Methods for aiding in the detection of nonalcoholic steatohepatitis

Abstract

Provided is a method for assisting in the detection of nonalcoholic steatohepatitis (NASH) with extremely low invasiveness compared to liver biopsy, which can be performed by a simple operation without relying on the skill of a technician. The present invention provides a method for assisting in the detection of NASH, which comprises: a) measuring the presence amount of LDL-TG contained in a test blood sample isolated from a living organism, b) measuring the presence amount of at least one component selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in the test blood sample, and c) determining the onset and / or the possibility of NASH by combining the use of the presence amount of the LDL-TG and the presence amount of the at least one component.

Description

TECHNICAL FIELD

[0001] The present application relates to a method for assisting in the detection of nonalcoholic steatohepatitis. BACKGROUND

[0002] Nonalcoholic fatty liver disease (hereinafter also referred to as "NAFLD") is a pathological state in which fatty liver is confirmed in histological diagnosis or imaging diagnosis, and viral hepatitis, autoimmune hepatitis, and alcoholic liver damage and the like are excluded. NAFLD has been increasing worldwide with the increase in the number of obese people. NAFLD is classified into nonalcoholic fatty liver (hereinafter also referred to as "NAFL") which is considered to hardly progress in the pathological state, and nonalcoholic steatohepatitis (hereinafter also referred to as "NASH") which is progressive and also a cause of the onset of cirrhosis or liver cancer. As characteristic pathological states of NASH, there are steatosis, ballooning degeneration, inflammation, fibrosis, and the like.

[0003] In order to discriminate various pathological states starting with steatosis of the liver, in order to distinguish between NAFL and NASH, definitive diagnosis based on liver biopsy is required. However, liver biopsy is an invasive examination, and is expensive, and thus the patient is consumed in various aspects. In addition, in liver biopsy, sampling errors and the like sometimes occur with a certain probability, and are dependent on the power of the technician who performs it, and the diagnosis of the collected specimen is also the same. Therefore, the patient needs to visit a specific institution in order to receive a certain level of medical care, and there is a problem that the burden on the patient is large.

[0004] On the other hand, as clinical examination items generally performed for liver diseases, there are aspartate aminotransferase (AST), alanine aminotransferase (ALT), AST / ALT ratio, and the like, or TNF-α, high-sensitivity CRP, ferritin, and the like as markers of inflammation, cytokines, cytokeratin 18 fragment as a marker of apoptosis, hyaluronic acid, collagen type IV 7S, and the like as markers of fibrosis (Patent Documents 1 and 2). However, there is no substance that can detect a biomarker that is considered to be suspected NASH in a plurality of pathological states related to NASH. That is, a biomarker that comprehensively reflects the pathological states related to NASH has not been found so far.

[0005] PRIOR ART DOCUMENTS

[0006] PATENT DOCUMENTS

[0007] Patent Document 1: Japanese Patent Application Laid-Open No. 2010-94131,

[0008] Patent Document 2: Japanese Patent Application Laid-Open No. 2018-80943. SUMMARY

[0009] PROBLEMS TO BE SOLVED BY THE INVENTION

[0010] An object of the present application is to provide a method for assisting in the detection of NASH with extremely low invasiveness compared to liver biopsy, which can be performed by a simple operation without depending on the skill of a technician.

[0011] MEANS FOR SOLVING THE PROBLEMS

[0012] The present inventors have conducted intensive studies, and as a result, have found that by using the presence amount of LDL-TG contained in a test blood sample isolated from a living organism in combination with the presence amount of at least one selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in the test blood sample, even if liver biopsy is not performed, NASH can be assisted in the detection by a simple operation, thereby completing the present application.

[0013] That is, the present application is as follows.

[0014] [1] A method for assisting in the detection of non-alcoholic steatohepatitis, comprising:

[0015] a) measuring the presence amount of LDL-TG contained in a test blood sample isolated from a living organism,

[0016] b) measuring the presence amount of at least one component selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in the test blood sample, and

[0017] c) determining the onset of non-alcoholic steatohepatitis and / or the possibility of the onset of non-alcoholic steatohepatitis by using the presence amount of the LDL-TG and the presence amount of the at least one component in combination.

