A multigene model for evaluating CD8-positive T cell anti-tumor immunity in breast cancer patients and its construction method

Abstract

The present invention discloses a multi-gene model for evaluating CD8-positive T-cell anti-tumor immunity in breast cancer patients and a method for constructing the same. The multi-gene model achieves quantitative scoring of anti-tumor immunity through multiple dimensions such as tumor recognition, cytotoxicity, IFN-gama secretion, proliferation and apoptosis. And after multi-angle verification from clinical and database perspectives, it has been confirmed that it has a good predictive effect on the prognosis of breast cancer patients, and its accuracy is higher than other indicators for predicting the efficacy of breast cancer immunotherapy, such as TIS, GeparSixto and CD8-positive T-cell infiltration levels.

Description

Technical Field

[0001] The present invention relates to the technical field of biological model construction, and in particular to a multi-gene model for evaluating CD8 positive T cell anti-tumor immunity of breast cancer patients and a construction method thereof. Background Art

[0002] The immune cold / hot classification of tumors is a key factor affecting the prognosis of tumor patients and the efficacy of immunotherapy. In theory, patients with immune-hot tumors have better anti-tumor ability and are more likely to benefit from immunotherapy. However, in current clinical practice, the immune cold / hot classification of tumors is determined by the infiltration level of lymphocytes. There are both lymphocytes that exert anti-tumor immunity and lymphocytes that suppress immunity in the tumor microenvironment. Therefore, the level of lymphocyte infiltration as a standard cannot actually reflect the patient's true tumor immune cold / hot situation. However, there are currently no other available diagnostic tools for accurately assessing the immune cold / hot situation of tumors in clinical practice.

[0003] CD8-positive T cells play a key role in anti-tumor immunity, and accurate assessment of their anti-tumor ability can largely reflect the immune cold / hot of the tumor. However, some previous evaluation methods, such as CD8-positive T cell abundance, cytotoxicity score, and interferon γ pathway score, were unable to accurately evaluate the anti-tumor immunity of CD8-positive T cells, resulting in poor performance in predicting the prognosis of breast cancer and the efficacy of immunotherapy. Anti-tumor immunity mediated by CD8-positive T cells is closely related to their ability to recognize tumors, kill cells, secrete immune factors, proliferate, and resist death. Only by integrating multiple dimensions of function can the anti-tumor immunity of CD8-positive T cells be more comprehensively and accurately evaluated. However, there is currently no evaluation tool that can integrate multiple CD8-positive T cell functional dimensions to quantify their anti-tumor immunity, so as to better predict the prognosis of breast cancer and the efficacy of immunotherapy in clinical practice.

[0004] In recent years, a variety of high-precision microenvironment analysis technologies have been developed and used, such as single-cell sequencing, which can more accurately evaluate the functions of microenvironment components. However, these technologies are expensive and not suitable for large-scale clinical practice. Based on the actual clinical situation, the development of a lower-cost, high-throughput and simple-to-operate multi-gene model for evaluating CD8-positive T cell anti-tumor immunity in breast cancer patients is of great significance for breast cancer risk prediction, immunotherapy efficacy and prognosis evaluation. Summary of the invention

[0005] The present invention aims to solve at least one of the above-mentioned technical problems existing in the prior art. To this end, the purpose of the present invention is to provide a multi-gene model for evaluating the anti-tumor immunity of CD8-positive T cells in breast cancer patients and a method for constructing the same. The multi-gene model in the present invention can accurately evaluate the five functional dimensions of CD8-positive T cells in the microenvironment, including tumor recognition, tumor killing, immune factor secretion, proliferation, and resistance to death (apoptosis), from multiple dimensions, and calculate a score by weighted integration to quantify the overall anti-tumor immunity of CD8-positive T cells, so as to more accurately predict the prognosis of breast cancer and the efficacy of immunotherapy in clinical practice.

[0006] The first aspect of the present invention provides a method for constructing a multi-gene CD8 positive T cell anti-tumor immune model, comprising the following steps:

[0007] (1) Extracting nucleic acid from the test sample, quantitatively detecting the expression of genes related to tumor recognition, cytotoxicity, IFN-gama secretion, proliferation and apoptosis, and calculating the scores of tumor recognition, cytotoxicity, IFN-gama secretion, proliferation and apoptosis based on the expression;

[0008] (2) Based on the scores of tumor recognition, cytotoxicity, IFN-gama secretion, proliferation, and apoptosis, an anti-tumor immunity scoring formula was constructed to obtain a multi-gene CD8-positive T cell anti-tumor immunity model;

[0009] Among them, for early breast cancer:

[0010] The anti-tumor immunity score is: 1×tumor recognition score + 0.2×cytotoxicity score + 0.2×IFN-gama secretion score + 0.1×proliferation score + apoptosis score × 1.

