Biomarker composition for detecting early gastric cancer and application of biomarker composition

Abstract

According to the present invention, four methylation sites (Cg26556436, Cg02063166, Cg05661282 and Cg03234186) are screened, the methylation sites show significantly increased methylation in the early gastric cancer, and the ROC curve and the hierarchical clustering result show that the early gastric cancer can be distinguished from the normal control by using the methylation sites. Therefore, the four methylation sites screened by the invention can be used as a biomarker combination for clinically detecting early gastric cancer. In addition, through a GC-mqMSP efficient diagnosis method, multiple quantification can be simultaneously performed on four methylation markers in a single reaction. Compared with the existing recognized CDO1 (AUC: 0.762) marker, the GC-mqMSP (AUC: 0.839) has an outstanding advantage in the detection of the early gastric cancer.

Description

Technical Field

[0001] This invention belongs to the field of molecular diagnostic technology, and specifically relates to the application of a biomarker. Background Technology

[0002] Gastric cancer is one of the five most common cancers worldwide and the third leading cause of cancer death, with a particularly high burden in East and Southeast Asia. Identified risk factors for gastric cancer include demographic variables (age, sex, race), Helicobacter pylori infection, smoking, and high-salt diets (Multi-omics analysis identifies LANCL2 as a potential biomarker for the diagnosis and prognosis of gastric cancer [J]. SciRep. 2025;15(1):18231.). Early detection, diagnosis, and treatment are effective strategies to reduce morbidity and mortality. Previous studies have shown that the 5-year and 10-year survival rates after endoscopic mucosal resection for early gastric cancer are both 99%, while the rate for late-stage cancer is only around 20% (Longterm outcomes after endoscopic mucosal resection for early gastric cancer [J]. Gastric Cancer. 2006;9(2):88-92.). Currently, the early detection rate of gastric cancer remains low. This is partly due to the lack of typical clinical symptoms in early-stage gastric cancer and a generally low level of public awareness regarding early diagnosis and treatment. It is also due to the lack of effective and convenient early diagnostic methods. The gold standard for gastric cancer screening is endoscopy, which can reduce the mortality rate by 65%. However, its invasiveness, high cost, painful procedure, and risks of complications such as bleeding and perforation limit its widespread clinical application. Emerging targeted and immunotherapies are also limited in their large-scale application due to high costs, low overall efficacy, and the risk of immune-related adverse events. Therefore, there is an urgent need to explore new strategies for the diagnosis and treatment of gastric cancer.

[0003] Early diagnosis of gastric cancer is crucial for prognosis. Molecular aberrations, especially epigenetic changes, are detected much earlier than histological changes (Epigenetic signatures in gastric cancer: current knowledge and future perspectives[J]. Expert Rev Mol Diagn. 2022;22(12):1063-1075.). Cell-free DNA (cfDNA) carries DNA methylation signals specific to malignant tumors and is therefore recognized as a relatively ideal non-invasive early cancer screening target. However, in practical applications, methylation biomarkers generally have better diagnostic performance for gastric cancer tissue, and there is a lack of reports on high-performance plasma cfDNA methylation biomarkers. Ren et al. identified 153 blood cfDNA methylation markers. Among them, when DOCK10, CABIN1 and KCNQ5 were used as a combination of markers to detect gastric cancer at different stages (stage I, stage II, stage III and stage IV), the sensitivities were 44%, 59%, 78% and 100%, respectively (Genome-Scale Methylation Analysis of Circulating Cell-Free DNA in Gastric Cancer Patients. Clin Chem. 2022;68(2):354-364.), which showed low sensitivity for the diagnosis of early gastric cancer. Grail, a leading company in the liquid biopsy field, released the latest research results of its Extracellular Genome Atlas (CCGA) project in March 2020. In a clinical cohort of 6,689 participants representing more than 50 cancer types, the project achieved an accuracy of 93% in cancer diagnosis and tissue tracing using ctDNA methylation. However, for early-stage (stage I and II) cancers, while the specificity was well controlled at around 99.9% to meet early screening requirements, the sensitivity was only 18% (stage I) and 43% (stage II), indicating significant room for improvement. The SiboV® RNF180 / Septin9 gene methylation detection kit and the Aklan® Reprimo / TCF4 / SDC2 gene methylation detection kit are currently approved gastric cancer gene methylation detection kits. The SiboV® RNF180 / Septin9 kit has a specificity of 85.07% and a sensitivity of 61.76% for gastric cancer detection, while the Aklan® Reprimo / TCF4 / SDC2 kit has a clinical specificity of 92.07% and a sensitivity of 80.77%.However, the detection sensitivity for early gastric cancer (stage I and stage II) is only 50.0% and 62.32%, respectively (Combining methylated SEPTIN9 and RNF180 plasma markers for diagnosis and early detection of gastric cancer[J]. Cancer communications (London, England). 2023; 43: 1275-9.)(Sensitive and specific multi-cancer detection and localization using methylation signatures in cell-free DNA. Ann Oncol[J].2020 Jun;31(6):745-759.).