[0018] [2] The method according to [1], wherein the LDL subfraction-C is small and dense (sd) LDL-C.

[0019] [3] The method according to [1], wherein the HDL subfraction-C is ApoE-carrying HDL-C and / or HDL2-C and / or HDL3-C.

[0020] [4] [1]~[3] The method wherein a model is prepared with the presence of LDL-TG and the presence of at least one component selected from LDL-C, LDL sub-C, HDL-C, HDL sub-C, ApoB, ApoE, total cholesterol, ALT and AST as variables, and if the score obtained by substituting the presence of LDL-TG and the presence of the at least one component in the tested blood sample into the model is higher than the score obtained for the blood sample of a patient with non-alcoholic fatty liver disease, the patient is determined to have a high probability of developing non-alcoholic fatty liver disease.

[0021] [5] A method for creating a model formula is a method for obtaining a model formula for an indicator used to assist in the detection of non-alcoholic steatohepatitis, wherein the presence of LDL-TG and at least one component selected from LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT and AST in blood samples of patients with non-alcoholic steatohepatitis and patients with non-alcoholic fatty liver are measured, and the measured presence amounts are used as variables to create a model formula.

[0022] [6] A assay kit for use in any of the methods for assisting in the detection of non-alcoholic steatohepatitis (NAH) as described in any one of [1] to [4], wherein the assay kit contains an assay reagent for LDL-TG and an assay reagent for at least one component selected from LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT and AST.

[0023] [7] [6] The kit described herein, wherein the LDL subfraction-C is small and dense (sd) LDL-C.

[0024] [8] [6] The kit described herein, wherein the HDL subfraction-C is HDL-C and / or HDL2-C and / or HDL3-C carrying ApoE.

[0025] The kit described in any one of [9] [1] to [4] is intended to be used as a kit for the auxiliary detection of non-alcoholic steatohepatitis.

[0026] The effects of the invention

[0027] According to the present invention, NASH can be detected safely and simply by minimizing the use of liver biopsies, which are burdensome or risky for patients, and without relying on the expertise of the technicians performing the biopsy. As a result, treatment for NASH can be initiated at an early stage, preventing or delaying progression to severe illness or death. Attached Figure Description

[0028] [ Figure 1 ] indicates a graph of the ROC curve when NASH and NAFL are discriminated by substituting the amounts of LDL-TG, sdLDL-C, HDL2-C, ApoE, and ALT in blood collected from a subject into the model formula in Example 20.

[0029] [ Figure 2 ] indicates a graph of the ROC curve when NASH and NAFL are discriminated by substituting the binary variables obtained by binarizing the amounts of LDL-TG, HDL2-C, ApoE, and total cholesterol in blood collected from a subject with the cutoff values of each item into the model formula in Example 36. DETAILED DESCRIPTION

[0030] The present application is a method for assisting in the detection of NASH by combining the presence amount of at least LDL-TG contained in a test blood sample isolated from a living organism with the presence amount of at least one component selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in the test blood sample. The method according to the present application can assist in the detection of NASH with high precision compared to the conventional method of using only the liver disease marker ALT or the method of using only LDL-TG.

[0031] The at least one component selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in the test blood sample can be selected as any one single component, or two or more components.

[0032] In the present application, LDL-TG refers to triglyceride (TG) in low density lipoprotein (hereinafter referred to as LDL). LDL-C refers to cholesterol (C) in low density lipoprotein (LDL). LDL subfraction is a fraction into which LDL is classified in accordance with size or density of particles, or constituent components, and examples thereof include sdLDL, large and light (lb) LDL, and the like, and LDL subfraction-C refers to cholesterol (C) in LDL subfraction. sdLDL-C refers to cholesterol (C) in LDL of small particles and high density (d = 1.044 to 1.063 g / mL). HDL-C refers to cholesterol (C) in high density lipoprotein (HDL). HDL subfraction is a fraction into which HDL is classified in accordance with size or density of particles, or constituent components, and examples thereof include HDL2, HDL3, ApoE-carrying HDL, and the like, and HDL subfraction-C refers to cholesterol (C) in HDL subfraction. HDL2-C refers to cholesterol (C) in HDL of large particles and low density (d = 1.063 to 1.125 g / mL), and HDL3-C refers to cholesterol (C) in HDL of high density (d = 1.125 to 1.210 g / mL). ApoE-carrying HDL-C refers to cholesterol (C) in high density lipoprotein containing a large amount of apolipoprotein E (hereinafter referred to as ApoE-carrying HDL). ApoB refers to apolipoprotein B. ApoE refers to apolipoprotein E.

