Testing method for sensitivity to immune checkpoint inhibitors

Abstract

The present invention addresses the problems of providing a biomarker that serves as an indicator of susceptibility to an immune checkpoint inhibitor in liver cancer and providing suitable treatments for individual liver cancer patients.　Liver cancer patients were stratified according to prognosis and tumor immune microenvironment, and a new relationship was clarified between fatty liver cancer and a tumor immune microenvironment that is immunopotentiating but also immunodepleting. Furthermore, fatty liver cancer patients were discovered to be susceptible to immunotherapy using immune checkpoint inhibitors. Employing the fat content in liver cancer tissue as an indicator of susceptibility to immune checkpoint inhibitors makes it possible to predict the success of an immune checkpoint inhibitor and to provide suitable treatments for individual liver cancer patients. The fat content can be calculated from an image of the liver cancer tissue or a signal for image display.

Description

[Technical Field]

[0001] The present invention relates to a method for testing sensitivity to immune checkpoint inhibitors, a testing device, and a testing program. This application claims priority from Japanese Patent Application No. 2022-074111, which is incorporated herein by reference. [Background technology]

[0002] Liver cancer is the fourth leading cause of cancer-related deaths worldwide (Non-Patent Document 1). Immunotherapy has become the standard treatment for hepatocellular carcinoma (HCC), but its effectiveness remains limited. HCC is the most common type of primary liver cancer and is a heterogeneous disease with various etiologies (Non-Patent Documents 2 and 3). Hepatitis C virus (HCV) is one of the major causes of HCC, but recent advances in surveillance and treatment have led to a global decline in the prevalence of HCV-associated HCC (HCV-HCC) (Non-Patent Document 4). On the other hand, the prevalence of non-viral HCC is rapidly increasing, with various causes including heavy alcohol consumption, non-alcoholic fatty liver disease (NAFLD), and diabetes mellitus (DM) (Non-Patent Document 5). To understand the diversity of HCC and develop targeted therapies, it is necessary to elucidate the molecular mechanisms underlying its carcinogenesis. Therefore, profiling of HCC at the gene and transcriptome levels has been conducted (Non-Patent Documents 6-8). However, the relationship between molecular and clinicopathological features in non-viral hepatocellular carcinoma has not been fully elucidated.

[0003] In recent years, immune checkpoint inhibitors (ICIs) have been shown to be highly effective against various types of solid tumors (Non-Patent Document 9). Immune checkpoint inhibitors include monoclonal antibodies against cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), programmed cell death protein 1 (PD-1), and its ligand PD-L1 (CD274). In 2020, the IMbrave150 trial demonstrated that combination therapy with atezolizumab (anti-PD-L1 antibody) and bevacizumab (anti-VEGF antibody) significantly improved progression-free survival (PFS) and overall survival (OS) compared with sorafenib in patients with unresectable hepatocellular carcinoma (HCC) (Non-Patent Document 10). Currently, combination immunotherapy is attracting attention in the field of HCC. While combination therapy with atezolizumab and bevacizumab is the first-line treatment for advanced hepatocellular carcinoma (HCC), many patients still experience ineffective results. On the other hand, if the selected drug is ineffective, the treatment method will be changed, such as changing the drug, but if this is repeated, liver reserve will gradually decrease. Therefore, it is important to select the appropriate drug for each patient.

[0004] The tumor immune microenvironment (TIME) has generally been stratified into immune-exclusion, immune-activated, and immune-exhausted subtypes based on the levels of tumor-infiltrating lymphocytes (TILs) and immune checkpoint expression (Non-Patent Documents 11 and 12). Meta-analyses have shown that TIL levels and PD-1 / PD-L1 expression are positively correlated with response to ICIs in various types of cancer (Non-Patent Documents 13 and 14). However, the heterogeneity of tumor immunity in hepatocellular carcinoma, particularly non-viral HCC, and its impact on response to combined immunotherapy have not been elucidated. [Prior art documents] [Non-patent literature]

[0005] [Non-Patent Document 1] CA Cancer J Clin. 2018 Nov;68(6):394-424.

Non-Patent Document 2

Non-Patent Document 3

Non-Patent Document 4

Non-Patent Document 5

Non-Patent Document 6

Non-Patent Document 7

Non-Patent Document 8

[0006] In view of the low efficacy of immunotherapy using immune checkpoint inhibitors in drug therapy for liver cancer, the present inventors aimed to provide a biomarker that serves as an indicator of sensitivity to immune checkpoint inhibitors. [Means for solving the problem]

[0007] To solve the above problems, the present inventors focused on the molecular abnormalities and high heterogeneity of the tumor immune microenvironment in liver cancer patients, and conducted extensive research. They stratified liver cancer patients based on prognosis and the tumor immune microenvironment, and revealed a new correlation between fatty liver cancer and an immune-enhanced yet immune-exhausted tumor immune microenvironment. Furthermore, they found that fatty liver cancer patients are sensitive to immunotherapy using immune checkpoint inhibitors, leading to the completion of the present invention.

[0008] That is, the present invention comprises the following: 1. A method for testing sensitivity to immune checkpoint inhibitors, comprising a step of measuring the fat content in liver cancer tissue. 2. The method according to item 1 above, wherein the fat content is calculated from an image obtained from the liver cancer tissue or a signal for displaying an image obtained from the liver cancer tissue. 3. The method according to the preceding paragraph 2, wherein the image is an MRI image. 4. The method according to the preceding item 3, wherein the MRI image is a chemical shift imaging image. 5. The method according to any one of items 1 to 4 above, further comprising a step of comparing the fat content with a standard value, and determining that the subject is highly sensitive or sensitive to an immune checkpoint inhibitor when the fat content is equal to or greater than the standard value. 6. The method according to item 5 above, wherein the reference value is a value selected from the range of 5% or more and 15% or less. 7. The method according to item 6 above, wherein the reference value for fat content measured by chemical shift imaging is 10%. 8. The method according to any one of the preceding items 1 to 7, wherein the immune checkpoint inhibitor is an anti-PD-L1 antibody or an anti-PD-1 antibody. 9. The method according to any one of the preceding items 1 to 8, wherein the liver cancer is hepatocellular carcinoma. 10. A device for testing sensitivity to an immune checkpoint inhibitor, comprising: a signal detection unit for detecting MR signals from liver cancer tissue; a fat content calculation unit that calculates the fat content in the liver cancer tissue from the detected MR signal; and An output section that compares the calculated fat content with a reference value and indicates the sensitivity level to immune checkpoint inhibitors. An apparatus comprising: 11. A program for testing sensitivity to immune checkpoint inhibitors, inputting MR signal data of liver cancer tissue; Calculating the fat content in the liver cancer tissue from the input MR signal data; and A step of comparing the calculated fat content with a reference value and outputting a sensitivity level. A program that executes. [Effects of the Invention]

[0009] The present invention makes it possible to predict the effectiveness of immune checkpoint inhibitors, enabling the selection of an appropriate treatment method for each individual liver cancer patient.

