Method for identifying non-alcoholic fatty liver disease, and biomarker
The method of detecting specific metabolites in body fluids using mass spectrometry and a disease discriminant formula effectively addresses the limitations of current NAFLD diagnosis, achieving high accuracy and reliability in identifying NAFLD.
Patent Information
- Application Number
- PCT/JP2024/017844
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-17
- Filing Date
- 2024-05-14
- Publication Date
- 2024-11-21
AI Technical Summary
Current non-invasive methods for diagnosing non-alcoholic fatty liver disease (NAFLD) have limitations in reliability and accuracy, particularly due to small sample sizes and variability in image diagnostic results, necessitating a more effective method for identifying NAFLD with improved identification performance.
A method involving the acquisition and detection of specific metabolites in body fluids, specifically alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, and uric acid, using mass spectrometry techniques like GC/MS, with a disease discriminant formula to determine NAFLD presence based on detected compound amounts.
This approach provides excellent identification performance for NAFLD, as demonstrated by high AUC values in ROC curve analysis, indicating accurate and reliable diagnosis with a large sample dataset.
Smart Images

Figure JP2024017844_21112024_PF_FP_ABST
Abstract
Description
Methods for identifying nonalcoholic fatty liver disease and biomarkers
[0001] The present invention relates to a method for identifying non-alcoholic fatty liver disease and a biomarker for use therein.
[0002] Nonalcoholic fatty liver disease (NAFLD) is a general term for fatty liver not caused by alcohol or viruses. NAFLD includes simple fatty liver, in which fat is simply deposited in liver cells, and nonalcoholic steatohepatitis (NASH), in which liver cells are destroyed and no longer function, leading to inflammation and fibrosis from fatty liver and the risk of progressing to cirrhosis or liver cancer. Because simple fatty liver can sometimes lead to nonalcoholic steatohepatitis, diagnosing NAFLD, including NASH and simple fatty liver, is important.
[0003] Liver biopsy has traditionally been the standard diagnostic method. However, because liver biopsy is an invasive method in which a needle is inserted into the liver to collect tissues and cells, alternative non-invasive methods are desired. Non-invasive methods include abdominal ultrasound, CT, and MRI. However, diagnostic imaging has limitations due to the variability of results depending on the subject's physique, the skill and experience of the operator and radiologist, and the size, performance, and cost of each diagnostic imaging device. It has been pointed out that diagnostic imaging may not be suitable for screening large populations.
[0004] To solve these problems, methods for diagnosing NAFLD have been proposed, which involve collecting a patient's body fluid and measuring metabolites therein. For example, Patent Document 1 discloses a method for determining the amount of a specific lipid metabolite from a subject's body fluid and correlating it with the presence of NAFLD. Patent Document 2 discloses a method for determining the amount of acetaminophen-glucuronide in a plasma sample from a subject administered acetaminophen and comparing that amount with a control. Patent Document 3 discloses a method for assessing liver disease by measuring the concentration of soluble LR11 in a blood-derived sample. Patent Document 4 discloses a method using full-length high-molecular-weight kininogen and / or a partial peptide derived from high-molecular-weight kininogen as a biomarker. However, the methods disclosed in Patent Documents 1 to 4 lack sufficient reliability because the number of subject samples is limited, at several tens or less.
[0005] Therefore, Patent Document 5 proposes a more reliable method for diagnosing NAFLD based on data results from a larger number of samples.
[0006] Japanese Patent Publication No. 2010-500566 Japanese Patent Publication No. 2012-504629 Japanese Patent Application Laid-Open No. 2014-167446 WO2009 / 051259 WO2020 / 066162
[0007] However, in the method of Patent Document 5, the AUC (area under the curve) of the receiver operating characteristic (ROC) curve is approximately 0.750 to 0.650, so further improvement of the AUC is desired. In other words, a method for identifying NAFLD with better identification performance is desired.
[0008] An object of the present invention is to provide a method for identifying NAFLD with excellent identification performance and a biomarker to be used therefor.
