Method for examining possibility of contracting hepatocellular cancer and kit for examining possibility of contracting hepatocellular cancer

Abstract

The purpose of the present invention is to provide a new method that makes it possible to diagnose hepatocellular carcinoma, especially early-stage hepatocellular carcinoma, with high sensitivity, on the basis of new factors instead of AFP or PIVKA-II, and without requiring genetic testing.　The method for examining the possibility that a subject contracts hepatocellular carcinoma includes measuring (a) or (b) in a blood sample isolated from a subject, where (a) is a glycated ferritin amount and the total ferritin amount and (b) is the ratio of the glycated ferritin to the amount of total ferritin or non-glycated ferritin.

Description

Method for detecting the possibility of hepatocellular carcinoma and kit for detecting the possibility of hepatocellular carcinoma

[0001] The present invention relates to a method for testing the possibility of hepatocellular carcinoma, particularly early stage hepatocellular carcinoma, and a kit for testing the possibility of hepatocellular carcinoma.

[0002] The annual death toll from hepatocellular carcinoma (HCC) in Japan is as high as 25,000, making it the fifth most common cause of cancer death by site. Although HCC is one of the more difficult-to-treat cancers, early detection and treatment are important because curative treatment is expected to improve survival rates in early-stage HCC. Known blood tests used to diagnose HCC include α-fetoprotein (AFP), AFP-L3, and protein induced by vitamin K absorption or antagonist-II (PIVKA-II).

[0003] AFP also increases in cancers other than hepatocellular carcinoma, chronic hepatitis, liver cirrhosis, neurodegenerative disease and non-seminoma testicular cancer, so it is considered undesirable to use AFP alone.In addition, in recent years, the number of patients with viral hepatocellular carcinoma has decreased, and the number of patients with non-viral hepatocellular carcinoma has tended to increase.Under these circumstances, compared with the case of viral hepatocellular carcinoma, the AFP value is low in non-viral hepatocellular carcinoma, so it is known that AFP has poor detection accuracy for non-viral hepatocellular carcinoma (see Non-Patent Document 1).

[0004] On the other hand, PIVKA-II is an abnormal prothrombin without coagulation activity that is produced when vitamin K is deficient. When used alone, PIVKA-II has higher sensitivity and specificity than AFP in diagnosing hepatocellular carcinoma. However, PIVKA-II is known to increase and give false positives when warfarin is administered, when cephalosporin antibiotics are used, in patients with obstructive jaundice, in extremely malnutrition, or in patients with alcoholic liver damage. On the other hand, caution is required when administering vitamin K preparations, as PIVKA-II levels normalize even in the presence of hepatocellular carcinoma.

[0005] This simultaneous combined testing of AFP and PIVKA-II is covered by health insurance in Japan for patients with high-risk factors for hepatocellular carcinoma, such as liver cirrhosis or chronic hepatitis, and the Liver Cancer Treatment Guidelines (see Non-Patent Document 2) also recommends measuring two or more tumor markers in the diagnosis of small hepatocellular carcinoma.

[0006] Here, in the above-mentioned liver cancer treatment guidelines, the sensitivity of cancer diagnosis based on AFP or PIVKA-II is less than 72%, and the diagnostic performance is insufficient. Furthermore, when focusing on early-stage hepatocellular carcinoma, AFP alone has a sensitivity of 53% and a specificity of 90%, PIVKA-II alone has a sensitivity of 61% and a specificity of 70%, and the combined use of both improves the sensitivity to 78%, but the specificity is low at 62%, and it has been reported that AFP and PIVKA-II alone have limitations in screening for early-stage hepatocellular carcinoma (see Non-Patent Document 3).

[0007] In recent years, methods for detecting hepatocellular carcinoma using gene methylation as an indicator have been studied. The present inventors have disclosed that a combination of methylated SEPT9 and AFP in serum CpG sequences is useful for diagnosing hepatocellular carcinoma (see Non-Patent Documents 4 and 5).

[0008] Chronic liver inflammation frequently results in intrahepatic iron overload, and iron-induced oxidative stress is known to cause liver damage and contribute to hepatocarcinogenesis (see Non-Patent Document 6). Ferritin is known to be a protein that reflects iron deficiency and iron overload in the body, and the relationship between ferritin itself and hepatocellular carcinoma has been studied (see Non-Patent Document 7). However, the focus has been on the protein itself, and the ferritin glycosylation chain has not been investigated. Rather, because it is negative in early-stage cancer, its diagnostic significance as a tumor marker has been considered limited (see Non-Patent Document 8). Meanwhile, the present inventors have conducted research focusing on ferritin as an iron metabolism marker and discovered that the glycated ferritin / total ferritin ratio is significantly reduced in hepatocellular carcinoma (see Non-Patent Document 9).

[0009] Japanese Patent Application Laid-Open No. 2008-283945

[0010] Kawai K et al., Clinical characteristics of non-B, non-C liver cirrhosis complicated with hepatocellular carcinoma: Survey of non-B, non-C liver cirrhosis in Japan 2011, Kyobunsha 2012; 27-31. Kanehara Publishing Co., Ltd. Liver Cancer Treatment Guidelines 2021 Edition, pp. 34-35. Marrero J et al. Gastroenterology. 2009; 137: 110-118. Kotoh Y, et al. Hepatol Commun 2020; 4: 461-470. Yamazaki A. et al., Hepatocellular carcinoma screening by liquid biopsy using highly sensitive DNA methylation analysis technology: Comparison of diagnostic performance between methylated SST and methylated SEPT9, Yamaguchi Medical Journal, Vol. 70, No. 3, pp. 89-98, 2021. Kato J, et al. Cancer Res. 2001;61(24):8697-702. Asakawa et al., History of Medicine, Vol. 138, No. 5: Page 367 (August 2, 1986) Sato et al., Journal of the Japanese Society of Gastroenterology 15(8):1379-1386 (1982) Ishiguro, A. et al., Proceedings of the 53rd Annual Meeting of the Japanese Society of Medical Laboratory Science, Page 489, General Presentation 150 (2021)