[0010] Since many liver cancer patients undergo diagnostic imaging rather than tumor biopsy in general practice, the method, device, and program of the present invention are easily applicable to clinical practice. Furthermore, the testing method of the present invention is a testing method that places little burden on patients. [Brief explanation of the drawings]

[0011] [Figure 1]Classification results of non-viral hepatocellular carcinoma based on the tumor immune microenvironment. (A) Comparison of the CIBERSORT scores of total immune cells and cytotoxic T cells between immune and other classes. (B) Comparison of the proportion of each molecular class between immune and other classes. (C) Comparison of the proportion of CTNNB1 or TP53 mutations between immune and other classes. (D) Comparison of the CIBERSORT scores of total immune cells between hepatocellular carcinoma with CTNNB1 mutations ("+" in the figure) and hepatocellular carcinoma without CTNNB1 mutations ("-" in the figure). In the figure, "immune class" refers to the immune class, and "others" refers to other classes. (Test Example 2) [Figure 2] Classification results of steatotic hepatocellular carcinoma based on the tumor immune microenvironment. (A) Representative image of steatotic hepatocellular carcinoma (T and NT represent tumor and non-tumor, respectively). (B) Comparison of CIBERSORT scores of total immune cells between steatotic hepatocellular carcinoma and non-steatotic hepatocellular carcinoma. (C) Comparison of T cell exhaustion signature, stromal signature, and enrichment scores of TGF-β signaling pathway between steatotic hepatocellular carcinoma and non-steatotic hepatocellular carcinoma. (D) Comparison of CIBERSORT scores of M2 macrophages between steatotic hepatocellular carcinoma and non-steatotic hepatocellular carcinoma. In the figure, "steatotic" refers to steatotic hepatocellular carcinoma, and "non-steatotic" refers to non-steatotic hepatocellular carcinoma. (Test Example 2) [Figure 3] Results of immunohistochemical analysis. Representative images of immunohistochemical staining of PD-L1, αSMA, and CD163 in steatotic and non-steatotic hepatocellular carcinoma are shown. In the figure, "steatotic" refers to steatotic hepatocellular carcinoma, and "non-steatotic" refers to non-steatotic hepatocellular carcinoma. (Test Example 2) [Figure 4]Results of analysis of the tumor immune microenvironment. (A) Comparison of the proportion of PD-L1-positive hepatocellular carcinoma between steatotic and non-steatotic hepatocellular carcinoma, (B) Comparison of the proportion of αSMA-positive areas between steatotic and non-steatotic hepatocellular carcinoma, and (C) Comparison of the proportion of CD163-positive areas between steatotic and non-steatotic hepatocellular carcinoma. In the figure, "steatotic" means steatotic hepatocellular carcinoma, and "non-steatotic" means non-steatotic hepatocellular carcinoma. (Test Example 2) [Figure 5] Results of spatial transcriptome analysis of fatty hepatocellular carcinoma. Enrichment scores for immune signature, stromal signature, and T cell exhaustion signature in each cluster are shown. (Test Example 2) [Figure 6] Results of spatial transcriptome analysis of fatty hepatocellular carcinoma. (A) Pie chart showing the proportion of each cluster in spots containing exhausted cytotoxic T lymphocytes (CTLs) defined as CD8A- and NR4A1-positive. (B) Violin diagram showing the expression levels of T cell exhaustion markers CD8A and NR4A1, M2 macrophage marker CD163, CAF markers VIM and TGFB1, and internal standard GAPDH in exhausted CTL spots and other spots. In the figure, "Exhausted CTL" refers to spots containing exhausted cytotoxic T lymphocytes, and "others" refers to other spots. (Test Example 2) [Figure 7] Results of total fatty acid profiling by lipidomics. In the figure, "Steatotic HCC" means fat-containing hepatocellular carcinoma, and "non-Steatotic HCC" means non-fat-containing hepatocellular carcinoma. (Test Example 3) [Figure 8]Results of evaluating the effect of lipid accumulation on tumor cells. (A) BODIPY staining images of Hep3B cells 24 hours after addition of bovine serum albumin (BSA) or palmitic acid (PA). (B) Relative PD-L1 (CD274) mRNA levels in Hep3B cells 24 hours after addition of BSA or PA. (C) Flow cytometry analysis of PD-L1 (CD274) protein levels in Hep3B cells 24 hours after addition of BSA or PA, shown as histograms (left) and mean fluorescence intensity (MFI) (right). In the figures, "-" indicates the addition of bovine serum albumin, and "+" indicates the addition of palmitic acid. (Test Example 3) [Figure 9] Results of evaluation of the effect of lipid accumulation on tumor cells. (A) Relative mRNA levels of CSF1, CXCL8, and TGFB1 in Hep3B cells 24 hours after addition of BSA or PA. (B) Relative mRNA levels of CD206 and IL10 in macrophages after 3 days of co-culture with BSA- or PA-added Hep3B cells. (C) Relative mRNA levels of TGFB1 in LX-2 cells after 3 days of co-culture with BSA- or PA-added Hep3B cells. In the figure, "-" indicates the addition of bovine serum albumin (no lipid accumulation), and "+" indicates the addition of palmitic acid (lipid accumulation). (Test Example 3) [Figure 10] Results of evaluation of the effect of lipid accumulation on tumor cells. The relative mRNA levels of PD-L1 (CD274), CSF1, CXCL8, TGFB1, CD206, and IL10 in steatotic and non-steatotic hepatocellular carcinoma are shown. In the figure, "steatotic" refers to steatotic hepatocellular carcinoma, and "non-steatotic" refers to non-steatotic hepatocellular carcinoma. (Test Example 3) [Figure 11] Correlation diagram between FFCSI and histological fat accumulation in 20 surgically resected hepatocellular carcinoma samples (Example 2). [Figure 12]Identification of fatty hepatocellular carcinoma by MRI. (A) (A1) In-phase T1-weighted gradient-echo MR image (showing a distinct high-signal area (arrow) just below the diaphragm side of liver slice VIII), (A2) Out-phase T1-weighted gradient-echo MR image corresponding to A1 (showing decreased signal intensity of the tumor (arrow)), (A3) Hepatic arterial phase Gd-EOB-DTPA-enhanced MR image (showing arterial enhancement of the tumor (arrow)), (A4) Hepatocellular phase Gd-EOB-DTPA-enhanced MR image 20 minutes later (showing decreased signal intensity of the tumor (arrow)), (B) Hematoxylin-eosin image of a tumor biopsy specimen. The bar in the figure represents 200 μm. (Example 3) [Figure 13] Results of evaluation of the therapeutic response of steatotic hepatocellular carcinoma to immunotherapy. (A) Kaplan-Meier analysis of progression-free survival time for patients stratified by the presence or absence of steatosis in hepatocellular carcinoma, and (B) the results of disease control rate (DCR) for patients stratified by the presence or absence of steatosis. In the figure, "Steatotic HCC" means steatotic hepatocellular carcinoma, and "non-Steatotic HCC" means non-steatotic hepatocellular carcinoma. (Example 3) DETAILED DESCRIPTION OF THE INVENTION

[0012] The present invention relates to a method for testing sensitivity to immune checkpoint inhibitors, a testing device, and a testing program.