[0009] The method for identifying non-alcoholic fatty liver disease according to a first aspect of the present invention comprises an obtaining step of obtaining a sample derived from the body fluid of a subject, a detecting step of detecting eight types of compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, and uric acid in the sample, and an using step of using all of the detected amounts of the eight types obtained by the detecting step.
[0010] The identification method and biomarker according to the first aspect of the present invention are excellent in the ability to identify non-alcoholic fatty liver disease.
[0011] FIG. 1 is a graph showing the performance of the disease discriminant in Example 1 (13 types of specific compounds) evaluated by an ROC curve; in the right figure, the sample cohort is the training set, and in the right figure, the sample cohort is the test set. In FIG. 1, the vertical axis represents sensitivity, and the horizontal axis represents non-specificity (1-specificity). FIG. 2 is a graph showing the performance of the disease discriminant in this Example evaluated by an ROC curve for a cohort different from that in FIG. 1. FIG. 3 is a graph showing the performance of the disease discriminant in Example 2 (11 types of specific compounds) evaluated by an ROC curve; in the right figure, the sample cohort is the training set, and in the right figure, the sample cohort is the test set. FIG. 4 is a graph showing the performance of the disease discriminant in Example 3 (8 types of specific compounds) evaluated by an ROC curve; in the right figure, the sample cohort is the training set, and in the right figure, the sample cohort is the test set.
[0012] 1. First Embodiment The first embodiment of the present invention is a method for identifying nonalcoholic fatty liver disease (hereinafter sometimes abbreviated as "NAFLD"), which uses all eight compounds consisting of 2-alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, and uric acid as biomarkers for identifying NAFLD. Specifically, the identification method of the first embodiment includes an acquisition step, a detection step, and a use step. Each step is described in detail below.
[0013] (Obtaining Step) In this step, a sample derived from the body fluid of a subject is obtained.
[0014] Examples of body fluids include blood, urine, saliva, sweat, sputum, and feces. Examples of blood include whole blood, plasma, and serum. Preferably, blood is used, and more preferably serum. Blood is minimally invasive, is a target sample for screening in health checkups, and is easily available.
[0015] The sample is temporarily stored before the detection step. The storage method may be frozen, refrigerated, or stored at room temperature. When serum or plasma is used as the sample, it may be stored in the form of whole blood, or in the form of serum or plasma obtained by centrifuging whole blood.
[0016] (Detection step) In this step, at least eight compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, and uric acid are detected from the sample. Hereinafter, these eight compounds are collectively referred to as specific compounds.
[0017] In this embodiment, "detecting a compound" refers to directly or indirectly detecting a specific compound contained in a sample for quantification, and specifically includes not only detecting the specific compound itself, but also performing processing such as ionization, dissociation, or derivatization on the specific compound, and detecting the processed specific compound (ionized product, dissociated product, derivative, etc.).
[0018] The detection method may be any method capable of quantifying a specific compound contained in a sample, and examples thereof include gas chromatography (GC), liquid chromatography (LC), mass spectrometry (MS), gas chromatography / mass spectrometry (GC / MS), liquid chromatography / mass spectrometry (LC / MS), nuclear magnetic resonance spectroscopy, immunological detection methods, etc. From the viewpoint of enabling highly sensitive detection of blood samples, preferred are methods using mass spectrometry (MS, LC / MS, GC / MS), and more preferred are GC / MS.
[0019] When mass spectrometry is employed, examples of ionization methods include electron ionization (EI), electrospray ionization (ESI), and atmospheric pressure chemical ionization (APCI). Examples of mass separation methods include quadrupole, magnetic sector, time-of-flight, ion trap, and ion cyclotron resonance. Other examples include tandem mass separation methods such as triple quadrupole (Q-Q), tandem time-of-flight (TOF-TOF), quadrupole-time-of-flight (Q-TOF), quadrupole-ion trap (Q-IT), quadrupole-ion cyclotron resonance (Q-ICR), and ion trap-time-of-flight (IT-TOF).
[0020] Before carrying out the detection step, the sample may be subjected to pretreatment. Examples of pretreatment include treatments for the purpose of purification, derivatization, etc. For example, when detecting serum or the like by mass spectrometry, it is preferable to carry out deproteinization treatment, derivatization, etc.