[0011] For advanced breast cancer:

[0012] The anti-tumor immunity score is: 0.3×tumor recognition score + 0.8×cytotoxicity score + 0.7×IFN-gama secretion score + 0.1×proliferation score + apoptosis score × 0.9.

[0013] The early breast cancer is clinical stage I-III breast cancer, and the late breast cancer is clinical stage IV breast cancer.

[0014] In some embodiments of the present invention, tumor recognition-related genes are composed of positively correlated genes and negatively correlated genes, wherein the positively correlated genes include HLA-DQB1, RNF19A, C12orf75, B4GALT5, OPTN, R3HDM1, UBD, MARCO, CXCR6, KHDRBS1, CD8B, RUNX3, ZFP69B, CCDC65 and GBP1; the negatively correlated genes include COA3, TMEM54, SCNN1A, RHOD, ADIRF and USP27X-AS1.

[0015] In some embodiments of the present invention, the cytotoxicity-related genes are composed of positively correlated genes and negatively correlated genes, wherein the positively correlated genes include HLA-DQB1, RNF19A, C12orf75, FASLG, ZNF683, CST1, GRAP2, GBP2, FNBP1, GZMA, GRSF1, IKZF4 and CCDC59; the negatively correlated genes include PRSS22, MPC2, NT5C3B, JPH1 and USP27X-AS1.

[0016] In some embodiments of the present invention, IFN-gama secretion-related genes are composed of positively correlated genes and negatively correlated genes, wherein the positively correlated genes include HLA-DQB1, B4GALT5, OPTN, FASLG, ZNF683, CST1, CD8A, HLA-DQA1, GPA33, MIIP, CCR5, FPR3, UXS1, LAG3 and IL15; the negatively correlated genes include CELSR2 and USP27X-AS1.

[0017] In some embodiments of the present invention, the growth-related genes are composed of positively correlated genes and negatively correlated genes, wherein the positively correlated genes include RNF19A, B4GALT5, OPTN, R3HDM1, UBD, CD8A, HLA-DQA1, and YME1L1; and the negatively correlated genes include NT5C3B, JPH1, and ADIRF.

[0018] In some embodiments of the present invention, apoptosis is composed of positively correlated genes and negatively correlated genes, wherein the positively correlated genes include HLA-DQB1, B4GALT5, OPTN, MARCO, CD8A, GPA33, CD70, GZMB, PLCG1, HSPA8, AKIRIN1 and JSRP1; and the negatively correlated genes include AGR3, TRIM45, KLC4, MED20 and MPC2.

[0019] In some embodiments of the present invention, the scores of the tumor recognition, cytotoxicity, IFN-gama secretion, proliferation and apoptosis dimensions are calculated as follows:

[0020] Score = average expression of positively correlated genes of the dimension - average expression of negatively correlated genes.

[0021] In some embodiments of the present invention, the scores of each dimension are specifically:

[0022] Tumor recognition score = (expression level of HLA-DQB1+RNF19A+C12orf75+B4GALT5+OPTN+R3HDM1+UBD+MARCO+CXCR6+KHDRBS1+CD8B+RUNX3+ZFP69B+CCDC65+GBP1) / 15-(expression level of COA3+TMEM54+SCNN1A+RHOD+ADIRF+USP27X-AS1) / 6;

[0023] Cytotoxicity score = (expression level of HLA-DQB1+RNF19A+C12orf75+FASLG+ZNF683+CST1+GRAP2+GBP2+FNBP1+GZMA+GRSF1+IKZF4+CCDC59) / 13-(expression level of PRSS22+MPC2+NT5C3B+JPH1+USP27X-AS1) / 5;

[0024] IFN-gama secretion score = (expression level of HLA-DQB1+B4GALT5+OPTN+FASLG+ZNF683+CST1+CD8A+HLA-DQA1+GPA33+MIIP+CCR5+FPR3+UXS1+LAG3+IL15) / 15-(expression level of CELSR2+USP27X-AS1) / 2;

[0025] Proliferation score = (expression level of RNF19A+B4GALT5+OPTN+R3HDM1+UBD+CD8A+HLA-DQA1+YME1L1) / 8-(expression level of NT5C3B+JPH1+ADIRF) / 3;

[0026] Apoptosis score = (expression level of HLA-DQB1+B4GALT5+OPTN+MARCO+CD8A+GPA33+CD70+GZMB+PLCG1+HSPA8+AKIRIN1+JSRP1) / 12-(expression level of AGR3+TRIM45+KLC4+MED20+MPC2) / 5.

