Method for predicting the efficacy of cancer therapy with immune checkpoint inhibitors

Abstract

To provide novel means capable of grasping an effective case of an immunization checkpoint inhibitor before start of treatment.SOLUTION: There is provided a method comprising: a step for using a tumor tissue sample and a periphery blood sample collected from a cancer patient before treatment of the patient is started by an immunization checkpoint inhibitor, the patient being an object of treatment effect prediction, for acquiring information of a TCR repertoire of a tumor invasion T cell in the tumor tissue and information of TCR repertoire of a CD8 positivity T cell in peripheral blood; a step for comparing the TCR repertoire of the tumor invasion T cell with the TCR repertoire of the CD8 positivity T cell in the periphery blood, for identifying a T cell clone detected in both TCR repertoires in common as a double clone; and a step for calculating a total frequency occupied by the double clone in the TCR repertoire of the CD8 positivity T cell in the peripheral blood, in which prediction of a cancer treatment effect by the immunization checkpoint inhibitor can be assisted.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a method for predicting the efficacy of cancer therapy with an immune checkpoint inhibitor. [Background technology]

[0002] Immune checkpoint inhibitors (ICBs) such as the anti-CTLA-4 antibody Ipilimumab and the anti-PD-1 antibody Nivolumab have been reported one after another with high levels of evidence showing treatment outcomes superior to conventional radiochemotherapy, ushering in a new era in which cancer immunotherapy will make a leap forward as a standard treatment. In particular, PD-1 / PD-L1 inhibitors, which promote the reactivation of tumor-specific T cells that have become exhausted locally in the tumor, have been approved in Japan for nine types of cancer, including melanoma and non-small cell lung cancer, and have also been shown to be effective in esophageal cancer, and the number of prescriptions is increasing year by year.

[0003] However, the response rate of PD-1 / PD-L1 inhibitors is often only 20-30% even for approved cancer types (Non-Patent Document 1), while some patients experience severe immune-related adverse events (irAE) (Non-Patent Document 2), and more recently, issues such as rapid tumor growth (hyper progressive disease; HPD) after the start of treatment have become apparent in 10-20% of patients (Non-Patent Document 3). Today, as the number of cancer types and prescriptions for PD-1 / PD-L1 inhibitors continues to increase, it is an urgent task to identify biomarkers that can detect effective cases, irAE, HPD, etc. before or early after treatment, and to develop patient stratification methods.

[0004] As a method for predicting and diagnosing the therapeutic response to PD-1 / PD-L1 inhibitors, several approaches evaluating the immune sensitivity of tumors or the immune status of the host are expected. Regarding the former, microsatellite instability (MSI) has been established, while the usefulness of tumor-mutation burden (TMB) and PD-L1 expression in tumor localization for prior prediction has not been established. Immunophenotyping by flow cytometry has been attempted as an immune monitoring method for evaluating the immune status of the host, but most of these methods are limited to the evaluation of nonspecific immune cell subsets that do not reflect the specificity of tumor antigens, and therefore their usefulness for both prior prediction and evaluation of therapeutic response has not been established.

[0005] T cell receptor (TCR) repertoire analysis is a new immune monitoring method that aims to evaluate T cell responses in individuals based on antigen specificity through determining a diverse collection of TCRs that define the antigen specificity of T cells (Non-Patent Document 4). In the field of tumor immunology, some clinical studies have reported that the therapeutic effects and adverse events of immune checkpoint inhibitors correlate with the TCR repertoires in peripheral blood and tumor local areas before treatment, and with the changes associated with treatment (Non-Patent Documents 5, 6). However, these previous studies have problems such as not separating CD4 / CD8 T cells from the peripheral blood samples used for analysis, and the correlation between TCR repertoire information and therapeutic effects is not strong enough to be established as a biomarker.

[0006] So far, the group of the present inventors has found that even when nucleic acids are extracted from tumor biopsy specimens without isolating CD4-positive T cells or CD8-positive T cells, the unfractionated tumor TCR repertoire can be annotated using the CD4-positive TCR repertoire and CD8-positive TCR repertoire of the peripheral blood or tumor-associated lymph nodes of the same patient as references, and the repertoire of tumor-infiltrating CD4-positive T cells and tumor-infiltrating CD8-positive T cells can be analyzed, and the effectiveness of cancer treatment can be determined by comparing the tumor-infiltrating T cell repertoire before and after treatment (Patent Document 1). However, the method reported in Patent Document 1 is a method for determining the effectiveness of a treatment that has already been received, and is not a method for predicting in advance whether a cancer treatment will be successful. [Prior art documents] [Patent documents]

[0007] [Patent Document 1] International Publication No. 2019 / 189383 [Non-patent literature]

[0008] [Non-Patent Document 1] Ribas A et al., Science, 2018, vol. 359, pp. 1350-1355. [Non-Patent Document 2] Postow MA et al., N Engl J Med, 2018, vol 378, pp. 158-168. [Non-Patent Document 3] Sasaki A et al., Gastric Cancer, 2019, vol. 22, pp. 793-802. [Non-Patent Document 4] D. Schrama et al., Semin Immunopathol., 2017, vol.39, no.3, pp.255-268 [Non-Patent Document 5] E. Cha et al., Sci Transl Med., 2014, vol.6, no.238, pp.238ra70 [Non-Patent Document 6] E. Yusko et al., Cancer Immunol Res., 2019, vol.7, no.3, 458-465 Summary of the Invention [Problem to be solved by the invention]

[0009] The present invention aims to provide a novel means for identifying cases in which immune checkpoint inhibitors will be effective before the start of treatment. [Means for solving the problem]

[0010] As a result of intensive research, the present inventors have discovered that by obtaining information on the TCR repertoire of peripheral blood CD8-positive T cells and the TCR repertoire of tumor-infiltrating T cells from peripheral blood samples and tumor tissue samples collected from cancer patients before the start of treatment with an immune checkpoint inhibitor, identifying T cell clones detected in common in both as "peripheral blood-tumor overlap clones," and calculating the total frequency of peripheral blood-tumor overlap clones present in the peripheral blood sample before treatment, it is possible to predict whether or not an immune checkpoint inhibitor will be effective in treating cancer, thereby completing the present invention.

