Method for determining microsatellite instability state, kit and use thereof

Abstract

The invention relates to the field of genomics. The invention more specifically relates to novel microsatellite markers and uses thereof, in particular for determining the microsatellite status of a tumor, in particular a tumor from a human subject. The specification provides a method for analyzing a DNA microsatellite locus. The invention also relates to tools, kits and systems that can be used to carry out such assays.

Description

Technical Field

[0001] The present invention relates to the field of genomics. More specifically, the present invention relates to novel microsatellite markers and their uses, particularly for determining the microsatellite status of tumors, particularly tumors from human subjects. The present invention provides methods for analyzing DNA microsatellite loci. It also relates to tools, kits, and systems that can be used to perform such analyses. Background Art

[0002] Microsatellites are short, repetitive DNA sequences, either coding or noncoding, consisting of tandemly repeated motifs of 1 to 6 bases, also known as "short tandem repeats." During replication, DNA slippage occurs at microsatellite loci, resulting in changes in repeat number and overall microsatellite length. The mismatch repair system (MMR), encoded by the MLH1, MSH2, PMS2, and MSH6 genes, is involved in repairing these errors [Evrard et al.].

[0003] Microsatellite instability (MSI) is a molecular marker of a defective mismatch repair system (dMMR). In fact, dMMR deficiency leads to a progressive accumulation of mutator phenotypes accompanied by genetic instability and is characterized by widespread changes in microsatellite length. The MSI phenotype is most commonly caused by hypermethylation of the MLH1 gene promoter in tumors, leading to loss of its expression [Cunningham JM et al.]. Colorectal cancer (CRC) and endometrial cancer (EC) are the tumors most associated with sporadic dMMR (15% to 20% and 20% to 30%, respectively) [Bonneville R et al., Levine DA]. A less common cause is the inherited transmission of mutant alleles to one of the MMR genes, leading to hereditary non-polyposis colorectal cancer (HNPCC), also known as Lynch syndrome.

[0004] In daily practice, there are two different, validated techniques for detecting the dMMR phenotype: immunohistochemistry (IHC) and microsatellite instability-polymerase chain reaction assay ("MSI-PCR") (Luchini C et al.). The first method identifies the loss of expression of at least one of the four proteins of the MMR system (MLH1, MSH2, MSH6 or PMS2), which reflects a loss of its function. The European Society for Medical Oncology (ESMO) guidelines recommend its use in the first instance in any sporadic cancer type belonging to the Lynch syndrome spectrum. The other method reveals the consequence of the dMMR phenotype, namely microsatellite instability, by analyzing microsatellite loci. MSI-PCR requires the amplification of the locus containing the microsatellite of interest and the analysis of the length of the PCR-generated amplicons by capillary electrophoresis. The MSI phenotype is defined as the instability of at least two microsatellites from the five quintuplet group (BAT-25, BAT-26, NR-21, NR-24, and NR-27 (also designated “MONO-27”), consistent with the Bethesda guidelines (Bacher et al., 2004) for HNPCC. This panel is recommended by international guidelines in cases of equivocal IHC or loss of only one MMR protein [Luchini C et al.]. Furthermore, due to the significant percentage (10%) of false-positive dMMR or MSI-PCR results in immunotherapy CRC trials, it is also recommended to systematically correlate IHC and MSI-PCR before using any immunotherapeutic agent in metastatic dMMR CRC [Levine DA]. For other types of non-colorectal tumors, such as EC tumors, the sensitivity of these tests is lower because they exhibit smaller repeat number changes than CRC, resulting in less obvious shifts in microsatellite markers [Wang et al., 2004]. Y]. Therefore, during MSI-PCR testing of these tumors, comparison of tumor and non-tumor tissue is mandatory, and PCR is strongly recommended as a confirmatory test in cases where dMMR status is determined by IHC or the results are equivocal [Evrard C et al].

[0005] Recent studies have highlighted that dMMR predicts response to immune checkpoint inhibitors (ICIs) in many cancer types. Pembrolizumab, an anti-programmed cell death protein 1 (anti-PD-1) immunotherapy, doubled progression-free survival (PFS) compared with chemotherapy in the phase III KEYNOTE-177 trial as first-line treatment for dMMR metastatic CRC [André T et al.]. The ongoing phase II GARNET trial has recently demonstrated the efficacy of another anti-PD-1 immunotherapy, dostarlimab, in pretreated dMMR / MSI metastatic EC and other solid tumors. Given these promising results, the US Food and Drug Administration (FDA) granted accelerated approval for dostarlimab for dMMR EC following a platinum-based regimen [Berton D et al.]. Furthermore, the phase II KEYNOTE-158 trial demonstrated the efficacy of pembrolizumab as second-line treatment for other types of dMMR cancers, such as EC, pancreatic and gastric cancers, and cholangiocarcinoma [Marabelle A et al].

[0006] With the development of immunotherapy as a new treatment option, the determination of MMR status has clearly become a key point in cancer management. However, conventional methods have diagnostic limitations, such as IHC cannot always detect the loss of mutant protein caused by missense mutations, and normal staining may occur even for some protein-truncating mutations [Shia J. et al.]. In addition, the cost impact, time and tissue consumption associated with IHC and MSI-PCR analysis (especially when non-tumor tissue is required) may represent limiting factors; in turn, these tests are rarely performed on tumors with low dMMR incidence, such as prostate cancer, which may also benefit from immunotherapy. Determination of MSI status by next-generation sequencing (MSI-NGS) may be an interesting alternative because it allows the analysis of a large number of samples simultaneously and can be associated with the search for somatic variants of therapeutic diagnostic interest, such as KRAS, NRAS and BRAF mutations in CRC or POLE and TP53 mutations in EC. There are many published studies on MSI-NGS algorithms, but these require sequencing of large microsatellite panels, cannot be used in routine practice, and therefore have not yet been approved by international guidelines [Long DR et al, Hempelmann JA et al, Middha S et al, Trabucco SE et al]. Therefore, there is a need for a highly sensitive and specific MSI-NGS method that can be used in routine practice.

[0007] Only a few studies have designed NGS methods for MSI determination, but only in colorectal cancer, such as the studies by Herbreteau G. et al. and Ratovomanana T et al. However, these methods may only be used for colorectal cancer and cannot be confidently applied to cancers with minimal microsatellite shifts (such as endometrial cancer). In addition, most of these methods require comparison with matched non-tumor samples. Therefore, to date, there has been no convincing progress in resolving MSI determination in cancer, especially non-colorectal cancer, through NGS.

[0008] To address this need for CRC as well as other tumors such as EC, the inventors describe herein a new biomarker set and a new NGS-compatible method for determining microsatellite status, in particular MSI status, in a highly sensitive and specific manner. Summary of the Invention

[0009] In recent years, the detection of microsatellite instability (MSI) phenotypes has become increasingly important due to the development of immune checkpoint inhibitors as effective therapies against MSI tumors. To date, these therapies are available for colorectal cancer (CRC) and endometrial cancer (EC), but will soon be expanded to other tumor types, necessitating the need to ensure our ability to detect MSI in a more timely and efficient manner.

[0010] This article presents novel markers with high sensitivity for determining the microsatellite stability (MSS) or microsatellite instability (MSI) status of a tumor. These markers are more sensitive than the popular standard pentad panel (i.e., BAT-25, BAT-26, NR-21, NR-24, and MONO-27) while retaining specificity for MSI, in the context of colon cancer, but not exclusively.

[0011] The present description relates in particular to a method of analyzing a set of microsatellite loci, such as a set of at least two or at least three, preferably at least four, microsatellite loci, in human DNA selected from the group comprising:

[0012] -CABIO-P05, defined as a 21T duplication located at 14q23.1 and starting at position chr14:58359108,

[0013] -CABIO-P07, defined as a 21T duplication located at 7q32 and starting at position chr7:131478596,

[0014] -CABIO-E01, defined as a 22T repeat located at Xq22.3 starting at position chrX:106849221,

[0015] -CABIO-E03, defined as a 23T repeat located at Xq21.2 starting at position chrX:85268269,

[0016] -CABIO-E04, defined as a 25T duplication located at 14q32.3 and starting at position chr14:103574079,

[0017] -CABIO-E05, defined as a 22T duplication located at 2p11.2 and starting at position chr2:86456417,

[0018] - CABIO-E06, defined as a 23T duplication located at 4q23 and starting at position chr4:99216136, and

[0019] - CABIO-E07, defined as a 21T repeat located at 20p13 and starting at position chr20:290564, with reference to the Homo sapiens reference genome assembly from the Genome Reference Consortium human Build 38 patch release 14, also referred to herein as GRCh38.p14 or simply GRCh38.

[0020] The present disclosure also relates to a method for assessing the microsatellite stability (MSS) or microsatellite instability (MSI) status of a tumor, comprising the following steps:

[0021] a) counting the number of indels in at least two, such as at least three, preferably at least four microsatellite loci in a tumor DNA sample, wherein the tumor is preferably a human tumor and the microsatellite loci are preferably selected from the group comprising CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07 as first described herein;

[0022] b) for each locus in the microsatellite loci, comparing the number of indels with a reference number of indels for the locus and determining the MSI status of the locus, if the indel count is equal to or higher (≥) the reference value for the locus, then the locus is considered unstable, and if the indel count is lower (<) the reference value for the locus, then the locus is considered stable;

[0023] c) calculating an MSID score, which ranges between 0 and 1 and consists of the total number of unstable loci relative to the total number of loci; and

[0024] d) determining the MSI status of the tumor as unstable if the MSID score is equal to or higher (≥) 0.5; as stable if the MSID score is equal to or lower (≤) 0.125, and as uncertain if the MSID score is between 0.125 and 0.5; and optionally

[0025] e) If the MSID score is between 0.125 and 0.5, repeating steps a)-d) until the status of the MSI can be determined.

[0026] In certain aspects described herein, the methods disclosed herein by the inventors are used to assess microsatellite instability in precancerous and / or cancerous cells; to detect a propensity to develop cancer; to assess the prognosis of cancer; to monitor cancer progression or regression; to predict, evaluate, and monitor response to cancer treatment; to select an appropriate cancer treatment for a subject in need thereof; to select a patient who is likely to respond to a cancer treatment; or to select a patient for enrollment in a clinical trial for a cancer treatment.

[0027] In a preferred embodiment, the method is a partially or fully computer-implemented method.

[0028] Also described herein is a computer-implemented method for training a classifier for determining the microsatellite stability (MSS) or microsatellite instability (MSI) status (also referred to herein as the "phenotype") of a biological sample, particularly a tumor, e.g., a human tumor, wherein the method comprises:

[0029] a) providing a training set of microsatellite loci or preprocessed information obtained from the training set as input to the classifier, each locus being obtained from a DNA sequence of interest, the training set comprising i) stable (MSS) microsatellite loci or subsequences thereof obtained from intact mismatch repair system (pMMR) or MSS cells, DNA, tumors or subjects known to have a microsatellite stable state or phenotype, and ii) unstable (MSI) microsatellite loci or subsequences thereof obtained from deficient mismatch repair system (dMMR) or MSI cells, DNA, tumors or subjects known to have a microsatellite unstable state or phenotype;

[0030] b) generating a classifier output for each microsatellite locus, said output classifying the microsatellite locus input as having a stable (MSS) or unstable (MSI) state or phenotype; and

[0031] c) for each microsatellite locus, evaluating the accuracy of the classifier for distinguishing between a stable (MSS) state or phenotype and an unstable (MSI) state or phenotype by comparing the output of the classifier to the known actual phenotype of the microsatellite locus or to a reference number of indels at the microsatellite locus;

[0032] If the classifier exhibits accuracy in counting insertions and deletions (also known as deletion insertions) at each of the microsatellite loci with a resolution of 1 bp (base pair), the classifier is considered to be an accurate classifier for determining the MSS or MSI status or phenotype of a biological sample, particularly a tumor.

[0033] The inventors also describe herein a computing system comprising:

[0034] - a memory storing at least one instruction of a classifier trained according to a computer-implemented method as described herein, and

[0035] - a processor accessing said memory to read said instructions and to execute the method of the invention as described herein.

[0036] Also described herein is a kit for analyzing microsatellite loci in genomic DNA, preferably human genomic DNA, comprising an oligonucleotide primer pair, preferably at least two oligonucleotide primer pairs, suitable for amplifying or co-amplifying a set of microsatellite loci in human genomic DNA, and / or an oligonucleotide probe, preferably at least two oligonucleotide probes, for detecting sequences in said set of microsatellite loci, and optionally a thermostable polymerase and / or control DNA isolated from normal, non-cancerous biological material and / or lacking mismatch repair genes. The set typically comprises several, e.g., 2 to 8, preferably at least 4, primer pairs suitable for the amplification of the following items, and / or 2 to 8, preferably at least 4 oligonucleotide probes for the detection of the following items: a set of markers as described herein, in particular a set comprising at least two microsatellite markers selected from CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07, preferably a set of at least four, e.g., at least five, six or seven microsatellite markers, or a set of all eight microsatellite markers.

[0037] The inventors also describe herein the use of such kits for analyzing microsatellite stability or instability. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 : Research workflow.

[0039] IHC: immunohistochemistry; CRC: colorectal cancer; EC: endometrial carcinoma, MMR: mismatch repair, dMMR: deficient MMR; pMMR: intact MMR; MSS: microsatellite stability; MSI: microsatellite instability; MSI-PCR: MSI testing by polymerase chain reaction; Octet CaBio-MSID: MSI testing by next-generation sequencing using an MSI detection tool.

[0040] Figure 2 : MSI classification using MSID in CRC using a conventional 5-marker panel (BAT-25, BAT-26, NR-21, NR-24, MONO-27).

[0041] The dot plots show the correlation between pMMR / MSS or dMMR / MSI status initially determined by a reference method (IHC+PCR) and MSI-NGS testing of a retrospective cohort of 303 CRC samples. After stratification by the reference method, 241 pMMR / MSS samples (left) and 62 dMMR / MSI samples (right) were reassessed using the MSID, showing the fraction of unstable microsatellite loci (MSID score values: 0, no unstable marker; 0.20, 1 unstable marker; 0.40, 2 unstable markers; 0.60, 3 unstable markers; 0.80, 4 unstable markers; 1, 5 unstable markers). Microsatellite status was considered stable if the MSID score was ≤0.2, unstable if the MSID score was ≥0.6, and uncertain if the MSID score was equal to 0.4.

[0042] MSI: microsatellite instability; MSS: microsatellite stability; MSID: MSI detection tool

[0043] Figure 3 : Intra-laboratory reproducibility of an 8-marker CaBio panel for MSI classification using MSID in FFPE samples.

[0044] The dot plot shows the reproducibility results obtained for 3 different CRC samples (1 pMMR / MSS and 2 dMMR / MSI) that were initially stratified by the reference method (IHC+PCR) and reassessed using MSID in 3 independent analysis procedures. The horizontal line and error bars represent the mean + SD, showing the fraction of unstable microsatellite loci (MSID score values: 0, no unstable marker; 0.125, 1 unstable marker; 0.25, 2 unstable markers; 0.375, 3 unstable markers; 0.50, 4 unstable markers; 0.625, 5 unstable markers; 0.75, 6 unstable markers; 0.875, 7 unstable markers; 1, 8 unstable markers). If the MSID score is ≤0.125, the microsatellite status is considered stable, if the MSID score is ≥0.5, the microsatellite status is considered unstable, and if the MSID score is equal to 0.25 and 0.375, the microsatellite status is considered uncertain.

