Biomarker for predicting treatment response to immunotherapy and gene prediction model using same

Abstract

The present invention relates to a biomarker for predicting a treatment response to immunotherapy and a gene prediction model using the same. The gene prediction model using the biomarker is based on the expression pattern of a main gene by which characteristics of a tumor microenvironment of a patient can be recognized, and thus can predict a treatment response. Therefore, the gene prediction model makes it possible to avoid unnecessary surgery and determine an optimal treatment plan.

Description

Biomarkers for predicting treatment response to immunotherapy and a genetic prediction model using them

[0001] This relates to a biomarker for predicting treatment response to immunotherapy and a genetic prediction model using the same.

[0002] Immunotherapy is a treatment that stimulates the body's immune system to fight cancer, targeting specific genetic characteristics of tumor cells. These treatments are rapidly evolving, particularly in the field of cancer treatment, and can be categorized into methods that use tumor vaccines, immune adjuvants, and immune substances (such as cytokines) to induce an active immune response within the body; immune checkpoint inhibitors that activate the immune response; and infusions of cytotoxic T cells or NK cells that actually destroy tumors.

[0003] Among these, immune checkpoint inhibitors are treatments that boost immunity by inhibiting the activity of checkpoint molecules, such as PD-L1 and CTLA-4, which block the activity of immune cells and prevent them from killing cancer cells. While immune checkpoint inhibitors offer the advantage of durable clinical benefit in patients who respond to treatment, the drawback is that only a small percentage of patients respond.

[0004] Meanwhile, immunotherapy does not appear to be effective in all patients receiving treatment, so it is crucial to determine whether immunotherapy is the appropriate treatment option for each patient. However, immunochemistry of PD-L1, a biomarker that predicts response to immune checkpoint inhibitors, has poor predictive power, necessitating the discovery of new biomarkers.

[0005] One aspect is to provide one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC for predicting treatment responsiveness to immunotherapy.

[0006] In one specific example, the biomarker may be a combination of any one of the following (i) to (iv): (i) TNFRSF18, CXCL9, and HLA-DQB1, (ii) TNFRSF18 and CD8A, (iii) TNFRSF18, CD8A, and VEGFC, and (iv) CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

[0007] Another aspect is to provide a biomarker panel or composition for predicting treatment responsiveness to immunotherapy comprising an agent measuring gene or protein expression levels of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

[0008] Another aspect is to provide a kit for predicting treatment responsiveness of an immunotherapy comprising the above biomarker panel or composition.

[0009] Another aspect is a method for providing information about treatment responsiveness to immunotherapy or a method for predicting treatment responsiveness to immunotherapy, comprising the step of measuring gene or protein expression levels of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC from an isolated biological sample.

[0010] Another aspect provides the use of a formulation for measuring gene or protein expression levels of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC for predicting treatment responsiveness to immunotherapy.

[0011] Another aspect is to provide a method for constructing a panel of biomarkers to predict treatment responsiveness to immunotherapy.

[0012] Another aspect provides a method for constructing a predictive model for predicting a therapeutic response to immunotherapy using a machine learning predictive model, the method including a step of analyzing the expression level of a biomarker according to a certain aspect.

[0013] Another aspect is to provide a system for predicting treatment response of immunotherapy using a machine learning prediction model including a data analysis unit that analyzes the expression level of biomarkers according to a certain aspect.

[0014] One aspect provides a biomarker panel or composition for predicting treatment responsiveness to immunotherapy. In one embodiment, the biomarker panel composition may include a formulation that measures the gene or protein expression level of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

[0015] As used herein, the term "biomarker panel" refers to any combination of biomarkers for predicting treatment responsiveness to immunotherapy, which may refer to the entire set, or any subset or subcombination thereof. In other words, a biomarker panel may refer to a set of biomarkers, or may refer to any form of biomarker being measured. Thus, if CD8A is part of a biomarker panel, for example, CD8A mRNA or CD8A protein may be considered part of the panel. While individual biomarkers are useful as diagnostic agents, sometimes a combination of biomarkers may provide greater value than a single biomarker in determining a specific condition. Specifically, detecting multiple biomarkers in a sample may increase the sensitivity and / or specificity of the test. Thus, in one embodiment, the biomarker panel can include 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or more types of biomarkers. In another embodiment, the biomarker panel consists of the minimum number of biomarkers to generate the maximum amount of information. Thus, in various embodiments, the biomarker panel consists of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 or more types of biomarkers. When a biomarker panel consists of "a set of biomarkers," no biomarkers are present other than those that make up the set. In one embodiment, the biomarker panel consists of one biomarker disclosed herein. In another embodiment, the biomarker panel consists of two biomarkers disclosed herein. In another embodiment, the biomarker panel consists of three biomarkers disclosed herein. In another embodiment, the biomarker panel comprises four or more biomarkers disclosed herein. The biomarkers of the present invention demonstrate statistically significant differences in predicting treatment response to immunotherapy.In one embodiment, a diagnostic test using these biomarkers alone or in combination exhibits a sensitivity and specificity of at least about 85%, at least about 90%, at least about 95%, at least about 98%, and at least about 100%. In one embodiment, the biomarkers may be a combination of any one of the following (i) to (iv): (i) TNFRSF18, CXCL9, and HLA-DQB1, (ii) TNFRSF18 and CD8A, (iii) TNFRSF18, CD8A, and VEGFC, and (iv) CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

[0016] In one specific example, measuring the expression level of the gene may include measuring the amount of mRNA, which is a process for confirming the presence and expression level of mRNA of the genes in a biological sample. Analytical methods for this may include reverse transcription polymerase chain reaction (RT-PCR), competitive RT-PCR, real-time RT-PCR, RNase protection assay (RPA), Northern blotting, DNA chips, etc.

