Multi-cancer detection assay

Abstract

The present invention in general relates to the field of cancer diagnosis and / or monitoring. More in particular, the current invention provides a multi-cancer diagnosis or detection method by determining the DNA methylation status in a sample from a subject suspected of cancer. In particular, the method of the present invention determines the methylation status of CpG dinucleotides across 3 or more selected CpG sites as disclosed herein.

Description

[0001] MULTI-CANCER DETECTION ASSAY

[0002] FIELD OF THE INVENTION

[0003] The present invention in general relates to the field of cancer diagnosis and / or monitoring. More in particular, the current invention provides a multi-cancer diagnosis or detection method by determining the DNA methylation status in a sample from a subject suspected of cancer. In particular, the method of the present invention determines the methylation status of CpG dinucleotides across 3 or more selected CpG sites as disclosed herein.

[0004] BACKGROUND TO THE INVENTION

[0005] Cancer remains one of the deadliest diseases worldwide. Breast, lung, colorectal, and prostate cancer are amongst the most common cancers, with each over 1 .4 million cases per year. Diagnosis occurs regularly in an advanced disease stage due to the lack of clear symptoms and the absence of screening programs for most cancer types. This is reflected in the percentages of late-stage diagnoses, for example, 68% and 59% in lung and colorectal cancer respectively.

[0006] Besides tumor specific mutation analysis, an interesting biomarker candidate for cancer detection is DNA methylation. Genome wide epigenetic reprogramming of tumors occurs early in carcinogenesis. Methylation patterns of many tumor types are widely dysregulated compared to that of healthy cells, but the tumor type specific patterns are very distinctive. Many researchers have been investigating the potential of the methylome in recent years but until now, there are few successful methylation biomarkers for cancer in a clinical setting. Previously, the potential of methylation as a diagnostic biomarker has been shown (Ibrahim et al. 2022) by demonstrating the possibility to discriminate 14 different cancer types from control tissue and each other using methylation biomarkers in silico, with high sensitivity and specificity.

[0007] However, there is still an unfulfilled need for effective, accurate and sensitive diagnostic biomarkers for multi-cancer to improve survival rate in these patients. Therefore, it was an object of the present invention to provide a screening tool that allows a high and accurate prediction of a subject likely suffering from (multi-)cancer. The present invention addresses this need by providing a method to determine the methylation pattern of CpG dinucleotides across a combination of CpG locations which, combined, discriminate a sample from a case with a (multi-)cancer condition with a sample from a normal healthy subject.

[0008] SUMMARY OF THE INVENTION

[0009] In a first aspect, the present invention provides a multi-cancer diagnosis or detection method comprising the step of: determining in a sample from a subject suspected of having cancer the methylation status of the CpG dinucleotides across at least 3 CpG locations selected from the group comprising: SEQ ID N°: 1 - 2316.

[0010] In a specific embodiment, said methylation status is measured in at least 5 CpG locations, in particular in at least 10 CpG locations, more in particular in at least 20 CpG locations, even more in particular in at least 50 CpG locations, most in particular in at least 100 CpG locations.

[0011] In another embodiment, the method further comprises a step of comparing the methylation status of said CpG locations in said sample with the methylation status of said CpG locations in at least one reference sample.

[0012] In a very specific embodiment, an alteration in the methylation status of said CpG locations in said sample compared to the methylation status of said CpG locations in at least one reference sample is indicative of said subject suffering from cancer.

[0013] In yet a further embodiment, said alteration in methylation status is an increase in the percentage of methylation compared to said reference sample.

[0014] In another embodiment, said increase in methylation status is at least 20%, preferably at least 25%, more preferably at least 30% increase in methylation compared to said reference sample.

[0015] In a further embodiment, the methylation status of said sample is at least 25%, preferably 30%, more preferably 40%, even more preferably 50% higher compared to the methylation status of a normal sample from a subject not suffering from cancer.

[0016] In yet another embodiment of the method of the present invention, the methylation status of the CpG dinucleotides is determined across at least the CpG locations selected from: set 1 : chr10:101287802-101287891 (SEQ ID N° 470), chr15:68121 108-68121 197 (SEQ ID N° 1024), chr2: 73147755-73147844 (SEQ ID N° 1482); or set 2: chr1 :197888429-197888518 (SEQ ID N° 337), chr1 :170630499-170630588 (SEQ ID N° 331 ), chr1 1 :31846774-31846863 (SEQ ID N 667); or set 3: chr2:73152605-73152694 (SEQ ID N° 1483), chr4:41882097-41882186 (SEQ ID N° 1761 ), chr10:109674269-109674358 (SEQ ID N°51 1 ); or set 4: chr2:73147755-73147844 (SEQ ID N° 1482), chr5:76923876-76923965 (SEQ ID N° 1880), chr7:8482030-84821 19 (SEQ ID N° 263); or

[0017] Set 5: chr1 :197888439-197888528 (SEQ ID N° 338), chr1 :17475757-17475846 (SEQ ID N° 1 ), chr 10:101287802-101287891 (SEQ ID N° 470); or

[0018] Set 6: chr5:76923876-76923965 (SEQ ID N° 1880), chr4:41882097-41882186 (SEQ ID N° 1761 ), chr2:73147755-73147844 (SEQ ID N° 1482); or

[0019] Set 7: chr2:87016651 -87016740 (SEQ ID N° 1490), chr2:73147755-73147844 (SEQ ID N° 1482), chr10:109674269-109674358 (SEQ ID N° 51 1 ); or

[0020] Set 8: chr1 :197888439-197888528 (SEQ ID N° 338), chr1 1 :320091 17-32009206 (SEQ ID N° 669), chr1 :170630499-170630588 (SEQ ID N° 331 ); or

[0021] Set 9: chr1 :197888429-197888518 (SEQ ID N° 337), chr13:95354716-95354805 (SEQ ID N° 894), chr1 :17475757-17475846 (SEQ ID N° 1 ); or

[0022] Set 10: chr1 :17475757-17475846 (SEQ ID N° 1 ), chr7:35301 123-35301212 (SEQ ID N° 2129), chr10:109674269-109674358 (SEQ ID N° 51 1 ); or

[0023] Set 1 1 : chr2:73147755-73147844 (SEQ ID N° 1482), chr10: 102894297-102894386 (SEQ ID N° 481 ), chr8:109095906-109095995 (SEQ ID N° 2182); or

[0024] Set 12: chr2:73147755-73147844 (SEQ ID N° 1482), chr10:102894297-102894386 (SEQ ID N° 481 ), chr8:109093380-109093469 (SEQ ID N° 2181 ); or

[0025] Set 13: chr15:68121108-68121 197 (SEQ ID N° 1024), chr10:102894297-102894386 (SEQ ID N° 481 ), chr8:109093380-109093469 (SEQ ID N° 2181 ); or

[0026] Set 14: chr15:68121108-68121 197 (SEQ ID N° 1024), chr10:102894297-102894386 (SEQ ID N° 481 ), chr8:109095906-109095995 (SEQ ID N° 2182); or

[0027] Set 15: chr10:101287802-101287891 (SEQ ID N° 470), chr2:73147755-73147844 (SEQ ID N° 1482), chr15:68121 108-68121 197 (SEQ ID N° 1024); or

[0028] Set 16: chr1 :197888429-197888518 (SEQ ID N° 337), chr10:101287802-101287891 (SEQ ID N° 470), chr2:73147755-73147844 (SEQ ID N° 1482); or

[0029] Set 17: chr10:109674269-109674358 (SEQ ID N° 51 1 ), chr2:73147755-73147844 (SEQ ID N° 1482), chr10:102894297-102894386 (SEQ ID N° 481 ); or

[0030] Set 18: chr10:109674269-109674358 (SEQ ID N° 51 1 ), chr10:101287802-101287891 (SEQ ID N° 470), chr2:73147755-73147844 (SEQ ID N° 1482); or

[0031] Set 19: chr5:76923876-76923965 (SEQ ID N° 1880), chr1 :197888429-197888518 (SEQ ID N° 337), chr2:73147755-73147844 (SEQ ID N° 1482); and / or

[0032] Set 20: chr5:76924134-76924223 (SEQ ID N° 1881 ), chr1 :197888429-197888518 (SEQ ID N° 337), chr2:73147755-73147844 (SEQ ID N° 1482); optionally in combination with the methylation status of the CpG dinucleotides determined across at least one or more CpG locations selected from the group comprising: SEQ ID N°: 1 - 2316.

[0033] In a very specific embodiment of the method of the present invention, the methylation status is determined of the CpG dinucleotides across at least the CpG locations of set 4: chr2:73147755- 73147844 (SEQ ID N° 1482), chr5:76923876-76923965 (SEQ ID N° 1880), chr7:8482030- 84821 19 (SEQ ID N° 263).

[0034] In a specific embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across the CpG locations of at least any one of set 1 to set 20, or any combination thereof; wherein an increase in the methylation status of at least one set compared to the methylation status of a reference sample is indicative for a subject suffering from cancer.

[0035] In another specific embodiment, said sample is a solid tissue biopsy and / or a body fluid biopsy, preferably a body fluid biopsy.

[0036] In a very specific embodiment, said body fluid biopsy is selected from the list comprising: blood, serum, plasma, saliva, urine, stool, or a combination thereof.

[0037] In another embodiment, said sample is a DNA sample.

[0038] In a very specific embodiment, the methylation status of the CpG dinucleotides determined across the CpG locations are measured by a method selected from the group comprising IMPRESS (Improved multiplex Methylation Profiling using Restriction Enzymes and smMIP sequencing ) or ddPCR (droplet digital PCR).