[0004] Tumor heterogeneity leads to differences in genes and phenotypes among different cells. This heterogeneity may prevent liquid biopsies from fully reflecting the overall picture of the tumor when detecting tumor-related substances (such as DNA and RNA). Since liquid biopsies are typically performed by detecting tumor-related markers in body fluids, the heterogeneity within tumors may cause uneven distribution of these markers in the body fluids, thus affecting the accuracy and sensitivity of the test.

[0005] Therefore, how to accurately detect it and take into account tumor heterogeneity factors in cfDNA methylation level has become an urgent problem to be solved in liquid biopsy technology. Summary of the Invention

[0006] This invention identifies a combination of DNA methylation biomarkers for diagnosing early gastric cancer. Early gastric cancer patients are diagnosed using GC-mqMSP detection technology, which demonstrates high sensitivity for early gastric cancer, superior to CDO1, the best-reported biomarker for early diagnosis. Based on this, the invention was completed.

[0007] In a first aspect, the present invention provides a biomarker combination for detecting early gastric cancer, the biomarker combination comprising a Cg26556436 methylation site located in the chr6_88875621_88875622 fragment, a Cg02063166 methylation site located in the chr6_88875650_88875651 fragment, a Cg05661282 methylation site located in the chr19_58220369_58220370 fragment, and a Cg03234186 methylation site located in the chr19_58220656_58220657 fragment; When the GC-mqMSP index of the four methylation sites Cg26556436, Cg02063166, Cg05661282, and Cg03234186 in the subject's biological sample is higher than 0.658, the subject is judged to have early gastric cancer. The formula for calculating the GC-mqMSP index is as follows: Where p is the GC-mqMSP exponent; x i For the i-th marker, deita-CT; a i b is the coefficient of the i-th marker; b is the cutoff value. .

[0008] Furthermore, the gastric cancer includes adenocarcinoma, signet ring cell carcinoma, adenosquamous carcinoma, and undifferentiated cell carcinoma.

[0009] Furthermore, the early gastric cancer includes stage I gastric cancer and stage II gastric cancer.

[0010] In a second aspect, the present invention provides a GC-mqMSP multiple methylation detection system for detecting early gastric cancer, the detection system comprising: The data input module, data processing module, and result output module are included. The data input module refers to inputting the methylation data of four methylation sites, Cg26556436, Cg02063166, Cg05661282 and Cg03234186, in the subject's biological sample; The data processing module performs logistic regression analysis on the methylation data input from the data input module to calculate the GC-mqMSP index. The calculation formula is as follows: Where p is the GC-mqMSP exponent; x i For the i-th marker, deita-CT; a i b is the coefficient of the i-th marker; b is the cutoff value. ; The result output module outputs the GC-mqMSP index obtained above. When the GC-mqMSP index output by the result output module is greater than 0.658, the subject is judged to have early gastric cancer.

