by the Ethics Committee for Clinical Research of this Institution. Informed consent was obtained from each subject and all investigations were performed in accordance with the principles embodied in the World Medical Association Declaration of Helsinki. Subjects were classified into 3 groups: metastatic patients (M; n=10), non-metastatic patients (N; n=10) and a group containing unclassifiable patients (U; n=10) according to the selection criteria shown in Table 2. Tumor staging was conducted according to the TNM Classification of Malignant Tumours (American Joint Committee on Cancer, AJCC. Chicago, Illinois. AJCC Cancer Staging Manual, 7th edition, published by Springer-Verlag New York, www.cancerstaging.org).

Briefly, we considered as unclassifiable patients those with T4 locally advanced disease and/or affected nodes but with no signs of distant metastasis affecting other organs or peritoneal carcinomatosis determined by PET-CT (MO).

Table 2. Patient selection criteria

Sample collection and processing

Blood samples were collected before surgery in EDTA tubes and centrifuged at 1800 x g for 10 minutes. Plasma obtained from the first centrifugation was centrifuged again at 3000 x g for 10 min, aliquoted and stored at -80°C until analysis. Library Preparation, Exome Capture and Sequencing

Circulating cell-free DNA was extracted from plasma with the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. Concentration, quantity and integrity of cfDNA were estimated prior to use. The size distribution of the cfDNA fragments was determined using an Agilent 2100 BioAnalyzer (Agilent Technologies, Santa Clara, CA, USA). Additional shearing was not performed since the majority of circulating DNA fragments in plasma is naturally short. Library preparation and specific exome capture were performed using the SeqCap EZ HGSC VCRome Kit (Roche NimbleGen, Basel, Switzerland). This protocol is based on the Roche NimbleGen SeqCap EZ Exome Library SR platform, for further details see Chen et al. "Whole-Exome Enrichment with the Roche NimbleGen SeqCap EZ Exome Library SR Platform" Chen et al., Cold Spring Harbor Protocols 2015 (7), 634-641 . The libraries were hybrid captured using biotinylated probes. Adapter DNA sequences were placed on both ends with a total length of 126 nucleotides and exomes were sequenced on an lllumina NextSeq500 platform (lllumina, Inc, San Diego, CA, USA) with 75 bp paired-end reads. Library preps and sequencing were performed at the Genomic Facility of the Scientific Park of Madrid, Spain.

Data analysis

Around 974 M of 2x75 nt reads were obtained, with an average read depth of 40-80x per sample (typically general sequencing depth is about 100x in SNP analysis). The coverage was calculated as the numbers of reads times the length divided by the human exome size. Quality analyses were performed over reads using FastQC software (FastQC [Available from: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/]).

Reads were aligned against H. sapiens genome (hg38; ftp://ftp.ensembl.org/pub/release- 89/fasta/homo_sapiens/dna/) using Bowtie (Langmead B et al. Genome Biol. 2009, 10(3), R25) with the following parameters: -v 3 -k 1 -best.

- Identification of Differentially Present Exons (DPE)

cfDNA sequencing data were processed as in a typical RNA-seq pipeline, but this strategy was aimed at detecting presence of exons instead of gene expression. For microarrays, the abundance of a particular transcript is measured as a fluorescence intensity, which is a continuous response, whereas for digital gene expression (DGE) data (as obtained by Next Generation Sequencing (NGS) methods) the abundance is observed as a count. Therefore, procedures that are successful for microarray data are not directly applicable to DGE data. The present analysis was performed using the R package 'EdgeR' (Robinson MD et al. Bioinformatics. 2010, 26(1 ), 139-40) which is specifically designed for the analysis of replicated count-based expression data. The counts were calculated using HTSeq-count software (Anders S et al Bioinformatics. 2015, 31 (2), 166-9).