[0033] In the present application, since a blood sample is used as a test sample, invasiveness is extremely low as compared with liver biopsy, and the state of the liver as a whole can be grasped as a whole without depending on a specific sample cut out. In addition, collection of the test blood sample is easy, and if a measurement kit for a corresponding automatic analysis device is used in measurement of the presence amount of LDL-TG or LDL-C, sdLDL-C, HDL-C, ApoE-carrying HDL-C, HDL2-C, HDL3-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in the blood sample, measurement can be performed by a simple operation to the same extent as routine diagnosis and treatment or health diagnosis and the like. As the blood sample in the present application, whole blood, plasma, and serum can be exemplified.

[0034] Hereinafter, in the present application, a specific method for assisting in detection of NASH will be described.

[0035] In the present application, the presence amount of LDL-TG contained in a test blood sample isolated from a living body is measured. In addition, the presence amount of at least one component selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in the test blood sample is measured. Then, the measured presence amount of LDL-TG and the presence amount of the at least one component are combined as an index for assisting in detection of NASH.

[0036] To obtain the result of assisting in the detection of NASH of the present application, a score is obtained by applying a statistical method known to those skilled in the art to the obtained data (the presence amount of each component in blood) and evaluating it. As a method of obtaining a score, there can be cited a method of selecting a combination of variables effective for assisting in the detection of NASH from a plurality of variables (the presence amount of each component in blood), making a model formula using the selected combination of variables, and applying the presence amount of each component in the blood sample to be tested.

[0037] As the value of the presence amount of each component in the blood sample to be tested used in making a model formula, there can be cited a continuous variable of a measured value itself, a continuous variable of a transformed value obtained by transforming a measured value such as logarithmic transformation, a binary variable obtained by dividing into 2 groups with an arbitrary value, a nominal variable or an ordinal variable obtained by dividing into 3 or more groups with a plurality of arbitrary values, any of which can be used and is not limited to any one. Specifically, in the case of using a binary variable, the measured value of each component is compared with an arbitrary value, and if it is small, the presence amount can be set to 0, and if it is large, the presence amount can be set to 1. Here, as an arbitrary value, for example, a cutoff value obtained by the Youden Index can be used by obtaining an ROC curve for discriminating NASH and NAFL using each component alone. The Youden Index is a method of setting a point where (sensitivity + specificity - 1) is the maximum value, that is, the point farthest from the slanting dotted line where AUC is 0.500 in the ROC curve, as a cutoff value.

[0038] As a selection method of a combination of variables effective for assisting in the detection of NASH, there can be cited a stepwise variable addition and subtraction method, a Losso method, an Elastic net method, and the like, but is not limited thereto.

[0039] After selecting an effective combination of variables, a model formula most suitable for assisting in the detection of NASH is made by a suitable method such as regression analysis. As a method of making a model formula, there can be cited a least square method, a linear discriminant analysis, a logistic regression, a maximum likelihood method, a Lasso regression, a Ridge regression, an Elastic net regression, and the like, but is not limited thereto.

[0040] In the present application, a model formula for obtaining a score as an index for assisting in the detection of NASH is represented by, for example, the following general formula (1).

[0041] Score = a + b x [amount of LDL-TG present] + c x [amount of LDL-C present] + d x [amount of sdLDL-C present] + e x [amount of HDL-C present] + f x [amount of ApoE-carrying HDL-C present] + g x [amount of HDL2-C present] + h x [amount of HDL3-C present] + i x [amount of ApoB present] + j x [amount of ApoE present] + k x [amount of total cholesterol present] + 1 x [amount of ALT present] + m x [amount of AST present] · · · (1)

[0042] In formula (1), a represents an intercept, b to m represent coefficients with respect to each of the measured values, and a to m can be numbers obtained by multiplying each of a to m by the same positive number.