[0017] Preparations capable of measuring the expression level of the above genes may include primers, probes, or antisense oligonucleotides. Since the nucleic acid information for these genes is available in GeneBank and elsewhere, those skilled in the art can design primers or probes that specifically amplify specific regions of these genes based on the sequences.

[0018] In one specific example, the target exon region for measuring the expression level of the gene may include at least 50 consecutive polynucleotide sequences, at least 80 consecutive polynucleotide sequences, at least 100 consecutive polynucleotide sequences, at least 120 consecutive polynucleotide sequences, at least 180 consecutive polynucleotide sequences, at least 240 consecutive polynucleotide sequences, at least 280 consecutive polynucleotide sequences, or at least 300 consecutive polynucleotide sequences, of any one polynucleotide sequence selected from the group consisting of SEQ ID NO: 1 (CXCL9), SEQ ID NO: 2 (CD8A), SEQ ID NO: 3 (TNFRSF18), SEQ ID NO: 4 (TIGIT), SEQ ID NO: 5 (HLA-DQB1), and SEQ ID NO: 6 (VEGFC), a polynucleotide sequence having at least 90%, at least 80%, at least 70% homology thereto, or a portion thereof.

[0019] The term "primer" as used herein includes any combination of primer pairs consisting of forward and reverse primers that recognize a target gene sequence, and more specifically, a primer pair that provides an analysis result with specificity and sensitivity.

[0020] The term "probe" used herein refers to a substance that can specifically bind to a target substance to be detected in a sample, and refers to a substance that can specifically confirm the presence of the target substance in the sample through said binding. The type of probe molecule is not limited to a substance commonly used in the art, but may preferably be PNA (peptide nucleic acid), LNA (locked nucleic acid), peptide, polypeptide, protein, RNA, or DNA. More specifically, the probe includes a biomaterial derived from or similar to a living organism or manufactured in vitro, and may be, for example, an enzyme, a protein, an antibody, a microorganism, an animal or plant cell and organ, a nerve cell, DNA, and RNA. DNA includes cDNA, genomic DNA, and oligonucleotides, RNA includes genomic RNA, mRNA, and oligonucleotides, and examples of proteins may include antibodies, antigens, enzymes, peptides, etc.

[0021] As used herein, the term "antisense oligonucleotide" refers to a DNA or RNA or derivative thereof containing a nucleic acid sequence complementary to the sequence of a specific gene (e.g., mRNA of a specific gene), which binds to the complementary sequence in the mRNA and inhibits the translation of the mRNA into protein. An antisense oligonucleotide sequence refers to a DNA or RNA sequence that is complementary to the mRNA of the genes and is capable of binding to the mRNA. This can inhibit the translation, translocation into the cytoplasm, maturation, or any other essential activity for the overall biological function of the gene mRNA. The length of the antisense oligonucleotide may be 6 to 100 bases, preferably 8 to 60 bases, and more preferably 10 to 40 bases.

[0022] In one specific example, measuring the expression or activity level of the protein may include a process of confirming the presence and expression (or activity) level of the protein in a biological sample. Analysis methods for this purpose may include protein chip analysis, immunoassay, ligand binding assay, MALDI-TOF (Matrix Desorption / Ionization Time of Flight Mass Spectrometry) analysis, SELDI-TOF (Sulface Enhanced Laser Desorption / Ionization Time of Flight Mass Spectrometry) analysis, radioimmunoassay, radioimmunodiffusion, aukteroni immunodiffusion, rocket immunoelectrophoresis, tissue immunostaining, complement fixation assay, two-dimensional electrophoresis analysis, liquid chromatography-mass spectrometry (LC-MS), liquid chromatography-mass spectrometry / mass spectrometry (LC-MS), western blot, and enzyme linked immunosorbent assay (ELISA).

[0023] The formulation capable of measuring the activity level of the above protein may comprise an antibody, an aptamer, an avidity multimer or a peptidomimeric.

[0024] The term "antibody" as used herein may refer to a specific protein molecule directed against an antigenic site. For the purposes of the present invention, the antibody refers to an antibody that specifically binds to the estrogen receptor, MUCIN5AC, or aromatase protein, and includes polyclonal antibodies, monoclonal antibodies, and recombinant antibodies. Antibodies can be readily produced using techniques well known in the art. In addition, the antibodies of the present specification include complete forms having two full-length light chains and two full-length heavy chains, as well as functional fragments of antibody molecules. Functional fragments of antibody molecules refer to fragments that retain at least an antigen-binding function, such as Fab, F(ab'), F(ab') 2, and Fv.

[0025] The term "treatment responsiveness" herein may mean whether a particular drug, for example, an anticancer agent, exhibits a therapeutic effect on an individual patient's cancer.

[0026]

[0027] In this specification, the term "predicting the treatment response of a cancer patient" may mean predicting in advance whether administration of a drug may be useful in treating cancer, and may mean predicting the treatment response to a drug by measuring the expression level of a gene.

[0028] The term "prediction" in this specification may mean to determine in advance a specific outcome, such as treatment responsiveness, by identifying a characteristic, such as the expression level of a specific gene.

