WO2022014804A1 - Marker for predicting risk of hepatic fibrosis, and information providing method using same - Google Patents

Abstract

The present invention relates to a marker for predicting the risk of hepatic fibrosis, and an information providing method using same and, more specifically, to a marker for predicting the risk of hepatic fibrosis, and an information providing method using same which enable a prediction with high sensitivity and specificity by distinguishing the risk in accordance with hepatic fibrosis stages, by performing a regression analysis using a genetic marker combination of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 to derive a hepatic fibrosis prediction score and determine the risk of hepatic fibrosis.

Description

A marker for predicting the risk of liver fibrosis and a method for providing information using the same

The present invention relates to a genetic marker, a composition for diagnosing the advanced stage of liver fibrosis or for predicting the risk of liver fibrosis, and a method for providing information using the same.

Hepatic fibrosis refers to a state in which damaged liver tissue is not restored to normal hepatocytes but is transformed into fibrous tissue such as collagen. do. If these injuries are not cured in the early stage and become chronic, hepatic stellate cells (HSCs) become Activated and activated hepatic stellate cells are differentiated into proliferative myofibroblasts expressing α-SMA (alpha smooth muscle actin), and the differentiated myofibroblasts are various cells such as collagen and elastin. As extracellular matrix (ECM) is secreted, liver fibrosis is induced. In addition, according to a recent report, hepatic fibrosis is also induced through epithelial mesenchymal transition (EMT), a novel mechanism related to liver fibrosis.

Hepatic fibrosis is an unavoidable pathological result of the liver accompanying chronic liver damage, and a decrease in liver function inevitably appears in that the liver is replaced by a fibrous tissue that cannot perform a specific function of the liver, such as metabolism or bile secretion of biological substances. Hepatic fibrosis is known to be reversible, composed of thin fibrils, and not to form nodules, so it may be possible to return to normal when the cause of liver damage is eliminated. However, if the fibrosis process continues repeatedly, the exchange and binding between ECMs increase, leading to irreversible liver cirrhosis, in which thick fibrils and regenerated nodules are generated. Since the degree of liver fibrosis reflects the degree of chronic liver function impairment, liver fibrosis provides important information in predicting the severity and prognosis of chronic liver disease.

Determining the progression stage of liver fibrosis is very important in determining the treatment for liver fibrosis or observing the prognosis after treatment. Since fibrosis is not a change that appears evenly throughout the liver, a biopsy in which a biopsy tissue is obtained and morphologically observed has a risk that some tissues for which the biopsy is performed may not represent the overall situation of the liver. However, the important thing is that the accuracy can be improved only when a serum marker is developed using the markers identified in the liver biopsy tissue. Josep M. Llovet of the Barcelona Liver Cancer Research Council (BCLC), Spain, who is leading the academic community related to liver research, published a review paper on the recent guidelines recommending tissue biomarker research through liver cancer biopsy in clinical trial design among future treatment strategies for liver cancer. recommended. This guideline is ‘EASL-EORTC Clinical Practice Guidelines: Management of Hepatocellular Carcinoma’ (J Hepatol. 2012 Apr;56(4):908-43). Accordingly, there is a need for a diagnostic method capable of diagnosing the liver fibrosis stage, which can be used for the development of a simple and high-accuracy serum marker, and which has better accuracy than conventionally known methods.

An object of the present invention is to provide a marker for diagnosing the progress of liver fibrosis that can specifically discriminate, diagnose, or detect the progress of liver fibrosis.

Another object of the present invention is to provide an information providing method for predicting the risk of liver fibrosis by performing a statistical analysis through the combination of the above markers, and diagnosing the progression stage of liver fibrosis based on this.

In addition, another object of the present invention is to provide an information providing method for predicting the risk of liver fibrosis by measuring the expression level of the marker.

In addition, another object of the present invention is to provide a composition for predicting or diagnosing the risk of liver fibrosis that can measure the expression level of the marker.

According to one aspect of the present invention, there is provided a marker composition for predicting the risk of liver fibrosis, comprising at least one gene selected from the group consisting of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17.

According to another aspect of the present invention, determining a gene combination using at least two genes selected from the group consisting of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17; measuring the expression level for each gene included in the determined gene combination;

obtaining a regression coefficient value for the expression level for each gene through regression analysis; and obtaining a liver fibrosis prediction score using the regression coefficient value; further comprising, an information providing method for predicting liver fibrosis risk is provided.

In addition, the method further comprises the step of obtaining a reference value through the ROC curve for the gene combination and comparing the liver fibrosis prediction score and the reference value, but if the liver fibrosis prediction score is higher than the reference value, the risk of liver fibrosis is determined to be high. If the liver fibrosis prediction score is lower than the reference value, it can be determined that the risk of liver fibrosis is low.

In addition, the gene combination may include at least two types selected from the group consisting of JAK2, MAP3K14 and SOX17.

In addition, the gene combination may include CSF3R and MAP3K14.

In addition, the gene combination may include JAK2, MAPK8IP2 and SOX17.

In addition, the gene combination includes SOX17 and may further include at least three kinds of genes selected from the group consisting of CSF3R, JAK2, MAP3K14 and MAPK8IP2.