Then, we used statistical methods based on generalized linear models (glm), suitable for multifactor experiments of any complexity to identify differences in exon presence (DPE) between the two groups. The glm functions can test for differential expression using either likelihood ratio tests (LRT) (Robinson MD et al. Bioinformatics. 2010, 26(1 ), 139-40) or quasi- likelihood F-tests (QLF) (Lund SP et al. Stat Appl Genet Mol Biol. 2012;1 1 (5)). We selected the DPE for a p-value of p_v≤ 0.005. MA plots for selected DPE were represented.

- DPE clustering and Principal Components Analysis (PCA)

We obtained normalized presence values of every exon for each sample using EdgeR (Counts Per Million, CPM). EdgeR uses the trimmed mean of M values (TMM) for data normalization (Robinson MD et al. Bioinformatics. 2010, 26(1 ), 139-40; McCarthy, J. D, et al., Nucleic Acids Research 2012, 40(10), 4288-4297). The TMM method estimates scale factors between samples that can be incorporated into normally used statistical methods for differential expression analysis.

Once normalized presence values were calculated, they were used to cluster the samples using the Ward's method (Ward JH Journal of the American Statistical Association. 1963, 58(301 ), 9).

Principal Components Analysis (PCA) was conducted with an in-house R script software, using the obtained data of normalized quantification of DPE. - Random Forest (RF) classification

RF classification was implemented with an R script using YandomForest' package (Liaw A. R News. 2002, 2(3), 5). Briefly, 2 samples from M and N were randomly selected and extracted from each group, respectively, using the 8 remaining samples (16 samples in total) as a 'training set' to generate a predictive algorithm. One hundred classifications were performed by iteration of this process and the mean value of the obtained probabilities was calculated. The accuracy of the resulting model was tested by checking its ability to correctly classify previously extracted samples into their corresponding groups of origin.

- Statistical analysis

Non-parametric Kolmogorov-Smirnov and Wilcoxon-Mann-Whitney tests for significance were performed using R software. Results

cfDNA extraction and whole-exome sequencing

Circulating cell-free DNA was successfully extracted from all plasma samples, obtaining a variable concentration of DNA that ranged from 13.76 to 1602.90 pg/μΙ-. Median DNA concentration was higher in metastatic patients in comparison to non-metastatic patients (Fig. 1A), although this difference was not statistically significant. Unclassifiable patients' DNA median concentration was slightly elevated with respect to both groups. Bioanalyzer plots revealed a characteristic cfDNA sizing distribution with a nucleosomal fragmentation pattern. We obtained cfDNA with median fragment lengths of 173 and 342 bp, once adapter sequencing lengths (126 nt) were substracted. One additional peak of 51 1 bp was only observed in 2 patients of our series (Fig. 1 B and Table 3).

Sizing distribution showed a typical nucleosomal laddering. The first and predominant peak had a length around 173 bp, probably corresponding to mononucleosomal DNA, whereas other cfDNA molecules were present in multiples of this size, characteristic of a di- and trinucleosomal fragmentation pattern, as previously described (Dietz S et al. PLoS One. 2016, 1 1 (8), e0161012). It has been suggested three possible sources of cfDNA: apoptosis, necrosis and active release (Bronkhorst AJ et al. Biochim Biophys Acta. 2016, 1863(1 ), 157-65), however, most of authors have emphazised the cell death as the main origin (Jahr S, RD, et al. Cancer Res. 2001 , 61 (4), 1659-65; Jiang P et al. Proc Natl Acad Sci USA. 2015, 1 12(1 1 ), E1317-25). Without willing to be bound by theory, in the present study, the oligonucleosomal laddering- pattern of size distribution and the trace quantities of sequences larger than 10000 bp, suggest that in the origin might have concurred apoptosis and, minimally, necrosis. These results are in agreement with previous observations (Heitzer E et al. Int J Cancer. 2013, 133(2), 346-56).

The total number of reads per patient ranged from 45 to 87 million reads with a read length of 76 bp (see Table 3 for further details). Quality analyses performed over reads using FastQC software indicated that base calling quality (Phred+33 quality score) was maintained in general good standard across all cycles with median and mean base quality over 28, although some bases' quality fell down to 22. As usual, a certain lack of accuracy was found in 10-1 1 first bases.