[0021] The deproteinization treatment is carried out, for example, by mixing an extraction solvent such as chloroform solvent, water, or the like with the sample, centrifuging the mixture, and removing the precipitate from the mixture after centrifugation.
[0022] Examples of derivatization include silylation such as trimethylsilylation, dimethylsilylation, and halomethylsilylation; esterification such as methyl esterification; acylation; and alkylation. From the viewpoints of the large number of target compounds that can be derivatized, the ability to react with many compounds having active hydrogen, and the ability to give peaks that facilitate analysis in mass spectra, silylation is preferred, and trimethylsilylation is more preferred. That is, it is preferred to use the detected amount of a derivative of a specific compound, particularly a trimethylsilylated derivative of a specific compound.
[0023] When trimethylsilylation is employed, for example, a methoxyamine solution is added to a freeze-dried sample to convert it into an oxime, followed by the addition of a trimethylsilylating agent such as N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) or N-methyl-N-trimethylsilyltrifluoroacetamide (MSTFA).
[0024] Examples of trimethylsilylated derivatives of specific compounds include bis(trimethylsilyl)alanine (Alanine-2TMS), bis(trimethylsilyl)isoleucine (Isoleucine-2TMS), tris(trimethylsilyl)glycine (Glycine-3TMS), tris(trimethylsilyl)serine (Serine-3TMS), tris(trimethylsilyl)glutamic acid (Glutamic acid-3TMS), tris(trimethylsilyl)asparagine (Asparagine-3TMS), tris(trimethylsilyl)tyrosine (Tyrosine-3TMS), and tetraxtris(trimethylsilyl)uric acid (Uric acid-4TMS). When a specific compound is derivatized, multiple derivatives (e.g., Alanine-2TMS and Alanine-3TMS) may be obtained for a single specific compound (e.g., Alanine). In this case, only one of the multiple derivatives may be selected arbitrarily, or multiple derivatives may be selected. When multiple derivatives are selected, a predetermined coefficient (described below) is set for each of the multiple derivatives in the use step described below. When setting the coefficient, the detected amounts of multiple derivatizations may be added together. Then, the detected amounts of the multiple derivatives and the corresponding multiple predetermined coefficients are substituted into the disease discriminant. Even when multiple derivatives are used from one type of specific compound, the number of selected specific compounds is treated as one type.
[0025] During pretreatment, it is preferable to mix a stable isotope and / or an internal standard of a specific compound with the sample. By mixing a stable isotope, the absolute amount of a specific compound contained in a body fluid such as blood can be measured. Furthermore, by mixing an internal standard, the accuracy of quantification by mass spectrometry can be improved. Examples of stable isotopes include compounds in which one or more carbon, hydrogen, nitrogen, or oxygen atoms in the molecule are substituted with stable carbon isotopes (C, C), stable hydrogen isotopes (H, H, D), stable nitrogen isotopes (N, N), or stable oxygen isotopes (O, O, O). The internal standard may be any substance that is not naturally contained in body fluids, such as 2-isopropylmalic acid, ribitol, or tropic acid.
[0026] The pretreated sample is detected by the above-mentioned detection method. The detection method conditions may be appropriately carried out according to the conditions recommended for the device used in the detection method (e.g., GC / MS device, LC / MS device).
[0027] This allows the detection amount of each specific compound contained in the sample to be obtained. The detection amount is determined appropriately depending on the detection method and detection device, and is output, for example, as concentration; peak area / peak intensity of a mass spectrum or mass chromatogram; light intensity, etc. In this case, the detection amount may be converted. For example, when the detection amount of a specific compound is output as a peak area of a mass spectrum, the peak area may be converted to concentration. That is, the "detection amount" in the first embodiment includes not only the detection amount directly output from the detection device, but also the amount obtained by converting the detection amount into another parameter. Examples of conversion include known methods such as the internal standard method and the absolute calibration curve method, and the conversion may be performed using any means, such as an analysis function attached to the detection device, a separate analysis device, or manually. In the first embodiment, it is preferable to use concentration as the detection amount, and it is particularly preferable to use the concentration (μg) of the specific compound present in 1 mL of serum.