[0039] BRIEF DESCRIPTION OF THE DRAWINGS

[0040] With specific reference now to the figures, it is stressed that the particulars shown are by way of example and for purposes of illustrative discussion of the different embodiments of the present invention only. They are presented in the cause of providing what is believed to be the most useful and readily description of the principles and conceptual aspects of the invention. In this regard no attempt is made to show structural details of the invention in more detail than is necessary for a fundamental understanding of the invention. The description taken with the drawings making apparent to those skilled in the art how the several forms of the invention may be embodied in practice.

[0041] Fig. 1 : Sample distribution of normalized counts. All nine different tissue types are displayed in distinct colors in the Tukey boxplot. For every tissue type, the sample distribution of the sum of the normalized counts for tumor (T) and normal-adjacent tissue (N) are shown. This value is the ratio of the sum of the counts of the CpG smMIPs and the sum of the counts of the reference smMIPs. Statistical analysis using the Mann-Whitney U test was performed to compare the average of T and N. Significance is indicated with asterisks (* = p-value < 0.05, “ = p-value < 0.01 , “* = p-value < 0.001 and ““ = p- value < 0.0001).

[0042] Fig. 2: ROC curves of the three selected target sites for both the IMPRESS assay and the ddPCR assay. The left panels are the ROC curves of the IMPRESS assay. The right panels are the ROC curves of the ddPCR assay. For each model, the 5-fold cross validation ROC curves (thin lines) and the mean ROC curve (bold line) are plotted. Both assays are targeting 3 CpG sites: location chr2:73147755- 73147844 - SEQ ID N° 1482 (target 1 -Fig. 2A), chr5:76923876-76923965 - SEQ ID N° 1880 (target 2 - Fig. 2B) and chr7:8482030-84821 19 - SEQ ID N° 263 (target 3 - Fig. 2C). AUC values are shown for each ROC curve.

[0043] Figs. 3-12: ROC curves of 10 combinations of each 3 selected target sites. Figures 3 to 12 represent combinations of three target locations which were in the list of top 200 best performing smMIP targets (non-exhaustively generated) and refer to respectively set 1 -10 as described in the application. For each model, the 5-fold cross validation ROC curves (thin lines) and the mean ROC curve (bold line) are plotted. The ROC curves with AUC values for each of the three target sites are shown as well as a common denominator AUC value.

[0044] Fig. 13: Density plot of the mean cross validated AUC (cvAUC) values for each single smMIP model. The distribution of cvAUC values is centered around 0.8 and is left-skewed. It is clear that the majority of smMIP models can make an accurate prediction.

[0045] Fig. 14-23: Cross-validated ROC curves of the 10 best performing three-probe models out of the top 100 probes (exhaustively generated). The panels are the ROC curves for which the 5-fold cross validation ROC curves (thick black lines) and the mean ROC curve (grey line) are plotted. Average AUC values are shown for each single-probe models ROC curve.

[0046] DETAILED DESCRIPTION OF THE INVENTION

[0047] As already detailed herein above, the present invention provides a method to determine the methylation status of a combination of CpG locations which, combined, discriminates a sample of a case with a (pan-)cancer condition from a sample of a normal healthy subject with high sensitivity and specificity.

[0048] Accordingly, in a first aspect, the present invention provides a method for diagnosing or detecting a multi-cancer condition, or stated otherwise the present invention provides a multi-cancer method for diagnosing or detecting cancer in a subject suspected of having cancer comprising the step of: determining in a sample from a subject suspected of having cancer the methylation status of CpG dinucleotides across at least 3 CpG locations selected from the group comprising: SEQ ID N°: 1 - 2316.

[0049] The inventors have thus developed a highly sensitive and targeted multi-cancer detection assay to discriminate tumor samples from normal samples using methylation biomarkers. For the selection of the biomarker targets, methylation data of tissue samples from eight of the most lethal cancer types worldwide and normal adjacent tissue were used. Selection of relevant detection biomarkers allows the assay to be either multi-cancer (capable of detecting multiple different types of cancer) or cancer specific. The method according to the invention has an increased sensitivity and specificity compared to current existing diagnostic methods to detect cancers (e.g. mutations and protein analyses). Since methylation occurs very early in cancer development, possibly before actual neoplastic transformation, the method according to the invention is particularly suitable for early cancer diagnosis.

[0050] The method of the invention also has the advantage of being applicable to various types of tissue such as solid tissue and / or tumor samples as well as liquid samples. As this technology will be much more sensitive compared to current protocols, it has also applications in the detection of methylated DNA fragments in liquid biopsies. Accordingly, in a specific embodiment, said sample is a solid tissue biopsy or a body fluid biopsy from a subject, preferably a body fluid biopsy. As stated already, this novel multicancer detection assay is specifically suitable as a multi-cancer detection assay, which allows for high resolution methylation detection in tissues, plasma or other biological matrices of cancer patients. In a very specific embodiment, said body fluid biopsy is selected from the list comprising: blood, serum, plasma, saliva, urine, stool, or a combination thereof. For example, methylation patterns of normal blood samples were included for the target selection, to end up with a biomarker panel that is also suitable for use in plasma derived liquid biopsies wherein even cell-free DNA can be used for the diagnosis.

[0051] In another embodiment, said sample is a DNA sample. A DNA sample in the context of the invention means any biological specimen that is obtained or retained for the purpose of extracting and analyzing DNA to perform the method of the present invention. ‘DNA sample’ includes DNA extracted from a test sample or reference sample. Alternatively, ‘DNA sample’ refers to a test sample or reference sample, wherein said sample contains DNA.

[0052] As used herein, the term “multi-cancer” or “pan-cancer” refers to a collection of multiple cancer conditions. Although all cancers are molecularly distinct, many share common genetic driver abnormalities. As used herein, the diagnosis or detection of a multi-cancer condition involves assessing the methylation status of genes common to many different cancers including both solid tumors and hematologic malignancies, regardless of tumor origin. In a particular embodiment, the method of the present invention is suitable to diagnose a patient suffering from one or more of the following cancers but not limited thereto: lung cancer, colorectal cancer, liver cancer, breast cancer, pancreatic cancer, head & neck cancer, esophageal cancer, prostate cancer.

[0053] In a specific embodiment, the method of the present invention is particularly suitable for detecting or diagnosing one or more cancers of the list comprising: lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), colorectal adenocarcinoma (CRC), liver hepatocellular carcinoma (LIHC), breast invasive carcinoma (BRCA), pancreatic adenocarcinoma (PAAD), head & neck squamous cell carcinoma (HNSC), esophageal carcinoma (ESCA), prostate adenocarcinoma (PRAD).

[0054] The term “multi-cancer biomarkers” or alternatively “pan-cancer biomarkers” is used in the context of robust biomarkers that can detect or diagnose cancer, based on shared methylation patterns between the different cancer types. Several of such methylation biomarkers have been identified focusing mainly on gene promoter markers or single CpG markers but show inconsistent performance across cancer stages and inadequacy for detecting residual disease. As described herein, next-gen sequencing was performed for multi-cancer profiling and validated with targeted digital PCR. The identified panel enabled the inventors to comprehensively analyze a defined set of aberrations, in particular a set of CpG locations as set forth in SEQ ID N°: 1 - 2316, that are associated with many common cancers. Thus, the method of the present invention is able to solve the problem of inadequacy and inaccuracy by specifically determining in a sample the methylation status of CpG dinucleotides across at least 3 CpG locations selected from the group comprising: SEQ ID N°: 1 - 2316. As used herein, the term “CpG” refers to a dinucleotide of DNA wherein a cytosine nucleotide is followed by a guanine nucleotide in the linear sequence of bases along its 5' to 3' direction. Cytosines in CpG dinucleotides can be methylated to form 5-methylcytosines. Methylation of the cytosine within a gene can change its expression, a mechanism that is part of a larger field of gene regulation called epigenetics. As used herein, the term “CpG site” refers to one specific dinucleotide genomic coordinate. If individual CpG sites occur with high frequency in a particular genomic region it is herein referred to as CpG islands (or CG islands). For example, in humans, about 70% of promoters located near the transcription start site of a gene (proximal promoters) contain a CpG island.

[0055] On the other hand, the term “CpG locations” as described herein, refers to a specific region of the genome of several base pairs (typically about 90) which comprises various numbers of CpG sites. For example, SEQ ID N°:2 is a genomic region on the positive strand of chromosome 1 demarcated by genomic coordinates 17475798 and 17475887 and has sequence GCCCGGCCCGGCCCGGGAGTGAGCCTGCAGCGAGGGATTAGCGCGGTAATAGCCGGGATTAG CGCAGGCTGCGAGCGCGTTAGTCACTAA which comprising 1 1 CpG sites or CpG dinucleotides. Accordingly, each defined CpG locations as defined in SEQ ID N°: 1 -2316 will comprise at least 1 or more CpG sites that are measured to determine the methylation status. As described herein, the nucleic acid sequences are based on Human Genome Build GRCh37 / hg19.

[0056] As used herein, the term “methylation status” is to be understood as the average methylation percentage of one or more CpG sites determined within one or more CpG locations. The methylation status is determined as:

[0057] Normalized count smMIP i in sample

[0058] When referring to the Beta value (Illumina methylation arrays) the methylation status is described as the following: b=M / (M+U+a), where M>0 and U>0 denote the methylated and unmethylated signal intensities, respectively, measured by the Illumina 450k array. The offset a>0 is usually set equal to 100 and is added to M+U to stabilize beta values when both M and U are small.

[0059] It is important to note that for arrays, single CpG sites are quantified individually and not “one or more CpG sites”.