The GC content varied between 46 and 50. Results showed a percentage of aligned reads ranging from 64.20 to 77.61 % (Table 3). Thus, we considered that it was unnecessary to trim or filter reads due to quality reasons. An schematic representation of the experimental workflow is shown in Figure 2. Table 3

DNA Peak 1 Peak 2 Peak 3

Patient ID Reads G+C (pg/ML) (bp) (bp) (bp) Group

1 741431 14 49 73.97 298 459 M

2 86815250 50 635.47 293 N

3 55673798 48 18.71 297 460 M

4 87651 134 48 566 302 U

5 50063772 48/47 28.84 295 464 M

6 60477358 49 100.27 295 462 N

7 54333354 48/47 73.55 298 469 N

8 77724428 48 151.34 296 464 U

9 50174462 47/46 33.99 297 475 N

10 56406050 47/46 23.99 297 469 N

1 1 55199556 49 1602.9 295 M

12 47482676 48 129.36 297 467 U

15 54307914 47 96.29 296 467 U

17 73159248 47 49.36 299 466 N

19 84446166 48 126.72 299 472 N

23 73253014 46 443.75 299 471 632 U

25 75216530 50 250.77 296 463 N

28 73314574 47 100.05 300 474 U

31 45252722 46 1 128.07 298 M

41 46336662 47 23.81 299 478 M

42 651 12448 48 45.97 297 466 N

43 72144422 48/49 122.37 303 478 M

46 60040810 48 721.52 304 453 M

48 83631372 49 91.48 310 466 M

50 53433962 46 13.76 307 481 N

57 69015600 48 80.94 301 469 U

58 63934496 48 24.57 304 474 U

63 52804622 46 342.28 302 472 641 U

66 78342024 48 108.48 304 481 U

70 68775536 49/48 63.89 299 466 M

Group M = 10

Group N = 10

Group U = 10

Median 98.17 299 468 637 Identifying Differentially Present Exons (DPE)

One of the main goals of this research was to compare M and N groups and define a set of exons that could allow distinguishing between both groups using the differential levels of these exons in cfDNA.

The DPE was analyzed with EdgeR using either likelihood ratio tests (LRT) or quasi-likelihood F-tests (QLF) tests, with a threshold of p-value < 0.005 for M~N comparison. A total of 366 and 297 exons were obtained, respectively, giving a global number of 379 exons (differentially present exons identified only by LRT, only by QLF and by both statistical methods). MA plots for selected DPE are represented in Figure 3, wherein are highlighted those exons which are differentially over-present (i.e. wherein the exon levels are significantly increased) in the N group or differentially over-present in the M group.

DPE clustering

Clusterization of normalized quantification values of the 379 identified DPE was performed by Ward's method. The resulting tree is included in Figure 4. As observed, patients are mostly grouped properly, keeping the M and the N samples separated.

Regarding Principal Components Analysis (PCA) of DPE, Figure 5 represents a bidimensional plot with the two first Principal Components. This figure shows that M and N groups are clearly separated and cluster properly. U patients are located between the limits of both groups, supporting the idea that patients belonging to the U group share characteristics with metastatic as well as non-metastatic patients. These results pointed us to the possibility to obtain a predictive algorithm to classify samples from patients. To achieve this goal, a Random Forest (RF) classification was obtained after 100 iterations extracting two randomly selected samples from each group (M and N) and generating a predictive model with the 16 remaining samples (eight per group) as a 'training set'. A checking test was performed to confirm whether the algorithm was able to classify extracted samples into their corresponding groups of origin, calculating the average probabilities of belonging to one group or another (Table 3). Thus, extracted samples were correctly identified, with the highest mean probability being 0.68.

