Test method for predicting effect of treatment of hepatocellular carcinoma

Abstract

[Problem] The problem addressed is to discover clinically useable non-invasive index that makes it possible to predict the effect of combination therapy by multiple drugs, including immune checkpoint inhibitors, in hepatocellular carcinoma. [Solution] Using the peripheral blood of a hepatocellular carcinoma patient as a sample, the effect of combination therapy by multiple drugs, including immune checkpoint inhibitors, is predicted by a test method that includes at least one of steps (a)-(d). For example, the effect of combination therapy by an immune checkpoint inhibitor and an angiogenesis inhibitor is predicted. Preferably, the method includes a step for comparing the measured value with a reference value. (a) A step for measuring the relative cell count of CD14+ cells, (b) a step for measuring the expression level of a gene such as FABP5 in CD14+ cells, (c) a step for measuring the relative cell count of CD4+ cells that are positive for CCR7, etc., (d) a step for measuring the expression level of a gene such as FOS in CD4+ cells.

Description

Testing methods for predicting the effectiveness of treatment for hepatocellular carcinoma

[0001] The present invention relates to a testing method and testing device for predicting the effectiveness of treatment for hepatocellular carcinoma using combination therapy with multiple drugs including immune checkpoint inhibitors.

[0002] Drug therapy for unresectable advanced hepatocellular carcinoma has made significant progress in recent years. In Japan, the drugs currently available for first-line treatment include a combination of atezolizumab (anti-PD-L1 antibody) and bevacizumab (anti-VEGF antibody), a combination of durvalumab (anti-PD-L1 antibody) and tremelimumab (anti-CTLA-4 antibody), durvalumab, lenvatinib, and sorafenib. Second-line treatment options include regorafenib, cabozantinib, and ramucirumab. There are a wide variety of options for both first-line and second-line treatment (Non-Patent Document 1). However, the response rate for each drug is approximately 30%, and due to reduced hepatic reserve, only a small number of patients can receive all drugs sequentially. Furthermore, the mechanisms of action of drugs vary, including immune checkpoint inhibitors (ICIs), multikinase inhibitors, and angiogenesis inhibitors.

[0003] Therefore, to improve the prognosis of liver cancer patients, "personalized medicine," which involves not only appropriate assessment of hepatic reserve but also the selection of the most appropriate drug for each cancer, is crucial. To achieve this, the development of biomarkers that can guide drug selection is essential. Genome sequencing of numerous liver cancer cases has identified numerous driver genes, primarily TERT, TP53, and CTNNB1, demonstrating high intertumor heterogeneity in liver cancer. Furthermore, liver cancer is classified into various subtypes based on driver gene abnormalities and differences in immune dynamics in the tumor microenvironment. The relationship between these molecular immune profiles and the efficacy of drug therapy has attracted attention. A research group including the present inventors reported that approximately 25% of liver cancers are hot tumors with high immune infiltration, which can be expected to be highly responsive to ICI, while liver cancers with CTNNB1 mutations are cold tumors with low intratumoral immune infiltration, which may result in poor efficacy (Non-Patent Document 2).

[0004] However, in actual clinical practice, treatment for liver cancer is based on a definitive diagnosis using diagnostic imaging, so there are usually few opportunities to collect tumor tissue before starting drug therapy, making it difficult to understand the immune dynamics within the tumor. While attempts have been reported to predict the therapeutic efficacy of immune checkpoint inhibitors based on the immune dynamics within the tumor for other cancers (Reference 3), there are no non-invasive markers for clinical use that can predict the efficacy of combination immunotherapy, such as combination therapy with atezolizumab (anti-PD-L1 antibody) and bevacizumab (anti-VEGF antibody), for hepatocellular carcinoma.

[0005] Liver Cancer Treatment Guidelines 2021, edited by the Japan Society of Hepatology, Kanehara Publishing Co., Ltd. Murai H et al., Hepaotlogy 2023 Jan 1;77(1):77-91. doi: 10.1002 / hep.32573. Kumagai S et al., Nat Immunol. 2020 Nov;21(11):1346-1358. doi: 10.1038 / s41590-020-0769-3.

[0006] The objective of this study is to find a clinically applicable non-invasive indicator that can predict the efficacy of combination therapy with multiple drugs, including immune checkpoint inhibitors, in hepatocellular carcinoma.

[0007] In order to solve the above problems, the present inventors have conducted research into a testing method using a patient's peripheral blood as a sample, based on the belief that the dynamics of the patient's systemic immune system influences the efficacy of a combination therapy of multiple drugs including an immune checkpoint inhibitor (hereinafter also referred to as combined immunotherapy). The present invention includes the following aspects.1. A testing method for predicting the efficacy of a treatment for hepatocellular carcinoma using a combination therapy of multiple drugs including an immune checkpoint inhibitor, using peripheral blood from a hepatocellular carcinoma patient as a sample, the testing method comprising at least one step selected from the group consisting of the following (a) to (d): (a) measuring the relative cell number of CD14-positive cells; (b) measuring the expression level of at least one gene selected from gene group P in CD14-positive cells; (c) measuring the relative cell number of CD4-positive cells positive for a protein encoded by at least one gene selected from gene group Q; and (d) measuring the expression level of at least one gene selected from gene group Q in CD4-positive cells. Here, the gene group P includes LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, IL1RN, FABP5, JARID2, AREG, CXCL16, CCL2, CTSL, CCL7, IL1B, G0S2, IER3, ASPH, CCL3, MAFB, IFI30, SGK1, H consisting of MOX1, CEBPD, RGCC, CST3, ATP2B1, FCN1, MARCKS, CLEC7A, FTH1, DUSP6, S100A11, CD14, S100A9, HLA-DRA, MTSS1, CD300E, ATP2B1-AS1, DOCK4, SERPINA1, PLAUR, AIF1, TYROBP, SAT1, S100A10, FTL, and GRN; 1. A testing method wherein the gene group Q consists of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, SELL, TCF7, FOS, SOCS3, FHIT, MYC, GIMAP5, TRABD2A, PIM2, ITGA6, ADTRP, EDA, LTB, RPS13, RPS3A, GAS5, NOSIP, RPS5, EEF1B2, RPL32, LDHB, RPL22, RPL5, RPS12, RPL30, RPS8, RPL11, EEF1G, RPS23, SNHG29, RPL35A, EIF3E, RPS28, RPL39, and RPL9. 2. The method according to item 1, wherein the multiple drugs including an immune checkpoint inhibitor are an immune checkpoint inhibitor and an angiogenesis inhibitor.3. The method of preceding item 1 or 2, wherein the immune checkpoint inhibitor is a drug that inhibits the PD-1 / PD-L1 pathway. 4. The method of preceding item 3, wherein the drug that inhibits the PD-1 / PD-L1 pathway is an anti-PD-L1 antibody or an anti-PD-1 antibody, and the angiogenesis inhibitor is an anti-VEGF antibody. 5. The method of preceding item 1, wherein the multiple drugs including the immune checkpoint inhibitor are two types of immune checkpoint inhibitors. 6. The method of preceding item 5, wherein the two types of immune checkpoint inhibitors are a drug that inhibits the PD-1 / PD-L1 pathway and a drug that inhibits CTLA-4 signaling. 7. The method of preceding item 6, wherein the drug that inhibits the PD-1 / PD-L1 pathway is an anti-PD-L1 antibody or an anti-PD-1 antibody, and the drug that inhibits CTLA-4 signaling is an anti-CTLA-4 antibody. 8. The method of any one of preceding items 1 to 7, comprising at least step (a). 9. 9. The method according to any one of preceding items 1 to 8, wherein the CD14-positive cells in step (a) are positive cells for a protein encoded by at least one gene selected from gene group P. 10. The method according to preceding item 9, wherein gene group P is gene group Pa2, which is a gene group consisting of LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, and IL1RN. 11. The method according to any one of preceding items 1 to 10, wherein step (a) comprises single-cell RNA sequencing analysis. 12. The method according to any one of preceding items 1 to 7, comprising at least step (c). 13. 13. The method according to any one of the preceding paragraphs 1 to 12, wherein the gene group Q in step (c) is gene group Qc2, which is a gene group consisting of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, and TCF7. 14. The method according to any one of the preceding paragraphs 1 to 13, wherein step (c) comprises single-cell RNA sequencing analysis. 15. The method according to any one of the preceding paragraphs 1 to 7, which comprises at least step (b).16. The method of any one of preceding paragraphs 1 to 15, wherein the gene group P in step (b) is gene group Pb2, which consists of FABP5, JARID2, PID1, AREG, LGALS2, CXCL16, and TIMP1. 17. The method of any one of preceding paragraphs 1 to 16, wherein step (b) comprises single-cell RNA sequencing analysis. 18. The method of any one of preceding paragraphs 1 to 7, which includes at least step (d). 19. The method of any one of preceding paragraphs 1 to 18, wherein the gene group Q in step (d) is gene group Qd2, which consists of at least one gene selected from the group consisting of FOS, LEF1, CCR7, TCF7, SELL, and TSHZ2. 20. 21. The method of any one of items 1 to 19, wherein step (d) comprises single-cell RNA sequencing analysis. 21. The method of any one of items 1 to 20, further comprising the following step (e) or step (f): (e) calculating the ratio of the relative cell number obtained in step (c) to the relative cell number obtained in step (a), or (f) calculating the ratio of the relative cell number obtained in step (a) to the relative cell number obtained in step (c). 22. The method of any one of items 1 to 21, wherein at least one of steps (a) to (f) comprises a step of comparing a measured value with a reference value, and the effect of treatment of hepatocellular carcinoma with combination therapy of multiple drugs including an immune checkpoint inhibitor is determined based on the comparison result. 23. 23. The method according to the preceding item 22, wherein the treatment of hepatocellular carcinoma with the combined therapy of an immune checkpoint inhibitor and an angiogenesis inhibitor is determined to be effective when the ratio of the relative cell numbers obtained in step (e) is higher than a reference value, or when the ratio of the relative cell numbers obtained in step (f) is lower than a reference value.24. A testing method for predicting the effectiveness of a treatment for hepatocellular carcinoma using a combination therapy of multiple drugs including an immune checkpoint inhibitor, using peripheral blood from a hepatocellular carcinoma patient as a sample, the testing method using at least one indicator selected from the group consisting of the following (1) to (6): (1) the relative cell count of CD14-positive cells; (2) the expression level of at least one gene selected from gene group P in CD14-positive cells; (3) the relative cell count of CD4-positive cells and positive for a protein encoded by at least one gene selected from gene group Q; (4) the expression level of at least one gene selected from gene group Q in CD4-positive cells; (5) the ratio of index (3) to index (1); (6) the ratio of index (1) to index (3). Here, the gene group P includes LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, IL1RN, FABP5, JARID2, AREG, CXCL16, CCL2, CTSL, CCL7, IL1B, G0S2, IER3, ASPH, CCL3, MAFB, IFI30, SGK1, H consisting of MOX1, CEBPD, RGCC, CST3, ATP2B1, FCN1, MARCKS, CLEC7A, FTH1, DUSP6, S100A11, CD14, S100A9, HLA-DRA, MTSS1, CD300E, ATP2B1-AS1, DOCK4, SERPINA1, PLAUR, AIF1, TYROBP, SAT1, S100A10, FTL, and GRN; 24. A testing method wherein the gene group Q consists of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, SELL, TCF7, FOS, SOCS3, FHIT, MYC, GIMAP5, TRABD2A, PIM2, ITGA6, ADTRP, EDA, LTB, RPS13, RPS3A, GAS5, NOSIP, RPS5, EEF1B2, RPL32, LDHB, RPL22, RPL5, RPS12, RPL30, RPS8, RPL11, EEF1G, RPS23, SNHG29, RPL35A, EIF3E, RPS28, RPL39, and RPL9. 25. The method according to item 24 above, wherein the multiple drugs including an immune checkpoint inhibitor are an immune checkpoint inhibitor and an angiogenesis inhibitor.26. The method of preceding item 25, wherein the immune checkpoint inhibitor is a drug that inhibits the PD-1 / PD-L1 pathway. 27. The method of preceding item 26, wherein the drug that inhibits the PD-1 / PD-L1 pathway is an anti-PD-L1 antibody or an anti-PD-1 antibody, and the angiogenesis inhibitor is an anti-VEGF antibody. 28. The method of preceding item 24, wherein the multiple drugs including the immune checkpoint inhibitor are two types of immune checkpoint inhibitors. 29. The method of preceding item 28, wherein the two types of immune checkpoint inhibitors are a drug that inhibits the PD-1 / PD-L1 pathway and a drug that inhibits CTLA-4 signaling. 30. The method of preceding item 29, wherein the drug that inhibits the PD-1 / PD-L1 pathway is an anti-PD-L1 antibody or an anti-PD-1 antibody, and the drug that inhibits CTLA-4 signaling is an anti-CTLA-4 antibody. 31. 31. The method according to any one of items 24 to 30 above, wherein the CD14-positive cells in the indicator (1) are positive cells for a protein encoded by at least one gene selected from the gene group P.32. A testing device for predicting the efficacy of a hepatocellular carcinoma treatment using a combination therapy of multiple drugs including an immune checkpoint inhibitor, using peripheral blood from a hepatocellular carcinoma patient as a sample, the device comprising at least one detection unit selected from the group consisting of the following (h) to (k): (h) a relative cell number detection unit that detects the relative cell number of CD14-positive cells; (i) an expression level detection unit that detects the expression level of at least one gene selected from gene group P in CD14-positive cells; (j) a relative cell number detection unit that detects the relative cell number of CD4-positive cells and positive for a protein encoded by at least one gene selected from gene group Q; and (k) an expression level detection unit that detects the expression level of at least one gene selected from gene group Q in CD4-positive cells. Here, the gene group P includes LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, IL1RN, FABP5, JARID2, AREG, CXCL16, CCL2, CTSL, CCL7, IL1B, G0S2, IER3, ASPH, CCL3, MAFB, IFI30, SGK1, H consisting of MOX1, CEBPD, RGCC, CST3, ATP2B1, FCN1, MARCKS, CLEC7A, FTH1, DUSP6, S100A11, CD14, S100A9, HLA-DRA, MTSS1, CD300E, ATP2B1-AS1, DOCK4, SERPINA1, PLAUR, AIF1, TYROBP, SAT1, S100A10, FTL, and GRN; the gene group Q consists of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, SELL, TCF7, FOS, SOCS3, FHIT, MYC, GIMAP5, TRABD2A, PIM2, ITGA6, ADTRP, EDA, LTB, RPS13, RPS3A, GAS5, NOSIP, RPS5, EEF1B2, RPL32, LDHB, RPL22, RPL5, RPS12, RPL30, RPS8, RPL11, EEF1G, RPS23, SNHG29, RPL35A, EIF3E, RPS28, RPL39, and RPL9.33. The device according to preceding paragraph 32, further comprising an output unit that compares the detected value with a reference value and indicates the effect of treatment of hepatocellular carcinoma by combination therapy with multiple drugs including an immune checkpoint inhibitor. 34. The device according to preceding paragraph 32 or 33, wherein the multiple drugs including the immune checkpoint inhibitor are an immune checkpoint inhibitor and an angiogenesis inhibitor. 35. The device according to preceding paragraph 34, wherein the immune checkpoint inhibitor is a drug that inhibits the PD-1 / PD-L1 pathway. 36. The device according to preceding paragraph 35, wherein the drug that inhibits the PD-1 / PD-L1 pathway is an anti-PD-L1 antibody or an anti-PD-1 antibody, and the angiogenesis inhibitor is an anti-VEGF antibody. 37. The device according to preceding paragraph 32 or 33, wherein the multiple drugs including the immune checkpoint inhibitor are two types of immune checkpoint inhibitors. 38. The device according to preceding paragraph 37, wherein the two types of immune checkpoint inhibitors are a drug that inhibits the PD-1 / PD-L1 pathway and a drug that inhibits CTLA-4 signaling. 39. 40. The device according to any one of items 32 to 39, wherein the CD14-positive cells in (h) are positive for a protein encoded by at least one gene selected from gene group P.

