Method for early detection of precancerous / cancerous lesions based on liquid biopsy

Abstract

The present invention relates to the field of cancer and methods for early detection of cancerous lesions. In particular, the present invention relates to methods for early detection of cancerous lesions by analyzing microsatellite instability in the genome of an individual from liquid biopsy.

Description

TECHNICAL FIELD

[0001] The present invention relates to the field of cancer and to methods for early detection of cancerous lesions.

[0002] In particular, the present invention relates to a method for early detection of cancerous lesions by analyzing microsatellite instability in the genome of an individual from a liquid biopsy. BACKGROUND

[0003] Lynch Syndrome (LS) is a genetic disorder (autosomal dominant) characterized by an increased risk of developing colorectal cancer (CRC) and other malignancies (albeit with a lower associated risk). It is caused by a germline mutation in a gene involved in the DNA repair mechanism (mismatch repair), such as MLH1, MSH2, MSH6, PMS2 or EPCAM (the latter caused by inactivation of the MSH2 gene). The prevalence of LS in the human population is estimated to be 1:2000 to 1:660. Tumors occurring in patients with Lynch Syndrome show microsatellite instability (MSI) and loss of expression of at least one MMR protein on tumor tissue (which can be confirmed by immunohistochemistry techniques).

[0004] LS is the most common inherited colorectal cancer syndrome, with an estimated risk of CRC in LS patients that is about 14 times higher than in the general population and with an earlier age of onset (usually before 50 years of age). In Italy, more than 50,000 new cases of CRC are diagnosed each year. Of these, about 3-5% are associated with LS.

[0005] LS is a rare condition, but undoubtedly has important social relevance. Patients usually develop cancer at a young age and have a high risk of developing metachronous tumors. In addition, their blood relatives perceive a high risk of developing cancer.

[0006] In addition to preventive colectomy in specific cases, close endoscopic surveillance (with colonoscopy every 1-2 years) starting in the early 20s is the only option to reduce morbidity and mortality in LS patients.

[0007] However, endoscopy as an invasive and costly procedure is generally not well accepted by patients, especially after ten or twenty years of monitoring from the start, when the risk increases. Moreover, it is to be considered that there is a large variability in the annual or biannual screening of patients and, like many screening processes, there can be over-diagnosis or, conversely, failure to detect cancer between intervals of monitoring procedures. Furthermore, in some cases, CRC development in Lynch syndrome patients skips the adenoma stage, with rapid and aggressive tumor growth occurring in the case of MMR deficiency. This can quickly lead to an aggressive phenotype. Since mismatch repair-deficient crypt foci (MMR-DCF) can grow under the intact mucosal surface, these types of lesions often escape detection by colonoscopic monitoring, directly evolving into overt cancer without a non-invasive precursor that is visible to the naked eye.

[0008] A less invasive test that accurately detects the presence of early lesions is essential for risk stratification, tailoring of monitoring programs and increasing patient acceptance. Molecular biomarkers that can be detected by minimally invasive testing in blood, stool and / or urine samples, are easy to handle and repeatable at short intervals, can overcome these limitations.

[0009] Nucleic acids released by tumor cells and circulating in the plasma have emerged as promising biomarkers for cancer diagnosis and monitoring, especially in liquid biopsies.

[0010] One characteristic of LS colorectal cancers is their MSI status. For a long time, MMR deficiency was considered a secondary event in the pathogenesis of LS cancers, but this view has recently been challenged. Ahadova et al. (2018) argue that there are different carcinogenic pathways in LS, with MMR deficiency usually representing an early possible initiating event. The most widely used commercial kit for identifying MSI in tumor tissue samples detects 5 microsatellite markers (pentaplex analysis): BAT 25, BAT 26, NR-21, NR-24 and MONO 27. Tumors that are unstable in two or more markers are classified as MSI. Moreover, Silveira et al. (2020) demonstrated that it is possible to identify MSI in circulating tumor DNA (ctDNA) in liquid biopsies of patients with advanced colorectal cancer, suggesting that MSI can be considered a possible candidate non-invasive biomarker in LS subjects.