According to another aspect of the present invention, measuring the expression levels of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 genes; obtaining a regression coefficient value for the expression level for each gene through logistic regression analysis; and obtaining a liver fibrosis prediction score using the regression coefficient value; further comprising, an information providing method for predicting liver fibrosis risk is provided.

According to another aspect of the present invention, measuring the expression level of mRNA of at least two genes selected from the group consisting of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 or a protein encoded by the gene, An information provision method for predicting the risk of liver fibrosis is provided.

In addition, the method may further include comparing the measured expression level of mRNA or protein with the expression level of a normal group.

According to another aspect of the present invention, comprising measuring the expression level of mRNA of at least two genes selected from the group consisting of CSF3R, JAK2, MAP3K14 and MAPK8IP2 or a protein encoded by the gene, the measured When the expression level of mRNA or protein is the same as or higher than that of the normal group, the risk of liver fibrosis is predicted to be low, and when the measured expression level of mRNA or protein is lower than the expression level of the normal group, the risk of liver fibrosis is high. An information provision method for predicting the risk of liver fibrosis is provided.

According to another aspect of the present invention, at least one selected from the group consisting of CSF3R, JAK2, MAP3K14 and MAPK8IP2 and mRNA of a gene comprising SOX17 or a protein encoded by the gene comprising the step of measuring the expression level However, when the expression level of at least one measured mRNA or protein selected from the group consisting of CSF3R, JAK2, MAP3K14 and MAPK8IP2 is equal to or higher than the expression level of the normal group, the risk of liver fibrosis is predicted to be low, and the CSF3R When the expression level of at least one measured mRNA or protein selected from the group consisting of , JAK2, MAP3K14 and MAPK8IP2 is lower than the expression level of the normal group, the risk of liver fibrosis is predicted to be high, and the measured mRNA or protein of the SOX17 gene Provided is an information providing method for predicting the risk of liver fibrosis, predicting that there is a risk of liver fibrosis when the expression level of is higher than the expression level of the normal group.

According to another aspect of the present invention, the mRNA of at least two genes selected from the group consisting of SF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 or an agent for measuring the expression level of a protein encoded by the gene comprises an agent, A composition for predicting the risk of liver fibrosis is provided.

In addition, the agent for measuring the expression level of the mRNA may be sense and antisense primers or probes complementary to the mRNA of the gene.

In addition, the agent for measuring the expression level of the protein may include an antibody that specifically binds to the protein, an interacting protein, a ligand, an oligopeptide, a peptide nucleic acid (PNA), a nanoparticle, or an aptamer.

According to the information providing method according to an aspect of the present invention, the risk according to the liver fibrosis stage can be classified and predicted with high sensitivity and specificity. Through this, information on the progression stage of liver fibrosis can be obtained with high accuracy, and treatment for liver fibrosis can be effectively performed by applying an appropriate treatment for the progression stage of liver fibrosis.

In addition, the marker according to another aspect of the present invention can be used as a target for studying the mechanism of progressive liver fibrosis, developing a therapeutic agent, or building a model for predicting therapeutic effect.

1 shows the entire process of the information providing method according to an embodiment of the present invention;

2 is a check of the mRNA expression level of the marker according to an aspect of the present invention by dividing the liver fibrosis stage;

3 is a ROC curve of a gene combination including all markers according to an aspect of the present invention.

The present invention relates to a marker capable of diagnosing liver fibrosis as well as predicting the risk of liver fibrosis based on the progression stage of the liver fibrosis and its use.

Hereinafter, the present invention will be described in more detail.

As used herein, the term marker refers to a molecule that is quantitatively or qualitatively associated with the existence of a biological phenomenon. The marker is a gene product, including a gene, RNA, polynucleotide, peptide, or protein directly or indirectly related to an antecedent mechanism of the phenomenon; related metabolites; It may include both antibodies or other identifying molecules such as antibody fragments.

As used herein, the term "gene signature" refers to a combination of markers that can predict the risk of liver fibrosis with good reliability.

The term liver fibrosis risk used in the present invention includes information on the progression stage of liver fibrosis in addition to the diagnosis of liver fibrosis disease. Based on the progression stage of liver fibrosis, it can be divided into a low-risk group in the early or early stage, and a high-risk group in the late stage. Here, the liver fibrosis progression stage can be classified as an early stage when it is stage 1 or 2, and a late stage when it is stage 3 or 4 according to the METAVIR liver fibrosis stage definition.

The term threshold or reference value used in the present invention can be obtained through statistical analysis, and can be understood as a number that can be determined whether a patient has or does not have a disease for a given condition. Hereinafter, it will be described as a reference value. In general, a reference value in a diagnostic sense can be set to achieve a desired sensitivity and specificity depending on factors such as populations, prevalence, and clinical outcome, calculated through algorithms or computerized data analysis, or can be established The reference value may be experimentally established through clinical studies such as those described in Examples to be described later. Depending on the disease prediction model used, it can be set to achieve maximum sensitivity, specificity or minimum error. In one embodiment according to the present invention, the reference value is set to achieve maximum sensitivity, specificity, accuracy, positive predictive value and negative predictive value.