[0011] Furthermore, the biological samples are selected from the subject's tissue and blood samples.

[0012] Furthermore, the biological sample is preferably a blood sample.

[0013] Furthermore, the method for obtaining the methylation data is selected from one of the following: bisulfite sequencing PCR, methylation-specific PCR, high-resolution melting curve analysis, or digital PCR.

[0014] Furthermore, the preferred method for obtaining the methylation data is methylation-specific PCR.

[0015] In one specific embodiment of the present invention, methylation-specific PCR is employed. Primers and probes targeting four methylation sites—Cg26556436, Cg02063166, Cg05661282, and Cg03234186—as well as the internal reference gene of β-actin are designed. Methylation-specific PCR is then performed on the biological samples of the subjects to obtain methylation data for the four methylation sites—Cg26556436, Cg02063166, Cg05661282, and Cg03234186—in the biological samples of the subjects.

[0016] Furthermore, the four methylation sites are the Cg26556436 methylation site located in the chr6_88875621_88875622 fragment, the Cg02063166 methylation site located in the chr6_88875650_88875651 fragment, the Cg05661282 methylation site located in the chr19_58220369_58220370 fragment, and the Cg03234186 methylation site located in the chr19_58220656_58220657 fragment.

[0017] Furthermore, the gastric cancer includes adenocarcinoma, signet ring cell carcinoma, adenosquamous carcinoma, and undifferentiated cell carcinoma.

[0018] Furthermore, the early gastric cancer includes stage I gastric cancer and stage II gastric cancer.

[0019] Thirdly, the present invention provides the use of reagents for detecting the combination of biomarkers described in the first aspect in the preparation of reagents for diagnosing early gastric cancer.

[0020] Furthermore, the biomarker combination includes four methylation sites: Cg26556436, Cg02063166, Cg05661282, and Cg03234186.

[0021] Furthermore, the four methylation sites are Cg26556436 methylation site located in the chr6_88875621_88875622 fragment, Cg02063166 methylation site located in the chr6_88875650_88875651 fragment, Cg05661282 methylation site located in the chr19_58220369_58220370 fragment, and Cg03234186 methylation site located in the chr19_58220656_58220657 fragment.

[0022] Furthermore, the gastric cancer includes adenocarcinoma, signet ring cell carcinoma, adenosquamous carcinoma, and undifferentiated cell carcinoma.

[0023] Furthermore, the early gastric cancer includes stage I gastric cancer and stage II gastric cancer.

[0024] Beneficial effects This invention identified four methylation sites (Cg26556436, Cg02063166, Cg05661282, and Cg03234186). These methylation sites showed significantly elevated methylation levels in patients with early-stage gastric cancer. ROC curves and hierarchical clustering results demonstrated that these methylation sites could distinguish early-stage gastric cancer from normal controls. Furthermore, the four methylation sites (AUC: 0.839) showed greater advantage than the recognized CDO1 (AUC: 0.762) biomarker in the differential diagnosis of early-stage gastric cancer. Therefore, the four methylation sites identified in this invention can be used as a combination of biomarkers for the clinical detection of early-stage gastric cancer. Attached Figure Description

[0025] Figure 1 A schematic diagram of the process for designing and screening specific methylation sites for research.

[0026] Figure 2 This study aimed to analyze the genomic methylation profile of gastric mucosa tissue.

[0027] Note: A is the study design flowchart; B is the principal component analysis of 707 differentially methylated probes; C is the chromosomal distribution bar chart of differentially methylated probes, where orange and blue bars represent hypermethylation and hypomethylation events, respectively; D is the comparison of methylation distribution around transcription start sites in gastric cancer patients and normal controls; E is the CpG site differential methylation volcano plot (showing Delta and Beta values), where orange indicates hypermethylation, blue indicates hypomethylation, and gray indicates CpG sites with no significant difference; F is the GO enrichment analysis of differentially methylated genes; G is the unsupervised hierarchical clustering heatmap of 24 candidate differentially methylated probes in 43 pairs of tissue samples.