[0043] In the case where a continuous variable using the measured value itself is used as the amount, in the above general formula (1), the intercept a is preferably -10 to 10, and further preferably -5 to 5. The coefficient b is preferably 0.1 to 1.5, and further preferably 0.2 to 1.0. The coefficient c is preferably -1.0 to 0, and further preferably -0.1 to 0. The coefficient d is preferably -1 to 0, and further preferably -0.5 to 0. The coefficient e is preferably -1.0 to 0, more preferably -0.2 to 0, and further preferably -0.1 to 0. The coefficient f is preferably -2 to 0, and further preferably -1.5 to 0. The coefficient g is preferably -1.0 to 0, and further preferably -0.3 to 0. The coefficient h is preferably -1.0 to 0, and further preferably -0.2 to 0. The coefficient i is preferably -1.0 to 0, and further preferably -0.2 to 0. The coefficient j is preferably 0 to 10, and further preferably 0 to 5. The coefficient k is preferably -1.0 to 0, and further preferably -0.1 to 0. The coefficient 1 is preferably -1.0 to 0, and further preferably -0.1 to 0. The coefficient m is preferably -1.0 to 0, and further preferably -0.2 to 0.

[0044] In the case where the binary variable obtained by dividing into two groups at an arbitrary value is used as the presence amount, for example, in the case where the arbitrary value is the value used in Comparative Examples 3 and 4, Examples 21 to 36, and Examples 53 to 57 described above, i.e., LDL-TG is 17.6 mg / dL, LDL-C is 116 mg / dL, sdLDL-C is 34.3 mg / dL, HDL-C is 47 mg / dL, HDL2-C is 43.8 mg / dL, HDL3-C is 22.7 mg / dL, ApoE-carrying HDL-C (AE-HDL) is 4.2 mg / dL, ApoB is 88.9 mg / dL, ApoE is 3.4 mg / dL, total cholesterol (TC) is 178 mg / dL, ALT is 32 units, and AST is 20 units, in the above general formula (1), the intercept a is preferably -100 to 100, and further preferably 0 to 40. The coefficient b is preferably 1 to 10, and further preferably 2 to 5. The coefficient c is preferably -5.0 to 0, and further preferably -3 to 0. The coefficient d is preferably -5 to 0, more preferably -4 to 0, and further preferably -3 to 0. The coefficient e is preferably -30 to 0, and further preferably -20 to 0. The coefficient f is preferably -30 to 0, and further preferably -20 to 0. The coefficient g is preferably -5 to 0, and further preferably -3 to 0. The coefficient h is preferably -30 to 0, and further preferably -20 to 0. The coefficient i is preferably -5 to 0, and further preferably -3 to 0. The coefficient j is preferably 0 to 10, and further preferably 0 to 3. The coefficient k is preferably -30 to 0, and further preferably -20 to 0. The coefficient 1 is preferably -5 to 0, and further preferably -2 to 0. The coefficient m is preferably -5 to 0, and further preferably -2 to 0.

[0045] The above general formula (1) is one example of the model formula in the present application, and a formula obtained by multiplying the intercept a and the coefficients b to m of the above general formula (1) by the same positive number can be used as the model formula, and a formula obtained by adding or subtracting an arbitrary number from the intercept a can be used as the model formula. In addition, the value that becomes the boundary when divided into the binary variable can be arbitrarily set. Therefore, the intercept or the coefficients of the model formula used in the present application are not limited to the above ranges.

[0046] As shown in the above general formula (1), the score is obtained by substituting the amount of presence of each item into the obtained model formula. The obtained score is compared with the reference value, and in the case where it is larger than the reference value, it is possible to indicate that the possibility of the onset of NASH is high, and in the case where it is smaller than the reference value, it is possible to indicate that the possibility of the onset of NASH is low. The reference value can be arbitrarily set according to the purpose, taking into consideration the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). For example, in the case where it is desired to reduce false negatives, a low reference value can be set, and in the case where it is desired to reduce false positives, a high reference value can be set. In the present application, the setting of the reference value is performed using the Youden index, which is considered to be well-balanced. Note that the reference value described in the present application is merely a simple example, and as described above, other values can be used as the reference value according to the purpose. In addition, the optimal reference value is different depending on the population of patients from which the data is collected. Furthermore, as described above, in the case where the coefficient and the intercept of the model formula are multiplied by the same positive number or the intercept of the model formula is added or subtracted by an arbitrary number, the optimal reference value also changes in conjunction.