[0013] Immune checkpoint inhibitors conceptually reactivate exhausted effector T cells and are more effective in tumor-infiltrating cytotoxic T cells and cancers with high expression of PD-L1 (JAMA Oncol. 2019 Aug 1;5(8):1195-1204.). Immune checkpoint inhibitors for which sensitivity can be assessed 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-1 antibody, anti-PD-L1 antibody, anti-PD-L2 antibody, anti-CTLA4 antibody, anti-TIM-3 antibody, anti-LAG-3 antibody, or anti-TIGIT antibody. Suitable immune checkpoint inhibitors include an anti-PD-L1 antibody and an anti-PD-1 antibody, with an anti-PD-L1 antibody being more preferred.

[0014] The testing method of the present invention can also evaluate sensitivity to combined therapy of an immune checkpoint inhibitor with another drug. Examples of other drugs include angiogenesis inhibitors, anticancer antibiotics, and hormone therapy agents. Examples of angiogenesis inhibitors include, but are not limited to, preparations containing at least one antibody against vascular endothelial growth factor or an antibody against its receptor, such as an anti-VEGF antibody or an anti-VEGFR2 antibody. Preferred other drugs include angiogenesis inhibitors, and more preferably, anti-VEGF antibodies.

[0015] The immune checkpoint inhibitor of the present invention is preferably an immune checkpoint inhibitor used in combination with other drugs, more preferably an immune checkpoint inhibitor used in combination with an angiogenesis inhibitor, even more preferably an anti-PD-L1 antibody or anti-PD-1 antibody used in combination with an angiogenesis inhibitor, and most preferably an anti-PD-L1 antibody used in combination with an anti-VEGF antibody.

[0016] In the present invention, the term "antibody" includes polyclonal antibodies, monoclonal antibodies, antigen-binding fragments of said antibodies, chimeric antibodies comprising said antigen-binding fragments, recombinant antibodies, and derivatives thereof.

[0017] Liver cancers that are the subject of the testing method of the present invention include metastatic liver cancer and primary liver cancer, and are preferably primary liver cancers such as hepatocellular carcinoma and cholangiocarcinoma, and more preferably hepatocellular carcinoma.

[0018] Sensitivity to immune checkpoint inhibitors can also be interpreted as responsiveness to immune checkpoint inhibitors. Cancer immunotherapy using immune checkpoint inhibitors is used in many liver cancer patients, but not all patients achieve the same results; some patients have high sensitivity and others have low sensitivity. Furthermore, even if a therapeutic effect is observed, some patients later develop resistance. Sensitivity to immune checkpoint inhibitors can be tested before or during treatment. If testing reveals high sensitivity, good anticancer effects can be achieved by initiating or continuing the administration of immune checkpoint inhibitors. If testing reveals low sensitivity, administration of other drugs or other treatment methods can be selected. This provides patients with an appropriate treatment method, enabling treatment to be continued without reducing hepatic reserve.

[0019] The testing method of the present invention uses fat content as an index of sensitivity to immune checkpoint inhibitors and is characterized by including a step of measuring fat content in liver cancer tissue. There are no particular limitations on the method for measuring fat content in liver cancer tissue. Preferably, the fat content of the present invention is calculated from an image obtained from liver cancer tissue or a signal for image display obtained from liver cancer tissue. Images include medical images such as MRI images, CT images, ultrasound images, stained images of tissue biopsies, and pathological tissue images. In the present invention, MRI images are images obtained by magnetic resonance imaging (MRI), preferably chemical shift imaging images. In the present invention, chemical shift imaging images are images obtained by chemical shift imaging. Known methods can be used to obtain images or signals for image display from liver cancer tissue, including, for example, abdominal MRI, abdominal CT (computed tomography), abdominal ultrasound, and HE staining of liver biopsy tissue sections. In the present invention, "liver cancer tissue" refers to tumor tissue in the liver, and is not particularly limited as long as it can be used as a test subject for sensitivity to immune checkpoint inhibitors. Preferably, it is tumor tissue in the liver of a test subject suffering from or suspected of suffering from liver cancer. The test subject is not particularly limited, but is preferably a human or a companion animal, more preferably a human. The term "cancer tissue" or "tumor tissue" refers to tissue containing at least one tumor cell, and may include connective tissue and blood vessels that support the tumor.

[0020] The fat content in the present invention is not particularly limited as long as it is an index indicating the proportion of fat in liver cancer tissue. The fat content may be, for example, the proportion of fat in the entire liver cancer tissue, or the proportion of fat measured, calculated, or estimated in a two-dimensional or three-dimensional region of a portion of liver cancer tissue. The fat content may be the ratio of the number of cells having lipid droplets to the number of cells in liver cancer tissue (the ratio of cells with lipid droplets, hereinafter also referred to as the steatosis rate). It may also be the proportion of the area or volume of lipid droplets obtained from an image. It may also be the proportion of fat components in liver cancer tissue. It may also be the proportion of fat calculated from signals for image display. It may also be the steatosis rate calculated from an image obtained from liver cancer tissue or the fat content calculated from signals for image display obtained from liver cancer tissue. More specific examples include the steatosis rate calculated from a stained image of a tissue biopsy, and the fat content calculated by decomposing the net MRI MR signal intensity into fat signal intensity and water signal intensity and then calculating fat signal intensity / (fat signal intensity + water signal intensity).

[0021] The method for obtaining images for measuring fat content in the testing method of the present invention is not particularly limited and may be invasive or non-invasive. However, since many patients with unresectable advanced liver cancer who are candidates for immunotherapy can be diagnosed and treated without tumor biopsy, non-invasive methods are preferred. Examples of invasive methods include tissue biopsy. Specific examples include quantification based on the proportion of cells containing large lipid droplets in stained images of liver cancer tissue biopsies. Non-invasive methods are not particularly limited and include, for example, MRI, CT, ultrasound, etc., with MRI being more preferred from the perspective of detection accuracy.

[0022] The method for measuring fat content using MRI is similar to existing methods, such as proton nuclear magnetic resonance spectroscopy ( 1H-MRS, 3-point Dixon (DIXON), multi-echo gradient-echo (MEGE), chemical shift imaging, and frequency-selective imaging can be applied (Magn Reson Med Sci. 2011;10(1):41-8; Radiographics. 2009 Jan-Feb;29(1):231-60.). A suitable method for measuring the fat content is a method using chemical shift imaging (CSI). More preferably, the fat content can be calculated by the following formula (1): fat fraction measured by CSI (FF) CSI ) can be calculated based on the intensities of the water and fat signals resolved from the MR signal, and more specifically, can be calculated from the ratio of the fat signal intensity to the sum of the fat signal intensity and water signal intensity.

[0023]

number

[0024] In formula (1), “S fat ” is the signal intensity of fat, “S water " represents the signal intensity of water. Equation (1) can be converted to the following equation (2). In equation (2), the fat content can be calculated from the signal intensity in a region of interest (ROI) common to the in-phase image and the opposite-phase image of the MR signal frequency obtained from the protons of water molecules and the protons of the methylene groups of fat molecules. Specifically, a region of interest is set on the tumor tissue of the MRI image, preferably the whole or a part of the tumor tissue of the MRI image, more preferably the whole tumor tissue of the MRI image, and the average signal intensity of the region of interest in the in-phase image is IP and the average signal intensity of the region of interest in the opposite-phase image is OP, which can be applied to equation (2). The part of the tumor tissue is not particularly limited, but may be, for example, a 1 mm 2 area within the outline of the tumor tissue. 2 ~10,000mm 2, preferably 20 mm 2 ~1000mm 2 The region of interest can be a portion having an area of ​​100 μm or less. In setting the region of interest, tumor tissue can be identified using an MRI image enhanced with a contrast agent. The contrast agent used to identify tumor tissue is not particularly limited as long as it can be used to detect liver cancer, and examples include gadoxetate sodium (Gd-EOB-DTPA) and superparamagnetic iron oxide (SPIO).