[0027] The second aspect of the present invention provides a breast cancer prognosis and / or treatment effect evaluation product, which includes a gene expression quantity quantification module and an analysis module; the analysis module contains the calculation algorithm and anti-tumor immunity scoring algorithm for the scores of tumor recognition, cytotoxicity, IFN-gama secretion, proliferation and apoptosis dimensions in the first aspect of the present invention.

[0028] In some embodiments of the present invention, the gene expression quantity quantification module is used to quantify the expression quantity of the target gene in the sample to be tested.

[0029] In some embodiments of the present invention, the gene expression quantity quantification module includes conventional quantification methods and products in the art.

[0030] In some embodiments of the present invention, the gene expression quantitative module includes but is not limited to quantitative PCR, fluorescence in situ hybridization (FISH), sequencing and real-time fluorescence quantitative PCR (qPCR).

[0031] In some embodiments of the present invention, the gene expression quantity quantification module includes but is not limited to commercially available detection kits, detection reagents, and detection chips.

[0032] In some embodiments of the present invention, the method of using the gene expression quantity quantification module is determined according to the specific method and product selected. For example, for specific methods, reference may be made to technical manuals in the field; for products, reference may be made to product instructions.

[0033] The third aspect of the present invention provides the use of the method for constructing the multi-gene CD8-positive T cell anti-tumor immune model described in the first aspect of the present invention in the prognosis and / or treatment effect evaluation of breast cancer.

[0034] In some embodiments of the present invention, the corresponding relationship between the breast cancer prognosis and / or treatment effect and the anti-tumor immunity score is:

[0035] When the anti-tumor immunity score is greater than or equal to -0.111, it is predicted that the sample is responsive to immunotherapy and has a good prognosis for breast cancer;

[0036] When the anti-tumor immunity score is less than -0.111, it is predicted that the sample has low or no response to immunotherapy and a poor prognosis for breast cancer.

[0037] In the present invention, the multi-gene model used is based on transcriptome sequencing / chip technology, so it is low-cost and has good repeatability, and can be effectively applied to clinical applications. Moreover, in the present invention, the multi-gene model is obtained by combining single-cell sequencing and transcriptome sequencing technology to achieve multi-angle training, which can effectively reflect the different functional dimensions of CD8-positive T cells, and can quantify CD8-positive T cell anti-tumor immunity after weighted integration, thereby more accurately predicting breast cancer prognosis and immunotherapy efficacy in clinical practice.

[0038] The beneficial effects of the present invention are:

[0039] 1. The multi-gene model of the present invention has a good predictive effect on the prognosis of breast cancer patients. It can effectively predict the prognosis of breast cancer patients in clinical examples and in two well-known independent large breast cancer cohorts (TCGA-BRCA and METABRIC); and it is superior to traditional lymphocyte infiltration, CD8-positive T cell infiltration and immune activation-related models.

[0040] 2. The polygenic model of the present invention can effectively predict the efficacy of immunotherapy for breast cancer patients, and in clinical examples, it shows that the overall anti-tumor immune score of CD8-positive T cells in patients who have obtained clinical benefits is significantly higher than that in patients who have not obtained clinical benefits. Moreover, in three other independent clinical trials, the overall anti-tumor immune score of CD8-positive T cells can effectively predict the benefit of immunotherapy for breast cancer patients, with AUCs of 0.822, 0.782, and 0.710, respectively, which is significantly better than other indicators for predicting the efficacy of immunotherapy for breast cancer, such as TIS, GeparSixto, and CD8-positive T cell infiltration levels. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 Immunofluorescence images of tumor recognition, cytotoxicity, and IFN-gama secretion.

[0042] Figure 2 Immunofluorescence images of proliferation and apoptosis.

[0043] Figure 3 This is a correlation analysis diagram between different functional dimension scores of CD8 positive T cells and the corresponding CD8 positive T cell functional dimensions evaluated by immunofluorescence.