[0011] When diagnosing hepatocellular carcinoma, genetic testing requires a genetic analysis device. Furthermore, the early detection of hepatocellular carcinoma significantly affects the subsequent prognosis. Therefore, an object of the present invention is to provide a new method for detecting hepatocellular carcinoma, particularly early-stage hepatocellular carcinoma, with high sensitivity, without genetic testing, based on a new factor that replaces AFP and PIVKA-II.

[0012] As a result of intensive research aimed at solving the above problems, the present inventors have found that the glycated / total ferritin ratio in a blood sample isolated from a subject can be measured, and the performance of detecting hepatocellular carcinoma can be improved based on the measured value. Furthermore, they have found that hepatocellular carcinoma, particularly early-stage hepatocellular carcinoma, can be detected using the absolute amount of glycated ferritin and the absolute amount of total ferritin in a blood sample isolated from a subject as factors, and have completed the present invention.

[0013] That is, the present invention is as follows: [1] A method for examining the possibility of the subject having hepatocellular carcinoma based on the measured values ​​of (a) the amount of glycated ferritin and the amount of total ferritin, or (b) the ratio of the amount of glycated ferritin to the amount of total ferritin or the amount of non-glycated ferritin in a blood sample isolated from the subject. [2] The method according to [1] above, comprising performing a statistical analysis based on the measured values ​​of (a) or (b) and the levels of PIVKA-II and / or AFP in the blood sample, and examining the possibility of the subject having hepatocellular carcinoma based on the value obtained by the statistical analysis. [3] The method according to [1] or [2] above, comprising performing a statistical analysis based on the measured values ​​of (a) or (b), the levels of PIVKA-II and / or AFP in the blood sample, and age, and examining the possibility of the subject having hepatocellular carcinoma based on the value obtained by the statistical analysis. [4] The method according to [2] or [3] above, wherein the statistical analysis uses logarithmically transformed values ​​of at least one of the measured values ​​(a) or (b) above and the PIVKA-II level and / or AFP level in the blood sample. [5] The method according to [1] or [2] above, wherein the subject has a history of cirrhosis but has not been diagnosed with hepatocellular carcinoma, and the possibility of the subject having early-stage hepatocellular carcinoma is tested within 6 months after the subject is diagnosed with cirrhosis or after being diagnosed with cirrhosis but has not progressed to hepatocellular carcinoma, in order to monitor progression to early-stage hepatocellular carcinoma. [6] The method according to [5] above, wherein the early-stage hepatocellular carcinoma is cancer at Stage 0 or A according to the BCLC staging system. [7] A kit for testing the possibility of a subject having hepatocellular carcinoma, comprising a substance that specifically binds to glycated ferritin.

[0014] According to the present invention, hepatocellular carcinoma, particularly early stage hepatocellular carcinoma, can be detected with high sensitivity by testing proteins or their sugar chains, which is simpler than genetic testing.

[0015] 1A shows the results of Receiver Operating Characteristic (ROC) analysis of the control group and all hepatocellular carcinoma patients (viral hepatocellular carcinoma patients and non-viral hepatocellular carcinoma patients), viral hepatocellular carcinoma patients, and non-viral hepatocellular carcinoma patients, using PIVKA-II and AFP as factors in Example 1. FIG. 1B shows the results of ROC analysis of the control group and all hepatocellular carcinoma patients, viral hepatocellular carcinoma patients, and non-viral hepatocellular carcinoma patients, using a combination of the logarithm of PIVKA-II (LogPIVKA-II), the logarithm of AFP (LogAFP), the glycated ferritin / total ferritin ratio (%GF), the logarithm of the actual measured value of glycated ferritin (LogGF), and the logarithm of the actual measured value of total ferritin (LogFER) as factors in Example 1. Figure 1C is a diagram showing an example of values ​​such as the regression coefficient (β) obtained by the multiple logistic regression analysis in Example 1. Figure 2A shows the results of an analysis in Example 2, in which age was added as a factor to the logarithmic values ​​of the measured values ​​of glycated ferritin and non-glycated ferritin, PIVKA-II or its logarithmic value, AFP or its logarithmic value, and the differentiation group was defined as all stages of hepatocellular carcinoma, and the control group was defined as cirrhosis. Figure 2B shows the results of an analysis in Example 2, in which age was added as a factor to the measured values ​​of glycated ferritin and non-glycated ferritin or their logarithmic values, PIVKA-II or its logarithmic value, AFP or its logarithmic value, and the differentiation group was defined as early hepatocellular carcinoma, and the control group was defined as cirrhosis. Figure 2C is a diagram showing an example of values ​​such as the regression coefficient (β) obtained by multiple logistic regression analysis in Example 2, in which the differentiation group was defined as all stages of hepatocellular carcinoma, and the control group was defined as cirrhosis. Figure 2D is a diagram showing an example of values ​​such as the regression coefficient (β) obtained by performing a multiple logistic regression analysis in Example 2, with the differentiation group being early hepatocellular carcinoma and the control group being liver cirrhosis. Figure 3A shows the results of an analysis in Example 3, in which various factors were combined and the differentiation group was hepatocellular carcinoma (HCC) and the control group was liver cirrhosis (LC). Figure 3B shows the results of combining various factors in Example 3, determining predetermined cutoff values, and determining sensitivity, specificity, and AUC values ​​for all cases and early cancer. Figure 3C is a diagram showing an example of values ​​such as the regression coefficient (β) obtained by the multiple logistic regression analysis in Example 3.Figure 4 shows the results of ROC analysis of the predicted value P calculated by the method of the present invention for assessing the possibility of having hepatocellular carcinoma in Example 4 (referred to as the "present model" in the figure; the same applies to Figure 5) and the GALAD score. Figure 5 shows the results of a comparison of the score distribution of the predicted value P calculated by the method of the present invention for assessing the possibility of having hepatocellular carcinoma in Example 4 and the GALAD score. Figure 6A shows the results of a comparison of the predicted value P calculated by the method of the present invention for assessing the possibility of having hepatocellular carcinoma for each stage of the Barcelona Clinical Hepatocellular Carcinoma Staging Classification in Example 4 and the sensitivity and specificity based on the GALAD score. FIG. 6B shows the results of comparing the predicted value P calculated by the method of the present invention for testing the possibility of having hepatocellular carcinoma, divided into early stage cancer (Early stage: Stages 0 and A), advanced stage cancer (Advanced stage: Stages B, C, and D), and all cancers in Example 4, with the sensitivity and specificity based on the GALAD score.