[0045] MSI: microsatellite instability; MSS: microsatellite stability; MSID: MSI detection tool; MSI-1: MSI-2: patients with different microsatellite instability levels

[0046] Figure 4 : Overview of microsatellite status determination in a representative selection of FFPE-CRC (A) and FFPE-EC (B) samples.

[0047] 30 CRCs and 26 ECs were first classified by two reference technologies: IHC analysis for expression of the MMR proteins MLH1 / PMS2 / MSH2 / MSH6 to detect pMMR or dMMR phenotypes, and MSI-PCR analysis for determination of MSS or MSI status. The samples were then blinded for MSS or MSI status by MSI-NGS. An MSI-NGS score of MSS (blue) or MSI (red) was given for each microsatellite locus (CaBio E01, E03, E04, E05, E06, E07, P05, P07). Samples were grouped according to IHC phenotype and MSI-PCR status.

[0048] MSI: microsatellite instability; MSS: microsatellite stability; MSID: MSI detection tool; dMMR: deficient mismatch repair; pMMR: intact mismatch repair; PCR: polymerase chain reaction; EC: endometrial carcinoma; CRC: colorectal carcinoma; IHC: immunohistochemistry

[0049] Figure 5 : MSI classification using MSID in CRC using the CaBio octet (“CaBio panel”).

[0050] Figure 3. Dot plots showing the correlation between pMMR / MSS or dMMR / MSI status initially determined by a reference method (IHC+PCR) and MSI-NGS testing of a retrospective cohort of 303 CRC samples. After stratification by the reference method, 241 pMMR / MSS samples (left) and 62 dMMR / MSI samples (right) were reassessed using the MSID, showing the fraction of unstable microsatellite loci (MSID score values: 0, no unstable marker; 0.125, 1 unstable marker; 0.25, 2 unstable markers; 0.375, 3 unstable markers; 0.50, 4 unstable markers; 0.625, 5 unstable markers; 0.75, 6 unstable markers; 0.875, 7 unstable markers; 1, 8 unstable markers). Microsatellite status was considered stable if the MSID score was ≤ 0.125, unstable if the MSID score was ≥ 0.5, and uncertain if the MSID score was equal to 0.25 and 0.375. Comparison of the MSID scores obtained for MSI tumors indicated a significant difference relative to MSS tumors (p = 2.81). -48 , Wilcoxon-Mann Withney rank-sum test, two-sided).

[0051] MSI: microsatellite instability; MSS: microsatellite stability; MSID: MSI detection tool

[0052] Figure 6 : Diagnostic performance of CaBio loci in MSI-NGS for predicting CRC compared with the reference method (IHC+PCR).

[0053] AB: AUC (area under the curve)-ROC (receiver operating characteristic) curves for evaluating the performance of individual loci in 41 CRC samples. C: Calculation of 95% confidence intervals (CI) for AUC, sensitivity (Se%), and specificity (Sp%) for each of the eight individual loci is given.

[0054] Figure 7 : MSI classification using MSID in EC using the CaBio octet (“CaBio panel”).

[0055] Figure 3. Dot plots showing the correlation between pMMR / MSS or dMMR / MSI status initially determined by a reference method (IHC+PCR) and MSI-NGS testing in a retrospective cohort of 88 EC samples. After stratification by the reference method, 60 pMMR / MSS samples (left) and 28 dMMR / MSI samples (right) were reassessed using the MSID, showing the fraction of unstable microsatellite loci (MSID score values: 0, no unstable marker; 0.125, 1 unstable marker; 0.25, 2 unstable markers; 0.375, 3 unstable markers; 0.50, 4 unstable markers; 0.625, 5 unstable markers; 0.75, 6 unstable markers; 0.875, 7 unstable markers; 1, 8 unstable markers). Microsatellite status was considered stable if the MSID score was ≤ 0.125, unstable if the MSID score was ≥ 0.5, and uncertain if the MSID score was equal to 0.25 and 0.375. Comparison of the MSID scores obtained for MSI tumors indicated a significant difference relative to MSS tumors (p = 2.89, 1.0). -18 , Wilcoxon-Mann Withney rank-sum test, two-sided).

[0056] MSI: microsatellite instability; MSS: microsatellite stability; MSID: MSI detection tool

[0057] Figure 8 : Diagnostic performance of CaBio loci in MSI-NGS for predicting EC compared with the reference method (IHC+PCR).

[0058] AB: AUC (area under the curve)-ROC (receiver operating characteristic) curves for evaluating the performance of individual loci in 11 EC samples. C: Calculation of 95% confidence intervals (CI) for AUC, sensitivity (Se%), and specificity (Sp%) for each of the eight individual loci is given. DETAILED DESCRIPTION

[0059] In order to make the present invention more readily understood, certain terms are defined herein. Additional definitions are set forth throughout the detailed description of the invention.

[0060] Unless otherwise defined, all technical terms, symbols, and other scientific terms used herein are intended to have the meanings commonly understood by those skilled in the art to which the invention pertains. In some cases, for clarity and / or ease of reference, terms with commonly understood meanings are defined herein, and the inclusion of such definitions herein is not necessarily to be construed as representing a departure from what is generally understood in the art. Techniques and procedures described or referenced herein are generally well understood and routinely employed by those skilled in the art using conventional methods.

[0061] The term "amplicon" or "amplicon molecule" refers to a nucleic acid molecule generated by amplification of a template nucleic acid molecule such as cfDNA or a nucleic acid molecule having a sequence complementary thereto, or a double-stranded nucleic acid including any such nucleic acid molecule.

[0062] The term "oligonucleotide primer" or "primer" refers to a nucleic acid molecule that is used for, can be used for, or is used for use in a nucleic acid molecule that generates an amplicon from a template nucleic acid molecule. Under conditions that allow amplification (e.g., in the presence of nucleotides and a DNA polymerase, and at a suitable temperature and pH), an oligonucleotide primer can provide a starting point for amplification from a template to which it hybridizes. Typically, an oligonucleotide primer is a single-stranded nucleic acid with a length between 5 and 200 nucleotides. One of ordinary skill in the art will appreciate that the optimal primer length for generating an amplicon from a template nucleic acid molecule can vary with conditions including temperature parameters, primer composition, and amplification method. As used herein, an oligonucleotide primer pair refers to a set of two oligonucleotide primers that are complementary to the first and second chains of a template double-stranded nucleic acid molecule, respectively. The first and second members of a pair of oligonucleotide primers can be referred to as a "forward" oligonucleotide primer and a "reverse" oligonucleotide primer, respectively, relative to a template nucleic acid strand, wherein the forward oligonucleotide primer is capable of hybridizing with a nucleic acid strand complementary to the template nucleic acid strand, the reverse oligonucleotide primer is capable of hybridizing with the template nucleic acid strand, and the position of the forward oligonucleotide primer relative to the template nucleic acid strand is 5' to the position of the reverse oligonucleotide primer sequence relative to the template nucleic acid strand. One of ordinary skill in the art will appreciate that designating the first and second oligonucleotide primers as forward and reverse oligonucleotide primers, respectively, is arbitrary, as these identifiers are largely dependent on whether a given nucleic acid strand or its complementary strand is used as a template nucleic acid molecule.

[0063] The term "probe" refers to a single-stranded or double-stranded nucleic acid molecule that can hybridize with a complementary target (such as a DNA or an amplicon) and that comprises a detectable portion. In some cases, for example, as described herein, a probe is a capture probe that can be used to detect, identify, and / or separate a target sequence such as a gene sequence. In various cases, for example, as described herein, the detectable portion / fragment of a probe can be, for example, an enzyme (see ELISA and enzyme-based histochemical assays), a fluorescent portion, a radioactive portion, or a portion associated with a luminescent / light signal.

[0064] The "sequence identity" between two sequences is described by the parameters "sequence identity", "sequence similarity" or "sequence homology". In the context of the present invention, the "sequence identity" between two sequences (A) and (B) is determined by comparing the two sequences in an optimally aligned manner over a comparison window. The alignment or comparison of sequences and the determination of the percent identity between two sequences can be accomplished using any method known in the art. For example, the comparison may involve a computational algorithm such as BLAST (Basic Local Alignment Search Tool).

[0065] Compare by method well known in the art, for example, use Needleman-Wunsch global alignment algorithm or Smith-Waterman local alignment algorithm.Analysis software uses similarity measurements owing to various disappearances and other modifications to mate similar sequences.Once total comparison is obtained, the sum of the nucleic acid residues of the same nucleic acid residues compared divided by the sum of the nucleic acid residues that the longest sequence in sequence (a) and (B) comprises can be obtained identity percentage, to compare two nucleotide sequences, for example, BLAST or EMBOSS Needle tool can be used.EMBOSS Needle uses the Needleman-Wunsch algorithm to create the best global comparison of two sequences.

[0066] When used herein to refer to a value, the term "about" refers to a value similar to the reference value in the context. Generally, one of ordinary skill in the art who is familiar with the context will understand the relative degree of variation encompassed by the term "about" in this context.

[0067] Unlike point mutations that only affect a single nucleotide, microsatellite mutations result in the acquisition or loss of an entire repeat unit (and sometimes two or more repeats simultaneously). Therefore, the mutation rate of microsatellite loci is expected to be different from other mutation rates, such as base substitution rates. One proposed reason for this length variation is replication slippage, which is caused by mismatches between the DNA strands being replicated during meiosis. DNA polymerase is the enzyme responsible for reading DNA during replication. It can slip while moving along the template strand and continue at the wrong nucleotide. When a repetitive sequence (such as CGCGCG) is replicated, DNA polymerase slippage is more likely to occur. Because microsatellites are composed of such repetitive sequences, DNA polymerase may make errors at a higher rate in these sequence regions, resulting in changes in the number of repeats and the total length of the microsatellite. Several studies have found evidence to support that slippage causes microsatellite mutations (Klintschar M et al.; Forster P et al.). Typically, slippage occurs approximately once every 1,000 generations in each microsatellite (Weber JL et al.). Therefore, slippage changes in repetitive DNA occur three orders of magnitude more often than point mutations in other parts of the genome (Jarne P et al.).

[0068] As explained in the Background section, "microsatellites" and "microsatellite regions", also called "short tandem repeats" ("STRs") or "simple sequence repeats" ("SSRs"), refer to repetitions of a pattern of one or more nucleotides (usually two (2) to six (6) or up to ten (10) nucleotide motifs), typically two (2) to fifty (50) repeats, with a minimum length of 5 or 6 bases. The term "tandem" means that the repeats are directly adjacent to each other. Microsatellites occur at many loci throughout the genome. Microsatellite sequences are either coding or non-coding sequences. Repeating units of one, two, three, four, five, and six nucleotides in a nucleotide sequence are referred to as mono-, di-, tri-, tetra-, penta-, and hexanucleotide motifs / repeats, respectively. For example, the sequence TATATATATA is a dinucleotide microsatellite, and GTCGTCGTCGTCGTC is a trinucleotide microsatellite (where "A" is adenine, "G" is guanine, "C" is cytosine, and "T" is thymine). Specific subclasses of microsatellites include homopolymers. As used herein, "homopolymer" refers to a microsatellite region that is a single nucleotide repeat of at least 6 (nucleo) bases; in other words, if viewed at the DNA level, it is a stretch of at least 6 consecutive guanine ("G"), adenine ("A"), cytosine ("C"), or thymine ("T") residues / base. Most specifically, when determining microsatellites, the genomic DNA of the subject (or the genomic DNA of a cancer present in the subject) is observed.

[0069] In the context of the present invention, the terms "microsatellite" and "microsatellite locus" preferably refer to a single nucleotide repeat or homopolymer, ie a repeat of only one of the four bases: A, T, C or G.

[0070] In preferred aspects described herein, the microsatellite is a repeat of "n" thymine ("T"), wherein "n" refers to the number of Ts, and wherein "n" is, for example, equal to 20, 21, 22, 23, 24, 25 or 26, preferably equal to 21, 22, 23 or 25.

[0071] In a preferred aspect of the invention, the microsatellite loci are selected from the group comprising or consisting of CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07, the loci being defined with reference to the Homo sapiens (human) reference genome assembly identified as GRCh38 (or hg38) from the Genome Reference Consortium human Build 38 patch release 14 (also known as GRCh38.p14) [see NCBI: GCA_000001405.15 GCF_000001405.26]. One of ordinary skill in the art will appreciate that the positions of the loci described herein may vary when referring to different reference genomes.

[0072] Therefore, the reference is the Homo sapiens reference genome assembly identified as GRCh38 or hg38:

[0073] -CABIO-P05 is a 21T duplication located at 14q23.1, starting at position chr14:58359108,

[0074] -CABIO-P07 is a 21T repeat located at 7q32, starting at position chr7:131478596,

[0075] -CABIO-E01 is a 22T repeat located at Xq22.3 starting at position chrX:106849221,

[0076] -CABIO-E03 is a 23T repeat located at Xq21.2 starting at position chrX:85268269,

[0077] -CABIO-E04 is a 25T repeat located at 14q32.3, starting at position chr14:103574079,

[0078] -CABIO-E05 is a 22T repeat located at 2p11.2, starting at position chr2:86456417,

[0079] - CABIO-E06 is a 23T repeat located at 4q23 starting at position chr4:99216136, and

[0080] -CABIO-E07 is a 21T repeat located at 20p13 starting at position chr20:290564.

[0081] CABIO-P05 can be amplified, for example, with the aid of the forward primer of SEQ ID NO:9 and / or the reverse primer of SEQ ID NO:10.

[0082] CABIO-P07 can be amplified, for example, with the aid of the forward primer of SEQ ID NO: 11 and / or the reverse primer of SEQ ID NO: 12.

[0083] CABIO-E01 can be amplified, for example, with the aid of the forward primer of SEQ ID NO: 13 and / or the reverse primer of SEQ ID NO: 14.

[0084] CABIO-E03 can be amplified, for example, with the aid of the forward primer of SEQ ID NO: 15 and / or the reverse primer of SEQ ID NO: 16.

[0085] CABIO-E04 can be amplified, for example, with the aid of the forward primer of SEQ ID NO: 17 and / or the reverse primer of SEQ ID NO: 18.

[0086] CABIO-E05 can be amplified, for example, with the aid of the forward primer of SEQ ID NO: 19 and / or the reverse primer of SEQ ID NO: 20.

[0087] CABIO-E06 can be amplified, for example, with the aid of the forward primer of SEQ ID NO: 21 and / or the reverse primer of SEQ ID NO: 22.

[0088] CABIO-E07 can be amplified, for example, with the aid of the forward primer of SEQ ID NO: 23 and / or the reverse primer of SEQ ID NO: 24.

[0089] The term "indel" (or "delins") refers to a class of mutations including insertions and deletions of single or multiple bases, and combinations thereof. It is classified as a small genetic variation with a length of from 1 to 10,000 base pairs. Indels insert or delete nucleotides from a sequence and are therefore different from point mutations that occur in the form of substitutions, which replace one of the nucleotides without changing the total number of nucleotides in the DNA sequence. In most known genomes, including humans, indel frequencies tend to be significantly lower than single nucleotide polymorphisms (SNPs), except near highly repetitive regions (including homopolymers and microsatellites). In the coding regions of the genome, unless the length of the indel is a multiple of 3, it will produce a frameshift mutation. Indels in microsatellite regions result in a net increase or loss of nucleotides.