[0044] Figure 4 This is a correlation analysis diagram of different functional dimension scores of CD8-positive T cells in the GSE176078 cohort and the corresponding CD8-positive T cell functional dimensions assessed by single-cell sequencing.

[0045] Figure 5This is a correlation analysis diagram of different functional dimension scores of CD8-positive T cells in the GSE161529 cohort and the corresponding CD8-positive T cell functional dimensions assessed by single-cell sequencing.

[0046] Figure 6 The relationship between the overall anti-tumor immunity score of CD8-positive T cells and prognosis of breast cancer patients in Sun Yat-sen Memorial Hospital, TCGA cohort and METABRIC cohort based on the method in the embodiment of the present invention.

[0047] Figure 7 This is a comparison chart of the ability of the overall anti-tumor immunity score of CD8-positive T cells obtained based on the method in the embodiment of the present invention for breast cancer patients in Sun Yat-sen Memorial Hospital, TCGA cohort and METABRIC cohort to predict prognosis with other immune-related markers.

[0048] Figure 8 This is the relationship between the overall anti-tumor immunity score of CD8 positive T cells and the efficacy of immunotherapy for breast cancer patients in Sun Yat-sen Memorial Hospital obtained based on the method in the embodiment of the present invention.

[0049] Fig. 9 This is a comparison chart of the overall anti-tumor immunity score of CD8-positive T cells obtained by the method in the embodiment of the present invention for the GSE194040 cohort and other immune-related markers in their ability to predict the efficacy of immunotherapy.

[0050] Fig.10 This is a comparison chart of the overall anti-tumor immunity score of CD8-positive T cells obtained by the method in the embodiment of the present invention for the GSE173839 cohort and other immune-related markers in terms of their ability to predict the efficacy of immunotherapy.

[0051] Fig.11 This is a comparison chart of the overall anti-tumor immunity score of CD8-positive T cells obtained by the method in the embodiment of the present invention for the PRJNA558949 cohort and other immune-related markers in their ability to predict the efficacy of immunotherapy. DETAILED DESCRIPTION

[0052] The present invention is further described in detail below by specific examples. Unless otherwise specified, the raw materials, reagents or devices used in the examples and comparative examples can be obtained from conventional commercial sources or can be obtained by prior art methods. Unless otherwise specified, the experiments or test methods are conventional methods in the art.

[0053] Example 1 Construction of a multigene model for evaluating CD8-positive T cell anti-tumor immunity in breast cancer patients

[0054] In this embodiment, a method for constructing a multigene model that can be used to evaluate the anti-tumor immunity of CD8-positive T cells in breast cancer patients is provided. The resulting multigene model is clinically used to evaluate the different functional dimensions of CD8-positive T cells in the breast cancer microenvironment, and to quantify the overall anti-tumor immunity of CD8-positive T cells, and to predict the prognosis of breast cancer patients and the efficacy of immunotherapy.

[0055] In this polygenic model, the following groups of functional dimension genes (including positively correlated genes and negatively correlated genes) are composed:

[0056] (1) Tumor recognition-related genes:

[0057] Positively correlated genes included HLA-DQB1, RNF19A, C12orf75, B4GALT5, OPTN, R3HDM1, UBD, MARCO, CXCR6, KHDRBS1, CD8B, RUNX3, ZFP69B, CCDC65, and GBP1;

[0058] Negatively correlated genes include COA3, TMEM54, SCNN1A, RHOD, ADIRF, and USP27X-AS1.

[0059] (2) Cell killing related genes:

[0060] Positively correlated genes included HLA-DQB1, RNF19A, C12orf75, FASLG, ZNF683, CST1, GRAP2, GBP2, FNBP1, GZMA, GRSF1, IKZF4, and CCDC59;

[0061] Negatively correlated genes include PRSS22, MPC2, NT5C3B, JPH1, and USP27X-AS1.

[0062] (3) Genes related to immune factor secretion:

[0063] Positively correlated genes included HLA-DQB1, B4GALT5, OPTN, FASLG, ZNF683, CST1, CD8A, HLA-DQA1, GPA33, MIIP, CCR5, FPR3, UXS1, LAG3, and IL15;

[0064] Negatively correlated genes include CELSR2 and USP27X-AS1.