[0016] The method of the present invention for testing the possibility that a subject has hepatocellular carcinoma is not particularly limited as long as it involves measuring (a) the amount of glycated ferritin and the amount of total ferritin, or (b) the ratio of the amount of glycated ferritin to the amount of total ferritin or non-glycated ferritin in a blood sample isolated from the subject, and testing the possibility that the subject has hepatocellular carcinoma based on the measured values, and is hereinafter referred to as the "method of testing the possibility of hepatocellular carcinoma." Furthermore, the kit for testing the possibility that a subject has hepatocellular carcinoma is not particularly limited as long as it contains a substance that specifically binds to glycated ferritin and is used to test the possibility that a subject has hepatocellular carcinoma, and is hereinafter referred to as the "kit for testing the possibility of hepatocellular carcinoma."

[0017] The subject in this specification is not particularly limited as long as it is a human, and examples thereof include subjects whose cancer status is unknown, subjects with a history of liver cirrhosis, non-hepatocellular carcinoma and non-cirrhotic subjects with other diseases, and healthy individuals. Such subjects also include subjects who have previously suffered from cancer and have since been cured of the cancer, but whose cancer status is unknown at the time of examination.

[0018] Examples of blood samples include serum, plasma, and blood. When measuring (a) or (b) above, the blood sample may be used in an amount of serum or plasma of about 0.01 to 10 mL, or may be 0.02 to 2 mL, or 0.025 to 1 mL. The blood sample may be used in an amount of whole blood of about 0.02 to 20 mL, or may be 0.04 to 4 mL, or 0.05 to 2 mL.

[0019] <Factors> Ferritin as used herein is a spherical iron storage protein capable of storing 2,500 iron molecules, and is abundantly distributed in organs such as the liver, spleen, and heart. The total amount of ferritin in a blood sample isolated from a subject herein refers to the amount of ferritin in the blood sample isolated from the subject, i.e., the sum of the amount of glycated ferritin and the amount of non-glycated ferritin. The glycated ferritin refers to ferritin bound to a sugar chain, and the non-glycated ferritin refers to ferritin without a sugar chain. The sugar chain preferably includes a sugar chain to which jack bean lectin (concanavalin A: ConA) can bind, in other words, a sugar chain having mannose and glucose at the non-reducing end.

[0020] The ratio of the amount of glycated ferritin to the amount of total ferritin or the amount of non-glycated ferritin (hereinafter simply referred to as "glycated ferritin ratio") in this specification can be determined by any of the following methods. <When the amount of total ferritin is used>

[0021]

[0022] <When using non-glycated ferritin>

[0023]

[0024] The total amount of ferritin in a blood sample can be measured by known methods or by using a commercially available ferritin measurement kit. For example, the blood sample can be centrifuged, and a substance that binds to ferritin can be added to the supernatant for measurement. The amount of non-glycated ferritin can be measured by known methods or by using a commercially available ferritin measurement kit. For example, a lectin-binding carrier that binds to glycated ferritin can be added to the blood sample, and the sample can be centrifuged at 11,000 rpm to precipitate glycated ferritin, and the non-glycated ferritin contained in the supernatant can be measured.

[0025] Specific examples of substances that bind to the sugar chains of ferritin include lectins that bind to the sugar chains of ferritin, such as concanavalin A (ConA), wheat germ lectin (WGA), kidney bean lectin (PHA), mushroom lectin (ABA), Japanese elderberry lectin (SSA), and castor bean lectin (RCA); anti-ferritin antibodies that bind to the sugar chains of ferritin; and aptamers that bind to the sugar chains of ferritin.

[0026] The anti-ferritin antibody that binds to the sugar chain of ferritin can be produced by techniques well known to those skilled in the art, such as the hybridoma method or the phage display method. Alternatively, the ferritin gene sequence may be obtained from a database such as the National Center for Biotechnology Information (NCBI), a recombinant ferritin protein may be synthesized by cloning, and a mouse may be immunized with the synthesized recombinant ferritin protein to produce the antibody. Furthermore, a commercially available antibody may be used.