[0011] That is, the present invention provides a method for assisting in prediction of the efficacy of cancer therapy by an immune checkpoint inhibitor, comprising the steps of: obtaining information on the T cell receptor (TCR) repertoire of tumor-infiltrating T cells in the tumor tissue and the TCR repertoire of CD8-positive T cells in the peripheral blood, using a tumor tissue specimen and a peripheral blood specimen collected from a cancer patient prior to the start of treatment with an immune checkpoint inhibitor for which therapeutic efficacy is to be predicted; comparing the TCR repertoire of the tumor-infiltrating T cells with the TCR repertoire of peripheral blood CD8-positive T cells and identifying T cell clones detected in common in both as overlapping clones; and calculating the total frequency of overlapping clones in the TCR repertoire of peripheral blood CD8-positive T cells, and when the total frequency of the overlapping clones is equal to or higher than a predetermined cutoff value, the method assists in prediction that the cancer patient will be effective in treating the cancer by the immune checkpoint inhibitor. Effect of the Invention

[0012] According to the present invention, by analyzing tumor tissue samples and peripheral blood samples collected from a cancer patient before the start of treatment with an immune checkpoint inhibitor, it is possible to predict whether or not the immune checkpoint inhibitor will have a cancer therapeutic effect in that cancer patient before the administration of the immune checkpoint inhibitor. By starting treatment with the immune checkpoint inhibitor for patients who are predicted to have a cancer therapeutic effect, and selecting a different treatment method for patients who are not predicted to have a cancer therapeutic effect, it is possible to minimize the risk that the immune checkpoint inhibitor will not have a therapeutic effect and only adverse effects such as irAE and HPD will occur. [Brief description of the drawings]

[0013] [Figure 1] 1 is an outline of a method for preparing a TCR variable region library for TCR repertoire analysis. [Diagram 2] This is a graph in which the rate of change in tumor size after the start of treatment in eight cases is plotted on the horizontal axis and the total frequency of overlapping clones in the peripheral blood CD8-positive T cell repertoire before the start of treatment is plotted on the vertical axis for each patient. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] The method of the present invention is a method for assisting in predicting the efficacy of cancer therapy using immune checkpoint inhibitors, and comprises the following steps. (TCR repertoire information acquisition process) A process of acquiring information on the T cell receptor (TCR) repertoire of tumor-infiltrating T cells in the tumor tissue and the TCR repertoire of CD8 positive T cells in the peripheral blood, using tumor tissue specimens and peripheral blood specimens collected from a cancer patient before starting treatment with an immune checkpoint inhibitor. (Overlapping clone identification step) A step of comparing the TCR repertoire of tumor-infiltrating T cells with the TCR repertoire of peripheral blood CD8-positive T cells, and identifying T cell clones detected in common to both as overlapping clones. (Step of calculating the total frequency of overlapping clones) A ​​step of calculating the total frequency of overlapping clones in the TCR repertoire of peripheral blood CD8-positive T cells. When the total frequency of overlapping clones calculated in the step of calculating the total frequency of overlapping clones is equal to or greater than a predetermined cutoff value, it is possible to predict that an immune checkpoint inhibitor will be effective in treating cancer in the cancer patient.

[0015] In the present invention, an immune checkpoint inhibitor is a substance that promotes T cell activation by inhibiting the function of an inhibitory immune checkpoint receptor or its ligand, more specifically, a substance that inhibits the activation of an inhibitory immune checkpoint receptor by inhibiting the binding between the inhibitory immune checkpoint receptor and its ligand, and can also be referred to as an "antagonist to an inhibitory immune checkpoint receptor or its ligand."

[0016] Immune checkpoint inhibitors include substances that inhibit receptor activation by binding to the receptor and interfering with the binding of a ligand to the receptor, and substances that inhibit receptor activation by binding to the ligand and interfering with the binding of the ligand to the receptor. Examples of immune checkpoint inhibitors include antagonistic antibodies that bind to inhibitory immune checkpoint receptors, antibodies that bind to inhibitory immune checkpoint ligands and interfering with binding to the receptor, and soluble polypeptides designed based on inhibitory immune checkpoint ligands that do not activate receptors.

[0017] Immune checkpoints are immune escape mechanisms that prevent the immune system from attacking the body itself. Immune checkpoint receptors exist on T cells and interact with ligands expressed on cancer cells and antigen-presenting cells. T cells recognize antigens presented on MHC molecules, become activated, and initiate an immune response, but T cell activation is regulated by immune checkpoint receptor-ligand interactions that occur in parallel. There are costimulatory and inhibitory immune checkpoint receptors, and T cell activation and immune responses are regulated by the balance between the two.

[0018] Cancer cells express ligands for inhibitory immune checkpoint receptors and use these receptors to escape destruction by cytotoxic T cells. Therefore, by administering an antagonist for the inhibitory receptor or its ligand, it is possible to prevent cancer cells from using the immune checkpoint mechanism and promote the killing of cancer cells by CD8+ T cells. Immune checkpoint inhibitors are drugs that exert anticancer effects through this mechanism, and various types are known.

