Gene marker combination, kit and system for evaluating risk of lymph node metastasis of breast cancer

Abstract

The invention discloses a gene marker combination, a kit and a system for breast cancer lymph node metastasis risk assessment, and belongs to the field of precise medical big data analysis. The gene marker combination is composed of 18 genes, namely ACE2, BPIFB1, CLDN19, CT83, FABP7, FGFBP1, KLK7, KLK8, OCA2, PGLYRP4, PI3, PICSAR, PRSS33, ROPN1, ROPN1B, SCRG1, SERPINA11 and SRARP in total. Based on expression data of the combination, a brand new breast cancer lymph node metastasis risk assessment system (MRA18) is constructed, and breast cancer samples can be divided into eight molecular subtypes with different lymph node metastasis risks. Experiments show that the typing method has extremely strong category distinguishing ability (macro AUC score: 0.981 + / -0.004) and steady comprehensive classification performance (balance accuracy: 0.769 + / -0.027), and survival curves and lymph node metastasis risk grade distribution between different molecular subtypes have significant differences. The invention further provides a corresponding detection kit and an intelligent metastasis risk assessment system integrated with the random forest classification model, and an effective auxiliary diagnosis tool is provided for accurate assessment of the preoperative lymph node metastasis risk of the breast cancer patient and formulation of an individualized operation scheme.

Description

Technical Field

[0001] This invention relates to the field of precision medicine big data analysis, specifically to a combination of gene biomarkers for assessing the risk of lymph node metastasis in breast cancer, a kit for detecting the expression level of the combination, and a system for analyzing the expression data of the combination. Background Technology

[0002] Breast cancer is a highly molecularly heterogeneous malignant tumor, and assessing the risk of lymph node metastasis is a crucial prerequisite for accurate diagnosis and treatment. Currently, methods such as preoperative imaging and intraoperative sentinel lymph node biopsy are primarily used to determine whether lymph node dissection is necessary and the extent of dissection. However, neither of these methods can detect micrometastases in lymph nodes, resulting in a high false-negative rate. Therefore, there is an urgent clinical need for a method that can accurately assess the risk of lymph node metastasis in breast cancer at the molecular level using tissue samples before surgery. This would overcome the limitations of existing technologies and provide a more reliable basis for developing personalized surgical plans. Summary of the Invention

[0003] This invention provides a complete solution from biomarker discovery and molecular subtype definition to clinical testing applications. Its core lies in the first-ever proposal and construction of a biomarker combination consisting of 18 genes and the definition of 8 breast cancer molecular subtypes. Based on this, corresponding testing kits and an intelligent risk assessment system have been developed for assessing the risk of preoperative lymph node metastasis in breast cancer patients.

[0004] In a first aspect, this invention provides a set of gene biomarkers for assessing the risk of lymph node metastasis in breast cancer, characterized in that the set consists of the following 18 gene biomarkers: ACE2, BPIFB1, CLDN19, CT83, FABP7, FGFBP1, KLK7, KLK8, OCA2, PGLYRP4, PI3, PICSAR, PRSS33, ROPN1, ROPN1B, SCRG1, SERPINA11, and SRARP. The MRA18 (18-Gene Metastatic Risk Assessment) metastasis risk assessment system of this invention refers to classifying breast cancer samples into eight molecular subtypes with different lymph node metastasis risks based on the expression data of the aforementioned 18 gene biomarkers and using a specific classification model.

[0005] In a second aspect, the present invention provides a detection kit for assessing the risk of lymph node metastasis in breast cancer, comprising reagents for detecting the expression levels of the 18 gene markers described in the first aspect.

[0006] In a third aspect, the present invention provides a system for assessing the risk of lymph node metastasis in breast cancer, characterized by comprising the following modules: (1) a data import and quality control module, used to acquire the expression data of the 18 gene markers in the sample to be tested; (2) an intelligent risk assessment calculation module, which embeds a trained machine learning classification model, wherein the model takes the expression data of the 18 genes as the sole input and automatically calculates the probability distribution of the risk level of lymph node metastasis within the 8 molecular subtypes defined in the present invention and the risk level of lymph node metastasis within the subtype; and (3) a result generation and reporting module, used to report the classification results and related interpretation information.

[0007] In a fourth aspect, the present invention provides the application of the above-mentioned combination of gene markers, kits, and systems in the preparation of products for assessing the risk of lymph node metastasis in breast cancer.