[0090] The presence of an indel can be determined by comparing it to DNA in which no indels are present (e.g., comparing DNA from a tumor sample to germline DNA from a subject with a tumor), or in the case of a monomorphic microsatellite or homopolymer, by comparing it to a reference genome. According to specific embodiments, indels are particularly contemplated to have a length between 1 and 5 or 6 nucleotides (e.g., the length of the microsatellite or homopolymer is 1 to 5 or 6 nucleotides longer or shorter than the normal known length of the microsatellite or homopolymer). Note that since indels can be a combination of insertions and deletions, the altered nucleic acid sequence may be greater than the length difference (e.g., a deletion of 5 nucleotides combined with an insertion of 3 nucleotides results in a change in length of 2, but the sequence of the microsatellite may also have changed). However, most typically, an indel will be an insertion or deletion, typically an insertion or deletion of 1 or 2 nucleotides.

[0091] As used herein, the term "microsatellite status" can be one of two categories: the presence of microsatellite instability (MSI, a clonal or somatic change in the number of repeat DNA nucleotide units in a microsatellite), or microsatellite stability (MSS, also referred to herein as "absence of MSI").

[0092] Microsatellite instability (MSI) refers to a condition of inherited hypermutability (susceptibility to mutations, particularly indels), which can be caused by impaired DNA mismatch repair ("MMR"), also known as defective DNA mismatch repair ("dMMR"). Given the correlation between the absence of an intact mismatch repair (MMR) system and the presence of MSI, diagnosing the presence of MSI (or determining MSI status) can be interpreted as diagnosing MMR deficiency. In other words, if microsatellite stability (MSS) reflects or indicates an efficient mismatch repair system, then MSI is a molecular marker of a defective MMR system.

[0093] If a cell, DNA, tumor or subject has a microsatellite instability phenotype, or in other words, if its DNA has / exhibits microsatellite instability, then the cell, DNA, tumor or subject is identified herein as an "MSI cell, DNA, tumor or subject" or a "dMMR cell, DNA, tumor or subject" ("defective mismatch repair system cell, DNA, tumor or subject").

[0094] If a cell, DNA, tumor or subject has a microsatellite stable phenotype, or in other words, if its DNA does not have / does not exhibit microsatellite instability, then the cell, DNA, tumor or subject is identified herein as a "MSS cell, DNA, tumor or subject" or a "pMMR cell, DNA, tumor or subject" ("intact MMR cell, DNA, tumor or subject").

[0095] The most common method of detecting MSI is to measure the length of the polymerase chain reaction amplicon comprising the entire microsatellite. This usually requires DNA, primer pairs (one of which is typically fluorescent end-labeled), a sequencer and suitable software. Alternatively, if the sequencing amplicon, the quantity of repeat units can be simply counted. MSI can also be diagnosed indirectly by the staining loss of one of the mismatch repair genes detected by immunohistochemistry ("IHC"), because this staining loss also points to the abnormality in mismatch repair. The feature of both immunohistochemistry and genetic methods is that there are a large number of false negatives, and therefore, in a conventional diagnostic environment, the combined assessment of immunohistochemistry and genetic level is carried out.

[0096] There are at least 700,000 microsatellites in the human genome. Because microsatellite markers were initially randomly selected by researchers based on their own experiments, a meeting was held at Bethesda to discuss these issues and make recommendations to promote consistency across studies. This led to the development of guidelines for a "gold standard" marker panel (called the "Bethesda" panel). This panel consists of three dinucleotide repeats (D2S123, D5S346, and D17S250) and two mononucleotide repeats (BAT-26 and BAT-25). Recent developments in MSI detection kits retain BAT-25 and BAT-26 and replace the three dinucleotide markers with more sensitive mononucleotide markers (NR-21, NR-24, and MONO-27).

[0097] There is a proposal to consider a tumor MSI positive if 40% or more of the tested markers are unstable. When a five-marker panel is used, this means that MSI is called when at least two of them are positive; however, in tumors with MSI, four or all five markers are usually positive. Tumors that test negative for all five markers are called microsatellite stable (MSS). Although the initial Bethesda panel and the "advanced" Bethesda panel are still considered standard, they are known to have a rather low sensitivity (also depending on which MMR gene is mutated). In the field of cancer, another significant shortcoming is that the Bethesda guidelines are specific to colon cancer, even though other cancers that show MSI are known.

[0098] The inventors now describe herein a method of analyzing human DNA for a set of at least two or at least three microsatellite loci, preferably a set of at least four microsatellite loci, selected from the group comprising:

[0099] -CABIO-P05, defined as a 21T duplication located at 14q23.1 and starting at position chr14:58359108,

[0100] -CABIO-P07, defined as a 21T duplication located at 7q32 and starting at position chr7:131478596,

[0101] -CABIO-E01, defined as a 22T repeat located at Xq22.3 starting at position chrX:106849221,

[0102] --CABIO-E03, defined as a 23T repeat located at Xq21.2 starting at position chrX:85268269,

[0103] -CABIO-E04, defined as a 25T duplication located at 14q32.3 and starting at position chr14:103574079,

[0104] -CABIO-E05, defined as a 22T duplication located at 2p11.2 and starting at position chr2:86456417,

[0105] - CABIO-E06, defined as a 23T duplication located at 4q23 and starting at position chr4:99216136, and

[0106] - CABIO-E07, defined as a 21T repeat located at 20p13 and starting at position chr20:290564, with reference to the Homo sapiens (human) reference genome assembly designated GRCh38 or hg38 from the Genome Reference Consortium human Build 38 patch release 14 (also known as GRCh38.p14) [see NCBI: GCA_000001405.15 GCF_000001405.26].

[0107] In a specific aspect, the method comprises analyzing five, six, seven, or all of the eight loci described herein above.

[0108] Analysis of nucleic acid sequences, in particular the detection of indels in said sequences, can be performed by any method well known to those skilled in the art, with or without an amplification step. For example, indels can be detected by direct sequencing using well-known methods.

[0109] As described above, microsatellites are typically analyzed by conventional PCR amplification and amplicon sizing (also referred to herein as "MSI-PCR"), sometimes followed by DNA sequencing, particularly Sanger sequencing.

[0110] Can be by extracting nuclear DNA from the cell of sample of interest, then by polymerase chain reaction (PCR) amplification the specific polymorphic region of the DNA extracted and analyze.Once these sequences are amplified, they are just or resolved by gel electrophoresis or by capillary electrophoresis, which will allow the analyst to determine how many repetitions the microsatellite sequence in question has.If by gel electrophoresis analysis DNA, then can for example by silver staining (low sensitivity, safety, cheap) or use intercalating dye such as ethidium bromide (quite sensitive, medium health risk, cheap) or fluorescent dye (high sensitivity, safety, expensive) to make DNA visualization.Conventionally also use fluorescent dye to construct the instrument for resolving microsatellite fragments by capillary electrophoresis.

[0111] Using the unique sequences of the flanking regions as primers, microsatellites can be amplified by PCR for identification. DNA is repeatedly denatured at high temperatures to separate the double strands, and then cooled to allow primer annealing and nucleotide sequence extension through the microsatellite. This process results in the production of enough DNA to be visible on agarose or polyacrylamide gels; amplification requires only a small amount of DNA because thermal cycling in this way produces an exponential increase in the number of replicated fragments (Griffiths, AJF et al.). With the abundance of PCR technology, primers flanking the microsatellite locus are simple and fast to use, but developing primers that function correctly is generally a tedious and expensive process.

[0112] In the context of the present invention, multiple polymerase chain reaction (multiplex PCR) can be carried out.Multiplex PCR refers to the use of PCR, the use of multiple primers and temperature-mediated DNA polymerase to amplify several different DNA sequences simultaneously (as in a single reaction, many independent PCR reactions are carried out together). The primer design of all primer pairs / groups and the annealing temperature of each primer pair must be optimized so that all primers can work under the same annealing temperature during PCR. It is carried out in a single PCR mixture to produce amplicons of different sizes specific to different DNA sequences. When visualized by gel electrophoresis, amplicon sizes (i.e., their base pair lengths) should be sufficiently different to form different bands. Alternatively, if the amplicon sizes overlap, primers such as those dyed with fluorescent dyes of different colors can be used to distinguish and visualize different amplicons. Commercial multiplexing kits for PCR are available and are used to amplify degraded DNA samples by many forensic laboratories. By once targeting multiple sequences, additional information can be obtained from a single test run, otherwise several times the reagents and more time will be needed to perform the single test run.

[0113] In a preferred aspect, the method for analyzing microsatellite locus of the present invention comprises the step of using suitable primers to amplify the set of microsatellite locus from at least one sample of DNA, preferably in a multiplex amplification reaction. Especially, the size of the dna fragmentation of amplification can be determined and / or the dna fragmentation of amplification can be checked order. In an advantageous embodiment, this group of microsatellite locus comprises at least two, for example three, four, five, six, seven or eight loci that are selected from CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07, preferably at least four, five, six or seven loci of this group, or all eight loci of this group.

[0114] In one specific aspect, at least one of the primers used in the method has a nucleic acid sequence selected from the group of primer sequences identified by SEQ ID NOs: 9 to 24. In another specific aspect, at least one of the primer pairs used in the method has a sequence pair selected from the group consisting of: SEQ ID NO: 9 and SEQ ID NO: 10; SEQ ID NO: 11 and SEQ ID NO: 12; SEQ ID NO: 13 and SEQ ID NO: 14; SEQ ID NO: 15 and SEQ ID NO: 16; SEQ ID NO: 17 and SEQ ID NO: 18; SEQ ID NO: 19 and SEQ ID NO: 20; SEQ ID NO: 21 and SEQ ID NO: 22; and SEQ ID NO: 23 and SEQ ID NO: 24.

[0115] In yet another specific aspect, the set of loci is co-amplified using a pair of specific oligonucleotide primers for each locus of interest, each having a sequence as shown below:

[0116] If the locus is CABIO-P05, SEQ ID NO: 9 and SEQ ID NO: 10,

[0117] If the locus is CABIO-P07, SEQ ID NO: 11 and SEQ ID NO: 12,

[0118] If the locus is CABIO-E01, SEQ ID NO: 13 and SEQ ID NO: 14,

[0119] If the locus is CABIO-E03, SEQ ID NO: 15 and SEQ ID NO: 16,

[0120] If the locus is CABIO-E04, SEQ ID NO: 17 and SEQ ID NO: 18,

[0121] If the locus is CABIO-E05, SEQ ID NO: 19 and SEQ ID NO: 20,

[0122] If the locus is CABIO-E06, SEQ ID NO: 21 and SEQ ID NO: 22, or

[0123] If the locus is CABIO-E07, SEQ ID NO:23 and SEQ ID NO:24, or sequences having at least 90% identity thereto (functionally equivalent, i.e., suitable for amplifying a specific locus) (depending on the nature of the target locus of interest, the locus being selected from the group consisting of CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06, and CABIO-E07).

[0124] The analysis of microsatellites, particularly indels in microsatellites, does not necessarily require the use of full PCR. In a specific aspect, the method of the present invention uses only extension of primers by a single fluorescently labeled dideoxyribonucleic acid molecule (ddNTP) complementary to the nucleotide to be studied. The nucleotide at the polymorphic site can be identified by detecting a primer that has been extended by one base and, for example, fluorescently labeled (e.g., Kobayashi et al., Mol. Cell. Probes, 9: 175-182, 1995).

[0125] A specific technique involves "enriching" the extracted DNA using oligonucleotide sequences composed of repeats complementary to those in microsatellites ("microsatellite enrichment"). The oligonucleotide probe hybridizes to the repeats in the microsatellite, and the probe / microsatellite complex is then pulled from the solution. The enriched DNA is then cloned as usual (Ostrander et al., Proc. Natl. Acad. Sci. USA, 15;89(8):3419-23, 1992).

[0126] Another method of performing nucleic acid sequence analysis according to the present invention involves capturing the target nucleic acid sequence of interest with a suitable detectable probe.

[0127] In a specific aspect, allele-specific oligonucleotide hybridization (ASO) technology (such as Saiki et al. or Stoneking et al.) can be used. This technology relies on differentiating two DNA molecules that differ by one base by hybridizing an oligonucleotide probe that is specific to one of the variants with an amplified product obtained from an amplified nucleic acid sample. The method generally uses short oligonucleotides, for example, 15-30 bases in length. The probe is designed to hybridize with a variant relative to another variant difference. The principles and guidelines for designing such probes are available in the art. Hybridization conditions should be sufficiently strict so that there are significant differences in the hybridization intensity between the alleles and a substantially binary response is produced, whereby the probe only hybridizes with one of the alleles. The amount and / or presence of a specific sequence is determined by measuring the amount of the sequence-specific oligonucleotide hybridized with the sample. Typically, oligonucleotides are labeled with markers such as fluorescent markers for detection. For example, sequence-specific oligonucleotides are applied to immobilized oligonucleotides representing sequences with different microsatellite lengths. After stringent hybridization and washing conditions, the fluorescence intensity of each microsatellite oligonucleotide is measured. Suitable assay schemes for detecting the hybrid formed between the target nucleic acid sequence in the probe and the sample are known in the art, and include immobilized target (dot blotting) schemes and immobilized probe (reverse dot blotting or line blotting) assay schemes. In the dot blotting scheme, the target DNA of amplification is fixed on a solid support such as a nylon membrane. Film-target complex is hatched under suitable hybridization conditions with the probe of labeling, unhybridized probe is removed by washing under suitable stringent conditions, and the presence of the probe of combination in the monitoring film is observed. In the reverse dot blotting (or line blotting) scheme, the probe is fixed on a solid support, such as a nylon membrane or a microtiter plate. Target DNA is usually labeled by incorporating a labeled primer during amplification. One or two of the labeled primers can be used. Film-probe complex is hatched under suitable hybridization conditions with the amplified target DNA of labeling, unhybridized target DNA is removed by washing under suitable stringent conditions, and the presence of the target DNA of combination in the monitoring film is observed.

[0128] In another specific aspect, the nucleic acid sequence analysis method of the present invention involves detectable hybridization probes that are sequence-specific probes that discriminate between sequences with and without indels.

[0129] In a specific aspect described herein, the probe may have the sequence described below:

[0130] If the locus is CABIO-P05, SEQ ID NO: 25,

[0131] If the locus is CABIO-P07, SEQ ID NO: 26,

[0132] If the locus is CABIO-E01, SEQ ID NO: 27,

[0133] If the locus is CABIO-E03, SEQ ID NO: 28,

[0134] If the locus is CABIO-E04, SEQ ID NO: 29,

[0135] If the locus is CABIO-E05, SEQ ID NO: 30,

[0136] If the locus is CABIO-E06, SEQ ID NO: 31, or

[0137] If the locus is CABIO-E07, SEQ ID NO:32, or (functionally equivalent, i.e., suitable for amplifying a specific locus) a sequence having at least 90%, e.g., 91%, 92%, 93% or 94%, or at least 95%, e.g., 96%, 97%, 98% or 99% sequence identity thereto (i.e., identity to one of SEQ ID NOs:25-32, depending on the nature of the target locus of interest selected from CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07).

[0138] In a specific aspect, the nucleic acid sequence analysis method of the present invention uses both sequence-specific primers and labeled probes that bind to amplification products.

[0139] Amplification products generated using polymerase chain reaction can be analyzed using denaturing gradient gel electrophoresis. Different sequences (alleles) can be identified based on the different sequence-dependent melting characteristics and electrophoretic migration of DNA in solution (see, for example, Erlich, ed., PCR Technology, Principles and Applications for DNA Amplification, WH Freeman and Co, New York, 1992, Chapter 7). Capillary electrophoresis can be used to distinguish microsatellite polymorphisms. Capillary electrophoresis conveniently allows identification of the number of repeats in a specific microsatellite sequence (allele). The application of capillary electrophoresis in DNA polymorphism analysis is well known to those skilled in the art (see, for example, Szantai et al., J Chromatogr A. (2005) 1079 (1-2): 41-9; Bjorheim and Ekstrom, Electrophoresis (2005) 26 (13): 2520-30 and Mitchelson, Mol Biotechnol. (2003) 24 (1): 41-68). Single-stranded conformational polymorphism analysis can be used to distinguish alleles of a target sequence by identifying base differences by changing the electrophoretic migration of single-stranded PCR products, as described in, for example, Orita et al., Proc. Nat. Acad. Sci. 86, 2766-2770 (1989). Amplified PCR products can be produced as described above and denatured by heating or otherwise to form single-stranded amplified products. Single-stranded nucleic acids can refold or form secondary structures that depend in part on the base sequence. The electrophoretic mobility differences of single-stranded amplified products may be related to base sequence differences between the target gene alleles.