As used in the present invention, the term sensitivity (Sensitivity) means the ability to detect a person in the late stage of liver fibrosis as a high-risk group, and the specificity (Specificity) means the ability to detect a person in the early stage of liver fibrosis as a low-risk group, Accuracy refers to the probability of judging a person in the late stage of liver fibrosis as a high-risk group and a person in an early stage of liver fibrosis as a low-risk group. It means the probability of being in the late stage, and the negative predictive value can be understood as meaning the probability of being in the actual pre-stage liver fibrosis among those in the low-risk group. The sensitivity, specificity, accuracy, positive predictive value, and negative predictive value may be calculated using the values in Table 1 below. In Table 1, the high-risk group and the low-risk group are divided based on the threshold obtained through the ROC curve.

	Advanced Stage (number)	Early Stage (number)	total (number)
high risk group (number)	A (True Positive)	B (False Positive)	Total (Test Positive)
low-risk group (number)	C (False Negative)	D (True Negative)	Total (Test Negative)
	Total (Advanced Stage)	Total (Early Stage)	Total

The sensitivity may be calculated as A/(A+C) using the values in Table 1 above. The specificity may be calculated as D/(B+D) using the values in Table 1 above. The accuracy may be calculated as A+D/(A+B+C+D) using the values in Table 1 above. The positive predictive value may be calculated as A/(A+B) using the values in Table 1 above. The speech predictive value may be calculated as D/(C+D) using the values in Table 1 above. In Table 1, Advanced Stage means the late stage according to the progress of liver fibrosis, and Early Stage means the early stage according to the progress of liver fibrosis.

In the present invention, terms used in statistical analysis related to examination or diagnosis for diseases mean commonly understood in the art in relation to regression analysis and ROC (Receiver Operation Characteristic) curves, except for the aforementioned terms. can be interpreted as

The term Liver fibrosis prediction score used in the present invention means a value for classifying the risk of liver fibrosis into a high-risk group or a low-risk group using a genetic signature, and It can be calculated using the regression coefficient value for each marker of univariate logistic regression analysis obtained through the ROC curve.

According to one aspect of the present invention, a marker capable of predicting or diagnosing the risk of liver fibrosis and a composition comprising the same are provided.

A marker according to an aspect of the present invention and a gene signature comprising the same may be derived through statistical analysis. Through statistical analysis, a marker for predicting or diagnosing the risk of liver fibrosis according to an aspect of the present invention and a method for providing information on a gene combination for performing an information providing method according to another aspect of the present invention using the same An example is shown in FIG. 1 . Referring to FIG. 1 , performing nCounter Assay to obtain expression data for expression level or expression pattern, and through correlation analysis, different expression according to liver fibrosis stage (Differentially Expressed Genes) Information on Candidate of Gene Signatures, which is the optimal gene combination, by performing machine learning including steps to obtain information on the gene combo analysis and cross validation A marker and a gene signature according to an aspect of the present invention may be derived through the steps of obtaining , and determining the gene signature through a regression analysis (univariate or multivariate regression analysis).

The nCounter Assay may be performed to derive markers related to the classification of liver fibrosis and its progression stages. Specifically, a normal group without hepatic fibrosis disease, a low-risk group with a low risk of liver fibrosis, and a high-risk group with a high risk of liver fibrosis were distinguished, and the expression or level of gene expression was checked for each group to determine the significant expression difference between the groups. It can be performed to derive a gene with Whether or not the expression of a gene or its expression level can be measured through an mRNA transcribed from the gene or a protein encoded by it, the mRNA or protein can be extracted from a sample obtained from a patient with a disease or a subject who wants to be tested. The sample may be a biological sample derived from the subject, for example, liver tissue or liver tissue-derived cells, but is not limited thereto. The sample may be from a fresh-frozen biopsy sample isolated in vitro or a formalin-fixed paraffin-embedded (FFPE) biopsy sample, but is not limited thereto. The biopsy sample may be a surgical tissue or a tissue derived from a biopsy.

The nCounter Assay is only a preferred exemplary means selected to quickly check the expression difference of the gene, and the means for checking the expression difference of the gene in the present invention is not limited thereto. For example, the measurement of mRNA expression or expression level is RT-PCR, competitive RT-PCR (RT-PCR), real-time RT-PCR (RT-PCR), in-situ hybridization (in-situ hybridization). hybridization), RNase protection assay, Northern blot, and at least one method selected from the group consisting of a DNA chip, but is not limited thereto. In another example, the determination of the presence or absence of a protein or measurement of its expression level is performed by immunohistochemistry staining, Western blot, ELISA (enzyme linked immunosorbent assay), radioimmunoassay, radioimmunodiffusion, At least selected from the group consisting of ouchterlony immunodiffusion, rocket immunoelectrophoresis, immunoprecipitation assay, complement fixation assay, FACS and protein chip It may be performed through one method, but is not limited thereto.

After obtaining the expression value of the gene, statistical analysis may be performed on it. A regression coefficient and a reference value may be obtained through the statistical analysis, and the risk of liver fibrosis may be predicted or diagnosed using the regression coefficient and the reference value.