[0019] Specific examples of inhibitory immune checkpoint receptors and their ligands targeted by immune checkpoint inhibitors include receptors such as PD-1, CTLA-4, LAG-3, TIM-3, BTLA, and KIR, and ligands such as PD-L1 (PD-1 ligand), PD-L2 (PD-1 ligand), GAL9 (TIM-3 ligand), and HVEM (BTLA ligand).

[0020] Therefore, the term immune checkpoint inhibitor includes PD-1 inhibitor, CTLA-4 inhibitor, LAG-3 inhibitor, TIM-3 inhibitor, BTLA inhibitor, and KIR inhibitor. Even when the receptor name is specified in this way, the above explanation of the term immune checkpoint inhibitor applies. The term "PD-1 inhibitor" includes a substance that binds to the receptor PD-1 or its ligand PD-L1 or PD-L2 and inhibits the binding of PD-1 to PD-L1 / PD-L2, thereby inhibiting the activation of PD-1, and includes, for example, an antagonistic antibody that binds to PD-1, an antibody that binds to PD-L1 or PD-L2 and inhibits the binding of PD-1 to PD-L1 or PD-L2, and a soluble polypeptide designed based on PD-L1 or PD-L2 that binds to PD-1 without activating PD-1. The same applies to the terms CTLA-4 inhibitor, LAG-3 inhibitor, TIM-3 inhibitor, BTLA inhibitor, and KIR inhibitor.

[0021] Specific examples of each inhibitor include: specific examples of PD-1 inhibitors include anti-PD-1 antibodies, anti-PD-L1 antibodies, and anti-PD-L2 antibodies; specific examples of CTLA-4 inhibitors include anti-CTLA-4 antibodies; specific examples of LAG-3 inhibitors include anti-LAG-3 antibodies; specific examples of TIM-3 antibodies include anti-TIM-3 antibodies and anti-GAL9 antibodies; specific examples of BTLA inhibitors include anti-BTLA antibodies and anti-HVEM antibodies; and specific examples of KIR inhibitors include anti-KIR antibodies.

[0022] Specific examples of immune checkpoint inhibitors, together with examples of known pharmaceuticals, include anti-PD-1 antibodies [e.g., nivolumab (Opdivo (registered trademark)), Cemiplimab (REGN-2810), pembrolizumab (KEYTRUDA (registered trademark), MK-3475), Spartalizumab (PDR-001), Tislelizumab (BGB-A317), AMP-514 (MEDI0680), Dostarlimab (ANB011, TSR-042), Toripalimab (JS001), Camrelizumab (SHR-1210), Genolimzumab (CBT-501), Sintilimab (IBI308), STI-A1110, ENUM 388D4, ENUM 244C8, GLS010, MGA012, AGEN2034, CS1003, HLX10, BAT-1306, AK105, AK103, BI754091, LZM009, CMAB819, Sym021, GB226, SSI-361, JY034, HX008, ABBV181, BCD-100, PF-06801591, CX-188 and JNJ-63723283, etc.], anti-PD-L1 antibodies [e.g., atezolizumab (Tecentriq®, RG7446, MPDL3280A), avelumab (BAVENCIO®, PF-06834635, MSB0010718C), durvalumab ( Imfinzi®, MEDI4736), BMS-936559, STI-1010, STI-1011, STI-1014, KN035, LY3300054, HLX20, SHR-1316, CS1001 (WBP3155), MSB2311, BGB-A333, KL-A167, CK-301, AK106, AK104, ZKAB001, FAZ053, CBT-502 (TQB2450), JS003 and CX-072, etc.], anti-PD-L2 antibodies [e.g. , rHIgM12B7], anti-CTLA-4 antibodies [e.g., Ipilimumab (YERVOY®), Tremelimumab, AGEN-1884], anti-TIM-3 antibodies [e.g., MBG453], anti-LAG-3 antibodies [e.g., BMS-986016, LAG525], anti-KIR antibodies [e.g., Lirilumab], PD-L1 fusion proteins, PD-L2 fusion proteins [e.g., AMP-224], PD-1 antagonists [e.g., AUNP-12, BMS-M Examples of the anti-PD-1 antibody include the compounds BMS-1, BMS-2, BMS-3, BMS-8, BMS-37, BMS-200, BMS-202, BMS-230, BMS-242, BMS-1001, BMS-1166, the compounds Incyte-1 to Incyte-6, CAMC-1 to CAMC-4, RG_1, and DPPA-1, PD-L1 / VISTA antagonists, CA-170, and CA-327. Antibodies containing the heavy and light chain complementarity determining regions (CDRs) or variable regions (VR) of the above known antibodies are also an embodiment of the immune checkpoint control agent. For example, a further embodiment of the anti-PD-1 antibody includes an antibody containing the heavy and light chain complementarity determining regions (CDRs) or variable regions (VR) of nivolumab.

[0023] Typical examples of immune checkpoint inhibitors targeted by the method of the present invention include known pharmaceuticals such as those exemplified above, but are not limited to these, and include various drugs that will be developed in the future as anticancer drugs, so long as they fall within the above-defined definition of "immune checkpoint inhibitors."

[0024] The tumor tissue samples and peripheral blood samples used in the present invention are samples taken from cancer patients before starting treatment with an immune checkpoint inhibitor for which therapeutic efficacy is to be predicted. The cancer patient may have undergone cancer therapy other than immune checkpoint inhibitors, or may have been administered an immune checkpoint inhibitor other than the immune checkpoint inhibitor for which therapeutic efficacy is to be predicted by the method of the present invention. For example, the method of the present invention can be applied to cancer patients who have been administered a CTLA-4 inhibitor in the past but have not been administered a PD-1 inhibitor, in order to predict the therapeutic efficacy of a PD-1 inhibitor. The cancer patient is not particularly limited as long as it is a mammal, but is typically a human cancer patient.