[0011] The need and importance of developing a non-invasive and effective test for identifying precancerous lesions in patients in need thereof is increasingly highlighted.

[0012] It is therefore an object of the present application to develop a diagnostic method relying on liquid biopsy. SUMMARY

[0013] In the complex situation of detecting the presence of early lesions, the inventors have developed a non-invasive detection method based on liquid biopsy, which enables the detection of the presence of early lesions. The detection method according to the present invention enables stratification of the risk of developing cancer, tailoring the monitoring program to patients with a higher risk of developing pre-neoplastic and (early) neoplastic lesions, and improving the acceptance of patients who would otherwise have to undergo invasive colonoscopy monitoring.

[0014] The present invention thus aims at a method for early detection of microsatellite instability associated with pre-neoplastic or neoplastic lesions in the genome of an individual with Lynch syndrome, comprising the steps of: a. providing a biological sample obtained from a liquid biopsy of said individual; b. isolating circulating cell-free nucleic acids (cfDNA) from the biological sample; c. assessing microsatellite instability of at least one microsatellite locus in said cfDNA, wherein said microsatellite instability is detected in three or more microsatellite loci selected from the group consisting of BAT25, BAT26, NR21, NR24 and Mono27, wherein said three or more loci comprise at least BAT25, BAT26 and NR24.

[0015] In a second aspect, the present invention relates to a biomarker for detecting a pre-neoplastic or neoplastic lesion in an individual with Lynch syndrome by assessing microsatellite instability in three or more loci selected from the group consisting of BAT25, BAT26, NR21, NR24 and Mono27. BRIEF DESCRIPTION OF DRAWINGS

[0016] The features and advantages of the present invention will become apparent from the specific embodiments described below, given by way of illustration and non-limitation, and from the appended Figures 1-6 in which: Figure 1 : Figure showing the inclusion criteria for patient recruitment as described in Example 1.

[0017] Figure 2 : Heatmap of available plasma samples for dPCR analysis and corresponding lesions detected by endoscopy. LS patients were classified according to the presence or absence of lesions at baseline as described in Example 1.

[0018] Figure 3: Analysis of circulating free DNA (cfDNA) concentration (ng / mL) measured in TapeStation and MSI dPCR analysis, as described in Example 5. Comparison of cfDNA concentration and percentage of wells with FAM and / or HEX amplification signal using MSI dPCR (A) Assay 1, (B) Assay 2 and (C) Assay 3. Spearman R and relative p-value of cfDNA concentration and MAF of the 5 microsatellites compared in the analysis. (D) Heatmap of Spearman R with relative p-value of MAF and cfDNA concentration of the 5 microsatellites compared in the analysis. (E) Distribution of cfDNA concentration and (F; G) MAF data according to the presence and type of lesion detected by colonoscopy at the time of plasma sample collection. Samples were grouped according to absence of any lesion (No lesion; n=52), presence of any lesion (Lesion; n=26), low grade disease (LgD; n=15), high grade disease (HgD; n=8) and adenocarcinoma (ADK; n=3). Median values are reported. Asterisks indicate significance p-value: .

[0019] Figure 4 : Distribution of MAF data according to sample collection time (T0 vs. T1-T2) in LS patients, (A) with colonoscopy-detected lesions and (B) without colonoscopy-detected lesions.

[0020] Figure 5 : Fitted ROC curve (black line) of 95% CI (blue line) and relative AUC to discriminate between 26 plasma samples collected with presence of colonoscopy-identified lesions and 52 plasma samples collected without presence of any lesion. MAF of each of the 5 microsatellites (A-E) and (F) MAF added together were considered as continuous data.