[0060] For example, the methylation status is determined in a sample from a subject suspected having (pan- )cancer by assessing the average percentage of methylation of CpG location 1 (CpG1 , CpG2, ... CpGn), CpG location 2 (CpG1 , CpG2, ... CpGn), CpG location 3 (CpG1 , CpG2, ... CpGn), etc.

[0061] In a specific embodiment, the methylation status of a (test) sample from a subject suspected of having cancer is compared to the methylation status of a normal or reference sample. Said normal or reference sample, may be a tissue biopsy from the same subject, but from an organ or site of an organ which is not expected to be tumor tissue. Alternatively, the normal or reference sample may be obtained from a healthy subject, or at least a subject not suspected to have cancer. The reference sample may also be a reference curve obtained from a set of reference samples, which would then not require the comparison with a fresh reference sample each time a new patient sample is to be analyzed. In this embodiment, the methylation status of the CpG sites in the test sample is preferably compared to the methylation status of the same CpG sites in the reference or normal sample.

[0062] In a specific embodiment, an alteration in the methylation status of the CpG dinucleotides across said CpG locations in the test sample compared to the methylation status of the CpG dinucleotides across said CpG sites in a reference or normal sample is indicative of said subject suffering from cancer.

[0063] Alternatively, the methylation status of specific CpG sites in the test sample may also be compared to least one reference site. As used herein, the term “reference site” refers to a genomic location without a CpG site. These reference sites are included to allow normalization of the methylation percentage per sample and to estimate the effective total amount of input DNA. In a specific embodiment, an alteration in the methylation status of said CpG locations compared to the methylation status of said reference site is indicative of said subject suffering from cancer.

[0064] For example, methylation status can be obtained from the CpG dinucleotides across at least 3 CpG location which are selected from the list comprising SEQ ID N°: 1 -2316 in the test sample and reference sample. Alternatively, methylation status can be obtained from the CpG dinucleotides across at least 3 CpG location selected from the list comprising SEQ ID N°: 1 -2316 compared to at least 1 , 2, 3, ...to about 600 reference site locations.

[0065] In a particular embodiment, DNA input may be normalized by the sum of CpG locations divided by the sum of reference locations for each sample.

[0066] In a further embodiment, said alteration in methylation status in the test sample is an increase in the percentage of methylation compared to said reference sample or reference site.

[0067] In case methylation-sensitive restriction enzymes (MSRE) are used, the reference site refers to a genomic location without a CpG site and without an enzyme recognition site. Alternatively, a reference sample may include similar CpG sites, however, would typically not be methylated in such sites. The MSREs cleave unmethylated DNA at their recognition sites, while methylated CpG sites block the restriction enzymes which results in unaffected CpG regions. In a particular embodiment, a combination of restriction enzymes can be used such as but not limited to Hpall, HpyCH4IV, Acil and / or HinP1 1. Each MSRE recognition site may have a CpG site in the middle (CACGG, AACGT, CACGC and GACGC). During digestion, unmethylated recognition sites are cleaved while methylated CpG sites blocked the restriction enzymes which resulted in uncut CpG sites. In another embodiment, said increase in methylation status is at least 20%, such as at least 21 %, at least 22%, at least 23%, at least 24%, preferably at least 25%, at least 26%, at least 27%, at least 28%, at least 29% more preferably at least 30%, at least 31 %, at least 32%, at least 33%, at least 34%, at least 35%, at least 36%, at least 37%, at least 38%, at least 39%, at least 40% compared to said reference sample or reference site.

[0068] In a particular embodiment, the methylation status is determined in a sample from a subject suspected of having cancer of the CpG dinucleotides across at least 3 CpG locations selected from the group comprising: SEQ ID N°: 1 - 2316.

[0069] In a specific embodiment, said methylation status of the CpG dinucleotides is measured across at least 3 CpG locations such as at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, at least 10, at least 1 1 , at least 12, at least 13, at least 14, at least 15, at least 16, at least 17, at least 18, at least 19, at least 20, at least 21 , at least 22, at least 23, at least 24, at least 25, at least 26, at least 27, at least 28, at least 29, at least 30, at least 31 , at least 32, at least 33, at least 34, at least 35, at least 36, at least 37, at least 38, at least 39, at least 40, at least 41 , at least 42, at least 43, at least 44, at least 45, at least 46, at least 47, at least 48, at least 49, at least 50 CpG locations selected from the list comprising SEQ ID N°: 1 -2316. In a particular embodiment, the methylation status is measured in 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, or at least 150, at least 200, at least 250, at least 300 CpG locations selected from the list comprising SEQ ID N°: 1 -2316.

[0070] In a very specific embodiment, the methylation status of the CpG dinucleotides is measured across at least 300 such as at least 310, at least 320, at least 330, at least 340, at least 350, preferably about 354 CpG locations selected from the list comprising SEQ ID N°: 1 -2316.

[0071] In another specific embodiment, the methylation status of the CpG dinucleotides is determined across at least 3 CpG locations, such as across at least 5 CpG locations, in particular across at least 10 CpG locations, in particular across at least 20 CpG locations, more in particular across at least 50 CpG locations, even more in particular across at least 100 CpG locations most in particular across 354 CpG locations, wherein the CpG location are selected from the group comprising: SEQ ID N°: 1 , 6, 9, 36, 54, 62, 77, 79, 82, 83, 84, 86, 88, 89, 105, 107, 1 10, 1 16, 1 18, 1 19, 125, 148, 150, 212, 215, 228, 231 , 233,

[0072] 239, 257, 261 , 262, 263, 264, 265, 297, 300, 301 , 304, 312, 313, 315, 320, 323, 324, 325, 328, 330,

[0073] 331 , 337, 339, 347, 350, 352, 354, 355, 388, 389, 404, 427, 430, 431 , 440, 441 , 442, 445, 447, 451 ,

[0074] 456, 460, 470, 472, 475, 477, 479, 481 , 495, 497, 500, 504, 507, 51 1 , 521 , 524, 530, 533, 534, 572,

[0075] 574, 575, 580, 596, 599, 601 , 602, 613, 614, 615, 617, 625, 627, 632, 635, 636, 637, 649, 651 , 653,

[0076] 657, 663, 665, 667, 671 , 684, 706, 708, 723, 724, 731 , 749, 750, 752, 756, 762, 772, 776, 783, 788,

[0077] 789, 790, 794, 804, 813, 814, 815, 816, 818, 819, 831 , 832, 839, 862, 871 , 874, 882, 888, 890, 892,

[0078] 893, 895, 913, 925, 936, 938, 943, 946, 948, 950, 953, 955, 961 , 965, 967, 969, 982, 986, 989, 992,

[0079] 1001 , 1018, 1024, 1034, 1038, 1040, 1043, 1070, 1077, 1079, 1094, 1098, 1 105, 1 106, 1 108, 1 1 17, 1 1 18, 1 126, 1 170, 1 177, 1 189, 1 191 , 1 192, 1214, 1219, 1239, 1242, 1243, 1247, 1254, 1255, 1260,

[0080] 1280, 1281 , 1286, 1298, 1299, 1304, 1306, 1322, 1333, 1335, 1345, 1346, 1359, 1382, 1393, 1400,

[0081] 1401 , 1407, 1416, 1418, 1423, 1424, 1426, 1428, 1429, 1431 , 1438, 1441 , 1462, 1467, 1482, 1483,

[0082] 1487, 1494, 1495, 1510, 1533, 1585, 1594, 1598, 1599, 1602, 1603, 1604, 1605, 1611 , 1613, 1614,

[0083] 1620, 1622, 1626, 1628, 1631 , 1634, 1656, 1658, 1664, 1666, 1682, 1742, 1746, 1751 , 1754, 1759,

[0084] 1761 , 1763, 1764, 1765, 1789, 1810, 1818, 1826, 1830, 1847, 1848, 1851 , 1865, 1871 , 1872, 1873,

[0085] 1880, 1881 , 1885, 1897, 1898, 1900, 1901 , 1906, 1908, 1909, 1910, 1912, 1951 , 1955, 1956, 1960,

[0086] 1963, 1973, 1979, 1981 , 1983, 1986, 1990, 1993, 2005, 2015, 2018, 2020, 2022, 2026, 2036, 2041 ,

[0087] 2042, 2048, 2053, 2054, 2058, 2064, 2065, 2069, 2079, 2081 , 2082, 2085, 2088, 2089, 2107, 2108, 2126, 2129, 2130, 2140, 2141 , 2149, 2151 , 2153, 2156, 2157, 2169, 2181 , 2182, 2186, 2191 , 2206, 2208, 2224, 2225, 2247, 2257, 2260, 2267, 2277, 2285, 2315.

[0088] In a preferred embodiment, the at least 3 CpG locations are located within the genomic human EMX1 , Chr5q14.1 or NXPH1 DNA, including the respective promoter region, and DNA within 5, 4, 3, 2 or 1 kb upstream and downstream thereof.