[0065] (4) Proliferation-related genes:

[0066] Positively correlated genes included RNF19A, B4GALT5, OPTN, R3HDM1, UBD, CD8A, HLA-DQA1, and YME1L1;

[0067] Negatively correlated genes included NT5C3B, JPH1, and ADIRF.

[0068] (5) Apoptosis-related genes:

[0069] Positively correlated genes included HLA-DQB1, B4GALT5, OPTN, MARCO, CD8A, GPA33, CD70, GZMB, PLCG1, HSPA8, AKIRIN1, and JSRP1;

[0070] Negatively correlated genes include AGR3, TRIM45, KLC4, MED20, and MPC2.

[0071] Take the sample to be tested, and measure the expression of each of the above genes by conventional technical means in the art. Then calculate the polygenic model score according to the following method:

[0072] (1) Calculate the scores of different functional dimensions, that is, the average expression of positively correlated genes in the functional dimension minus the average expression of negatively correlated genes. In this embodiment, the specific calculation formula is as follows:

[0073] Tumor recognition score = (expression level of HLA-DQB1+RNF19A+C12orf75+B4GALT5+OPTN+R3HDM1+UBD+MARCO+CXCR6+KHDRBS1+CD8B+RUNX3+ZFP69B+CCDC65+GBP1) / 15-(expression level of COA3+TMEM54+SCNN1A+RHOD+ADIRF+USP27X-AS1) / 6;

[0074] Cytotoxicity score = (expression level of HLA-DQB1+RNF19A+C12orf75+FASLG+ZNF683+CST1+GRAP2+GBP2+FNBP1+GZMA+GRSF1+IKZF4+CCDC59) / 13-(expression level of PRSS22+MPC2+NT5C3B+JPH1+USP27X-AS1) / 5;

[0075] IFN-gama secretion score = (expression level of HLA-DQB1+B4GALT5+OPTN+FASLG+ZNF683+CST1+CD8A+HLA-DQA1+GPA33+MIIP+CCR5+FPR3+UXS1+LAG3+IL15) / 15-(expression level of CELSR2+USP27X-AS1) / 2;

[0076] Proliferation score = (expression level of RNF19A+B4GALT5+OPTN+R3HDM1+UBD+CD8A+HLA-DQA1+YME1L1) / 8-(expression level of NT5C3B+JPH1+ADIRF) / 3;

[0077] Apoptosis score = (expression level of HLA-DQB1+B4GALT5+OPTN+MARCO+CD8A+GPA33+CD70+GZMB+PLCG1+HSPA8+AKIRIN1+JSRP1) / 12-(expression level of AGR3+TRIM45+KLC4+MED20+MPC2) / 5.

[0078] (2) The overall anti-tumor immunity of CD8-positive T cells is integrated and quantified by weighting the scores of different functional dimensions, where the corresponding formula is as follows:

[0079] For early breast cancer (clinical stage I-III):

[0080] The anti-tumor immunity score is: 1×tumor recognition score + 0.2×cytotoxicity score + 0.2×immune factor secretion score + 0.1×proliferation score + apoptosis score × 1.

[0081] For advanced breast cancer (clinical stage IV):

[0082] The anti-tumor immunity score is: 0.3×tumor recognition score + 0.8×cytotoxicity score + 0.7×IFN-gama secretion score + 0.1×proliferation score + apoptosis score × 0.9.

[0083] Among them, when the anti-tumor immunity score is greater than or equal to -0.111, it is predicted that the sample is responsive to the efficacy of immunotherapy and has a good prognosis for breast cancer;

[0084] When the anti-tumor immunity score is less than -0.111, it is predicted that the sample has low or no response to immunotherapy and a poor prognosis for breast cancer.

[0085] The anti-tumor immune score achieved by the method in this embodiment is applicable to any type of sample, including surgically removed tissue samples or puncture biopsy tissue samples, and the detection molecule is a nucleic acid.

[0086] Example 2 Verification of the effect of the polygenic model in evaluating different functional dimensions of CD8 positive T cells

[0087] In this example, a retrospective study method was used to randomly select some primary breast cancer patients from Sun Yat-sen Memorial Hospital of Sun Yat-sen University as samples for validation testing. The inclusion criteria for primary breast cancer patients were: (1) primary breast cancer; (2) paired pathological sections and transcriptome sequencing data. The relevant steps and results are shown below.