[0021] Figure 6 : Fitted ROC curve (black line) of 95% CI (blue line) and relative AUC to discriminate between 18 plasma samples with presence of colonoscopy-identified lesions at baseline and 18 matched controls without lesions at baseline. MAF of each of the 5 microsatellites (A-E) and (F) MAF added together were considered as continuous data.

[0022] Figure 7: Performance of digital PCR detection was assessed. (A) Limit of blank (LoB) and (B) limit of detection (LoD) of the five markers were assessed. LoB was defined as the upper limit of the 95% confidence interval of the mean false positive MAF values of 25 wild-type DNA samples. LoD was determined by serial dilution experiments using MSI positive and negative DNA, in triplicate, and defined as the lowest expected MAF over LoB with corresponding observed positive signal in all three replicates. Mean values with standard deviation are reported.

[0023] Figure 8: Assess the utility of blood MSI (bMSI) to differentiate between patients with lesions. (A) ROC curves for cell-free DNA (cfDNA) and (B) blood microsatellite instability (bMSI) to differentiate between patients with and without colorectal lesions at baseline. (C-F) Spaghetti plots report the distribution of bMSI time trends over time according to the patient cluster distribution map of the presence of lesions. Points and dark gray dotted lines indicate the numerical values of the patients and their trend over time. DETAILED DESCRIPTION

[0024] Therefore, it is the object of the present application to provide a method for early detection of microsatellite instability associated with pre-cancerous or cancerous lesions in the genome of an individual suffering from Lynch syndrome, comprising the following steps: a. providing a biological sample obtained from a liquid biopsy of said individual; b. isolating circulating cell-free nucleic acids (cfDNA) from the biological sample; c. assessing microsatellite instability of at least one microsatellite locus in said cfDNA, wherein said microsatellite instability is detected in three or more loci selected from the group consisting of BAT25, BAT26, NR21, NR24 and Mono27, wherein said three or more loci comprise at least BAT25, BAT26 and NR24.

[0025] Liquid biopsies for the analysis of circulating tumor DNA (ctDNA) have emerged as promising tools for the diagnosis and monitoring of therapeutic response to several cancer types, including CRC (Gilson, P., Merlin, J. L. & Harle, A. Detection of microsatellite instability: State of the art and future applications in circulating tumour DNA (ctdna). Cancers vol. 13 (2021)). Recently, Silveira et al. reported that it is possible to detect MSI in liquid biopsies of patients with advanced CRC by digital PCR (dPCR) (Silveira, A. B. et al. High-Accuracy Determination of Microsatellite Instability Compatible with Liquid Biopsies. Clin. Chem. 66, 606-613 (2020)). However, none of these studies assessed the effectiveness of MSI in liquid biopsies for early detection and screening of CRC in high-risk populations.

[0026] The results of the retrospective case-control study evaluated the feasibility of MSI detection in liquid biopsies in samples collected during a CRC screening program, which led to the development of a method for the early detection of microsatellite instability. The participants of the study were LS patients who were monitored for up to 5 years. At each round of screening, all patients underwent endoscopy and plasma samples were collected for the analysis of molecular biomarkers, such as MSI.

[0027] The non-invasive method of the present application, based on liquid biopsy, enables the detection of the presence of early precancerous and cancerous lesions, reducing the need for invasive colonoscopy.

[0028] In a preferred embodiment, the method of the present application enables the early detection of microsatellite instability associated with precancerous or cancerous lesions in the genome of an individual suffering from Lynch syndrome, wherein said precancerous or cancerous lesions are metachronous lesions.

[0029] Through this detection, the physician can stratify the risk of cancer in the patient and tailor the monitoring program for high-risk patients.

[0030] The method of the present application exploits the detection of MSI in liquid biopsy. In a preferred aspect, the present application relates to a method for the early detection of microsatellite instability in the genome of an individual, wherein said microsatellite instability is detected in one or more loci selected from the group consisting of BAT25, BAT26, NR21, NR24 and Mono27, more preferably from the group consisting of BAT25, BAT26 and NR24.

