Construction method and system of chronic hepatitis B liver inflammation grading prediction model and application

Abstract

The invention provides a construction method and system of a chronic hepatitis B liver tissue inflammation grading prediction model and application. The construction method of the chronic hepatitis B liver tissue inflammation grading prediction model comprises the following steps: S1, screening out a plurality of patients which are diagnosed as chronic hepatitis B through liver tissue biopsy pathology and have clear liver tissue inflammation grading according to inclusion and exclusion standards, and collecting clinical data and serum samples of the patients; s2, performing serum protein N-carbohydrate chain fingerprint spectrum difference analysis on the serum samples of the patients with different ALT levels by adopting an N-carbohydrate fingerprint spectrum technology; s3, on the basis of conventional biochemical indexes and / or N-sugar markers, constructing liver tissue inflammation grading evaluation models under different ALT levels by adopting a binary logistic regression analysis method; and S4, respectively evaluating and selecting the model with the optimal efficiency as a chronic hepatitis B liver tissue inflammation grading prediction model. The chronic hepatitis B liver tissue inflammation grading method can be used for quickly, accurately and noninvasively grading the chronic hepatitis B liver tissue inflammation.

Description

Technical Field

[0001] The present invention relates to the technical field of N-glycome fingerprint (NGFP) detection, and in particular to the application of N-glycome markers in chronic hepatitis B liver tissue inflammation grading. Background Art

[0002] Hepatitis B Virus (HBV) infection is one of the major challenges in the field of global public health. According to statistics from the World Health Organization (WHO), the global hepatitis B surface antigen (HBsAg) positivity rate was 3.3% in 2022, and the number of chronic HBV-infected people reached 254 million, of which about 1.1 million people died from HBV-related diseases such as cirrhosis and liver cancer. CHB is one of the main causes of cirrhosis and liver cancer. Although there are currently a variety of antiviral treatments, choosing the right time for treatment is still a key factor in determining the efficacy. The active inflammatory phase of hepatitis is considered to be the best time for treatment.

[0003] At present, liver biopsy is still the gold standard for determining the degree of liver tissue inflammation, but its invasiveness and poor patient acceptance limit its widespread application. Clinically, the level of alanine aminotransferase (ALT) is one of the important determinants for the initiation of antiviral treatment for chronic hepatitis B, but it has limitations in reflecting the degree of liver tissue inflammation. The increase in ALT may be caused by a variety of non-liver factors, such as muscle damage or drug effects, which reduces its specificity. In patients with chronic liver disease, ALT levels may be normal even if there is inflammation in the liver. In addition, ALT levels are easily affected by factors such as diet and weight changes, showing volatility. Therefore, it is of great clinical significance to find a non-invasive, simple and sensitive indicator and diagnostic tool for liver tissue inflammatory activity. Summary of the invention

[0004] In order to overcome the above technical defects, the first aspect of the present invention provides a method for constructing a chronic hepatitis B liver tissue inflammation grading prediction model, which comprises:

[0005] Step S1: screening a number of patients who were diagnosed with chronic hepatitis B by liver tissue biopsy pathology and had a clear liver tissue inflammation grade according to the inclusion and exclusion criteria, collecting the clinical data of the patients, and collecting serum samples from the patients;

[0006] Step S2: using N-glycomic fingerprint profiling (NGFP) technology to perform serum protein N-glycan fingerprint differential analysis on serum samples of patients with different alanine aminotransferase (ALT) levels, and obtaining N-glycan markers at different ALT levels;

[0007] Step S3: Based on conventional indicators and / or N-glycan markers, binary logistic regression analysis method is used to construct liver tissue inflammation grading assessment models at different ALT levels;

[0008] Step S4: Use the area under the receiver operating characteristic curve (ROC) to evaluate the diagnostic efficacy of the above models, and select the model with the best efficacy as the prediction model for chronic hepatitis B liver tissue inflammation grading.

[0009] Further, in step S2, the N-sugar markers when serum ALT ≤ 40 IU / L are Peak1, Peak2 and Peak5; and the N-sugar markers when ALT > 40 IU / L are Peak1, Peak2, Peak3, Peak4, Peak5 and Peak8.

[0010] Furthermore, in step S3, conventional indicators include gender (Gender), age (Age), total bilirubin (TBIL) and ALT.