[0027] The type of anti-ferritin antibody that binds to the sugar chain of ferritin may be any type of antibody, including human antibodies, chimeric antibodies, humanized antibodies, and F(ab') ２ , Fab, diabody, Fv, ScFv, or Sc(Fv) ２Examples of antibody fragments include those described above. The above antibodies may be polyclonal or monoclonal. Chimeric or humanized antibodies can be produced by genetic engineering in accordance with standard methods. Antibody fragments can also be produced by digesting full-length antibodies with pepsin or papain, for example.

[0028] The substance that binds to the sugar chain of ferritin may be labeled with a labeling substance. Examples of such labeling substances include enzymes such as peroxidase (e.g., horseradish peroxidase), alkaline phosphatase, β-D-galactosidase, glucose oxidase, glucose-6-phosphate dehydrogenase, alcohol dehydrogenase, malate dehydrogenase, penicillinase, catalase, apoglucose oxidase, urease, luciferase, and acetylcholinesterase; fluorescent substances such as fluorescein isothiocyanate, phycobiliprotein, rare earth metal chelates, dansyl chloride, and tetramethylrhodamine isothiocyanate; green fluorescent protein (GFP), cyan fluorescent protein (CFP), and blue fluorescent protein (Blue Fluorescence Protein); Examples of the fluorescent agent include fluorescent proteins such as yellow fluorescent protein (BFP), yellow fluorescent protein (YFP), red fluorescent protein (RFP), luciferase, radioisotopes such as 3H, 14C, 125I, or 131I, metal colloids, non-metal colloids, dye particles such as dye sols or dispersed dyes, latex particles or colored microparticles, biotin, avidin, or chemiluminescent substances.

[0029] AFP levels and PIVKA-II levels can be measured by known techniques, including enzyme immunoassay (EIA), enzyme-linked immunosorbent assay (ELISA), fluorescent immunoassay (FIA), chemiluminescent immunoassay (CLIA), chemiluminescent enzyme immunoassay (CLEIA), and immunochromatography.

[0030] <Statistical Analysis> Examples of statistical analysis include multivariate analysis such as logistic regression analysis, linear discriminant analysis, or multiple regression analysis. Furthermore, for continuous variables, the Mann-Whitney U test can be used for two-group testing, the Kruskal-Wallis test and Dunn's test can be used for multiple comparisons, and the chi-square test and Fisher's exact test can be used for categorical variable testing. Factors necessary for assessing the possibility of having hepatocellular carcinoma, such as sex, age, alanine aminotransferase (ALT), aspartate aminotransferase (AST), and FIB-4 index (e.g., platelet count), may also be added to the statistical analysis.

[0031] The above-mentioned respective factors, for example, any one, two or more, three or more, four or more, or all of the glycated ferritin amount and total ferritin amount, glycated ferritin ratio, AFP level, and PIVKA-II level, may be logarithmically transformed or power transformed, and a normally distributed value may be used. Note that the above-mentioned "glycated ferritin amount and total ferritin amount" means two amounts, glycated ferritin amount and total ferritin amount, and since the two amounts, glycated ferritin amount and total ferritin amount, are used as one set, in this specification, glycated ferritin amount and total ferritin amount are considered to be one factor.

[0032] The combination of values ​​to be normally distributed is not particularly limited, and specific examples include the following. When only one factor is normally distributed, only one of the following factors can be normally distributed: (a) the amount of glycated ferritin and the amount of total ferritin, or (b) the glycated ferritin ratio, AFP level, and PIVKA-II level in a blood sample isolated from a subject. When two factors are normally distributed, only two of the following factors can be normally distributed: (a) the amount of glycated ferritin and the amount of total ferritin, or (b) the glycated ferritin ratio, AFP level, and PIVKA-II level in a blood sample isolated from a subject. When three factors are normally distributed, only three of the following factors can be normally distributed: (a) the amount of glycated ferritin and the amount of total ferritin, or (b) the glycated ferritin ratio, AFP level, and PIVKA-II level in a blood sample isolated from a subject. When the four factors are normally distributed, all four factors, (a) the amount of glycated ferritin and the amount of total ferritin, or (b) the glycated ferritin ratio, AFP level, and PIVKA-II level, in a blood sample isolated from a subject can be normally distributed. Note that normalizing the amount of glycated ferritin and the amount of total ferritin may mean normalizing both the amount of glycated ferritin and the amount of total ferritin, or may mean normalizing only one of them.

[0033] When logistic regression analysis is performed as the statistical analysis, for example, when age, normally distributed total ferritin (LogFER) value, normally distributed glycated ferritin (LogGF), normally distributed AFP level (LogAFP), and normally distributed PIVKA-II level (LogPIVKA-II) are used as independent variables to calculate a predicted value (P: Predicted Value), the calculation can be based on the following formula (III).

[0034]

[0035] In the above formula (III), β0 is the intercept (constant term), β1 to β5 are regression coefficients, Age is substituted for age, LogFER is substituted for normally distributed total ferritin (FER) values, LogGF is substituted for normally distributed glycated ferritin (GF) values, LogAFP is substituted for normally distributed AFP values, and LogPIVKA-II is substituted for normally distributed PIVKA-II values. Log means normally distributed. When other factors are used as independent variables, the predicted value P can be obtained by further adding or replacing them to the independent variables of the above formula (III).

[0036] The above formula (III) can be obtained by measuring predetermined factors in blood samples separated from the discrimination group and the control group in advance, creating a prediction model formula by logistic regression analysis, and then verifying the discriminant usefulness of the prediction model formula using AUC value, AIC value, etc., and selecting the formula.