The statistical analysis may be performed through regression analysis. The regression analysis may be performed through an ROC curve. Univariable or multivariable analysis can be performed, and the markers according to an aspect of the present invention derived through the above-described method are CSF3R (Colony Stimulating Factor 3 Receptor), JAK2 (Janus Kinase 2), MAP3K14 (Mitogen-Activated Protein Kinase Kinase Kinase 14), MAPK8IP2 (Mitogen-Activated Protein Kinase 8 Interacting Protein 2), and SOX17 (SRY-Box Transcription Factor 17). It contains at least one gene selected from the group consisting of. For the gene signature, a gene combination excellent in AUC, sensitivity and specificity may be selected from among gene combinations including the above genes. The gene signature may be selected to include at least two genes selected from the group consisting of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17, and preferably a gene combination comprising CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 can be selected.

According to another aspect of the present invention, there is provided an information providing method for predicting or diagnosing the risk of liver fibrosis using the marker.

The information providing method according to a preferred embodiment of the present invention comprises the steps of determining a gene combination using at least two genes selected from the group consisting of derived genetic markers CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17; Measuring the expression level of the gene included in the determined gene combination; obtaining each regression coefficient value for each gene through regression analysis; obtaining a liver fibrosis prediction score using the regression coefficient value; obtaining a reference value through regression analysis for a gene combination including all of the genes; and comparing the liver fibrosis prediction score with the reference value.

In order to improve the reliability of predicting the risk of liver fibrosis, it is recommended to use the derived genetic marker as a gene combination including two or more of the derived genes rather than using a single one. Preferably, the gene combination includes at least two selected from the group consisting of JAK2, MAP3K14 and SOX17, or preferably includes CSF3R and MAP3K14. Alternatively, it is preferable to include JAK2, MAPK8IP2 and SOX17. More preferably, the gene combination includes SOX17 and further includes at least three genes selected from the group consisting of CSF3R, JAK2, MAP3K14 and MAPK8IP2.

The expression of the gene may be measured or confirmed using the mRNA of the gene or the expression of a protein encoded by the gene.

The regression analysis may be performed using an ROC curve. Regression coefficient values for each gene can be obtained through univariate logistic regression analysis. The reference value may be obtained through multivariate logistic regression analysis. In detail, the reference value may be obtained through multivariate logistic regression analysis using an ROC curve.

If the liver fibrosis prediction score is higher than the reference value, it may be determined that the risk of liver fibrosis is high in the later stage of progression. If the liver fibrosis prediction score is lower than the reference value, it may be determined that the risk of liver fibrosis is low in the initial stage of progression.

The information providing method according to another example of the present invention measures the mRNA of at least two genes selected from the group consisting of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 or the expression level of a protein encoded by the gene, and measured Comparing the expression level of mRNA or protein with the expression level measured in the normal group may be performed.

The expression levels of CSF3R, JAK2, MAP3K14 and MAPK8IP2 are the same or higher than in the normal group in the early stage of liver fibrosis, and their expression levels are lowered in the late stage of liver fibrosis. In addition, when comparing the early and late stages of liver fibrosis, the expression level is lowered in the late stage compared to the early stage of liver fibrosis. Considering these expression characteristics, if the expression levels of CSF3R, JAK2, MAP3K14 and MAPK8IP2 measured in the test subject are the same as or higher than the expression level of the normal group, the risk of liver fibrosis is predicted to be low, and the measured expression level is the normal group If it is lower than the expression level of , it can be predicted that the risk of liver fibrosis is high.

The expression level of SOX17 is higher in the late stage than in the early stage of liver fibrosis, and when the expression level of SOX17 is higher than the expression level of the normal group, it can be predicted that there is a risk of liver fibrosis. At this time, it can be predicted that the higher the expression level than the normal group, the higher the risk of liver fibrosis.

According to another aspect of the present invention, there is provided a composition for predicting the risk of liver fibrosis using a marker according to an aspect of the present invention.

The composition comprises at least one selected from the group consisting of SF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17, preferably, mRNA of at least two genes or an agent for measuring the expression level of a protein encoded by the gene. can

The agent for measuring the mRNA expression level may be, for example, sense and antisense primers or probes complementary to mRNA of a gene, but is not limited thereto. The agent for measuring the expression level of the protein may include, for example, an antibody that specifically binds to the protein, an interacting protein, a ligand, an oligopeptide, a peptide nucleic acid (PNA), a nanoparticle, or an aptamer. It is not limited.

Hereinafter, preferred embodiments are presented to help the understanding of the present invention, but these examples are merely illustrative of the present invention and do not limit the appended claims, and are within the scope and spirit of the present invention. It is obvious to those skilled in the art that various changes and modifications are possible, and it is natural that such variations and modifications fall within the scope of the appended claims.