[0028] Figure 3 To identify and validate DNA methylation markers associated with gastric cancer.

[0029] Note: AC. β-value distribution of four candidate methylation markers in 43 pairs of tumor / adjacent tissues (A), plasma cfDNA from 79 healthy controls and 70 gastric cancer patients (B), and the TCGA database (C); DE. Association between four candidate markers and clinical characteristics in tissue samples (D) and plasma cfDNA (E); FG. ROC curves of four methylation markers in tissue samples (F) and plasma cfDNA (G) from the bisulfite-targeted sequencing cohort (all stages and stage I-II gastric cancer). H represents unsupervised hierarchical clustering of the four methylation markers based on differential methylation levels between gastric cancer patients and normal controls in the bisulfite-targeted sequencing cohort.

[0030] Figure 4 For the development and clinical application of GC-mqMSP.

[0031] Note: A shows the cfDNA methylation levels detected by GC-mqMSP in 106 gastric cancer patients and 70 healthy controls; B shows the ROC curve comparison of GC-mqMSP performance for gastric cancer diagnosis with that of individual markers (Cg26556436, Cg02063166, Cg05661282, Cg03234186); Left side of CD: training set (C, 74 gastric cancer vs. 49 control groups) and validation set (D, 32) In the gastric cancer vs. 21 control group, methylation levels were quantified by GC-mqMSP. Middle: AUC value and confusion matrix of GC-mqMSP classification; Right: Supervised hierarchical clustering of differential methylation features (training set: n=123; validation set: n=53); E is the distribution of GC-mqMSP values ​​in gastric cancer (n=106), lung cancer (LC, n=41), cancer (CRC, n=23) and normal control (n=70) on the left, Youden index = 0.658 (dashed line), Middle: Confusion matrix between gastric cancer and non-gastric cancer; Right: AUC (0.912) used to distinguish gastric cancer from other cancers.

[0032] Figure 5 The effectiveness of GC-mqMSP in diagnosing gastric cancer in an external validation cohort.

[0033] Note: AB. Left panel: Methylation levels quantitatively analyzed by GC-mqMSP (A) and CDO1 (B) in an external validation cohort (45 gastric cancer cases vs. 13 controls). Middle panel: AUC values ​​and confusion matrix of GC-mqMSP classification. Right panel: Differential methylation characterization profile.

[0034] Figure 6 This refers to the system ratio of sample to reagent.

[0035] Figure 7 This refers to the system ratio of sample to reagent.

[0036] Figure 8 This refers to the system ratio of sample to reagent.

[0037] Figure 9 This is an MS-qPCR reaction system.

[0038] Figure 10 Primer and probe sequence information for targeting the four DMPs. Detailed Implementation

[0039] The specific embodiments of the present invention will be further described below. It should be noted that these descriptions are for the purpose of aiding understanding the present invention, but do not constitute a limitation thereof. Furthermore, the technical features involved in the embodiments described below can be combined with each other as long as they do not conflict with each other.

[0040] Unless otherwise specified, the experimental methods used in the following embodiments are conventional methods, and the experimental materials used in the following embodiments are all available through conventional commercial channels.

[0041] Terminology Explanation Post-treatment progression tissue refers to the phenomenon where, after treatment, changes in disease or pathological state lead to alterations in tissue structure, function, or pathological characteristics. This phenomenon can involve various diseases and treatments, including cancer, inflammatory diseases, and complications after organ transplantation. In cancer treatment, post-treatment progression tissue typically manifests as tumor recurrence or metastasis.