[0025]

number

[0026] In equation (2), "IP" indicates the in-phase signal intensity (IP) expressed by the following equation (3), and "OP" indicates the opposed-phase signal intensity (OP) expressed by the following equation (4) (Radiographics. 2009 Jan-Feb; 29(1): 231-60.).

[0027]

number

[0028] In formula (3), “S water ” is the signal strength of water, “S fat " represents the signal intensity of fat.

[0029]

number

[0030] In formula (4), “S water ” is the signal strength of water, “S fat " represents the signal intensity of fat.

[0031] The testing method of the present invention can further include a step of comparing the fat content with a reference value. The testing method of the present invention can determine the sensitivity level to an immune checkpoint inhibitor by comparing the measured fat content with a predetermined reference value. Examples of sensitivity levels include high sensitivity, sensitive, low sensitivity, resistant, and moderate sensitivity. The sensitivity level can also be expressed numerically in stages, such as sensitivity level 1 and sensitivity level 2. For example, if the calculated fat content is equal to or greater than a predetermined reference value, the sensitivity level can be determined to be high sensitivity or sensitive to an immune checkpoint inhibitor. If the calculated fat content is less than the predetermined reference value, the sensitivity level can be determined to be low sensitivity, high resistance, or resistant to an immune checkpoint inhibitor.

[0032] The reference value for determining whether a tumor is highly sensitive or sensitive to an immune checkpoint inhibitor is a value selected from the range of 1% to 40%, for example, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 17%, 20%, 25%, 30%, and 40%. A preferred value is a value selected from the range of 3% to 20%, more preferably 5% to 15%, and even more preferably 5% to 10%. The reference value for determining whether a tumor is highly sensitive, resistant, or resistant to an immune checkpoint inhibitor is a value selected from the range of 1% to 40%, for example, 40%, 30%, 25%, 20%, 17%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, and 1%. The value is preferably selected from the range of 3% or more and 20% or less, more preferably 5% or more and 15% or less, and even more preferably 5% or more and 10% or less.

[0033] A reference value for fat content can be set for each method of measuring fat content. By comparing the reference value with the fat content, the sensitivity level to immune checkpoint inhibitors (high sensitivity, low sensitivity, resistance, moderate sensitivity, etc.) can be determined. The reference value for the steatosis rate (the ratio of the number of cells with lipid droplets to the number of cells in liver cancer tissue (the ratio of cells with lipid droplets)) and the fat content FF measured by chemical shift imaging are shown below. CSI The setting of the reference value in (%) will be further explained.

[0034] The reference value for the lipid conversion rate can be set as follows: The reference value for determining whether a subject is highly sensitive or sensitive to an immune checkpoint inhibitor is a value selected from the range of 1% to 40%, for example, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 17%, 20%, 25%, 30%, or 40%. A value selected from the range of preferably 3% to 15%, more preferably 4% to 10%, and even more preferably 5% can be used. The reference value for determining whether a tumor is insensitive, highly resistant, or resistant to an immune checkpoint inhibitor is a value selected from the range of 1% to 40%, for example, 40%, 30%, 25%, 20%, 17%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, and 1%. A value selected from the range of preferably 3% to 15%, more preferably 4% to 10%, and even more preferably 5% can be used.

[0035] Fat content FF measured by chemical shift imaging CSIThe reference value in (%) can be set as follows: The reference value for determining whether a patient is highly sensitive or sensitive to an immune checkpoint inhibitor is a value selected from the range of 3% to 40%, for example, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 17%, 20%, 25%, 30%, or 40%. A value selected from the range of preferably 5% to 20%, more preferably 5% to 15%, even more preferably 8% to 12%, and most preferably 10%. The reference value for determining whether a tumor is insensitive, highly resistant, or resistant to an immune checkpoint inhibitor is a value selected from the range of 3% to 40%, for example, 40%, 30%, 25%, 20%, 17%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, and 3%. A value selected from the range of preferably 5% to 20%, more preferably 5% to 15%, even more preferably 8% to 12%, and most preferably 10%. FF CSI (%) has a strong positive correlation with the fat content. CSI The reference value for (%) may be the same as the reference value for the fat content rate described above.

[0036] Although the scope of the present invention is not limited to the examples described below, the present inventors have demonstrated in the examples that fatty hepatocellular carcinoma, in which the proportion of cells containing large lipid droplets in liver cancer tissue is equal to or greater than a reference value, is characterized by an immune-exhausted tumor immune microenvironment with high PD-L1 expression, and further demonstrated that fatty hepatocellular carcinoma is highly sensitive to immune checkpoint inhibitor therapy.

[0037] (Device for testing susceptibility to immune checkpoint inhibitors) The device for testing sensitivity to immune checkpoint inhibitors of the present invention has one or more of the following configurations. a signal detection unit for detecting MR signals from liver cancer tissue; a fat content calculation unit that calculates the fat content in the liver cancer tissue from the detected MR signal; and An output section that compares the calculated fat content with a reference value and indicates the level of sensitivity to immune checkpoint inhibitors.

[0038] The signal detection unit detects MR signals generated by nuclear magnetic resonance. The detected MR signals are transmitted and received as signal data to the fat content calculation unit by a known data transmission / reception means. The fat content calculation unit calculates the fat content from the MR signals detected by the signal detection unit. A preferred calculation method involves decomposing the net MR signal into a water signal and a fat signal, and calculating the fat content from the fat signal intensity relative to the sum of the fat signal intensity and the water signal intensity. The output unit compares the fat content calculated by the fat content calculation unit with a predetermined reference value and indicates the sensitivity level to the immune checkpoint inhibitor based on the comparison result. The sensitivity level may be the presence or absence of sensitivity, or may be quantified or stratified as a sensitivity level according to the level of fat content. The testing device of the present invention may be a standalone device or a device externally connected to a known MRI device. When the testing device is externally connected to the MRI device, the signal detection unit is included in the MRI device, and the fat content calculation unit and output unit are included in a computer externally connected to the MRI device and equipped with a CPU, a storage medium, etc. The embodiments of the respective components of the inspection device of the present invention are the same as those described in the inspection method.

[0039] (Program for testing sensitivity to immune checkpoint inhibitors) The program for testing susceptibility to immune checkpoint inhibitors of the present invention carries out any one or more of the following steps. inputting MR signal data of liver cancer tissue; Calculating the fat content in the liver cancer tissue from the input MR signal data; and A step of comparing the calculated fat content with a reference value and outputting a sensitivity level.