Example

<Collection of samples>

Specimens were collected from a total of 94 patients. An extracorporeally isolated fresh-frozen biopsy sample (surgical tissue) was used as a sample for the experiment. Written consent was obtained from the patient before participation in the trial, and approval from the Clinical Trial Review Committee of Ajou University Hospital (approval number: AJIRB-BMR-KSP-18-444) was obtained. When the collected specimens were classified according to the stage of liver fibrosis disease, there were 17 normal specimens, 12 stage 1 liver fibrosis, 12 stages 2 stage 2, 25 stage 3 stage 4 specimens, and a total of 77 specimens with liver fibrosis disease. The stages of liver fibrosis were classified according to the METAVIR liver fibrosis stage definition, and the degree of fibrosis was evaluated by histological analysis by a pathologist. Data obtained from patients were used to collect significant fibrosis-related variables in patient characteristics and univariate analyses.

<Statistical analysis of clinical information>

Among 94 patients, the clinical information of 77 samples with liver fibrosis disease was analyzed and confirmed by the univariable logistic regression method, and the analysis results are shown in Table 2 below.

variable (Variable)	number of patients (n)	coef	Odds Ratio	se(coef)	z	P-value	lower .95	upper .95
age (<55 years vs ≥55 years)	77	-0.0711	0.93	0.5097	-0.139	0.8891	0.34	2.53
gender (male vs female)	77	-0.3600	0.70	0.5878	-0.612	0.5402	0.22	2.24
HBV (None vs Yes)	74	1.7525	5.77	0.6048	2.898	0.0038	1.76	18.88
HCV (None vs Yes)	73	15.7653	7027141.58	1696.7344	0.009	0.9926	0.00	Inf
AFP (<100ng/mL vs ≥100ng/mL)	77	0.5008	1.65	0.4970	1.008	0.3140	0.62	4.37

Referring to Table 2, as a result of univariate regression analysis by items other than gene combinations, it was found that the low-risk and high-risk groups for liver fibrosis were significantly different depending on the presence or absence of HBV (Hepatitis B), and HBV also showed that progressive liver fibrosis It can be seen that the distinction between

<RNA extraction>

Total RNA was extracted from surgical tissues (n=94) using RNeasy mini kit (Qiagen, Hilden, Germany) and DNase I treatment (Qiagen, Hilden, Germany) (Table 1). Total RNA integrity was verified using Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA), and total RNA concentration was measured using Nanodrop 2000 (Thermo Fisher scientific, Waltham, MS, USA).

<Gene expression analysis>

Sample gene expression was performed using the nCounter PanCancer Pathway Panel (Nanostring Technologies, Seattle, WA) using nCounter MAX (Nanostring, Technologies, Seattle, WA, USA). The panel analyzed 770 genes, including 40 control genes, and each reaction contained 100 ng of total RNA and reporter and capture probes in 15 μl aliquots. Quality control and normalization of raw data was performed using nSolver Analysis Software v 4.0 (Nanostring Technologies, Seattle, WA, USA). Among the analyzed 770 genes, 730 endogenous genes and 40 housekeeping genes were identified. These genes were normalized using nSolver software. For the program setting for standardization, select Geometric mean in Positive Control Normalization, Range as 0.3-3, and Standard in CodeSet Content (Reference or Housekeeping) Normalization, Codeset Content for Endogenous genes, Normalization Codes for Housekeeping genes. was selected, geometric mean was selected, and the range was set to 0.1-10.

Data standardization was performed with the above-described settings, and the values of standardized expression levels of 730 genes were obtained.

When the sample size of the group is small and the normality is not satisfied, the Wilcoxon rank sum test, a method of analyzing the difference between two different groups, was used for evaluation, and p<0.05 was considered statistically significant. did

Logistic regression analysis is an analysis method to analyze the effect of a quantitative variable that can be scored as an analysis for use in future predictive models by expressing the relationship between the dependent variable and the independent variable as a specific function on the dichotomous variable, and p<0.05 were considered to be academically significant.

<Identification of genes with different expression levels according to liver fibrosis stage>

Stages 1 and 2 of liver fibrosis were classified as a low-risk group for liver fibrosis, and stages 3 and 4 of liver fibrosis were classified as a high-risk group, and genes with differences in gene expression were identified between the low-risk and high-risk groups. As a result of analyzing the different gene expression between the low-risk group (n=24) and the high-risk group (n=53) in a cohort of patients with liver fibrosis (n=77), 71 genes out of 730 endogenous genes were statistically significantly differentially expressed. was found (P<0.05). Specifically, out of 71 genes, the expression of 10 genes was increased in the high-risk group than in the low-risk group, and the expression of 61 genes was decreased in the high-risk group than in the low-risk group. Some statistical analysis results of these 71 genes are shown in Table 3 below.

SEQ	gene (Gene)	number of samples (n)	Logistic Regression P-value	(FS1,2 vs 3,4) Wilcoxon P-value	Fold Change
8	SOX17	77	1.62E-03	4.61E-05	2.09
29	JAK2	77	2.48E-04	8.55E-09	-2.33
37	CSF3R	77	4.26E-05	1.14E-06	-2.59
40	MAPK8IP2	77	3.79E-04	5.68E-03	-2.62
71	MAP3K14	77	2.41E-03	1.26E-07	-6.37

<Perform statistical combinatorial genetic analysis>

In order to find the optimal gene combination for diagnosing the liver fibrosis stage, combinatorial statistical analysis was performed on 71 genes with different expression according to the liver fibrosis stage. Combination statistical analysis was performed through logistic regression combination analysis, ROC (Receiver Operation Characteristic) combination analysis, and cross validation combination analysis (300 cross-analysis). Here, each combination analysis was performed from Single to Combo10, which means the number of combined genes. Specifically, Single means one gene and Combo10 means a combination of 10 genes.