[0042] The basic information of the differential methylation sites Cg26556436, Cg02063166, Cg05661282 and Cg03234186 described in this invention is as follows: Example 1: Screening of DNA methylation biomarkers related to early gastric cancer diagnosis Displacement criteria and sample selection Inclusion criteria: age 40-80 years; participants were able to undergo gastroscopy and other related examinations; patients with gastric cancer and precancerous lesions were required to have not undergone surgery, radiotherapy, or chemotherapy before blood collection; healthy controls were required to have negative gastroscopy and / or pathology results and to have no other cancers; other cancer controls (including colorectal cancer, lung cancer, breast cancer, and esophageal cancer) were required to have primary, single-type cancers and to have not undergone surgery, radiotherapy, or chemotherapy before blood collection; participants fully understood the nature of this study, the related examination methods, and the risks that may be involved in participating in this study, and were able to understand and voluntarily sign the informed consent form.

[0043] Exclusion criteria: Small blood sample volume, plasma less than 1.5 ml, which is not suitable for cfDNA extraction; incomplete blood sample information (including sample collection time, sample number, subject age, gender, medical history, imaging results, gastroscopy and pathology results, etc.); substandard blood sample testing quality control, including samples that do not meet the requirements for sample collection and processing, samples that have undergone repeated freeze-thaw cycles, hemolysis, or microbial contamination, samples tested using reagents with quality problems (including expired, contaminated, or improperly stored reagents, etc.), or other human errors during the test process; pregnant women, breastfeeding women, or women of childbearing age who are planning to conceive; individuals who cannot undergo screening and / or follow-up; and any circumstances deemed unsuitable for inclusion by the research physician.

[0044] A total of 170 patients were included according to the inclusion and exclusion criteria, including 106 patients with gastric cancer, 23 patients with colorectal cancer, and 41 patients with lung cancer. An additional 70 healthy controls were recruited. The 106 gastric cancer patients and 70 healthy controls were further randomly assigned to training and validation sets. The training set included 49 healthy controls and 74 gastric cancer patients; the validation set included 21 healthy controls and 32 gastric cancer patients.

[0045] Screening differentially methylation sites from public databases To explore gastric cancer-specific DNA methylation markers, genomic DNA methylation analysis was performed using the Gene Expression Omnibus (GEO), The Cancer Genome Atlas Program (TCGA), and The Human Protein Atlas (HPA) databases (see [link to database]). Figure 1 ).

[0046] This invention first used methylation data from normal gastric tissue from GEO (n=61) and mRNA expression data from gastric cancer from TCGA (T=352 vs n=31), as well as methylation data from gastric cancer from TCGA (T=385) and six other cancer types (bladder cancer, T=410 vs n=21; breast cancer, T=777 vs n=95; colorectal cancer, T=328 vs n=33; esophageal cancer, T=177 vs n=17; lung cancer, T=820 vs n=73; liver cancer, T=386 vs n=48) to identify the top 500 differentially methylated positions (DMPs) with Δβ (q≤0.05). DMPs were detected in the six cancer types (q≤0.05, |Δβ|>0.2), generating 811 DMPs. Subsequently, DMPs associated with gene expression repression in gastric cancer (Δβ>0.2, q≤0.01) were selected based on |log2FC|< -3, resulting in 851 DMPs. The integration of these three sets of data yielded 2131 DMPs.

[0047] Further analysis of the methylation status of 43 paired tumor and adjacent tissues (cohort 1) was performed using Wilcoxontests for differential methylation, yielding 1,149,300 DMPs. Intersecting these 2,131 DMPs with the 1,149,300 DMPs from the paired tissue analysis identified 707 DMPs, of which 286 were located within the genomic body. Principal component analysis used these 707 DMPs to demonstrate a clear separation between gastric cancer and normal controls (see [link to analysis]). Figure 2 B). Genomic distribution revealed more frequent hypomethylation events, and DMPs were distributed differently across chromosomes. Figure 2 C). Notably, compared to the control group, gastric cancer patients also exhibited high methylation levels in the gene body regions following the transcription start sites (TSS) (see [link]). Figure 2 D). Pathway analysis of differentially methylated genes showed enrichment in RNA transcription-related processes (see D). Figure 2 F). In addition, 80 DMPs with moderate to high expression levels in normal tissues recorded in the HPA database were selected. After prioritizing the DMPs, 24 DMPs were selected as target sites from the top five candidate genes by weight (see F). Figure 2 G).