[0011] Further, in step S3,

[0012] (1) When ALT ≤ 40 IU / L, the liver tissue inflammation grading assessment models are:

[0013] HIS-R A =0.468×Gender+0.051×Age+0.001×TBIL+0.068×ALT;

[0014] HIS-G A =41.67-0.315×Peak1+0.226×Peak2-0.444×Peak5+0.690×Gender-0.010×Age

[0015] HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757

[0016] (2) When ALT>40IU / L, the liver tissue inflammation grading assessment models are:

[0017] HIS-R B =0.134+0.862×Gender-0.014×Age+0.017×TBIL+0.002×Age

[0018] HIS-G B =16.39-0.19×Peak1+0.96×Peak2-0.26×Peak3+0.53×Peak4-0.20×Peak5-0.30×Peak8+1.48×Gender-0.03×Age

[0019] HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284

[0020] Among them, the values ​​of Peak1, Peak2, Peak3, Peak4, Peak5, and Peak8 are obtained by detecting the patient's serum with NGFP technology to obtain the N-glycan map, extracting the peak height values ​​of the thirteen N-glycans, dividing the peak height value of each peak by the sum of the heights of all peaks, and calculating the relative content of each peak in %; Gender is the patient's gender, with 0 for males and 1 for females; AGE is the patient's age, accurate to one decimal place; TBIL is the patient's serum total bilirubin level, obtained by automatic biochemical analyzer, in μmol / L; ALT is the patient's serum alanine aminotransferase level, obtained by automatic biochemical analyzer, in IU / L.

[0021] Further, in step S4,

[0022] When ALT≤40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0023] HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757

[0024] When ALT>40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0025] HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284.

[0026] The second aspect of the present application provides a chronic hepatitis B liver tissue inflammation grading prediction system, which comprises:

[0027] An input module, the input module is used to obtain data of N-glycogen markers, gender, age, TBIL and ALT of the current patient;

[0028] A grading prediction module, which is used to use a pre-built chronic hepatitis B liver tissue inflammation grading prediction model to perform chronic hepatitis B liver tissue inflammation grading prediction and obtain a grading prediction result;

[0029] An output module, wherein the output module is used to output the hierarchical prediction result.

[0030] Furthermore, the method for constructing the chronic hepatitis B liver tissue inflammation grading prediction model comprises the following steps:

[0031] Step S1: screening a number of patients who were diagnosed with chronic hepatitis B by liver tissue biopsy pathology and had a clear liver tissue inflammation grade according to the inclusion and exclusion criteria, collecting the clinical data of the patients, and collecting serum samples from the patients;

[0032] Step S2: using N-glycan fingerprint technology to perform serum protein N-glycan fingerprint difference analysis on serum samples of patients with different ALT levels to obtain N-glycan markers at different ALT levels;

[0033] Step S3: Based on conventional biochemical indices and / or N-glycan markers, binary logistic regression analysis was used to construct liver tissue inflammation grading assessment models at different ALT levels;

[0034] Step S4: Use the area under the receiver operating characteristic curve to evaluate the diagnostic efficacy of the above models respectively, and select the model with the best efficacy as the chronic hepatitis B liver tissue inflammation grading prediction model.

[0035] Further, the N-sugar markers are Peak1, Peak2, Peak3, Peak4, Peak5 and Peak8;

[0036] When ALT≤40 IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0037] HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757

[0038] When ALT>40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0039] HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284

[0040] Among them, HIS-G A =41.67-0.315×Peak1+0.226×Peak2-0.444×Peak5+0.690×Gender-0.010×Age; HIS-G B =16.39-0.19×Peak1+0.96×Peak2-0.26×Peak3+0.53×Peak4-0.20×Peak5-0.30×Peak8+1.48×Gender-0.03×Age; the values ​​of Peak1, Peak2, Peak3, Peak4, Peak5 and Peak8 are obtained by detecting the patient's serum with NGFP technology to obtain the N-glycan map, extracting the peak height values ​​of the thirteen N-glycans, dividing the peak height value of each peak by the sum of the heights of all peaks, and calculating the relative content of each peak in %; Gender is the patient's gender, with 0 for males and 1 for females; AGE is the patient's age, accurate to one decimal place; TBIL is the patient's serum total bilirubin level, obtained by automatic biochemical analyzer, in μmol / L; ALT is the patient's serum alanine aminotransferase level, obtained by automatic biochemical analyzer, in IU / L.

[0041] The third aspect of the present application provides the use of a combined marker in evaluating the grade of chronic hepatitis B liver inflammation, wherein the combined marker consists of an N-sugar marker, gender, age, TBIL and ALT, and the combined marker is used to evaluate the grade of chronic hepatitis B liver tissue inflammation by constructing a model.

[0042] Furthermore, when ALT≤40 IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0043] HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757

[0044] When ALT>40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0045] HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284

[0046] Among them, HIS-G A =41.67-0.315×Peak1+0.226×Peak2-0.444×Peak5+0.690×Gender-0.010×Age;

[0047] HIS-G B =16.39-0.19×Peak1+0.96×Peak2-0.26×Peak3+0.53×Peak4-0.20×Peak5-0.30×Peak8+1.48×Gender-0.03×Age;

[0048] The values ​​of Peak1, Peak2, Peak3, Peak4, Peak5, and Peak8 are obtained by detecting the patient's serum with NGFP technology to obtain the N-glycan map, extracting the peak height values ​​of the thirteen N-glycans, dividing the peak height value of each peak by the sum of the heights of all peaks, and calculating the relative content of each peak in %; Gender is the patient's gender, with 0 for males and 1 for females; AGE is the patient's age, accurate to one decimal place; TBIL is the patient's serum total bilirubin level, obtained by fully automatic biochemical analyzer, in μmol / L; ALT is the patient's serum alanine aminotransferase level, obtained by fully automatic biochemical analyzer, in IU / L.