[0040] The step of inputting MR signal data involves inputting MR signal data generated by nuclear magnetic resonance. The step of calculating the fat content calculates the fat content from the input MR signal data. A preferred calculation method involves decomposing net MR signal data into water signal data and fat signal data, and calculating the fat content from the fat signal intensity relative to the sum of the fat signal intensity and the water signal intensity. The step of outputting the sensitivity level involves comparing the calculated fat content with a preset reference value and outputting the sensitivity level to an immune checkpoint inhibitor according to the comparison result. Sensitivity to an immune checkpoint inhibitor can be tested by installing the testing program of the present invention in a known MRI device or a device externally connected thereto. Embodiments of each configuration of the testing program of the present invention are the same as those described in the embodiments of the testing method. [Example]

[0041] The present invention will be specifically described below with reference to examples to deepen understanding of the present invention, but these examples are not intended to limit the scope of the present invention. The following tests 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.

[0042] 1. Method (DNA and RNA extraction) Genomic DNA and total RNA were extracted from tissue samples using the DNeasy Blood & Tissue Kit (QIAGEN, Venlo, The Netherlands) and the RNeasy Mini Kit (QIAGEN), respectively, as previously described (Proc Natl Acad Sci U S A. 2018;115:E10417-E10426.) The integrity of the DNA and RNA obtained was confirmed using a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA).

[0043] (RNA sequence analysis) Total RNA was isolated from liver tissue as previously described (Gastroenterology. 2010 Jun;138(7):2487-98, 2498.e1-7.). Library preparation was performed using the TruSeq Stranded mRNA Sample Prep Kit (Illumina, San Diego, CA, USA) on the Apollo Library Prep System (TaKaRa). Sequencing was performed on the Illumina HiSeq 3000 platform in 75-base single-end mode. Sequenced reads were mapped to the human reference genome sequence (hg19) using TopHat v2.1.1. Count data were normalized using the relative log expression (RLE) method or trimmed mean of M values ​​(TMM) method.

[0044] (Genome sequence analysis) After library preparation, custom PCR primer sets for amplicon sequencing were designed using Ion AmpliSeq Designer, targeting the coding regions of 69 genes previously reported to be frequently mutated in HCC (Table 1) and the TERT promoter region. Genomic DNA extracted from 55 pairs of tumor and normal tissues was used for library construction. Two different methods were used for library preparation: the Ion AmpliSeq Library Kit 2.0 (Thermo Fisher Scientific, Waltham, MA, USA) and the Ion AmpliSeq Kit for Chef DL8 (Thermo Fisher Scientific). For the former, multiplex PCR was performed using 30 ng of DNA as template. The amplified products were further ligated to barcode adapters (IonXpress Barcode Adapters; Thermo Fisher Scientific) for sample identification and purified using Agencourt AMPure XP Reagent (Beckman Coulter, Brea, CA, USA). For the latter, 10 ng of DNA was used as template. Multiplex PCR, barcode adapter ligation (IonCode Barcode Adapters; Thermo Fisher Scientific), and sample purification were performed automatically using the Ion Chef system (Thermo Fisher Scientific). Emulsion PCR was performed with 14–16 samples per Ion PI chip v3 using the Ion PI HI-Q Chef Kit (Thermo Fisher Scientific), followed by library sequencing. Libraries were sequenced by Ion Proton using the Ion PI Hi-Q Sequencing 200 Kit.The FASTQ files obtained from sequencing were aligned to the human reference genome sequence (GRCh38) using Burrows-Wheeler Alignment tool v0.7.12 (Bioinformatics. 2009 Jul 15;25(14):1754-60). The alignments were converted from Sequence Alignment Map (SAM) format to Binary Alignment Map (BAM) files using SAMtools v1.9 (Bioinformatics. 2009 Aug 15;25(16):2078-9). Realignment and recalibration of base quality scores were performed using Genome Analysis Toolkit version 4.1. For single-nucleotide variant (SNV) analysis, pileup files were created from the BAM files using SAMtools, and somatic mutations were detected using VarScan2 v2.4 (Genome Res. 2012 Mar;22(3):568-76). Among the filtered tumor tissue-specific mutations, those with a minor allele frequency of less than 0.05 and significant at P<0.05 by Fisher's exact test were identified as candidate mutated genes. The filtered variants were annotated using ANNOVAR (Nucleic Acids Res. 2010 Sep;38(16):e164.) and compared with the dbSNP (build 150) and COSMIC (ver. 90) databases. Functional prediction information for nonsynonymous mutations was obtained from the SIFT and Polyphen databases. To detect copy number variations (CNVs), DeCON (Wellcome Open Res. 2016 Nov 25;1:20.) was used to compare copy number variations between BAM files of a set of 55 normal tissue samples and one tumor sample.

[0045] [Table 1]

[0046] (Bioinformatics analysis) Unsupervised hierarchical clustering of RNA-seq data, principal component analysis (PCA), differentially expressed gene detection, and generally applicable gene-set enrichment (GAGE) pathway analysis focusing on Gene Ontology (GO) biological processes were performed using iDEP92 (http: / / bioinformatics.sdstate.edu / idep92 / ). Immune class was determined by nearest neighbor template prediction (NTP) analysis using the immune subclass gene classifier (Gastroenterology. 2017 Sep;153(3):812-826.). Single-sample gene set enrichment analysis (ssGSEA) and NTP were performed using Gene Pattern (https: / / cloud.genepattern.org / gp / pages / index.jsf). CIBERSORT analysis of RNA-seq data was performed in absolute mode using the LM22 signature matrix on the website (https: / / cibersort.stanford.edu / ).

[0047] (immunohistochemistry) Liver samples were embedded in paraffin blocks and stained with hematoxylin and eosin according to standard methods. Steatotic HCC was defined as HCC with lipid droplets in more than 5% of tumor cells. Immunohistochemistry for CD163, αSMA, and PD-L1 was performed in a subset of patients from a cohort of 113 patients. Immunohistochemistry was performed on 3-μm-thick formalin-fixed, paraffin-embedded (FFPE) tissue sections after microwave-induced antigen retrieval using 10 mM TRIS-EDTA (pH 9.0 for CD163, pH 6.0 for αSMA and PD-L1). The primary antibodies used were anti-CD163 (Proteintech, Rosemont, IL, USA), anti-αSMA (Abcam, Cambridge, UK), and anti-PD-L1 (Abcam, clone 28-8). For CD163 and αSMA, staining-positive areas were quantified using a BZ-X700 (KEYENCE, Osaka, Japan) and used for comparative analysis of fat-containing and non-fat-containing HCC. PD-L1 expression was assessed in neoplastic HCC cells. As previously reported for HCC and other cancers (Hepatology. 2016 Dec;64(6):2038-2046), the percentage of neoplastic cells showing membranous staining was recorded, and tumors with at least 1% positive cells were classified as positive.

[0048] (Cell lines and reagents) The Hep3B human hepatoma cell line, the THP-1 human monocyte-derived cell line, and the LX-2 human hepatic stellate cell line were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). These cell lines were cultured in DMEM or RPMI-1640 medium supplemented with 10% fetal calf serum and antibiotics. All cells were confirmed to be pathogen- and mycoplasma-free. Bovine serum albumin (BSA, fat-free) was purchased from Wako Pure Chemical Industries, Ltd. and used as a palmitic acid (PA) control. Palmitic acid was purchased from Sigma-Aldrich, dissolved in ethanol to prepare a stock solution, and diluted in cell growth medium for the assay. Phorbol 12-myristate 13-acetate (PMA) was purchased from Sigma-Aldrich, dissolved in DMSO to prepare a stock solution, and diluted in cell growth medium for the assay. PMA was used as a stimulator for the differentiation of THP-1 cells into macrophages.