[0028] (Using step) In this step, at least all of the eight types of detection amounts obtained in the detecting step are used. Specifically, all of the eight types of detection amounts of the specific compound are used, and based on the results of the use, it is determined (determined) whether or not the subject has NAFLD. Through this use, the presence or absence of NAFLD in the subject is identified.
[0029] For example, the amounts of all eight specific compounds detected in a sample from a subject are used in a predetermined formula, specifically, substituted into a disease discriminant, and the result is used to determine whether the subject is affected by NAFLD.
[0030] The disease discriminant includes the sum of values obtained by multiplying each of the detected amounts of eight specific compounds by a predetermined coefficient for each specific compound (specifically, the second term on the right-hand side of formula (1) described below). Specifically, each detected amount of a compound (eight types) is multiplied by a specific coefficient predetermined for each compound. As a result, eight values are obtained for each compound, and these values are added together to derive a sum, which is then substituted into the disease discriminant.
[0031] Examples of such disease discriminant formulas include the following formulas (1) and (2).
[0032]
[0033] The result of calculation using the prevalence discriminant, i.e., the prevalence probability, is used as a criterion for determining whether or not a subject has NAFLD. For example, the prevalence probability is expressed as a probability of having NAFLD in the range of 0 to 100%.
[0034] The constant K and the predetermined coefficients (eight types of coefficients) of each compound can be obtained, for example, by acquiring data obtained in advance from measurements of a large number of samples from healthy subjects and NAFLD patients, and calculating and verifying this data using a statistical machine learning method (such as LASSO). In the first embodiment, the inventors have already collected data from a large number of healthy subjects and NAFLD patients, and the results of calculations are listed in Tables 2 to 4 in the Examples section described below. These constants K and the coefficients of each compound can be used.
[0035] In addition to using the probability of disease calculated by the above formulas (1) and (2), for example, (i) a value obtained by comparing the amount of a specific compound detected in a sample from a subject (the subject's value) with a value obtained by comparing the amount of a specific compound detected in a sample from a healthy subject or a NAFLD patient (the healthy subject's or NAFLD patient's value), and (ii) determining whether or not the subject has NAFLD based on whether the subject's value is higher or lower than the healthy subject's or NAFLD patient's value. Specifically, (i) the value obtained by substituting the amount detected in the subject into the second term on the right side of the above formula (1) is compared with the value obtained by substituting the amount detected in the healthy subject or the NAFLD patient into the second term on the right side of the above formula (1). Then, (ii) if the subject's value is equal to or higher (or lower) than the value of a NAFLD patient (threshold for NAFLD disease), the subject is determined to have NAFLD, and if the subject's value is equal to or lower (or higher) than the value of a healthy person (normal threshold), the subject is determined not to have NAFLD. Part or all of this using step may be processed by a computer.
[0036] The identification method of the first embodiment can accurately determine whether a subject has NAFLD, with excellent identification performance. As a result, the presence or absence of NAFLD in a subject can be accurately identified. Therefore, for example, if a medical professional presents the results of this identification method to a physician, the physician can use the results to make a definitive diagnosis of whether the subject has NAFLD and use them to guide subsequent treatment. Specifically, a treatment step can be performed after this identification method, particularly after the use step. Examples of treatment steps include well-known methods for treating NAFLD, such as dietary therapy, exercise therapy, drug therapy, and surgical therapy. Dietary therapy, exercise therapy, and surgical therapy, which are effective for weight loss, are effective, and medications used in drug therapy include antioxidants and antidiabetic drugs. These therapies are applied in appropriate combinations depending on the subject's condition.
[0037] 2. Other Embodiments In the first embodiment, the detection results of eight specific compounds are used, but the present invention is sufficient as long as it includes the detection amounts of the eight specific compounds, and for example, the detection amounts of other compounds may also be added. In other words, the detection amounts of nine or more compounds may be used. Examples are shown below.