[0047] In the present application, as an index indicating the detection accuracy of NASH, the area under the ROC curve (AUC) and the positive diagnosis rate (the proportion of patients in which NASH or NAFL can be correctly determined among the patients to be examined) are used. With respect to the AUC, an AUC of 0.80 or higher is high accuracy, 0.85 or higher is very high accuracy, and 0.90 or higher is extremely high accuracy. Therefore, in the present application, the AUC is also preferably 0.80 or higher, more preferably 0.85 or higher, and particularly preferably 0.90 or higher. With respect to the positive diagnosis rate, a positive diagnosis rate of 75% or higher is high accuracy, 85% or higher is very high accuracy, and 95% or higher is extremely high accuracy. Therefore, in the present application, the positive diagnosis rate is also preferably 75% or higher, more preferably 85% or higher, and particularly preferably 95% or higher.

[0048] Among the components in the blood sample to be examined, in the case where the combination of LDL-TG, sdLDL-C, HDL-C, ApoE, and ALT (Example 20 described below) is used in the example using the amount of presence of the measurement value itself, in the example using a binary variable obtained by dividing the measurement value into two groups as the amount of presence, and in the case where the combination of LDL-TG, HDL2-C, ApoE, and total cholesterol (Example 36 described below) is used, an AUC of 0.90 or higher and a positive diagnosis rate of 95% can be achieved, and thus NASH can be detected with extremely high accuracy.

[0049] Note that in the present application, even if the above formula is not used, in the case where the amount of LDL-TG contained in the blood sample under test is combined with the amount of at least one selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in the blood sample under test, the possibility of NASH onset can be determined by the combination of the amount of each component being above or below the cutoff value. The cutoff value can be obtained, for example, using the Youden index from the ROC curve when NASH and NAFL are discriminated using each component alone.

[0050] Specifically, in the case where the cutoff values are set to the values used in Examples 37 to 47, i.e., LDL-TG is 17.6 mg / dL, LDL-C is 116 mg / dL, sdLDL-C is 34.3 mg / dL, HDL-C is 47 mg / dL, HDL2-C is 43.8 mg / dL, HDL3-C is 22.7 mg / dL, ApoE-carrying HDL-C (AE-HDL) is 4.2 mg / dL, ApoB is 88.9 mg / dL, ApoE is 3.4 mg / dL, total cholesterol (TC) is 178 mg / dL, ALT is 32 units, and AST is 20 units, in the case where the amount of LDL-TG is above the cutoff value and the amount of any one of the test items selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, total cholesterol, ALT, and AST is below the cutoff value, or in the case where the amount of LDL-TG contained in the blood sample under test is above the cutoff value and the amount of ApoE is above the cutoff value, it can be determined that the possibility of NASH onset is high. However, the cutoff values can be arbitrarily set depending on the purpose of prioritizing sensitivity or specificity, and the combination is not limited thereto. Generally, however, the cutoff values can be set within ±50% of the above values (the values described in Table 5) used in Examples 37 to 47, and more preferably within ±30%.

[0051] In this way, by combining the amount of LDL-TG contained in the blood sample under test and the amount of at least one selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in the blood sample under test, NASH and NAFL, which have conventionally required liver biopsy, can be easily and accurately assisted by a simple operation.

[0052] In the present application, there is also provided a method for detecting LDL-TG in a blood sample from a person suspected of having non-alcoholic steatohepatitis or suffering from non-alcoholic steatohepatitis, in comparison with at least one component selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST.

[0053] That is, there is also provided a method for detecting LDL-TG in a blood sample from a person suspected of having non-alcoholic steatohepatitis or suffering from non-alcoholic steatohepatitis, in comparison with at least one component selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST, which comprises:

[0054] a step of obtaining a blood sample from a person, and

[0055] a step of measuring the amount of LDL-TG present in the blood sample and the amount of at least one component selected from the group consisting of LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST;

[0056] wherein a model equation is prepared using the amount of LDL-TG present and the amount of the at least one component as variables, and the score obtained by substituting the amount of LDL-TG present and the amount of the at least one component contained in the blood sample into the model equation is higher than the score obtained from a blood sample from a patient suffering from non-alcoholic fatty liver.