[0049] (Palmitic acid added) Hep3B cells were cultured in 6-well plates at 1.0 × 10 5 The cells were plated at a density of 100 cells / well. The next day, the cells were supplemented with palmitic acid at a concentration of 400 μM for 24 hours.

[0050] (Co-culture system of fat-accumulating Hep3B cells with macrophages or fibroblasts) THP-1 cells were plated in 6-well plates at 1.0 × 10 cells per well 1 day before PMA supplementation. 5 The cells were plated at a density of 1.0 × 10 cells / well. The next day, the cells were induced to differentiate into macrophages by supplementing with PMA at a concentration of 10 ng / ml for 24 hours. The next day, the palmitic acid-containing supernatant was removed, and the cells were then co-cultured with adipose-accumulating Hep3B cells in a transwell assay for 72 hours. LX-2 cells were plated at a density of 1.0 × 10 cells / well in a 6-well plate one day before co-culture. 5 The cells were plated at a density of 1000 cells / well. The next day, the palmitic acid-containing supernatant was removed, and the cells were then co-cultured with adipose-accumulating Hep3B cells in a transwell assay for 72 hours.

[0051] (Real-time PCR) Total RNA was extracted from cells using the RNeasy Mini Kit (Qiagen) according to the manufacturer's instructions. RNA was reverse transcribed using SuperScript VILO Master Mix (Thermo Fisher Scientific). qPCR was performed as previously described (Gastroenterology. 2016 Aug;151(2):324-337.e12.) using the QuantStudio 12K Flex RT-PCR System with TaqMan Gene Expression Assay probes (Thermo Fisher Scientific).

[0052] (Flow cytometry) One million cells were mixed with 0.25 ml of phycoerythrin (PE)-conjugated anti-CD274 antibody (0.5 mg / ml) (BioLegend, San Diego, CA, USA) and incubated for 15 minutes at room temperature in the dark. After washing, the cells were analyzed using a Becton Dickinson FACS Canto II flow cytometer (BD Pharmingen, San Diego, CA, USA).

[0053] (Lipidomics-based total fatty acid profiling) Liver fatty acids were measured using a gas chromatography-flame ionization detector (GC-FID) as previously described (Bio Protoc. 2020 May 5;10(9):e3613). Briefly, frozen liver samples were weighed and ground using an oat mill (Tokken Co., Ltd.). Total lipids were extracted from the ground tissue powder using the Bligh and Dyer method (Can J Biochem Physiol. 1959 Aug;37(8):911-7). Methyl tricosanoate (C23:0) was added to the recovered lipids as an internal standard. Fatty acids were then methylated using a Fatty Acid Methylation Kit (Nacalai Tesque, Inc.). Fatty acid methyl esters were measured using a GC-2010 Plus system (Shimadzu Corporation) equipped with a flame ionization detector (FID). The separation capillary column used was a FAMEWAX, 30 m, 0.25 mm inner diameter, 0.25 μm (Restek Corporation). The carrier gas flow rate was 45 cm / sec linear velocity. The oven temperature was initially set at 140 °C, increased to 200 °C at a rate of 11 °C / min, increased to 225 °C at a rate of 3 °C / min, and finally increased to 240 °C at a rate of 20 °C / min, where it was maintained for 5 min. The injection volume was 2 μL in split injection mode. Each fatty acid methyl ester was identified and quantified using a mixture of fatty acid methyl ester standards (Supelco 37 Component FAME Mix and 2-(diisopropylamino)ethyl methacrylate (DPA)(n-3) (Sigma-Aldrich); DPA(n-6) (NU-CHEK PREP, INC., Elysian, MN, USA); DTA(n-6) (Cayman)) for calibration. The values ​​of each fatty acid were normalized to the value of C23:0.

[0054] (statistical analysis) Data are presented as mean ± standard deviation. Statistical analysis was performed using the Mann-Whitney U test to assess differences between unpaired groups. One-way analysis of variance (ANOVA) followed by the Kruskal-Wallis test was used for multiple comparisons. Fisher's exact test was used to analyze categorical data. Correlations were assessed using the Pearson product-moment correlation coefficient. The Kaplan-Meier method and log-rank test were used to analyze differences in overall survival (OS) or progression-free survival (PFS). Univariate and multivariate logistic regression analyses were used to analyze factors associated with immune class in non-viral hepatocellular carcinoma. Odds ratios and 95% confidence intervals (CI) are shown. Unless otherwise noted, a p value of <0.05 indicates statistical significance. Prism ver. 8.4.2 for Mac (GraphPad Prism, RRID: SCR_002798, San Diego, CA, USA) and SPSS software version 24 (IBM, Armonk, NY, USA) were used for the analysis.

[0055] 2. Test Example (Test Example 1) Stratification by multi-omics profiling We studied 113 patients who underwent curative hepatectomy for non-viral hepatocellular carcinoma (HCC) between 2005 and 2018 at the Japan Cancer Research Foundation Cancer Institute Hospital and Kagoshima University Hospital. These patients were confirmed to be free of chronic liver disease, including viral and autoimmune hepatitis. For RNA and DNA sequencing, snap-frozen HCC tissues were used, and formalin-fixed, paraffin-embedded (FFPE) HCC tissues and surrounding liver tissue were used for histological analysis. Tumor RNA sequencing was performed to understand the molecular abnormalities of non-viral HCC. Unsupervised hierarchical cluster analysis of transcriptomics classified the 113 non-viral HCC patients into three molecular classes (class I, II, and III). Class I (n = 36), class II (n = 46), and class III (n = 31). Class I patients had the poorest prognosis, while class III patients had the best prognosis (p < 0.05). Next, we performed cancer genome sequencing of 55 patients with non-viral HCC using a panel of 69 genes for which genetic abnormalities have been reported in multiple cases of HCC (Cell. 2017 Jun 15;169(7):1327-1341.e23.). Somatic mutations were detected in 50 of the 55 cases, with frequent mutations observed in the TERT promoter region (58%), CTNNB1 (36%), and TP53 (18%) (Figure 1C). Integrated analysis revealed that class I was closely associated with TP53 mutations (p<0.05), and class III was closely associated with CTNNB1 mutations (p<0.05). Patients with non-viral HCC harboring TP53 mutations had the poorest prognosis, while patients with CTNNB1 mutations had the best prognosis (TP53 vs. CTNNB1 p<0.05).