[0088] (1) Sample collection and processing:

[0089] A total of 53 HLA-A2+ primary breast cancer tumor samples diagnosed with invasive breast cancer in Zhongshan Memorial Hospital, Sun Yat-sen University from 2017 to 2021 were collected as validation samples. All samples were from patients who provided informed consent, and all related procedures were approved by the Internal Review and Ethics Committee of Zhongshan Memorial Hospital.

[0090] After tumor tissue transcriptome sequencing, all samples were standardized using the hg38 reference genome as a reference and organized into a transcripts per million (TPM) expression matrix.

[0091] (2) Immunofluorescence: At the same time, pathological sections were prepared and co-stained with Tetramer, Granzyme B, IFN-gama, Ki67, TUNEL and CD8 to stain CD8-positive T cells with different functions (tumor recognition, cytotoxicity, IFN-gama secretion, proliferation and apoptosis) on the pathological sections. Then, senior doctors in the pathology department counted the CD8-positive T cells with strong activity in different functional dimensions to evaluate the strength of different functional dimensions of CD8-positive T cells in each patient (such as Figure 1 and Figure 2 shown).

[0092] (3) Calculating the scores of different functional dimensions of CD8 positive T cells according to the method based on the polygene model in the above-mentioned embodiment of the present invention.

[0093] (4) The different functional dimension scores of CD8-positive T cells (tumor recognition, cell killing, immune factor secretion, proliferation and apoptosis) were correlated with the corresponding CD8-positive T cell functional dimensions obtained by immunofluorescence evaluation.

[0094] The results are as follows Figure 3 As shown, the correlation coefficients obtained were as follows: tumor recognition: 0.74; cytotoxicity: 0.69; IFN-gama secretion: 0.74; proliferation: 0.65; apoptosis: 0.8.

[0095] (5) External validation: In order to illustrate the effectiveness of this validation, the inventors collected two paired single-cell sequencing and transcriptome sequencing databases, GSE161529 and GSE176078, from the database for external validation. The data of CD8-positive T cells were extracted from the single-cell sequencing data, and then the AUCell method was used to estimate the abundance of CD8-positive T cells with strong activity in different functional dimensions (tumor recognition, cytotoxicity, IFN-gama secretion, proliferation and apoptosis). Based on the transcriptome sequencing data, the multi-gene model method in the embodiment of the present invention was used to calculate the scores of different functional dimensions of CD8-positive T cells (tumor recognition, cell killing, immune factor secretion, proliferation and apoptosis), and the scores were correlated with the corresponding CD8-positive T cell functional dimensions evaluated by single-cell sequencing.

[0096] The results are as follows Figure 4 and Figure 5 The correlation coefficients obtained are as follows: tumor recognition: 0.74 and 0.75; cytotoxicity: 0.69 and 0.77; IFN-gama secretion: 0.74 and 0.66; proliferation: 0.65 and 0.73; apoptosis: 0.69 and 0.76. Among them, the former correlation coefficient is the result of GSE161529, and the latter is the result of GSE176078.

[0097] Example 3: Verification of the effect of the polygenic model in evaluating the prognosis of breast cancer

[0098] In this example, a retrospective study method was used to randomly select breast cancer patients from Sun Yat-sen Memorial Hospital of Sun Yat-sen University as samples for validation testing. The inclusion criteria for breast cancer patients were: (1) primary breast cancer; (2) transcriptome sequencing data were available. The exclusion criteria were: (1) missing survival information; (2) missing clinical pathology information. The relevant steps and results are shown below.

[0099] (1) Sample collection and processing:

[0100] A total of 170 primary breast cancer tumor samples were collected from patients diagnosed with invasive breast cancer at the Zhongshan Memorial Hospital affiliated to Sun Yat-sen University from 2017 to 2021. All samples were obtained from patients who had provided informed consent, and all related procedures were approved by the Internal Review and Ethics Committee of Zhongshan Memorial Hospital.

[0101] After tumor tissue transcriptome sequencing, all samples were standardized using the hg38 reference genome as a reference and organized into a transcripts per million (TPM) expression matrix.

[0102] (2) Calculating the scores of different functional dimensions of CD8 positive T cells according to the method based on the polygene model in the above-mentioned embodiment of the present invention.

[0103] (3) Calculation of the overall anti-tumor immunity score of CD8-positive T cells: Weighted integration of the scores of different functional dimensions of CD8-positive T cells was performed to quantify the overall anti-tumor immunity of CD8-positive T cells.