[0057] In addition, in the case of detecting non-alcoholic steatohepatitis with the aid of the method of the present application described above, non-alcoholic steatohepatitis can be treated by administering an effective amount of a non-alcoholic steatohepatitis treatment drug to a patient in whom non-alcoholic steatohepatitis has been detected. As the treatment drug, pioglitazone can be cited in the case of diabetes complications, and vitamin E or the like can be cited in the case of non-diabetes complications.

[0058] In the present application, as the method for measuring the amounts of LDL-C, sdLDL-C, HDL2-C, HDL3-C, HDL-C, ApoE-carrying HDL-C, ApoB, ApoE, total cholesterol, ALT, and AST contained in a blood sample, the methods known in the art can be used. For example, as the method for measuring each lipid item, the following methods can be mentioned: a method in which the target lipoprotein is separated by a fractionation operation such as ultracentrifugation, electrophoresis, or high-performance liquid chromatography, and then the triglyceride or cholesterol is quantified by a quantification operation; or a method in which the triglyceride or cholesterol in all lipoproteins except for the target lipoprotein is removed in the first step, and then the triglyceride or cholesterol in the target lipoprotein is measured in the second step. As the method for measuring ApoB or ApoE, the ELISA method using an antibody specific thereto or the immunoturbidimetry method can be mentioned. As the method for measuring ALT, the LDH-UV method (JSCC Standard Method) can be mentioned, and as the method for measuring AST, the MDH-UV method (JSCC Standard Method) can be mentioned.

[0059] Specifically, as the method for measuring the amount of LDL-TG, the method described in WO2013 / 157642 can be used, as the method for measuring the amount of LDL-C, the method described in WO98 / 47005 can be used, as the method for measuring the amount of sdLDL-C, the method described in WO08 / 105486 can be used, as the method for measuring the amount of HDL-C, the method described in WO98 / 26090 can be used, as the method for measuring the amounts of HDL2-C and HDL3-C, the method described in Ito Y et al. (2014) “Development of a homogeneous assay for measurement of high-density lipoprotein-subclass cholesterol,” Clinica Chimica Acta, 427: 86-93. can be used, and as the method for measuring the amount of ApoE-carrying HDL-C, the method described in Japanese Patent Application Publication No. 2014-030393 can be used. As long as it is a method for measuring the amount of each item, the method is not limited to the above-described methods.

[0060] Hereinafter, the present application will be specifically described based on Examples, but the present application is not limited to the following Examples. Example

[0061] For the LDL-TG amount, LDL-C amount, sdLDL-C amount, HDL-C amount, HDL2-C amount, HDL3-C amount, ApoE-carrying HDL-C amount, ApoB amount, ApoE amount, total cholesterol amount, ALT amount, and AST amount in blood collected from a total of 42 people consisting of 8 NAFL patients and 34 NASH patients, measurement was performed using an automatic analyzer. In the measurement of the LDL-C amount, HDL-C amount, ApoB amount, ApoE amount, total cholesterol amount, ALT amount, and AST amount, each reagent for measurement for an automatic analyzer commonly used in a clinical examination site was used. In the measurement of the LDL-TG amount, LDL-TG-EX “SEIKEN” (manufactured by DENKA SEIKEN Co., Ltd.) was used, in the measurement of the sdLDL-C amount, sdLDL-EX “SEIKEN” (manufactured by DENKA SEIKEN Co., Ltd.) was used, in the measurement of the HDL2-C amount and HDL3-C amount, the method of Ito Y et al. (2014) “Development of a homogeneous assay for measurement of high-density lipoprotein-subclass cholesterol,” Clinica Chimica Acta, 427: 86-93. was used, and in the measurement of the ApoE-carrying HDL-C amount, the method of Japanese Patent Application Publication No. 2014-030393 was used.

[0062] (Comparative Examples 1 and 2, and Examples 1 to 20)

[0063] In Comparative Examples 1 and 2, and Examples 1 to 20, the items shown in Table 1 were used, and as the amount of presence thereof, the measurement value itself was used as a continuous variable.