[0037] <Testing for the Possibility of Hepatocellular Carcinoma> Testing for the possibility of hepatocellular carcinoma can be performed by measuring (a) the amount of glycated ferritin and the amount of total ferritin, and (b) the glycated ferritin ratio (a) or (b) in a blood sample collected from the subject, and then testing the possibility of the subject having hepatocellular carcinoma based on the measured values ​​or a predetermined cutoff value. Specifically, (a) the amount of glycated ferritin and the amount of total ferritin, and (b) the glycated ferritin ratio (a) or (b) in blood samples separated from a control patient (a control patient is defined as a patient without cirrhosis and without cancer), a cirrhotic patient, and a cancer patient, respectively, are measured in advance. If the measured values ​​of (a) or (b) in the subject are higher than the measured values ​​of (a) or (b) in the control patient or the cirrhotic patient, or are close to the measured values ​​of (a) or (b) in the cancer patient, it can be determined that the subject is highly likely to have hepatocellular carcinoma. Conversely, if the measured value of (a) or (b) in the subject is equal to or less than the measured value of (a) or (b) in the control patient or the patient with cirrhosis, or is close to the measured value of (a) or (b) in the control patient or the patient with cirrhosis, it can be said that the subject is unlikely to be suffering from hepatocellular carcinoma.

[0038] In testing the possibility that a subject has hepatocellular carcinoma, (a) the amount of glycated ferritin and the amount of total ferritin, and (b) the glycated ferritin ratio (a) or (b) are measured in blood samples collected from each of a control patient, a cirrhosis patient, and a cancer patient, and statistical analysis is performed based on the measured values ​​of (a) or (b) for each patient, and, if necessary, the PIVKA-II level and / or AFP level in each patient's blood sample and age, to calculate a cutoff value using the statistical analysis. The possibility that the subject has hepatocellular carcinoma can also be tested by comparing the obtained cutoff value with the measured values ​​of (a) the amount of glycated ferritin and the amount of total ferritin, and (b) the glycated ferritin ratio (a) or (b) in each blood sample collected from the subject, and, if necessary, the PIVKA-II level and / or AFP level in each patient's blood sample and age.

[0039] Examples of such cutoff values ​​include the mean value of "the value obtained by the statistical analysis in the control subjects (control patients and / or cirrhosis patients)", the mean value ± standard deviation (SD), the mean value ± 2SD, the mean value ± 3SD, the median, the median ± SD, the median ± 2SD, the median ± 3SD, etc. Furthermore, to increase sensitivity (the proportion of subjects with cancer who can be correctly determined to be positive) and specificity (the proportion of subjects without cancer who can be correctly determined to be negative), the cutoff value can be calculated using the Youden Index by creating an ROC (Receiver Operating Characteristic) curve using statistical analysis software based on the data of "measurements of (a) or (b) in blood samples collected from liver cancer patients, and, if necessary, values ​​obtained by statistical analysis using the PIVKA-II level and / or AFP level in the blood samples and age as factors" and the data of "measurements of (a) or (b) in blood samples collected from control subjects, and, if necessary, values ​​obtained by statistical analysis using the PIVKA-II level and / or AFP level in the blood samples and age as factors."

[0040] Statistical analysis may be performed to create the formula (III) or the like, and the subject's factors may be input into the formula to determine the predicted value P for testing. The closer the predicted value P is to 1, the higher the likelihood of hepatocellular carcinoma, while the closer the predicted value P is to 0, the lower the likelihood of hepatocellular carcinoma. Therefore, hepatocellular carcinoma can be tested based on this value (predicted value P). The value obtained by the statistical analysis in the control subject may be a value measured each time, or may be a value analyzed in advance. Furthermore, it is preferable that the subject and control subjects to be compared use blood samples prepared by substantially the same method or obtained by substantially the same measurement method.

[0041] <Early Hepatocellular Carcinoma, Viral Hepatocellular Carcinoma> Hepatocellular carcinoma also includes early hepatocellular carcinoma. Early hepatocellular carcinoma refers to hepatocellular carcinoma staged at 0 or A in the BCLC staging system (Barcelona Clinical Liver Cancer Staging Classification) described in Non-Patent Document 4 above. In other words, "the possibility that a subject has hepatocellular carcinoma" can be interpreted as "the possibility that a subject has hepatocellular carcinoma, including hepatocellular carcinoma staged at 0 or A in the BCLC staging system." Hepatocellular carcinoma also includes viral hepatocellular carcinoma. Viral hepatocellular carcinoma refers to primary hepatocellular carcinoma arising from chronic hepatitis and cirrhosis caused by persistent infection with a hepatitis virus. Examples of hepatoviruses include hepatitis A virus (HAV), hepatitis B virus (HBV), hepatitis C virus (HCV), hepatitis D virus (HDV), and hepatitis E virus (HEV).

[0042] In the present specification, a subject has a history of cirrhosis but has not been diagnosed with hepatocellular carcinoma. To monitor progression to early-stage hepatocellular carcinoma, the subject's likelihood of developing early-stage hepatocellular carcinoma may be assessed within six months, five months, four months, or three months after the subject is diagnosed with cirrhosis, or after the subject is diagnosed with cirrhosis but is subsequently found to have no progression to hepatocellular carcinoma. Patients with cirrhosis typically require monitoring for hepatocellular carcinoma every three to six months using abdominal ultrasound or other methods. This method for assessing the likelihood of developing hepatocellular carcinoma can also be used to monitor patients with cirrhosis for the likelihood of developing early-stage hepatocellular carcinoma.