[0048] Identification of gastric cancer-specific DNA methylation markers in tissue and plasma cfDNA To identify the most prominent and specific DNA methylation markers in gastric cancer tissues and plasma, methylation analysis was performed on cfDNA from 79 normal controls and 70 gastric cancer patients.

[0049] Test methods 1.1 Detection of plasma cfDNA methylation levels Collect blood samples from the patient and extract plasma cfDNA according to the instructions for use of the MagMAX™ Cell-Free DNA Isolation Kit (Thermo Fisher Scientific): (1) According to the volume of the specimen, add the corresponding volumes of reagents in sequence (see Figure 6 Add to the water bath. During the water bath, follow the instructions. Figure 7 Prepare the Binding Solution / Beads mixture, and then place it in an ice bath for 5 minutes after completion. (2) According to Figure 8 Add reagents as shown: Mix plasma with Binding Solution / Beads solution to allow cfDNA to bind to magnetic beads, transfer centrifuge tube to magnetic rack and let stand for 5 min, then discard supernatant.

[0050] (3) Washing with Wash Buffer: Resuspend the magnetic beads in 1 mL MagMAX™ Wash Solution. Transfer all the magnetic bead suspension to a new 1.5 mL centrifuge tube. Let the magnetic beads aggregate on a magnetic rack. Use the supernatant to wash back the residual magnetic beads on the original tube wall. Combine all the wash solutions into the same new tube, let stand for 2 min, and discard the supernatant. Add 1 mL Wash Solution, vortex to mix, let stand on a magnetic rack for 2 min, and discard the supernatant.

[0051] (4) Rinsing with 80% alcohol: Remove from the rack, resuspend the magnetic beads with 1 mL of 80% ethanol, vortex, let stand on the magnetic rack for 2 minutes, discard the supernatant, repeat the above operation to complete the 80% alcohol rinsing again, and then dry at room temperature.

[0052] (5) Elution of cfDNA: Add 25 µL of Elution Solution, vortex to elute, let stand on a magnetic rack for 2 min, and transfer 23 µL of supernatant to a new 1.5 mL centrifuge tube.

[0053] (6) cfDNA sulfidation: The zymo EZ DNA Methylation-Lightning™ Kit was used. 130 µL of Lightning Conversion Reagent was added to 20 µL of DNA sample and mixed well. The mixture was then aliquoted into two PCR tubes, 75 µL each, and placed in a PCR instrument at 98 °C for 8 min, followed by an incubation at 54 °C for 60 min. After the incubation, the tubes were stored at 4 °C. (7) Purification of the transformation product: Assemble the centrifuge column, add 600 μL of M-Binding Buffer, mix the samples from the two reaction tubes, and transfer them to the centrifuge column for mixing; transfer the mixed solution to a Zymo-Spin IC Column and centrifuge; add 100 μL of M-Wash buffer to the column and centrifuge; add 200 μL of L-Desulphonation Buffer to the column, incubate at room temperature for 15 min, and then centrifuge; add 200 μL of M-Wash buffer to the column and centrifuge, repeating once; transfer the column to a new 1.5 mL EP tube and let it stand, then add 16 μL of Low TE, incubate at room temperature for 2 min, centrifuge for 2 min, discard the adsorption column, and collect the sample.

[0054] (8) MS-qPCR: The reaction system is as follows Figure 9 As shown, the PCR program is as follows: pre-denaturation at 95℃ for 30s, denaturation at 95℃ for 5s, annealing / extension at 60℃ for 30s, and repeat the denaturation, annealing / extension steps 45 times.

[0055] (9) Quantitative detection of the methylation level of target cfDNA using fluorescence signals.