[0008] In a fifth aspect, the present invention provides an electronic device and a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, enables the implementation of the functions of the lymph node metastasis risk assessment system described above. Beneficial effects

[0009] Compared with existing technologies, the beneficial effects of this invention are as follows: (1) Precise analysis of the molecular mechanisms driving lymph node metastasis in breast cancer: This invention, for the first time, defines a novel breast cancer lymph node metastasis risk assessment system, MRA18, based on a carefully selected combination of 18 gene markers, accurately classifying breast cancer into 8 molecular subtypes with significantly different expression patterns. DAVID functional clustering analysis of this combination of 18 gene markers shows that these genes mainly participate in the regulation and remodeling of the tumor microenvironment, confirming their high correlation with lymph node metastasis risk from a biological functional perspective. For example... Figure 1 As shown, Kaplan-Meier survival curve analysis revealed significant differences in the survival curves of these molecular subtypes (Multivariate log-rank test: p<0.05). Among them, samples with a high risk of lymph node metastasis accounted for 21.0% in the R1C1_R2C1_R3C1 molecular subtype and 6.3% in the R1C2_R2C2_R3C3 molecular subtype, with a difference of up to 3.33 times. (2) Molecular subtyping is highly correlated with the risk of lymph node metastasis: such as Figure 2As shown, this invention clearly reveals and confirms that there are significant differences in the distribution of lymph node metastasis risk levels among different molecular subtypes (Chi-square test: p<0.05), indicating that the MRA18 classification system can analyze molecular subtypes with different lymph node metastasis risks at the molecular level. (3) The combination of gene markers is concise, efficient, and easy to translate into clinical applications: This invention screens 18 core genes from a large number of gene markers. This combination can efficiently distinguish different molecular subtypes while greatly reducing the cost of detection and analysis, and has good prospects for industrialization and clinical translation. (4) Provides an end-to-end solution: This invention connects the entire process from biomarker discovery and verification to the development of detection products and intelligent analysis systems, providing a high-performance and accessible auxiliary diagnostic solution for the precise molecular diagnosis of breast cancer. Attached Figure Description

[0010] Figure 1 The Kaplan-Meier survival curves between molecular subtypes are shown in this embodiment of the invention. The survival curves of different molecular subtypes show significant differences (p<0.05), indicating that the MRA18 typing system has value in assessing the risk of lymph node metastasis.

[0011] Figure 2 This embodiment of the invention illustrates a comparison of the distribution of lymph node metastasis risk levels among molecular subtypes. The distribution of lymph node metastasis risk levels among different molecular subtypes shows significant differences (p<0.05), confirming that molecular subtypes are highly correlated with the risk of lymph node metastasis.

[0012] Figure 3 This embodiment of the invention demonstrates the z-score heatmap of the branching genes driving the K-Means clustering splitting process. The expression differences of the branching genes that each clustering split depends on are very significant among the split sub-clusters.

[0013] Figure 4 The t-SNE visualization comparison diagram between molecular subtypes is shown in this embodiment of the invention. The high aggregation within the molecular subtype indicates that the molecular characteristics of the samples within the molecular subtype are highly similar. Detailed Implementation

[0014] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. The following embodiments are only used to explain the present invention and do not constitute a limitation on the scope of protection of the present invention. Example 1