[0029] As used herein, the term "cancer" refers to a physiological condition in animals that is typically characterized by abnormal or uncontrolled cell growth. Cancer and cancer pathology may be associated with, for example, metastasis, interference with normally functioning surrounding cells, release of cytokines or other secreted products at abnormal levels, suppression or enhancement of inflammatory or immunological responses, neoplasia, premalignancy, malignancy, invasion of surrounding or distant tissues or organs, such as lymph nodes, etc.

[0030] The cancer may be a solid cancer. Additionally, the cancer may be a gastrointestinal cancer or a non-gastrointestinal cancer.

[0031] The above gastrointestinal cancer is a malignant tumor occurring in the gastrointestinal tract, such as the esophagus, stomach, small intestine, or large intestine. The gastrointestinal cancer may be, for example, one or more cancers selected from the group consisting of esophageal cancer, gallbladder cancer, liver cancer, bile duct cancer, pancreatic cancer, stomach cancer, small intestine cancer, large intestine cancer, colon cancer, anal cancer, and rectal cancer, but is not limited thereto. In one example, it may be colon cancer.

[0032] The above non-gastrointestinal cancer includes, without limitation, malignant tumors occurring in organs other than the gastrointestinal tract or digestive system, and may be, for example, leukemia, acute myeloid leukemia, neuroblastoma, retinoblastoma, lung cancer, head and neck cancer, salivary gland cancer, melanoma, laryngeal cancer, prostate cancer, breast cancer, bladder cancer, kidney cancer, multiple myeloma, cervical cancer, thyroid cancer, ovarian cancer, urethral cancer, skin cancer, osteosarcoma, glioblastoma, brain tumor, or lymphoma, but is not limited thereto.

[0033] In this specification, the terms “immunotherapy,” “immune checkpoint inhibitor,” or “immuno-oncology agent” are used interchangeably and may refer to a substance that inhibits the activity of an immune checkpoint protein that suppresses the differentiation, proliferation, and activity of immune cells. The immunotherapeutic agent may be at least one selected from the group consisting of NK cell therapy, T cell therapy, CAR-immune cell therapy, DC vaccine, CTL therapy, anti-PD-L1, anti-PD-1, and anti-CTLA-4. The above-mentioned immuno-oncology agent may be an antibody to any one selected from the group consisting of 2B4, 4-1BB (CD137), AaR, B7-H3, B7-H4, BAFFR, BTLA, CD2, CD7, CD27, CD28, CD30, CD40, CD80, CD83 ligand, CD86, CD160, CD200, CDS, CEACAM, CTLA-4, GITR, HVEM, ICAM-1, KIR, LAG-3, LAIR1, LFA-1 (CD 11 a / CD 18), LIGHT, NKG2C, NKp80, OX40, PD-1, PD-L1, PD-L2, SLAMF7, TGFRp, TIGIT, Tim3, and VISTA. More specifically, it may be any one selected from the group consisting of, but is not limited to, anti-CTLA-4 antibody, anti-PD-1 antibody, anti-PD-L1 antibody, anti-PD-L2 antibody, anti-B7-H4 antibody, anti-HVEM antibody, anti-TIM3 antibody, anti-GAL9 antibody, anti-LAG3 antibody, anti-VISTA antibody, anti-KIR antibody, anti-BTLA antibody, and anti-TIGIT antibody.In one specific example, the immunotherapy anticancer agent is Atezolizumab, Avelumab, Durvalumab, Nivolumab, Pembrolizumab, Abagovomab, Adecatumumab, Afutuzumab, Alemtuzumab, Anatumomab mafenatox, Mepolizumab, Apolizumab, Blinatumomab, BMS-936559, Catumaxomab, Cemiplimab, Epacadostat, Epratuzumab, It may be any one selected from the group consisting of indoximod, Inotuzumab, Ozogamicin, intelumumab, Ipilimumab, Isatuximab, Lambrolizumab, MED 14736, MPDL3280A, Obinutuzumab, Ocaratuzumab, Ofatumumab, Olatatumab, Pidilizumab, Rituximab, Ticilimumab, Samalizumab, and Tremelimumab, but is not limited thereto.

[0034] In one specific example, the biomarker may be a combination of any one of the following (i) to (iv): (i) TNFRSF18, CXCL9, and HLA-DQB1, (ii) TNFRSF18 and CD8A, (iii) TNFRSF18, CD8A, and VEGFC, and (iv) CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

[0035]

[0036] Another aspect provides a kit for predicting treatment response to immunotherapy comprising a formulation that measures gene or protein expression levels of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

[0037] In one specific example, the kit may be a reverse transcription polymerase chain reaction (RT-PCR) kit, a DNA chip kit, a microarray kit, an enzyme-linked immunosorbent assay (ELISA) kit, a protein chip kit, a rapid kit, or a multiple reaction monitoring (MRM) kit.