[0049] Compared with the prior art, the above technical solution has the following beneficial effects:

[0050] This application uses N-sugar markers, gender, age, TBIL and ALT as joint markers, and uses binary logistic regression method to construct liver tissue inflammation grading assessment model under different ALT levels, and screens out the assessment model with the best model effect as the final chronic hepatitis B liver tissue inflammation grading prediction model. The chronic hepatitis B liver tissue inflammation grading prediction model uses different model formulas to calculate the model calculation value according to the ALT level in the patient's serum. When the model calculation value exceeds the optimal cutoff value, it indicates that the patient's liver tissue has significant inflammatory necrosis (≥G2 grade), and antiviral treatment should be considered clinically; when the model calculation value is less than the optimal cutoff value, it indicates that the patient's liver tissue has no or mild inflammation. The technical solution of this application can quickly, accurately and non-invasively grade the degree of inflammation in chronic hepatitis B liver tissue. Only a trace amount of peripheral blood is required to efficiently identify the liver tissue inflammation risk group (≥G2 grade), which significantly improves the accuracy and safety of non-invasive diagnosis of chronic hepatitis patients, and effectively reduces the need for unnecessary invasive examinations and reduces the potential risks associated with invasive examinations. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1The difference analysis of N-glycan structure abundance among the three groups of liver tissue inflammation G0-1, G2, and G3-G4;

[0052] Figure 2a-2e This is the difference analysis of N-glycan fingerprints at different ALT levels in serum. Figure 2a As shown, serum ALT levels were poorly correlated with the inflammatory activity of liver tissue; Figure 2b It showed that 53.33% of the patients in the normal ALT group had significant liver tissue inflammation; Figure 2c It showed that 27.87% of patients in the ALT elevated group had mild inflammation in liver tissue; Figure 2d It was shown that when serum ALT ≤ 40 IU / L, as the degree of liver tissue inflammation and damage intensified, the abundance of Peak1 and Peak2 showed a significant upward trend, while the abundance of Peak5 decreased accordingly, showing a clear downward trend ( Figure 2d ); Figure 2e It shows that when ALT>40IU / L, the abundance of Peak1-Peak4 increased significantly, while the abundance of Peak5 and Peak8 decreased significantly.

[0053] Figure 3 To evaluate the diagnostic efficiency of the liver tissue inflammation grading assessment model at different ALT levels for significant liver tissue inflammation. DETAILED DESCRIPTION

[0054] The advantages of the present invention are further described below in conjunction with the accompanying drawings and specific embodiments. Those skilled in the art should understand that the following specific description is illustrative rather than restrictive, and should not be used to limit the scope of protection of the present invention.

[0055] Example 1 Construction of a chronic hepatitis B liver tissue inflammation grading prediction model

[0056] This embodiment provides a method for constructing a chronic hepatitis B liver tissue inflammation grading prediction model, which includes steps S1 to S4:

[0057] Step S1: screening a number of patients who were diagnosed with chronic hepatitis B by liver tissue biopsy pathology and had a clear liver tissue inflammation grade according to the inclusion and exclusion criteria, collecting the clinical data of the patients, and collecting serum samples from the patients;

[0058] A total of 221 serum samples from patients diagnosed with chronic hepatitis B (CHB) by liver biopsy were collected in this study. Inclusion criteria for the patients included: hepatitis B surface antigen (HBsAg) that was continuously positive for more than 6 months; or hepatitis B e antigen (HBeAg) that was positive and HBV DNA level was >2000 IU / mL; or HBeAg that was negative and HBV DNA level was >20000 IU / mL. Exclusion criteria included: (1) incomplete clinical information; (2) diagnosed with other chronic liver diseases (such as autoimmune liver disease); (3) combined infection with other hepatitis viruses other than HBV; (4) patients with a history of alcohol consumption (more than 20 g per day) or unstable diabetes; (5) patients with decompensated cirrhosis or hepatocellular carcinoma; (6) patients with serious diseases such as heart, lung, kidney, brain, and blood dysfunction; (7) patients whose blood samples were not obtained.

[0059] Clinical data include but are not limited to age, gender, HBV DNA load, hepatitis B surface antigen (HBsAg), platelet counts (PLT), prothrombin time (PT), total bilirubin (TBIL), albumin (ALB), aspartate aminotransferase (AST), alanine aminotransferase (ALT), γ-glutamyl transpeptidase (GGT), and alkaline phosphatase (ALP).