[0056] 2. Results Plasma cfDNA methylation sequencing data from 79 normal controls and 70 gastric cancer patients were compared with 24 DMPs screened from the database using LASSO regression analysis to establish the association between tissue and plasma DMPs. Four DMPs that met the criteria of q≤0.05 and |Δβ|>0.2 and overlapped between plasma and tissue were identified. These four DMPs, namely Cg26556436, Cg02063166, Cg05661282, and Cg03234186, were further validated in TCGA (see [link to TCGA]). Figure 3 AC).

[0057] The results showed that, compared with the normal control group, the methylation levels of these four hypermethylation markers were consistently higher in gastric cancer patients, as observed in both tissue and plasma cfDNA (see [link to study]). Figure 3Notably, all four DMPs showed significantly elevated methylation in stage I-II tumors (see AB). Figure 3 DE). ROC curves and hierarchical clustering showed that each biomarker could distinguish early gastric cancer from normal controls (see DE). Figure 3 FH).

[0058] Example 2: Establishment of a GC-mqMSP multimethylation detection system Based on the four DMPs (Cg26556436, Cg02063166, Cg05661282, and Cg03234186) screened in Implementation 1, five methylation probes with different fluorescent labels were designed to target the methylation sites of Cg26556436, Cg02063166, Cg05661282, and Cg03234186, as well as the β-actin internal reference gene. Primer and probe sequences are shown below. Figure 10 As shown. Cg05661282 was labeled with 5'6-FAM and 3'BHQ1; Cg03234186 was labeled with 5'TET and 3'BHQ1; Cg26556436 was labeled with 5'VIC and 3'BHQ1; Cg02063166 was labeled with 5'CY3 and 3'BHQ2; and β-actin was labeled with 5'CY5 and 3'BHQ3.

[0059] To verify the specificity of the primers and probes described above, cfDNA methylation sequencing analysis was performed on independent training and validation sets according to the experimental method described in Example 1. Additionally, during MS-qPCR, an additional reaction system was added, in which the above five primer-probe pairs were added, allowing for multiplex quantification of four methylation markers simultaneously in a single reaction system.

[0060] The CT values ​​obtained from qPCR were processed using the Delta Cycle Threshold (DCT) = (Target CT - Internal Reference CT) / (45 - Internal Reference CT). Data statistics were performed using Python 3.8. Logistic Regression analysis was conducted using the Logistic Regression model from the scikit-learn library (version 1.0.2), followed by a sigmoid transformation to obtain the GC-mqMSP index, calculated using the following formula: Where p is the GC-mqMSP exponent; x i Let a be the deita-CT value of the i-th marker; i is the coefficient of the i-th DMPs (see Table 1); b is the cutoff value (0.82645).

[0061] The results are as follows Figure 4 As shown in Figure A, the methylation level in the gastric cancer group was significantly higher than that in the control group. The AUC (0.933) of the quadruple-labeled GC-mqMSP system was significantly higher than that of each individual label (0.608–0.873) (see Figure A). Figure 4 B). GC-mqMSP demonstrated robust performance in terms of AUC (0.932 and 0.927), sensitivity (90.54% and 84.38%), and specificity (83.67% and 90.48%) on the training and validation sets, respectively. Figure 4 CD. The optimal cutoff value determined using the Youden index was 0.658, and the GC-mqMSP classification was highly consistent with the pathological diagnosis. Figure 4 CD).

[0062] Table 1. Coefficients of the four DMPs Example 3: Performance Verification of the GC-mqMSP System To evaluate the performance of the GC-mqMSP detection method in early diagnosis, an independent validation cohort was established, including 45 patients with gastric cancer and 13 healthy controls. Previous studies have shown that CDO1 is an important biomarker for the diagnosis, prognosis prediction, and metastasis assessment of gastric cancer (Cancer-specific promoter DNA methylation of Cysteine ​​dioxygenase type 1 (CDO1) gene as an important prognostic biomarker of gastric cancer[J]. PLoS One. 2019;14(4):e0214872.). Therefore, CDO1 was selected as a control indicator in this study. CDO1 methylation score and GC-mqMSP index were detected simultaneously in this cohort.