[0015] Discovery of core gene biomarkers and establishment of a new lymph node metastasis risk assessment system: (1) Data preparation: Publicly available multi-omics data of 875 breast cancer samples were obtained from the TCGA database, including mRNA expression profiles, DNA methylation profiles and microRNA expression profiles. Publicly available multi-omics data of 1073 breast cancer samples were obtained from the UCSC Xena platform, including mRNA expression profiles and clinical data. (2) Feature screening: Multi-omics data integration and classification method based on hierarchical attention (see application number: 202511477493.2) was used to perform multi-omics integration analysis on 875 breast cancer samples and screen out the 49 most important mRNA features. (3) Unsupervised clustering and definition of new molecular subtypes: Using the above 49 mRNA features, multi-level K-Means clustering analysis was performed on 1073 breast cancer samples. By analyzing the core differentially expressed genes (i.e., "branching genes") that drive the splitting of each layer of clusters, and taking the union of branching genes at all levels, the minimum branching gene set consisting of 18 genes was finally determined, which is used to define the entire MRA18 subtyping system. This ensures that it plays a key role in distinguishing all 8 molecular subtypes. The biological function of this gene set mainly involves the regulation and remodeling of the tumor microenvironment, providing important information for elucidating the molecular mechanism driving lymph node metastasis in breast cancer. (4) Analysis of molecular characteristics of new subtypes: such as Figure 3As shown, the eight newly defined molecular subtypes show significant differences in the expression patterns of 18 genes. First, based on the expression differences of CT83, OCA2, PICSAR, ROPN1, and ROPN1B genes, 1073 breast cancer samples were divided into two clusters: R1C1 (n=866) and R1C2 (n=207). Then, based on the expression differences of BPIFB1, SERPINA11, and SRARP genes, the R1C1 cluster was further divided into two clusters: R1C1_R2C1 (n=338) and R1C1_R2C2 (n=528). Based on the expression differences of PRSS33, ROPN1, ROPN1B, and SCRG1 genes, the R1C2 cluster was further divided into two clusters: R1C2_R2C1 (n=93) and R1C2_R2C2 (n=114). Finally, based on the expression differences of FABP7, FGFBP1, and PI3 genes, the R1C1_R2C1 cluster was further divided into two clusters. The data were divided into two clusters: R1C1_R2C1_R3C1 (n=236) and R1C1_R2C1_R3C2 (n=102). Based on the differential expression of CLDN19, FABP7, KLK7, KLK8, and ROPN1 genes, the R1C1_R2C2 cluster was further divided into R1C1_R2C2_R3C1 (n=144) and R1C1_R2C2_R3C2 (n=102). Based on the expression differences of ACE2, FABP7, FGFBP1, PGLYRP4, PI3, ROPN1, and ROPN1B genes, the R1C2_R2C2 cluster was divided into three clusters: R1C2_R2C2_R3C1 (n=44), R1C2_R2C2_R3C2 (n=54), and R1C2_R2C2_R3C3 (n=16). After three rounds of K-Means clustering analysis, eight molecular subtypes of breast cancer were finally defined, namely: R1C1_R2C1_R3C1, R1C1_R2C1_R3C2, R1C1_R2C2_R3C1, R1C1_R2C2_R3C2, R1C2_R2C1, R1C2_R2C2_R3C1, R1C2_R2C2_R3C2, and R1C2_R2C2_R3C3. (5) Visualization of the new subtypes using t-SNE: such as Figure 4 As shown, t-SNE dimensionality reduction visualization indicates that samples belonging to the same molecular subtype are tightly clustered in two-dimensional space, intuitively verifying that this MRA18 typing system can identify breast cancer subgroups with similar molecular characteristics. Example 2

[0016] Performance verification of the new subtyping method: (1) Model construction: Using only the expression data of the 18 genes determined in Example 1 as input features, a random forest classifier was constructed, with the MRA18 molecular subtype labels (8 classes) defined in this invention as the prediction target. (2) Stability evaluation: To comprehensively evaluate the model stability, the dataset was split into a training set (60%), a validation set (20%), and a test set (20%) using 10 different random seeds, while maintaining a consistent proportion of each molecular subtype in different subsets. For each data split, the model was initialized for training using 10 different random seeds. Finally, statistical analysis was performed based on the results of 100 independent experiments. (3) Stability assessment results: The MRA18 classification model showed strong class discrimination ability on the validation set (macro AUC score: 0.981 ± 0.004). More importantly, in the 8-class classification task with extremely imbalanced sample size, the model maintained robust comprehensive classification ability (balanced accuracy: 0.769 ± 0.027), proving that it can fairly identify each molecular subtype, rather than only favoring the molecular subtype with large sample size. Example 3

[0017] Construction of the detection kit (taking qPCR as an example): (1) Primer and probe design: Based on the specific sequences of the 18 core gene markers of this invention, specific primers and TaqMan hydrolysis probes were designed and synthesized using professional software. All sequences were verified by bioinformatics comparison to ensure their specificity and avoid cross-reaction. (2) The kit mainly includes the following components: RNA reverse transcription premix, qPCR reaction premix, specific primer and probe premix for 18 target genes and 3 internal reference genes, positive control and negative control. (3) The detection process is as follows: First, total RNA is extracted from fresh frozen tissue or paraffin-embedded tissue of breast cancer. After passing the quality assessment, it is reverse transcribed into cDNA. Then, the cDNA is mixed with the qPCR reaction premix and qPCR amplification is performed. By analyzing the Ct value of the target gene and correcting it with the internal reference gene, the relative expression data is obtained. Finally, the obtained data is used as the input of the evaluation system in Example 4. Example 4