[0025] The tumor tissue sample and the peripheral blood sample are collected from the same cancer patient at the same time. "At the same time" means within a few days, typically on the same day. The tumor tissue sample may be a tumor biopsy sample collected from a cancer patient at a medical institution. The peripheral blood sample may be blood collected at a medical institution.

[0026] The type of cancer is not particularly limited as long as it is a solid cancer. In the following examples, the analysis results of patients with esophageal cancer, colorectal cancer (colon cancer), and stomach cancer are described, but the present invention is not limited to these digestive cancers, and may be used to treat other digestive cancers such as esophageal adenocarcinoma, esophagogastric junction cancer, small intestine cancer, liver cancer, biliary tract cancer (e.g., gallbladder cancer, bile duct cancer, ampullary cancer), and pancreatic cancer, as well as malignant melanoma, lung cancer (e.g., small cell lung cancer, and non-small cell lung cancer such as lung adenocarcinoma, lung squamous cell carcinoma, and large cell lung cancer), tongue cancer, head and neck cancer, renal cell cancer (e.g., clear cell carcinoma, papillary renal cell carcinoma, chromophobe cell carcinoma, collecting duct carcinoma), breast cancer, ovarian cancer, serous ovarian cancer, ovarian clear cell adenocarcinoma, nasopharyngeal cancer, uterine cancer (e.g., The present invention encompasses a variety of solid cancers, including cervical cancer, endometrial cancer, and uterine cancer), anal cancer (e.g., anal canal cancer), urothelial cancer (e.g., bladder cancer, upper urinary tract cancer, ureteral cancer, renal pelvis cancer, and urethral cancer), prostate cancer, fallopian tube cancer, primary peritoneal cancer, pleural mesothelioma, skin cancer (e.g., uveal melanoma and Merkel cell carcinoma), testicular cancer (germ cell tumor), vaginal cancer, vulvar cancer, penile cancer, endocrine system cancer, thyroid cancer, parathyroid cancer, adrenal cancer, spinal tumor, brain tumor, glioblastoma, gliosarcoma, squamous cell carcinoma, bone and soft tissue sarcoma (e.g., Ewing's sarcoma, childhood rhabdomyosarcoma, and uterine leiomyosarcoma), and Kaposi's sarcoma.

[0027] A. TCR repertoire information acquisition process In the TCR repertoire information acquisition step, T cell receptor (TCR) repertoire analysis is performed on the tumor tissue sample to obtain information on the TCR repertoire of tumor-infiltrating T cells in the tumor tissue (hereinafter referred to as "tumor TCR repertoire"), and TCR repertoire analysis is performed on the peripheral blood sample to obtain information on the TCR repertoire of CD8-positive T cells in the peripheral blood (hereinafter referred to as "peripheral blood CD8-positive TCR repertoire").

[0028] The peripheral blood CD8-positive TCR repertoire can be obtained by separating and recovering CD8-positive T cells from a peripheral blood specimen using a commercially available kit or the like in a standard manner to prepare an mRNA sample, and then using this mRNA sample to prepare a TCR variable region library for repertoire analysis.

[0029] The tumor TCR repertoire may be a TCR repertoire of tumor-infiltrating T cells including T cells that are not CD8 positive (e.g., CD4 positive T cells), or a TCR repertoire of tumor-infiltrating CD8 positive T cells. In the present invention, either of these can be used as the tumor TCR repertoire. In addition, there are cases where a target patient has multiple cancer lesions (e.g., a primary lesion and one or more metastatic lesions, or two or more recurrent lesions) and multiple different tumor tissue specimens are obtained from the same patient. In this case, all of the multiple tumor tissue specimens may be analyzed. A TCR variable region library is prepared for each of the multiple tumor tissue specimens, and the information of the T cell clones of each tumor tissue obtained by comparing them with the reference sequence and extracting the functional TCR sequence can be pooled and used as the tumor TCR repertoire of the patient.

[0030] When using the TCR repertoire of tumor-infiltrating T cells, including T cells that are not CD8 positive, as the tumor TCR repertoire, a TCR variable region library may be prepared using an mRNA sample prepared from a tumor tissue specimen without fractionating the cells, and repertoire analysis may be performed, but an mRNA sample prepared by separating and recovering CD8 positive or CD4 positive T cells (CD4 / 8 T cells) from a tumor tissue specimen may also be used. Cell separation can be performed by a standard method using a cell sorter.

[0031] When the TCR repertoire of tumor-infiltrating CD8-positive T cells is used as the tumor TCR repertoire, CD8-positive T cells may be separated and collected from a tumor tissue specimen by a standard method using a cell sorter to prepare an mRNA sample, and a TCR variable region library may be prepared using this mRNA sample of CD8-positive T cells to perform repertoire analysis; alternatively, a TCR variable region library may be prepared from an mRNA sample prepared from a tumor tissue specimen without fractionating cells according to the repertoire analysis method described in WO 2019 / 189383, and repertoire analysis may be performed to obtain information on the unfractionated tumor TCR repertoire, and then the unfractionated tumor TCR repertoire may be annotated using the peripheral blood CD8-positive TCR repertoire as a reference, thereby obtaining information on the tumor-infiltrating CD8-positive T cell repertoire. However, in the present invention, since the information on the unfractionated tumor TCR repertoire can be used as the tumor TCR repertoire as is, it is simpler to omit the annotation of the unfractionated tumor TCR repertoire using the peripheral blood CD8 positive TCR repertoire as a reference.