[0063] Test results as follows Figure 5As shown, although the overall diagnostic efficacy of GC-mqMSP is comparable to that of CDO1 (AUC = 0.899, 95% CI: 0.774-1.000; AUC = 0.884, 95% CI: 0.773-0.995 for CDO1), it demonstrates a key advantage when specifically analyzing early-stage tumors: for the diagnostic accuracy of stage I gastric cancer, the GC-mqMSP detection method is significantly superior to CDO1 (AUC = 0.839, 95% CI: 0.669-1.000; AUC = 0.762, 95% CI: 0.565-0.959), highlighting its enhanced ability in early lesion detection.

[0064] Furthermore, this invention also evaluated the specificity of the GC-mqMSP system in 64 other cancer patients (23 with colorectal cancer and 41 with lung cancer). The results are as follows... Figure 4 As shown in E, GC-mqMSP effectively distinguished between gastric cancer and non-gastric cancer malignancies, with an AUC of 0.912. This experimental result indicates that the methylation sites Cg26556436, Cg02063166, Cg05661282, and Cg03234186 have low specificity for other cancer types.

[0065] The above results confirm that the combination of biomarkers screened by this invention, namely the four methylation sites Cg26556436, Cg02063166, Cg05661282 and Cg03234186, has unique advantages in early diagnosis.

Claims

1. A biomarker combination for detecting early gastric cancer, the biomarker combination comprising a Cg26556436 methylation site located in the chr6_88875621_88875622 fragment, a Cg02063166 methylation site located in the chr6_88875650_88875651 fragment, a Cg05661282 methylation site located in the chr19_58220369_58220370 fragment, and a Cg03234186 methylation site located in the chr19_58220656_58220657 fragment; When the GC-mqMSP index of the four methylation sites Cg26556436, Cg02063166, Cg05661282, and Cg03234186 in the subject's biological sample is higher than 0.658, the subject is judged to have early gastric cancer. The formula for calculating the GC-mqMSP index is as follows: in, p is the GC-mqMSP exponent; x i For the i-th marker deitaCT; a i is the coefficient of the i-th marker; b is the cutoff value.

2. The combination of biomarkers as described in claim 1, wherein gastric cancer includes adenocarcinoma, signet ring cell carcinoma, adenosquamous carcinoma, and undifferentiated cell carcinoma, and wherein the early gastric cancer includes stage I gastric cancer and stage II gastric cancer.

3. A GC-mqMSP multiple methylation detection system for detecting early gastric cancer, the detection system comprising: The data input module, data processing module, and result output module are included. The data input module refers to inputting the methylation data of four methylation sites, Cg26556436, Cg02063166, Cg05661282 and Cg03234186, in the subject's biological sample; The data processing module performs logistic regression analysis on the methylation data input from the data input module to calculate the GC-mqMSP index. The calculation formula is as follows: Where p is the GC-mqMSP exponent; x i For the i-th marker deitaCT; a i Here, b is the coefficient of the i-th marker; b is the cutoff value. The result output module outputs the GC-mqMSP index obtained above. When the GC-mqMSP index output by the result output module is greater than 0.658, the subject is judged to have early gastric cancer.

4. The system of claim 3, wherein the biological sample is selected from tissue samples and blood samples of the subject.

5. The system of claim 3, wherein the method for acquiring the methylation data is selected from one of bisulfite sequencing PCR, methylation-specific PCR, high-resolution melting curve analysis, or digital PCR.

6. The use of a reagent for detecting the combination of biomarkers according to claim 1 in the preparation of a reagent for diagnosing early gastric cancer.

7. The application as described in claim 6, wherein the biomarker combination comprises four methylation sites: Cg26556436, Cg02063166, Cg05661282, and Cg03234186.

8. The application as described in claim 6, wherein the gastric cancer includes adenocarcinoma, signet ring cell carcinoma, adenosquamous carcinoma, and undifferentiated cell carcinoma, and the early gastric cancer includes stage I gastric cancer and stage II gastric cancer.

Images