[0008] The method of the present invention makes it possible to predict the efficacy of combination therapy with multiple drugs, including immune checkpoint inhibitors, non-invasively, without the need to collect liver cancer tissue. Performing the test of the present invention before the start of the combination therapy enables optimal drug selection. Performing the test of the present invention during the combination therapy enables the selection of a change in drug. While there are many drug options for drug therapy for hepatocellular carcinoma, the response rate for each drug is only about 30%, and only a small number of patients can use all drugs sequentially due to reduced hepatic reserve. Therefore, the present invention enables optimal drug selection and contributes to improving patient prognosis.

[0009] The present invention can also stratify hepatocellular carcinoma patients. For example, the present invention can divide patients into those who will respond to a combination therapy of an agent that inhibits the PD-1 / PD-L1 pathway and an angiogenesis inhibitor, and those who will respond to a combination therapy of an agent that inhibits the PD-1 / PD-L1 pathway and an agent that inhibits CTLA-4 signaling.

[0010] In clinical settings, liver cancer treatment is based on a definitive diagnosis using diagnostic imaging, so there are usually few opportunities to collect tumor tissue before starting drug therapy. The present invention makes it possible to predict the efficacy of combination therapy using peripheral blood as a sample, thereby reducing the burden on patients.

[0011] Figure 1 shows the results of CITE-seq (UMAP analysis) performed on 231,383 cells obtained from 20 PBMC samples taken before and during treatment from five responders (hereinafter referred to as the responder group) and five non-responders (hereinafter referred to as the non-responder group) of atezolizumab and bevacizumab combination therapy. Cells were classified into 24 subtypes based on differentially expressed genes. The two major cell groups, the left group representing T cells and NK cells, and the right group representing monocyte-related cells, are shown in Figure 2. Figure 3 shows the relative cell counts of cluster 4 [CD14 monocytes] (hereinafter referred to as cluster CD14) shown in Figures 1 and 4(A). The relative cell counts in the responder group (CRPR) were significantly lower than those in the non-responder group (PD). The vertical axis, "Frequency," indicates the ratio (frequency) of cluster CD14 cells to the total number of analyzed PBMC cells. "Pre" indicates the pretreatment sample, and "Post" indicates the treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 4(A) shows the same UMAP analysis results as Figure 1. The circled area in the monocyte-related cell group on the right indicates cluster CD14. In Figure 4(B), the circled area indicates cluster 4 [CD4T] (hereinafter also referred to as cluster CD4T). Cluster CD4T was obtained by reclustering 124,058 cells belonging to the T cell and NK cell groups based on their intercellular expression profiles. Figure 5 shows the relative cell counts of cluster CD4T shown in Figure 4(B). The relative cell counts in the responder group (CRPR) were significantly higher than in the non-responder group (PD). The vertical axis, "Frequency," indicates the ratio (frequency) of cluster CD4 T cell counts to the number of T cells and NK cells. "Pre" indicates the pre-treatment sample, and "Post" indicates the treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 6 shows the results of comparing the relative cluster CD14 cell counts in pre-treatment samples from patients receiving atezolizumab and bevacizumab combination therapy with the reference value of 0.099, dividing the patients into a high-value group (5 cases) and a low-value group (5 cases), and comparing the progression-free survival time. The horizontal axis indicates the number of days after the start of treatment, and the vertical axis indicates the progression-free survival rate.The progression-free survival period was significantly extended in the low-value group. Statistical analysis was performed using the log-rank test. Figure 7 shows the results of comparing the relative cell counts of cluster CD4 T cells in pretreatment samples from patients receiving atezolizumab and bevacizumab combination therapy with a baseline value of 0.074, dividing the patients into high-value groups (5 cases) and low-value groups (5 cases), and comparing progression-free survival. The horizontal axis represents the number of days since the start of treatment, and the vertical axis represents the progression-free survival rate. The progression-free survival period was significantly extended in the high-value group. Statistical analysis was performed using the log-rank test. Figure 8 shows the results of comparing the ratio of the relative cell counts of cluster CD4 T cells to the relative cell counts of cluster CD14 cells in pretreatment samples from patients receiving atezolizumab and bevacizumab combination therapy with a baseline value of 0.65, dividing the patients into high-value groups (5 cases) and low-value groups (5 cases), and comparing progression-free survival. The horizontal axis indicates the number of days since the start of treatment, and the vertical axis indicates the progression-free survival rate. The progression-free survival period was significantly longer in the high-value group. Statistical analysis was performed using the log-rank test. Figure 9 shows the relative cell counts of CD14-positive cells in PBMCs from patients before and during treatment with atezolizumab and bevacizumab combination therapy. The relative cell counts in the responder group (CRPR) were significantly lower than in the non-responder group (PD). (A) shows the relative cell counts of cells with an intracellular CD14-encoding mRNA level greater than 1, and (B) shows the relative cell counts of cells with an anti-CD14 antibody level greater than 1 bound to the cell surface antigen CD14. Freq on the vertical axis indicates the ratio (frequency) of CD14-positive cells to the total PBMC cell count. Pre indicates the pretreatment sample, and Post indicates the treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 10 shows the relative cell counts of CD14-positive cells and positive for each protein encoded by each gene (A) to (F) in PBMCs of patients before and during treatment with atezolizumab and bevacizumab combination therapy. The genes are (A) LYZ, (B) CXCL8, (C) TIMP1, (D) LGALS2, (E) CD68, and (F) LGALS3. In all cases, the relative cell counts in the responder group (CRPR) were significantly lower than in the non-responder group (PD). CD14 positivity and positive for the proteins encoded by each gene were detected by mRNA. The freq on the vertical axis indicates the ratio (frequency) of each cell count to the total PBMC cell count.Pre indicates pretreatment samples, and Post indicates treatment samples. Graphs show mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 11 shows the relative number of CD14-positive cells and positive for each protein encoded by the genes (A) to (F) in PBMCs from patients before and during treatment with atezolizumab and bevacizumab combination therapy. The genes are: (A) LYZ, (B) CXCL8, (C) TIMP1, (D) LGALS2, (E) CD68, and (F) LGALS3. In both cases, the relative number of cells in the responder group (CRPR) was significantly lower than in the non-responder group (PD). CD14 positivity was detected using an anti-CD14 antibody, and positive proteins encoded by each gene were detected using mRNA. Freq on the vertical axis indicates the ratio (frequency) of each cell number to the total PBMC cell count. "Pre" indicates the pre-treatment sample, and "Post" indicates the on-treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 12 shows the relative number of CD14-positive cells and positive for each protein encoded by the genes (A) to (D) in PBMCs from patients before and during treatment with atezolizumab and bevacizumab combination therapy. The genes are (A) CXCL2, (B) PID1, (C) APOBEC3A, and (D) PPARG. In all cases, the relative number of cells in the responder group (CRPR) was significantly lower than in the non-responder group (PD). CD14 positivity and positive for the proteins encoded by each gene were detected by intracellular mRNA. The "freq" on the vertical axis indicates the ratio (frequency) of each cell count to the total PBMC cell count. "Pre" indicates the pre-treatment sample, and "Post" indicates the on-treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 13 shows the relative cell counts of CD14-positive, CD3-negative cells that are positive for each protein encoded by each gene (A) to (F) in PBMCs from patients before and during treatment with atezolizumab and bevacizumab combination therapy. The genes are (A) CXCL2, (B) PID1, (C) PHLDA1, (D) APOBEC3A, (E) IL1RN, and (F) PPARG. Both figures show that the relative cell counts in the responder group (CRPR) were significantly lower than in the non-responder group (PD).CD14-positive and CD3-negative cells were detected using antibodies against each gene, and positive cells for the proteins encoded by each gene were detected by mRNA analysis. The freq on the vertical axis indicates the ratio of each cell count to the total PBMC cell count (frequency). Pre indicates the pretreatment sample, and Post indicates the treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 14 shows the relative number of CD4-positive cells and positive cells for each protein encoded by each gene (A) to (G) in PBMCs from patients before and during treatment with atezolizumab and bevacizumab combination therapy. The genes are: (A) CCR7; (B) LEF1; (C) MAL; (D) TSHZ2; (E) PRKCQ-AS1; (F) RCAN3; (G) TCF7. In both cases, the relative cell counts in the responder group (CRPR) were higher than in the non-responder group (PD). CD4 positivity was detected using