[0031] As mentioned above, a great advantage of the method of the present application is that precancerous lesions can be identified by analyzing a liquid biopsy, preferably said liquid biopsy is a blood sample, a urine sample, a stool sample, a pap test, a cerebrospinal fluid or any other body fluid.

[0032] In the method for the early detection of microsatellite instability in the genome of an individual according to the present application, the step c. of assessing microsatellite instability of at least one microsatellite locus in cfDNA is performed by PCR, preferably by digital PCR (dPCR), more preferably said digital PCR is a droplet digital polymerase chain reaction.

[0033] Microsatellite instability in the genome of an individual indicates that the individual suffers from cancer, in most cases the type of cancer is colorectal cancer. It is also described that subjects with Lynch syndrome have a slightly increased risk of developing tumors of the stomach, small intestine, ureter and renal pelvis, pancreas, central nervous system, hepatobiliary system and skin.

[0034] Microsatellite instability in an individual's genome also indicates that the individual is a patient with Lynch syndrome or has a family history of LS.

[0035] As will be discussed and detailed in the Examples section, microsatellite instability in an individual's genome corresponds to the presence of one or more mutations in a microsatellite sequence. Microsatellites exhibiting instability in LS patients with precancerous lesions are microsatellites at loci BAT25, BAT26, NR21, NR24, and / or Mono27, with a preferred aspect being microsatellites at loci BAT25, BAT26, and NR24.

[0036] Patients with LS may not show any lesions during colonoscopy analysis, but they may have precancerous lesions that are present but difficult to identify. In such cases, the method of this invention analyzes liquid biopsy, where one or more MSIs from one or more of five loci (BAT25, BAT26, NR21, NR24, and Mono27) can identify the presence of lesions.

[0037] The inventors were surprised to discover that there exists a detection limit (LoD) applicable to each of the five loci. This is further illustrated in Example 3 and Figure 7, where it can be seen that the LoD is 0.02% for BAT26, 0.05% for BAT25, 0.03% for NR24, 0.06% for NR21, and 0.20% for Mono27.

[0038] The presence of microsatellite instability (MSI) can be detected using many methods known to those skilled in the art of molecular biology. A preferred method is to detect microsatellite instability in the genome of an individual with Lynch syndrome by detecting the lack of a double-positive signal from two probes binding to the same target microsatellite sequence. Depending on the microsatellite mutation state, the failure of one of the two probes to bind indicates the presence of MSI. In other words, when a wild-type DNA target molecule is present, both probes bind to the genome, producing a double-positive signal. When a mutant DNA molecule is present, only one of the two probes binds, producing a single-positive signal.

[0039] Therefore, the inventors have discovered that LS-related tumors are characterized by plasma MSI at loci BAT25, BAT26, NR21, NR24 and / or Mono27, and that the detection of plasma MSI is associated with precancerous or cancerous lesions, preferably precancerous or cancerous lesions of the colorectal cancer.

[0040] In a second aspect, the present application relates to a biomarker for detecting a precancerous or cancerous lesion in an individual with Lynch syndrome by assessing microsatellite instability in three or more loci selected from the group consisting of BAT25, BAT26, NR21, NR24 and Mono27. In a preferred aspect, the precancerous or cancerous lesion is a metachronous lesion.

[0041] In a preferred aspect, in the biomarker according to the present application, the loci are BAT25, BAT26 and NR24.

[0042] Surprisingly, the limit of detection (LoD) of the five loci of the biomarker is: 0.02% for BAT26, 0.05% for BAT25, 0.03% for NR24, 0.06% for NR21 and 0.20% for Mono27.

[0043] The various embodiments and aspects of the application as described above and as claimed in the claims section are experimentally supported in the following examples.

[0044] Examples Reference is now made to the following examples, which together with the above description illustrate some embodiments of the application.