[0140] Nucleic acid analysis (particularly insertion / deletion detection) methods typically employ labeled oligonucleotides. Oligonucleotides can be labeled by incorporating a label detectable by spectroscopy, photochemistry, biochemistry, immunochemistry, or chemical means. Useful labels include fluorescent dyes, radioactive labels (e.g., 32P), electron-dense reagents, enzymes (such as peroxidase or alkaline phosphatase), biotin or haptens, and proteins for which corresponding antisera or monoclonal antibodies are available. Labeling techniques are well known in the art (see, e.g., Sambrook et al.). Microsatellite locus marker sets / groups as herein described can be detected using any of these techniques or other techniques, and the marker set is independent of the technique used.

[0141] According to Tytgat, O. et al., repetitive DNA is not easily analyzed by next-generation DNA sequencing (NGS) methods, which struggle with homopolymeric tracts. Therefore, microsatellites are typically analyzed by conventional PCR amplification and determination of amplicon size. The use of PCR means that microsatellite length analysis is susceptible to PCR limitations, just like any other PCR-amplified DNA locus.

[0142] Advantageously, the nucleic acid sequence analysis method of the present invention that allows determination of microsatellite status is compatible with NGS (next generation sequencing) (also known as "massively parallel sequencing"), particularly second generation high throughput sequencing. Many NGS platforms differ in their engineering configuration and sequencing chemistry. They share a technical paradigm of massively parallel sequencing of single DNA molecules in spatially separated, clonally amplified DNA templates or flow cells. This design differs from Sanger sequencing, also known as capillary sequencing or first generation sequencing, which is based on electrophoretic separation of chain termination products produced in a single sequencing reaction.

[0143] DNA sequencing using commercially available NGS platforms is typically performed in the following steps. First, a DNA sequencing library is typically generated by clonal amplification using in vitro PCR. Second, DNA is sequenced by synthesis, such that the DNA sequence is determined by adding nucleotides to the complementary chain rather than by chain termination chemistry. Third, the spatially separated amplified DNA templates are sequenced simultaneously in a massively parallel manner without the need for physical separation steps.

[0144] Most NGS platforms follow these steps, but each platform employs a different strategy.

[0145] Two approaches are used to prepare DNA templates for NGS reactions: amplified templates derived from single DNA molecules and single DNA molecule templates. For imaging systems that cannot detect individual fluorescence events, amplification of the DNA template is necessary. The three most common amplification methods are emulsion PCR (ePCR), rolling circle amplification, and solid-phase amplification. The final distribution of the template can be spatially random or on a grid.

[0146] In the emulsion PCR method, a DNA library is first generated by random fragmentation of genomic DNA. Single-stranded DNA fragments (templates) are attached to the surface of beads using adapters or linkers, and one bead is attached to a single DNA fragment from the DNA library. The surface of the beads contains oligonucleotide probes whose sequences are complementary to the adapters that bind the DNA fragments. The beads are then separated into water-oil emulsion droplets. In the aqueous water-oil emulsion, each of the droplets that capture a bead is a PCR microreactor that produces an amplified copy of a single DNA template.

[0147] Amplification of a population of single DNA molecules by rolling circle amplification in solution is followed by capture on a grid of spots smaller in size than the DNA to be immobilized.

[0148] In the context of generating DNA clones, forward and reverse primers are covalently attached to a glass slide in a flow cell at high density. The ratio of primers to template on the support defines the surface density of amplification clusters. The flow cell is exposed to reagents for polymerase-based extension, and priming occurs when the free / distal ends of the ligated fragments "bridge" to complementary oligonucleotides on the surface. Repeated denaturation and extension result in local amplification of DNA fragments at millions of individual locations on the flow cell surface. Solid-phase amplification generates up to 20 billion spatially separated template clusters, providing free ends that subsequently hybridize with universal sequencing primers to initiate the sequencing reaction.

[0149] A similar but non-clonal surface amplification method, called "bridge amplification," was adapted for clonal amplification by Church and Mitra (1999) in 1997.

[0150] The scheme that needs DNA amplification is usually implemented very loaded down with trivial details, and may introduce sequencing errors.The preparation of unimolecule template is a more direct scheme, and it does not need PCR, and PCR can introduce errors in the template of amplification.Usually one of at least three different methods is used that unimolecule template is fixed on solid support.In the first method, the single primer molecule of spatial distribution is covalently attached to solid support.By starting material random fragmentation into small size (for example, about 200-250bp) and adding common adapter at the fragment end to prepare template, then this template is hybridized with fixed primer.In the second method, by triggering and extending single strand, unimolecule template from fixed primer, the unimolecule template of spatial distribution is covalently attached to solid support.Then common primer is hybridized with this template.In either method, DNA polymerase can be attached to the template configuration of fixed initiation to start NGS reaction.In the third method, the single polymerase molecule of spatial distribution is connected to the solid support that the template molecule of initiation is combined. Larger DNA molecules (up to tens of thousands of base pairs) can be used with this technology, and unlike the first two methods, the third method can be used with real-time approaches, leading to potentially longer read lengths.

[0151] In a specific aspect, the template preparation method is based on the use of emPCR (or clonal-emPCR), gridded DNA nanospheres (or gridded rolling circle nanospheres), DNA colony generation (or bridge amplification or clonal-bridge amplification), or single molecules.

[0152] Sequencing methods suitable for NGS platforms include, for example, pyrosequencing, reversible dye terminators, oligonucleotide 8-mer chain ligation, oligonucleotide 9-mer chain ligation, natural dNTP proton detection, or phosphate-linked fluorescent nucleotides.

[0153] Pyrosequencing is a non-electrophoretic bioluminescent method that measures the release of inorganic pyrophosphate by converting it proportionally into visible light through a series of enzymatic reactions. Unlike other sequencing methods that use modified nucleotides to terminate DNA synthesis, pyrosequencing manipulates DNA polymerase through the single addition of a limited number of dNTPs. Once the complementary dNTP is incorporated, the DNA polymerase extends the primer and pauses. DNA synthesis resumes after the next complementary dNTP is added in the cycle. The order and intensity of the light peaks are recorded as a flow graph, which reveals the actual DNA sequence.

[0154] The sequencing by reversible terminator chemical method uses the dNTP of reversible terminator combination in the cyclic method that comprises nucleotide incorporation, fluorescence imaging and cutting.When adding each dNTP and then cutting to allow incorporation into the next base, the fluorescent-labeled terminator is imaged.These nucleotides are chemically sealed so that each incorporation is a unique event.Each base incorporation step is followed by an imaging step, and then chemical removal is carried out through the blocked group to prepare each chain for the next incorporation by DNA polymerase.This series of steps continues the cycle of a specific number, as determined by the user-defined instrument setting.Initially, 3' blocking groups are designed to be enzymatic or chemically reversed.Sequencing by reversible terminator chemistry can be four-color cycles or single-color cycles.Can use " virtual terminator ", it is the unblocked terminator with the second nucleoside analog that serves as an inhibitor.These terminators have the appropriate modification for termination or inhibitory groups so that DNA synthesis stops after adding single base.

[0155] In the ligation sequencing method mediated by ligase, the sequence extension reaction is not performed by a polymerase, but by DNA ligase and a base-coded probe or a two-base-coded probe. In its simplest form, a fluorescently labeled probe hybridizes with the complementary sequence of the template it is adjacent to. DNA ligase is then added to connect the dye-labeled probe to the primer. The unconnected probe is washed away and then fluorescence imaging is performed to determine the identity of the connected probe. This cycle can be repeated by using a cleavable probe to remove the fluorescent dye and regenerate the 5'-PO4 group for subsequent connection cycles (chain connection) or by removing a new primer and hybridizing it with the template (non-chain connection).

[0156] The real-time sequencing method involves imaging the continuous incorporation of dye-labeled nucleotides during DNA synthesis: a single DNA polymerase molecule is attached to the bottom surface of a single zero-mode waveguide detector (ZMW detector), which can obtain sequence information as phosphate-linked nucleotides are incorporated into the growing primer chain. Unique DNA polymerases can be used that are better at incorporating phosphate-linked nucleotides and allow resequencing of closed circular templates.

[0157] The inventors compared the method of the present invention with NGS sequencing of the five most commonly used analytical microsatellites in accordance with the Bethesda and ESMO international guidelines. They also compared NGS sequencing of microsatellite panels (including both the present panel and the reference panel in the field) with conventional techniques of immunohistochemistry ("IHC") and "MSI-PCR", and described herein the results showing the superiority of the method of the present invention, and in particular the superiority of the panel of up to 8 microsatellite loci disclosed herein in terms of sensitivity and specificity.

[0158] In a preferred aspect, the method of analyzing microsatellite loci of the present invention comprises the step of sequencing a set of microsatellite loci as described herein, preferably (as taught herein above) a set of at least 2 to 8, even more preferably at least 4, in particular 4 to 8 microsatellite loci.

[0159] In a specific aspect, this method can further include the step of a collection of this microsatellite locus of amplification together. As will be appreciated by those skilled in the art, sequence-specific amplification methods can be carried out in the reaction using multiple sequence-specific primers to target specific microsatellite sequences. Primers used for this multiple application are usually labeled by distinguishable markers or are selected to make the amplification products produced from the alleles distinguishable by size. Therefore, for example, two alleles in a single sample can be identified using single amplification by gel analysis of amplification products. As in the case of sequence-specific probes, sequence-specific oligonucleotide primers can be accurately complementary to one of the polymorphic sequences in the hybridization region, or can have some mispairings in the position beyond the 3 ' end of the oligonucleotide, and the mispairings occur at the non-polymorphic sites in the two nucleotide sequences.

[0160] Thus, any primer of SEQ ID NOs: 9 to 24 or any probe of SEQ ID NOs: 25 to 32 described herein, or sequences having at least 90%, e.g., 91%, 92%, 93% or 94%, or at least 95%, e.g., 96%, 97%, 98% or 99% sequence identity thereto, and any combination thereof, may be advantageously used in the context of such methods.

[0161] This specification relates particularly to methods, in particular computer-implemented methods for assessing (or determining) the microsatellite stability (MSS) or microsatellite instability (MSI) status of a tumor. The method comprises the following steps:

[0162] a) counting the number of indels in at least two, such as at least three, preferably at least four, five, six, seven or eight microsatellite loci in a DNA sample, preferably a tumor DNA sample, wherein the tumor is preferably a human tumor and the microsatellite loci are preferably selected from the group comprising: CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07 as described herein;

[0163] b) for each of the at least two microsatellite loci, comparing the number of indels to a reference number of indels for that locus and determining the MSI status of the locus, if the indel count is equal to or higher (higher than) (≥) the reference value for that locus, deeming the locus unstable, and if the indel count is lower (<) the reference value for that locus, deeming the locus stable;

[0164] c) calculating an MSID score, which ranges between 0 and 1 (inclusive), consisting of the total number of unstable loci relative to the total number of loci; and

[0165] d) determining the MSI status of the tumor as unstable if the MSID score is equal to or higher (higher than) (≥) 0.5; as stable if the MSID score is equal to or lower (≤) 0.125, and as uncertain if the MSID score is between 0.125 and 0.5; and optionally

[0166] e) If the MSID score is between 0.125 and 0.5, repeating steps a)-d) until the status of the MSI can be determined.

[0167] In a preferred aspect, the method comprises amplifying at least two, such as at least three, preferably at least four microsatellite loci, such as five, six, seven or eight loci. The group of at least two loci is preferably selected from the group comprising CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07 disclosed for the first time herein.

[0168] In a specific aspect of the invention, the reference numbers of the indels for each locus of interest identified herein are provided below (= instability threshold: mean + 1 SD):

[0169] If the locus is CABIO-P05, the reference number for the indel is 26;

[0170] If the locus is CABIO-P07, the reference number for the indel is 31;

[0171] If the locus is CABIO-E01, the reference number for the indel is 32;

[0172] If the locus is CABIO-E03, the reference number for the indel is 24;

[0173] If the locus is CABIO-E04, the reference number for the indel is 28;

[0174] If the locus is CABIO-E05, the reference number for the indel is 26;

[0175] If the locus is CABIO-E06, the indel reference number is 31; and

[0176] If the locus is CABIO-E07, the reference number for the indel is 17.

[0177] In a specific and preferred aspect of the invention, the disease is cancer.All or part of the set of eight microsatellite loci described herein can be used to assess / detect microsatellite status in cancer.

[0178] The markers described herein can advantageously be used to determine the microsatellite (MSI or MSS) status independent of the cancer type. Thus, in principle, the microsatellite status can be diagnosed using the markers provided herein for each type of cancer.

[0179] As used herein, the term "cancer" refers to different diseases involving unregulated cell growth, also referred to as malignant tumors. The term "tumor" is used as a synonym in this application. It is expected that the term covers all solid tumor types (cancer, sarcoma, blastoma), but also clearly covers non-solid cancer types, such as blood diseases, lymphomas or myelomas. Therefore, a "tumor DNA sample" can also be a blood sample from a person suffering from leukemia. Typically, a tumor DNA sample has been separated from a subject, particularly a subject suffering from cancer, at some point. Optionally, it has undergone one or more forms of pre-treatment (e.g., cracking, classification, separation, purification) in order to sequence the DNA, although it is also envisioned that the DNA from an untreated sample is sequenced.

[0180] In one specific aspect, the cancer is a mismatch repair (MMR-) deficient tumor or cancer (known to be associated with microsatellite instability), in particular a tumor or cancer with a high dMMR (defective mismatch repair system) prevalence, i.e., greater than 15% (>15%), such as a gastrointestinal tumor, such as a colorectal or gastric tumor, or an endometrial tumor. Colorectal cancer can be, for example, colon cancer, rectosigmoid junction cancer, rectal cancer, anus and / or anal canal cancer. In another specific aspect, the cancer is a cancer with a low dMMR prevalence (<1%), such as prostate cancer and kidney cancer.

[0181] In a specific aspect, the cancer is ovarian cancer.

[0182] In a preferred aspect, the cancer is colorectal cancer or endometrial cancer.

[0183] In another aspect, the disease is a condition, such as an autosomal dominant genetic condition, e.g., Lynch syndrome (also known as "HNPCC syndrome"), which is associated by physicians with a higher risk of developing colorectal cancer or other cancers, including endometrial cancer, ovarian cancer, gastric cancer, small intestine cancer, hepatobiliary tract cancer, upper urinary tract cancer, brain cancer, and skin cancer. The increased risk of these cancers is due to inherited mutations that impair DNA mismatch repair.

[0184] In a preferred aspect, tumor DNA is obtained from a biological sample of a subject suffering from a disease, particularly cancer, or suspected of suffering from such a disease, and the biological sample is a solid, fluid or semi-fluid sample, particularly a sample suitable for detecting tumor cells.

[0185] In one aspect, the biological sample is a biopsy, in particular a solid or liquid biopsy.