[0038] 2-1. 13 Types of Specific Compounds In this embodiment, in addition to eight types of compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, and uric acid, 2-hydroxybutyric acid, proline, decanoic acid, α-ketoglutaric acid, and glucose are used as biomarkers for identifying NAFLD. That is, 13 types of compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, uric acid, 2-hydroxybutyric acid, proline, decanoic acid, α-ketoglutaric acid, and glucose are used as biomarkers for identifying NAFLD. These 13 types of compounds are collectively referred to as specific compounds.
[0039] The present embodiment is similar to the above embodiment except that these 13 types of compounds are used as biomarkers. That is, the identification method of the present implementation steps includes an acquisition step, a detection step, and a use step, as described above in the first embodiment. This embodiment also achieves the same adoption effects as the first embodiment.
[0040] In the detection step, examples of the trimethylsilylated derivatives of the 13 specific compounds include, in addition to the trimethylsilylated derivatives of the above-mentioned eight specific compounds, bis(trimethylsilyl)2-hydroxybutyric acid (2TMS), bis(trimethylsilyl)proline (Proline-2TMS), trimethylsilyldecanoic acid (Decanoic acid-TMS), bis(trimethylsilyl)α-ketoglutaric acid-methyloxime-2TMS, and pentakis(trimethylsilyl)glucose-methyloxime-5TMS. In the use step, the amounts of at least 13 compounds are used. Specifically, the amounts (e.g., concentrations) of the at least 13 specific compounds (or their derivatives) are multiplied by a predetermined coefficient for each specific compound, and the sum is used in the disease discriminant.
[0041] 2-2. 11 Types of Specific Compounds In this embodiment, in addition to eight types of compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, and uric acid, lactic acid, succinic acid, and creatinine are used as biomarkers for identifying NAFLD. That is, 11 types of compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, uric acid, lactic acid, succinic acid, and creatinine are used as biomarkers for identifying NAFLD. These 11 types of compounds are collectively referred to as specific compounds.
[0042] The present embodiment is similar to the above embodiment except that these 11 types of compounds are used as biomarkers. That is, the identification method of the present implementation steps includes an acquisition step, a detection step, and a use step, as described above in the first embodiment. This embodiment also achieves the same adoption effects as the first embodiment.
[0043] In the detection step, examples of the trimethylsilylated derivatives of the 11 specific compounds include, in addition to the trimethylsilylated derivatives of the above-mentioned eight specific compounds, bis(trimethylsilyl)lactic acid (Lactic acid-2TMS), bis(trimethylsilyl)succinic acid (Succinic acid-2TMS), and tris(trimethylsilyl)creatinine (Creatinine-3TMS). In the use step, the amounts of at least 11 specific compounds detected are used. Specifically, the amounts of the at least 11 specific compounds detected are multiplied by a predetermined coefficient for each specific compound, and the sum is used in the disease discriminant formula.
[0044] 3. Aspects It will be appreciated by those skilled in the art that the exemplary embodiments described above are examples of the following aspects.
[0045] (Item 1) A method for identifying non-alcoholic fatty liver disease according to one embodiment may include an obtaining step of obtaining a sample derived from a subject's body fluid, a detecting step of detecting eight compounds in the sample, consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, and uric acid, and an using step of using all of the eight detected compounds obtained by the detecting step. This identification method can determine the presence or absence of NAFLD in a subject with excellent identification performance. Therefore, the presence or absence of NAFLD can be accurately identified.
[0046] (Item 2) In the identification method described in Item 1, in addition to the eight compounds, lactic acid, succinic acid, and creatinine may be detected, and all of the eleven detected amounts obtained in the detection step may be used in the using step. This identification method makes it possible to more reliably determine whether or not a person is affected by NAFLD.
[0047] (Item 3) In the identification method according to item 1, in addition to the eight compounds, 2-hydroxybutyric acid, proline, decanoic acid, α-ketoglutaric acid, and glucose may be detected, and the amounts of all 13 compounds detected by the detection step may be used in the using step. This identification method makes it possible to more reliably determine whether or not a subject is affected by NAFLD.