[0043] The kit for testing the possibility that the subject in question has hepatocellular carcinoma may contain a substance that specifically binds to the above-mentioned glycated ferritin, and may further contain a secondary antibody that reacts with the substance that specifically binds to glycated ferritin, a substrate reagent that produces color or fluorescence, and instructions describing how to measure (a) the amount of glycated ferritin and total ferritin, and (b) the glycated ferritin ratio in a blood sample.

[0044] The contents of all patent and non-patent literature cited herein are hereby incorporated by reference in their entirety.

[0045] The present invention will be explained in more detail below with reference to examples, but the technical scope of the present invention is not limited to these examples.

[0046] Example 1 Analysis of Glycated / Total Ferritin Ratio, or Measured Values ​​of Glycated Ferritin and Total Ferritin First, the diagnostic performance of hepatocellular carcinoma was verified by calculating the glycated / non-glycated ferritin ratio, measured values ​​of glycated ferritin, and measured values ​​of total ferritin in the serum of 138 hepatocellular carcinoma (HCC) patients, 101 control patients, and 96 patients with liver cirrhosis, as shown in Table 1. Among the hepatocellular carcinoma patients, 47 had early hepatocellular carcinoma (BCLC stage 0 / A), 84 had viral hepatocellular carcinoma (HBV, HCV, HBV + HCV), and 54 had non-viral hepatocellular carcinoma (nBnC). The progression of hepatocellular carcinoma was determined using the criteria of the BCLC staging system, with Stage 0 / A being considered early stage hepatocellular carcinoma. Of the 47 patients, 29 had viral hepatocellular carcinoma and 18 had non-viral hepatocellular carcinoma.

[0047]

[0048] (Collection of Blood Samples) The blood samples collected from each subject were centrifuged at 2000 g for 10 minutes to separate 3 to 4 mL of serum from 8 mL of blood, which was dispensed into 2 mL tubes and stored frozen at −20° C. until use.

[0049] Glycated and non-glycated ferritin in serum was separated using a modified version of the method for separating glycated and non-glycated ferritin (M Worwood, et al. Clin Sci 1979;56:83-7) that utilizes ConA-coupled Sepharose beads (ConA: Sigma-Aldrich), which specifically bind to sugars. Sepharose-4B beads (Sigma-Aldrich) were used instead of ConA under the same conditions to determine the total ferritin concentration.

[0050] Specifically, ConA-conjugated Sepharose beads or Sepharose-4B beads were washed three times with 10 mM phosphate buffer (PBS) pH 7.4. Next, each of the beads was diluted 3.8-fold with 10 mM phosphate buffer (PBS) to prepare a separation reagent. Next, 238 μL of each of the separation reagents and 13 μL of serum from subjects (138 hepatocellular carcinoma patients and 197 non-cancer patients) were added to a 1.5 mL tube, mixed, and allowed to react at room temperature for 2 hours. During this time, the mixture was stirred every 15 minutes using a vortex mixer. The mixture was then centrifuged at 11,000 rpm for 1 minute, and the ferritin content of the supernatant was measured. When ConA-coupled Sepharose beads were used, non-glycated ferritin was found in the supernatant, and glycated ferritin + ConA was found in the sediment. When Sepharose-4B beads were used, total ferritin (glycated ferritin + non-glycated ferritin) was found in the supernatant, and Sepharose-4B beads were found in the sediment. Ferritin levels in each sample were measured using the AIA-Pack CL Ferritin (Tosoh Corporation) reagent on an AIA-CL2400 (Tosoh Corporation). The ratio of glycated / total ferritin was calculated using the following formula (IV). Hereinafter, the ratio of glycated / total ferritin is also referred to as "%GF." The actual value of glycated ferritin was calculated based on the actual values ​​of non-glycated ferritin and total ferritin.

[0051]

[0052] (Measurement of AFP Levels) The AFP concentration (ng / mL) in 0.005 mL of serum was measured by the CLEIA method. The test reagent used was AIA-Pack CL AFP (Tosoh Corporation), and measurements were performed according to the attached protocol. The measurement device used was AIA-CL2400 (Tosoh Corporation).

[0053] (Measurement of PIVKA-II Level) The PIVKA-II concentration (mAU / mL AU: arbitrarily unit standard) in 0.005 mL of serum was measured as the PIVKA-II level using the CLEIA method. The test reagent used was AIA-Pack CL PIVKA-II (Tosoh Corporation), and measurements were performed according to the attached protocol. The measurement device used was AIA-CL2400 (Tosoh Corporation).

[0054] (Statistical Analysis) Statistical analysis was performed on the relationship between the measured factors (glycated / total ferritin ratio, AFP, PIVKA-II) and the clinical pathological background of the subjects. GraphPad Prism ver. 9 (GraphPad Software) and StatFlex ver. 7 (Artec) were used for the analysis. A chi-square test was performed to test categorical variables. For multivariate analysis, multiple logistic regression analysis was performed using intercepts and regression coefficients according to the same formula as formula (III) above. Receiver Operating Characteristic (ROC) analysis was performed to compare the discriminant performance of each factor combination between the hepatocellular carcinoma patient group and the non-cancer group. The Area Under Curve (AUC) was calculated based on the ROC curve. A statistically significant difference was determined when p<0.05. Furthermore, a cutoff value was set using the Youden Index, and the sensitivity and specificity at that time were calculated.

[0055] First, a prediction model formula was created by performing multiple logistic regression analysis, and then values ​​were substituted into the prediction model formula to calculate the predicted value P. Based on the predicted value P, an ROC analysis was performed between the non-cancer group and the hepatocellular carcinoma patients. The results are shown in Figures 1A and 1B.