[0032] In the analysis of the tumor TCR repertoire and the peripheral blood CD8-positive TCR repertoire, the sequence of the region including at least the CDR3 region in the variable region of the TCR is analyzed. The variable region may be the variable region of the α chain, the variable region of the β chain, or the variable regions of both. That is, in the present invention, at least the CDR3 region of the α chain variable region, at least the CDR3 region of the β chain variable region, or both of them may be analyzed. In addition, in the present invention, the term variable region has the same meaning as the VJ region in the case of the α chain and the VDJ region in the case of the β chain. Preferable specific examples of the "region including at least the CDR3 region" to be analyzed include a region including the CDR3 region, V region, and J region of the α chain, and / or a region including the CDR3 region, V region, D region, and J region of the β chain. As described below, when poly-C or poly-G tailing is performed to give the second strand synthesis primer a complementary homopolymer moiety of G or C at the 3' end and a molecular barcode is introduced upstream of the complementary homopolymer moiety, a next-generation sequencer can sequence beyond the homopolymer sequence, which has the advantage of being able to correct errors caused by PCR or sequencing of the TCR repertoire.

[0033] Below, the steps from preparation of a TCR variable region library to acquisition of TCR repertoire information will be explained.

[0034] 1. Preparation of Tumor TCR Variable Region Library 1-1. Preparation of unfractionated tumor TCR variable region library The mRNA prepared from unfractionated tumor tissue contains many mRNAs expressed in cells other than T cells, such as tumor cells, and cDNAs encoding various amino acid sequences other than TCR are synthesized, but by using techniques such as (1) suppressing the production of by-products by adding homopolymeric tails by terminal deoxynucleotidyl transferase reaction in the presence of ddNTPs, and (2) using constant region-specific primers with specific binding molecules bound to the 5' ends in the TCR cDNA amplification step, a TCR variable region library can be prepared even from unfractionated tumor mRNA samples with a low content of T cell-derived mRNA. The preparation of an unfractionated tumor TCR variable region library will be described below with reference to FIG. 1 as appropriate.

[0035] Preparation of an mRNA sample from an unfractionated tumor tissue specimen may involve extracting total RNA before extracting mRNA, or directly extracting mRNA from the specimen without extracting total RNA. One example of a general method for extracting mRNA is a method using a solid support such as beads immobilized with an mRNA capture nucleic acid having an oligo dT portion, and the present invention also preferably uses a method using a solid support. The solid support used may be beads or a plate such as a slide glass or a well plate, but beads are preferably used. Beads may be either non-magnetic or magnetic, but magnetic beads, which are easier to handle, are particularly preferred.

[0036] The nucleic acid for capturing mRNA usually has an appropriate adaptor sequence (first adaptor, Adaptor 1 in Figure 1) on the support side. After separating and recovering the support on which the mRNA has been captured, the cDNA is reverse transcribed with the mRNA still captured on the support. After the reverse transcription reaction, the mRNA is degraded with RNase, or simultaneously with the mRNA decomposition, a homopolymer tail is added to the 3' end of the cDNA.

[0037] In the homopolymer tail addition reaction, the reaction is carried out by terminal deoxynucleotidyl transferase (TdT) in the presence of ddNTP. As the ddNTP to be used, any of ddATP, ddCTP, ddGTP, and ddTTP can be used. However, it is preferable to align the base chain with the substrate that forms the homopolymer tail, and since the minimum homopolymer tail length required for the second strand synthesis reaction and the required length of the 3'-end complementary homopolymer portion of the primer used for second strand synthesis can be suppressed, poly-C tail addition by ddCTP+dCTP or poly-G tail addition by ddGTP+dGTP are preferable. Figure 1 shows an example of poly-C tail addition. As a divalent cation that significantly increases the efficiency of the homopolymer addition reaction by TdT, either cobalt or manganese can be used. The amount of ddNTP to be added may be about 1:10 to 1:100 in terms of the ratio to dNTP (dATP, dTTP, dCTP, or dGTP), and may be, for example, 1:10 to 1:80, 1:10 to 1:60, 1:10 to 1:40, 1:15 to 1:80, 1:15 to 1:60, or 1:15 to 1:40. By adding ddNTP, ddNTP is stochastically incorporated and the homopolymer extension reaction is stopped, so that the tolerance of reaction time extension by TdT is greatly increased. Therefore, it is also possible to add RNase H to the TdT reaction solution and perform the mRNA degradation reaction and the homopolymer tail addition reaction simultaneously.

[0038] Next, a complementary DNA strand of the cDNA is synthesized using a primer for second strand synthesis having a homopolymer portion complementary to the homopolymer tail at the 3' side and a second adapter sequence at the 5' side (second strand synthesis). When the homopolymer tail added to the cDNA is polyC or polyG, the chain length of the 3' complementary homopolymer portion in the primer for second strand synthesis is preferably 9 bases, but may be other lengths of 6 bases or more (for example, about 6 to 20 bases). When the homopolymer tail is polyA or polyT, the chain length of the 3' complementary homopolymer portion is usually about 15 to 30 bases, for example, about 18 to 25 bases. An anchor sequence of about 1 to 2 bases may be added to the 3' end of the complementary homopolymer portion. By adding an anchor sequence, the probability of the primer for second strand synthesis annealing to the starting point of the homopolymer tail added to the cDNA by TdT can be increased. The anchor sequence can be H or HN (H = A, T or C; N = any base (A, T, C or G)) when the complementary homopolymer portion is polyG, D or DN (D = A, T or G; N = any base (A, T, C or G)) when the complementary homopolymer portion is polyC, B or BN (B = T, G or C; N = any base (A, T, C or G)) when the complementary homopolymer portion is polyA, or V or VN (V = A, G or C; N = any base (A, T, C or G)) when the complementary homopolymer portion is polyT.