All statistical analyzes in this study were performed using the open source statistical programming environment R language (Version 3.4.3). In the training cohort, differentially expressed genes (DEGs) were classified as overexpressed or underexpressed through Wilcoxon rank sum test (p <0.05 and |fold-change| > 2), the low-risk group (Fibrosis Stage 1,2) was compared with the high-risk group (Fibrosis Stage 3,4). DEGs were further shortlisted using univariate logistic regression. The number of selected DEGs was combined and analyzed, and the total number of gene combinations was calculated using Equation 1 below.

In Equation 1, n is the total number of selected DEGs, and k is the number of genes included in combination.

Multivariate logistic regression analysis was performed to determine the relevance of gene signatures and clinicopathological features (p<0.05). Sensitivity and specificity were obtained and used to obtain the optimal gene combination, and in addition, accuracy, positive predictive value, and negative predictive value were additionally obtained and utilized.

The candidate gene signature of the optimal gene combination was obtained under the conditions of p < 0.05, AUC > 0.800, sensitivity > 80%, and specificity > 80%, and the optimal gene combination ( gene combinations) were identified. If the AUC is 0.800 or less or the sensitivity and specificity are 80% or less, there is a problem in that the reliability of the diagnosis and prediction of a disease using the same is lowered. An optimal gene combination was obtained under one condition. The training cohort was divided into two folds (training set and test set) and the results were confirmed by applying the reference values from the training set to the test set. . Accuracy was calculated based on p <0.05 for the test set.

As a result of statistical analysis of candidate combinatorial genes, it was confirmed that a combination with significantly excellent prediction accuracy was confirmed by satisfying values of 90% or more in all items with sensitivity of 92.45%, specificity of 91.67%, and accuracy of 92.21%, and the results are shown in Table 4.

genetic combination	number of genes	Logistic Regression Continuous P-value	ROC AUC	Pre Vali- dation accuracy	Logistic Regression Discrete P-value	threshold	responsiveness	specificity	accuracy	positivity predictability	voice predictability
CSF3R_JAK2_MAP3K14_MAPK8IP2_SOX17	5	5.80E-04	0.951	77.33	5.68E-08	-8.436595	92.45	91.67	92.21	96.08	84.62

Referring to Table 4, it was confirmed that the sensitivity, specificity, accuracy, and predictability were significantly higher when the combined analysis results of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 five genes were analyzed. In addition, as a result of analyzing the gene expression of the five genes in each stage of liver fibrosis, it was confirmed that the expression was statistically significantly different (p<0.05). The results of confirming the expression levels of the above five genes at the level of liver fibrosis at the mRNA level are shown in FIG. 2 .

Referring to FIG. 2 , the expression of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 genes was different in the normal group and in subjects with liver fibrosis. Looking at the change in expression of each gene between the normal group and the early stage of liver fibrosis, it was found that the expression of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 genes increased compared to the normal group in the early stage of liver fibrosis. Looking at the changes in the expression of each gene in the early and late stages of liver fibrosis, the expression of CSF3R, JAK2, MAP3K14 and MAPK8IP2 genes was decreased in the late stage than in the early stage of liver fibrosis, and the SOX17 gene was expressed in the later stage than in the early stage. appeared to increase. Through this, it can be confirmed that the expression of the above-mentioned five genes appears differently depending on the stage of liver fibrosis.

Next, using 5 genes CSF3R, JAK2, MAP3K14, MAPK8IP2, and SOX17, a total of 31 combinations were generated as a combination of Single to Combo5, and the accuracy, sensitivity and specificity were compared for each gene combination, and the results are shown in Table 5 below. . In the gene combinations shown in Table 5 below, C is CSF3R, J is JAK2, M3 is MAP3K14, M8 is MAPK8IP2, and S is SOX17 as an abbreviation. For example, C-J refers to the gene combination of CSF3R and JAK2. In addition, in Table 5 below, YI (Youden Index) is to find a position where the sum of specificity and sensitivity is the maximum to obtain a threshold from the ROC curve, and through a formula subtracting 1 from the sum of sensitivity and specificity saved In the evaluation of a diagnostic or predictive model, it is desirable that the sum of sensitivity and specificity is high, and YI can be used as an important indicator to confirm the excellence of a model to which a gene combination is applied. The AUC (Area Under the Curve) of ROC is a performance evaluation standard that can show the correlation between overall sensitivity and specificity, and was referred to to evaluate the accuracy of the model.