[0045] Example 1 : Study design High-risk CRC patients, defined as subjects affected by LS and with identified pathogenic germline mutations in MMR genes, were enrolled and signed an informed consent.

[0046] Patients were enrolled at baseline (TO) to receive a colonoscopy and, according to standard screening guidelines, were invited to return for a second (T1 : 12-24 months after TO) and third (T2: 12-24 months after T1 ) examination. Plasma samples were collected from all patients at TO and, if possible, at other time points. During each endoscopy, precancerous and / or cancerous lesions were identified and annotated in a dedicated database (HeGint / Progeny). For each patient, genetic, molecular and clinical data and tumor family history were also collected.

[0047] LS patients with lesions at TO and a 1 : 1 matched control group for age and sex were selected for molecular analysis. The study was in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the IRCCS Istituto Nazionale Tumori, Milan, Italy.

[0048] Results Patient characteristics CRC screening study recruited 87 LS patients at baseline, of which 66 returned at T1 and 51 at T2 Figure 1 In the current study, available plasma samples from 18 baseline diseased patients and 1 : 1 matched controls were selected. There were no major differences in age, gender and MMR gene mutation status between the case and control groups (Table 1). In total, 78 samples were analyzed considering all time points, of which 25 were collected in the context of a colonoscopy finding a lesion Figure 2 ): 14 low grade dysplasia adenoma (LgD), 8 high grade dysplasia adenoma (HgD) and 3 adenocarcinoma (ADK). Three plasma samples of LS patients that were negative at T0 and T1 but developed LgD at T2 were also analyzed and considered as case reports.

[0049] Table 1: Stratification of patient characteristics according to the presence or absence of lesions found by colonoscopy

[0050] Included one additional LS patient (male, 43 years old, MLH1 mutation), without lesions at T0 and T1 and with lesions at T2 Example 2: Plasma sample collection and DNA extraction Plasma was collected and stored at baseline endoscopy and further screenings. Blood was drawn from screening volunteers using 10 mL Vacutainer tubes with spray-coated K2EDTA, plasma was isolated by two centrifugation steps at 1,258 x g and 4°C for 10 minutes and stored at -80°C in 1.5 mL cryovials. Circulating cell-free DNA (cfDNA) was extracted from up to 2 mL plasma using Maxwell RSC cfDNA Plasma Kit (Promega) following standard protocol and eluted in 50 uL elution buffer. The cfDNA fraction was quantified by electrophoresis using the TapeStation system (Agilent) with Cell-Free DNA Screening Tape. Data were analyzed using TapeStation Analysis Software 4.1.1 and considered 50-700 bp DNA fragments.

[0051] Example 3: MSI status by dPCR Stability of the 5 microsatellites was analyzed using the MSI multiplex dd PCR Expert Design Assay (Bio-Rad), the ProFlex 2x Flat PCR system (Thermo Fisher) and the QuantStudio 3D Digital PCR instrument (Thermo Fisher). Bio-Rad MSI probes were labeled with FAM or HEX fluorophores and competed for the detection of BAT25 and BAT26 (Assay 1), NR21 and NR24 (Assay 2) or Mono27 (Assay 3). In the presence of WT sequences, both probes bind the microsatellite, resulting in a double positive signal (FAM+ and HEX+). Conversely, in the presence of mutant sequences, one probe drops out, resulting in a single positive signal (FAM+ or HEX+). Briefly, a reaction mix of 15.8 μL containing 0.8 μL 20X Bio-Rad Assay, 8.4 μL 2X QuantStudio 3D Digital PCR Master Mix and 6.6 μL eluted cfDNA was prepared. All samples were loaded into a 20,000-well chip using an automated chip loader according to the manufacturer's specifications. Thermal cycling was performed under defined amplification conditions: 96°C for 10 minutes, 40 cycles of 55°C for 1 minute, 98°C for 30 minutes followed by an extension step of 60°C for 2 minutes.