[0048] (4) In the identification method according to any one of items 1 to 3, the using step may include at least summing values obtained by multiplying the detected amounts of the eight types of compounds by predetermined coefficients for each of the eight types of compounds. This identification method makes it possible to more reliably determine the presence or absence of NAFLD.
[0049] (Item 5) In the identification method according to any one of Items 1 to 4, the detection step may include obtaining concentrations of the eight compounds or derivatives thereof. This identification method makes it possible to more reliably determine whether or not a person is affected by NAFLD.
[0050] (Item 6) In the identification method according to item 5, the derivative may be derivatized by silylation, acylation, esterification, or alkylation. This identification method allows for the detection of the derivative with high sensitivity, thereby enabling a more accurate determination of the presence or absence of NAFLD.
[0051] (Item 7) In the identification method according to any one of Items 1 to 6, the detection step may be performed by gas chromatography, liquid chromatography, mass spectrometry, gas chromatography / mass spectrometry, liquid chromatography / mass spectrometry, nuclear magnetic resonance spectroscopy, or an immunological detection method. This identification method uses a detection method with high sensitivity, making it possible to more accurately determine whether or not a patient has NAFLD.
[0052] (Item 8) In the identification method according to any one of Items 1 to 7, the body fluid may be blood. According to this identification method, measurement samples are easily available and the method is minimally invasive to the subject, thereby reducing the burden of diagnosis on the subject.
[0053] (Item 9) A biomarker for identifying non-alcoholic fatty liver disease according to one embodiment may contain eight compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, and uric acid. This biomarker can determine with excellent discrimination whether a subject has NAFLD.
[0054] (Item 10) The biomarker for identifying non-alcoholic fatty liver disease described in Item 9 may contain 11 compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, uric acid, lactic acid, succinic acid, and creatinine. This biomarker can determine the presence or absence of NAFLD in a subject with excellent discrimination ability.
[0055] (Item 11) The biomarker for identifying non-alcoholic fatty liver disease described in Item 9 may contain 13 compounds consisting of alanine, 2-hydroxybutyric acid, isoleucine, proline, glycine, serine, decanoic acid, α-ketoglutaric acid, glutamic acid, asparagine, glucose, tyrosine, and uric acid. This biomarker can determine with excellent discrimination performance whether a subject has NAFLD.
[0056] The present invention will now be described in detail with reference to examples, but the scope of the present invention is not limited to these examples.
[0057] <Example 1> In the following example, GC / MS was performed using serum obtained from patients diagnosed with NAFLD based on abdominal ultrasound findings (NAFLD group) and serum obtained from healthy individuals (normal group), and the analysis results of searching for components in the serum that have different concentrations between the NAFLD group and the normal group are described.
[0058] (Subjects) Serum samples were collected from 3,733 Japanese people who met the inclusion criteria and were randomly selected from among those who underwent comprehensive medical checkups and visited the hospital between October 2015 and September 2016. The inclusion criteria were that the subjects did not fall under any of the following (1) to (5).
[0059] (1) Those who have expressed their refusal to participate in the study. (2) Those with an alcohol intake of 30 g or more per day (equivalent to ethanol) for men and 20 g or more per day (equivalent to ethanol) for women. (3) Those with a history of currently receiving treatment for liver disease. (4) Those with a history of currently receiving treatment for at least one of the following: diabetes, dyslipidemia, or hyperuricemia. (5) Those with a history of currently receiving treatment for malignant disease.
[0060] All individuals who met the above inclusion criteria had abdominal ultrasound findings obtained during a medical checkup, and with the consent of the hospital's research ethics committee, their clinical information was compared with the results of serum analysis. Of the 3,733 individuals in the NAFLD and normal groups, 2,252 (60.3%) were women, and the median age was 51 years (interquartile range: 43-61 years).