[0089] In a particular embodiment of the method of the present invention, the methylation status of the CpG dinucleotides is determined across at least the CpG locations selected from: set 1 : chr10:101287802-101287891 (SEQ ID N° 470), chr15:68121 108-68121 197 (SEQ ID N° 1024), chr2: 73147755-73147844 (SEQ ID N° 1482); or set 2: chr1 :197888429-197888518 (SEQ ID N° 337), chr1 :170630499-170630588 (SEQ ID N° 331 ), chr1 1 :31846774-31846863 (SEQ ID N° 667); or set 3: chr2:73152605-73152694 (SEQ ID N° 1483), chr4:41882097-41882186 (SEQ ID N° 1761 ), chr10:109674269-109674358 (SEQ ID N° 51 1 ); or set 4: chr2:73147755-73147844 (SEQ ID N° 1482), chr5:76923876-76923965 (SEQ ID N° 1880), chr7:8482030-84821 19 (SEQ ID N° 263); or

[0090] Set 5: chr1 :197888439-197888528 (SEQ ID N° 338), chr1 :17475757-17475846 (SEQ ID N° 1 ), chr 10:101287802-101287891 (SEQ ID N° 470); or

[0091] Set 6: chr5:76923876-76923965 (SEQ ID N° 1880), chr4:41882097-41882186 (SEQ ID N° 1761 ), chr2:73147755-73147844 (SEQ ID N° 1482); or

[0092] Set 7: chr2:87016651 -87016740 (SEQ ID N° 1490), chr2:73147755-73147844 (SEQ ID N° 1482), chr10:109674269-109674358 (SEQ ID N° 51 1 ); or

[0093] Set 8: chr1 :197888439-197888528 (SEQ ID N° 338), chr1 1 :320091 17-32009206 (SEQ ID N° 669), chr1 :170630499-170630588 (SEQ ID N° 331 ); or

[0094] Set 9: chr1 :197888429-197888518 (SEQ ID N° 337), chr13:95354716-95354805 (SEQ ID N° 894), chr1 :17475757-17475846 (SEQ ID N° 1 ); and / or

[0095] Set 10: chr1 :17475757-17475846 (SEQ ID N° 1 ), chr7:35301 123-35301212 (SEQ ID N° 2129), chr10:109674269-109674358 (SEQ ID N° 511 ). Reference is made to Figures 3-12 of the application as filed. In another embodiment of the method of the present invention, the methylation status of the CpG dinucleotides is determined across at least the CpG locations selected from:

[0096] Set 1 1 : chr2:73147755-73147844 (SEQ ID N° 1482), chr10: 102894297-102894386 (SEQ ID N° 481 ), chr8:109095906-109095995 (SEQ ID N° 2182); or

[0097] Set 12: chr2:73147755-73147844 (SEQ ID N° 1482), chr10:102894297-102894386 (SEQ ID N° 481 ), chr8:109093380-109093469 (SEQ ID N° 2181 ); or

[0098] Set 13: chr15:68121108-68121 197 (SEQ ID N° 1024), chr10:102894297-102894386 (SEQ ID N° 481 ), chr8:109093380-109093469 (SEQ ID N° 2181 ); or

[0099] Set 14: chr15:68121108-68121 197 (SEQ ID N° 1024), chr10:102894297-102894386 (SEQ ID N° 481 ), chr8:109095906-109095995 (SEQ ID N° 2182); or

[0100] Set 15: chr10:101287802-101287891 (SEQ ID N° 470), chr2:73147755-73147844 (SEQ ID N° 1482), chr15:68121 108-68121 197 (SEQ ID N° 1024); or

[0101] Set 16: chr1 :197888429-197888518 (SEQ ID N° 337), chr10:101287802-101287891 (SEQ ID N° 470), chr2:73147755-73147844 (SEQ ID N° 1482); or

[0102] Set 17: chr10:109674269-109674358 (SEQ ID N° 51 1 ), chr2:73147755-73147844 (SEQ ID N° 1482), chr10:102894297-102894386 (SEQ ID N° 481 ); or

[0103] Set 18: chr10:109674269-109674358 (SEQ ID N° 51 1 ), chr10:101287802-101287891 (SEQ ID N° 470), chr2:73147755-73147844 (SEQ ID N° 1482); or

[0104] Set 19: chr5:76923876-76923965 (SEQ ID N° 1880), chr1 :197888429-197888518 (SEQ ID N° 337), chr2:73147755-73147844 (SEQ ID N° 1482); and / or

[0105] Set 20: chr5:76924134-76924223 (SEQ ID N° 1881 ), chr1 :197888429-197888518 (SEQ ID N° 337), chr2:73147755-73147844 (SEQ ID N° 1482).

[0106] Reference is made to Figures 14-23 of the application as filed.

[0107] Evidently, in these embodiments wherein the methylation status of the CpG dinucleotides is determined across at least 3 CpG locations (one set), it can be optionally combined with determining the methylation status of the CpG dinucleotides determined across at least one or more other sets; or one or more CpG locations selected from the group comprising: SEQ ID N°: 1 - 2316.

[0108] In another particular embodiment of the method of the present invention, the methylation status of the CpG dinucleotides is determined across at least the 3 CpG locations selected from any one of set 1 to set 1 13 as defined herein, or any combination thereof.

[0109] In a very specific embodiment of the method of the present invention, the methylation status of the CpG dinucleotides is determined across at least the 3 CpG locations of set 4: chr2:73147755-73147844 (SEQ ID N° 1482), chr5:76923876-76923965 (SEQ ID N° 1880), chr7:8482030-84821 19 (SEQ ID N° 263).

[0110] In a specific embodiment, the method of the present invention comprises determining the methylation status of CpG dinucleotides across at least 3 CpG locations, wherein an increase in the methylation status of at least one set compared to a reference sample or reference site is indicative for a subject suffering from cancer.

[0111] In a specific embodiment, the method of the present invention comprises determining the methylation status of CpG dinucleotides across at least 3 CpG locations of any one of set 1 to 1 13 or any combination thereof; wherein an increase in the methylation status of at least one set compared to a reference sample or reference site is indicative for a subject suffering from cancer.

[0112] In another specific embodiment, the method of the present invention comprises determining the methylation status in CpG dinucleotides across 3 or more CpG locations selected from the following list: SEQ ID N° 1 , SEQ ID N° 263, SEQ ID N° 328, SEQ ID N° 331 , SEQ ID N° 332, SEQ ID N° 337, SEQ ID N° 338, SEQ ID N° 404, SEQ ID N° 447, SEQ ID N° 470, SEQ ID N°: 481 , SEQ ID N°51 1 , SEQ ID N 667, SEQ ID N° 669, SEQ ID N° 894, SEQ ID N° 1024, SEQ ID N° 1286, SEQ ID N° 1482, SEQ ID N° 1483, SEQ ID N° 1487, SEQ ID N° 1490, SEQ ID N° 1761 , SEQ ID N° 1880, SEQ ID N° 1881 , SEQ ID N° 2129, SEQ ID N°: 2181 , SEQ ID N°: 2182.

[0113] In a particular embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least 3 CpG locations wherein at least one CpG location is chr2:73147755-73147844 (SEQ ID N° 1482), optionally in combination with at least one or more other CpG locations selected from the list comprising: SEQ ID N°: 1 -2316, in particular SEQ ID N°: 1 , 6, 9, 36, 54, 62, 77, 79, 82, 83, 84, 86, 88, 89, 105, 107, 1 10, 1 16, 1 18, 1 19, 125, 148, 150, 212, 215, 228,

[0114] 231 , 233, 239, 257, 261 , 262, 263, 264, 265, 297, 300, 301 , 304, 312, 313, 315, 320, 323, 324, 325,

[0115] 328, 330, 331 , 337, 339, 347, 350, 352, 354, 355, 388, 389, 404, 427, 430, 431 , 440, 441 , 442, 445,

[0116] 447, 451 , 456, 460, 470, 472, 475, 477, 479, 481 , 495, 497, 500, 504, 507, 51 1 , 521 , 524, 530, 533,

[0117] 534, 572, 574, 575, 580, 596, 599, 601 , 602, 613, 614, 615, 617, 625, 627, 632, 635, 636, 637, 649,

[0118] 651 , 653, 657, 663, 665, 667, 671 , 684, 706, 708, 723, 724, 731 , 749, 750, 752, 756, 762, 772, 776,

[0119] 783, 788, 789, 790, 794, 804, 813, 814, 815, 816, 818, 819, 831 , 832, 839, 862, 871 , 874, 882, 888,

[0120] 890, 892, 893, 895, 913, 925, 936, 938, 943, 946, 948, 950, 953, 955, 961 , 965, 967, 969, 982, 986,

[0121] 989, 992, 1001 , 1018, 1024, 1034, 1038, 1040, 1043, 1070, 1077, 1079, 1094, 1098, 1 105, 1 106, 1 108, 1 1 17, 1 1 18, 1 126, 1 170, 1 177, 1 189, 1 191 , 1 192, 1214, 1219, 1239, 1242, 1243, 1247, 1254, 1255,

[0122] 1260, 1280, 1281 , 1286, 1298, 1299, 1304, 1306, 1322, 1333, 1335, 1345, 1346, 1359, 1382, 1393,

[0123] 1400, 1401 , 1407, 1416, 1418, 1423, 1424, 1426, 1428, 1429, 1431 , 1438, 1441 , 1462, 1467, 1483,

[0124] 1487, 1494, 1495, 1510, 1533, 1585, 1594, 1598, 1599, 1602, 1603, 1604, 1605, 1611 , 1613, 1614,

[0125] 1620, 1622, 1626, 1628, 1631 , 1634, 1656, 1658, 1664, 1666, 1682, 1742, 1746, 1751 , 1754, 1759,

[0126] 1761 , 1763, 1764, 1765, 1789, 1810, 1818, 1826, 1830, 1847, 1848, 1851 , 1865, 1871 , 1872, 1873,

[0127] 1880, 1881 , 1885, 1897, 1898, 1900, 1901 , 1906, 1908, 1909, 1910, 1912, 1951 , 1955, 1956, 1960,

[0128] 1963, 1973, 1979, 1981 , 1983, 1986, 1990, 1993, 2005, 2015, 2018, 2020, 2022, 2026, 2036, 2041 ,

[0129] 2042, 2048, 2053, 2054, 2058, 2064, 2065, 2069, 2079, 2081 , 2082, 2085, 2088, 2089, 2107, 2108, 2126, 2129, 2130, 2140, 2141 , 2149, 2151 , 2153, 2156, 2157, 2169, 2181 , 2182, 2186, 2191 , 2206, 2208, 2224, 2225, 2247, 2257, 2260, 2267, 2277, 2285, 2315. In a further embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least 3 CpG locations, wherein one CpG location is chr2:73147755-73147844 (SEQ ID N°: 1482) and at least one other CpG location is selected from the list comprising chr10: 102894297-102894386 (SEQ ID N° 481 ), chr10:101287802-101287891 (SEQ ID N°: 470), chr10:109674269-109674358 (SEQ ID N° 51 1 ), chr15:68121 108-68121 197 (SEQ ID N° 1024), chr1 :197888429-197888518 (SEQ ID N° 337), chr5:76923876-76923965 (SEQ ID N° 1880), in particular SEQ ID N° 481 , optionally in combination with at least one further CpG location selected from the list comprising SEQ ID N°: 1 - 2316.