[0038] In addition, the kit may further comprise one or more other component compositions, solutions or devices suitable for the analysis method. For example, the kit may be a kit comprising essential elements necessary for performing a reverse transcription polymerase reaction. The reverse transcription polymerase reaction kit comprises a pair of primers specific for each marker gene. The primers are nucleotides having a sequence specific to the nucleic acid sequence of each gene, and are about 7 bp to 50 bp in length, more preferably about 10 bp to 30 bp in length. It may also comprise a primer specific to the nucleic acid sequence of a control gene. In addition, the reverse transcription polymerase reaction kit may comprise a test tube or other appropriate container, a reaction buffer (with various pH and magnesium concentrations), deoxynucleotides (dNTPs), an enzyme such as Taq polymerase and reverse transcriptase, a DNAse, RNAse inhibitor DEPC-water, sterile water, etc. In addition, for example, the DNA chip kit may include a substrate to which cDNA or oligonucleotides corresponding to a gene or a fragment thereof are attached, and reagents, agents, enzymes, etc. for producing a fluorescent label probe. In addition, the substrate may include cDNA or oligonucleotides corresponding to a control gene or a fragment thereof. In addition, for example, the kit may be a diagnostic kit including essential elements required for performing LISA. The ELISA kit includes an antibody specific for the protein. The antibody is an antibody with high specificity and affinity for each marker protein and little cross-reactivity to other proteins, and may be a monoclonal antibody, a polyclonal antibody, or a recombinant antibody. In addition, the ELISA kit may include an antibody specific for a control protein.Other ELISA kits may include reagents capable of detecting bound antibodies, such as labeled secondary antibodies, chromophores, enzymes (e.g., conjugated to antibodies), and their substrates or other substances capable of binding to antibodies. Furthermore, for example, the kit may be a rapid kit containing the essential components necessary for rapid testing that yields analytical results. Rapid kits include antibodies specific for proteins. The antibodies have high specificity and affinity for each marker protein and little cross-reactivity with other proteins, and may be monoclonal, polyclonal, or recombinant antibodies. Rapid kits may also include antibodies specific for control proteins. Other rapid kits may include reagents capable of detecting bound antibodies, such as nitrocellulose membranes immobilized with specific antibodies and secondary antibodies, membranes bound to antibody-bound beads, absorbent pads, sample pads, and other materials. In addition, for example, the kit may be an MRM (Multiple reaction monitoring) kit, which is an MS / MS mode that includes the essential elements required to perform mass analysis. While SIM (Selected Ion Monitoring) is a method that utilizes ions generated by a single collision in the source section of a mass analyzer, MRM is a method that selects specific ions from among the ions that have been broken once more, passes them through the source of another MS that is connected in series once more for collision, and then utilizes the ions obtained from these.

[0039] Another aspect provides a method for providing information about treatment responsiveness to immunotherapy or a method for predicting treatment responsiveness to immunotherapy, comprising the step of measuring gene or protein expression levels of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC from an isolated biological sample.

[0040] In one specific example, the method may include a step of determining that the treatment response of the immunotherapy is high when the expression level of a gene or protein of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC is higher or lower than the expression level of a control sample. The control may be used interchangeably with the term negative control. The control may not respond to the immunotherapy or may have a low response.

[0041] For example, if the expression level of a gene or protein of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, and IL10 is higher than the expression level of a control sample, the treatment response of the immunotherapy can be determined to be high.

[0042] In addition, if the expression level of one or more genes or proteins of biomarkers selected from the group consisting of CDH2 and VEGFC is lower than the expression level of the control sample, the treatment responsiveness of the immunotherapy can be determined to be high.

[0043]

[0044] In one specific example, the method may further include a step of performing immunotherapy or administering an immunotherapy agent when the treatment response to immunotherapy is determined to be high.

[0045] If the expression level is increased compared to the expression level measured in the control group, the step of determining that the subject is likely to be responsive to the immunotherapy or to show a prognosis of recovery after administration of the immunotherapy may further be included.

[0046] In one specific embodiment, the method may further include a step of determining whether to treat the cancer patient with an immunotherapy based on the predicted treatment responsiveness. In one specific embodiment, the step of predicting the treatment responsiveness may be determining that the cancer patient has a high treatment responsiveness to the immunotherapy if the measured gene or protein expression level of the gene is increased compared to other cancer patients. In one specific embodiment, the step of determining whether to treat the cancer patient with an immunotherapy may be determining that the cancer patient receives immunotherapy if the cancer patient is predicted to have a high treatment responsiveness.

[0047] In one specific example, the patient may have received or is receiving standard treatment. For example, the method may be used to predict whether the patient will respond well to immunotherapy by measuring the expression of the gene or protein in a biological sample obtained from a patient receiving standard treatment, thereby establishing a future treatment plan.

[0048] In one specific example, the method for measuring the expression level of the gene may include reverse transcription polymerase reaction, competitive reverse transcription polymerase reaction, real-time reverse transcription polymerase reaction, RNase protection assay, Northern blotting, or DNA chip.

[0049] In one specific example, the method for measuring the activity level of the protein may include Western blotting, ELISA, radioimmunoassay, radioimmunodiffusion, Ouchterlony immunodiffusion, rocket immunoelectrophoresis, immunohistochemical staining, immunoprecipitation assay, complement fixation assay, FACS, or protein chip.

[0050] The term “biological sample” as used herein means a sample obtained from a subject. The biological sample may include whole blood, plasma, serum, red blood cells, white blood cells (e.g., peripheral blood monocytes), ductal fluid, ascites, pleural eflux, nipple aspirate, lymph (e.g., disseminated tumor cells in a lymph node), bone marrow aspirate, saliva, urine, stool (i.e., excretions), sputum, bronchial washings, tears, fine needle aspirates (e.g., harvested by random mammary fine needle aspiration), any other body fluid, a tissue sample (e.g., tumor tissue) such as a tumor biopsy (e.g., a puncture biopsy) or a lymph node (e.g., a sentinel lymph node biopsy), a tissue sample (e.g., tumor tissue) such as a surgical resection of a tumor, and cellular extracts thereof. In one embodiment, the sample may be whole blood or a component thereof, such as plasma, serum, or a cell pellet. In one embodiment, the sample may be obtained by isolating circulating cells of a solid tumor from whole blood or a cellular fraction thereof using any technique known in the art. In one embodiment, the sample may be, for example, a formalin-fixed, paraffin-embedded (FFPE) tumor tissue sample from a solid tumor. In one embodiment, the sample may be a tumor lysate or extract prepared from frozen tissue obtained from a subject with cancer.