[0061] (Pretreatment) After blood was collected from the subject, serum was extracted by centrifugation and frozen at -80°C within 6 hours of collection. The frozen serum was thawed by leaving it at 25°C for 5 minutes. The thawed serum was centrifuged, and the supernatant was collected into a tube containing a specified amount of 40 stable isotopes (Table 1 lists only the stable isotopes corresponding to 13 of the 40 metabolites being measured). The collected serum was mixed with 250 μL of a chloroform mixed solvent (methanol / water / chloroform = 2.5:1:1) and 6 μL of an internal standard solution (2-isopropylmalic acid; 0.1 mg / mL), shaken, and centrifuged. Next, 140 μL of water was added, shaken, and centrifuged to obtain a sample containing the internal standard and stable isotope. 180 μL of this sample was concentrated using a centrifugal evaporator at room temperature for 60 minutes, left to stand at −80° C. for 30 minutes to freeze, and then dried overnight using a freeze dryer.
[0062]
[0063] (Derivatization) The dried sample was mixed with 80 μL of a pyridine solution containing methoxyamine hydrochloride (methoxyamine hydrochloride; 20 mg / mL) and 40 μL of N-methyl-N-trimethylsilyltrifluoroacetamide, followed by shaking and centrifugation. 50 μL of the resulting supernatant was placed in the autosampler of the GC / MS.
[0064] (GC / MS Measurement) Gas chromatography mass spectrometry was carried out on the sample under the following conditions. <Gas chromatography conditions> System: GCMS-TQ8040 (Shimadzu Corporation) Column oven temperature: 80°C Injection temperature: 280°C Injection mode: Split introduction Carrier gas: Helium Carrier gas flow rate: 39 cm / sec Separation column: DB-5 (Agilent Technologies) Stationary phase: (5%-phenyl)-methylpolysiloxane (non-polar) Column length: 30.0 m, column inner diameter: 0.25 mm, column film thickness: 1.00 μm Column temperature program: 80°C (0 min) / 80°C (2.5 min) / 280°C (18.5 min) / 280°C (23.0 min) <Mass spectrometry conditions> System: GCMS-TQ8040 (Shimadzu Corporation) Ion source temperature: 200°C Interface temperature: 250°C Detector voltage: 0.2 kV Measurement mode: MRM mode Scanning range: m / z 82-500
[0065] (Calculation of detection amount) A mass spectrum at each retention time was output from the detection intensity obtained by scanning m / z at each retention time. Based on the mass spectrum data of a serum sample obtained in the past, metabolites (40 types) corresponding to each peak in the mass spectrum and the stable isotopes corresponding to those metabolites were identified. The detection intensity corresponding to each metabolite was standardized using the detection intensity of the stable isotope corresponding to each metabolite, and the concentration (unit: μg / mL) in the sample was calculated as the detection amount.
[0066] (Data Analysis) Based on clinical diagnostic information and the detected amounts obtained by GC / MS, combinations of metabolites that can distinguish between normal and diseased groups were investigated using LASSO, a machine learning method. Specifically, before running LASSO, subjects were divided into two groups, a training set and a test set, and a disease discriminant was developed using the training set and validated using the test set. In running LASSO, logistic regression with 10-fold cross-validation was applied.
[0067] As a result, 13 compounds were selected as discriminable metabolites, as shown in Table 2 below, and their regression coefficients were calculated to derive disease discriminants as shown in the following formulas (1) and (2). Regarding the disease discriminant, the judgment score (formula (1) below) is calculated by adding up a predetermined constant (first term on the right-hand side) and the sum of the detected amounts of each compound multiplied by each coefficient (second term on the right-hand side). Since this judgment score is expressed as logarithmic odds (expressed as ln{p / (1-p)} where p is probability), the disease probability (formula (2) below) is derived.
[0068]
[0069]
[0070] In Table 2, TMS represents a trimethylsilylated derivative. In Equation 1, the second term on the right side is the sum of 13 compounds, the constant K and the coefficient of each compound are as shown in Table 2, and the detection amount of each compound is the concentration (μg) of the compound present in 1 mL of serum.
[0071] The resulting probability of having NAFLD was used as a basis for determining whether each subject currently had NAFLD. Specifically, the probability of having NAFLD was expressed as 0 to 100%.
[0072] Receiver Operating Characteristic (ROC) analysis was performed on the disease discriminant, and the results are shown in Figure 1. For quantitative evaluation, the AUC was calculated, and the AUC was 0.86, a very good value, for both the training set and the test set. This shows that the discrimination method of the present example has excellent discrimination performance. In addition, since it is based on data from a large number of healthy subjects and NAFLD patients (3,733 individuals), it is highly reliable.