[0130] In a specific embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least 3 CpG locations, wherein one CpG location is chr2:73147755-73147844 (SEQ ID N°: 1482), in combination with at least two other CpG locations selected from the list comprising: SEQ ID N°: 481 and 2182, alternatively SEQ ID N°: 481 and 2181 , alternatively SEQ ID N°: 470 and SEQ ID N°: 1024, alternatively, SEQ ID N°: 337and 470, alternatively, SEQ ID N°: 51 1 and 481 , alternatively SEQ ID N°: 51 1 and 470, alternatively SEQ ID N°: 1880 and 337, alternatively SEQ ID N°: 1881 and SEQ ID N°: 337, alternatively SEQ ID N° 1490 and 51 1 , alternatively SEQ ID N° 1880 and 1761 , alternatively SEQ ID N° 1880 and 263, alternatively SEQ ID N° 470 and 1024, optionally in combination with at least one or more CpG locations selected from the list comprising SEQ ID N°: 1 - 2316.

[0131] In another embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least 3 CpG locations, wherein one CpG location is selected from the list comprising: SEQ ID N°: 1 , 337, 470, 481 , 51 1 , 1024 or 1880, and at least two other CpG locations selected from the list comprising SEQ ID N°: 1 - 2316.

[0132] In a specific embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least CpG location SEQ ID N°: 337, in particular in combination with CpG location SEQ ID N°: 1482 and one further CpG locations selected from the list comprising SEQ ID N°: 1 - 2316, but more in particular CpG location SEQ ID N°: 337 in combination with at least two other CpG locations selected from the combinations comprising: SEQ ID N° 331 and 667, alternatively SEQ ID N° 894 and 1 , alternatively SEQ ID N°: 470 and 1482, alternatively SEQ ID N°: 1880 and 1482, alternatively SEQ ID N°: 1881 and 1482.

[0133] In a specific embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least CpG location SEQ ID N°: 470, in particular in combination with CpG location SEQ ID N°: 1482 and one further CpG locations selected from the list comprising SEQ ID N°: 1 - 2316, but more in particular CpG location SEQ ID N°: 470 in combination with at least two other CpG locations selected from the combinations comprising: SEQ ID N° 1024 and 1482, alternatively SEQ ID N° 338 and 1 , alternatively SEQ ID N°: 1482 and 1024, alternatively SEQ ID N°: 337 and 1482, alternatively SEQ ID N°: 1881 and 1482.

[0134] In a specific embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least CpG location SEQ ID N°: 481 , in particular in combination with CpG location SEQ ID N°: 1482 and one further CpG locations selected from the list comprising SEQ ID N°: 1 - 2316, but more in particular CpG location SEQ ID N°: 481 in combination with at least two other CpG locations selected from the combinations comprising: SEQ ID N° 1482 and 2182, alternatively SEQ ID N° 1482 and 2181 , alternatively SEQ ID N°: 1024 and 2181 , alternatively SEQ ID N°: 1024 and 2182, alternatively SEQ ID N°: 51 1 and 1482.

[0135] In a specific embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least CpG location SEQ ID N°: 51 1 , in particular in combination with CpG location SEQ ID N°: 1482 and one further CpG locations selected from the list comprising SEQ ID N°: 1 - 2316, but more in particular CpG location SEQ ID N°: 51 1 in combination with at least two other CpG locations selected from the combinations comprising: SEQ ID N° 1483 and 1761 , alternatively SEQ ID N° 1490 and 1482, alternatively SEQ ID N°: 1 and 2129, alternatively SEQ ID N°: 1482 and 481 , alternatively SEQ ID N°: 470 and 1482.

[0136] In a specific embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least CpG location SEQ ID N°: 1024, in particular in combination with CpG location SEQ ID N°: 1482 and one further CpG locations selected from the list comprising SEQ ID N°: 1 - 2316, but more in particular CpG location SEQ ID N°: 1024 in combination with at least two other CpG locations selected from the combinations comprising: SEQ ID N° 470 and 1482, alternatively SEQ ID N° 481 and 2181 , alternatively SEQ ID N°: 481 and 2182, alternatively SEQ ID N°: 470 and 1482.

[0137] In a specific embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least CpG location SEQ ID N°: 1880, in particular in combination with CpG location SEQ ID N°: 1482 and one further CpG locations selected from the list comprising SEQ ID N°: 1 - 2316, but more in particular CpG location SEQ ID N°: 1880 in combination with at least two other CpG locations selected from the combinations comprising: SEQ ID N° 1482 and 263, alternatively SEQ ID N° 1761 and 1482, alternatively SEQ ID N°: 337 and 1482.

[0138] In a specific embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least CpG location SEQ ID N°: 1 , in particular in combination with two further CpG locations selected from the list comprising SEQ ID N°: 1 - 2316, but more in particular CpG location SEQ ID N°: 1 in combination with at least two other CpG locations selected from the combinations comprising: SEQ ID N° 338 and 470, alternatively SEQ ID N° 337 and 894, alternatively SEQ ID N°: 2129 and 51 1 . Following CpG locations occurred more than 1 ,000 times in the top 20,000 three-probe models (sets): SEQ ID N° 328, SEQ ID N° 331 , SEQ ID N° 332, SEQ ID N° 337, SEQ ID N° 404, SEQ ID N° 447, SEQ ID N° 470, SEQ ID N°: 481 , SEQ ID N°511 , SEQ ID N° 1024, SEQ ID N° 1286, SEQ ID N° 1482, SEQ ID N° 1483, SEQ ID N° 1487, SEQ ID N° 1761 , SEQ ID N° 2129, SEQ ID N°: 2181 , SEQ ID N°: 2182. Therefore, in particular embodiments, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least 3 CpG locations, wherein at least one or more CpG locations are selected from the list comprising: SEQ ID N° 328, SEQ ID N° 331 , SEQ ID N° 332, SEQ ID N° 337, SEQ ID N° 404, SEQ ID N° 447, SEQ ID N° 470, SEQ ID N°: 481 , SEQ ID N°51 1 , SEQ ID N° 1024, SEQ ID N° 1286, SEQ ID N° 1482, SEQ ID N° 1483, SEQ ID N° 1487, SEQ ID N° 1761 , SEQ ID N° 2129, SEQ ID N°: 2181 , SEQ ID N°: 2182, optionally in combination with determining the methylation status of the CpG dinucleotides determined across at least one or more further CpG locations selected from the group comprising: SEQ ID N°: 1 - 2316.

[0139] In a particular embodiment, the method of the present invention comprises determining the methylation status of the CpG dinucleotides across at least 3 CpG locations, wherein at least one or more CpG locations are selected from the list comprising: SEQ ID NO°: 1482, SEQ ID NO°: 51 1 , SEQ ID NO°: 337, SEQ ID NO°: 1024, SEQ ID NO°: 1482,1761 , SEQ ID NO°: 481 , SEQ ID NO°: 1483, SEQ ID NO°: 1487, SEQ ID NO°: 470.

[0140] In a preferred embodiment, said CpG locations are equal to or less than 100, 99, 98, 97, 96, 94, 93, 92, 91 , 90, 89, 88, 87, 86 85, 84, 83, 82, 81 , 80, 79, 78, 77, 76, 75, 74, 73, 72, 71 , 70, 69, 68, 67, 66, 65, 64, 63, 62, 61 or 60 bp long, preferably equal to or less than 100, 97, 92, 90 bp long. Preferred ranges for the region are 50 to 100 bp, more preferably 60 to 100 bp and most preferably about 90 bp. In case of each of the aforementioned lengths, said CpG location comprises at least 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 1 1 , 12, 13, 14 or 15 CpG sites of the genomic DNA

[0141] In a very specific embodiment, the methylation status of the CpG locations are measured by any of the methods selected from the following table: In yet a specific embodiment, the methylation status of the CpG locations or reference sites is measured by IMPRESS or ddPCR (droplet digital PCR).

[0142] As used herein, IMPRESS is a methylation detection technique which combines methylation-sensitive restriction enzymes (MSRE) and single-molecule Molecular Inversion Probes (smMIP) to capture DNA methylation. Detailed description of the technology is cited in WO2022053637. MSREs have been used for a very long time for analysis of methylation in specific regions of the genome. MSREs are used herein to digest genomic DNA. These MSREs are very specific and cut the DNA only when it is unmethylated and not when it is methylated, (see details in WO2022053637) . Alternatively, MSREs which are able to cut the DNA only when it is methylated and not when it is unmethylated may also suitably be used within the context of the present invention. Using MSREs instead of bisulfite, which is a widely used compound used in methylation analysis, has strong advantages. The main advantage is that MSREs do not degrade DNA, in contrast to bisulfite. smMIPs are very efficient to capture and enrich carefully chosen informative regions of the genome and are extremely suitable for multiplex analysis of thousands of genomic regions.