[0060] The pathological evaluation method for liver tissue biopsy diagnosis includes: liver tissue samples are obtained by liver puncture biopsy or surgery, the tissue length is at least 1-2 cm, and it contains 6-8 portal vein areas. The samples are fixed, embedded, sliced ​​and stained (hematoxylin-eosin staining, Masson staining and reticular fiber staining), and the structure and pathological characteristics of the liver tissue are observed under a microscope to perform a histological evaluation of the degree of liver tissue inflammation. The histological evaluation is blindly interpreted by two pathologists, and the consistency rate of the two experts is higher than 90%. If the results are inconsistent or seriously inconsistent with the clinical information, they will be interpreted by a third senior pathologist. The Scheuer scoring system is used to grade the inflammatory activity, including G0 (no inflammatory activity), G1 (mild inflammatory activity, mainly concentrated in the portal vein area), G2 (moderate inflammatory activity, more significant inflammatory cell infiltration), G3 (severe inflammatory activity, extensive infiltration and accompanied by hepatocellular necrosis) and G4 (very severe inflammatory activity, extensive and severe hepatocellular necrosis and liver structural destruction).

[0061] 221 patients with pathologically confirmed liver tissue inflammation grade were included in this study. Their baseline characteristics are shown in Table 1. Among them, 83 patients (37.55%) were in G0-G1, 71 patients (32.13%) were in G2, 39 patients (17.65%) were in G3, and 28 patients (12.67%) were in G4. Among the different inflammation grades, the differences in PLT, PT, TBIL, ALB, ALT, AST, GGT and ALP were statistically significant (p < 0.05).

[0062] Table 1 Baseline information of enrolled patients

[0063]

[0064] Step S2: using N-glycan fingerprint technology to perform serum protein N-glycan fingerprint difference analysis on serum samples of patients with different ALT levels to obtain N-glycan markers at different ALT levels;

[0065] pass Figure 1It can be found that the abundance of N-glycan structures is different among the three groups of liver tissue inflammation G0-1, G2, and G3-G4: with the aggravation of liver tissue inflammation, the α-1,6-core fucosylated two-antennary structure peak1 (non-galactose-α-1,6-core fucosylated two-antennary structure, NGA2F), peak2 (non-galactose-α-1,6-core fucosylated bisected two-antennary structure, NGA2FB), and peak3 (monogalactose-α-1,6-core fucosylated two-antennary structure, NG1A2F) in the serum of CHB patients gradually increased, while the abundance of non-fucosylated sugar chain structures Peak5 (dialactose two-antennary structure, NA2), Peak8 (trigalactose three-antennary structure, NA3), and Peak11 (tetragalactose four-antennary structure, NA4) showed a trend of gradual decrease.

[0066] This application uses serum ALT 40IU / L as the cut-off point defined by relevant professional guidelines, and constructs N-glycan diagnostic models for normal and abnormal ALT to judge different pathological stages of liver tissue inflammation. The relevant professional guidelines refer to the "Guidelines for the Prevention and Treatment of Chronic Hepatitis B (2022 Edition)", "EASL 2017 Clinical Practice Guidelines on the management of hepatitis B virus infection" and "Expert Consensus on the Diagnosis and Treatment of Chronic HBV Infected Patients with Persistently Normal ALT--2021".

[0067] The present application uses N-glycomic fingerprint profiling (NGFP) technology based on a first-generation DNA sequencer to analyze the N-glycomic profile of serum proteins. This technology can accurately display the height of at least 13 peaks, each of which represents a different N-glycomic structure. The detailed information of these structures is shown in Table 2. The core of the technology is oligosaccharide chain separation and analysis technology based on capillary electrophoresis. The operating steps of N-glycomic fingerprint detection mainly include: (1) release of protein sugar chains; (2) fluorescent dye labeling of sugar chains; (3) removal of sialic acid at the end of sugar chains; (4) DNA sequencer detection.

[0068] Table 2 Description of NGFP detection structure

[0069]

[0070] The serum ALT level is a key determinant of initiating antiviral therapy. We first performed a correlation analysis between the serum ALT level and the grade of liver tissue inflammation to evaluate the effectiveness of ALT as a marker of liver tissue inflammation. Figure 2aAs shown, serum ALT levels are poorly correlated with liver tissue inflammatory activity. Based on the differences in ALT levels, we divided patients into two subgroups: normal ALT group and elevated ALT group. The normal ALT group included patients whose ALT levels did not exceed one times the upper limit of normal (ULN), that is, below 40 IU / L. The elevated ALT group included all individuals with ALT levels exceeding 40 IU / L. The analysis showed that 53.33% of patients in the normal ALT group had significant liver tissue inflammation ( Figure 2b ), while 27.87% of the patients in the ALT elevated group had mild liver inflammation ( Figure 2c ).