[0052] DNA from tumor tissue samples from two MSI CRC patients was used as positive controls and elution buffer alone as a negative control. Tumor tissue DNA was quantified using a NanoDrop 2000 (Thermo Fisher) spectrophotometer and appropriately diluted to obtain a final concentration of 1-5 ng / μL. Thresholds for FAM and HEX positivity were established on the negative control. Positive outliers in the negative control were subtracted from each sample run. Two expert researchers analyzed the data using the QuantStudio 3D Analysis Suite online tool (Thermo Fisher). Disputes were expected to be resolved by face-to-face meetings. Chips with <10,000 wells that were quality tested (QT) eligible were excluded and repeated. Presence of ctDNA was assessed as mutant allele fraction (MAF) for each microsatellite. MAF was calculated as the number of mutant calls (FAM+ or HEX+) over the total number of mutant and wild type (WT) calls (FAM+ and / or HEX+).

[0053] To verify the overall performance of the assay, LoB was first evaluated using an extra sample of 25 WT and defined as the upper limit of the 95% CI of the mean false positive MAF. LoB for the five markers was 0.01% for BAT26, 0.05% for BAT25 and NR21, 0.03% for NR24 and 0.17% for Mono27. Figure 7A Then, repeatability and LoD of the assay were established by serial dilutions in triplicate of WT and mutant DNA mixed with expected MAF ranging from 25% to 0.02%. For each marker, all replicates beyond the expected LoB had a positive signal, with an rs higher than 0.9 between expected and observed MAF. Figure 7B Finally, LoD was evaluated for BAT26 at 0.02%, for BAT25 at 0.05%, for NR24 at 0.03%, for NR21 at 0.06% and for Mono27 at 0.20%.

[0054] Example 4: Calculations and statistical analysis For data analysis, cfDNA amount, percentage of dPCR wells with amplification signal and MAF values were considered as continuous data. Spearman R and relative p-value were used to correlate non-parametric continuous data. Student T, Mann Whitney and Kruskal-Wallis tests were used as appropriate to assess differences between parametric or non-parametric distributed data. All tests were two-sided and p-value < 0.05 was considered statistically significant. Area under the curve (AUC) of the receiver operating characteristic (ROC) curve was used to assess the performance of dPCR markers to distinguish patients with any colorectal lesion from controls at all time points and at T0 only.

[0055] For dPCR data analysis, the average number of target molecules per partition (l) was calculated as l = -ln(w / n) taking "w" and "n" as the number of negative and total partitions in each reaction, respectively. The average reaction volume in each partition was taken as "v" and then the total number of copies, WT copies and mutant DNA copies per dPCR mix per microliter were calculated as copies / pL = l / v. Mutant allele frequency (MAF) was calculated as the ratio of mutant copies per pL to total copies (mutant and WT) per pL.

[0056] Twenty-five additional control samples (18 cfDNA and 7 genomic DNA (gDNA)) with WT microsatellite sequences were used to assess the limit of blank (LoB) for each marker. LoB was defined as the upper limit of the 95% CI of the mean false positive MAF values. The limit of detection (LoD) and the reproducibility of the assay for each marker were assessed by serial dilution experiments using DNA of positive controls and cfDNA of MSI-negative patients (triplicates). LoD was defined as the lowest expected MAF over the LoB, for which a corresponding positive signal was observed in all three replicates. Experiments were performed according to the minimum information published by the guidelines for quantitative digital PCR experiments (MIQE) and all required information is available in the main text or in the supplementary files.

[0057] MAF values were considered as continuous data and bMSI was defined as the sum of the MAF values of the five markers. To facilitate the evaluation of the changes in the results of the detection related to the monitoring period and the presence of lesions, we classified the dPCR MAF data as positive (any value above the LoB) and negative.

[0058] Example 5: Microsatellite MAF in plasma Microsatellite MAF in plasma was independent of cfDNA concentration and lesion type.