[0186] A biological sample can be a solid tissue biopsy. A tissue biopsy requires solid material from the subject's body. This biopsy is typically taken from a solid tumor or from a tissue or organ suspected of containing tumor cells. A tissue biopsy is typically used when the location of a tumor is suspected or confirmed and is available for extraction.

[0187] Additionally, the biological sample can be a liquid biopsy. A liquid biopsy sample is, for example, a blood, plasma, serum, sputum, bronchial fluid, or pleural effusion sample.

[0188] In a specific and preferred aspect, the biological sample is a sample comprising or consisting of tumor DNA.A "tumor DNA sample" refers to any sample that can be used as a basis for sequencing and in which DNA from a cancer is present.

[0189] Various methods for extracting / isolating nucleic acids, such as DNA, from a sample are known in the art. Nucleic acids can be isolated using, for example, but not limited to, standard DNA purification techniques such as organic extraction or solid phase extraction. DNA extraction can be performed using kits available in the art, such as, for example, the KAPA Express extraction kit (Kapa Biosystems, Wilmington, MA) or 16FFPE LEV DNA Purification Kit (Promega Corporation, Madison, WI). Since the level of purification of the extracted (DNA) material may have an impact on the MSID score, in a preferred aspect of the present invention, when performing MSI analysis using the method of the present invention, a known extraction method with a higher purity grade is used. For example, DNA extraction was performed using the 16FFPELEV DNA Purification Kit (Promega Corporation, Madison, WI) or any other known kit showing similar performance.

[0190] As used herein, the term "subject" refers to an individual vertebrate, more particularly an individual mammal, most particularly and preferably an individual human. As used herein, a "subject" is typically a human, but can also be a mammal, particularly a domesticated animal such as a cat, dog, rabbit, guinea pig, ferret, rat, mouse, etc., or a farm animal such as a horse, cow, pig, goat, sheep, llama, etc. The subject can also be a non-mammalian vertebrate, such as a fish, reptile, amphibian, or bird; essentially, any animal that can develop cancer meets this definition.

[0191] In a specific aspect, the subject has or is suspected of having cancer (as defined herein).For example, the subject has an MMR-deficient tumor.

[0192] In tumor cells with impaired replication control, microsatellites may be gained or lost at particularly high frequencies during each round of mitosis. Consequently, tumor cell lines may display a genetic fingerprint distinct from that of the host tissue and may exhibit loss of heterozygosity.

[0193] In another specific aspect, the subject is a patient having a tumor that has been identified as sensitive to a particular therapeutic agent or therapy, as determined by the microsatellite status of the tumor.

[0194] In specific aspects described herein, the methods disclosed herein by the inventors are used to assess microsatellite instability in precancerous and / or cancerous cells; to detect a predisposition to develop cancer; to assess the prognosis of cancer; to monitor cancer progression or regression; to predict, assess, or monitor response to cancer treatment; to select an appropriate cancer treatment for a subject in need thereof; to select a patient who is likely to respond to a cancer treatment; or to select a patient for enrollment in a clinical trial for a cancer treatment.

[0195] The method generally comprises the steps of analyzing a set of at least four microsatellite loci of human DNA, the microsatellite loci being selected from the group comprising CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06, and CABIO-E07, as first described above. In a preferred aspect, the method is a method of assessing the microsatellite stability (MSS) or microsatellite instability (MSI) status of a tumor as first described above. The method preferably comprises the steps of:

[0196] a) counting the number of indels in at least four microsatellite loci in a tumor DNA sample, wherein the microsatellite loci are selected from the group comprising: CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06, and CABIO-E07 as described herein;

[0197] b) for each of the at least four microsatellite loci, comparing the number of indels to a reference number of indels for that locus and determining the MSI status of the locus, if the indel count is ≥ the reference value for that locus, deeming the locus unstable, and if the indel count is < the reference value for that locus, deeming the locus stable;

[0198] c) calculating an MSID score, which ranges between 0 and 1 and consists of the total number of unstable loci relative to the total number of loci; and

[0199] d) determining the MSI status of the tumor as unstable if the MSID score is ≥ 0.5; as stable if the MSID score is ≤ 0.125, and as uncertain if the MSID score is between 0.125 and 0.5; and optionally

[0200] e) If the MSID score is between 0.125 and 0.5, repeating steps a)-d) until the status of the MSI can be determined.

[0201] As used herein, the terms "diagnose" and "diagnose" refer to determining (and / or the qualitative or quantitative probability / likelihood) whether a subject has or will develop a disease, disorder, condition, or state. For example, in diagnosing cancer, diagnosis can include determining the risk, type, stage, malignancy, or other classification of the cancer. In some cases, for example, as described herein, diagnosis can be or include determining a prognosis and / or a likely response to one or more general or specific therapeutic agents or regimens.

[0202] "Diagnosing the microsatellite status of a tumor," "diagnosing the microsatellite MSI or MSS status of a tumor," or "diagnosing the microsatellite status of a tumor in a subject" are all considered synonymous herein. Determining (or diagnosing) microsatellite status generally means concluding that MSI or MSS is present based on the presence of one or more indels detected in the microsatellite region under investigation, or concluding that microsatellite instability is absent based on the absence of indels detected in the microsatellite region under investigation. Thus, "determining the presence of an indel" in a microsatellite region means assessing or detecting the presence or absence of an indel in the microsatellite region. Similarly, determining the presence of an indel in at least two microsatellite regions means assessing or detecting the presence or absence of an indel in each of the at least two microsatellite regions.

[0203] The term "treatment" refers to any action intended to improve a patient's health, such as treatment, prevention, prophylaxis, and delaying the symptoms of a disease or disease. It refers to the curative treatment and / or preventative treatment of a disease. Curative treatment is defined as treatment that results in a cure or treatment that alleviates, improves, and / or eliminates, reduces, and / or stabilizes a disease or disease symptom or the pain it causes directly or indirectly. Preventative treatment includes treatment that results in disease prevention and treatment that reduces and / or delays the progression and / or morbidity of a disease or its risk of occurrence. In certain aspects, such terms refer to the improvement or eradication of a disease, disorder, infection, or the symptoms associated therewith. In other aspects, the term refers to minimizing the spread or worsening of a disease. Treatment according to the present invention does not necessarily mean 100% or complete treatment. On the contrary, there are different degrees of treatment that a person of ordinary skill in the art would consider to have potential benefits or therapeutic effects. Preferably, the term "treatment" refers to the application or administration of a composition comprising one or more active agents to a subject suffering from a disorder / disease.

[0204] In one specific aspect, the treatment is cancer therapy. Preferably, the anti-cancer treatment is selected from the group consisting of resection, chemotherapy, radiotherapy, or immunotherapy. Preferably, the therapeutic compound is a chemotherapeutic or immunotherapeutic compound. Chemotherapeutic compounds can be, but are not limited to, alkylating agents, antimetabolites, plant alkaloids, topoisomerase inhibitors, and antitumor antibiotics. Immunotherapeutic compounds can be, for example, but are not limited to, antibodies, cytokines, or interferons.

[0205] In a specific aspect described herein, the methods disclosed herein are used to detect or identify MMR-deficient tumors, predict, assess, or monitor response to cancer treatment, or to select patients who are likely to respond to cancer treatment.

[0206] For example, published work has shown that tumors known to be mismatch repair (MMR-) deficient or cancers associated with microsatellite instability (as described herein) have different responses to standard treatments, typically standard chemotherapy such as 5-fluorouracil (5-fluoracil) for colorectal cancer, or cisplatin and carboplatin for endometrial cancer, as well as alkylating agents such as temozolomide, and emerging targeted therapies. For example, preclinical studies have shown that patients with these familial dMMR cancers exhibit increased sensitivity to anti-EGFR (e.g., gefitinib, erlotinib, cetuximab, panitumumab) and VEGF-targeted therapies. The exact cause of this heterogeneity is unknown, but determining the status of microsatellites (MSI or MSS), or in other words, the presence or absence of mutations that result from MMR deficiency, is crucial for determining treatment outcome.

[0207] The method preferably includes a step of assessing the microsatellite stability (MSS) or microsatellite instability (MSI) status of the tumor, which involves analyzing a set of at least two, preferably at least four, microsatellite loci of human DNA, selected from the group comprising CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07, as first described above.

[0208] In certain cases, the method preferably comprises the additional step of optimizing treatment based on the microsatellite status (i.e., based on the discovery of whether the cancerous tumor is MSI or MSS), by determining, in particular, that if the tumor is identified as an MSS tumor, the tumor does not respond (if the tumor has been previously exposed to the specific therapeutic agent / therapy) or will not respond (if the tumor has never been previously exposed to the specific therapeutic agent / therapy) to the specific therapeutic agent / therapy, and, conversely, if the tumor is identified as an MSI tumor, the tumor responds (if the tumor has been previously exposed to the specific therapeutic agent / therapy) or will respond (if the tumor has never been previously exposed to the specific therapeutic agent / therapy), i.e., cancer cells will die, stop growing, or have reduced proliferation when treated therewith.

[0209] In a specific aspect, the method preferably further comprises an additional step for predicting a response to an immune checkpoint inhibitor ("ICI", e.g., pembrolizumab, dositalizumab, and similar compounds), detection of MSI or dMMR predicting a (clinical) response to ICI, whereas detection of MSS or pMMR predicts the absence of a (clinical) response (e.g., resistance) to ICI. In certain cancer types, such as, in particular, EC, CRC, cholangiocarcinoma, and pancreatic cancer, this additional step for predicting a response to ICI is preferably performed prior to any therapeutic treatment of a subject with cancer [André T et al., Berton D et al., Marabelle A et al.].

[0210] In another specific aspect, the method further comprises the additional step of selecting an appropriate cancer treatment for a subject in need thereof based on the microsatellite status (i.e., based on whether the cancerous tumor is found to be MSI or MSS). For example, in this regard, adjuvant chemotherapy with a fluoropyrimidine plus oxaliplatin improved disease-free survival in MMR-deficient CRC compared to systemic fluoropyrimidine therapy alone [Tougeron et al].

[0211] Selecting the most appropriate (optimized) treatment for a given subject (patient), i.e., a therapeutic product to which the MSS or MSI tumor is sensitive, even if the MSS or MSI tumor is resistant to standard treatment, is an approach to limit, and preferably overcome, resistance that may be observed when using conventional therapies such as targeted therapies.

[0212] In another specific aspect, the method further comprises the step of treating the subject, preferably with an appropriate or optimized cancer therapy.

[0213] In these methods, the cancer is preferably selected from a gastrointestinal cancer, such as colorectal cancer or gastric cancer, or endometrial cancer.

[0214] Also disclosed herein is a method for generating a personalized cancer treatment report, the method comprising obtaining a sample from a subject having or suspected of having cancer, determining the microsatellite status of the tumor using the method first described herein by the inventors, and generating a personalized cancer treatment report to remember the presence or absence of microsatellite instability in the subject (i.e., in the subject's tumor).

[0215] In a particular aspect, the cancer treatment report includes one or more of: (i) a suspected genetic predisposition, (ii) information regarding prognosis, resistance to treatment, or possible therapy options; (iii) information regarding the likely effectiveness of therapy options; (iv) the acceptability of therapy options, or the desirability of applying the selected therapy to the subject; or (v) information regarding drug administration.

[0216] Further provided herein is a method for screening cancer cells for sensitivity or resistance to a particular treatment, particularly treatment with a test compound, comprising determining the microsatellite status of the cancer cells using the methods disclosed herein for the first time by the inventors, and correlating the status with the sensitivity or resistance of the cancer cells to the treatment, particularly treatment with the test compound. According to a specific aspect, screening for sensitivity or resistance is used to select an appropriate cancer treatment for a subject suffering from cancer, or to stratify or classify subjects for clinical trials.

[0217] Although the methods described herein can in principle be performed in vivo, ex vivo and in vitro, it is specifically envisaged that they are performed in vitro.

[0218] In a preferred embodiment, the method of the present invention is a partially or fully computer-implemented method.

[0219] The term "computer-implemented method" refers to a method involving programmable apparatus / device, in particular a computer, a computer network or a readable medium carrying a computer program, wherein at least one step of the method is performed by using at least one computer program. A computer-implemented method may further include at least one step that is not performed by using a computer program.

[0220] In a specific and preferred aspect, the method of the invention makes it possible to distinguish between the MSS state / profile and the MSI state / profile for a particular tumor.

[0221] In another specific aspect, the method of the present invention enables, for a specific biological sample (particularly a tumor sample) from a subject, to distinguish between a healthy or diseased (particularly cancerous) state / profile of the biological sample (and therefore the subject) based on the MSS or MSI status / profile of the biological sample.

[0222] Also described herein is a computer-implemented method for training a classifier for (accurately) determining the microsatellite stability (MSS) or microsatellite instability (MSI) status (also referred to herein as "phenotype") of a biological sample, particularly a tumor, e.g., a human tumor, wherein the method comprises:

[0223] a) providing a training set of microsatellite loci or preprocessed information obtained from the training set as input to the classifier, each locus being obtained from a DNA sequence of interest, the training set comprising i) stable (MSS) microsatellite loci or subsequences thereof obtained from intact mismatch repair system (pMMR) or MSS cells, DNA, tumors or subjects known to have a microsatellite stable state or phenotype, and ii) unstable (MSI) microsatellite loci or subsequences thereof obtained from deficient mismatch repair system (dMMR) or MSI cells, DNA, tumors or subjects, particularly tumor cells, known to have a microsatellite unstable state or phenotype;

[0224] b) generating a classifier output for each microsatellite locus, said output classifying the microsatellite locus input as having a stable (MSS) or unstable (MSI) state or phenotype; and

[0225] c) for each microsatellite locus, evaluating the accuracy of the classifier for distinguishing between a stable (MSS) state or phenotype and an unstable (MSI) state or phenotype by comparing the output of the classifier to the known actual phenotype of the microsatellite locus or to a reference number of indels at the microsatellite locus;

[0226] If the classifier demonstrates accuracy in counting indels (also known as deletion insertions) at a resolution of 1 bp (base pair) (or instability threshold) for each microsatellite locus, the classifier is considered an accurate classifier for determining the MSS or MSI status or phenotype of a biological sample, particularly a tumor. The instability threshold is calculated as the mean + 1SD (standard deviation) of the indel counts in the specific MSS training set of microsatellite loci.

[0227] As used herein, the term "classifier" refers to an algorithm that implements classification, i.e., a "classifier" can determine a likelihood score or probability that an object is classified into a group of objects (e.g., a group of MSS profiles) rather than one or more other groups of objects (e.g., a group of MSI profiles), and a "classifier" maps the input object to a category (e.g., a benign or malignant profile). The term "classifier" can refer to one or more classifiers. For example, multiple classifiers can be trained, and they can process data in parallel and / or in a pipelined manner. For example, the output of one type of classifier (e.g., from an intermediate layer of a neural network) can be fed as input to another type of classifier.

[0228] The example of the classifier that can use in the context of the present invention comprises for example but not limited to the neural network and the supervised machine learning classifier of various frameworks (for example artificial, depth, convolution, fully connected), such as support vector machine (SVM) classifier, random forest classifier, decision tree classifier, K-nearest neighbor classifier (KNN), logistic regression classifier, nearest neighbor classifier, Gaussian mixture model (GMM), nearest centroid classifier and linear regression classifier.This is not an exhaustive list, and those skilled in the art will be able to identify the similar algorithms that can be used equivalently, although they are not specifically mentioned at this. The details and the functional rules of the algorithm have been extensively described in the literature. An important contribution is the input data set (that is, the set of the nucleotide sequence that comprises microsatellite or the pre-treated information obtained from the set of such nucleotide sequence) that is provided for the classifier. Based on this input data, any suitable supervised or unsupervised machine learning technique can be used to create suitable model. Therefore, the selection of appropriate algorithm is secondary, and can be performed with many different ways and various combinations that are obvious to those skilled in the art.