[0071] In this example, CHB patients were divided into two groups to analyze the association between the abundance of N-glycan structures in the serum of chronic hepatitis B patients and liver tissue inflammation at different ALT levels. When serum ALT ≤ 40 IU / L, as the degree of liver tissue inflammation and damage intensified: the abundance of Peak1 (galactose-free-α-1,6-core-fucosylated bi-antenna structure, NGA2F) and Peak2 (galactose-free-α-1,6-core-fucosylated bi-antenna structure, NGA2FB) containing α-1,6-core fucosylated bi-antenna structure showed a significant upward trend, while the abundance of Peak5 (dual galactose bi-antenna structure, NA2) decreased accordingly, showing a significant downward trend ( Figure 2d ). When ALT>40IU / L, a more complex pattern of changes was observed. The abundance of Peak1-Peak4 (Peak3 and Peak4 are isomers, monogalactose-α-1,6-core fucosylated biantennary structure, NG1A2F) showed a significant increase, a trend similar to the change in the abundance of some N-sugars when ALT is normal, but the range is wider. The abundance of Peak5 (dual galactose biantennary structure, NA2) and Peak8 (trigalactose triantennary structure, NA3) decreased significantly, showing a different trend from that when ALT is normal, especially the more complex multi-antennary sugar chain structure of Peak8 ( Figure 2e ).

[0072] In summary, there is a clear correlation between the N-glycan structure and its abundance in serum and the severity of liver tissue inflammation: when serum ALT ≤ 40 IU / L, the N-glycan markers are Peak 1, Peak 2 and Peak 5; when ALT > 40 IU / L, the N-glycan markers are Peak 1, Peak 2, Peak 3, Peak 4, Peak 5 and Peak 8.

[0073] Step S3: Based on conventional indicators and / or N-glycan markers, binary logistic regression analysis method is used to construct liver tissue inflammation grading assessment models at different ALT levels;

[0074] Conventional parameters included gender, age, TBIL, and ALT.

[0075] Continuous variables with normal distribution were expressed as mean ± standard deviation, and t-test or analysis of variance were used for inter-group comparison; continuous variables with non-normal distribution were expressed as median (interquartile range), and non-parametric tests were used for inter-group comparison. Pearson correlation analysis was used for correlation analysis of normally distributed data, and Spearman correlation analysis was used for correlation analysis of non-normally distributed data.

[0076] Methods: A total of 221 patients with chronic hepatitis B were stratified according to the grade of liver tissue inflammation. A serum N-glycan model was constructed using binary logistic regression (LR) analysis, and the area under the receiver operating characteristic curve (ROC) was used to determine the diagnostic efficacy. The sensitivity, specificity, positive predictive value, and negative predictive value of each diagnostic model were calculated at the optimal cutoff value. In all analyses, P < 0.05 was considered statistically significant. SPSS 26.0 (SPSS, Chicago, IL) was used for data analysis.

[0077] There is a clear correlation between the abundance of N-glycan structures in serum and the severity of liver tissue inflammation. At the same time, the abundance of these N-glycan structures is also significantly affected by the level of alanine aminotransferase (ALT). We still use serum ALT40IU / L (1 times the reference value) as the cutoff point to construct a diagnostic model for N-glycans when ALT is normal or elevated to distinguish different pathological stages of liver tissue inflammation. The N-glycan fingerprint is combined with the widely used liver function evaluation indicators (total bilirubin TBIL and alanine aminotransferase ALT) in clinical practice and the patient's gender and age information, and a stepwise logistic regression analysis method is used to construct a diagnostic model for liver tissue inflammation. This embodiment defines three different versions of logistic regression: (1) HIS-R A / B ( H eptic I nflammation S core- R outine) model only includes conventional indicators; (2) HIS-G A / B ( H eptic I nflammation S core- G lycomic) model focuses on the integration of N-glycan markers; (3) HIS-G A+ / B+ ( H epticI nflammation S core- G The lycomic Plus model combines N-glycan markers with conventional indicators.

[0078] (1) When ALT ≤ 40 IU / L, the liver tissue inflammation grading assessment models are:

[0079] HIS-R A =0.468×Gender+0.051×Age+0.001×TBIL+0.068×ALT;

[0080] HIS-G A =41.67-0.315×Peak1+0.226×Peak2-0.444×Peak5+0.690×Gender-0.010×Age

[0081] HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757

[0082] (2) When ALT>40IU / L, the liver tissue inflammation grading assessment models are:

[0083] HIS-R B =0.134+0.862×Gender-0.014×Age+0.017×TBIL+0.002×Age

[0084] HIS-G B =16.39-0.19×Peak1+0.96×Peak2-0.26×Peak3+0.53×Peak4-0.20×Peak5-0.30×Peak8+1.48×Gender-0.03×Age