[0104] The weights of the scores for different functional dimensions are as follows:

[0105] Early breast cancer: 1 (tumor recognition), 0.2 (cell killing), 0.2 (IFN-gama secretion), 0.1 (proliferation), and 1 (apoptosis);

[0106] Advanced breast cancer: 0.3 (tumor recognition), 0.8 (cell killing), 0.7 (IFN-gama secretion), 0.1 (proliferation) and 0.9 (apoptosis).

[0107] (4) Survival analysis: Based on the weights obtained in step (3), the corresponding formula of the multi-gene model in the above embodiment was used to calculate the overall anti-tumor immunity score of CD8-positive T cells for each breast cancer patient. The patients were divided into three groups according to the tertiles: strong, medium, and weak overall anti-tumor immunity of CD8-positive T cells.

[0108] After correcting the molecular typing and staging according to the actual situation of breast cancer patients in our hospital, patients with high CD8-positive T cell overall anti-tumor immunity scores had better prognosis (high immunity score vs. low immunity score, HR = 0.41, 95% CI: 0.09-1.00, such as Figure 6 And the predictive ability is better than other biomarkers (such as Figure 7 shown).

[0109] (5) External validation: In order to illustrate the effectiveness of this validation, the inventors collected transcriptome sequencing data and survival data from the two breast cancer cohorts of TCGA and METABRIC from the database for external validation. The overall anti-tumor immunity score of CD8-positive T cells was calculated according to the above method and brought into the corresponding formula of the multi-gene model in the above embodiment to calculate the overall anti-tumor immunity score of CD8-positive T cells for each breast cancer patient. According to the tertiles, the patients were divided into three groups of strong, medium and weak overall anti-tumor immunity of CD8-positive T cells in the two databases.

[0110] The results showed that in the two breast cancer cohorts of TCGA and METABRIC, after correction of molecular typing and staging, patients with strong overall anti-tumor immune scores of CD8-positive T cells had better prognosis (TCGA: high immune score vs. low immune score, HR = 0.61, 95% CI: 0.40-0.93; METABRIC: high immune score vs. low immune score, HR = 0.71, 95% CI: 0.59-0.86, Figure 6 And the predictive ability is better than other biomarkers (such as Figure 7 shown).

[0111] Example 4 Validation of the effect of polygenic model in evaluating the efficacy of immunotherapy for breast cancer

[0112] In this example, a retrospective study method was used to randomly select breast cancer patients from Sun Yat-sen Memorial Hospital of Sun Yat-sen University as samples for validation testing. The inclusion criteria for breast cancer patients were: (1) breast cancer patients receiving immunotherapy; (2) with transcriptome sequencing data. The exclusion criteria were: (1) missing efficacy information; (2) missing clinical pathology information. The relevant steps and results are shown below.

[0113] (1) Sample collection and processing are the same as in the above embodiment.

[0114] (2) Calculating the scores of different functional dimensions of CD8 positive T cells according to the method based on the polygene model in the above-mentioned embodiment of the present invention.

[0115] (3) Calculation of the overall anti-tumor immunity score of CD8-positive T cells: Weighted integration of the scores of different functional dimensions of CD8-positive T cells was performed, that is, the overall anti-tumor immunity of CD8-positive T cells was quantified (the anti-tumor immunity score was obtained).

[0116] (4) To evaluate the correlation between the overall anti-tumor immunity score of CD8-positive T cells and the efficacy of immunotherapy: Three patients who benefited from immunotherapy and one patient who did not benefit from immunotherapy were selected as samples to analyze the differences in their overall anti-tumor immunity of CD8-positive T cells.

[0117] The results are as follows Figure 8 As shown, it can be seen that the overall anti-tumor immunity score of CD8-positive T cells in patients who achieved clinical benefit was significantly higher than that in patients who did not achieve clinical benefit.

[0118] (5) External validation: Three breast cancer immunotherapy databases containing transcriptome data (GSE194040, GSE173839, and PRJNA558949) were further collected for validation analysis.

[0119] The results show that the overall anti-tumor immunity of CD8 positive T cells obtained by the method in the embodiment of the present invention can significantly predict the efficacy of immunotherapy for breast cancer patients. The AUCs for the three groups of data are 0.822, 0.782 and 0.710 respectively (e.g. Figures 9-11 And its predictive ability is better than other biomarkers.