[0039] The second strand synthesis primer may include a molecular barcode composed of random bases between the complementary homopolymer portion and the second adapter sequence. The second strand synthesis primer of this embodiment has a structure in which each portion is linked in the order of [second adapter]-[molecular barcode]-[complementary homopolymer] from the 5' side. The chain length of the molecular barcode is not particularly limited, but is typically about 12 to 30 bases. The molecular barcode may have a configuration in which a specific fixed sequence is inserted into the random base sequence to provide resistance to sequence errors. A typical example of the molecular barcode is a molecular barcode with a chain length of about 12 to 30 bases containing three or four types of mixed bases.

[0040] After the second strand synthesis, a nucleic acid amplification reaction is performed by PCR using a high-fidelity enzyme to amplify a sequence containing the full-length V(D)J (variable region) of TCR from the cDNA on the support (TCR cDNA amplification). In this step, PCR is performed using a primer set for the second adaptor and a TCR constant region-specific primer set for the TCR constant region. Prior to the TCR cDNA amplification, the cDNA on the support may be comprehensively amplified (total cDNA amplification) using a primer set set for the first and second adaptors, and the total cDNA amplification product may be used as a template for TCR cDNA amplification, but the production of by-products is suppressed by adding a homopolymer tail by the TdT reaction using ddNTPs, so the total cDNA amplification step can be omitted.

[0041] The TCR constant region-specific primer may be set to the β constant region when the target of analysis is the β chain, and may be set to the α chain constant region when the target of analysis is the α chain. When both the α chain and the β chain are to be analyzed, a primer set to the α chain constant region and a primer set to the β chain constant region may be mixed and used as a set with a primer of the second adapter sequence.

[0042] The amplification step of the TCR cDNA is preferably carried out by nested PCR. The primer to be combined with the primer consisting of the second adaptor sequence may be a constant region-specific primer set within the constant region in both the first (1st PCR in FIG. 1) and second (2nd PCR in FIG. 1) rounds of nested PCR. When one of the primers is the same in the first and second rounds, this is sometimes called semi-nested PCR. By using a set of a primer set in the adaptor and a constant region-specific primer, the TCR cDNA can be amplified without PCR bias. For the convenience of adding an adaptor sequence for sequencing to the target of sequence analysis in a later step, it is preferable to use a constant region-specific primer with a specific binding molecule such as biotin bound to the 5' end in the final PCR in this amplification step (2nd PCR in FIG. 1).

[0043] After amplifying the TCR cDNA, it is fragmented to a length suitable for the sequencer to be used, and then DNA fragments to which adapter sequences (third and fourth adapters) for sequencing have been added are prepared. For example, the TCR cDNA fragment is fragmented to about 250 bp with an appropriate DNA cleavage enzyme to adjust the size, the protruding ends of the fragments are blunted and modified, and the third adapter (Adaptor P1 in Figure 1) is added. In the 2nd PCR, if a primer set for the constant region is used with a specific binding molecule such as biotin bound to the 5' end, Adaptor P1 is not added to the constant region side end of the fragment containing the constant region sequence because the specific binding molecule is present at the constant region side end, while Adaptor P1 is added to both ends of the fragment not containing the constant region sequence because the specific binding molecule is not bound to the fragment.

[0044] Next, PCR is performed using a primer targeting the third adaptor (P1 primer in FIG. 1) and a primer targeting the constant region on the DNA fragment (A-BC-TRC primer in FIG. 1). At this time, the fragment with adaptor P1 added to both ends is not amplified by PCR because it undergoes self-annealing within the DNA molecule during the PCR process, and therefore only fragments containing the constant region sequence are selectively amplified. The latter primer targeting the constant region has a fourth adaptor (lonA in FIG. 1) at the 5' end. A desired sequence (for example, a barcode for identifying which tissue sample the library is derived from) may be included between the fourth adaptor and the sequence targeting the constant region. In this way, a TCR variable region library with adaptors for sequencing added to both ends can be obtained.

[0045] 1-2. Preparation of TCR variable region library from tumor-infiltrating CD8+ T cells When CD8-positive T cells are separated and collected from a tumor tissue specimen to prepare an mRNA sample and a TCR variable region library is prepared, the TCR mRNA content in the mRNA sample is high. Therefore, there is no need to implement measures such as (1) suppressing the production of by-products by performing a homopolymer tail addition by a terminal deoxynucleotidyl transferase reaction in the presence of ddNTP, and (2) using a constant region-specific primer with a specific binding molecule bound to the 5' end in the TCR cDNA amplification step, and the tumor-infiltrating CD8-positive TCR variable region library can be prepared by a conventional method that does not include these measures. However, even when preparing a tumor-infiltrating CD8-positive TCR variable region library, these measures may be adopted and the library may be prepared by the same method as in 1.

[0046] 2. Preparation of TCR variable region library from peripheral blood CD8 positive T cells For peripheral blood samples, the same procedure as described in 1-2 above applies since it is necessary to separate and collect CD8-positive T cells to prepare a TCR variable region library. A peripheral blood CD8-positive TCR variable region library may be prepared by a conventional method that does not include the techniques described in (1) and (2), or the library may be prepared by the same method as described in 1 above, employing these techniques.

[0047] 3. Acquisition of sequence data for TCR variable region library Sequencing of the TCR variable region library may be performed using a sequencer generally called a next-generation sequencer, specific examples of which include the Ion Proton system and Ion S5 / Ion S5 XL system from Thermo Fisher Scientific, and the MiSeq system from Illumina.