an anti-CD4 antibody, and positivity for the proteins encoded by each gene was detected using mRNA. The freq on the vertical axis indicates the ratio (frequency) of each cell count to the total PBMC cell count. Pre indicates the pretreatment sample, and Post indicates the treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 15 shows the relative cell counts of CD4-positive, HLA-DR-negative, and positive for each protein encoded by each gene (A) to (G) in PBMCs from patients before and during treatment with atezolizumab and bevacizumab combination therapy. The genes are: (A) CCR7, (B) LEF1, (C) MAL, (D) TSHZ2, (E) PRKCQ-AS1, (F) RCAN3, and (G) TCF7. In each case, the relative cell counts in the responder group (CRPR) were higher than in the non-responder group (PD). CD4-positive and HLA-DR-negative cells were detected using antibodies against each gene, and positive cells for the proteins encoded by each gene were detected using mRNA. The freq on the vertical axis indicates the ratio (frequency) of each cell count to the total PBMC cell count. Pre indicates the pretreatment sample, and Post indicates the treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 16 shows the expression levels of each gene (A) to (G) in CD14-positive cells in pretreatment samples from atezolizumab and bevacizumab combination therapy.The genes are: (A) FABP5, (B) JARID2, (C) PID1, (D) AREG, (E) LGALS2, (F) CXCL16, (G) TIMP1. In all cases, gene expression levels in the responder group (CRPR) were significantly lower than in the non-responder group (PD). CD14 positivity was detected by mRNA. "Pre" indicates pretreatment samples. Statistical analysis was performed using the Mann-Whitney U test. Figure 17 shows the expression levels of each gene (A) to (F) in CD4-positive T cells in pretreatment samples from atezolizumab and bevacizumab combination therapy. The genes are: (A) FOS, (B) LEF1, (C) CCR7, (D) TCF7, (E) SELL, (F) TSHZ2. In both cases, gene expression levels in the responder group (CRPR) were significantly higher than in the non-responder group (PD). CD4-positive T cells were detected as CD4-positive HLA-DR-negative cells using antibodies against each. "Pre" indicates pretreatment samples. Statistical analysis was performed using the Mann-Whitney U test. Figure 18 shows the relative cell counts of Cluster 4 [CD14 Monocyte] (hereinafter also referred to as Cluster CD14 [Dur / Tre]) in patients receiving durvalumab and tremelimumab combination therapy. Cluster CD14 [Dur / Tre] is one of the 24 subtypes identified by UMAP analysis results obtained by performing CITE-seq on 14 samples of PBMCs before and during treatment from four responders and five non-responders. The relative cell counts in the responder group (CRPR) are higher than in the non-responder group (PD). The vertical axis, freq, indicates the ratio (frequency) of cluster CD14 [Dur / Tre] cell counts to the number of analyzed PBMC cells. Pre indicates pre-treatment samples, and Post indicates treatment samples. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 19 shows the progression-free survival rate of patients with high relative cluster CD14 cell counts in pre-treatment samples (high group). Atez / Bev indicates combination therapy of atezolizumab and bevacizumab, and Dur / Tre indicates combination therapy of durvalumab and tremelimumab. The horizontal axis indicates the number of days after the start of treatment, and the vertical axis indicates progression-free survival rate. Atez / Bev did not extend progression-free survival, but Dur / Tre did.Statistical analysis was performed using the log-rank test. Figure 20 shows the progression-free survival rate of patients with low relative cluster CD14 cell counts in pre-treatment samples (low group). Atez / Bev indicates the combination of atezolizumab and bevacizumab, and Dur / Tre indicates the combination of durvalumab and tremelimumab. The horizontal axis represents the number of days after the start of treatment, and the vertical axis represents the progression-free survival rate. Atez / Bev extended progression-free survival, but Dur / Tre did not. Statistical analysis was performed using the log-rank test. Figure 21 shows the relative cell counts of CD14-positive cells in PBMCs of patients before and during treatment with durvalumab and tremelimumab combination therapy. It shows that the relative cell counts in the responder group (CRPR) were higher than in the non-responder group (PD). Cells with an intracellular CD14-encoding mRNA level greater than 1 were defined as CD14-positive cells. The vertical axis, freq, indicates the ratio (frequency) of CD14-positive cells to the total PBMC cell count. Pre indicates the pretreatment sample, and Post indicates the treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 22 shows the relative numbers of CD14-positive cells and positive for each protein encoded by the genes (A) to (H) in PBMCs from patients before and during treatment with durvalumab and tremelimumab combination therapy. The genes are: (A) PID1; (B) APOBEC3A; (C) PPARG; (D) LYZ; (E) CXCL8; (F) TIMP1; (G) LGALS2; (H) LGALS3. Both figures show that the relative cell counts in the responder group (CRPR) were higher than in the non-responder group (PD). CD14 positivity was detected using an anti-CD14 antibody, and positivity for the proteins encoded by each gene was detected by mRNA. The freq on the vertical axis indicates the ratio of each cell count to the total PBMC cell count (frequency). Pre indicates the pretreatment sample, and Post indicates the treatment sample. The graph shows the mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test. Figure 23 shows the expression levels of each gene (A) to (F) in CD14-positive cells in pretreatment samples from durvalumab and tremelimumab combination therapy. The genes are: (A) FABP5; (B) JARID2; (C) PID1; (D) AREG; (E) CXCL16; and (F) TIMP1.In both cases, the gene expression levels in the responder group (CRPR) were higher than in the non-responder group (PD). CD14 positivity was detected by mRNA. Statistical analysis was performed using the Mann-Whitney U test.

[0012] Predicting the efficacy of hepatocellular carcinoma treatment with a combination therapy of multiple drugs, including immune checkpoint inhibitors, can be said to be predicting the patient's sensitivity to the combination therapy. It can also be said to be predicting the patient's responsiveness to the combination therapy. Although the combination therapy is used in many hepatocellular carcinoma patients, not all patients achieve the same effect, with some patients exhibiting high sensitivity and others exhibiting low sensitivity. Furthermore, even if a therapeutic effect is observed, some patients later develop resistance.

[0013] The testing method of the present invention is a testing method for predicting the therapeutic effect of the combination therapy, and uses peripheral blood from a patient with hepatocellular carcinoma as a sample. Preferably, the testing method of the present invention uses peripheral blood collected from the patient before the start of the combination therapy or during the combination therapy. If the test determines that the combination therapy is effective, a good anti-cancer effect can be obtained by starting or continuing the combination therapy. If the test determines that the combination therapy is not effective, administration of another drug or another treatment method can be selected. Thus, the present invention provides an appropriate treatment method for patients, making it possible to continue treatment without reducing hepatic reserve function.

[0014] Immune checkpoint inhibitors are conceptually drugs that reactivate exhausted effector T cells and are more effective in cancers with high expression of tumor-infiltrating cytotoxic T cells and PD-L1 (JAMA Oncol. 2019 Aug 1;5(8):1195-1204.). The immune checkpoint inhibitors used in the combination therapy evaluated by the testing method of the present invention are not particularly limited, and examples include at least one antibody against an immune checkpoint molecule or an antibody against its ligand, such as an anti-PD-L1 antibody, anti-PD-1 antibody, anti-PD-L2 antibody, anti-CTLA-4 antibody, anti-TIM-3 antibody, anti-LAG-3 antibody, or anti-TIGIT antibody. Suitable immune checkpoint inhibitors include drugs that inhibit the PD-1 / PD-L1 pathway. Anti-PD-L1 antibodies and anti-PD-1 antibodies are more preferred, and anti-PD-L1 antibodies are even more preferred. Examples of anti-PD-L1 antibodies include atezolizumab, durvalumab, and avelumab.

[0015] In the present invention, a combination therapy of multiple drugs including an immune checkpoint inhibitor can be described as a treatment using one of the immune checkpoint inhibitors in combination with another drug. In this specification, such a combination therapy of multiple drugs is also referred to as a composite immunotherapy. There are no limitations on the combination of the immune checkpoint inhibitor with the other drug. Preferred combination therapies include a combination therapy of an immune checkpoint inhibitor and an angiogenesis inhibitor, and a combination therapy of two types of immune checkpoint inhibitors.

[0016] The angiogenesis inhibitor of the present invention is not particularly limited as long as it is a drug that inhibits the generation or growth of blood vessels around or within a tumor. Suitable angiogenesis inhibitors include vascular endothelial growth factor (VEGF) inhibitors. Preferred VEGF inhibitors include antibodies against VEGF or its receptor, such as anti-VEGF antibodies and anti-VEGFR2 antibodies. More preferred are anti-VEGF antibodies. Examples of anti-VEGF antibodies include bevacizumab and ranibizumab. Preferred VEGF inhibitors include drugs that inhibit the tyrosine kinase activity of VEGF receptors. Examples of VEGF receptor tyrosine kinase inhibitors include lenvatinib, sorafenib, etc.