[0051]

[0052] Another aspect provides a method for constructing a panel of biomarkers to predict treatment responsiveness to immunotherapy.

[0053] Specifically, the method may include a step of obtaining treatment response information before or after immunotherapy from a sample isolated from an individual; a step of obtaining data by measuring gene expression levels of a plurality of individual biomarkers from the sample; and a step of selecting a gene associated with treatment response to immunotherapy from the data.

[0054] The step of selecting genes related to the treatment response of the above immunotherapy can be performed by generating expression data information of individual biomarker genes through machine learning analysis.

[0055] Biomarkers for predicting the therapeutic response of the above immunotherapy may be selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

[0056] Another aspect is a system for predicting the treatment response of immunotherapy provided based on the patient's gene expression analysis data.

[0057] An information receiving unit that collects the patient's genetic analysis data;

[0058] Includes a prediction unit that predicts the treatment response of a patient's immunotherapy using information input through the above information receiving unit,

[0059] The above-mentioned gene provides a system for predicting treatment response of immunotherapy, wherein the gene is at least one selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

[0060] Another aspect provides a recording medium comprising the above system.

[0061] Biomarkers specific to each aspect of the disease can effectively predict treatment response to immunotherapy. Furthermore, genetic prediction models utilizing these biomarkers can predict treatment response based on the expression patterns of key genes that can characterize the patient's tumor microenvironment, thereby avoiding unnecessary surgery and determining the optimal treatment plan.

[0062] Figure 1 shows the results of PCA analysis of gene expression patterns in patients with cervical cancer (CX), ovarian cancer (OV), and endometrial cancer (UT).

[0063] Figure 2 shows the results of predicting the response of an immune checkpoint inhibitor in solid cancer using nine biomarkers according to one specific example.

[0064] Figure 3 shows the results of predicting the response of an immune checkpoint inhibitor in solid cancer using three biomarkers according to one specific example.

[0065] Figure 4 shows the results showing the relationship between CXCL9 gene expression and treatment prognosis in solid cancer.

[0066] Figure 5 shows the results showing the relationship between VEGFC gene expression and treatment prognosis in solid cancer.

[0067] Figure 6 shows the results of predicting the response to an immune checkpoint inhibitor in cervical cancer using nine biomarkers according to one specific example.

[0068] Figure 7 shows the results of predicting the response to an immune checkpoint inhibitor in cervical cancer using three biomarkers according to one specific example.

[0069] Figure 8 shows the results of predicting the response of an immune checkpoint inhibitor to progressive disease in cervical cancer using nine biomarkers according to one specific example.

[0070] Figure 9 shows the results of predicting the response of an immune checkpoint inhibitor to progressive disease in cervical cancer using three biomarkers according to one specific example.

[0071] Figure 10 shows the results of predicting the response of an immune checkpoint inhibitor to progressive disease in ovarian cancer using three biomarkers according to one specific example.

[0072] Hereinafter, preferred examples are presented to aid in understanding the present invention. However, the following examples are provided solely to facilitate a better understanding of the present invention, and the scope of the present invention is not limited by the following examples.

[0073]

[0074] Example 1. Preparation of patient samples and confirmation of clinical characteristics

[0075] mRNA was extracted from parafil blocks (Formalin-Fixed Paraffin-Embedded, FFPE) of surgical tissues from 39 solid tumor patients receiving immune checkpoint inhibitor treatment at Bundang CHA Hospital, and the expression level was measured using panel sequencing.

[0076] The clinical characteristics of the samples of 39 patients with solid tumors are shown in Table 1 below.

[0077] Clinical characteristics Patient statistics Mean age = 56 (range: 30–83 years) Cancer Cervical cancer 20 (51.3%) Endometrial cancer 5 (12.8%) Ovarian cancer 14 (35.9%) Stage II 7 (18.0%) Stage III 24 (61.5%) Stage IV 8 (20.5%) Treatment response Complete response 3 (7.7%) Partial response 8 (20.5%) Stable disease 13 (33.3%) Progressive disease 15 (38.5%)

[0078] As shown in Table 1, the cancer type distribution of patients with solid tumors was cervical cancer (20 patients, 51.3%), endometrial cancer (5 patients, 12.8%), and ovarian cancer (14 patients, 35.9%). Thirty-two patients, excluding 7 patients in stage 2, had stage 3 or 4 disease. Treatment response included 3 patients with complete response, 8 patients with partial response, 13 patients with stable disease, and 15 patients with progressive disease. Eleven patients with complete response and partial response were classified as the durable clinical benefit (DCB) group responding to immune checkpoint inhibitor treatment, and 28 patients whose cancer worsened without responding to treatment were classified as the non-durable clinical benefit (NDB) group. mRNA was extracted from Formalin-Fixed Paraffin-Embedded (FFPE) tumor tissues of 39 patients with solid tumors whose response groups were confirmed, and gene expression was measured using NGS panel sequencing. In addition, mRNA was extracted from FFPE tumor tissues using the truXTRAC FFPE total NA Ultra kit (Covaris). The extracted RNA quality was measured using the 4200 Tapestation, and a cDNA library was generated using a cDNA synthesis kit (Celemics) and a library preparation kit (Celemics). The DNA library was captured and amplified using a target enrichment kit (Celemics), and the generated PCR products were sequenced using the NextSeq platform (Illumina).