[0048] 4. Extraction of Functional TCR Sequences From all the reads obtained by sequencing, reads having a functional TCR sequence are extracted. The extraction method is not particularly limited, but can be performed, for example, as follows. First, all the reads are compared with a known TCR database (e.g., IMGT library (http: / / www.imgt.org / ), etc.), and reads that are considered to have a TCR sequence based on sequence identity are extracted as TCR sequence detection reads. Next, the TCR sequence detection reads are analyzed, and reads that can identify the sequence of the variable region (VDJ sequence in the case of a β chain, VJ sequence in the case of an α chain) are extracted as variable region identification reads. This analysis can be performed using MiXCR (https: / / milaboratory.com / software / mixcr / ). Next, from the variable region identification reads, reads that do not contain a stop codon are extracted as functional TCR sequence identification reads.

[0049] 5. Obtaining sequence information of T cell clones by repertoire analysis Using the sequence information of the extracted functional TCR sequence identification lead, the repertoire structure (type of clones, frequency of presence of each clone) is analyzed to obtain sequence information of unfractionated intratumoral T cell clones and CD4-positive T cell clones or CD8-positive T cell clones in the second sample. This analysis can be performed using various known analysis software.

[0050] B. Overlapping clone identification process In the overlapping clone identification step, sequence information of intratumoral T cell clones (unfractionated intratumoral T cell clones or tumor-infiltrating CD8-positive T cell clones) is compared with sequence information of peripheral blood CD8-positive T cell clones. Some of the peripheral blood CD8-positive T cell clones have the same TCR sequence as the intratumoral T cell clones. The clones having the same TCR sequence are identified as peripheral blood-tumor overlapping CD8-positive T cell clones (hereinafter, "overlapping clones").

[0051] C. Calculation of total frequency of overlapping clones After identifying the overlapping clones, the total frequency of the overlapping clones in the peripheral blood CD8+ TCR repertoire is calculated using the following formula: Total frequency (%) of overlapping clones in the TCR repertoire of peripheral blood CD8+ T cells = [Number of CD8+ T cells belonging to overlapping clones in peripheral blood] / [Number of CD8+ T cells in peripheral blood] × 100

[0052] When the calculated total frequency of overlapping clones in peripheral blood before treatment is equal to or higher than a predetermined cutoff value, it can be predicted that the cancer patient is responsive to an immune checkpoint inhibitor, i.e., the cancer patient will be able to obtain a cancer therapeutic effect (such as tumor shrinkage and improved survival rate) by the immune checkpoint inhibitor, and when the calculated total frequency is less than the predetermined cutoff value, it can be predicted that the cancer patient is unresponsive to an immune checkpoint inhibitor, i.e., the cancer patient will not obtain a cancer therapeutic effect by the immune checkpoint. The cutoff value can be set within a range of 7% to 12%, for example, within a range of 8% to 11.5%, or 8.5% to 11.5%. By starting treatment with an immune checkpoint inhibitor for patients who are predicted to obtain a cancer therapeutic effect by the immune checkpoint inhibitor and selecting a different treatment method for patients who are not predicted to obtain a therapeutic effect, it is possible to minimize the risk of only causing adverse effects such as irAE and HPD without obtaining the therapeutic effect of the immune checkpoint inhibitor. EXAMPLES

[0053] The present invention will be described in more detail below with reference to examples, although the present invention is not limited to the following examples.

[0054] <Method> Peripheral T cells and tumor biopsy samples were collected from eight patients with advanced gastrointestinal cancer (four patients with esophageal cancer, three with colon cancer, and one with gastric cancer) before and after anti-PD-1 antibody administration, and their TCR repertoires were analyzed. For peripheral blood T cells, blood was drawn before the start of anti-PD-1 antibody administration, and CD8-positive T cells were purified using magnetic beads. For tumor biopsies, samples collected before the start of anti-PD-1 antibody administration were used.

[0055] Library preparation for TCR repertoire analysis was performed by modifying the method described in WO 2019 / 189383. The outline of the library preparation method is shown in Figure 1. RNA with polyA sequences was recovered from the lysate of CD8 positive T cells isolated and purified from peripheral blood samples and from the total RNA of unfractionated tumor biopsy tissue, and reverse transcription and addition of an artificial adapter sequence (Adaptor 2 in Figure 1) to the reverse transcription product were performed by modifying the method described in a previous study (Shichino S, Ueha S, Hashimoto S, Otsuji M, Abe J, Tsukui T, et al. Transcriptome network analysis identifies protective role of the LXR / SREBP-1c axis in murine pulmonary fibrosis. JCI Insight. 2019;4.). The cDNA synthesized on the magnetic beads was subjected to a terminal deoxynucleotidyl transferase reaction using ddCTP in a ratio of ddCTP:dCTP = 1:20 in a reaction solution containing CoCl2 to add a poly C tail to the end of the cDNA. For second strand synthesis, trP1-9G, in which the 3' complementary homopolymer portion is 9 bases long, was used. Next, a PCR reaction was performed using a primer (trP1) set to an artificial adapter sequence (Adaptor 2 in Figure 1) and a common primer for the TCR constant region (C region) (Trac_ex and Trbc_ex were combined for the 1st PCR, and Trac_in-Bio and Trbc_in-Bio were combined for the 2nd PCR, and used as a common primer for the TCR C region), to amplify the variable region including the CDR3 region of the TCR α chain and β chain. After amplification, the purified TCR gene products were enzymatically fragmented and then adapter sequences for sequencing (P1 Adaptor-1 & P1 Adaptor-2, Adaptor P1 in Figure 1) were added. Fragmentation of these gene products and addition of adapter sequences were performed using the NEBNext FS DNA Library Prep Kit (New England Biolabs, #E7805).A barcode sequence (BC) was added to the TCR gene product after the addition of the adapter sequence for sequencing (primers: trP1 & IonA-BC-Trbc, 3rd PCR in Figure 1) to add a barcode sequence (BC) to distinguish samples during sequencing. The TCR library created by the above procedure was sequenced using an Ion Genestudio S5 Sequencer (Thermo Fisher Scientific, # A27759, # A27766, # A38194). The sequences of the primers, adapters, etc. used are shown in Table 1 below.