[0017] The combination therapy of an immune checkpoint inhibitor and an angiogenesis inhibitor according to the present invention is a treatment in which the immune checkpoint inhibitor and an angiogenesis inhibitor are administered in combination to a patient, and the combination of each drug is not particularly limited. Preferably, it is a combination therapy of an anti-PD-L1 antibody and an anti-VEGF antibody, and more preferably, it is a combination therapy of atezolizumab and bevacizumab.

[0018] The combination therapy of two types of immune checkpoint inhibitors in the present invention is a treatment in which two types of immune checkpoint inhibitors are administered in combination to a patient, and there are no limitations on the combination of immune checkpoint inhibitors. A preferred example is a combination therapy of a drug that inhibits the PD-1 / PD-L1 pathway and a drug that inhibits CTLA-4 signaling. Preferred drugs that inhibit the PD-1 / PD-L1 pathway include anti-PD-L1 antibodies and anti-PD-1 antibodies, and more preferred are anti-PD-L1 antibodies. Examples of anti-PD-L1 antibodies include atezolizumab, durvalumab, and avelumab. Preferred drugs that inhibit CTLA-4 signaling include anti-CTLA-4 antibodies. Preferred anti-CTLA-4 antibodies include tremelimumab and ipilimumab.

[0019] More preferably, the combination therapy of two types of immune checkpoint inhibitors includes a combination therapy of an anti-PD-L1 antibody or an anti-PD-1 antibody and an anti-CTLA-4 antibody, and even more preferably, a combination therapy of durvalumab and tremelimumab.

[0020] In the present invention, the term "antibody" includes polyclonal antibodies, monoclonal antibodies, antigen-binding fragments of antibodies, chimeric antibodies containing the antigen-binding fragments, recombinant antibodies, and derivatives thereof.

[0021] In the present invention, a positive cell refers to a cell expressing a certain protein at an amount recognizable by an antibody against the protein, or a cell in which mRNA encoding the certain protein is detected or a certain amount or more of the mRNA is detected within the cell. A cell in which mRNA encoding a certain protein is detected within the cell refers, for example, to a cell in which the mRNA is detected by PCR. A cell in which mRNA encoding a certain protein is detected at an amount greater than or equal to 1 is, for example, a cell in which the gene expression level of the protein is detected at an amount greater than or equal to 1. A cell in which the gene expression level is detected at an amount greater than or equal to 1 is a cell in which, when the total number of UMIs in the mRNA of each cell is normalized to 10,000, the logarithmic transformation (natural logarithm) of the number of UMIs in the mRNA of the gene is greater than or equal to 1. For example, a CD14-positive cell refers to a cell in which CD14 is expressed at an amount recognizable by an anti-CD14 antibody, or a cell in which mRNA encoding CD14 is detected or a certain amount or more of the mRNA is detected within the cell.

[0022] In the present invention, a negative cell refers to a cell in which a certain protein is not recognized by an antibody against that protein, or a cell in which mRNA encoding a certain protein is not or barely detectable within the cell. For example, a cell in which mRNA encoding a certain protein is barely detectable within the cell is a cell in which, when the total number of UMIs in the mRNA of the gene in question is normalized to 10,000, the logarithmic transformation (natural logarithm) of the number of UMIs in the mRNA of that gene is less than 1.

[0023] Each protein-positive cell can be detected by measuring the protein or mRNA present in the cell. That is, CD14-positive cells can be described as cells in which CD14 protein is detected or cells in which mRNA encoding CD14 protein is detected. CD4-positive cells can be described as cells in which CD4 protein is detected or cells in which mRNA encoding CD4 protein is detected. Cells positive for a protein encoded by at least one gene selected from gene group P can be described as cells in which a protein encoded by at least one gene selected from gene group P is detected or cells in which mRNA of the gene is detected. Cells positive for a protein encoded by at least one gene selected from gene group Q can be described as cells in which a protein encoded by at least one gene selected from gene group Q is detected or cells in which mRNA of the gene is detected.

[0024] The testing method of the present invention is a testing method that includes a step of measuring at least one of four indicators focusing on the cell counts of CD14-positive cells and CD4-positive cells in peripheral blood and the gene expression characteristics of these cells. The testing method of the present invention also uses at least one of the four indicators or the ratio of these indicators as an indicator.

[0025] One of the indicators of the test method of the present invention is the relative cell number of CD14-positive cells. Hereinafter, this indicator will also be referred to as indicator (1). The test method of the present invention may include a step (a) of measuring indicator (1). The CD14-positive cells are not limited by other cellular characteristics, so long as they are cells that are positive for the cell surface antigen CD14. Preferred CD14-positive cells include CD14-positive cells further specified by other cellular characteristics. For example, these include cells that are CD14-positive and positive or negative for other proteins.

[0026] The CD14-positive cells in indicator (1) are preferably CD14-positive cells positive for a protein encoded by at least one gene selected from gene group P. Gene group P includes LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, IL1RN, FABP5, JARID2, AREG, CXCL16, CCL2, CTSL, CCL7, IL1B, GOS2, IER3, ASPH, CCL3, MAFB, IFI30, SGK1, and HMOX. 1, CEBPD, RGCC, CST3, ATP2B1, FCN1, MARCKS, CLEC7A, FTH1, DUSP6, S100A11, CD14, S100A9, HLA-DRA, MTSS1, CD300E, ATP2B1-AS1, DOCK4, SERPINA1, PLAUR, AIF1, TYROBP, SAT1, S100A10, FTL, and GRN.

[0027] Preferred CD14-positive cells in indicator (1) include CD14-positive CD3-negative cells. Preferred CD14-positive CD3-negative cells include CD14-positive CD3-negative cells that are positive for a protein encoded by at least one gene selected from gene group P. The genes included in gene group P are as described above.

[0028] In index (1), among the genes contained in gene group P, preferred genes are selected from genes belonging to gene group P1. Gene group P1 is a gene group consisting of LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, IL1RN, FABP5, JARID2, AREG, CXCL16, CCL2, CTSL, CCL7, IL1B, G0S2, IER3, ASPH, and CCL3. More preferred genes are selected from genes belonging to gene group Pa2. Gene group Pa2 is a gene group consisting of LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, and IL1RN.

[0029] One of the indicators of the test method of the present invention is the expression level of at least one gene selected from gene group P in CD14-positive cells. The expression level of a gene in CD14-positive cells refers to the expression level of a gene in a CD14-positive cell group, and can also be expressed as the expression level of a gene in a fraction extracted, identified, or separated by extracting, identifying, or isolating CD14-positive cells from a sample. The genes contained in gene group P are as described above. Hereinafter, this indicator will also be referred to as indicator (2). The test method of the present invention may include step (b) of measuring indicator (2). The CD14-positive cells may be cells further identified based on other cellular characteristics. For example, the expression level of at least one gene selected from gene group P in cells that are CD14-positive and positive or negative for other proteins is a preferred embodiment of indicator (2). The expression level of at least one gene selected from gene group P in CD14-positive and CD3-negative cells is a more preferred embodiment of indicator (2).

[0030] Among the genes included in gene group P, genes preferred for expression level measurement are selected from genes belonging to gene group P1. The genes belonging to gene group P1 are as described above. More preferred genes are selected from genes belonging to gene group Pb2. Gene group Pb2 is a gene group consisting of FABP5, JARID2, PID1, AREG, LGALS2, CXCL16, and TIMP1.

[0031] One of the indicators of the testing method of the present invention is the relative number of CD4-positive cells positive for a protein encoded by at least one gene selected from gene group Q. Gene group Q consists of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, SELL, TCF7, FOS, SOCS3, FHIT, MYC, GIMAP5, TRABD2A, PIM2, ITGA6, ADTRP, EDA, LTB, RPS13, RPS3A, GAS5, NOSIP, RPS5, EEF1B2, RPL32, LDHB, RPL22, RPL5, RPS12, RPL30, RPS8, RPL11, EEF1G, RPS23, SNHG29, RPL35A, EIF3E, RPS28, RPL39, and RPL9. Hereinafter, this indicator will also be referred to as indicator (3). The testing method of the present invention may include a step (c) of measuring an indicator (3).

[0032] Preferred cells that are CD4-positive and positive for a protein encoded by at least one gene selected from gene group Q include cells further specified by other cellular characteristics. For example, these cells may further include other protein-positive or -negative cells. Preferred cells include CD4-positive T cells that are positive for a protein encoded by at least one gene selected from gene group Q. The CD4-positive T cells may be CD4-positive HLA-DR-negative cells or CD4-positive CD3-positive cells.

[0033] In indicator (3), among the genes contained in gene group Q, preferred genes are selected from genes belonging to gene group Q1. Gene group Q1 is a gene group consisting of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, SELL, TCF7, FOS, SOCS3, FHIT, MYC, GIMAP5, TRABD2A, PIM2, ITGA6, ADTRP, and EDA. More preferred genes are selected from genes belonging to gene group Qc2. Gene group Qc2 is a gene group consisting of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, and TCF7.

[0034] One of the indicators of the testing method of the present invention is the expression level of at least one gene selected from gene group Q in CD4-positive cells. Here, the expression level of a gene in CD4-positive cells refers to the expression level of a gene in a CD4-positive cell group, and can also be expressed as the expression level of a gene in a fraction extracted, identified, or separated by extracting, identifying, or isolating CD4-positive cells from a sample. The genes contained in gene group Q are as described above. Hereinafter, this indicator will also be referred to as indicator (4). The testing method of the present invention may include a step (d) of measuring indicator (4). The CD4-positive cells may be cells further identified based on other cellular characteristics. For example, the expression level of at least one gene selected from gene group Q in cells that are CD4-positive and positive or negative for other proteins is one preferred embodiment of indicator (4). A preferred example of indicator (4) is the expression level of at least one gene selected from gene group Q in CD4-positive T cells. The CD4-positive T cells may be CD4-positive CD3-positive cells. Furthermore, the CD4-positive T cells may be CD4-positive HLA-DR-negative cells.

[0035] Among the genes included in gene group Q, genes preferred for expression level measurement are selected from genes belonging to gene group Q1. The genes belonging to gene group Q1 are as described above. More preferred genes are selected from genes belonging to gene group Qd2. Gene group Qd2 is a gene group consisting of FOS, LEF1, CCR7, TCF7, SELL, and TSHZ2.

[0036] One preferred indicator for the testing method of the present invention is the ratio of indicator (3) to indicator (1). Hereinafter, this indicator will also be referred to as indicator (5). Index (5) = Index (3) / Index (1) One preferred indicator for the testing method of the present invention is the ratio of index (1) to index (3). Hereinafter, this indicator will also be referred to as index (6). Index (6) = Index (1) / Index (3) There are no particular limitations on the combination of indicator (3) and index (1). Any measured values ​​of relative cell numbers can be combined. The testing method of the present invention may comprise step (a) of measuring indicator (1), step (c) of measuring index (3), and step (e) of calculating index (5) or step (f) of calculating index (6). A preferred embodiment of indicator (5) is, for example, the ratio of the relative cell number of CD4-positive and CCR7-positive cells to the relative cell number of CD14-positive cells.