[0085] HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284

[0086] Among them, the values ​​of Peak1, Peak2, Peak3, Peak4, Peak5, and Peak8 are obtained by detecting the patient's serum with NGFP technology to obtain the N-sugar chain map, extracting the peak height values ​​of the thirteen N-sugar chains, dividing the peak height value of each peak by the sum of the heights of all peaks, and calculating the relative content of each peak in %; Gender is the patient's gender, with a value of 0 for males and 1 for females; AGE is the patient's age, accurate to one decimal place, such as 60 years and 8 months = 60 years + (8 / 12) years = 60.7 years; TBIL is the patient's serum total bilirubin (TBIL) level, obtained by automatic biochemical analyzer, in μmol / L; ALT is the patient's serum alanine aminotransferase (ALT) level, obtained by automatic biochemical analyzer, in IU / L.

[0087] Step S4: Use the area under the receiver operating characteristic curve to evaluate the diagnostic efficacy of the above models respectively, and select the model with the best efficacy as the chronic hepatitis B liver tissue inflammation grading prediction model.

[0088] For HIS-R in step S3 A Model, HIS-R B Model, HIS-G A Model, HIS-G B Model, HIS-G A+ Model and HIS-G B+ The model effectiveness was evaluated for each model, and the results are shown in Table 3 and Figure 3 .

[0089] Table 3 Evaluation of diagnostic efficiency of liver tissue inflammation grading assessment model at different ALT levels

[0090]

[0091] Through Table 3 and Figure 3 It can be seen that:

[0092] (1)HIS-G A / B Superior to conventional indicator models, with good ability to discriminate liver tissue inflammation

[0093] When ALT≤40IU / L, NGFP model HIS-G A Compared with conventional indicators, it showed higher discrimination ability: in distinguishing significant inflammation of liver tissue (≥G2), NGFP model HIS-G A The area under the ROC curve reached 0.805, while HIS-R A The area under the ROC curve of the model was only 0.724, and the area under the ROC curve of the single indicator ALT was only 0.641.

[0094] When ALT>40IU / L, the advantage of NGFP model is still significant. B The area under the ROC curve for identifying significant liver inflammation (≥G2) was 0.810, which was higher than that of HIS-R B Model with single indicator ALT.

[0095] (2) Model HIS-G A+ / B+ Has the best diagnostic performance for liver tissue inflammation

[0096] When ALT ≤ 40 IU / L, HIS-G A+ The area under the ROC curve for identifying significant liver inflammation (≥G2) reached 0.904, which was significantly higher than that of HIS-R alone. A 、HIS-G A Model or single indicator ALT.

[0097] When ALT>40IU / L, HIS-G B+ The area under the ROC curve for identifying significant liver inflammation (≥G2) also reached 0.838, which was higher than that of HIS-R alone. B 、HIS-G B Model or single indicator ALT.

[0098] Therefore, we selected HIS-G, which has the best performance. A+ Model and HIS-G B+ The model is used as a prediction model for liver tissue inflammation grade in chronic hepatitis B:

[0099] When ALT≤40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0100] HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757

[0101] When ALT>40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0102] HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284.

[0103] Among them, HIS-G A+ The optimal cutoff value of the model was 0.093; HIS-G B+The optimal cutoff value of the model is 0.689. When the model calculation value exceeds the optimal cutoff value, it indicates that the patient's liver tissue has significant inflammation, and antiviral treatment should be considered clinically. If the model calculation value is less than the optimal cutoff value, it indicates that the patient's liver tissue has no or mild inflammation.

[0104] Example 2 Chronic Hepatitis B Liver Tissue Inflammation Grading Prediction System

[0105] This embodiment provides a chronic hepatitis B liver tissue inflammation grading prediction system, which includes: an input module, a grading prediction module and an output module.

[0106] The input module is used to obtain the data of N-glycogen markers, gender, age, TBIL and ALT of the current patient.

[0107] The N-sugar markers are Peak1, Peak2, Peak3, Peak4, Peak5 and Peak8.

[0108] The grading prediction module is used to use a pre-built chronic hepatitis B liver tissue inflammation grading prediction model to predict the grading of chronic hepatitis B liver tissue inflammation and obtain the grading prediction results.

[0109] When ALT≤40 IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0110] HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757

[0111] When ALT>40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is:

[0112] HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284

[0113] Among them, HIS-G A =41.67-0.315×Peak1+0.226×Peak2-0.444×Peak5+0.690×Gender-0.010×Age;

[0114] HIS-G B =16.39-0.19×Peak1+0.96×Peak2-0.26×Peak3+0.53×Peak4-0.20×Peak5-0.30×Peak8+1.48×Gender-0.03×Age;

[0115] The values ​​of Peak1, Peak2, Peak3, Peak4, Peak5, and Peak8 are obtained by detecting the patient's serum with NGFP technology to obtain the N-glycan map, extracting the peak height values ​​of the thirteen N-glycans, dividing the peak height value of each peak by the sum of the heights of all peaks, and calculating the relative content of each peak in %; Gender is the patient's gender, with 0 for males and 1 for females; AGE is the patient's age, accurate to one decimal place; TBIL is the patient's serum total bilirubin level, obtained by fully automatic biochemical analyzer, in μmol / L; ALT is the patient's serum alanine aminotransferase level, obtained by fully automatic biochemical analyzer, in IU / L.