[0120] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A method for constructing a multi-gene CD8 positive T cell anti-tumor immune model, comprising the following steps: (1) Extracting nucleic acids from test samples, quantitatively detecting the expression of genes related to tumor recognition, cytotoxicity, IFN-gama secretion, proliferation, and apoptosis, and calculating the scores of tumor recognition, cytotoxicity, IFN-gama secretion, proliferation, and apoptosis based on the expression levels; (2) Based on the scores of tumor recognition, cytotoxicity, IFN-gama secretion, proliferation, and apoptosis, an anti-tumor immunity scoring formula was constructed to obtain a multi-gene CD8-positive T cell anti-tumor immunity model; in, For early breast cancer: The antitumor immunity score was: 1× tumor recognition score + 0.2× cytotoxicity score + 0.2× IFN-gama secretion score + 0.1× proliferation score + apoptosis score × 1; For advanced breast cancer: The antitumor immunity score is: 0.3×tumor recognition score + 0.8×cytotoxicity score + 0.7×IFN-gama secretion score + 0.1×proliferation score + apoptosis score × 0.9; Wherein, the early breast cancer is breast cancer of clinical stage I-III, and the late breast cancer is breast cancer of clinical stage IV; Among them, tumor recognition-related genes are composed of positively correlated genes and negatively correlated genes. Among them, the positively correlated genes included HLA-DQB1, RNF19A, C12orf75, B4GALT5, OPTN, R3HDM1, UBD, MARCO, CXCR6, KHDRBS1, CD8B, RUNX3, ZFP69B, CCDC65, and GBP1; Negatively correlated genes included COA3, TMEM54, SCNN1A, RHOD, ADIRF, and USP27X-AS1; cytotoxicity-related genes consisted of positively correlated genes and negatively correlated genes. Among them, the positively correlated genes included HLA-DQB1, RNF19A, C12orf75, FASLG, ZNF683, CST1, GRAP2, GBP2, FNBP1, GZMA, GRSF1, IKZF4, and CCDC59; Negatively correlated genes included PRSS22, MPC2, NT5C3B, JPH1, and USP27X-AS1; IFN-gama secretion-related genes consist of positively correlated genes and negatively correlated genes. Among them, the positively correlated genes included HLA-DQB1, B4GALT5, OPTN, FASLG, ZNF683, CST1, CD8A, HLA-DQA1, GPA33, MIIP, CCR5, FPR3, UXS1, LAG3, and IL15; Negatively correlated genes included CELSR2 and USP27X-AS1; Proliferation-related genes consist of positively correlated genes and negatively correlated genes. Among them, positively correlated genes included RNF19A, B4GALT5, OPTN, R3HDM1, UBD, CD8A, HLA-DQA1, and YME1L1; Negatively correlated genes included NT5C3B, JPH1, and ADIRF; Apoptosis-related genes consist of positively correlated genes and negatively correlated genes. Among them, the positively correlated genes included HLA-DQB1, B4GALT5, OPTN, MARCO, CD8A, GPA33, CD70, GZMB, PLCG1, HSPA8, AKIRIN1, and JSRP1; Negatively correlated genes included AGR3, TRIM45, KLC4, MED20, and MPC2; The calculation method for calculating the scores of tumor recognition, cytotoxicity, IFN-gama secretion, proliferation and apoptosis dimensions based on the expression amount is: The score of a dimension = the average expression level of the positively correlated genes of the dimension - the average expression level of the negatively correlated genes of the dimension.

2. A breast cancer prognosis and / or treatment effect evaluation product, characterized in that: The breast cancer prognosis and / or treatment effect evaluation product includes a gene expression quantity quantification module and an analysis module; the analysis module contains the calculation algorithm for calculating the scores of tumor recognition, cytotoxicity, IFN-gama secretion, proliferation and apoptosis dimensions based on the expression quantity in claim 1, and the anti-tumor immunity scoring algorithm in claim 1.

3. Application of a construction method in the evaluation of prognosis and / or treatment effect of breast cancer, characterized in that: The construction method is the construction method of the multi-gene CD8 positive T cell anti-tumor immunity model according to claim 1.

4. The use according to claim 3, characterized in that: The corresponding relationship between the breast cancer prognosis and / or treatment effect and the anti-tumor immunity score is: When the anti-tumor immunity score is greater than or equal to -0.111, it is predicted that the sample is responsive to immunotherapy and has a good prognosis for breast cancer; When the anti-tumor immunity score is less than -0.111, it is predicted that the sample has low or no response to immunotherapy and a poor prognosis for breast cancer.

Images