[0037] The relative cell number measured by the testing method of the present invention may be a relative value or an absolute value. The relative cell number in the present invention can be expressed in other words as the frequency of cell presence, the cell ratio, or the true cell number.

[0038] The method for measuring the relative cell number in the present invention is not particularly limited. Measurement can be performed using any commonly used cell number measurement method. Cell quantification methods such as flow cytometry, ELISA, RT-PCR (reverse transcription PCR), RNA-seq (RNA sequencing), scRNA-seq (single cell RNA-seq), and magnetic cell sorting (MACS) can be used. Preferably, flow cytometry can be used.

[0039] In flow cytometry, ELISA, and the like, anti-CD14 antibodies, antibodies against proteins encoded by genes belonging to gene group P, anti-CD4 antibodies, antibodies against proteins encoded by genes belonging to gene group Q, and the like can be used. For example, by using peripheral blood mononuclear cells (PBMCs) as a sample and the above antibodies in a flow cytometer, the number of cells that are CD14-positive and positive for proteins encoded by genes belonging to gene group P can be measured. The ratio of this cell number to the total number of PBMCs can then be calculated. Also, by using PBMCs as a sample and the above antibodies in a flow cytometer, the number of cells that are CD4-positive and positive for proteins encoded by genes belonging to gene group Q can be measured. The ratio of this cell number to the total number of PBMCs can then be calculated.

[0040] Each protein-positive cell can be detected using flow cytometry, ELISA, Western blot, RT-PCR, RNA-seq, magnetic cell sorting (MACS), RNA-seq, scRNA-seq, etc.

[0041] In the present invention, the method for measuring gene expression levels is not limited, and known methods can be used. Gene expression levels can be measured by measuring the amount of mRNA or protein. mRNA levels can be measured by RT-qPCR (quantitative reverse transcription PCR), RNA-seq, etc. For example, the expression level of at least one gene selected from gene group P or gene group Q can be measured by detecting CD14-positive cells or CD4-positive cells and performing RT-qPCR. RNA-seq may also be used. RNA-seq may also be scRNA-seq. Gene expression levels can also be measured by quantifying the protein encoded by the gene. Protein levels can be measured by, for example, Western blot, ELISA, flow cytometry, magnetic cell sorting (MACS), etc.

[0042] A preferred embodiment of the testing method of the present invention includes a step of comparing the measured relative cell number or expression level with a reference value. The effectiveness of a combination therapy of multiple drugs, including an immune checkpoint inhibitor, for treating hepatocellular carcinoma can be assessed based on the comparison results.

[0043] When the combination therapy is a combination therapy of an immune checkpoint inhibitor and an angiogenesis inhibitor, the combination therapy can be determined as follows: When the relative cell number (index (1)) measured in step (a) is lower than a reference value, the combination therapy of an immune checkpoint inhibitor and an angiogenesis inhibitor can be determined to be effective. When the expression level (index (2)) measured in step (b) is lower than a reference value, the combination therapy of an immune checkpoint inhibitor and an angiogenesis inhibitor can be determined to be effective. When the relative cell number (index (3)) measured in step (c) is higher than a reference value, the combination therapy of an immune checkpoint inhibitor and an angiogenesis inhibitor can be determined to be effective. When the expression level (index (4)) measured in step (d) is higher than a reference value, the combination therapy of an immune checkpoint inhibitor and an angiogenesis inhibitor can be determined to be effective. When the ratio of the relative cell numbers obtained in step (e) (index (5)) is higher than a reference value, the combination therapy of an immune checkpoint inhibitor and an angiogenesis inhibitor can be determined to be effective. When the ratio of the relative cell counts (index (6)) obtained in the step (f) is lower than the reference value, it can be determined that the combination therapy of the immune checkpoint inhibitor and the angiogenesis inhibitor is effective.

[0044] When the combination therapy is a combination therapy of a drug that inhibits the PD-1 / PD-L1 pathway and a drug that inhibits CTLA-4 signaling, it can be determined as follows: When the relative cell number (index (1)) measured in the step (a) is higher than a reference value, it can be determined that the combination therapy of a drug that inhibits the PD-1 / PD-L1 pathway and a drug that inhibits CTLA-4 signaling is effective. When the expression level (index (2)) measured in the step (b) is higher than a reference value, it can be determined that the combination therapy of a drug that inhibits the PD-1 / PD-L1 pathway and a drug that inhibits CTLA-4 signaling is effective.

[0045] In the present invention, "effective" refers to the improvement of hepatocellular carcinoma lesions by combination therapy with multiple drugs, including immune checkpoint inhibitors. Specifically, "effective" refers to a complete response (CR) or partial response (PR) based on the RECIST (Response Evaluation Criteria in Solid Tumors) standard. While there are revised versions of the RECIST guidelines, the evaluation criteria of either guideline may be used.

[0046] The reference value in the present invention can be appropriately determined for each specific measurement method or calculation method. The reference value in the present invention can be determined from measured values ​​obtained by analyzing peripheral blood samples obtained from multiple hepatocellular carcinoma patients and / or healthy individuals using the same method. Preferably, the reference value in the present invention can be determined from measured values ​​obtained by analyzing peripheral blood samples obtained from multiple hepatocellular carcinoma patients using the same method. For example, the reference value in the present invention can be the median or average of the measured values ​​obtained from multiple hepatocellular carcinoma patients, or can be a value determined taking into consideration the measured values ​​obtained from multiple hepatocellular carcinoma patients and an evaluation of the patient's lesions. One preferred reference value can be the median of the measured values ​​obtained from multiple hepatocellular carcinoma patients.

[0047] A preferred embodiment of the testing method of the present invention is a method comprising the step (a) above.

[0048] A preferred embodiment of step (a) is a step of detecting CD14-positive cells from a blood sample and measuring the relative cell number (presence frequency) of CD14-positive cells. The measured relative cell number can be compared with a reference value to determine the effect of the combination therapy.

[0049] A preferred embodiment of step (a) is a step of detecting cells from a blood sample that are CD14-positive and positive for a protein encoded by at least one gene selected from gene group P, and measuring the relative cell number (presence frequency). The measured relative cell number can be compared with a reference value to determine the effect of the combination therapy. The at least one gene selected from gene group P is preferably selected from gene group P1, more preferably selected from gene group Pa2.

[0050] A preferred embodiment of step (a) is a step of detecting cells from a blood sample that are CD14 positive, CD3 negative, and positive for a protein encoded by at least one gene selected from gene group P, and measuring the relative cell number (presence frequency). The measured relative cell number can be compared with a reference value to determine the effect of the combination therapy. The at least one gene selected from gene group P is preferably selected from gene group P1, more preferably selected from gene group Pa2.

[0051] A preferred embodiment of step (a) is a step of measuring the relative cell number (cell frequency) of a cell population containing CD14-positive CCL2-positive cells by scRNA-seq analysis. The relative cell number can be compared with a reference value to determine the effect of the combination therapy.

[0052] A preferred embodiment of the testing method of the present invention is a method comprising the step (c) above.

[0053] A preferred embodiment of step (c) is a step of detecting CD4-positive cells from a blood sample that are positive for a protein encoded by at least one gene selected from gene group Q, and measuring the relative cell number (presence frequency). The measured relative cell number can be compared with a reference value to determine the effect of the combination therapy. The at least one gene selected from gene group Q is preferably selected from gene group Q1, and more preferably selected from gene group Qc2.

[0054] A preferred embodiment of step (c) may be a step of detecting cells from a blood sample that are CD4 positive, HLA-DR negative, and positive for a protein encoded by at least one gene selected from gene group Q, and measuring the relative cell number (presence frequency). The measured relative cell number can be compared with a reference value to determine the effect of the combination therapy. The at least one gene selected from gene group Q is preferably selected from gene group Q1, and more preferably selected from gene group Qc2.

[0055] A preferred embodiment of step (c) may be a step of measuring the relative cell number (cell frequency) of a cell population containing CD4-positive, CCR7-positive cells by scRNA-seq analysis. The relative cell number can be compared with a reference value to determine the effect of the combination therapy. More preferably, the method may be a method of measuring the relative cell number (cell frequency) of a cell population containing CD4-positive, CCR7-positive, TCF7-positive cells by scRNA-seq analysis. The relative cell number can be compared with a reference value to determine the effect of the combination therapy.

[0056] A preferred embodiment of the testing method of the present invention is a method comprising the steps (a) and (c), and further comprising the step (e) or the step (f).

[0057] Step (e) is a step of calculating the ratio of the relative cell number of CD4-positive cells positive for a protein encoded by at least one gene selected from gene group Q to the relative cell number of CD14-positive cells. A preferred embodiment can be a step of calculating the ratio of the relative cell number of CD4-positive CCR7-positive cells to the relative cell number of CD14-positive cells. A preferred embodiment can be a step of calculating the ratio of the relative cell number of CD4-positive LEF1-positive cells to the relative cell number of CD14-positive cells. The ratio of the relative cell numbers can be compared with a reference value to determine the effect of the combination therapy.

[0058] A preferred embodiment of step (e) may be a step of calculating the ratio of the relative cell numbers of the cell populations obtained by scRNA-seq analysis. For example, this may be a step of calculating the ratio of the relative cell number of a cell population containing CD4-positive CCR7-positive cells to the relative cell number of a cell population containing CD14-positive CCL2-positive cells obtained by scRNA-seq analysis. The ratio of the relative cell numbers may be compared with a reference value to determine the effect of the combination therapy.

[0059] Step (f) is a step of calculating the ratio of the relative cell number of CD14-positive cells to the relative cell number of CD4-positive cells positive for a protein encoded by at least one gene selected from gene group Q. One preferred embodiment can be a step of calculating the ratio of the relative cell number of CD14-positive cells to the relative cell number of CD4-positive CCR7-positive cells. One preferred embodiment can be a step of calculating the ratio of the relative cell number of CD14-positive cells to the relative cell number of CD4-positive LEF1-positive cells, etc. The ratio of the relative cell numbers can be compared with a reference value to determine the effect of the combination therapy.

[0060] A preferred embodiment of step (f) may be a step of calculating the ratio of the relative cell numbers in the cell population obtained by scRNA-seq analysis. For example, it may be a step of calculating the ratio of the relative cell numbers in a cell population containing CD14-positive CCL2-positive cells to a cell population containing CD4-positive CCR7-positive cells obtained by scRNA-seq analysis. The ratio of the relative cell numbers may be compared with a reference value to determine the effect of the combination therapy.

[0061] A preferred embodiment of step (b) is a step of detecting CD14-positive cells from a blood sample and measuring the expression level of at least one gene selected from gene group P in the CD14-positive cell group. The measured value can be compared with a reference value to determine the effect of the combination therapy.

[0062] Among the genes contained in gene group P, genes that are preferable to be the target of expression level measurement can be selected from gene group P1, and more preferable genes can be selected from gene group Pb2.