[0229] Preferably, the classifier is selected from random forest (RF) classifier, support vector machine (SVM) classifier, decision tree classifier, K- nearest neighbor classifier (KNN), logistic regression classifier, nearest neighbor classifier, Gaussian mixture model (GMM) classifier, nearest centroid classifier, linear regression classifier and neural network, such as artificial, deep, convolutional or fully connected neural network. More preferably, the classifier is selected from support vector machine (SVM) classifier, random forest (RF) classifier and neural network, particularly convolutional neural network (CNN). Even more preferably, the classifier is a random forest classifier.

[0230] The classifier uses some training data to understand how a given input object belongs to one category / classification or another. The classifier can be provided with a training set of biological samples from subjects (such as pMMR / MSS and / or dMMR / MSI subjects, in particular cancer subjects (typically cancer dMMR / MSI subjects)), the biological samples comprising or consisting of DNA sequences, in particular DNA sequences comprising or consisting of microsatellite loci, in particular (benign and / or malignant) tumor DNA. In a preferred aspect, the microsatellite loci are from human DNA and are selected from the group comprising: CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07, as first described above.

[0231] In a preferred embodiment, the unstable (MSI) or stable (MSS) microsatellite loci used to prepare the training set are obtained from the DNA of cancerous tumor cells, the cancer being any cancer as identified above, such as colorectal cancer or endometrial cancer. Alternatively, the classifier can be provided with pre-processed information obtained from such a DNA sequence training set.

[0232] The accuracy of the classifier can be assessed using any method known to those skilled in the art. In particular, the accuracy of the classifier can be assessed by calculating the inherent instability threshold (also referred to as "peak threshold" in the MSID software) for each microsatellite locus. The instability threshold is a measure of the experimental cutoff value for assessing each locus and is defined as the mean + 1 SD (standard deviation) of the indel counts in the MSS training set for the microsatellite locus.

[0233] A classifier is considered accurate if it provides an instability value equal to or greater than (≥) 1 bp relative to the instability threshold at each locus of a known unstable specimen ( / biological sample). This result depicts perfect classification, i.e., indicates that the two classes, pMMR / MSS and dMMR / MSI, are fully distinguishable. An instability value equal to or greater than (≥) 1 bp at the instability threshold is considered to allow good separation between pMMR / MSS and dMMR / MSI phenotypes ( / states / profiles / conditions).

[0234] A classifier was considered accurate if it provided an instability value at each locus of the stable samples that was (strictly) below (<) 1 bp relative to the instability threshold.

[0235] If the classifier exhibits insufficient accuracy, the training method, i.e. steps a) to c), may be repeated, in particular by some modifications, such as by increasing the number of pMMR or MSS and / or dMMR or MSI microsatellite profiles in the training set of DNA sequences, by using a different training set of DNA sequences, for example by modifying some parameters of the classifier, until a satisfactory accuracy is achieved, as defined above.

[0236] The pMMR or MSS and / or dMMR or MSI microsatellite profile can be obtained from a biological sample / specimen (such as a sample of a cell, a tumor, a subject, etc., in particular a cancerous sample). Preferably, the microsatellite profile of the DNA sequence of interest or a subsequence thereof is determined using the method for determining a microsatellite profile disclosed above.

[0237] Therefore, a possibility to increase the accuracy of the classifier is to increase the number of subsequence sets of the DNA of interest used in order to reach an instability value ≥ 1 bp relative to the instability threshold.

[0238] The training set of microsatellite loci must include a set of at least 20 non-repeated pMMR / MSS microsatellite profiles (each microsatellite locus has a known pMMR / MSS profile) and a set of at least 20 non-repeated dMMR / MSI microsatellite profiles (each microsatellite locus has a known dMMR / MSI profile). The set is a set that has been previously validated by methodologically independent methods known to the skilled person, such as immunohistochemistry and / or PCR-based fragment size determination.

[0239] The evaluation of the accuracy of the classifier performed in step c) of the method described above is preferably based on classifying the microsatellite loci as MSS or MSI using a test set comprising microsatellite loci obtained from pMMR / MSS cells and microsatellite loci obtained from dMMR / MSI cells (preferably cancer cells), wherein the test set is different from the training set, the stable (MSS) or unstable (MSI) status of each microsatellite locus is known, and the microsatellite status of each locus in the test set is obtained and processed using the same method as that used to obtain and process the microsatellite status of each locus in the training set.

[0240] In a preferred embodiment, the methods of the invention utilize a classifier trained to determine the microsatellite stable (MSS) or unstable (MSI) profile, status, condition, or phenotype of a tumor as described herein.

[0241] In a specific embodiment, the present invention relates to an in vitro method for determining the microsatellite profile, status, condition or phenotype of a tumor, wherein said method comprises the following steps:

[0242] (i) providing a DNA sequence of interest from a tumor of a subject, or preprocessed information obtained from said DNA sequence, as input to a classifier trained to distinguish between MSS and MSI profiles, states, conditions or phenotypes, said DNA sequence of interest being a DNA sequence comprising microsatellites, in particular a DNA sequence comprising at least two (preferably at least four, five, six, seven or eight) microsatellite loci selected from the group / group comprising CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07 or consisting thereof, and

[0243] (ii) using the classifier to identify the microsatellite profile, status, condition or phenotype of the DNA sequence of interest of the tumor of the subject, as an output of the classifier, if the MSID score is equal to or lower than ≤0.125, then it is identified as an MSS profile, or if the MSID score is equal to or higher than ≥0.5, then it is identified as an MSI profile.

[0244] The mismatch repair system (MMR), encoded by the MLH1, MSH2, PMS2, and MSH6 genes, is involved in repairing these errors [Evrard et al. In a preferred embodiment, the method of the present invention further comprises the step of evaluating the expression or loss of expression of at least one of the MLH1, MSH2, MSH6, and PMS2 proteins of the mismatch repair system (MMR) system.

[0245] In another preferred embodiment, the method involves next generation sequencing (NGS) and further optionally comprises assessing the expression or loss of expression of at least one of the MLH1, MSH2, MSH6 and PMS2 proteins of the mismatch repair (MMR) system.

[0246] The method of determining whether a subject's cancer exhibits microsatellite instability according to the present invention can be performed once or multiple times during the subject's lifetime. Thus, the emergence and evolution of microsatellite instability can be monitored.

[0247] In a specific embodiment, the subject suffering from cancer is a subject who has received / exposed to an anti-cancer treatment such as resection surgery, chemotherapy, radiation therapy, or immunotherapy.

[0248] In one specific aspect, DNA from a subject or from a subject's tumor is provided to determine whether the subject's cancer exhibits microsatellite instability, to detect a propensity to develop cancer, to assess the prognosis of cancer, to monitor cancer progression or regression; to predict, assess or monitor response to cancer treatment, to select an appropriate cancer treatment, to select patients who are able to respond to cancer treatment, and to select patients for enrollment in clinical trials for the treatment of cancer.

[0249] The methods for determining whether a subject has cancer according to the present invention can also be performed after the first line of treatment, for example, six months, one year, two years, three years, four years, five years, or ten years after the first line of treatment, to monitor cancer progression or regression. DNA from the subject can be provided one or more times during the second line of treatment.

[0250] The efficacy of the first and / or second line of treatment can be assessed by monitoring the response to anticancer therapy, particularly to a therapeutic compound. This method can be performed one or more times during the first, second and / or subsequent treatments.

[0251] The inventors also describe herein a computing system comprising:

[0252] - a memory storing at least one instruction of a classifier trained according to a computer-implemented method as described herein, in particular a method of training a classifier for accurately distinguishing between pMMR / MSS microsatellite profiles and dMMR / MSI microsatellite profiles, and

[0253] - a processor that accesses the memory to read the instructions and executes the methods of the present invention as described herein, in particular methods for: assessing microsatellite instability in precancerous and / or cancerous cells; for detecting a propensity to develop cancer; for assessing the prognosis of cancer; for monitoring cancer progression or regression; for predicting, assessing or monitoring response to cancer treatment; for selecting an appropriate cancer treatment for a subject in need thereof; for selecting a patient who is likely to respond to a cancer treatment; or for selecting a patient for inclusion in a clinical trial for the treatment of cancer.

[0254] Also described herein is a kit for analyzing genomic DNA, preferably microsatellite loci of human genomic DNA, comprising a tool for genotyping at least two, such as three, preferably at least four, five, six, seven or eight of a set of microsatellite markers. Thus, the kit can comprise oligonucleotide primers for co-amplifying a set of human DNA microsatellite loci, and / or oligonucleotide probes for detecting specific sequences in the set of microsatellite loci, particularly insertions and deletions, and optionally a thermostable polymerase and / or control DNA isolated from normal non-cancerous biological material and / or lacking mismatch repair genes.

[0255] The kit typically comprises several, such as two to eight, for example three, preferably at least four, five, six, seven or eight primer pairs suitable for amplifying the following and / or several, such as two to eight, such as three, preferably at least four, five, six, seven or eight oligonucleotide probes for detecting the following: a set of markers as described herein, in particular a set comprising at least two, such as three, preferably at least four, five, six, seven or eight microsatellite markers identified herein as CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and CABIO-E07.

[0256] The inventors also describe herein the use of such kits for analyzing microsatellite stability or instability.

[0257] In a specific aspect, at least one, preferably at least two, oligonucleotide primers or probes for several (at least two, such as three, four, five, six or seven) or each of the several microsatellite loci of the set / group described for the first time herein are fluorescently labeled.

[0258] The present invention also relates to the use of a kit according to the invention for amplifying all or part of the microsatellite regions described herein, preferably CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and / or CABIO-E07, in particular by PCR multiplex amplification, preferably in the context of NGS, for example for determining the microsatellite (MSS or MSI) status.

[0259] The following examples are provided to better illustrate certain aspects and advantages of the present invention. They should be considered as merely illustrative and not restrictive. The present application is limited only by the claims.

[0260] Example

[0261] Materials and methods

[0262] Research samples

[0263] In this study, colorectal cancer (CRC) and endometrial cancer (EC) samples were collected from patients undergoing microsatellite evaluation in routine practice at Poitiers University Hospital.

[0264] The KAPA Express extraction kit (Kapa Biosystems, Wilmington, MA) or DNA was extracted using the FFPE LEV DNA purification kit (Promega Corporation, Madison, WI). All samples used in this study underwent routine analysis. For pMMR / dMMR phenotype analysis, IHC using four MMR proteins (MSH2, MSH6, PMS2, and MLH1) was performed with the following antibodies: anti-MSH2 (clone G219-1129, Reagent test kit for revealing), anti-MSH6 (clone 44BD San Jose, Reagent test kit for disclosure), anti-PMS2 (clone EPR 3947, Ready-to-use; kit for revealing and amplifying) and anti-MLH1 (clone M1, Tucson, Tucson, kit For the analysis of MSS / MSI status, the MSI-PCR method was used (Promega MSI Analysis System version 1.2, Promega Corporation, Madison, WI) as previously described [Guyot D'Asnières De Salins A et al].

[0265] MSI-NGS analysis was performed using an in-house designed 8-marker panel (“CaBio”) or a 5-marker panel (for CRC only). Figure 1 ). These assays were performed as a blinded study.

[0266] Microsatellite marker screening

[0267] Previous reports have shown that mononucleotide repeats are the most sensitive markers for MSI detection in humans, with a mutation rate of 96% in MMR-deficient tumors [Buhard O et al., Bacher JW et al. (2005), Bacher JW et al. (2004), Dietmaier W et al.]. Based on this evidence, the inventors conducted a study of a panel of mononucleotide markers compatible with NGS sequencing that could be promisingly applied to determine MSI status in colorectal and endometrial tumors. Retrieval from the MSDB v3.0.2 database (Microsatellite Database, CSIR-Center for Cell and Molecular Biology, Uppal Road, Hyderabad-500007, Telangana, India) allowed the identification of 706,849 simple sequence repeats (SSRs) [Avvaru AK et al.]. After the screening process, 13 SSR motifs were retained, enhancing the operability of NGS sequencing. Using computational processing, the inventors specifically identified an octet (8 marker) panel, designated the "CaBio Octet" (also identified herein as "CaBio") (Table 1), that exhibited very favorable sensitivity and specificity levels, compatible with conventional diagnostics. The inventors successfully reproduced these tests using only 7, 6, 5, and the last 4 markers, respectively, arbitrarily selected from the original set of 8 markers.

[0268] Table 1 : Detailed information on 8 microsatellite markers from the CaBio panel

[0269]

[0270] Molecular screening of microsatellite loci

[0271] Using Fluidigm Access Array TM System (Fluidigm Corporation, South San Francisco, CA, USA) carries out the amplification of mononucleotide repeat markers.According to the explanation of the manufacturer, the PCR amplification (target enrichment) of the specific microsatellite marker based on Access array is carried out on LP 48.48IFC, which allows 48 samples of parallel amplification.During the target enrichment step, the sample is barcoded and labeled with adapter sequence, which allows multiplex sequencing.Next generation sequencing is carried out using Illumina NextSeq 550 system (Illumina, San Diego, CA, USA).This group includes 29 genes (71 exons) with KRAS, NRAS, BRAF, POLE and PIK3CA genes.New 8 markers are included in this group.The MSID algorithm is used for analyzing data sets and for the classification of microsatellite instability.

[0272] MSID algorithm and MSI classification

[0273] The sequencing reads obtained from each sample were aligned on the human genome GRCh37 (hg19) using BWA software (version 0.7.17) [Li H, Durbin R (2009)]. The alignment process was performed using the MEM algorithm of BWA using the default options. The generated BAM files were processed using the mpileup algorithm using SAMtools software version 1.9 [Li H, Handsaker B. et al. R (2009)]. The final BAM and pileup files obtained for each sample were used as input for the software MSID (MSI detection tool) developed internally by the inventors. MSID includes 3 algorithms: MSID insertion and deletion counter, MSID baseline, and MSID detector. The generated pileup files were processed using the MSID insertion and deletion counter to count the number of insertions and deletions (insertions and deletions). The MSID insertion and deletion counter generates a VCF file for each sample containing all insertions and deletions. As a first step, 20 MSS / pMMR control samples were used to generate a baseline. This step is performed using an MSID baseline that involves an MSID indel counter for indel counts at each locus in the MSS / pMMR control samples. Subsequently, the mean and standard deviation (SD) of the indel counts for each locus are calculated. Finally, a text file is generated containing the following information for each locus: name, genomic coordinates, mean and SD of indel counts, instability threshold (mean + 1SD), and the number of MSS / pMMR control samples used in the calculation. After completing the baseline, compare it to the To be testedThe sample is compared. This step is performed by obtaining the insertion / deletion number and the MSID detector using the MSID insertion / deletion counter, for comparing the locus in each sample with the baseline. For each locus, if the insertion / deletion count is higher than the baseline value, the corresponding locus is considered as "unstable", otherwise the corresponding locus is considered as "stable". Subsequently, the MSID detector collects the quantity of unstable and stable loci, and generates the MSID scoring (=unstable locus / total locus). The MSID scoring is between 0 and 1. When using a 5-marker conventional panel, if the MSID scoring ≤ 0.2, the sample state is considered as "stable", and if the MSID scoring ≥ 0.6, the sample state is considered as "unstable", and if the MSID scoring = 0.40, the sample state is considered as "uncertain" (must be reassessed in the latter case). For analysis using octaplex CaBio-MSID, the sample status was considered "stable" if the MSID score was ≤0.125, "unstable" if the MSID score was ≥0.5, and "uncertain" if the MSID score was between 0.125 and 0.5, in this example equal to (=) 0.25 and 0.375 (and in the latter case had to be reassessed).