[0056] In Figure 1A, "Total" refers to viral hepatocellular carcinoma patients and non-viral hepatocellular carcinoma patients, "Viral" refers to viral hepatocellular carcinoma patients, and "Non-viral" refers to non-viral hepatocellular carcinoma patients. Figure 1A shows the results of a conventional method using PIVKA-II and AFP as factors, and Figure 1B shows the results of an ROC analysis using a combination of the logarithm of PIVKA-II (LogPIVKA-II), the logarithm of AFP (LogAFP), the glycated / total ferritin ratio (%GF), the logarithm of the actual measured value of glycated ferritin (LogGF), and the logarithm of the actual measured value of total ferritin (LogFER) (combination of ○ marks in the factor column in Figure 1B). The logarithm values ​​are values ​​normally distributed by logarithmic transformation.

[0057] The multiple logistic regression analysis used a formula similar to the above formula (III), and the factors marked with a circle in the factor column in Figures 1A and 1B were used as independent variables. In No. 2 of Figure 1B, when multiple logistic regression analysis was performed by combining LogPIVKA-II, LogAFP, and %GF as independent factors, the intercept (β of order 0), regression coefficient (β of orders 1 to 3), standard error of the β coefficient (SEβ), z score (z) indicating the significance of the explanatory variable, significance probability (P) of the explanatory variable, e β The diagnostic ability (odds ratio) of the explanatory variables and the confidence intervals (lower limit and upper limit of 95% Cl) of the odds ratio are shown in Figure 1C. Note that the regression coefficients (β) in the formula when the multiple logistic regression analysis in Figure 1A was performed and the regression coefficients (β) in the formula when the multiple logistic regression analysis was performed using a combination of factors other than No. 2 in Figure 1B are omitted.

[0058] 1A and 1B, it was confirmed that the AUC value is improved by using the glycated ferritin ratio (No. 2 in FIG. 1B), or the actual measured values ​​of glycated ferritin and total ferritin (No. 3-5 in FIG. 1B) compared to the use of conventional PIVKA-II or AFP, and that the AUC value is particularly high in non-viral cancers. Furthermore, by using the actual measured values ​​of glycated ferritin and total ferritin without using both PIVKA-II and AFP as factors, it was confirmed that the AUC value is equivalent to that when conventional PIVKA-II or AFP is used for viral cancers, and higher than that when PIVKA-II or AFP is used for non-viral cancers. Furthermore, it was confirmed that by using the actual measured values ​​of glycated ferritin and total ferritin with conventional PIVKA-II or AFP, it is possible to obtain high AUC values ​​for both viral and non-viral cancers.

[0059] Example 2: Analysis with Age Added as a Factor In Example 2, age was added as an additional factor and multiple logistic regression analysis was performed to create a prediction model formula. Values ​​were then substituted into the prediction model formula to calculate the predicted value P, and ROC analysis was performed based on the predicted value P. The differentiation group was hepatocellular carcinoma or early hepatocellular carcinoma, and the control group was cirrhosis. The factors used were PIVKA-II or its logarithm, AFP or its logarithm, age, the logarithm of the real value of total ferritin (LogFER), and the logarithm of the real value of glycated ferritin (LogGF). The results of the ROC analysis are shown in Figures 2A and 2B. The multiple logistic regression analysis used a formula similar to the above formula (III), and the combination of factors shown in Figures 2A and 2B was used as independent variables. In Figure 2A, when a multiple logistic regression analysis was performed by combining LogPIVKA-II, LogAFP, Age, LogGF, and LogFER as independent factors, the intercept (β of order 0), regression coefficient (β of orders 1 to 5), standard error of the β coefficient (SEβ), z score (z) indicating the significance of the explanatory variable, significance probability (P) of the explanatory variable, and e β The diagnostic ability (odds ratio) of the explanatory variables and the confidence interval (95% Cl lower limit, 95% Cl upper limit) of the odds ratio are shown in Figure 2C. In Figure 2B, when a multiple logistic regression analysis was performed by combining LogPIVKA-II, LogAFP, Age, LogGF, and LogFER as independent factors, the intercept (β of order 0), regression coefficient (β of orders 1 to 5), standard error of the β coefficient (SEβ), z score (z) indicating the significance of the explanatory variables, significance probability (P) of the explanatory variables, and e β The diagnostic ability (odds ratio) of the explanatory variables and the confidence interval (95% Cl lower limit, 95% Cl upper limit) of the odds ratio are shown in Figure 2D. Note that the regression coefficient (β) and other factors in the formulas used for multiple logistic regression analysis using combinations of other factors in Figures 2A and 2B are omitted.

[0060] 2A and 2B, it was confirmed that the AUC value increased when age was added as a factor, and that the AUC value increased further when the logarithm of the real values ​​of total ferritin and glycated ferritin were added to age. It was also confirmed that it is possible to detect early-stage hepatocellular carcinoma with a high AUC value, not only for all types of hepatocellular carcinoma but also for early-stage hepatocellular carcinoma.