[0018] Implementation of the Lymph Node Metastasis Risk Assessment System: Develop a software system for assessing the risk of lymph node metastasis in breast cancer, and realize automated analysis from data to report. The core functional modules include: (1) Data import and quality control module: Supports manual input or batch import of expression data of 18 genes obtained by the kit in Example 3, and performs data integrity and rationality checks. (2) Intelligent risk assessment calculation module: Embedded with a pre-trained high-performance random forest classification model as described in Example 2. This module receives the standardized 18-dimensional expression data vector as the only input, and automatically calculates the probability distribution of the sample belonging to the 8 molecular subtypes defined in this invention and the lymph node metastasis risk level within the subtype. (3) Result generation and reporting module: Used to report the classification results and related interpretation information, and generate a structured Chinese test report. The system can be packaged as an independent desktop software and installed on the local server of the hospital pathology department or a third-party independent medical testing laboratory, or it can be deployed on a private cloud platform that complies with medical data security specifications, and provide online analysis services through a browser or API interface. Clinical application value

[0019] Hospital pathology departments or independent third-party medical testing laboratories can use the kits of this invention to detect the expression data of 18 genes in breast cancer patient tissue samples. Using the accompanying lymph node metastasis risk assessment system, they can obtain the MRA18 subtype classification results and the probability distribution of lymph node metastasis risk levels for each subtype. These results can provide clinicians with molecular typing results that are unavailable through traditional imaging and sentinel lymph node biopsy techniques, assisting them in developing more effective lymph node dissection surgical plans and ultimately potentially improving patient prognosis. This invention features a small number of biomarkers, mature detection technology, convenient model deployment, and automated analysis processes, demonstrating clear prospects for industrialization and clinical translation.

[0020] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A combination of gene biomarkers for assessing the risk of lymph node metastasis in breast cancer, characterized in that, It consists of 18 gene markers: ACE2, BPIFB1, CLDN19, CT83, FABP7, FGFBP1, KLK7, KLK8, OCA2, PGLYRP4, PI3, PICSAR, PRSS33, ROPN1, ROPN1B, SCRG1, SERPINA11, and SRARP. The expression data of this combination is the feature set necessary to realize the MRA18 typing system defined in this invention.

2. A diagnostic kit for assessing the risk of lymph node metastasis in breast cancer, characterized in that, The kit contains reagents for detecting the expression levels of the 18 gene biomarkers of claim 1, and is configured to detect the expression levels of the 18 gene biomarkers.

3. The detection kit according to claim 2, characterized in that, The kit is based on nucleic acid detection technology, including but not limited to quantitative real-time PCR or high-throughput sequencing technology.

4. A breast cancer lymph node metastasis risk assessment system, characterized in that, The system includes the following modules: (1) a data import and quality control module, used to receive expression data of the 18 gene markers described in claim 1; (2) an intelligent risk assessment calculation module, which has an embedded classification model that uses the expression data of the 18 genes as a necessary input feature, and automatically calculates the probability distribution of the risk level of lymph node metastasis within the 8 molecular subtypes defined in this invention and the risk level of lymph node metastasis within the subtype; and (3) a result generation and reporting module, used to report the classification results and related interpretation information.

5. The breast cancer lymph node metastasis risk assessment system according to claim 4, characterized in that, The classification model is a machine learning model, including but not limited to random forest models or support vector machines.

6. The application of the gene marker combination according to claim 1 in the risk assessment of lymph node metastasis in breast cancer.

7. The application of the detection kit according to claim 2 or 3 in the risk assessment of lymph node metastasis in breast cancer.

8. The application of the risk assessment system described in claim 4 or 5 in the risk assessment of lymph node metastasis in breast cancer.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it performs the functions of the breast cancer lymph node metastasis risk assessment system according to any one of claims 4 or 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it performs the functions of the breast cancer lymph node metastasis risk assessment system according to any one of claims 4 or 5.

Images