SEQ	gene combination	number of genes	ROC AUC	responsiveness	specificity	accuracy	YI
One	C-J-M3-M8-S	5	0.951	92.45	91.67	92.21	0.8412
2	J-M3-M8-S	4	0.932	96.23	83.33	92.21	0.7956
3	J-M8-S	3	0.926	96.23	83.33	92.21	0.7956
4	J-M3-S	3	0.925	96.23	79.17	90.91	0.754
5	C-J-M3-S	4	0.932	92.45	83.33	89.61	0.7578
6	J-S	2	0.913	92.45	83.33	89.61	0.7578
7	C-J	2	0.892	98.11	70.83	89.61	0.6894
8	C-J-S	3	0.926	92.45	79.17	88.31	0.7162
9	C-J-M8_S	4	0.945	84.91	91.67	87.01	0.7658
10	M3-M8-S	3	0.894	90.57	79.17	87.01	0.6974
11	C-J-M3	3	0.909	92.45	75	87.01	0.6745
12	C-M3-M8	3	0.886	96.23	66.67	87.01	0.629
13	C-M8	2	0.859	96.23	66.67	87.01	0.629
14	C-M3-M8-S	4	0.932	81.13	95.83	85.71	0.7696
15	C-M3-S	3	0.921	84.91	87.5	85.71	0.7241
16	C-M3	2	0.887	86.79	83.33	85.71	0.7012
17	C-J-M3-M8	4	0.91	90.57	75	85.71	0.6557
18	C-M8-S	3	0.927	84.97	83.33	84.42	0.6824
19	J-M3	2	0.924	84.91	83.33	84.42	0.6824
20	M3-S	2	0.902	84.91	83.33	84.42	0.6824
21	M8-S	2	0.874	86.79	79.17	84.42	0.6596
22	J-M3-M8	3	0.909	81.13	87.5	83.12	0.6863
23	C-S	2	0.906	75.47	91.67	80.52	0.6714
24	C-J-M8	3	0.887	77.36	87.5	80.52	0.6486
25	M3-M8	2	0.829	86.19	66.67	80.52	0.5346
26	J-M8	2	0.885	64.15	95.83	74.03	0.5998
27	S	One	0.792	62.26	91.67	71.43	0.5393
28	M8	One	0.698	79.25	25	62.34	0.0425
29	J	One	0.912	86.79	0	59.74	-0.1321
30	M3	One	0.878	86.79	0	59.74	-0.1321
31	C	One	0.848	86.79	0	59.74	-0.1321

Referring to Table 5 above, when using at least two or more of the above-mentioned gene group together, AUC, sensitivity compared to the case of using five CSF3R, JAK2, MAP3K14, MAPK8IP2 or SOX17 genes alone (seq.27-31) , specificity, accuracy, and YI values were significantly increased. Many excellent gene combinations satisfying sensitivity, specificity, and accuracy>80.0 and AUC>0.8 were identified, and when all five genes CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 were combined (seq.1), AUC and YI were the most It was confirmed that it was excellent.

<Deduction of liver fibrosis prediction score through statistical analysis of gene signature candidates>

Liver fibrosis prediction score (Liver fibrosis) using the logistic regression coeffieient values for each gene obtained from univariate logistic regression for the selected CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 five gene combinations prediction score) was obtained. The regression coefficient values for each gene are shown in Table 6 below.

SEQ	gene	regression coefficient
One	CSF3R	-0.007283
2	JAK2	-0.005178
3	MAP3K14	-0.002994
4	MAPK8IP2	-0.059009
5	SOX17	0.027697

The liver fibrosis prediction score using the gene-specific regression coefficients in Table 6 above can be calculated using the formula (-0.007283)×CSF3R+(-0.005178)×JAK2+(-0.002994)×MAP3K14+(-0.059009)×MAPK8IP2+0.027697×SOX17. have. Here, each gene means the expression level of each gene in the sample to be tested to predict the risk by checking the progress of liver fibrosis, and nSolver It is a value that has been normalized using the 4.0 software.

Early or late judgment of liver fibrosis can be made using the reference value obtained through the ROC curve analysis for the above five gene combinations, and through this, the risk according to the progress of liver fibrosis is determined, and Prediction may be performed. The results of the ROC curve analysis performed to obtain the reference value are shown in FIG. 3 .

The threshold derived through the ROC curve analysis was confirmed to be -8.436595. If the liver fibrosis prediction score of the sample to be tested exceeded the derived reference value, it was classified as a high-risk group, and if it was less than or equal to the derived reference value, it was classified as a low-risk group, and progressive liver fibrosis was predicted for the high-risk group.

The regression coefficient and the derived reference value of Table 6 are values obtained based on the gene expression analysis value obtained through the nCounter assay in an embodiment of the present invention, and the measuring method for obtaining the RNA expression value is different from this example Of course, when the numerical value of the obtained expression value is changed, the regression coefficient and the reference value obtained therefrom may also be different from the aforementioned numerical values.

Multivariate analysis explains the degree of actual influence of each factor by correcting the effects of several risk factors that affect one dependent variable with one model. All of these become the corrected Odds Ratio.