[0116] The output module is used to output the hierarchical prediction results.

[0117] For example, the method for predicting the grade of chronic hepatitis B liver tissue inflammation using the chronic hepatitis B liver tissue inflammation grade prediction system is:

[0118] 1. The patient's serum is tested through N-sugar fingerprint (NGFP) and routine biochemical tests (including total bilirubin TBIL and alanine aminotransferase ALT) to obtain corresponding test values, and the above values ​​are input into the input module.

[0119] 2. The graded prediction module uses different model formulas to calculate the model calculation value according to the ALT level in the patient's serum. For example: If the patient's serum ALT level does not exceed 40IU / L, substitute each test value into HIS-G A+ Calculate and get HIS-G A+ The calculated value is 0.26. Because 0.26>0.093, it indicates that the patient's liver tissue is in a significant inflammatory state, and antiviral treatment should be considered clinically.

[0120] 3. The output module outputs the calculated value (0.26) and its corresponding inflammation grade prediction result (significant inflammation in liver tissue) and clinical measures (antiviral treatment needs to be considered).

[0121] It should be noted that the embodiments of the present invention have better practicability and do not impose any form of limitation on the present invention. Any technician familiar with the field may use the technical content disclosed above to change or modify it into an equivalent effective embodiment. However, any modification or equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A method for constructing a chronic hepatitis B liver tissue inflammation grading prediction model, characterized in that: include: Step S1: screening a number of patients who were diagnosed with chronic hepatitis B by liver tissue biopsy pathology and had a clear liver tissue inflammation grade according to the inclusion and exclusion criteria, collecting the clinical data of the patients, and collecting serum samples from the patients; Step S2: using N-glycan fingerprinting NGFP technology to perform serum protein N-glycan fingerprinting difference analysis on serum samples of patients with different alanine aminotransferase ALT levels, and obtaining N-glycan markers at different ALT levels; Step S3: Based on conventional biochemical indices and / or N-glycan markers, binary logistic regression analysis was used to construct liver tissue inflammation grading assessment models at different ALT levels; Step S4: Use the area under the receiver operating characteristic curve to evaluate the diagnostic efficacy of the above models respectively, and select the model with the best efficacy as the chronic hepatitis B liver tissue inflammation grading prediction model.

2. The method for constructing a chronic hepatitis B liver tissue inflammation grading prediction model according to claim 1, characterized in that: In step S2, the N-sugar markers when serum ALT ≤ 40 IU / L are Peak1, Peak2 and Peak5; and the N-sugar markers when ALT > 40 IU / L are Peak1, Peak2, Peak3, Peak4, Peak5 and Peak8.

3. The method for constructing a chronic hepatitis B liver tissue inflammation grading prediction model according to claim 2, characterized in that: In step S3, conventional indicators include gender, age, total bilirubin and ALT.

4. The method for constructing a chronic hepatitis B liver tissue inflammation grading prediction model according to claim 3, characterized in that: In step S3, (1) When ALT ≤ 40 IU / L, the liver tissue inflammation grading assessment models are: Conventional indicator model HIS-R A =0.468×Gender+0.051×Age+0.001×TBIL+0.068×ALT; NGFP model HIS-G A =41.67-0.315×Peak1+0.226×Peak2-0.444×Peak5+0.690×Gender-0.010×Age NGFP model enhanced version HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757 (2) When ALT>40IU / L, the liver tissue inflammation grading assessment models are: Conventional indicator model HIS-R B =0.134+0.862×Gender-0.014×Age+0.017×TBIL+0.002×Age NGFP model HIS-G B =16.39-0.19×Peak1+0.96×Peak2-0.26×Peak3+0.53×Peak4-0.20×Peak5-0.30×Peak8+1.48×Gender-0.03×Age NGFP model enhanced version HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284 Among them, the values ​​of Peak1, Peak2, Peak3, Peak4, Peak5, and Peak8 are obtained by detecting the patient's serum with NGFP technology to obtain the N-glycan map, extracting the peak height values ​​of the thirteen N-glycans, dividing the peak height value of each peak by the sum of the heights of all peaks, and calculating the relative content of each peak in %; Gender is the patient's gender, with a value of 0 for males and 1 for females; AGE is the patient's age, with one decimal place after the decimal point; TBIL is the patient's serum total bilirubin level, obtained by automatic biochemical analyzer, in μmol / L; ALT is the patient's serum alanine aminotransferase level, obtained by automatic biochemical analyzer, in IU / L.