[0056] (Test Example 2) Classification of non-viral hepatocellular carcinoma based on the tumor immune microenvironment (TIME) To classify non-viral hepatocellular carcinoma (HCC) based on the tumor immune microenvironment, we performed nearest template prediction (NTP) analysis of tumor transcriptomes from 113 patients in Study 1. NTP analysis identified an immune class, a reported subtype of HCC characterized by strong intratumoral immune cell infiltration (Gastroenterology. 2017 Sep;153(3):812-826). Of the 113 tumors, 43 were classified into this immune class, and CIBERSORT analysis showed significantly higher levels of estimated total intratumoral immune cells and cytotoxic T lymphocytes (CTLs) (Figure 1A). * in Figure 1A indicates p<0.05 for the immune class vs. other classes. This immune class was enriched in class II tumors and showed significantly lower CTNNB1 mutation frequency (Figure 1B, C). * in Figure 1B and Figure 1C indicates p<0.05 for the immune class vs. other classes. Consistently, non-viral HCC with CTNNB1 mutations showed significantly less intratumoral immune cell infiltration (Figure 1D), consistent with a previous report (Clin Cancer Res. 2019 Apr 1;25(7):2021-2023). *In Figure 1D, * indicates p<0.05 for HCC with CTNNB1 mutations vs. HCC without CTNNB1 mutations.

[0057] We explored clinicopathological factors characterizing this immune class and found that steatosis in HCC is strongly associated with this immune class. Indeed, fat-containing HCC, which accounts for 23% of non-viral HCCs (Figure 2A), showed significantly higher levels of total intratumoral immune cells than non-fat-containing HCC (Figure 2B). * in Figure 2B indicates p<0.05 for non-fat-containing HCC vs. fat-containing HCC. Interestingly, ssGSEA and pathway analysis revealed significantly higher levels of T cell exhaustion and stromal signatures, along with TGF-β signaling activation, in fat-containing HCC (Figure 2C). These findings were all hallmarks of tumor immune exhaustion ( Nat Med. 2010 Oct;16(10):1147-51; NatCommun4,2612(2013); Cancer Cell. 2018 Apr;33(4):547-562; Immunity. 2014 Sep;41(3):427-439). * in Figure 2C indicates p<0.05 for non-steatotic versus steatotic HCC. Consistently, steatotic HCC showed upregulation of various immune checkpoints, transcription factors involved in T cell exhaustion, and markers of cancer-associated fibroblasts (CAFs). Furthermore, CIBERSORT analysis demonstrated robust infiltration of M2 macrophages in steatotic HCC, along with upregulation of cytokines and chemokines involved in macrophage M2 polarization and immunosuppression ( Figure 2D ). In Figure 2D, * indicates p<0.05 for non-steat-containing HCC versus steat-containing HCC. Immunohistochemical analysis demonstrated upregulation of PD-L1 on tumor cells and strong infiltration of CAFs and M2 macrophages in non-viral HCC (Figure 3 and Figures 4A-C). In Figures 4A-C, n = 5-13 per group, and * indicates p<0.05 for non-steat-containing HCC versus steat-containing HCC. Overall, steat-containing HCC exhibited an immune-enhanced yet immune-exhausted tumor immune microenvironment characterized by T cell exhaustion, infiltration of M2 macrophages and CAFs, high PD-L1 expression, and activated TGF-β signaling.

[0058] To further characterize the immune-exhausted tumor immune microenvironment in steatotic HCC, we investigated the topography of M2 macrophages, CAFs, and CTLs in steatotic HCC using spatial gene expression analysis on the Visium platform. Steatotic HCC samples were embedded in OCT compound (TissueTek, Sakura) at -80°C in 10mm x 10mm cryomolds and sectioned at 10µm thickness using a Leica CM3050 S microscope. Visium libraries were prepared according to the instructions (Visium Spatial Gene Expression User Guide; CG000239_VisiumSpatialGeneExpression_UserGuide_Rev_A.pdf). Tissues were permeabilized for 3 minutes, which was determined as the optimal time in a tissue optimization time course experiment. Libraries were sequenced to full sequencing depth on a NovaSeq 6000 System (Illumina) using the NovaSeq S4 Reagent Kit (200 cycles, catalog number 20027466, Illumina). Raw FASTQ files and histological images were processed using Space Ranger software v1.2.1 (https: / / support.10xgenomics.com / spatial-gene-expression / software / pipelines / latest / installation). For visualization of spatial expression using histological images, raw Visium files from each sample were imported into Loupe Browser software v4.0.0 (https: / / support.10xgenomics.com / spatial-gene-expression / software / downloads / latest). An average of 288,702 sequence reads were obtained, identifying a median of 3,300 genes per spot.

[0059] Tumor sections were divided into 1,768 spots, and transcriptome data were obtained from each spot. Graph-based clustering divided all spots into six clusters, and cluster 2 showed high expression of immune cell populations (Figure 5). ssGSEA revealed that stromal and T cell exhaustion signatures were enriched in cluster 2 (Figure 5). * in Figure 5 indicates p<0.05 for cluster 2 vs. other clusters. We identified CD8A and NR4A1 double-positive spots as spots containing exhausted CTLs and found that nearly half of these spots were included in cluster 2 (Figure 6A). Next, we compared the transcriptome profiles of spots containing and not containing exhausted CTLs. We found that spots containing exhausted CTLs showed increased expression of the M2 macrophage marker CD163 and the CAF marker VIM, in addition to elevated TGFB1 levels (Figure 6B). * in Figure 6B indicates p<0.05 for spots containing exhausted CTLs vs. other spots. These results suggest that M2 macrophages and CAFs produce TGF-β, promoting the exhaustion of nearby CTLs and forming immune-exhausted TIME in fatty liver cancer.

[0060] (Test Example 3) Evaluation of the effect of lipid accumulation on tumor cells We investigated the mechanistic relationship between the intratumoral state and immune exhaustion TIME in steatotic HCC. To this end, we first performed lipidomics-based total fatty acid profiling and found that palmitic acid (C16:0) and palmitoleic acid (C16:1n-7) levels were increased in steatotic HCC compared with non-steatotic HCC (Figure 7). Therefore, we investigated the effect of palmitic acid accumulation on steatotic HCC cells in vitro. Addition of palmitic acid to Hep3B cells induced lipid accumulation (Figure 8A), upregulation of PD-L1 (CD274) expression at the mRNA and surface protein levels (Figures 8B and 8C), and upregulation of CSF1, CXCL8, and TGFB1 expression levels (Figure 9A). Therefore, we further investigated the effect of palmitic acid accumulation in tumor cells on surrounding macrophages and fibroblasts in vitro. Palmitic acid-treated Hep3B cells upregulated the expression levels of CD206 and IL10 in co-cultured human macrophage cell lines (Figure 9B) and TGFB1 in co-cultured human hepatic stellate cell lines (Figure 9C). In Figures 8B and C and Figures 9A-C, n = 3 per group. * indicates p < 0.05 for the palmitic acid-treated group vs. the BSA-treated (control) group. Upregulation of these immunosuppressive cytokines and chemokines was also observed in steatotic hepatocellular carcinoma in vivo (Figure 10). * indicates p < 0.05 for non-steatotic hepatocellular carcinoma vs. steatotic hepatocellular carcinoma. These data suggest that palmitic acid accumulation in tumor cells promotes immunosuppression and contributes to the development of immune-exhausted TIME in steatotic hepatocellular carcinoma.