[0063] A preferred embodiment of step (d) may be a step of detecting CD4-positive cells from a blood sample and measuring the expression level of at least one gene selected from gene group Q in the CD4-positive cell population. An even more preferred embodiment may be a step of detecting CD4-positive T cells from a blood sample and measuring the expression level of at least one gene selected from gene group Q in the CD4-positive T cell population. CD4-positive T cells can be detected from a blood sample, for example, by flow cytometry, magnetic cell sorting, RNA-seq, scRNA-seq, or the like, using the cell surface antigens CD4 and CD3 as markers, or the cell surface antigens CD4 and HLA-DR as markers. The measured values ​​are compared with reference values ​​to determine the effect of the combination therapy.

[0064] Among the genes contained in gene group Q, genes that are preferred as targets for expression level measurement are selected from gene group Q1, and more preferred genes are selected from gene group Qd2.

[0065] The testing device of the present invention for predicting the effect of a treatment for hepatocellular carcinoma by a combination therapy of multiple drugs including an immune checkpoint inhibitor from a patient's peripheral blood comprises at least one detection unit selected from the group consisting of the following (h) to (k): (h) a relative cell number detection unit that detects the relative cell number of CD14-positive cells, (i) an expression level detection unit that detects the expression level of at least one gene selected from gene group P in CD14-positive cells, (j) a relative cell number detection unit that detects the relative cell number of CD4-positive cells and positive for a protein encoded by at least one gene selected from gene group Q, and (k) an expression level detection unit that detects the expression level of at least one gene selected from gene group Q in CD4-positive cells.

[0066] The gene group P includes LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, IL1RN, FABP5, JARID2, AREG, CXCL16, CCL2, CTSL, CCL7, IL1B, G0S2, IER3, ASPH, CCL3, MAFB, IFI30, SGK1, HMO It consists of X1, CEBPD, RGCC, CST3, ATP2B1, FCN1, MARCKS, CLEC7A, FTH1, DUSP6, S100A11, CD14, S100A9, HLA-DRA, MTSS1, CD300E, ATP2B1-AS1, DOCK4, SERPINA1, PLAUR, AIF1, TYROBP, SAT1, S100A10, FTL, and GRN. The gene group Q consists of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, SELL, TCF7, FOS, SOCS3, FHIT, MYC, GIMAP5, TRABD2A, PIM2, ITGA6, ADTRP, EDA, LTB, RPS13, RPS3A, GAS5, NOSIP, RPS5, EEF1B2, RPL32, LDHB, RPL22, RPL5, RPS12, RPL30, RPS8, RPL11, EEF1G, RPS23, SNHG29, RPL35A, EIF3E, RPS28, RPL39, and RPL9.

[0067] The device of the present invention preferably further includes an output unit that compares the detected value (measured value) with a reference value and indicates the effect of the treatment of hepatocellular carcinoma by a combination therapy of multiple drugs including an immune checkpoint inhibitor. There are no limitations on the manner in which the treatment effect indicated by the output unit can be displayed. For example, it may be a numerical value indicating the possibility of response, or it may be displayed as "successful" or "ineffective."

[0068] The detection unit may include a detection means such as flow cytometry, ELISA, RNA-seq, RT-PCR, etc. The embodiments of each configuration of the testing device of the present invention are the same as the embodiments described in the testing method.

[0069] The present invention will be specifically described below using examples to deepen understanding of the present invention, but these examples are not intended to limit the scope of the present invention. The following studies conform to the principles of the Declaration of Helsinki, were approved by the Ethics Committee of Osaka University Hospital, and were conducted with the consent of all patients. The modified RECIST proposed by Lencioni et al. in 2010 was used to evaluate the lesions in the cases in the examples.

[0070] Example 1: Patients with unresectable advanced hepatocellular carcinoma receiving combined immunotherapy from Osaka University and 12 affiliated institutions were enrolled in this study. Whole blood samples were collected from patients before the start of combined immunotherapy (1 day before or on the day of treatment) and during treatment (6 weeks after treatment initiation). Peripheral blood mononuclear cells (PBMCs) were isolated from the whole blood by density gradient centrifugation using lymphocyte separation solution (Cat. 20828-15, NACALAI TESQUE, INC., Kyoto, Japan) and stored in liquid nitrogen. Plasma components were stored at -80°C.

[0071] PBMCs from patients receiving a combination therapy of atezolizumab and bevacizumab were analyzed. 20 samples of PBMCs from five responders (hereinafter referred to as the responder group) and five non-responders (hereinafter referred to as the non-responder group) were analyzed, including those from pre-treatment and treatment. Responders were cases evaluated as complete response (CR) and partial response (PR), while non-responders were cases evaluated as progressive disease (PD). Non-responders were evaluated 6 weeks after the start of treatment, and response was evaluated at the best time.

[0072] The 20 samples were subjected to single-cell RNA sequencing (CITE-seq) for simultaneous quantification of mRNA and protein. 10,000-20,000 cells from each sample were loaded onto the sequencer. Cells deemed dead based on mitochondrial genetic information were then excluded, leaving a total of 231,383 cells for analysis. The method was performed according to a previously published report (Stoeckius et al. Nat Methods. 2017 Sep; 14(9): 865-868.). CITE-seq (Cellular Indexing of Transcriptomes and Epitopes by Sequencing) is an analytical method developed at the New York Genome Center that integrates single-cell RNA-seq and proteomics to simultaneously obtain RNA and protein information. Specific methods, reagents, and equipment are described below.

[0073] (Cell Staining with DNA Barcoded Antibodies) Approximately 10 million peripheral blood mononuclear cells were lysed in 45 μL of Cell Staining Buffer (Cat. 420201, BioLegend, USA), and 5 μL of Human TruStain FcX Fc Receptor Inhibitor (Cat. 422301, BioLegend, USA) was added. The cells were then incubated at 4°C for 10 minutes. Cell surface antigen staining was then performed with DNA barcoded antibodies at 4°C for 30 minutes. Following the 10x Genomics (CA, USA) manual (CG000149 | Rev D), the cells were washed once with cold PBS (containing 0.04% BSA) and three times with cold PBS (containing 1% BSA), and the concentration was adjusted to 1,000 cells / μL with PBS (containing 0.04% BSA).

[0074] The list of DNA barcoded antibodies for CITE-seq is shown below. The following antibodies were mixed in accordance with the TotalSeq Antibody Cocktail Spike-in Guidance (BioLegend, USA).

[0075] TotalSeq-C Human TBNK Cocktail (Cat. 399903, BioLegend, USA); TotalSeq-C0161 anti-human CD11b Antibody (Cat. 301359, Clone ICRF44, BioLegend, USA); TotalSeq-C0392 anti-human CD15 (SSEA-1) Antibody (Cat. 323053, Clone W6D3, BioLegend, USA); TotalSeq-C0159 anti-human HLA-DR Antibody (Cat. 307663, Clone L243, BioLegend, USA);

[0076]

[0077] (Single-Cell RNA Sequencing) Single-cell cell suspensions were processed using the Chromium Controller (10x Genomics) according to the Chromium Single Cell 5' Reagent Kits User Guide. The reagents listed in Table 2 below were used during the processing process. According to the 10x Genomics manual, approximately 15,000 live cells were loaded into the Chromium Controller to form 10,000 single-cell gel bead emulsions, which were used for library preparation and sequencing. The oil droplets containing the single cells and barcoded beads (GEMs) were then reverse transcribed, and cell barcodes and UMIs (unique tags for each transcript) were added. The cDNA was then amplified to create single-cell libraries according to the protocol. The quantity and quality of the libraries were confirmed using an Agilent Tapestation system and Qubit Fluorometer (Thermo Fisher Scientific, MA USA), and sequencing was performed in paired-end mode using a DNBSEQ-G400RS (MGI Tech Japan, Tokyo Japan) at 20,000 reads / cell for gene expression and 5,000 reads / cell for BCR, TCR, and cell surface proteins. The resulting raw reads were processed using Cell Ranger ver. 7.0.0 (10x Genomics) and aligned to the human reference genome sequence (version GRCh38).

[0078]

[0079] (Single-cell RNA-seq Data Analysis) For each sample, a count matrix was created using Cell Ranger ver. 7.0.0, and Seurat ver. 4.3.0 and R ver. 4.2.2 were used to process the single-cell RNA-seq data. Cell Ranger output files were converted to a UMI matrix using the Read10x function. This UMI matrix was used to create a Seurat object. The total number of UMIs per cell was normalized to 10,000, and the results were log-transformed and normalized. A quality check was performed for each sample, and cells with a proportion of mitochondrial genes greater than 10%, extremely low numbers of detected genes (<1,000), or greater than 4,500 detected genes were generally excluded. Protein data were normalized and noise-reduced using the DSB (Denoised and Scaled by Background) method (Mule MP et al. Nat. Commun. 2022). The dsb normalization function was used with the parameters denoise_counts = FALSE and use.isotype.control = FALSE. K-nearest neighbor graphs were constructed using the FindNeighbors and FindCluster functions. Cell clusters were quality controlled based on nearest neighbor and cell surface proteins. Cells were then further clustered within each major cell population using weighted Nearest Neighbors analysis.

[0080] (Integration of single-cell RNA sequencing data) To correct for batch effects or inter-experimental errors in the datasets, data integration was performed using reciprocal principal component analysis (rPCA) according to the developer's manual (https: / / satijalab.org / seurat / articles / integration_rpca). Briefly, each dataset was linearly transformed and PCA analysis was performed using highly variable genes (HVG) selected by the Seurat functions FindVariableFeatures and SelectIntegrationFeatures. Anchors were then created using the Seurat function FindIntegrationAnchors, and each dataset was integrated using the Seurat function IngegrateData. The data was scaled using HVG to have a mean expression level of 0 and a variance of 1, followed by principal component analysis and graphing using UMAP.

[0081] The results of the UMAP analysis are shown in Figure 1. The cells were divided into 24 clusters based on differentially expressed genes (DEGs). Some of the differentially expressed genes (DEGs) are listed below: PTGDS, GNLY, GZMB, FGFBP2, PRF1, S100A8, S100A9, S100A12, VCAN, NAMPT, LTB, IL7R, LDHB, ZFP36L2, MAL, CD8A, GZMH, CCL5, CD8B, NKG7, CCL2, LYZ, CXCL2, PID1, and MAF. B, CCR7, TCF7, LEF1, TSHZ2, DOCK4, TIMP1, CTSL, LST1, IFITM3, CDKN1C, IFI30, SER PINA1, S100A11, CD79A, AFF3, BANK1, MS4A1, CD74, PRKCH, SYNE1, SYTL3, CEMIP2, C NOT6L, IL32, STMN1, FOXP3, CTLA4, CD27, PDE3B, CAMK4, ANK3, INPP4B, ARHGAP15, H LA-DRA, HLA-DQA1, HLA-DPB1, HLA-DPA1, HLA-DQB1, PPBP, NRGN, PF4, GP9, TUBB1, N ELL2, LINC02446, NEAT1, LRMDA, JARID2, ARHGAP26, SLC8A1, IGHM, FCHSD2, PLD4, I TM2C, TCF4, IGKC, MYL9, HBB, HBA2, HBA1, ALAS2, CA1, IGHA1, JCHAIN, IGHA2, IGLC2

[0082] Each cluster is numbered in Figure 1. The cells classified into each cluster are shown below.