[0079]

[0080] Example 2. Prediction of response to immune checkpoint inhibitor therapy in patients with solid tumors.

[0081] To predict the treatment response to immune checkpoint inhibitors in patients with solid tumors, gene expression in samples from the patients was analyzed using Principal Component Analysis (PCA).

[0082] As a result, as shown in Figure 1, uterine, ovarian, and endometrial cancers showed no differences in gene expression patterns depending on the cancer type. Based on the above results, it was confirmed that a predictive model could be constructed to distinguish response groups to immune checkpoint inhibitors using overall gene expression data, without distinguishing patient data by cancer type.

[0083] Afterwards, genes differentially expressed in the response groups were derived using group comparisons (two-group t-test p<0.05), prognostic analysis (overall survival p<0.05), and data information of expression values ​​(median absolute deviation, MAD >1) between the above response groups.

[0084] Additionally, to build a model for predicting the treatment response of immunotherapy, a machine learning model was built using the logistic regression method and the random forest method.

[0085] Afterwards, the main genes were selected using the prediction model constructed for the above-mentioned selected genes, and as a result, CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC were selected as shown in Table 2 below.

[0086] Gene Symbol Well-known Name t-test p-value overall survival p-value MAD CD8A 0.007 ns 1.53 CD8B ns ns 1.09 CCL 5 RANTES 0.002 3 ns 1.12 CXCL 9 MIG 7.9 x 10-5 0.00882.68HLA-DQB10.0024ns2.08TNFRSF18GITR0.0023ns1.24IL10nsns1.03CDH2N-cadherinnsns1.88VEGFCns0.0141.07

[0087] Ns: non-significant The performance of the prediction model constructed with the nine genes selected above was evaluated. Specifically, two comparison groups were established through a fair performance comparison of the constructed prediction models. The first comparison group used the expression of the PDL1 (CD274) gene, which is one of the methods mainly used to predict immune checkpoint inhibitor response. The second comparison group used the INF-γ signature gene list (CXCR6, TIGIT, CD274, HLA-DQA1, HLADRB1, HLA-E, LAG3, CXCL9, NKG7, CCL5, STAT1, CD8A, CD27), which is mainly used to predict activated immune response in cancer patients. The prediction models used were two models: logistic regression, which is a prediction model suitable for comparing two groups, and the random forest model, which shows excellent performance in classification problems. The better model was selected among them, and the results are shown in Figure 2.

[0088] As shown in Fig. 2, the predictive power of the nine marker genes according to one specific example for treatment response was 0.84 with an AUC value, which showed a remarkable predictive power for treatment response compared to the predictive power of PD-L1 expression (AUC=0.77) or INF-γ signature (AUC=0.72).

[0089] In addition, among the nine genes, three genes (TNFRSF18, CXCL9, and HLA-DQB1) with the best predictive power for treatment response were identified through Lasso regression analysis, and these genes were modeled using the same method as above to confirm their predictive power for treatment response, and the results are shown in Figure 3.

[0090] As shown in Fig. 3, the predictive power of three marker genes according to one specific example for treatment response was 0.92, showing a remarkable predictive power for treatment response compared to the predictive power of PD-L1 expression (AUC=0.77) or INF-γ signature (AUC=0.72).

[0091] In particular, among the 9 genes, CXCL9 and VEGFC were related to the patient's prognosis. Specifically, it was confirmed that the higher the expression of CXCL9, the higher the overall survival rate and the better the prognosis, and the higher the expression of VEGFV, the lower the survival rate and the worse the prognosis (see Figures 4 and 5).

[0092] As a result of the above, it was found that the genetic marker according to one specific example can be more useful in predicting the treatment response to immunotherapy than any existing marker.

[0093]

[0094] Example 3. Prediction of response to immune checkpoint inhibitor therapy in patients with cervical cancer.

[0095] To predict treatment response to immune checkpoint inhibitors in cervical cancer patients, modeling was conducted on 20 patients. The clinical characteristics of these 20 patients are presented in Table 3.

[0096] Clinical characteristics Patient statistics Age: Mean (range) 55 years (30–70) Stage II 6 (30.0%) Stage III 9 (45.0%) Stage IV 5 (25.0%) Treatment response Complete response 3 (15.0%) Partial response 5 (25.0%) Stable disease 5 (25.0%) Progressive disease 7 (35.0%)

[0097] As shown in Table 3, 14 patients, excluding 6, had stage 3 or 4 disease. Treatment response included 3 patients with complete response, 5 patients with partial response, 5 patients with stable disease, and 7 patients with progressive disease. The proportion of response group was 40.0%, which was higher than that of all solid tumor patients (28.2%). A treatment response prediction model for cervical cancer patients was constructed using two methods. First, a model to predict response and non-response groups was constructed using the same method as in Example 2. The performance of the prediction model was measured using the LOOCV method. As a result, as shown in Figure 6, the predictive power of PD-L1 expression in 20 cervical cancer patients showed a performance corresponding to random prediction power, with an AUC value of 0.56. However, in the case of the INF-γ signature, although the random forest model showed better results than the logistic regression model, the logistic regression model showed an AUC value of 0.32, indicating that its predictive power was actually worse than that of the random forest model. This result shows that in cervical cancer, the predictive power of the INF-γ signature was higher in the non-responder group than in the responder group, and the prediction of treatment responsiveness was reversed.