[0064] [Table 1]

[0065]

[0066] The model formula obtained by logistic regression for each combination is shown in Table 2. In the item name shown in the model formula, a score value is obtained by substituting the measurement value of each item.

[0067] [Table 2]

[0068]

[0069] The area under the ROC curve (AUC) when NASH was positive, the criterion value set by the Youden index, the sensitivity, specificity, PPV (positive predictive value), NPV (negative predictive value), and positive diagnosis rate when the criterion value and score value were compared are shown in Table 3-1 and Table 3-2. In addition, as a representative thereof, the ROC curve of Example 20 is shown in Figure 1

[0070] [Table 3-1]

[0071]

[0072] [Table 3-2]

[0073]

[0074] As shown in Table 3-1 and Table 3-2, the AUC of Examples 1 to 20 were all larger, and the positive diagnosis rate was all higher, compared with Comparative Example 1 using liver disease marker ALT alone or Comparative Example 2 using LDL-TG alone, thereby indicating that NASH or NAFL can be predicted with a higher probability.

[0075] (Comparative Examples 3 and 4, Examples 21 to 36)

[0076] In Comparative Examples 3 and 4, Examples 21 to 36, the items shown in Table 4 were used, and as the amount of presence thereof, a binary variable obtained by dividing into two groups at the cutoff value obtained by the Youden index when the ROC curve was obtained from each item alone was used.

[0077] [Table 4]

[0078]

[0079] The cutoff value of each item used in Comparative Examples 3 and 4, Examples 21 to 36 is shown in Table 5.

[0080] [Table 5]

[0081]

[0082] The model formula obtained by logistic regression of each combination is shown in Table 6. In the item name shown in the model formula, 1 is substituted when the measured value of each item is larger than the cutoff value, and 0 is substituted when it is smaller, thereby obtaining the score value.

[0083] [Table 6]

[0084]

[0085] ​The area under the ROC curve (AUC) when NASH is positive, the criterion value set by the Youden index, the sensitivity, specificity, PPV (positive predictive value), NPV (negative predictive value), and positive diagnosis rate when the criterion value and score value are compared are shown in Table 7-1 and Table 7-2. In addition, as a representative thereof, the ROC curve of Example 36 is shown in Figure 2

[0086] [Table 7-1]

[0087]

[0088] [Table 7-2]

[0089]

[0090]

[0091] As shown in Table 7-1 and Table 7-2, the AUC of Examples 21 to 36 are all larger, and the positive diagnosis rate is all higher, compared with Comparative Example 3 in which the liver disease marker ALT alone was used or Comparative Example 4 in which LDL-TG alone was used, thereby indicating that NASH or NAFL can be predicted with a higher probability.

[0092] (Examples 37 to 47)

[0093] In Examples 37 to 47, using the items shown in Table 8, evaluation was performed using a combination of the presence amount of each component contained in the blood sample to be examined being above or below the cutoff value, without using a model formula. The cutoff value of each item used the value shown in Table 5.

[0094] [Table 8]

[0095]

[0096] The combination of the measured values of each item evaluated in Examples 37 to 47, the corresponding number of persons, and the proportion of NASH patients in the corresponding persons are shown in Table 9.

[0097] [Table 9]

[0098]

[0099] As shown in Table 9, the proportion of NASH patients in the corresponding persons of the combination of measured values of Examples 37 to 47 is high, thereby indicating that the possibility of NASH onset can be determined by the combination of measured values of these items.

[0100] (Examples 48 to 52)

[0101] ​In Examples 48 to 52, the items shown in Table 10 were used as the presence amount, and the measured value itself was used as a continuous variable.

[0102] [Table 10]

[0103]

[0104] The model formula obtained by logistic regression in each combination is shown in Table 11. In the item name shown in the model formula, a score value is obtained by substituting the measured value of each item.

[0105] [Table 11]

[0106]

[0107] The area under the ROC curve (AUC) when NASH is positive, the set reference value, the sensitivity, specificity, PPV (positive predictive value), NPV (negative predictive value), and positive diagnosis rate when the reference value and the score value are compared to make a determination are shown in Table 12.