[0083] Of the results of the analysis of 12 types of cell surface antigens, CD14-positive cells and CD4-positive cells are shown in Figure 2. Two large cell groups are shown in Figure 1. Cell surface antigen analysis revealed that the cell group shown on the left is a T cell / NK cell group, and the cell group shown on the right is a monocyte-related cell group.

[0084] Example 2: When the pretreatment samples of the response group and the non-response group were compared using a graph of the UMAP analysis results, it was clear that there was a difference in the cell count in cluster 4 [CD14 monocyte] (hereinafter also referred to as cluster CD14). The relative cell counts of cluster CD14 in the response group and non-response group are shown in Figure 3. The relative cell count of cluster CD14 in the response group was significantly lower than that in the non-response group.

[0085] Table 4 shows the results of comparing genes expressed in cluster CD14 with PBMC cell groups excluding cluster CD14. Cluster CD14 was a cell population with high expression levels of CCL2 and other CCL2-like molecules. In Table 4, avg_log2 FC is the log fold-change between the two groups in average expression levels, pct.1 is the percentage of cells in which features were detected within cluster CD14, and pct.2 is the percentage of cells in which features were detected within the PBMC cell group excluding cluster CD14.

[0086]

[0087] Example 3: T cell and NK cell groups were extracted based on cell surface expression profiles and gene expression profiles, and 124,058 cells belonging to these cell groups were re-clustered based on the intercellular gene expression profiles, resulting in classification into 22 subtypes (Figure 4). Comparison of pre-treatment samples from the responder and non-responder groups using a graph of the UMAP analysis results after reclustering revealed a clear difference in cell counts in Cluster 4 [CD4T] (hereinafter also referred to as Cluster CD4T). The relative cell counts of Cluster CD4T in the responder and non-responder groups are shown in Figure 5. The relative cell count of Cluster CD4T in the responder group was significantly higher than that in the non-responder group.

[0088] Table 5 shows the results of comparing genes expressed in cluster CD4T with T cell and NK cell groups excluding cluster CD4T. Cluster CD4T was a cell population with high expression levels of CCR7, TCF7, etc. In Table 5, avg_log2 FC is the log fold-change between the two groups in average expression levels, pct.1 is the proportion of cells in which features were detected within cluster CD4T, and pct.2 is the proportion of cells in which features were detected within the T cell and NK cell group excluding cluster CD4T.

[0089]

[0090] Example 4 The relative cell count of cluster CD14 in each patient's pretreatment sample was compared with a reference value of 0.099, and the patients were divided into a high value group (5 cases) and a low value group (5 cases), and the progression-free survival time was compared. The reference value was the median of the relative cell counts of the 10 cases. The progression-free survival time of the low value group was significantly longer than that of the high value group. The progression-free survival rate is shown in Figure 6. Note that evaluation of one case in the low value group was completed at 200 days.

[0091] The relative cell count of cluster CD4 T cells in each patient's pretreatment sample was compared with the baseline value of 0.074, and the patients were divided into a high value group (5 cases) and a low value group (5 cases), and the progression-free survival time was compared. The baseline value was the median relative cell count of the 10 cases. The progression-free survival time was significantly longer in the high value group than in the low value group. The progression-free survival rate is shown in Figure 7. Note that evaluation of one case in the high value group was completed at 200 days.

[0092] Using the relative cell counts measured above, the ratio of the relative cell count of cluster CD4T to the relative cell count of cluster CD14 was calculated for each patient. The obtained values ​​were compared with the reference value of 0.65, and patients were divided into a high value group (5 cases) and a low value group (5 cases), and the progression-free survival time was compared. The reference value was the median of the relative cell count ratios of the 10 cases. The progression-free survival time of the high value group was significantly longer than that of the low value group. The progression-free survival rate is shown in Figure 8. Note that evaluation of one case in the high value group was completed at 200 days.

[0093] These results suggest that the relative cell counts of cluster CD14 and cluster CD4T may be a novel biomarker for predicting the efficacy of combined immune checkpoint inhibitor and anti-angiogenic therapy in hepatocellular carcinoma.

[0094] Example 5: CD14-positive cells were extracted from the library obtained in Example 1, and the relative cell counts (frequency of presence) were compared between the responder and non-responder groups. The results are shown in Figure 9. Figure 9(A) shows the relative cell count of cells with an intracellular CD14-encoding mRNA level greater than 1, and Figure 9(B) shows the relative cell count of cells with an anti-CD14 antibody level bound to the cell surface antigen CD14 greater than 1. The relative cell count in the responder group was significantly lower than that in the non-responder group. Note that cells with an mRNA level greater than 1 encoding CD14 are those for which the total UMI count per cell is normalized to 10,000 and the logarithmic transformation (natural logarithm) of the CD14 mRNA UMI count is greater than 1. Furthermore, cells with an anti-CD14 antibody level bound to the cell surface antigen CD14 greater than 1 are those for which the total UMI count per cell is normalized to 10,000 and the logarithmic transformation (natural logarithm) of the anti-CD14 antibody UMI count is greater than 1. The same applies to the amounts of mRNA of genes and the amounts of antibodies against cell surface antigens in the following examples.

[0095] Example 6: CD14-positive cells with mRNA levels of LYZ, CXCL8, TIMP1, LGALS2, CD68, and LGALS3 greater than 1 were extracted from the library obtained in Example 1, and the relative cell numbers (presence frequencies) were compared between the responder and non-responder groups. The results are shown in Figures 10 and 11. In Figure 10, the condition for extracting CD14-positive cells was that the intracellular mRNA level encoding CD14 was greater than 1. In Figure 11, the condition for extracting CD14-positive cells was that the amount of anti-CD14 antibody bound to the cell surface antigen CD14 was greater than 1. The relative cell numbers in the responder group were significantly lower than those in the non-responder group.

[0096] Example 7 Cells that were CD14 positive and had mRNA levels of CXCL2, PID1, APOBEC3A, and PPARG greater than 1 were extracted from the library obtained in Example 1, and the relative cell numbers (presence frequencies) were compared between the responder group and the ineffective group. The results are shown in Figure 12. The condition for extracting CD14-positive cells was that the intracellular mRNA level encoding CD14 be greater than 1. The relative cell number in the responder group was significantly lower than that in the ineffective group.

[0097] Example 8 Cells that were CD14-positive, CD3-negative, and had mRNA levels of CXCL2, PID1, PHLDA1, APOBEC3A, IL1RN, and PPARG greater than 1 were extracted from the library obtained in Example 1, and the relative cell numbers (presence frequencies) were compared between the responder and non-responder groups. The results are shown in Figure 13. The extraction conditions for CD14-positive, CD3-negative cells were that the amount of anti-CD14 antibody bound to the cell surface antigen CD14 was greater than 2, and the amount of antibody bound to the cell surface antigen CD3 was less than 1. The relative cell numbers in the responder group were significantly lower than those in the non-responder group.

[0098] Example 9 CD4-positive cells with mRNA levels of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, and TCF7 greater than 1 were extracted from the library obtained in Example 1, and the relative cell numbers (presence frequencies) were compared between the responder group and the ineffective group. The results are shown in Figure 14. The condition for extracting CD4-positive cells was that the amount of anti-CD4 antibody bound to the cell surface antigen CD4 be greater than 1. The relative cell number in the responder group was higher than that in the ineffective group.

[0099] Example 10 Cells that were CD4-positive, HLA-DR-negative, and had mRNA levels of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, and TCF7 greater than 1 were extracted from the library obtained in Example 1, and the relative cell numbers (presence frequencies) were compared between the responder group and the ineffective group. The results are shown in Figure 15. The extraction conditions for CD4-positive, HLA-DR-negative cells were that the amount of anti-CD4 antibody bound to the cell surface antigen CD4 was greater than 1, and the amount of antibody bound to the cell surface antigen HLA-DR was less than 1. The relative cell numbers in the responder group were higher than those in the ineffective group.

[0100] Example 11 Pre-treatment CD14-positive cells were extracted from the library obtained in Example 1, and the expression levels of FABP5, JARID2, PID1, AREG, LGALS2, CXCL16, and TIMP1 were compared between the response group and the ineffective group. The results are shown in Figure 16. The expression level of each gene was significantly lower in the response group than in the ineffective group. Cells with an intracellular CD14-encoding mRNA level greater than 1 were defined as CD14-positive cells.

[0101] Example 12 Pre-treatment CD4-positive T cells were extracted from the library obtained in Example 1, and the expression levels of FOS, LEF1, CCR7, TCF7, SELL, and TSHZ2 were compared between the responder group and the non-responder group. The results are shown in Figure 17. The expression level of each gene in the responder group was significantly higher than that in the non-responder group. Cells in which the amount of anti-CD4 antibody bound to the cell surface antigen CD4 was greater than 1 and the amount of anti-HLA-DR antibody bound to the cell surface antigen HLA-DR was less than 1 were defined as CD4-positive T cells.

[0102] Example 13: PBMCs collected in Example 1 and stored in liquid nitrogen were analyzed from patients who received a combination therapy of durvalumab and tremelimumab. Nine pre-treatment PBMC samples and five on-treatment PBMC samples from four responders (hereinafter also referred to as the responder group) and five non-responders (hereinafter also referred to as the non-responder group) were analyzed. Responders were cases evaluated as complete response (CR) and partial response (PR), while non-responders were cases evaluated as progressive disease (PD). Non-responders were evaluated 6 weeks after the start of treatment, and response was evaluated at the best time.

[0103] The above 14 samples were subjected to single-cell RNA sequencing (CITE-seq) for simultaneous quantification of mRNA and protein in the same manner as in Example 1. A total of 114,866 cells were analyzed.

[0104] As in Example 1, graphing was performed using UMAP, and similarly to Example 1, the data were divided into 24 clusters based on differentially expressed genes (DEGs).

[0105] Example 14 As in Example 2, the relative cell count for cluster 4 [CD14 Monocyte] (hereinafter also referred to as cluster CD14 [Dur / Tre]) obtained in Example 13 was analyzed. The relative cell counts of cluster CD14 [Dur / Tre] in the responder group and ineffective group are shown in Figure 18. The relative cell count in the responder group was higher than that in the ineffective group.

[0106] Example 15: The relationship between the relative cell count of cluster CD14 [Dur / Tre] and progression-free survival was investigated. Because one patient who received durvalumab and tremelimumab combination therapy did not have a recorded progression-free survival period, pretreatment samples from eight patients were used. The relative cell count of cluster CD14 [Dur / Tre] in each patient's pretreatment sample was compared with the baseline value of 0.054, and the patients were divided into a high value group (4 cases) and a low value group (4 cases). The progression-free survival period for the high value group is shown in Figure 19, and the progression-free survival period for the low value group is shown in Figure 20. The baseline value was the median of the relative cell counts of the eight cases. The progression-free survival period for the high value group was longer than that for the low value group.