Subsequently, we analyzed in which group U patients were classified by this algorithm, but the probabilities obtained were near 0.5 in all cases (data not shown). Of the U group, there were two patients (p04 and p58) who developed metastasis during the follow-up period and two other patients (p63 and p66) which were at very high risk of local recurrence and peritoneal metastasis according to eligibility criteria, so they underwent prophylactic treatment, namely second-look surgery plus hyperthermic intraperitoneal chemotherapy (HIPEC), see Cortes- Guiral et al., World J Gastroenterol 2017, 23(3), 377-381. These four patients were then correctly classified by our DPE algorithm as belonging to the M group.

Claims

. An in vitro method for identifying exons which are differentially present in patients having metastasis (DPE) with respect to patients having localized cancer disease, wherein said method comprises the following steps:

i. quantifying individual exons by whole-exome sequencing of the circulating cell-free DNA (cfDNA) by next-generation sequencing in a blood, plasma or serum sample obtained from a cancer patient or group of patients with localized disease (N group); ii. quantifying individual exons by whole-exome sequencing of the cfDNA by next- generation sequencing in a blood, plasma or serum sample obtained from a cancer patient or group of patients with metastasized disease (M group);

iii. comparing the quantification values in i) with the quantification values in ii) and identify as differentially present exons those which abundance differs in a statistically significant manner between both groups.

The method according to claim 1 , wherein exome capturing for whole-exome sequencing is conducted by a method comprising hybridization of cfDNA in said blood, plasma or serum sample with probes substantially complementary to substantially all the coding DNA sequences in the patient's species genome.

The method according to any of the precedent claims, wherein exon quantification is conducted by counting the number of sequence reads obtained by whole-exome sequencing that map to known exons when these are aligned to a reference genome sequence.

The method according to any of the preceding claims, wherein quantification values of DPE have been normalized, preferably data normalization has been conducted by the trimmed mean of M values (TMM) method.

An in vitro method of prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease, the method comprising:

i. quantifying individual exons identified as differentially present in metastatic patients

(DPE) according to a method as defined in any of claims 1 to 4 in a blood, plasma or serum sample obtained from said patient; and ii. comparing the quantification values in the patient's sample obtained in i) with the quantification values in a reference sample, wherein said reference sample is isolated from a patient or group of patients suffering from localized cancer disease; wherein when the quantification values in the patient's sample are increased or decreased in comparison with those in the reference sample the patient has a high risk of suffering from metastasis.

6. The method according to claim 5, wherein DPE are quantified by a method selected from the group consisting of next generation sequencing, quantitative PCR (qPCR), PCR- pyrosequencing, PCR-ELISA, DNA microarrays, branched DNA, dot-blot, Fluorescence In

Situ Hybridization assay (FISH), and multiplex versions of said methods.

7. The method according to any of claims 5 or 6, wherein DPE are quantified by whole-exome sequencing using next generation sequencing.

8. The method according to any of the precedent claims, wherein said patient is a human patient.

9. The method according to any of the precedent claims, wherein said cancer is a solid tumor selected from the group consisting of lung cancer, sarcoma, malignant melanoma, 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, mesothelioma and lymphoma.

10. The method according to any of the precedent claims, wherein said cancer is colorectal cancer, preferably adenocarcinoma.

1 1 . The method according to any of the precedent claims, wherein said differentially present exons are selected from the group consisting of the 379 exons found to be differentially present in colorectal cancer defined in Table 1 .

12. An in vitro method for selecting a treatment for a patient having localized cancer disease wherein said method comprises selecting a treatment according to the classification of said patient according to its prognosis or risk of metastasis by a method as defined in any of claims 5 to 1 1.

13. A method according to any of the preceding claims, wherein said method further comprises storing the method results in a data carrier, preferably wherein said data carrier is a computer readable medium. 14. A computer implemented method, wherein the method is as defined in any of claims 1 to 13.

15. Use of a kit for the prognosis or for predicting the risk of suffering from metastasis in a patient having localized cancer disease on the basis of the differential presence of exons in circulating cell-free DNA (cfDNA) according to a method of any of claims 5 to 13, said kit comprising:

a) a reagent for quantifying individual exons in a blood, plasma or serum sample;

b) optionally, instructions for the use of said reagent(s) in determining the levels of said exons, in a blood, plasma or serum sample.