[0073] In addition, as a separate cohort, 899 sera samples were randomly selected from Japanese patients undergoing a comprehensive medical checkup at a different time point. These samples met the above inclusion criteria. ROC analysis was also performed on this separate cohort, and the results are shown in Figure 2. The AUC was 0.90.
[0074] Example 2: The values of 34 metabolites were subjected to the LASSO method in the same manner as in Example 1 (cohort of 3,733 individuals), except that six metabolites with unstable elements (leucine-TMS, decanoic acid-TMS, 1-hexadecanol-TMS, uridine-4TMS, 2-aminoadipic acid-3TMS, and histidine-3TMS) were excluded from the 40 metabolites to further enhance analytical stability. As a result, 11 compounds were selected as distinguishable metabolites, as shown in Table 3, and their regression coefficients were calculated.
[0075]
[0076] Using the above 11 types of metabolites, ROC analysis was performed in the same manner as in Example 1, and the results are shown in Figure 3. The AUC in this case was 0.85 for both.
[0077] Example 3 Eight specific compounds common to the 13 compounds in Example 1 and the 11 compounds in Example 2 were selected as distinguishable metabolites (see Table 4 below).
[0078]
[0079] Using the above eight compounds, ROC analysis was performed in the same manner as in Example 1, and the results are shown in Figure 4. The AUC was 0.85 for both compounds.
Claims
1. A method for identifying non-alcoholic fatty liver disease, comprising: an obtaining step of obtaining a sample derived from a subject's body fluid; a detecting step of detecting eight types of compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, and uric acid in the sample; and an using step of using all of the eight detected compounds obtained by the detecting step.
2. The method of claim 1, wherein in the detecting step, in addition to the eight compounds, lactate, succinate and creatinine are detected, and in the using step, all of the detected amounts of the eleven compounds obtained in the detecting step are used.
3. The method of claim 1, wherein in the detecting step, 2-hydroxybutyric acid, proline, decanoic acid, α-ketoglutaric acid, and glucose are detected in addition to the eight compounds, and in the using step, all of the detected amounts of the thirteen compounds obtained in the detecting step are used.
4. The identification method according to claim 1, wherein the using step includes at least adding up values obtained by multiplying the detected amounts of the eight types of compounds by predetermined coefficients for each of the eight types of compounds.
5. The method of claim 1, wherein the detection step obtains the concentrations of the eight types of compounds or their derivatives.
6. The method of claim 5, wherein the derivative is derivatized by silylation, acylation, esterification or alkylation.
7. The method of claim 1, wherein the detecting step is performed by gas chromatography, liquid chromatography, mass spectrometry, gas chromatography / mass spectrometry, liquid chromatography / mass spectrometry, nuclear magnetic resonance spectroscopy, or immunological detection.
8. The method of claim 1, wherein the body fluid is blood.
9. A biomarker for identifying non-alcoholic fatty liver disease, containing eight compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine and uric acid.
10. The biomarker of claim 9, comprising 11 compounds consisting of alanine, isoleucine, glycine, serine, glutamic acid, asparagine, tyrosine, uric acid, lactic acid, succinic acid, and creatinine.
11. The biomarker of claim 9, comprising 13 compounds consisting of alanine, 2-hydroxybutyrate, isoleucine, proline, glycine, serine, decanoic acid, α-ketoglutaric acid, glutamic acid, asparagine, glucose, tyrosine, and uric acid.
Citation Information
Patent Citations
Markers for non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatitis (NASH) and their usage.
JP2010500566A
Glucuronide-conjugated acetaminophen as a marker of liver damage
JP2012504629A
Evaluation method and diagnostic kit of liver disease
JP2014167446A
Biomarker for diagnosis of liver disease
WO2009051259A1
Method for evaluating NASH, device for evaluating NASH, program for evaluating NASH, system for evaluating NASH, information communication terminal device, and method for searching for substance used to prevent or improve nash
JP2013040923A