[0143] IMPRESS is particularly suitable in the method of the present invention to determine methylation status and to discriminate tumor samples from normal samples.

[0144] EXAMPLES

[0145] EXAMPLE 1

[0146] In the present example, a novel high-multiplex methylation detection technique called IMPRESS (Improved multiplex Methylation Profiling using Restriction Enzymes and smMIP sequencing) is used. MSREs have been used for a very long time for the analysis of methylation in specific regions of the genome. smMIPs are very efficient to capture and enrich carefully chosen informative regions of the genome and are extremely suitable for multiplex analysis of thousands of genomic regions. In the present case, this technique is used for the development of a diagnostic biomarker assay discriminating tumor samples from normal samples. For the selection of the biomarker targets, methylation data of tissue samples from eight of the most lethal cancer types worldwide and normal adjacent tissue were used. In addition, methylation patterns of normal blood samples were included for the target selection, to end up with a biomarker panel that is also suitable for use in plasma derived liquid biopsies.

[0147] Materials and Methods

[0148] 1. IMPRESS DNA methylation detection technique

[0149] The IMPRESS technique is a combination of MSRE digestion and smMIP sequencing and is described in WO2022053637. In brief, the first step is a combined digestion of the DNA with four MSREs. The MSREs cleave unmethylated DNA at their recognition sites, while methylated CpG sites block the restriction enzymes which results in unaffected CpG regions. The methylated CpG regions are captured by the smMIPs through hybridization of the smMIP binding arms. Elongation and ligation of the smMIP created a circular DNA fragment. All remaining linear fragments were degraded by an exonuclease reaction. Thereafter, the circular fragments are amplified by PCR. Finally, all samples are pooled, purified, and sequenced by Next Generation Sequencing. In each sample, lambda phage DNA was spiked in as internal digestion control.

[0150] 2. Development of multi-cancer biomarker panel

[0151] 2. 1 Target sites selection

[0152] For the development of the multi-cancer detection assay, a panel with candidate methylation biomarkers was built using online available 450K methylation array data (Table 1 ). Methylation data processing and analysis were performed based on the methods previously described by Ibrahim etal.

[0153] Table 1 | Datasets and samples used for target selection and sequencing.

[0154] 2.2 smMIP design

[0155] Using the “MIPGEN” software, smMIPs were designed for both DNA strands (i.e. double-tiled) for each selected target site. smMIPs in our design contain (a) a common smMIP backbone of 30nt, (b) single molecule tags of 5nt on each side, and (c) two binding arms of circa 20nt that were specifically designed for each target to have an insert size of 50nt. The single molecule tags differ per smMIP copy and allow filtering for PCR duplicates. Afterwards smMIPs covering SNPs and / or repeats were removed, and a final selection was made.

[0156] 3. Classifier model construction based on NGS data

[0157] For each smMIP, linear discriminant analysis was carried out using the “Ida” function from the “MASS” package in the software package R (version 4.0.2) (23). A model was first constructed and then validated using the “ROCR” package. Five-fold cross validation was carried out with a randomization restriction to proportionally represent the tumor types across the five folds. Predictive accuracy of the LDA models was expressed using the Area under the ROC curve (AUC).

[0158] Finally, the least efficient smMIPs were removed with a cutoff of 1 ,000 cumulative counts in all undigested samples, since a minimal number of counts is needed to make a robust classifier. All smMIP models with a cross-validated AUC below 0.8 were removed for the final model. In case of double-tiled smMIPs, the best performing smMIP was selected. All remaining single smMIP models were combined into the final model. The prediction cutoff for each single smMIP model was determined by the lowest sum of false positives and false negatives. The combination of all single predictions was then assessed by a ROC curve and the prediction cutoff was determined based on the highest overall accuracy to end up with the final classifier model.

[0159] 4. Biomarker detection using ddPCR

[0160] A duplex and triplex ddPCR assay (according to the standard protocol) including one and two target sites as well as a reference site respectively were developed. These targets overlap with three smMIP targets of our classifier model. Following primer and probe sequences were used:

[0161] N*, NFQ-MGB = non-fluorescent quencher - minor groove binder

[0162] In brief, an assay mix containing primers, probes and Tris-HCI (pH8) was made. Subsequently, the detection mix was composed of 1 1 pL 2x Bio-Rad ddPCR™ Supermix for probes (no dUTP), 1 ,1 pL of the assay mix, 2 pL DNA and 7,9 pL miliQ H2O to a final volume of 22 pL. For both assays, target and reference primer concentrations as well as the reference probe concentration were 900 nM in the final detection mix. Around 20 000 nano-liter sized droplets were generated according to the manufacturer’s protocol using the automated QX200™ Droplet generator (Bio-Rad). The resulting droplets were transferred to a 96-well plate and sealed. Target sequences were amplified using a Veriti ™ thermal cycler (Applied Biosystems). Temperatures and times for the activation and inactivation steps were based on the recommended protocol from Bio-Rad for the ddPCR™ Supermix for Probes (no dUTP). Ramp rates of 2.5°C / sec were used based on Bio-Rad guidelines. The amplification temperature was set at the optimized temperature. Samples were immediately analyzed post amplification in the QX200™ Droplet Reader (Bio-Rad).

[0163] The ddPCR assays were used to assess 103 fresh frozen tumor samples and 109 fresh frozen normal adjacent tissue samples (Table 1 ). These samples were the same as in the IMPRESS experiment, except for the samples with insufficient concentration. For each target site, sensitivity and specificity were calculated based on the methylation level of each sample.

[0164] 5. Statistics and calculations

[0165] For the power calculations, online datasets were used to obtain the mean and the standard deviation of the methylation level of the selected targets for the tumor and the normal group. We used the target with the smallest Cohen’s D effect size. A sample size with 67 cases and an equal number of controls holds 80% power to detect any difference between the tumor group (methylation level = 0.50 ± 0.24) and the normal group (methylation level = 0.30 ± 0.20), corrected for multiple testing (1791 CpG sites) with a two-sided test. In the present example, 1 1 1 tumor samples and 149 normal samples were used holding a power of 99%.

[0166] Differences in average methylation levels between tumor and normal adjacent samples within one tissue type were tested using the “Mann-Whitney U” test. The performance of the IMPRESS and ddPCR was expressed in terms of specificity and sensitivity. Differences in sensitivity and specificity were tested for significance using a “differences in proportions” test. To measure the repeatability of our technique, the association between normalized counts from two separate sequencing runs was determined by calculating “Pearson correlation coefficients”. In addition, a “Bland-Altman” analysis was performed using the normalized counts of two separate runs. The limit of methylation detection was determined by first calculating the breakpoint of a segmented linear regression that fitted the input data the best using the “segmented” package in R. Afterwards, the intersection of the 95% confidence interval of the first part with the regression of the second part was determined and appointed as the lower limit of detection.

[0167] For all analyses, p-values lower than 0.05 were considered significant. All statistical tests were performed in GraphPad Prism (version 9.5.1 ) or R (version 4.0.2). Normalized counts for each smMIP were calculated as follows:

[0168] Results

[0169] 1. Development of multi-cancer biomarker panel

[0170] For the in silica selection of potential biomarkers, online available methylation data of eight of the most lethal cancer types worldwide and blood samples were used (Table 1 ). In total, 1 ,791 hypermethylated CpG sites (data not shown) were selected based on the following parameters: (a) a minimum average methylation level of 0.5 in the eight cancer types, (b) a maximum average methylation level of 0.2 for the normal adjacent tissue and blood, and (c) the presence of a restriction site for one of the four used MSREs. Normal blood datasets were included for biomarker selection, resulting in a biomarker panel that is suitable for liquid biopsies as well. Secondly, a total of 2,739 reference sites were selected from the human genome. These reference sites are included to estimate the effective total amount of input DNA and allow normalization of the results per sample. Reference sites were chosen to (a) not contain a recognition site of the selected MSREs (1 ,000 sites), or (b) not contain a CpG site (1 ,739 sites). These regions are never cleaved by the enzymes and are therefore always captured by the smMIPs. Lastly, both CpG (10 per MSRE site) and reference sites (15 without recognition site, 15 without CpG site) were selected in lambda phage DNA. Lambda phage DNA is never methylated and is used as an internal control for the enzymatic digestion reaction.

[0171] For these 1 ,791 hypermethylated CpG sites, 2,739 reference sites, and 70 lambda phage DNA sites, smMIPs were designed for both DNA strands (data not shown). After the removal of smMIPs covering SNPs and / or repeats, 2,331 CpG smMIPs and 600 reference smMIPs (300 without recognition site and 300 without CpG site) were selected for human targets. For lambda phage DNA, 12 CpG smMIPs (3 per restriction site) and 10 reference smMIPs (5 without recognition site and 5 without CpG site) were selected. 2. Multi-cancer detection assay

[0172] 2. 1 Data exploration

[0173] To evaluate the biomarker panel, the IMPRESS technique was performed on fresh frozen tissue and blood samples. First, we prepared a sequencing library with 1 1 1 tumor samples, 1 14 normal adjacent tissue samples and 35 whole blood samples targeted by a total of 2,953 smMIPs (2,331 CpG smMIPs, 600 reference smMIPs and 22 lambda smMIPs). Sequencing was done on the Illumina NextSeq system. In total, the output of the sequencer consisted of 475,1 15,304 paired reads. After the first analysis of the raw data, read counts for all smMIPs for each sample were obtained. Based on the characterization experiments, a minimum read count threshold of 5,000 counts per sample was determined, and all samples met this requirement.