[0098] Meanwhile, the treatment response prediction model of nine marker genes according to one specific example showed better performance in the random forest model compared to the logistic regression model, and the AUC value indicating the predictive power was 0.72, showing a remarkable treatment response predictive power compared to the predictive power of PD-L1 expression or INF-γ signature.

[0099] In addition, as shown in Fig. 7, in the case of a prediction model made with three marker genes according to one specific example, logistic regression showed better predictive power than the random forest model with AUC = 0.82.

[0100]

[0101] Next, we classified the patients into three groups—complete response, partial response, and stable disease—excluding the seven patients with progressive disease, and then conducted a predictive analysis. From a clinical perspective, although the stable disease group did not actively respond to treatment, their disease did not worsen as much as the progressive disease group, so they can be classified as a responder group in the sense that the patients' disease was maintained without worsening.

[0102] As a result, as shown in Figure 8, the PD-L1 expression pattern showed slightly better predictive power than the random forest model, with an AUC value of 0.60. Furthermore, the INF-γ signature showed better performance with the logistic regression model than with the random forest model, but similar performance to the random forest model, with an AUC value of 0.55.

[0103] Meanwhile, when a prediction model was created using 9 marker genes according to a specific example, random forest showed better performance than logistic regression, and the AUC value was 0.84, indicating better predictive power than existing methods.

[0104] In addition, as shown in Fig. 9, in the case of a prediction model made with three marker genes according to one specific example, logistic regression showed better performance than random forest, but the prediction performance was not as good as that of a prediction model made with nine genes, with an AUC value of 0.76.

[0105] When examining the expression patterns of the three genes by group, the CXCL9 and TNFRSF18 genes showed a pattern of lower expression in the progressive disease group, but only the TNFRSF18 gene showed statistical significance at p=0.03. In addition, when identifying genes related to prognosis in cervical cancer, we were able to confirm that VEGFC was closely related to prognosis. Although not statistically significant, the expression of VEGFC was higher in the progressive disease group, and it was confirmed that the higher the expression of VEGFC, the worse the prognosis.

[0106] Therefore, it was found that when a prediction model was created using marker genes according to one specific example, it predicted the treatment response to immune checkpoint inhibitors much better than existing methods.

[0107]

[0108] Example 4. Prediction of response to immune checkpoint inhibitor therapy in ovarian cancer patients.

[0109] To predict treatment response to immune checkpoint inhibitors in ovarian cancer, modeling was conducted on 14 patients. The clinical characteristics of the patients are presented in Table 4.

[0110] Clinical characteristics Patient statistics Age: Mean (range) 58 years (32–79) Stage II 0 (0.0%) Stage III 12 (85.7%) Stage IV 2 (14.3%) Treatment response Complete response 0 (0.0%) Partial response 2 (14.3%) Stable disease 7 (50.0%) Progressive disease 5 (35.7%)

[0111] A prediction model was constructed in the same manner as Example 3, except that patients were classified into five patients with progressive disease and the remaining nine patients were responders, and the performance of the models was compared. As a result, as shown in Figure 10, the PD-L1 expression pattern showed a much worse predictive power than random prediction, with an AUC value of 0.29. In cases where the AUC value is lower than 0.5, it indicates that PD-L1 expression is higher in the progressive disease group than in the responder group, which is the opposite of what was expected. The INF-γ signature showed better predictive power in the random forest model than in the logistic regression model, and the AUC value was 0.74, showing some predictive power. In addition, in the case of a prediction model created with nine marker genes according to one specific example, the random forest showed better performance than the logistic regression, and the AUC value was 0.69, showing no higher predictive power than any of the INF-γ signatures. However, in the case of a prediction model made with three marker genes according to one specific example, the random forest model showed better performance, and it was confirmed that it showed a higher predictive power than the INF-γ signature with an AUC value of 0.80.

[0112] Therefore, when a prediction model was created using three marker genes according to one specific example, it was found to better predict the treatment response to immune checkpoint inhibitors than the existing method.

[0113]

[0114] Example 5. Marker gene discovery using RT-PCR data

[0115] The expression levels of six marker genes, CXCL9, CD8A, TNFRSF18, TIGIT, HLA-DQB1, and VEGFC, according to one specific example were measured using RT-PCR, and marker genes were discovered in the same manner as in Example 1 and Example 2, except that four samples with poor RNA quality were removed (NDB: 28, DCB: 7). In summary, genes differentially expressed in the response group were CXCL9 (target exon region: chr4:76927301-76927427, SEQ ID NO: 1), CD8A (target exon region: chr2:87,017,451-87,017,804, SEQ ID NO: 2), TNFRSF18 (target exon region: chr1:1,140,750-1,140,872, SEQ ID NO: 3), TIGIT (target exon region: chr3:114014392-114014721, SEQ ID NO: 4), HLA-DQB1 (target exon region: chr6:32629744-32630025, SEQ ID NO: 5) and VEGFC (target exon region: Using the RT-PCR expression values ​​of chr4:177,608,341-177,608,674, sequence number 6), group comparisons between response groups (two-group t-test p<0.05), prognostic analysis (overall survival p<0.05), and data information of expression values ​​(median absolute deviation, MAD >1) were analyzed, and the results are shown in Table 5 below.