5. The method for constructing a chronic hepatitis B liver tissue inflammation grading prediction model according to claim 4, characterized in that: In step S4, When ALT≤40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is: HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757 When ALT>40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is: HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284。 6. A chronic hepatitis B liver tissue inflammation grading prediction system, characterized in that: include: An input module, the input module is used to obtain data of N-glycogen markers, gender, age, TBIL and ALT of the current patient; A grading prediction module, which is used to use a pre-built chronic hepatitis B liver tissue inflammation grading prediction model to perform chronic hepatitis B liver tissue inflammation grading prediction and obtain a grading prediction result; An output module, wherein the output module is used to output the hierarchical prediction result.

7. The chronic hepatitis B liver tissue inflammation grading prediction system according to claim 6, characterized in that: The method for constructing the chronic hepatitis B liver tissue inflammation grading prediction model comprises the following steps: Step S1: screening a number of patients who were diagnosed with chronic hepatitis B by liver tissue biopsy pathology and had a clear liver tissue inflammation grade according to the inclusion and exclusion criteria, collecting the clinical data of the patients, and collecting serum samples from the patients; Step S2: using N-glycan fingerprint technology to perform serum protein N-glycan fingerprint difference analysis on serum samples of patients with different ALT levels to obtain N-glycan markers at different ALT levels; Step S3: Based on conventional biochemical indices and / or N-glycan markers, binary logistic regression analysis was used to construct liver tissue inflammation grading assessment models at different ALT levels; Step S4: Use the area under the receiver operating characteristic curve to evaluate the diagnostic efficacy of the above models respectively, and select the model with the best efficacy as the chronic hepatitis B liver tissue inflammation grading prediction model.

8. The chronic hepatitis B liver tissue inflammation grading prediction system according to claim 6 or 7, characterized in that: The N-sugar markers are Peak1, Peak2, Peak3, Peak4, Peak5 and Peak8; When ALT≤40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is: HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757 When ALT>40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is: HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284 Among them, HIS-G A =43.67-0.315×Peak1+0.226×Peak2-0.444×Peak5+0.690×Gender-0.010×Age; HIS-G B =16.39-0.19×Peak1+0.96×Peak2-0.26×Peak3+0.53×Peak4-0.20×Peak5-0.30×Peak8+1.48×Gender-0.03×Age; the values ​​of Peak1, Peak2, Peak3, Peak4, Peak5 and Peak8 are obtained by detecting the patient's serum with NGFP technology to obtain the N-glycan map, extracting the peak height values ​​of the thirteen N-glycans, dividing the peak height value of each peak by the sum of the heights of all peaks, and calculating the relative content of each peak in %; Gender is the patient's gender, with 0 for males and 1 for females; AGE is the patient's age, accurate to one decimal place; TBIL is the patient's serum total bilirubin level, obtained by automatic biochemical analyzer, in μmol / L; ALT is the patient's serum alanine aminotransferase level, obtained by automatic biochemical analyzer, in IU / L.

9. Use of a combined marker in assessing the grading of chronic hepatitis B hepatitis, characterized in that: The combined marker consists of N-sugar markers, gender, age, TBIL and ALT, and the combined marker is used to evaluate the inflammation grade of liver tissue in chronic hepatitis B by constructing a model.

10. Use of the combined marker according to claim 9 in evaluating the grade of chronic hepatitis B hepatitis, characterized in that: When ALT≤40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is: HIS-G A+ =HIS-G A +0.032×TBIL-0.099×ALT-15.757 When ALT>40IU / L, the chronic hepatitis B liver tissue inflammation grading prediction model is: HIS-G B+ =HIS-G B +0.005×TBIL+0.006×ALT-4.284 Among them, HIS-G A =41.67-0.315×Peak1+0.226×Peak2-0.444×Peak5+0.690×Gender-0.010×Age; HIS-G B =16.39-0.19×Peak1+0.96×Peak2-0.26×Peak3+0.53×Peak4-0.20×Peak5-0.30×Peak8+1.48×Gender-0.03×Age; The values ​​of Peak1, Peak2, Peak3, Peak4, Peak5, and Peak8 are obtained by detecting the patient's serum with NGFP technology to obtain the N-glycan map, extracting the peak height values ​​of the thirteen N-glycans, dividing the peak height value of each peak by the sum of the heights of all peaks, and calculating the relative content of each peak in %; Gender is the patient's gender, with 0 for males and 1 for females; AGE is the patient's age, accurate to one decimal place; TBIL is the patient's serum total bilirubin level, obtained by fully automatic biochemical analyzer, in μmol / L; ALT is the patient's serum alanine aminotransferase level, obtained by fully automatic biochemical analyzer, in IU / L.

Images