[0174] The efficiency of the MSRE digest was verified in each sample by the spiked-in lambda phage DNA as an internal control. A threshold of 5% non-digested fragments was set. One out of 260 samples exceeded this threshold (9.7%) and was removed from further analyses. On average, only 1 .3% of the DNA in each sample was not properly digested. In total, 19 underperforming CpG smMIPs with no counts in any sample were removed from the analysis. Finally, normalization was done per sample to correct for the amount of input DNA. This was done by dividing the counts of every CpG smMIP by the sum of all reference smMIP counts, resulting in a final dataset with counts of 2,312 CpG targeting smMIPs for 259 samples. The normalized count is assumed to be higher in samples methylated for our targets (i.e. tumor samples) and lower in samples unmethylated for our targets (i.e. normal samples). An overview of the sample distribution of the sum of the normalized CpG smMIP counts is given in Figure 1 . Tumor samples show higher and more spread normalized counts, while normal samples show lower and more similar normalized counts. The blood samples have the lowest normalized counts compared to all other tissue types. Within each tissue type, the average normalized counts of tumor and normal samples are significantly different.

[0175] 2.2 Selection of the most efficient and discriminating smMIPs

[0176] To determine the final biomarker panel for the classifier model, the most efficient and discriminating smMIPs were selected. Using the final dataset with counts of 2,312 CpG targeting smMIPs for 259 samples, a single smMIP linear discriminant analysis (LDA) model was constructed using 5-fold cross validation and the mean cross validated AUC (cvAUC) was calculated for each smMIP. We used a cvAUC cutoff of 0.8 for selecting the best differentiating smMIPs, and a cutoff of 1 ,000 cumulative counts in all undigested samples as a measure for smMIP efficiency. This resulted in a set of 51 1 CpG smMIPs. Additionally, for CpG sites targeted by multiple smMIPs (i.e. double-tiled), the best performing one was selected. The 51 1 remaining CpG smMIPs targeted 358 CpG sites. Of these sites, 153 were targeted double-tiled. The difference in cvAUC between smMIPs targeting the same CpG site was less than 0.05 for 84.3% of the multi-targeted CpG sites. The correlation coefficient of these cvAUC values was r=0.668. Finally, 358 single-tiled CpG smMIPs remained for further analysis. For the reference smMIPs, only the efficient smMIPs were selected, with the same count cutoff of 1 ,000 cumulative counts in all undigested samples. As a result, 529 reference smMIPs were selected. 2.3 Classifier model

[0177] The remaining 358 CpG smMIP models were then combined into a single model by first selecting the cutoff for every single model for which the sum of false positives and false negatives was the lowest. Then all predictions were combined, and the cutoff was selected based on the highest overall accuracy, which was achieved when 1 14 single smMIP models agreed on a tumor classification. This final model has a sensitivity of 0.95, a specificity of 0.91 , and an accuracy of 0.92 (Table 2).

[0178] In addition, we investigated the results per cancer type (Table 2). Accuracy per cancer type ranged from 0.47 to 1 , with esophageal cancer being discerned with perfect accuracy, while colorectal tumors performed significantly worse than all other types. Sensitivity is very high overall, with only 6 false negatives among liver, pancreas and head and neck tumors. False positives are attributed to five tumor types, with colorectal tumors among the highest (specificity 0.10), which skews the overall specificity. However, when exclusively investigating colorectal samples, the cutoff can be adjusted to 282 single smMIP models to obtain a classification accuracy of 1 . Interestingly, healthy blood samples never showed up as false positives.

[0179] Table 2 | Metrics of our classifier model.

[0180] 3. Biomarker detection using ddPCR vs IMPRESS

[0181] We confirmed our findings with the gold standard ddPCR technology. Therefore, we selected three target sites of our final classifier model (see above). For this selection, we started with the top 200 best performing smMIP targets, for which we aimed to design ddPCR primers and probes. Finally, we selected the three targets with the best performing primers and probes for the development and optimization of two ddPCR assays. The assays were executed on 103 tumor samples and 109 normal adjacent tissue samples that were used for the IMPRESS experiments. For each of the targets, a single model was built to evaluate the sensitivity and the specificity. In Figure 2 the ROC curves for the three target sites are shown both for the IMPRESS assay and for the ddPCR assay.

[0182] The sensitivity and specificity of the single ddPCR models were compared with those from the single smMIP models for the three targets. Figure 2 and Table 3 show that only minimal differences between both technologies are found, admittedly favoring the IMPRESS technique. These results indicate that the IMPRESS technique performs at least equally good as the gold standard ddPCR.

[0183] Table 3| Sensitivities and specificities for the three selected target sites in both the IMPRESS assay and the ddPCR assay. The sensitivities and specificities are calculated for each cancer type separately an|d for the overall analysis. Statistics was executed by the differences in proportions test and significant differences are indicated, ns = not significant * = p-value £ 0.05, ’* = p-value £ 0.01, “ = p-value £ 0.001 and **“ = p-value £ 0.0001.

[0184] 3.1 Other combinations of CpG target locations

[0185] Figures 3-12 represent combinations of three target locations which were in the list of top 200 best performing smMIP targets (non-exhaustively generated) and refer to respectively set 1 -10 as described in the application. For each model, the 5-fold cross validation ROC curves (thin lines) and the mean ROC curve (bold line) are plotted. The ROC curves with AUC values for each of the three target sites are shown as well as a common denominator AUC value.

[0186] In particular, set 1 refers to chr10:101287802-101287891 (SEQ ID N° 470), chr15:68121 108-68121 197 (SEQ ID N° 1024), chr2: 73147755-73147844 (SEQ ID N° 1482) with an AUC value of 0.975; set 2 refers to chr1 :197888429-197888518 (SEQ ID N° 337), chr1 :170630499-170630588 (SEQ ID N° 331 ), chr1 1 :31846774-31846863 (SEQ ID N 667) with an AUC value of 0.974; set 3 refers to chr2:73152605-73152694 (SEQ ID N° 1483), chr4:41882097-41882186 (SEQ ID N° 1761 ), chr10:109674269-109674358 (SEQ ID N°51 1 ) with an AUC value of 0.970; set 4 refers to chr2:73147755-73147844 (SEQ ID N° 1482), chr5:76923876-76923965 (SEQ ID N° 1880), chr7:8482030-84821 19 (SEQ ID N° 263) with an AUC value of 0.955;

[0187] Set 5 refers to chr1 :197888439-197888528 (SEQ ID N° 338), chr1 :17475757-17475846 (SEQ ID N° 1 ), chr 10:101287802-101287891 (SEQ ID N° 470) with an AUC value of 0.970;

[0188] Set 6 refers to chr5:76923876-76923965 (SEQ ID N° 1880), chr4:41882097-41882186 (SEQ ID N° 1761 ), chr2:73147755-73147844 (SEQ ID N° 1482) with an AUC value of 0.971 ;

[0189] Set 7 refers to chr2:87016651 -87016740 (SEQ ID N° 1490), chr2:73147755-73147844 (SEQ ID N° 1482), chr10:109674269-109674358 (SEQ ID N° 51 1 ) with an AUC value of 0.972;

[0190] Set 8 refers to chr1 :197888439-197888528 (SEQ ID N° 338), chr1 1 :320091 17-32009206 (SEQ ID N° 669), chr1 :170630499-170630588 (SEQ ID N° 331 ) with an AUC value of 0.974;

[0191] Set 9 refers to chr1 :197888429-197888518 (SEQ ID N° 337), chr13:95354716-95354805 (SEQ ID N° 894), chr1 :17475757-17475846 (SEQ ID N° 1 ) with an AUC value of 0.971 ;

[0192] Set 10 refers to chr1 :17475757-17475846 (SEQ ID N° 1 ), chr7:35301 123-35301212 (SEQ ID N° 2129), chr10:109674269-109674358 (SEQ ID N° 51 1 ) with an AUC value of 0.971 . 4. Potential for liquid biopsies

[0193] To test whether the technique holds the potential as a multiplex tool for methylation detection in liquid biopsy samples, several characterization experiments were performed.

[0194] To further explore the lower limits of input DNA and the feasibility of the technique to study liquid biopsies, four cell-free DNA (cfDNA) samples were tested analogous to the previous experiment with 5ng of input. Additionally, undigested samples were included to serve as positive controls. Results showed that digested samples had an average normalized count of 0.24 while the undigested samples had an average normalized count of 2.24. This indicates that the samples were effectively digested by the MSREs as well as efficiently captured by the smMIPs and sequenced.

[0195] To mimic the presence of circulating tumor DNA (ctDNA) in cfDNA, DNA from three tumor cell lines was sheared into fragments of 150-500bp and spiked into cfDNA samples in different percentages between 0% and 100%. The calculated percentages based on the normalized counts were closely correlated to their expected value, with a correlation coefficient r of 0.97, 0.99 and 0.98 for the three different cell lines. This indicates the applicability of the technique for the quantification of methylation. The calculated percentage of 20% spiked-in DNA (lowest percentage tested) ranged from 21 % to 30%, while those of 0% spiked-in DNA (only cfDNA) ranged from 6% to 13%. This means samples with 20% spike-in of cell line DNA had on average a threefold higher percentage of normalized counts than cfDNA samples without spike-in.

[0196] In summary, the examples clearly provide validated support for a multi-cancer methylation biomarker panel which results in the development of a multi-cancer diagnosis and detection assay. With an overall cross-validated accuracy of 0.92, the final model performed very well in classifying samples. The overview of the sample distribution (Figure 1 ) shows a spread of tumor samples for most of the tumor types, while normal tissues are grouped closer together. There was no correlation between tumor cell percentage and normalized counts.