[0061] In the above test examples, the inventors used multi-omics profiling to stratify non-viral hepatocellular carcinoma according to prognosis or tumor immune microenvironment (TIME), and clarified the association between intratumoral adipose tissue and immune-exhausted immunotherapy-susceptibility TIME in hepatocellular carcinoma.

[0062] 3. Working Example (Example 1) Measurement of fat content using histopathological images Hepatocellular carcinoma (HCC) tissue sections surgically resected from 20 HCC patients were stained with hematoxylin and eosin (HE), and the number of lipid droplet-containing cells was counted. The ratio of the number of lipid droplet-containing cells to the total number of cells in the HCC tissue was calculated as the fat content (steatosis rate, histological fat accumulation) (Figure 11).

[0063] (Example 2) Measurement of fat content in hepatocellular carcinoma by MRI The 20 HCC patients described in Example 1 underwent abdominal MRI examinations including chemical shift imaging (CSI) before cancer tissue resection surgery. The maximum tumor signal intensity for each patient was obtained by drawing a region of interest (ROI) on the same level in-phase and anti-phase images. The ROI was manually drawn around the tumor outline using Horos™ software (Nimble Co. LLC, d / b / a Purview). The fat fraction measured by CSI (FF) was calculated. CSI ) was calculated using the following formula (2). Hepatocellular carcinoma from five patients with a fat content of 10% or more was determined to be sensitive to immune checkpoint inhibitors (Figure 11).

[0064]

number

[0065] In equation (2), "IP" represents the average signal intensity of the region of interest in the in-phase image, and "OP" represents the average signal intensity of the region of interest in the out-of-phase image (Radiographics. 2009 Jan-Feb;29(1):231-60.).

[0066] Steatosis rate measured by histopathological images and FF measured by chemical shift imaging CSI A strong positive correlation was observed between the two (Figure 11), confirming that MRI is a reliable tool for identifying fat-containing hepatocellular carcinoma.

[0067] Example 3: Identification of Fat-Containing Hepatocellular Carcinoma by MRI and Evaluation of Therapeutic Response to Immunotherapy Between October 2020 and September 2021, combination immunotherapy using atezolizumab (anti-PD-L1 antibody) and bevacizumab (anti-VEGF antibody) was performed for hepatocellular carcinoma (HCC) at Osaka University Hospital and six affiliated hospitals. Thirty patients who underwent abdominal MRI including chemical shift imaging (CSI) before the start of combination immunotherapy were retrospectively enrolled for the immunotherapy trial. Enrollment criteria were patients with measurable liver lesions, patients without significant liver iron deposition, and patients who underwent initial treatment response evaluation. The maximum tumor signal intensity for each patient was obtained by drawing a region of interest (ROI) on in-phase and out-of-phase images at the same level. The ROI was manually outlined using Horos™ software (Nimble Co. LLC, d / b / a Purview). Fat fraction measured by CSI (FF) was calculated. CSI ) was calculated using the above formula (2). FF CSI Fat-containing hepatocellular carcinoma was defined as a tumor with a fat content of 10% or more. As a result, 7 out of 30 patients were classified as having fat-containing hepatocellular carcinoma. There was no significant difference in clinical background between fat-containing and non-fat-containing hepatocellular carcinoma. CSI Chemical shift imaging images of a 35% hepatocellular carcinoma (Figures 12A1-2), Gd-EOB-DTPA-enhanced MR images (Figures 12A3-4), and hematoxylin and eosin images of a tumor biopsy specimen (Figure 12B) are shown.

[0068] During atezolizumab plus bevacizumab combination therapy, dynamic contrast-enhanced CT scans were performed every 6 weeks to assess treatment response, which was assessed using modified Response Evaluation Criteria in Solid Tumors (mRECIST) criteria.

[0069] Patients with steatotic hepatocellular carcinoma did not progress during the 5.4-month observation period. Patients with steatotic hepatocellular carcinoma showed significantly longer progression-free survival (PFS) than patients with non-steatotic hepatocellular carcinoma (Figure 13A). * in Figure 13A indicates p<0.05 for non-steatotic hepatocellular carcinoma vs. steatotic hepatocellular carcinoma. Patients with steatotic hepatocellular carcinoma also had a higher disease control rate (DCR) than patients with non-steatotic hepatocellular carcinoma (Figure 13B). Specifically, FF before the initiation of atezolizumab and bevacizumab combination therapy was significantly higher than that before the initiation of atezolizumab and bevacizumab combination therapy. CSI In patients with FF of 10% or more, the DCR {(CR + PR + SD) / (CR + PR + SD + PD)} was 100%, and treatment was effective in all patients. CSI In patients with a CR of less than 10%, the DCR was 56.5%, meaning that nearly half of the patients did not achieve an adequate therapeutic effect. Patients with fat-containing hepatocellular carcinoma are sensitive to immune checkpoint inhibitors, and it has been shown that intratumoral fat content may be a novel biomarker for predicting the efficacy of immune checkpoint inhibitor treatment in hepatocellular carcinoma.

[0070] The present inventors demonstrated that chemical shift imaging can identify fat-containing hepatocellular carcinoma (HCC). Identification of fat-containing HCC by chemical shift imaging can be easily applied clinically when combined with gadoxetate sodium-enhanced MRI, which is used to diagnose advanced HCC. Patients with fat-containing HCC identified by chemical shift imaging showed significantly longer progression-free survival than patients with non-fat-containing HCC when treated with anti-PD-L1 and anti-VEGF antibodies. These results suggest that intratumoral fat accumulation could be a novel biomarker for predicting the efficacy of immune checkpoint inhibitor therapy for advanced HCC. [Industrial Applicability]

[0071] To predict drug efficacy and provide appropriate treatment in immunotherapy for liver cancer patients.

Claims

1. A testing method that uses the fat content in liver cancer tissue as an indicator of sensitivity to immune checkpoint inhibitors, and determines that the tissue is highly sensitive or sensitive to immune checkpoint inhibitors when the fat content is equal to or higher than a standard value.

2. The method according to claim 1 , wherein the fat content is calculated from an image obtained from the liver cancer tissue or a signal for displaying an image obtained from the liver cancer tissue.

3. The method of claim 2 , wherein the image is an MRI image.

4. The method of claim 3 , wherein the MRI image is a chemical shift imaging image.

5. The method according to claim 1 , wherein the reference value is a value selected from the range of 5% to 15%.

6. 6. The method of claim 5, wherein the reference value for fat content as measured by chemical shift imaging is 10%.

7. The method of claim 1, wherein the immune checkpoint inhibitor is an anti-PD-L1 antibody or an anti-PD-1 antibody.

8. The method of claim 1, wherein the liver cancer is hepatocellular carcinoma.

9. A device for testing sensitivity to an immune checkpoint inhibitor, a signal detection unit for detecting MR signals from liver cancer tissue; a fat content calculation unit that calculates the fat content in the liver cancer tissue from the detected MR signal; and An output section that compares the calculated fat content with a reference value and indicates the sensitivity level to immune checkpoint inhibitors. An apparatus comprising:

10. A program for testing sensitivity to an immune checkpoint inhibitor, inputting MR signal data of liver cancer tissue; calculating the fat content in the liver cancer tissue from the input MR signal data; and A step of comparing the calculated fat content with a reference value and outputting a sensitivity level. A program that executes.

Images