[0274] result

[0275] Study population and analyses performed

[0276] In this study, the inventors studied a total of 303 cases of colorectal cancer (CRC) and 88 cases of endometrial cancer (EC) selected from the Poitiers University Hospital ( Figure 1 ).

[0277] Of the 303 CRC samples, 241 were pMMR / MSS and 62 were dMMR / MSI tumors, and for EC samples, 60 were pMMR / MSS and 28 were dMMR / MSI tumors ( Figure 1). CRC samples were obtained after biopsy (n=144) or colectomy (n=159), and EC samples were obtained after biopsy (n=7) or hysterectomy (n=81), with a median tumor cell content of 65% for CRC and 60% for EC (18 data missing). At the time of analysis, the median age of CRC and EC patients was 74 and 72 years, respectively. Details on IHC and molecular alterations are reported in Table 2. Analysis of the CRC cohort showed that 52.1% of the samples carried only one molecular alteration, mainly on the KRAS gene (30.4%), without POLE mutations. In contrast, only 19.8% of CRC patients had multiple molecular alterations, including 43.6% on KRAS, 15.5% on BRAF, and 14.2% on PIK3CA. Analysis of the EC cohort revealed a low percentage of patients harboring single molecular alterations (27.3%), whereas 27.3% of patients were found to have multiple molecular alterations, including 21.6% with PIK3CA mutations, 14.8% with CTNNB1 mutations, and 5.7% with POLE mutations.

[0278] Table 2 : Tumor Characterization: IHC and Most Common Molecular Alterations

[0279]

[0280]

[0281] CRC: colorectal cancer; EC: endometrial cancer; MMR: mismatch repair; pMMR: intact mismatch repair; IHC: immunohistochemistry

[0282] Evaluation of a conventional 5-marker panel for MSI-NGS classification in colorectal cancer

[0283] MSI-NGS classification of a CRC cohort (303 tumors previously tested by reference methods IHC and PCR) using the conventional 5-marker panel confirmed microsatellite status in 237 / 241 (98.3%) pMMR / MSS and 57 / 62 (91.9%) dMMR / MSI samples ( Figure 2 In pMMR / MSS samples, 4 discordant cases (1.7%) were observed with MSID scores > 0.2 (MSID score = 0.40, 2 unstable markers), whereas 5 discordant cases (8.1%) were observed in dMMR / MSI samples with MSID scores < 0.6 (1 case with MSID score = 0, no unstable marker; 2 cases with MSID score = 0.2, 1 unstable marker; and 2 cases with MSID score = 0.4, 2 unstable markers).

[0284] This 5-marker MSI-NGS classification was not applied to EC samples because preliminary testing showed low sensitivity for detecting dMMR / MSI status, with 6 discordant cases (30%) among 20 dMMR / MSI samples tested (data not shown), clearly highlighting the lack of sensitivity of this 5-marker conventional panel in detecting MSI in EC samples.

[0285] Validation of the 8-marker CaBio panel for MSI-NGS classification in formalin-fixed paraffin-embedded (FFPE) tumor samples

[0286] To overcome the poor performance of these five conventional microsatellite markers in non-colorectal cancers, especially in EC classification, the inventors identified and characterized an 8-marker panel (CaBio Octet) for determining microsatellite instability by MSI-NGS in colorectal cancer as well as non-colorectal cancers (see Materials and Methods).

[0287] To validate the reproducibility of the 8-marker CaBio panel, three CRC samples were tested in triplicate, showing reproducible results for pMMR / MSS and dMMR / MSI tumors ( Figure 3 ).

[0288] The performance of the CaBio octaplex panel was evaluated in 56 tumors (30 CRC and 26 EC). Figure 4 Provides support for CRC( Figure 4 A) and EC( Figure 4 B) Summary of MSI-NGS results obtained for each microsatellite marker in . Scoring results obtained by MSI-NGS using dMMR / MSI CRC and EC samples consistently showed at least 4 of the 8 instability markers, indicating an acceptable level of reliability for the CaBio panel and the MSID algorithm in the systematic determination of MSI status in both CRC and EC.

[0289] Validation and performance evaluation of the 8-marker CaBio panel for MSI-NGS classification of colorectal cancer

[0290] The octameric CaBio-MSID panel was implemented instead of the conventional 5-marker panel to re-evaluate a previous CRC cohort (303 tumors). MSID analysis confirmed microsatellite status in 237 / 241 (98.3%) pMMR / MSS and 61 / 62 (98.4%) dMMR / MSI samples ( Figure 5). Discordance between MSI-NGS and the reference method was observed in 4 pMMR / MSS CRC samples with MSID scores > 0.125 (=1.7%) (M6627 and M7047 samples had MSID scores = 0.25, 2 unstable markers; M6623 and M6759 samples had MSID scores = 0.375, 3 unstable markers), while for dMMR / MSI samples with MSID scores < 0.5, only 1 discordant case was observed (=1.6%) (M7093 sample had MSID score = 0.25, 2 unstable markers). Therefore, all 5 discordant samples were classified as "uncertain" microsatellite status. IHC slides of M6627, M7047, M6623 and M6759 were reread by different pathologists, confirming normal staining of MSH2, MSH6, MLH1 and PMS2 proteins. Rereading of slides M7093 confirmed isolated loss of PMS2 and confirmed the unstable microsatellite status of the tumor.

[0291] A different method that ensures higher purity levels is then used ( All five discrepant samples were re-extracted using the 16FFPE Plus LEV DNA Purification Kit, Promega, see Materials and Methods. Re-evaluation of M6627, M7047, M6623, and M6759 after kit re-extraction revealed MSID scores between 0 and 0.125 (Table 3), confirming the stable microsatellite status, and also confirmed the MSI status of M7093 (MSID score = 0.5).

[0292] Table 3 : Details of the discordant CRC cases

[0293]

[0294] IHC: immunohistochemistry; MSI: microsatellite instability; PCR: polymerase chain reaction; NGS: next-generation sequencing; pMMR: pristine mismatch repair; MSS: microsatellite stability; MSID: MSI detection tool

[0295] * No loss of expression of 4 MMR proteins (MSH2, MSH6, MLH1, and PMS2) The overall performance of the CaBio panel and the MSID algorithm on CRC was analyzed (Tables 4 and Figure 6 ).

[0296] Table 4 Performance characteristics of the CaBio panel for MSS / MSI diagnosis of CRC tumors

[0297]

[0298] MSS: microsatellite stability; MSID: MSI detection tool

[0299] When defining the upper limit of the MSS classification, various thresholds have been considered to calculate the performance, and for informational purposes, the values from 3 different thresholds (1, 2 or 3 unstable markers allowing MSS classification) have been reported in Table 4. In the current study, the MSID analysis has been performed by fixing the maximum threshold to 1 unstable marker for MSS classification (MSID score ≤ 0.125), showing sensitivity = 100% and specificity = 98.3%. In addition, taking into account the uncertainty of some pMMR / MSS samples (4 samples in this study) that showed 2 or 3 unstable markers, the inventors decided to set the minimum threshold to 4 unstable markers for MSI classification (MSID score ≥ 0.5). Therefore, when the CaBio-MSID results were compared with the pMMR / dMMR and MSS / MSI status determined from the validated technology, the overall performance obtained for CRC indicated a sensitivity of 98.4%, a specificity of 98.4%, a positive predictive value of 93.8% and a negative predictive value of 99.6%.

[0300] Validation and performance evaluation of the 8-marker CaBio panel for MSI-NGS classification in endometrial cancer

[0301] To determine whether the octameric CaBio-MSID can be applied to determine microsatellite status in EC tumors, the inventors performed MSI-NGS testing on a retrospective cohort of 88 EC samples. MSID analysis confirmed microsatellite status in 60 / 60 (100%) pMMR / MSS and 25 / 28 (89.3%) dMMR / MSI samples ( Figure 7 ).

[0302] Table 5 and Figure 8 The overall performance of the CaBio panel and the MSID algorithm on EC tumors is shown. In the same manner as for the MSS / MSI classification of CRC tumors, three different thresholds have been reported in Table 6. For CRC tumors, the same maximum threshold for MSS classification and minimum threshold for MSI classification have been utilized. When the CaBio-MSID results were compared with pMMR / dMMR and MSS / MSI status determined from validated techniques, the overall performance achieved for EC indicated a sensitivity of 89.3%, a specificity of 100%, a positive predictive value of 100%, and a negative predictive value of 95.2%.

[0303] Table 5: Performance characteristics of the CaBio panel for MSS / MSI diagnosis of EC tumors.

[0304]

[0305] MSS: microsatellite stability; MSID: MSI detection tool

[0306] discuss

[0307] Conventional 5-marker panels for MSI status determination have long been established and are mostly used in combination with MSI-PCR methods. This 5-marker panel has only been validated for CRC. An IVD (in vitro diagnostic) version of the 5-marker panel commercially manufactured by Promega (OncoMate TM The MSIDx analysis system) is widely used and shows satisfactory performance (97.3% sensitivity and 97.2% specificity) [PROMEGA MSI Analysis System, Version 1.2. Instructions for Use of Product]. However, this 5-marker panel has not yet been validated for non-CRC tumors. According to current projections, by 2030, there will be more than 515,000 new cases of EC worldwide, 23.7% higher than in 2020, and approximately 20% to 30% of them will require assessment of MSI status (https: / / gco.iarc.fr). Since 2013, the Cancer Genome Atlas (TCGA) classification has defined four tissue prognostic groups focusing on MSI status [Cancer Genome Atlas Research Network, Kandoth C et al.; Stelloo E et al.]. The "ultra-mutated" group (7%) involves patients with a high tumor mutation burden (TMB) and POLE mutations that are associated with a good prognosis. The "hyper-mutated" group (28%) consisted of patients with high TMB and MSI status caused by MLH1 promoter methylation. The third group was called "low copy number" (39%) and involved patients with low TMB who primarily had CTNNB1 mutations. The last group included patients with "serous" tumors (25%), with high somatic copy number alterations, mainly TP53 mutations [Stelloo E et al.]. In addition, due to the effectiveness of immunotherapy for this type of tumor, the determination of MSI status is a major new issue in EC [Berton D et al.; Marabelle A. et al.; Oaknin A. et al.].

[0308] In the context of the present invention, the inventors used their octet CaBio-MSID to classify 303 CRCs and 88 ECs compared with reference tests (IHC and MSI-PCR quintet). Evaluation of the diagnostic performance of the octet CaBio-MSID for classification of MSI status was satisfactory in CRC tumors, with a sensitivity of 98.4% and a specificity of 98.4%, and was also satisfactory in EC tumors, with a sensitivity of 89.3% and a specificity of 100%.

[0309] With OncoMate TM Compared with the MSIDx analysis system, the octameric CaBio-MSID showed higher sensitivity (98.4% vs 97.3%) and specificity (98.4% vs 97.2%) for MSI-NGS classification of the CRC cohort. In the literature, there are few studies implementing MSI-NGS algorithms, but most require matched normal samples, which is a limitation when testing many patients to ensure timely results [Ratovomanana T. et al.; Kautto EA et al.; Zhao L. et al.]. In published studies, the sensitivity level for detecting MSI in CRC ranges from 76.1% to 100%, while the specificity level fluctuates between 72.5% and 100% [Bacher JW et al. (2005); Bacher JW et al. (2004); Dietmaier W. et al.; Avvaru AK et al.]. The inventors' results are ranked favorably in the upper part of the range for both sensitivity and specificity.

[0310] Unlike the ESMO guidelines for CRC, there are currently no formally evaluated microsatellite markers for MSI testing in EC. According to the inventors' results, MSI-NGS classification using the octameric CaBio-MSID showed better performance than commonly used methods such as the Idylla MSI test (automated PCR), with a sensitivity of 89.3% vs 72.7% (both with 100% specificity) [Ukkola I. et al.]. Another study based on a probe capture method reported lower sensitivity than the octameric CaBio-MSID, with a sensitivity of 75% (and a specificity of 100%) [Waalkes A. et al.].

[0311] In fact, the octameric CaBio-MSID contains eight markers and is therefore easily compatible with NGS, allowing for simultaneous assessment of microsatellite status and detection of variants. The inventors successfully reproduced these tests using only 7, 6, 5, and the last 4 markers, respectively, arbitrarily selected from the original set of 8 markers.

[0312] References

[0313] -Amos W(September 2010)."Mutation biases and mutation rate variationaround very short human microsatellites revealed by human-chimpanzee-orangutan genomic sequence alignments".Journal of Molecular Evolution.71(3):192–201).

[0314] -AndréT,Shiu K.K,Kim T.W,Jensen B.V,Jensen L.H,Punt C,Smith D,…KEYNOTE-177Investigators(2020).Pembrolizumab in Microsatellite-Instability-High Advanced Colorectal Cancer.New England J Med 383(23),2207–2218.

[0315] -Avvaru A.K,Sharma D,Verma A,Mishra R.K,Sowpati D.T(2020)MSDB:AComprehensive,Annotated Database of Microsatellites.Nucleic Acids Res.,48,D155–D159.

[0316] -Bacher J.W,Abdel Megid W.M,Kent-First M.G,Halberg R.B(2005)Use ofMononucleotide Repeat Markers for Detection of Microsatellite Instability inMouse Tumors.Mol.Carcinog.,44,285–292.

[0317] -Bacher J.W,Flanagan L.A,Smalley R.L,Nassif N.A,Burgart L.J,HalbergR.B,Megid W.M.A,Thibodeau,S.N(2004)Development of a Fluorescent MultiplexAssay for Detection of MSI-High Tumors.Dis.Markers,20,237–250.

[0318] -Berton D,Banerjee S.N,Curigliano G,Cresta S,Arkenau H.-T,AbdeddaimC,Kristeleit R.S,…AndréT(2021).Antitumor Activity of Dostarlimab in Patientswith Mismatch Repair-Deficient / Microsatellite Instability–High Tumors:ACombined Analysis of Two Cohorts in the GARNET Study.J.Clin.Oncol.39,2564–2564.

[0319] -Bonneville R,Krook M.A,Kautto E.A,Miya J,Wing M.R,Chen H.-Z,ReeserJ.W,Yu L,Roychowdhury S(2017).Landscape of Microsatellite Instability Across39Cancer Types.JCO Precis.Oncol.PO.17.00073

[0320] -Buhard O,Suraweera N,Lectard A,Duval A,Hamelin R(2004)Quasimonomorphic Mononucleotide Repeats for High-Level MicrosatelliteInstability Analysis.Dis.Markers,20,251–257.

[0321] -Cancer Genome Atlas Research Network,Kandoth C,Schultz N,CherniackA.D,Akbani R,Liu Y,Shen H,…Levine DA(2013)Integrated GenomicCharacterization of Endometrial Carcinoma.Nature 497,67–73.

[0322] -Church and Mitra;Dec 1999.“In situ localized amplification andcontact replication of many individual DNA molecules.”Nucleic Acids Res.27(24)

[0323] -Cunningham J.M,Christensen E.R,Tester D.J,Roche P.C,Burgart L.J,Thibodeau,S.N(1998)Hypermethylation of the hMLH1 Promoter in Colon Cancerwith Microsatellite Instability.Cancer Res 58(15):3455–3460.

[0324] -Dietmaier W,Wallinger S,Bocker T,Kullmann F,Fishel R,Rüschoff,J(1997)Diagnostic Microsatellite Instability:Definition and Correlation withMismatch Repair Protein Expression.Cancer Res.,57,4749–4756.