[0059] For the molecular analysis, we started from a fixed volume (6.6 μΐ^) of elution buffer containing 0.2 ng to 1.5 ng of cfDNA. By observing the dPCR data, the number of wells with any amplification signal, either from WT or mutated DNA molecules, reflected the cfDNA content in the plasma sample, which was measured by electrophoresis on a TapeStation (Agilent, Santa Clara, CA, USA) (Figure 1A-C). Figure 3 A-C in Figure 1.

[0060] To assess whether the MAF of the 5 microsatellites analyzed were independent from each other and from the amount of starting material, a Spearman correlation matrix was generated. As shown in D in Figure 1, only BAT25 MAF was positively correlated with BAT26 MAF and there was no significant correlation between any of the 5 markers and the cfDNA amount. Figure 3

[0061] The distribution of cfDNA and MAF values was then assessed by stratifying the samples according to the presence, type of detected lesion and screening time point. When comparing samples collected in the presence or absence of any lesion, or comparing the 3 different types of lesions, there were no differences in cfDNA concentration (E in Figure 2). Figure 3 F in Figure 2), but when observing the type of lesion (G in Figure 2), the MAF values of BAT25, BAT26 and NR24 were significantly higher in the presence of a lesion than in the absence of a lesion. Figure 3 F in Figure 2), but when observing the type of lesion (G in Figure 2), the MAF values of BAT25, BAT26 and NR24 were significantly higher in the presence of a lesion than in the absence of a lesion.​Figure 3 or when samples are stratified according to screening time point (Fig. 6) Figure 4 ) was not significant.

[0062] Example 6: Liquid biopsy distinguishes LS patients with colorectal lesions To assess the ability of liquid biopsy to distinguish between patients with and without lesions, ROC curve analysis was considered for all available samples. The amount of cfDNA measured by electrophoresis did not show discriminating power, with an AUC of 0.522 (95% CI: 0.36; 0.69) (Fig. 6A). On the other hand, as shown in Fig. 6B, the AUC values of the fitted ROC curves for the 5 microsatellites were 0.713 (BAT25), 0.68 (BAT26), 0.579 (NR21), 0.731 (NR24) and 0.594 (Mono27). A higher AUC value was obtained by the sum of the 5 markers, equal to 0.818. When considering only baseline matched samples, the results were similar (AUC: 0.81) (Fig. 6C). Figure 6 Figure 5 ). When considering the 3 plasma samples of LS patients who were negative at T0 and T1 and developed LgD at T2, a gradual increase of the total MAF value was observed at the approach of disease onset: 0.94% at T0, 1.4% at T1 and 1.94% at T2. Figure 6

[0063] Example 7: Diagnostic value of blood MSI To assess the ability of liquid biopsy to distinguish between patients with and without lesions, we assessed the AUC of cfDNA and bMSI successively. While the amount of cfDNA did not show discriminating power, with an AUC of 0.61 (95% CI: 0.42; 0.80) (Fig. 7A), bMSI was able to distinguish patients according to the presence of lesions, with an AUC of 0.80 (95% CI: 0.66; 0.94) (Fig. 7B). This result was maintained after cross-validation (AUC-CV: 0.74; 95% CI: 0.58; 0.91). Figure 8 Figure 8 For exploratory purposes, we checked the sensitivity and specificity of bMSI, by selecting a cut-off value that guaranteed at least 75% sensitivity, the specificity of bMSI was 72%. To assess the possible utility of this tool, we assessed its positive predictive value (PPV) and negative predictive value (NPV) by considering the T0 prevalence of lesions registered in our entire series: this led to a PPV of 42% and a NPV of 93%.