[0061] [Example 3] Analysis by Combination of Various Factors In Examples 1 and 2, since the AUC value was high even with the combination of conventional AFP and PIVKA-II as factors, various combinations were investigated. A prediction model formula was created by performing multiple logistic regression analysis, and then the predicted value P was calculated. Based on the predicted value P, the differentiation group was hepatocellular carcinoma (HCC), and the control group was liver cirrhosis (LC). The results of the AUC value were obtained by ROC analysis, which is shown in Figure 3A. A prediction value P was calculated by performing multiple logistic regression analysis, and based on the predicted value P, the differentiation group was hepatocellular carcinoma (HCC), and the control group was liver cirrhosis (LC). The results of the AUC value were obtained, and the sensitivity and specificity were calculated using the cutoff value based on the Youden index. The multiple logistic regression analysis used a formula similar to the above formula (III), and the combination of factors shown in Figures 3A and 3B was used as independent variables. In Figure 3B, when a multiple logistic regression analysis was performed combining LogPIVKA-II, LogAFP, LogGF, LogFER, and Age as independent factors (non-cancer vs. early cancer), the intercept (β of order 0), regression coefficient (β of orders 1 to 5), standard error of the β coefficient (SEβ), z score (z) indicating the significance of the explanatory variable, significance probability (P) of the explanatory variable, and e β The diagnostic ability (odds ratio) of the explanatory variables and the confidence interval (95% Cl lower limit, 95% Cl upper limit) of the odds ratio are shown in Figure 3C. In addition, regression coefficients (β) and the like in the equations when multiple logistic regression analysis was performed using combinations of other factors in Figures 3A and 3B are omitted.

[0062] 3A and 3B, it was confirmed that the AUC value increased when age was added as a factor, and that the AUC value increased further when the logarithm of the real values ​​of total ferritin and glycated ferritin were added to age. It was also confirmed that the AUC value was high not only for all cancers but also for early-stage cancers, and that it was possible to differentiate between cirrhosis and hepatocellular carcinoma.

[0063] Example 4 Comparison with Other Factors of Hepatocellular Carcinoma While the prior art was examined above using AFP and PIVKA-II, the complementarity of AFP, AFP-L3, and PIVKA-II in hepatocellular carcinoma has been recognized. This complementarity has been statistically analyzed, and the GALAD score has been proposed by an international collaborative study as a score for detecting hepatocellular carcinoma. The GALAD score is composed of five parameters: G: gender, A: age, L: AFP-L3, A: AFP, and D: DCP, and is calculated using the following formula (V):

[0064]

[0065] First, age, the logarithm of total ferritin (LogFER), the logarithm of glycated ferritin (LogGF), the logarithm of AFP (LogAFP), the logarithm of PIVKA-II (LogPIVKA-II) as independent variables were used to create the formula (III). The predicted value P calculated based on the formula (III) and the GALAD score (gender, age, AFP-L3, the logarithm of AFP, the logarithm of PIVKA-II) were analyzed. The results are shown in Figure 4. In formula (III), β0: -14.75, β1 (Age): 0.085, β2 (LogFER): 4.32, β3 (LogGF): -4.90, β4 (LogAFP): 2.49, β5 (LogPIVKA-II): 1.33.

[0066] From FIG. 4, it was confirmed that the AUC value of the present model was statistically significantly higher than that of the GALAD score (p<0.05).

[0067] Next, the distribution of the predicted value P and GALAD score is shown in Figure 5. In Figure 5, X represents non-cancer (control patients and cirrhotic patients), and 0-D represents cancer according to the BCLC staging system (Barcelona Clinical Liver Cancer Staging: Villanueva A et al. Hepatology. 2015; 61: 1945-1956.) described in Non-Patent Document 4, with 0 and A representing early hepatocellular carcinoma. Gray dots indicate positive cases. The cutoff values ​​(present model: 0.397, GALAD: -0.5382) were both calculated using the Youden index.

[0068] Furthermore, the results of comparing the present model with the GALAD score for each stage of the Barcelona Clinical Liver Cancer Staging Classification are shown in Figure 6A, and the results of comparing the sensitivity and specificity based on the present model and the GALAD score for early cancer (Early stage: stages 0 and A), advanced cancer (Advanced stage: stages B, C and D), and all cancers are shown in Figure 6B.

[0069] 6A and 6B, it was confirmed that the present model had higher sensitivity and specificity than the GALAD score, and had high sensitivity and specificity even for hepatocellular carcinoma at stage 0, the earliest stage.

Claims

1. A method for examining the possibility that a subject is suffering from hepatocellular carcinoma, comprising measuring (a) the amount of glycated ferritin and the amount of total ferritin, and (b) the ratio of the amount of glycated ferritin to the amount of total ferritin or the amount of non-glycated ferritin in a blood sample isolated from a subject, based on the measured values.

2. The method according to claim 1, further comprising performing a statistical analysis based on the measured values ​​of (a) or (b) and the PIVKA-II level and / or AFP level in the blood sample, and examining the possibility that the subject is suffering from hepatocellular carcinoma based on the value obtained by the statistical analysis.

3. The method according to claim 1 or 2, further comprising performing a statistical analysis based on the measured value of (a) or (b), the PIVKA-II level and / or the AFP level in the blood sample, and age, and examining the possibility that the subject is suffering from hepatocellular carcinoma based on the value obtained by the statistical analysis.

4. The method according to claim 2 or 3, characterized in that the statistical analysis uses logarithmically transformed values ​​of at least one of the measured values ​​(a) or (b) and the PIVKA-II level and / or AFP level in the blood sample.

5. A method according to any one of claims 1 to 4, wherein the subject has a history of cirrhosis and has not been diagnosed with hepatocellular carcinoma, and the possibility of the subject suffering from early-stage hepatocellular carcinoma is examined within six months of the subject being diagnosed with cirrhosis or within six months of the subject being diagnosed with cirrhosis but not progressing to hepatocellular carcinoma in order to monitor progression to early-stage hepatocellular carcinoma.

6. The method according to claim 5, wherein the early stage hepatocellular carcinoma is stage 0 or A according to the BCLC staging system.

7. A kit for testing the possibility that a subject is affected by hepatocellular carcinoma, the kit comprising a substance that specifically binds to glycated ferritin.

Images