After performing univariate analysis on the gene signature of the optimal gene combination of five gene combinations, CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17, multivariate analysis was performed on statistically significant factors. Specifically, the relationship between the gene signature and HBV was evaluated and the results are shown in Tables 7 and 8. Table 7 below shows the results of univariate analysis on the gene signature, and Table 8 below shows the results of multivariate analysis on the gene signature and HBV.

variable	n	coef	Odds Ratio	se(coef)	z	P-value	lower .95	upper .95
Liver fibrosis prediction score (low vs high)	77	4.9034	134.75	0.9033	5.429	5.68E-08	22.94	791.37

variable	n	coef	Odds Ratio	se (coef)	z	P-value	lower .95	upper .95
Liver fibrosis prediction score (low vs high)	74	5.164	174.91 (19.03-1603.36)	1.130	4.568	4.91E-06	19.08	1603.36
HBV		2.444	11.52 (1.00-132.63)	1.247	1.961	4.99E-02	1.00	132.63

Referring to Tables 7 and 8, it was confirmed that the gene signature and HBV derived according to an aspect of the present invention are factors capable of independently predicting progressive liver fibrosis. In addition, it was confirmed that progressive liver fibrosis can be independently diagnosed with a gene signature alone.

Claims (15)

1. A marker composition for predicting the risk of liver fibrosis, comprising at least one gene selected from the group consisting of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17.
2. determining a gene combination using at least two genes selected from the group consisting of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17;

   measuring the expression level for each gene included in the determined gene combination;

   obtaining a regression coefficient value for the expression level for each gene through regression analysis; and

   Obtaining a liver fibrosis prediction score using the regression coefficient value; further comprising, information providing method for predicting liver fibrosis risk.
3. 3. The method of claim 2,

   Comprising the step of obtaining a reference value through the ROC curve for the gene combination and comparing the liver fibrosis prediction score and the reference value,

   If the liver fibrosis prediction score is higher than the reference value, the liver fibrosis risk is determined to be high, and if the liver fibrosis prediction score is lower than the reference value, the liver fibrosis risk is determined to be low. Information providing method for predicting the risk of liver fibrosis .
4. 3. The method of claim 2,

   The gene combination includes at least two types selected from the group consisting of JAK2, MAP3K14 and SOX17, an information providing method for predicting the risk of liver fibrosis.
5. 3. The method of claim 2,

   The gene combination includes CSF3R and MAP3K14, an information providing method for predicting the risk of liver fibrosis.
6. 3. The method of claim 2,

   The gene combination includes JAK2, MAPK8IP2 and SOX17, an information providing method for predicting the risk of liver fibrosis.
7. 3. The method of claim 2,

   The gene combination includes SOX17 and further comprises at least three genes selected from the group consisting of CSF3R, JAK2, MAP3K14 and MAPK8IP2. An information providing method for predicting the risk of liver fibrosis.
8. measuring the expression levels of CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 genes;

   obtaining a regression coefficient value for the expression level for each gene through logistic regression analysis; and

   Obtaining a liver fibrosis prediction score using the regression coefficient value; further comprising, information providing method for predicting liver fibrosis risk.
9. CSF3R, JAK2, MAP3K14, MAPK8IP2 and SOX17 comprising the step of measuring the expression level of the mRNA of at least two genes selected from the group consisting of or a protein encoded by the gene, information providing method for predicting the risk of liver fibrosis .
10. 10. The method of claim 9,

    Information providing method for predicting the risk of liver fibrosis, further comprising the step of comparing the measured expression level of mRNA or protein with the expression level of the normal group.
11. CSF3R, JAK2, comprising the step of measuring the expression level of at least two genes selected from the group consisting of MAP3K14 and MAPK8IP2 mRNA or a protein encoded by the gene,

    When the measured mRNA or protein expression level is equal to or higher than that of the normal group, the risk of liver fibrosis is predicted to be low, and when the measured mRNA or protein expression level is lower than the expression level of the normal group, the risk of liver fibrosis is Predicting high, information providing method for predicting the risk of liver fibrosis.
12. At least one selected from the group consisting of CSF3R, JAK2, MAP3K14 and MAPK8IP2 and measuring the expression level of the mRNA of a gene comprising SOX17 or a protein encoded by the gene,

    When the expression level of at least one measured mRNA or protein selected from the group consisting of CSF3R, JAK2, MAP3K14 and MAPK8IP2 is equal to or higher than the expression level of the normal group, the risk of liver fibrosis is predicted to be low, and the measured mRNA or If the expression level of the protein is lower than the expression level of the normal group, the risk of liver fibrosis is predicted to be high,

    When the measured mRNA or protein expression level of the SOX17 gene is higher than the expression level of the normal group, predicting that there is a risk of liver fibrosis, an information providing method for predicting the risk of liver fibrosis.
13. SF3R, JAK2, MAP3K14, MAPK8IP2 and a composition for predicting the risk of liver fibrosis, comprising an agent for measuring the expression level of mRNA of at least two genes selected from the group consisting of or SOX17 genes or proteins encoded by the genes.
14. 14. The method of claim 13,

    The agent for measuring the expression level of the mRNA is characterized in that the sense and antisense primers or probes that complementarily bind to the mRNA of the gene, the composition.
15. 14. The method of claim 13,

    The agent for measuring the expression level of the protein is characterized in that it comprises an antibody, interacting protein, ligand, oligopeptide, PNA (peptide nucleic acid), nanoparticles or aptamer that specifically binds to the protein, composition .

Images