[0197] The final model performed very well with 0.95 sensitivity and 0.91 specificity. This highlights the great performance of the biomarker panel of the present invention and shows that there are methylation differences for all samples.

[0198] EXAMPLE 2

[0199] Model construction for all three probe combinations out of the top 100 probes

[0200] Materials and methods

[0201] First, the 100 probes with the highest single-probe cross-validated AUC values, calculated as described in “Classifier model construction based on NGS data”, were selected (exhaustively generated). Using these best performers, all models using combinations of three different probes were made, amounting to 161 ,700 models. The prediction cut-off for each single-probe model was again determined by the lowest sum of false positives and false negatives. The combination of the three single predictions was then assessed by a ROC curve, AUC, sensitivity, and specificity. For the AUC values of all models, summary statistics were generated, and a density graph was created.

[0202] Finally, we looked at the distribution of the probes within the three-probe models, to check for overrepresentation of certain probes. This was done looking at the top 20,000 three-probe models and counting how many times every probe occurred within this selection (Table 4).

[0203] Table 4: CpG locations occuring more than 1000 times in the best 20000 sets:

[0204] Results

[0205] By generating all different three-probe models from the top 100 probes, we made 161 ,700 models. All these models performed very well, the minimum AUC was 85.2%, the maximum 97.8% and the mean 93.9%, the density graph is normally distributed (Figures 13). Table 5 displays the top 100 of these models while Figure 14-23 shows cross-validated ROC plots of the 10 best performing three-probe models (resp. set 11 to set 20), along with the performance of the single-probe models used for the creation of the three-probe model. When looking at the distribution of the single probes within the top three-probe models, it was clear that there are some probes overrepresented in the top models compared to all other probes. The most overrepresented probe is ‘chr2:73147755_73147844’ (SEQ ID N° 1482), occurring approximately 5.5 times more frequently than the mean probe (Table 4). Table 5 displays the top 100 (i.e. set 1 1 to set 1 13) of at least 3 CpG location models and their respective AUC values (it is noted that set 15, set 23 and set 102 are duplicates and correspond with respectively set 1 , set 2 and set 10).

[0206] REFERENCES

[0207] Ibrahim J, op de Beeck K, Fransen E, Peeters M, van Camp G. Genome-wide DNA methylation profiling and identification of potential pan-cancer and tumor-specific biomarkers. Mol Oncol. John Wiley and Sons Ltd; 2022;16:2432-47.

Claims

CLAIMS1 . A multi-cancer detection or diagnosis method comprising the step of: determining in a sample from a subject suspected of having cancer, the methylation status of the CpG dinucleotides across at least 3 CpG locations selected from the group comprising: SEQ ID N°: 1 - 2316.

2. The method of claim 1 , wherein at least one CpG location is chr2:73147755-73147844 (SEQ ID N°1482).

3. The method according to claim 2, wherein one or more further CpG location is selected from the list comprising chr10: 102894297-102894386 (SEQ ID N° 481 ), chr10:_101287802-101287891 (SEQ ID N°: 470), chr10:109674269-109674358 (SEQ ID N° 51 1 ), chr15:68121 108-68121 197 (SEQ ID N° 1024), chr1 :197888429-197888518 (SEQ ID N° 337), chr5:76923876-76923965 (SEQ ID N° 1880), in particular SEQ ID N° 481 .

4. The method of any one of claims 1 -3, wherein said methylation status of the CpG dinucleotides is measured across at least 5 CpG locations, in particular across at least 10 CpG locations, more in particular across at least 20 CpG locations, even more in particular across at least 50 CpG locations, most in particular across at least 100 CpG locations.

5. The method of any one of claims 1 -4, further comprising comparing said methylation status of saidCpG locations in said sample with the methylation status of said CpG locations in at least one reference sample.

6. The method of claim 5, wherein an alteration in the methylation status of said CpG locations in said sample compared to the methylation status of said CpG locations in at least one reference sample is indicative of said subject suffering from cancer.

7. The method of claim 6, wherein said alteration in methylation status is an increase in the percentage of methylation compared to said reference sample.

8. The method of claim 7, wherein said increase is at least 20%, preferably at least 25% increase in methylation compared to said reference sample.

9. The method of claim 8, wherein said increase is at least 30% increase in methylation compared to said reference sample.

10. The method of claim 1 , wherein the at least 3 CpG locations are CpG locations of set 1 : chr10:101287802-101287891 (SEQ ID N° 470), chr15:68121 108-68121 197 (SEQ ID N° 1024), chr2: 73147755-73147844 (SEQ ID N° 1482); or set 2: chr1 :197888429-197888518 (SEQ ID N° 337), chr1 :170630499-170630588 (SEQ ID N° 331 ), chr1 1 :31846774-31846863 (SEQ ID N° 667); or set 3: chr2:73152605-73152694 (SEQ ID N° 1483), chr4:41882097-41882186 (SEQ ID N° 1761 ), chr10:109674269-109674358 (SEQ ID N°51 1 ); or set 4: chr2:73147755-73147844 (SEQ ID N° 1482), chr5:76923876-76923965 (SEQ ID N° 1880), chr7:8482030-84821 19 (SEQ ID N° 263); orSet 5: chr1 :197888439-197888528 (SEQ ID N° 338), chr1 :17475757-17475846 (SEQ ID N° 1 ), chr 10:101287802-101287891 (SEQ ID N° 470); orSet 6: chr5:76923876-76923965 (SEQ ID N° 1880), chr4:41882097-41882186 (SEQ ID N° 1761 ), chr2:73147755-73147844 (SEQ ID N° 1482); orSet 7: chr2:87016651 -87016740 (SEQ ID N° 1490), chr2:73147755-73147844 (SEQ ID N° 1482), chr10:109674269-109674358 (SEQ ID N° 51 1 ); orSet 8: chr1 :197888439-197888528 (SEQ ID N° 338), chr1 1 :320091 17-32009206 (SEQ ID N° 669), chr1 :170630499-170630588 (SEQ ID N° 331 ); orSet 9: chr1 :197888429-197888518 (SEQ ID N° 337), chr13:95354716-95354805 (SEQ ID N° 894), chr1 :17475757-17475846 (SEQ ID N° 1 ); orSet 10: chr1 :17475757-17475846 (SEQ ID N° 1 ), chr7:35301 123-35301212 (SEQ ID N° 2129), chr10:109674269-109674358 (SEQ ID N° 51 1 ); orSet 1 1 : chr2:73147755-73147844 (SEQ ID N° 1482), chr10: 102894297-102894386 (SEQ ID N° 481 ), chr8:109095906-109095995 (SEQ ID N° 2182); orSet 12: chr2:73147755-73147844 (SEQ ID N° 1482), chr10:102894297-102894386 (SEQ ID N° 481 ), chr8:109093380-109093469 (SEQ ID N° 2181 ); orSet 13: chr15:68121108-68121 197 (SEQ ID N° 1024), chr10:102894297-102894386 (SEQ ID N° 481 ), chr8:109093380-109093469 (SEQ ID N° 2181 ); orSet 14: chr15:68121108-68121 197 (SEQ ID N° 1024), chr10:102894297-102894386 (SEQ ID N° 481 ), chr8:109095906-109095995 (SEQ ID N° 2182); orSet 15: chr10:101287802-101287891 (SEQ ID N° 470), chr2:73147755-73147844 (SEQ ID N° 1482), chr15:68121 108-68121 197 (SEQ ID N° 1024); orSet 16: chr1 :197888429-197888518 (SEQ ID N° 337), chr10:101287802-101287891 (SEQ ID N° 470), chr2:73147755-73147844 (SEQ ID N° 1482); orSet 17: chr10:109674269-109674358 (SEQ ID N° 51 1 ), chr2:73147755-73147844 (SEQ ID N° 1482), chr10:102894297-102894386 (SEQ ID N° 481 ); orSet 18: chr10:109674269-109674358 (SEQ ID N° 51 1 ), chr10:101287802-101287891 (SEQ ID N° 470), chr2:73147755-73147844 (SEQ ID N° 1482); orSet 19: chr5:76923876-76923965 (SEQ ID N° 1880), chr1 :197888429-197888518 (SEQ ID N° 337), chr2:73147755-73147844 (SEQ ID N° 1482); and / orSet 20: chr5:76924134-76924223 (SEQ ID N° 1881 ), chr1 :197888429-197888518 (SEQ ID N° 337), chr2:73147755-73147844 (SEQ ID N° 1482).1 1 . The method of claim 10, wherein the at least 3 CpG locations are CpG locations of set 4: chr2:73147755-73147844 (SEQ ID N° 1482), chr5:76923876-76923965 (SEQ ID N° 1880), chr7:8482030-84821 19 (SEQ ID N° 263).

12. The method of claim 10, wherein said method comprises determining the methylation status of theCpG dinucleotides measured across at least the CpG locations of any one of set 1 to set 20, or any combination thereof; wherein an increase in methylation status in said sample of at least one set compared to the methylation status of a reference sample is indicative for a subject suffering from cancer.

13. The method of any one of claims 1 -12, wherein said sample is solid tissue biopsy and / or a body fluidbiopsy, preferably a body fluid biopsy, more preferably a body fluid biopsy selected from the list comprising: blood, serum, plasma, saliva, urine, stool, or a combination thereof.

14. The method of any one of claims 1 -13, wherein said sample is a DNA sample.

15. The method of any one of claims 1 -14, wherein the methylation status of the CpG dinucleotides measured across the at least 3 CpG locations are measured by a method selected from the group comprising IMPRESS (Improved multiplex Methylation Profiling using Restriction Enzymes and chr sequencing) or ddPCR (droplet digital PCR).

Images