[0056] [Table 1]

[0057] The TCR sequence information obtained by sequencing was aligned with standard sequences (reference sequences) of the V, D, and J regions using an algorithm called MiXCR (Bolotin DA, Poslavsky S, Mitrophanov I, Shugay M, Mamedov IZ, Putintseva E V., et al. MiXCR: Software for comprehensive adaptive immunity profiling. Nat Methods. 2015;12:380-1.), and then assembled into T cell clones defined by common V region, J region, and CDR3 nucleotide sequences.

[0058] The TCR repertoire obtained in this way was analyzed using the program VDJtools-1.2.1 (Shugay M, Bagaev D V., Turchaninova MA, Bolotin DA, Britanova O V., Putintseva E V., et al. VDJtools: Unifying Post-analysis of T Cell Receptor Repertoires. PLoS Comput Biol. 2015;11:1-16.). When multiple biopsy tissues were obtained from the same patient, the TCR repertoire obtained by pooling the T cell clones detected in each biopsy tissue was defined as the "tumor TCR repertoire" of that patient. For this operation, the "PoolSamples" function of the VDJtools mentioned above was used.

[0059] T cell clones detected in common between the TCR repertoire of peripheral blood CD8 positive T cells and the tumor TCR repertoire of the same patient were identified as "peripheral blood-tumor overlap clones (hereafter, overlap clones)." Assuming that the total amount of overlap clones present in peripheral blood reflects the therapeutic effect of anti-PD-1 antibody administration, the total frequency of overlap clones in the peripheral blood CD8 positive T cell repertoire was calculated using the following formula. Total frequency (%) of overlapping clones in the TCR repertoire of peripheral blood CD8+ T cells = [Number of CD8+ T cells belonging to overlapping clones in peripheral blood] / [Number of CD8+ T cells in peripheral blood] × 100

[0060] <Result> After administration of anti-PD-1 antibody, PR (partial response) was observed in 2 of 8 cases, SD (stable disease) in 3 cases, and PD (progress disease) in 3 cases. The rate of change in tumor size before and after the start of treatment in each case is shown in Table 2 below. The rate of change in tumor size after the start of treatment is taken on the horizontal axis, and the total frequency of overlapping clones in the peripheral blood CD8-positive T cell repertoire before the start of treatment is taken on the vertical axis. When each patient was plotted, as shown in Figure 2, a correlation was observed between the two. When the cutoff value was set at around 10% of the total frequency of overlapping clones in the peripheral blood CD8-positive T cell repertoire before the start of treatment, a reduction in tumor size was observed after the start of treatment in 3 of 4 cases that were above the cutoff value, while all 4 cases that were below the cutoff value showed an increase in tumor size or a reduction in tumor size of less than 10%. This result indicates that it is possible to predict whether or not PD-1 inhibitor therapy will be effective in treating cancer, based on the total overlapping clone frequency in the peripheral blood of cancer patients before starting PD-1 inhibitor therapy. Specifically, by setting the cutoff value for the total overlapping clone frequency within the range of approximately 7-12%, for example, within the range of 8-11.5% or 8.5-11.5%, patients with a total overlapping clone frequency equal to or greater than the cutoff value can be identified as patients who are likely to benefit from PD-1 inhibitor therapy.

[0061] [Table 2]

Claims

1. A method for assisting in prediction of a cancer therapeutic effect by an immune checkpoint inhibitor, comprising: A step of obtaining information on the T cell receptor (TCR) repertoire of tumor-infiltrating T cells in the tumor tissue and the TCR repertoire of CD8-positive T cells in the peripheral blood, using a tumor tissue specimen and a peripheral blood specimen collected from a cancer patient before the start of treatment with an immune checkpoint inhibitor to be predicted for therapeutic effect; Comparing the TCR repertoire of tumor-infiltrating T cells and the TCR repertoire of peripheral blood CD8-positive T cells, and identifying T cell clones that are commonly detected in both as overlapping clones; and A step of calculating the total frequency of overlapping clones in the TCR repertoire of peripheral blood CD8-positive T cells When the total frequency of the overlapping clones is equal to or greater than a predetermined cutoff value, prediction that the cancer patient will be treated with the immune checkpoint inhibitor is assisted.

2. The method of claim 1, wherein the immune checkpoint inhibitor is a PD-1 inhibitor.

3. 3. The method of claim 2, wherein the PD-1 inhibitor is an anti-PD-1 antibody, an anti-PD-L1 antibody, or an anti-PD-L2 antibody.

4. The method according to any one of claims 1 to 3, wherein the total frequency of the overlapping clones is calculated by the following formula, and a cutoff value is set within the range of 7% to 12%. Total frequency (%) of overlapping clones in the TCR repertoire of peripheral blood CD8+ T cells = [Number of CD8+ T cells belonging to overlapping clones in peripheral blood] / [Number of CD8+ T cells in peripheral blood] x 100

5. The method according to any one of claims 1 to 4, wherein the TCR repertoire of tumor-infiltrating T cells in the tumor tissue is a TCR repertoire of tumor-infiltrating T cells including T cells that are not CD8 positive, or a TCR repertoire of tumor-infiltrating CD8 positive T cells.

Images