[0064]

[0065] ​​​​By stratifying patients according to the presence of lesions, we observed a temporal trend pattern of bMSI. In LS patients with lesions at T0 but without lesions at T1 and T2, 8 out of 10 patients (80%) had a decrease of bMSI value compared to baseline (C in Table 1). Figure 8 Conversely, in 7 patients with metachronous lesions detected at T1 (D in Table 1) or T2 (E in Table 1), bMSI values showed an ambiguous temporal trend. Finally, when considering LS patients who never developed lesions, the range of bMSI values at T0, T1 and T2 remained comparable (F in Table 1). Figure 8 Figure 8 Figure 8

[0066] Discussion These results demonstrate that MSI can be measured in LS patients by liquid biopsy and can be used for CRC screening. There was no difference in plasma cfDNA content when comparing samples collected in the presence or absence of lesions.

[0067] The diagnostic method of the present invention enables the analysis of MSI in liquid biopsy using dPCR-based tools in the context of CRC screening. This minimally-invasive detection has an excellent potential and a wide applicability. LS-associated tumors are characterized by the loss of expression of at least one MMR protein and MSI, which is often an early event also occurring in the context of precancerous lesions (e.g. adenomas).

[0068] If specificity is an important parameter when analyzing low-input biomarkers (e.g. ctDNA), sensitivity is even more important for detecting early-stage disease when, in the context of screening studies, the disease is still curable by surgical resection. In this retrospective study, we were able to detect mutations with a frequency as low as 0.2% starting from 2 mL of plasma. Preliminary analyses suggest that mutations with a frequency < 0.01% can be measured starting from 3-4 mL of plasma.

[0069] From the above description and the above examples, the advantages achieved by the method described and obtained according to the present invention are evident.​​​

Claims

1. A method for early detection of microsatellite instabilities in the genome of individuals with Lynch syndrome that are associated with precancerous or cancerous lesions, comprising the following steps: a. Provide biological samples obtained from a liquid biopsy of the individual; b. Isolate circulating cell-free nucleic acids (cfDNA) from the biological sample; c. Assessing microsatellite instability at at least one microsatellite locus in the cfDNA, wherein the microsatellite instability is detected in three or more microsatellite loci selected from the group consisting of BAT25, BAT26, NR21, NR24 and Mono27, wherein the three or more loci include at least BAT25, BAT26 and NR24.

2. The method according to claim 1, wherein the liquid biopsy is a blood sample, urine sample, or fecal sample.

3. The method according to any one of claims 1 or 2, wherein step c, assessing microsatellite instability, is performed by PCR.

4. The method according to claim 3, wherein, The PCR mentioned is an exfoliated digital polymerase chain reaction (dPCR).

5. The method according to any one of claims 1 to 4, wherein the precancerous lesion or cancerous lesion is a colorectal precancerous lesion or cancerous lesion.

6. The method according to any one of claims 1 to 5, wherein the precancerous lesion or cancerous lesion is a metachronous lesion.

7. The method according to any one of claims 1 to 6, wherein the detection limits (LoD) of the five loci are: 0.02% for BAT26, 0.05% for BAT25, 0.03% for NR24, 0.06% for NR21, and 0.20% for Mono27.

8. The method according to any one of claims 1 to 7, wherein the microsatellite instability in the genome of an individual suffering from Lynch syndrome corresponds to the presence of one or more mutations in the microsatellite sequence.

9. The method according to any one of claims 1 to 8, wherein the microsatellite instability in the genome of an individual with Lynch syndrome is detected by the absence of a double-positive signal from two probes that bind to the microsatellite sequence.

10. A biomarker for detecting precancerous or cancerous lesions in individuals with Lynch syndrome by assessing microsatellite instability at three or more loci selected from the group consisting of BAT25, BAT26, NR21, NR24 and Mono27.

11. The biomarker of claim 10, wherein the locus is BAT25, BAT26, and NR24.

12. The biomarker according to claim 10, wherein the detection limits (LoD) of the five loci are: 0.02% for BAT26, 0.05% for BAT25, 0.03% for NR24, 0.06% for NR21, and 0.20% for Mono27.

13. The biomarker according to any one of claims 10 to 12, wherein the precancerous lesion or cancerous lesion is a metachronous lesion.

Images