[0116] Gene Symbol Well-known Name t-test p-value overall survival p-value MAD X CL 9 MIG 0.000 4 0.01 6 2.89 CD 8 A 0.00 4 0.06 4 1.5 TN FR SF 18 GI TR 0.00 9 0.11 69 TIG IT 0.02 0.2 1 1.48 HLA-DQ B 10.5 0.8 6 1.57 VEG F C 0.5 3 0.7 3 1.56

[0117] As shown in Table 5 above, the expression of CXCL9, CD8A, TNFRSF18, and TIGIT genes was significantly higher in DCB, and it was found that the group with high expression levels of the four selected genes tended to have a good prognosis. In addition, although VEGFC was high in NDB only in some patients and did not give a significant result in the statistical analysis t-test, it can provide information to distinguish the responder and non-responder group when added to the modeling. Afterwards, for three of the six selected genes (CD8A, TNFRSF18, and VEGFC), the predictive power of treatment response was measured using a logistic regression model in the same manner as in Example 2, excluding four samples (i.e., a total of 35 samples). As a result, the combination of TNFRSF18 and CD8A showed an AUC value of 0.85, and the combination of TNFRSF18, CD8A, and VEGFC showed an AUC value of 0.85. These results showed that the combination of the above two and three genes also showed higher predictive power in predicting treatment response than the predictive power of PD-L1 expression or INF-γ signature.

[0118] In addition, sensitivity, specificity, precision, negative predictive value (NPV), and accuracy according to the DCB prediction probability cut-off (0.2) were calculated according to the following formulas, and the results are shown in Table 6 below.

[0119] Sensitivity = True Positive / (True Positive + False Negative)

[0120] Specificity = True Negative / (True Negative + False Positive)

[0121] Precision = True Positive / (True Positive + False Positive)

[0122] Negative predictive value (NPV) = True Negative / (True Negative + False Negative)

[0123] Accuracy = (True Positive + True Negative) / (True Positive + False Positive + True Negative + False Negative)

[0124] DCB probability cut off(n=35)TNFRSF18 and CD8ATNFRSF18, CD8A, and VEGFC0.2Sensitivity: 0.86 (6 / 7)Specificity: 0.75 (21 / 28)Precision: 0.46 (6 / 13)NPV: 0.95 (21 / 22)Accuracy: 0.77 (27 / 35)Sensitivity : 0.86 (6 / 7)Specificity : 0.71 (20 / 28)Precision : 0.43 (6 / 14)NPV : 0.95 (20 / 2)Accuracy : 0.74 (26 / 35)

[0125] As shown in Table 6 above, the combination of TNFRSF18 and CD8A and the combination of TNFRSF18, CD8A, and VEGFC according to one specific example have high sensitivity and specificity and can be usefully used to predict the therapeutic response to immune checkpoint inhibitors.

[0126]

[0127] The foregoing description of the present invention is provided for illustrative purposes only. Those skilled in the art will readily appreciate that the present invention can be readily modified into other specific forms without altering the technical spirit or essential characteristics of the present invention. Therefore, the embodiments described above should be understood as illustrative in all respects and not restrictive.

Claims

1. A biomarker panel for predicting treatment response to immunotherapy, comprising an agent measuring gene or protein expression levels of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

2. A biomarker panel for predicting treatment response to immunotherapy, wherein the agent for measuring the expression level of the gene according to claim 1 is a primer, a probe or an antisense oligonucleotide.

3. A biomarker panel for predicting therapeutic response to immunotherapy, wherein the agent for measuring the expression level of the protein according to claim 1 is an antibody, an aptamer, an avidity multimer, or a peptidomimerics.

4. In claim 1, the biomarker is a biomarker panel for predicting treatment response to immunotherapy, wherein the biomarker is a combination of any one of the following (i) to (iv); (i) TNFRSF18, CXCL9 and HLA-DQB1, (ⅱ) TNFRSF18 and CD8A, (ⅲ) TNFRSF18, CD8A and VEGFC, and (iv) CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

5. A biomarker panel for predicting treatment response to immunotherapy, for predicting treatment response to immunotherapy in a patient with solid cancer according to claim 1.

6. A kit for predicting treatment response to immunotherapy, comprising a formulation for measuring gene or protein expression levels of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

7. A method for providing information on treatment responsiveness of an immunotherapy, comprising the step of measuring the gene or protein expression level of one or more biomarkers selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC from an isolated biological sample.

8. A method according to claim 7, further comprising a step of determining that the therapeutic response to immunotherapy is high when the expression or activity level of any one or more genes or proteins selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC is lower or higher than the expression or activity level of the genes or proteins of a normal control sample.

9. A method for providing information on therapeutic responsiveness of immunotherapy according to claim 7, wherein the biomarker is a combination of TNFRSF18, CXCL9, and HLA-DQB1, or a combination of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

10. A method according to claim 7, wherein the method further comprises a step of determining whether to administer immunotherapy to the patient based on the predicted therapeutic responsiveness.

11. In a system for predicting the therapeutic response to immunotherapy provided based on the patient's gene expression analysis data, An information receiving unit that collects the patient's genetic analysis data; Includes a prediction unit that predicts the treatment response to immunotherapy for a patient using information input through the above information receiving unit, A system for predicting treatment response to immunotherapy, wherein the above genes are at least one selected from the group consisting of CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

12. A system for predicting therapeutic response to immunotherapy according to claim 11, wherein the gene is a combination of any one of the following (i) to (iv); (i) TNFRSF18, CXCL9 and HLA-DQB1, (ⅱ) TNFRSF18 and CD8A, (ⅲ) TNFRSF18, CD8A and VEGFC, and (iv) CD8A, CD8B, CCL5, CXCL9, HLA-DQB1, TNFRSF18, IL10, CDH2, and VEGFC.

Images