[0325] -Evrard C,Alexandre J(2021)Predictive and Prognostic Value ofMicrosatellite Instability in Gynecologic Cancer(Endometrial and Ovarian).Cancers 13,2434.

[0326] -Evrard C,Tachon G,Randrian V,Karayan-Tapon L,Tougeron D(2019)Microsatellite Instability:Diagnosis,Heterogeneity,Discordance,and ClinicalImpact in Colorectal Cancer.Cancers 11(10),1567.

[0327] -Guyot D’Asnières De Salins A,Tachon G,Cohen R,Karayan-Tapon L,JuncaA,Frouin E,Godet J,…Tougeron D(2021)Discordance between Immunochemistry ofMismatch Repair Proteins and Molecular Testing of Microsatellite Instabilityin Colorectal Cancer.ESMO Open,6,100120.

[0328] -Forster P,Hohoff C,Dunkelmann B,Schürenkamp M,Pfeiffer H,Neuhuber F,Brinkmann B(March 2015)."Elevated germline mutation rate in teenage fathers".Proceedings.Biological Sciences.282(1803):20142898.

[0329] -Griffiths,A.J.F.,Miller,J.F.,Suzuki,D.T.,Lewontin,R.C.&Gelbart,W.M.(1996).Introduction to Genetic Analysis,5th Edition.W.H.Freeman,New York.

[0330] -Hempelmann J.A,Lockwood C.M,Konnick E.Q,Schweizer M.T,AntonarakisE.S,Lotan T.L,Montgomery B,…Pritchard C.C(2018).Microsatellite Instabilityin Prostate Cancer by PCR or Next-Generation Sequencing.J.Immunother.Cancer,6,29.

[0331] -Herbreteau G,Airaud F, E,Vallée A,Bézieau S,Théoleyre S,Blons H,…Denis M.G(2021).MEM:An Algorithm for the Reliable Detection ofMicrosatellite Instability(MSI)on a Small NGS Panel in ColorectalCancer.Cancers,13,4203.

[0332] -Jarne P,Lagoda PJ(October 1996)."Microsatellites,from molecules topopulations and back".Trends in Ecology&Evolution.11(10):424–9).

[0333] -Kautto E.A,Bonneville R,Miya J,Yu L,Krook M.A,Reeser J.W,Roychowdhury S(2016)Performance Evaluation for Rapid Detection of Pan-CancerMicrosatellite Instability with MANTIS.Oncotarget,8,7452–7463.

[0334] -Klintschar M,Dauber EM,Ricci U,Cerri N,Immel UD,Kleiber M,Mayr WR(October 2004)."Haplotype studies support slippage as the mechanism ofgermline mutations in short tandem repeats".Electrophoresis.25(20):3344–8.

[0335] -Kobayashi et al.,Mol.Cell.Probes,9:175-182,1995

[0336] -Levine DA.Integrated genomic characterization of endometrialcarcinoma.Nature[Internet].2013[cited 2020 Oct 6];497:67–73.Available from:https: / / www.ncbi.nlm.nih.gov / pmc / articles / PMC3704730 /

[0337] -Luchini C,Bibeau F,Ligtenberg M.J.L,Singh N,Nottegar A,Bosse T,Miller R,…Scarpa A(2019).ESMO Recommendations on Microsatellite InstabilityTesting for Immunotherapy in Cancer,and Its Relationship with PD-1 / PD-L1Expression and Tumour Mutational Burden:A Systematic Review-BasedApproach.Ann.Oncol.30,1232–1243.

[0338] -Li H,Durbin R(2009)Fast and Accurate Short Read Alignment withBurrows–Wheeler Transform.Bioinformatics,25,1754–1760.

[0339] -Li H,Handsaker B,Wysoker A,Fennell T,Ruan J,Homer N,Marth G,…DurbinR(2009).The Sequence Alignment / Map Format and SAMtools.Bioinformatics,25,2078–2079.

[0340] -Long D.R,Waalkes A,Panicker V.P,Hause R.J,Salipante S.J(2020).Identifying Optimal Loci for the Molecular Diagnosis of MicrosatelliteInstability.Clin.Chem.66,1310–1318.

[0341] -Marabelle A,Le D.T,Ascierto P.A,Di Giacomo A.M,De Jesus-Acosta A,Delord J.P,Geva R,…Diaz L.A Jr(2020).Efficacy of Pembrolizumab in PatientsWith Noncolorectal High Microsatellite Instability / Mismatch Repair-DeficientCancer:Results From the Phase II KEYNOTE-158 Study.J.Clin.Oncol.38,1–10.

[0342] -Middha S,Zhang L,Nafa K,Jayakumaran G,Wong D,Kim H.R,Sadowska J,…Hechtman J.F(2017).Reliable Pan-Cancer Microsatellite Instability Assessmentby Using Targeted Next-Generation Sequencing Data.JCO Precis.Oncol.1,PO.17.00084.

[0343] -Oaknin A,Gilbert L,Tinker A.V,Brown J,Mathews C,Press J,SabatierR,…Pothuri B(2022)Safety and antitumor activity of dostarlimab in patientswith advanced or recurrent DNA mismatch repair deficient / microsatelliteinstability-high(dMMR / MSI-H)or proficient / stable(MMRp / MSS)endometrial cancer:interim results from GARNET-a phase I,single-arm study.J Immunother Cancer.10(1):e003777.

[0344] -Ostrander EA,Jong PM,Rine J,Duyk G.Construction of small-insertgenomic DNA libraries highly enriched for microsatellite repeatsequences.Proc Natl Acad Sci U S A.1992 Apr 15;89(8):3419-23.doi:10.1073 / pnas.89.8.3419.PMID:1314388;PMCID:PMC48879.

[0345] -PROMEGA MSI Analysis System,Version 1.2.Instructions for Use ofProduct.

[0346] -Ratovomanana T,Cohen R,Svrcek M,Renaud F,Cervera P,Siret A,Letourneur Q,…Duval A(2021)Performance of Next-Generation Sequencing for theDetection of Microsatellite Instability in Colorectal Cancer With DeficientDNA Mismatch Repair.Gastroenterology,161,814-826.e7.

[0347] -Saiki et al.,Nature 324,163-166,1986

[0348] -Stelloo E,Bosse T,Nout R.A,MacKay H.J,Church D.N,Nijman H.W,LearyA,…Creutzberg,C.L(2015).Refining Prognosis and Identifying TargetablePathways for High-Risk Endometrial Cancer;a TransPORTEC Initiative.ModPathol,28(6):836-44.

[0349] -Stoneking et al.,Am.J.Hum.Genet.48:70-382,1991

[0350] -Shia J(2008).Immunohistochemistry versus Microsatellite InstabilityTesting For Screening Colorectal Cancer Patients at Risk For HereditaryNonpolyposis Colorectal Cancer Syndrome.J.Mol.Diagn.10,293–300.

[0351] -Tougeron D,Mouillet G,Trouilloud I,Lecomte T,Coriat R,Aparicio T,DesGuetz G,Lécaille C,Artru P,Sickersen G,Cauchin E,Sefrioui D,Boussaha T,FerruA,Matysiak-Budnik T,Silvain C,Karayan-Tapon L,Pagès JC,Vernerey D, Bonnetain F, Michel P, J,Zaanan A.Efficacy of Adjuvant Chemotherapy in Colon CancerWith Microsatellite Instability:A Large Multicenter AGEO Study.J Natl CancerInst.2016 Feb 1;108(7).doi:10.1093 / jnci / djv438.PMID:26839356.

[0352] -Trabucco SE,Gowen K,Maund SL,Sanford E,Fabrizio DA,Hall MJ,Yakirevich E,…Sun JX(2019).A Novel Next-Generation Sequencing Approach toDetecting Microsatellite Instability and Pan-Tumor Characterization of1000Microsatellite Instability–High Cases in 67,000 PatientSamples.J.Mol.Diagn.21,1053–1066.

[0353] -Tytgat,Olivier;Gansemans,Yannick;Weymaere,Jana;Rubben,Kaat;Deforce,Dieter;Van Nieuwerburgh,Filip(2020).“Nanopore Sequencing of a ForensicSTRMultiplex Reveals Loci Suitable for Single-Contributor STR Profiling”.Genes.11(4):381.

[0354] -Ukkola I,Nummela P,Pasanen A,Kero M, A, S,Bützow R, A(2021)Detection of Microsatellite Instability with Idylla MSIAssay in Colorectal and Endometrial Cancer.Virchows Arch.,479,471–479.

[0355] -Waalkes A,Smith N,Penewit K,Hempelmann J,Konnick E.Q,Hause R.J,Pritchard C.C,Salipante S.J(2018)Accurate Pan-Cancer Molecular Diagnosis ofMicrosatellite Instability by Single-Molecule Molecular Inversion ProbeCapture and High-Throughput Sequencing.Clin.Chem.,64,950–958.

[0356] -Wang Y,Shi C,Eisenberg R,Vnencak-Jones C.L(2017)Differences inMicrosatellite Instability Profiles between Endometrioid and ColorectalCancers:A Potential Cause for False-Negative Results?JMD 19(1),57–64.

[0357] -Weber JL,Wong C(August 1993).“Mutation of human short tandemrepeats”.Human Molecular Genetics.2(8):1123–8.).

[0358] -Zhao L,Shan G,Li L,Yu Y,Cheng G,Zheng X(2020)A Robust Method for theRapid Detection of Microsatellite Instability in ColorectalCancer.Oncol.Lett.,20,1982–1988.

Claims

1. A method of analyzing a set of at least four microsatellite loci in human DNA selected from a group comprising: -CABIO-P05, defined as a 21T duplication located at 14q23.1 and starting at position chr14:58359108, -CABIO-P07, defined as a 21T duplication located at 7q32 and starting at position chr7:131478596, -CABIO-E01, defined as a 22T repeat located at Xq22.3 starting at position chrX:106849221, -CABIO-E03, defined as a 23T repeat located at Xq21.2 starting at position chrX:85268269, -CABIO-E04, defined as a 25T duplication located at 14q32.3 and starting at position chr14:103574079, -CABIO-E05, defined as a 22T duplication located at 2p11.2 and starting at position chr2:86456417, - CABIO-E06, defined as a 23T duplication located at 4q23 and starting at position chr4:99216136, and - CABIO-E07, defined as a 21T repeat located at 20p13 and starting at position chr20:290564, with reference to the Homo sapiens reference genome assembly from the Genome Reference Consortium human Build 38 patch release 14 (GRCh38.p14).

2. The method of claim 1, wherein the method comprises the step of sequencing the set of at least four microsatellite loci.

3. The method of claim 1 or 2, wherein the method comprises the step of co-amplifying the set of at least four microsatellite loci with primers, wherein at least one of the primers has a nucleic acid sequence selected from the group of primer sequences identified by SEQ ID NOs: 9-24 or a sequence having at least 90% identity thereto.

4. A method for assessing the microsatellite stability (MSS) or microsatellite instability (MSI) status of a tumor, comprising the following steps: a) counting the number of indels in at least four microsatellite loci in a tumor DNA sample, wherein the microsatellite loci are selected from the group consisting of CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06, and CABIO-E07 of claim 1; b) for each of the at least four microsatellite loci, comparing the number of indels to a reference number of indels for the locus and determining the MSI status of the locus, if the indel count is equal to or higher (≥) the reference value for the locus, deeming the locus unstable, and if the indel count is lower (<) than the reference value for the locus, deeming the locus stable; c) calculating an MSID score, which ranges between 0 and 1 and consists of the total number of unstable loci relative to the total number of loci; and d) determining the MSI status of the tumor as unstable if the MSID score is equal to or higher (≥) 0.5; as stable if the MSID score is equal to or lower (≤) 0.125, and as uncertain if the MSID score is between 0.125 and 0.5; and optionally e) If the MSID score is between 0.125 and 0.5, repeating steps a)-d) until the status of the MSI can be determined.

5. The method of any one of claims 1 to 4, wherein the method is used to assess microsatellite instability in precancerous and / or cancerous cells; to detect a propensity to develop cancer; to evaluate the prognosis of cancer; to monitor cancer progression or regression; to predict, evaluate or monitor response to cancer treatment; to select an appropriate cancer treatment for a subject in need thereof; to select a patient who is able to respond to a cancer treatment; or to select a patient for enrollment in a clinical trial for a cancer treatment.

6. The method according to claim 4 or 5, wherein the method is a partially or fully computer-implemented method.

7. A computer-implemented method for training a classifier for determining the microsatellite stability (MSS) or microsatellite instability (MSI) status or phenotype of a tumor, wherein the method comprises: a) providing a training set of microsatellite loci or preprocessed information obtained from the training set as input to the classifier, each locus being obtained from a DNA sequence of interest, the training set comprising i) stable (MSS) microsatellite loci or subsequences thereof obtained from intact mismatch repair system (pMMR) or MSS cells known to have a microsatellite stable state or phenotype, and ii) unstable (MSI) microsatellite loci or subsequences thereof obtained from deficient mismatch repair system (dMMR) or MSI cells known to have a microsatellite unstable state or phenotype; b) generating an output of the classifier for each microsatellite locus, the output classifying the microsatellite locus input as having a stable (MSS) or unstable (MSI) state or phenotype; and c) for each microsatellite locus, evaluating the accuracy of the classifier for distinguishing between stable (MSS) and unstable (MSI) states or phenotypes by comparing the output of the classifier to the known actual phenotype of the microsatellite locus or to a reference number of indels at the microsatellite locus; Wherein the classifier is considered an accurate classifier for determining the MSS or MSI status or phenotype of a tumor if the classifier exhibits accuracy in counting indels at each of the microsatellite loci with a resolution of 1 base pair.

8. The method of claim 7, wherein the unstable (MSI) microsatellite loci used to prepare the training set are obtained from DNA of cancerous tumor cells, and the cancer is selected from colorectal cancer, endometrial cancer, prostate cancer, kidney cancer, gastric cancer, bile duct cancer, pancreatic cancer, lung cancer or brain cancer.

9. The method of claim 6, wherein the method uses a classifier trained using the method of claim 7 or 8 to determine the microsatellite stable (MSS) or unstable (MSI) status of a tumor.

10. The method according to any one of claims 7 to 9, wherein the classifier is selected from a random forest (RF) classifier, a support vector machine (SVM) classifier, a decision tree classifier, a K-nearest neighbor classifier (KNN), a logistic regression classifier, a nearest neighbor classifier, a Gaussian mixture model (GMM) classifier, a nearest centroid classifier, a linear regression classifier, and a neural network, such as an artificial, deep, convolutional, or fully connected neural network.

11. The method of any one of claims 1 to 10, wherein the method involves next generation sequencing (NGS) and further optionally comprises assessing the expression or loss of expression of at least one of the MLH1, MSH2, MSH6 and PMS2 proteins of the mismatch repair (MMR) system.

12. A computing system comprising: - a memory storing at least one instruction of a classifier trained according to the method of any one of claims 7 to 10, and - a processor accessing the memory to read the instructions and execute the method according to claim 6 or 11.

13. A kit for analyzing microsatellite loci of human genomic DNA, comprising oligonucleotide primers for co-amplifying a set of microsatellite loci of human genomic DNA, characterized in that The method comprises at least two primer pairs suitable for amplifying CABIO-P05, CABIO-P07, CABIO-E01, CABIO-E03, CABIO-E04, CABIO-E05, CABIO-E06 and / or CABIO-E07 according to claim 1 or any combination thereof, and optionally a thermostable polymerase and / or a control DNA isolated from normal non-cancerous biological material and / or lacking mismatch repair genes.

14. Use of the kit according to claim 13 for analyzing microsatellite stability or instability.

Images