[0108] [Table 12]

[0109]

[0110] As shown in Table 12, the AUC of Examples 48 to 52 were all larger, and the positive diagnosis rate was all higher, compared to Comparative Example 1 in which the conventional liver disease marker ALT alone was used or Comparative Example 2 in which LDL-TG alone was used, which are shown in Table 3-1, thereby indicating that NASH or NAFL can be predicted with a higher probability.

[0111] (Examples 53 to 57)

[0112] In Examples 53 to 57, the items shown in Table 13 were used as the presence amount, and a binary variable obtained by dividing each item into two groups using the cutoff value obtained by the Youden index when the ROC curve was obtained for each item alone was used. The cutoff value of each item used the value shown in Table 5.

[0113] [Table 13]

[0114]

[0115] The model formula obtained by logistic regression in each combination is shown in Table 14. In the item name shown in the model formula, 1 was substituted when the measured value of each item was larger than the cutoff value, and 0 was substituted when it was smaller, thereby obtaining a score value.

[0116] [Table 14]

[0117]

[0118] The area under the ROC curve (AUC) when NASH is positive, the set reference value, the sensitivity, specificity, PPV (positive predictive value), NPV (negative predictive value), and positive diagnosis rate when the reference value and score value are compared are shown in Table 15.

[0119] [Table 15]

[0120]

[0121] As shown in Table 15, the AUC of Examples 53 to 57 are all larger, and the positive diagnosis rate is all higher, compared to Comparative Example 3 using only the liver disease marker ALT or Comparative Example 4 using only LDL-TG shown in Table 7-1, thereby indicating that NASH or NAFL can be predicted with a higher probability.

Claims

1. The use of reagents for the determination of LDL-TG and reagents for the determination of at least one component selected from LDL-C, LDL subfraction-C, HDL-C, HDL-C carrying ApoE, HDL2-C, HDL3-C, ApoB, ApoE, total cholesterol, ALT, and AST in the manufacture of a kit for the auxiliary detection of non-alcoholic steatohepatitis, wherein: a) The determination of LDL-TG uses reagents to determine the amount of LDL-TG present in the tested blood sample isolated from a living organism. b) The reagent for determining at least one component selected from LDL-C, LDL subfraction-C, HDL-C, HDL-C carrying ApoE, HDL2-C, HDL3-C, ApoB, ApoE, total cholesterol, ALT, and AST in the tested blood sample is used to determine the presence of at least one component selected from LDL-C, LDL subfraction-C, HDL-C, HDL-C, HDL-C carrying ApoE, HDL2-C, HDL3-C, ApoB, ApoE, total cholesterol, ALT, and AST. This excludes determining the onset of non-alcoholic steatohepatitis and / or the likelihood of developing non-alcoholic steatohepatitis by combining the presence of LDL-TG and the presence of LDL-C, or by combining the presence of LDL-TG and the presence of HDL-C carrying ApoE.

2. The use as described in claim 1, wherein, The LDL sub-C is a small and dense (sd) LDL-C.

3. The use as described in claim 1 or 2, wherein, A model is constructed using the presence of LDL-TG and the presence of at least one component selected from LDL-C, LDL subfraction-C, HDL-C, HDL-C carrying ApoE, HDL2-C, HDL3-C, ApoB, ApoE, total cholesterol, ALT, and AST as variables. If the score obtained by substituting the presence of LDL-TG and the presence of the at least one component in the tested blood sample into the model is higher than the score obtained from the blood sample of a patient with non-alcoholic fatty liver disease, the patient is considered to have a higher probability of developing non-alcoholic fatty liver disease.

4. A method for creating models, which is a method for obtaining models of indicators used to assist in the detection of non-alcoholic steatohepatitis, wherein... The presence of LDL-TG and at least one component selected from LDL-C, LDL subfraction-C, HDL-C, HDL subfraction-C, ApoB, ApoE, total cholesterol, ALT, and AST in blood samples from patients with non-alcoholic steatohepatitis and patients with non-alcoholic fatty liver disease was determined, and the determined presence was used as a variable to create a model.

5. A assay kit, which is an assay kit for use according to any one of claims 1 to 3, wherein, The reagents for the determination of LDL-TG and the reagents for the determination of at least one component selected from HDL2-C, HDL3-C, ApoE, ALT and AST.

Images