[0107] Patients with high cluster CD14 relative cell counts did not respond to atezolizumab and bevacizumab combination therapy and did not extend progression-free survival, but did respond to durvalumab and tremelimumab combination therapy, which extended progression-free survival (Figure 19). On the other hand, patients with low cluster CD14 relative cell counts responded to atezolizumab and bevacizumab combination therapy and extended progression-free survival, but did not respond to durvalumab and tremelimumab combination therapy, which did not extend progression-free survival (Figure 20). Therefore, a patient's pre-treatment cluster CD14 relative cell count can be used to select a treatment method (atezolizumab and bevacizumab combination therapy or durvalumab and tremelimumab combination therapy) before starting treatment.

[0108] Example 16: CD14-positive cells were extracted from the library obtained in Example 13, and the relative cell numbers (presence frequency) were compared between the responder group and the non-responder group. Cells with an intracellular amount of mRNA encoding CD14 greater than 1 were defined as CD14-positive cells. The results are shown in Figure 21. The relative cell numbers in the responder group were higher than in the non-responder group. This result is the opposite of the result in Example 5, and therefore it is thought that patients who respond to the combination therapy of atezolizumab and bevacizumab and patients who respond to the combination therapy of durvalumab and tremelimumab can be separated by the relative cell numbers of CD14-positive cells.

[0109] Example 17: CD14-positive cells with mRNA levels of PID1, APOBEC3A, PPARG, LYZ, CXCL8, TIMP1, LGALS2, and LGALS3 greater than 1 were extracted from the library obtained in Example 13, and the relative cell numbers (prevalence frequencies) were compared between the responder and non-responder groups. Cells with an anti-CD14 antibody level greater than 1 bound to the cell surface antigen CD14 were defined as CD14-positive cells. The results are shown in Figure 22. The relative cell numbers in the responder group were higher than in the non-responder group. This result is the opposite of the results in Examples 6 and 7, suggesting that patients who respond to the combination therapy of atezolizumab and bevacizumab and patients who respond to the combination therapy of durvalumab and tremelimumab can be distinguished based on the relative cell numbers of these cells.

[0110] Example 18: Pre-treatment CD14-positive cells were extracted from the library obtained in Example 13, and the expression levels of FABP5, JARID2, PID1, AREG, CXCL16, and TIMP1 were compared between the responder and non-responder groups. Cells with an amount of anti-CD14 antibody bound to the cell surface antigen CD14 greater than 1 were defined as CD14-positive cells. The results are shown in Figure 23. The expression level of each gene in the responder group was higher than that in the non-responder group. This result is the opposite of the results in Example 11, suggesting that patients who respond to the combination therapy of atezolizumab and bevacizumab and patients who respond to the combination therapy of durvalumab and tremelimumab can be distinguished based on the expression levels of these genes in CD14-positive cells.

[0111] The results of the above examples indicate that the relative cell numbers or gene expression levels of cell populations selected using the mRNA or proteins of the above genes as markers can serve as novel biomarkers for predicting the efficacy of combined therapy with multiple drugs, including immune checkpoint inhibitors, in treating hepatocellular carcinoma.

[0112] To predict drug efficacy and provide appropriate treatment for patients with hepatocellular carcinoma receiving combination therapy with multiple drugs, including immune checkpoint inhibitors.

Claims

1. A testing method for predicting the effect of a combination therapy of multiple drugs including an immune checkpoint inhibitor on hepatocellular carcinoma using peripheral blood of a patient with hepatocellular carcinoma as a sample, the testing method comprising at least one step selected from the group consisting of the following (a) to (d): (a) measuring the relative cell number of CD14-positive cells; (b) measuring the expression level of at least one gene selected from gene group P in CD14-positive cells; (c) measuring the relative cell number of CD4-positive cells and cells positive for a protein encoded by at least one gene selected from gene group Q; and (d) measuring the expression level of at least one gene selected from gene group Q in CD4-positive cells. Here, the gene group P is LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, IL1RN, FABP5, JARID2, AREG, CXCL16, CCL2, CTSL, CCL7, IL1B, G0S2, IER3, ASPH, CCL3, MAFB, IFI30, SGK1, H MOX1, CEBPD, RGCC, CST3, ATP2B1, FCN1, MARCKS, CLEC7A, FTH1, DUSP6, S100A11, CD14, S100A9, HLA-DRA, MTSS1, CD300E, ATP2B1-AS1, DOCK4, SERPINA1, PLAUR, AIF1, TYROBP, SAT1, S100A10, FTL, and GRN; wherein the gene group Q consists of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, SELL, TCF7, FOS, SOCS3, FHIT, MYC, GIMAP5, TRABD2A, PIM2, ITGA6, ADTRP, EDA, LTB, RPS13, RPS3A, GAS5, NOSIP, RPS5, EEF1B2, RPL32, LDHB, RPL22, RPL5, RPS12, RPL30, RPS8, RPL11, EEF1G, RPS23, SNHG29, RPL35A, EIF3E, RPS28, RPL39, and RPL9.

2. The method of claim 1, wherein the multiple drugs including an immune checkpoint inhibitor are an immune checkpoint inhibitor and an angiogenesis inhibitor.

3. The method according to claim 2, wherein the immune checkpoint inhibitor is a drug that inhibits the PD-1 / PD-L1 pathway.

4. The method according to claim 3, wherein the drug that inhibits the PD-1 / PD-L1 pathway is an anti-PD-L1 antibody or an anti-PD-1 antibody, and the angiogenesis inhibitor is an anti-VEGF antibody.

5. The method of claim 1, wherein the multiple drugs comprising an immune checkpoint inhibitor are two types of immune checkpoint inhibitors.

6. The method according to claim 5, wherein the two types of immune checkpoint inhibitors are a drug that inhibits the PD-1 / PD-L1 pathway and a drug that inhibits CTLA-4 signaling.

7. The method according to claim 1, wherein the CD14 positive cells in step (a) are positive cells for a protein encoded by at least one gene selected from the gene group P.

8. The method according to claim 7, wherein the gene group P is gene group Pa2, the gene group Pa2 consisting of LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, and IL1RN.

9. The method according to claim 1, wherein the gene group Q in step (c) is gene group Qc2, which is a gene group consisting of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, and TCF7.

10. The method according to claim 1, wherein the gene group P in step (b) is gene group Pb2, which is a gene group consisting of FABP5, JARID2, PID1, AREG, LGALS2, CXCL16, and TIMP1.

11. The method according to claim 1, wherein the gene group Q in step (d) is gene group Qd2, which is a gene group consisting of FOS, LEF1, CCR7, TCF7, SELL, and TSHZ2.

12. The method according to claim 1, further comprising the following steps (e) or (f): (e) calculating the ratio of the relative cell number obtained in step (c) to the relative cell number obtained in step (a); and (f) calculating the ratio of the relative cell number obtained in step (a) to the relative cell number obtained in step (c).

13. A method according to any one of claims 1 to 12, wherein at least one of steps (a) to (f) includes a step of comparing a measured value with a reference value, and the effect of treating hepatocellular carcinoma with a combination therapy of multiple drugs including an immune checkpoint inhibitor is determined based on the result of the comparison.

14. A method for predicting the effect of a combination therapy of multiple drugs including an immune checkpoint inhibitor on hepatocellular carcinoma using peripheral blood of a patient with hepatocellular carcinoma as a sample, the method using at least one indicator selected from the group consisting of the following (1) to (6): (1) the relative cell number of CD14-positive cells; (2) the expression level of at least one gene selected from gene group P in CD14-positive cells; (3) the relative cell number of CD4-positive cells and positive for a protein encoded by at least one gene selected from gene group Q; (4) the expression level of at least one gene selected from gene group Q in CD4-positive cells; (5) the ratio of said indicator (3) to said indicator (1); (6) the ratio of said indicator (1) to said indicator (3). Here, the gene group P is LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, IL1RN, FABP5, JARID2, AREG, CXCL16, CCL2, CTSL, CCL7, IL1B, G0S2, IER3, ASPH, CCL3, MAFB, IFI30, SGK1, H MOX1, CEBPD, RGCC, CST3, ATP2B1, FCN1, MARCKS, CLEC7A, FTH1, DUSP6, S100A11, CD14, S100A9, HLA-DRA, MTSS1, CD300E, ATP2B1-AS1, DOCK4, SERPINA1, PLAUR, AIF1, TYROBP, SAT1, S100A10, FTL, and GRN; wherein the gene group Q consists of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, SELL, TCF7, FOS, SOCS3, FHIT, MYC, GIMAP5, TRABD2A, PIM2, ITGA6, ADTRP, EDA, LTB, RPS13, RPS3A, GAS5, NOSIP, RPS5, EEF1B2, RPL32, LDHB, RPL22, RPL5, RPS12, RPL30, RPS8, RPL11, EEF1G, RPS23, SNHG29, RPL35A, EIF3E, RPS28, RPL39, and RPL9.

15. A testing device for predicting the effect of a hepatocellular carcinoma treatment using a combination therapy of multiple drugs including an immune checkpoint inhibitor, using peripheral blood from a hepatocellular carcinoma patient as a sample, the device comprising at least one detection unit selected from the group consisting of the following (h) to (k): (h) a relative cell number detection unit that detects the relative cell number of CD14-positive cells; (i) an expression level detection unit that detects the expression level of at least one gene selected from gene group P in CD14-positive cells; (j) a relative cell number detection unit that detects the relative cell number of CD4-positive cells and positive for a protein encoded by at least one gene selected from gene group Q; and (k) an expression level detection unit that detects the expression level of at least one gene selected from gene group Q in CD4-positive cells. Here, the gene group P is LYZ, CXCL8, TIMP1, LGALS2, CD68, LGALS3, CXCL2, PID1, APOBEC3A, PPARG, PHLDA1, IL1RN, FABP5, JARID2, AREG, CXCL16, CCL2, CTSL, CCL7, IL1B, G0S2, IER3, ASPH, CCL3, MAFB, IFI30, SGK1, H MOX1, CEBPD, RGCC, CST3, ATP2B1, FCN1, MARCKS, CLEC7A, FTH1, DUSP6, S100A11, CD14, S100A9, HLA-DRA, MTSS1, CD300E, ATP2B1-AS1, DOCK4, SERPINA1, PLAUR, AIF1, TYROBP, SAT1, S100A10, FTL, and GRN; The gene group Q consists of CCR7, LEF1, MAL, TSHZ2, PRKCQ-AS1, RCAN3, SELL, TCF7, FOS, SOCS3, FHIT, MYC, GIMAP5, TRABD2A, PIM2, ITGA6, ADTRP, EDA, LTB, RPS13, RPS3A, GAS5, NOSIP, RPS5, EEF1B2, RPL32, LDHB, RPL22, RPL5, RPS12, RPL30, RPS8, RPL11, EEF1G, RPS23, SNHG29, RPL35A, EIF3E, RPS28, RPL39, and RPL9.

Images